216 files changed, 7431 insertions, 2744 deletions

diff --git a/src/audio_core/common.h b/src/audio_core/common.h
index 1ab537588..e6b95769f 100644
--- a/src/audio_core/common.h
+++ b/src/audio_core/common.h
@@ -15,7 +15,9 @@ constexpr ResultCode ERR_INVALID_PARAMETERS{ErrorModule::Audio, 41};
 constexpr ResultCode ERR_SPLITTER_SORT_FAILED{ErrorModule::Audio, 43};
 } // namespace Audren
 
-constexpr u32_le CURRENT_PROCESS_REVISION = Common::MakeMagic('R', 'E', 'V', '9');
+constexpr u8 BASE_REVISION = '0';
+constexpr u32_le CURRENT_PROCESS_REVISION =
+    Common::MakeMagic('R', 'E', 'V', static_cast<u8>(BASE_REVISION + 0xA));
 constexpr std::size_t MAX_MIX_BUFFERS = 24;
 constexpr std::size_t MAX_BIQUAD_FILTERS = 2;
 constexpr std::size_t MAX_CHANNEL_COUNT = 6;
diff --git a/src/common/bit_util.h b/src/common/bit_util.h
index f50d3308a..f37538e06 100644
--- a/src/common/bit_util.h
+++ b/src/common/bit_util.h
@@ -57,4 +57,11 @@ requires std::is_integral_v<T>
     return static_cast<T>(1ULL << ((8U * sizeof(T)) - std::countl_zero(value - 1U)));
 }
 
+template <size_t bit_index, typename T>
+requires std::is_integral_v<T>
+[[nodiscard]] constexpr bool Bit(const T value) {
+    static_assert(bit_index < BitSize<T>(), "bit_index must be smaller than size of T");
+    return ((value >> bit_index) & T(1)) == T(1);
+}
+
 } // namespace Common
diff --git a/src/common/common_types.h b/src/common/common_types.h
index 4cec89fbd..99bffc460 100644
--- a/src/common/common_types.h
+++ b/src/common/common_types.h
@@ -46,13 +46,3 @@ using GPUVAddr = u64; ///< Represents a pointer in the GPU virtual address space
 
 using u128 = std::array<std::uint64_t, 2>;
 static_assert(sizeof(u128) == 16, "u128 must be 128 bits wide");
-
-// An inheritable class to disallow the copy constructor and operator= functions
-class NonCopyable {
-protected:
-    constexpr NonCopyable() = default;
-    ~NonCopyable() = default;
-
-    NonCopyable(const NonCopyable&) = delete;
-    NonCopyable& operator=(const NonCopyable&) = delete;
-};
diff --git a/src/common/fs/fs_util.cpp b/src/common/fs/fs_util.cpp
index 9f8671982..0068112e6 100644
--- a/src/common/fs/fs_util.cpp
+++ b/src/common/fs/fs_util.cpp
@@ -16,6 +16,10 @@ std::u8string BufferToU8String(std::span<const u8> buffer) {
     return std::u8string{buffer.begin(), std::ranges::find(buffer, u8{0})};
 }
 
+std::u8string_view BufferToU8StringView(std::span<const u8> buffer) {
+    return std::u8string_view{reinterpret_cast<const char8_t*>(buffer.data())};
+}
+
 std::string ToUTF8String(std::u8string_view u8_string) {
     return std::string{u8_string.begin(), u8_string.end()};
 }
@@ -24,6 +28,10 @@ std::string BufferToUTF8String(std::span<const u8> buffer) {
     return std::string{buffer.begin(), std::ranges::find(buffer, u8{0})};
 }
 
+std::string_view BufferToUTF8StringView(std::span<const u8> buffer) {
+    return std::string_view{reinterpret_cast<const char*>(buffer.data())};
+}
+
 std::string PathToUTF8String(const std::filesystem::path& path) {
     return ToUTF8String(path.u8string());
 }
diff --git a/src/common/fs/fs_util.h b/src/common/fs/fs_util.h
index 1ec82eb35..1620d38c9 100644
--- a/src/common/fs/fs_util.h
+++ b/src/common/fs/fs_util.h
@@ -38,6 +38,15 @@ concept IsChar = std::same_as<T, char>;
 [[nodiscard]] std::u8string BufferToU8String(std::span<const u8> buffer);
 
 /**
+ * Same as BufferToU8String, but returns a string view of the buffer.
+ *
+ * @param buffer Buffer of bytes
+ *
+ * @returns UTF-8 encoded std::u8string_view.
+ */
+[[nodiscard]] std::u8string_view BufferToU8StringView(std::span<const u8> buffer);
+
+/**
  * Converts a std::u8string or std::u8string_view to a UTF-8 encoded std::string.
  *
  * @param u8_string UTF-8 encoded u8string
@@ -58,6 +67,15 @@ concept IsChar = std::same_as<T, char>;
 [[nodiscard]] std::string BufferToUTF8String(std::span<const u8> buffer);
 
 /**
+ * Same as BufferToUTF8String, but returns a string view of the buffer.
+ *
+ * @param buffer Buffer of bytes
+ *
+ * @returns UTF-8 encoded std::string_view.
+ */
+[[nodiscard]] std::string_view BufferToUTF8StringView(std::span<const u8> buffer);
+
+/**
  * Converts a filesystem path to a UTF-8 encoded std::string.
  *
  * @param path Filesystem path
diff --git a/src/common/host_memory.cpp b/src/common/host_memory.cpp
index 28949fe5e..c465cfc14 100644
--- a/src/common/host_memory.cpp
+++ b/src/common/host_memory.cpp
@@ -327,8 +327,8 @@ private:
     bool IsNiechePlaceholder(size_t virtual_offset, size_t length) const {
         const auto it = placeholders.upper_bound({virtual_offset, virtual_offset + length});
         if (it != placeholders.end() && it->lower() == virtual_offset + length) {
-            const bool is_root = it == placeholders.begin() && virtual_offset == 0;
-            return is_root || std::prev(it)->upper() == virtual_offset;
+            return it == placeholders.begin() ? virtual_offset == 0
+                                              : std::prev(it)->upper() == virtual_offset;
         }
         return false;
     }
diff --git a/src/common/intrusive_red_black_tree.h b/src/common/intrusive_red_black_tree.h
index 3173cc449..b296b639e 100644
--- a/src/common/intrusive_red_black_tree.h
+++ b/src/common/intrusive_red_black_tree.h
@@ -4,6 +4,8 @@
 
 #pragma once
 
+#include "common/alignment.h"
+#include "common/common_funcs.h"
 #include "common/parent_of_member.h"
 #include "common/tree.h"
 
@@ -15,32 +17,33 @@ class IntrusiveRedBlackTreeImpl;
 
 }
 
+#pragma pack(push, 4)
 struct IntrusiveRedBlackTreeNode {
+    YUZU_NON_COPYABLE(IntrusiveRedBlackTreeNode);
+
 public:
-    using EntryType = RBEntry<IntrusiveRedBlackTreeNode>;
+    using RBEntry = freebsd::RBEntry<IntrusiveRedBlackTreeNode>;
 
-    constexpr IntrusiveRedBlackTreeNode() = default;
+private:
+    RBEntry m_entry;
 
-    void SetEntry(const EntryType& new_entry) {
-        entry = new_entry;
-    }
+public:
+    explicit IntrusiveRedBlackTreeNode() = default;
 
-    [[nodiscard]] EntryType& GetEntry() {
-        return entry;
+    [[nodiscard]] constexpr RBEntry& GetRBEntry() {
+        return m_entry;
     }
-
-    [[nodiscard]] const EntryType& GetEntry() const {
-        return entry;
+    [[nodiscard]] constexpr const RBEntry& GetRBEntry() const {
+        return m_entry;
     }
 
-private:
-    EntryType entry{};
-
-    friend class impl::IntrusiveRedBlackTreeImpl;
-
-    template <class, class, class>
-    friend class IntrusiveRedBlackTree;
+    constexpr void SetRBEntry(const RBEntry& entry) {
+        m_entry = entry;
+    }
 };
+static_assert(sizeof(IntrusiveRedBlackTreeNode) ==
+              3 * sizeof(void*) + std::max<size_t>(sizeof(freebsd::RBColor), 4));
+#pragma pack(pop)
 
 template <class T, class Traits, class Comparator>
 class IntrusiveRedBlackTree;
@@ -48,12 +51,17 @@ class IntrusiveRedBlackTree;
 namespace impl {
 
 class IntrusiveRedBlackTreeImpl {
+    YUZU_NON_COPYABLE(IntrusiveRedBlackTreeImpl);
+
 private:
     template <class, class, class>
     friend class ::Common::IntrusiveRedBlackTree;
 
-    using RootType = RBHead<IntrusiveRedBlackTreeNode>;
-    RootType root;
+private:
+    using RootType = freebsd::RBHead<IntrusiveRedBlackTreeNode>;
+
+private:
+    RootType m_root;
 
 public:
     template <bool Const>
@@ -81,149 +89,150 @@ public:
                                              IntrusiveRedBlackTreeImpl::reference>;
 
     private:
-        pointer node;
+        pointer m_node;
 
     public:
-        explicit Iterator(pointer n) : node(n) {}
+        constexpr explicit Iterator(pointer n) : m_node(n) {}
 
-        bool operator==(const Iterator& rhs) const {
-            return this->node == rhs.node;
+        constexpr bool operator==(const Iterator& rhs) const {
+            return m_node == rhs.m_node;
         }
 
-        bool operator!=(const Iterator& rhs) const {
+        constexpr bool operator!=(const Iterator& rhs) const {
             return !(*this == rhs);
         }
 
-        pointer operator->() const {
-            return this->node;
+        constexpr pointer operator->() const {
+            return m_node;
         }
 
-        reference operator*() const {
-            return *this->node;
+        constexpr reference operator*() const {
+            return *m_node;
         }
 
-        Iterator& operator++() {
-            this->node = GetNext(this->node);
+        constexpr Iterator& operator++() {
+            m_node = GetNext(m_node);
             return *this;
         }
 
-        Iterator& operator--() {
-            this->node = GetPrev(this->node);
+        constexpr Iterator& operator--() {
+            m_node = GetPrev(m_node);
             return *this;
         }
 
-        Iterator operator++(int) {
+        constexpr Iterator operator++(int) {
             const Iterator it{*this};
             ++(*this);
             return it;
         }
 
-        Iterator operator--(int) {
+        constexpr Iterator operator--(int) {
             const Iterator it{*this};
             --(*this);
             return it;
         }
 
-        operator Iterator<true>() const {
-            return Iterator<true>(this->node);
+        constexpr operator Iterator<true>() const {
+            return Iterator<true>(m_node);
         }
     };
 
 private:
-    // Define accessors using RB_* functions.
-    bool EmptyImpl() const {
-        return root.IsEmpty();
+    constexpr bool EmptyImpl() const {
+        return m_root.IsEmpty();
     }
 
-    IntrusiveRedBlackTreeNode* GetMinImpl() const {
-        return RB_MIN(const_cast<RootType*>(&root));
+    constexpr IntrusiveRedBlackTreeNode* GetMinImpl() const {
+        return freebsd::RB_MIN(const_cast<RootType&>(m_root));
     }
 
-    IntrusiveRedBlackTreeNode* GetMaxImpl() const {
-        return RB_MAX(const_cast<RootType*>(&root));
+    constexpr IntrusiveRedBlackTreeNode* GetMaxImpl() const {
+        return freebsd::RB_MAX(const_cast<RootType&>(m_root));
     }
 
-    IntrusiveRedBlackTreeNode* RemoveImpl(IntrusiveRedBlackTreeNode* node) {
-        return RB_REMOVE(&root, node);
+    constexpr IntrusiveRedBlackTreeNode* RemoveImpl(IntrusiveRedBlackTreeNode* node) {
+        return freebsd::RB_REMOVE(m_root, node);
     }
 
 public:
-    static IntrusiveRedBlackTreeNode* GetNext(IntrusiveRedBlackTreeNode* node) {
-        return RB_NEXT(node);
+    static constexpr IntrusiveRedBlackTreeNode* GetNext(IntrusiveRedBlackTreeNode* node) {
+        return freebsd::RB_NEXT(node);
     }
 
-    static IntrusiveRedBlackTreeNode* GetPrev(IntrusiveRedBlackTreeNode* node) {
-        return RB_PREV(node);
+    static constexpr IntrusiveRedBlackTreeNode* GetPrev(IntrusiveRedBlackTreeNode* node) {
+        return freebsd::RB_PREV(node);
     }
 
-    static const IntrusiveRedBlackTreeNode* GetNext(const IntrusiveRedBlackTreeNode* node) {
+    static constexpr IntrusiveRedBlackTreeNode const* GetNext(
+        IntrusiveRedBlackTreeNode const* node) {
         return static_cast<const IntrusiveRedBlackTreeNode*>(
             GetNext(const_cast<IntrusiveRedBlackTreeNode*>(node)));
     }
 
-    static const IntrusiveRedBlackTreeNode* GetPrev(const IntrusiveRedBlackTreeNode* node) {
+    static constexpr IntrusiveRedBlackTreeNode const* GetPrev(
+        IntrusiveRedBlackTreeNode const* node) {
         return static_cast<const IntrusiveRedBlackTreeNode*>(
             GetPrev(const_cast<IntrusiveRedBlackTreeNode*>(node)));
     }
 
 public:
-    constexpr IntrusiveRedBlackTreeImpl() {}
+    constexpr IntrusiveRedBlackTreeImpl() = default;
 
     // Iterator accessors.
-    iterator begin() {
+    constexpr iterator begin() {
         return iterator(this->GetMinImpl());
     }
 
-    const_iterator begin() const {
+    constexpr const_iterator begin() const {
         return const_iterator(this->GetMinImpl());
     }
 
-    iterator end() {
+    constexpr iterator end() {
         return iterator(static_cast<IntrusiveRedBlackTreeNode*>(nullptr));
     }
 
-    const_iterator end() const {
+    constexpr const_iterator end() const {
         return const_iterator(static_cast<const IntrusiveRedBlackTreeNode*>(nullptr));
     }
 
-    const_iterator cbegin() const {
+    constexpr const_iterator cbegin() const {
         return this->begin();
     }
 
-    const_iterator cend() const {
+    constexpr const_iterator cend() const {
         return this->end();
     }
 
-    iterator iterator_to(reference ref) {
-        return iterator(&ref);
+    constexpr iterator iterator_to(reference ref) {
+        return iterator(std::addressof(ref));
     }
 
-    const_iterator iterator_to(const_reference ref) const {
-        return const_iterator(&ref);
+    constexpr const_iterator iterator_to(const_reference ref) const {
+        return const_iterator(std::addressof(ref));
     }
 
     // Content management.
-    bool empty() const {
+    constexpr bool empty() const {
         return this->EmptyImpl();
     }
 
-    reference back() {
+    constexpr reference back() {
         return *this->GetMaxImpl();
     }
 
-    const_reference back() const {
+    constexpr const_reference back() const {
         return *this->GetMaxImpl();
     }
 
-    reference front() {
+    constexpr reference front() {
         return *this->GetMinImpl();
     }
 
-    const_reference front() const {
+    constexpr const_reference front() const {
         return *this->GetMinImpl();
     }
 
-    iterator erase(iterator it) {
+    constexpr iterator erase(iterator it) {
         auto cur = std::addressof(*it);
         auto next = GetNext(cur);
         this->RemoveImpl(cur);
@@ -234,16 +243,16 @@ public:
 } // namespace impl
 
 template <typename T>
-concept HasLightCompareType = requires {
-    { std::is_same<typename T::LightCompareType, void>::value } -> std::convertible_to<bool>;
+concept HasRedBlackKeyType = requires {
+    { std::is_same<typename T::RedBlackKeyType, void>::value } -> std::convertible_to<bool>;
 };
 
 namespace impl {
 
     template <typename T, typename Default>
-    consteval auto* GetLightCompareType() {
-        if constexpr (HasLightCompareType<T>) {
-            return static_cast<typename T::LightCompareType*>(nullptr);
+    consteval auto* GetRedBlackKeyType() {
+        if constexpr (HasRedBlackKeyType<T>) {
+            return static_cast<typename T::RedBlackKeyType*>(nullptr);
         } else {
             return static_cast<Default*>(nullptr);
         }
@@ -252,16 +261,17 @@ namespace impl {
 } // namespace impl
 
 template <typename T, typename Default>
-using LightCompareType = std::remove_pointer_t<decltype(impl::GetLightCompareType<T, Default>())>;
+using RedBlackKeyType = std::remove_pointer_t<decltype(impl::GetRedBlackKeyType<T, Default>())>;
 
 template <class T, class Traits, class Comparator>
 class IntrusiveRedBlackTree {
+    YUZU_NON_COPYABLE(IntrusiveRedBlackTree);
 
 public:
     using ImplType = impl::IntrusiveRedBlackTreeImpl;
 
 private:
-    ImplType impl{};
+    ImplType m_impl;
 
 public:
     template <bool Const>
@@ -277,9 +287,9 @@ public:
     using iterator = Iterator<false>;
     using const_iterator = Iterator<true>;
 
-    using light_value_type = LightCompareType<Comparator, value_type>;
-    using const_light_pointer = const light_value_type*;
-    using const_light_reference = const light_value_type&;
+    using key_type = RedBlackKeyType<Comparator, value_type>;
+    using const_key_pointer = const key_type*;
+    using const_key_reference = const key_type&;
 
     template <bool Const>
     class Iterator {
@@ -298,183 +308,201 @@ public:
                                              IntrusiveRedBlackTree::reference>;
 
     private:
-        ImplIterator iterator;
+        ImplIterator m_impl;
 
     private:
-        explicit Iterator(ImplIterator it) : iterator(it) {}
+        constexpr explicit Iterator(ImplIterator it) : m_impl(it) {}
 
-        explicit Iterator(typename std::conditional<Const, ImplType::const_iterator,
-                                                    ImplType::iterator>::type::pointer ptr)
-            : iterator(ptr) {}
+        constexpr explicit Iterator(typename ImplIterator::pointer p) : m_impl(p) {}
 
-        ImplIterator GetImplIterator() const {
-            return this->iterator;
+        constexpr ImplIterator GetImplIterator() const {
+            return m_impl;
         }
 
     public:
-        bool operator==(const Iterator& rhs) const {
-            return this->iterator == rhs.iterator;
+        constexpr bool operator==(const Iterator& rhs) const {
+            return m_impl == rhs.m_impl;
         }
 
-        bool operator!=(const Iterator& rhs) const {
+        constexpr bool operator!=(const Iterator& rhs) const {
             return !(*this == rhs);
         }
 
-        pointer operator->() const {
-            return Traits::GetParent(std::addressof(*this->iterator));
+        constexpr pointer operator->() const {
+            return Traits::GetParent(std::addressof(*m_impl));
         }
 
-        reference operator*() const {
-            return *Traits::GetParent(std::addressof(*this->iterator));
+        constexpr reference operator*() const {
+            return *Traits::GetParent(std::addressof(*m_impl));
         }
 
-        Iterator& operator++() {
-            ++this->iterator;
+        constexpr Iterator& operator++() {
+            ++m_impl;
             return *this;
         }
 
-        Iterator& operator--() {
-            --this->iterator;
+        constexpr Iterator& operator--() {
+            --m_impl;
             return *this;
         }
 
-        Iterator operator++(int) {
+        constexpr Iterator operator++(int) {
             const Iterator it{*this};
-            ++this->iterator;
+            ++m_impl;
             return it;
         }
 
-        Iterator operator--(int) {
+        constexpr Iterator operator--(int) {
             const Iterator it{*this};
-            --this->iterator;
+            --m_impl;
             return it;
         }
 
-        operator Iterator<true>() const {
-            return Iterator<true>(this->iterator);
+        constexpr operator Iterator<true>() const {
+            return Iterator<true>(m_impl);
         }
     };
 
 private:
-    static int CompareImpl(const IntrusiveRedBlackTreeNode* lhs,
-                           const IntrusiveRedBlackTreeNode* rhs) {
+    static constexpr int CompareImpl(const IntrusiveRedBlackTreeNode* lhs,
+                                     const IntrusiveRedBlackTreeNode* rhs) {
         return Comparator::Compare(*Traits::GetParent(lhs), *Traits::GetParent(rhs));
     }
 
-    static int LightCompareImpl(const void* elm, const IntrusiveRedBlackTreeNode* rhs) {
-        return Comparator::Compare(*static_cast<const_light_pointer>(elm), *Traits::GetParent(rhs));
+    static constexpr int CompareKeyImpl(const_key_reference key,
+                                        const IntrusiveRedBlackTreeNode* rhs) {
+        return Comparator::Compare(key, *Traits::GetParent(rhs));
     }
 
     // Define accessors using RB_* functions.
-    IntrusiveRedBlackTreeNode* InsertImpl(IntrusiveRedBlackTreeNode* node) {
-        return RB_INSERT(&impl.root, node, CompareImpl);
+    constexpr IntrusiveRedBlackTreeNode* InsertImpl(IntrusiveRedBlackTreeNode* node) {
+        return freebsd::RB_INSERT(m_impl.m_root, node, CompareImpl);
     }
 
-    IntrusiveRedBlackTreeNode* FindImpl(const IntrusiveRedBlackTreeNode* node) const {
-        return RB_FIND(const_cast<ImplType::RootType*>(&impl.root),
-                       const_cast<IntrusiveRedBlackTreeNode*>(node), CompareImpl);
+    constexpr IntrusiveRedBlackTreeNode* FindImpl(IntrusiveRedBlackTreeNode const* node) const {
+        return freebsd::RB_FIND(const_cast<ImplType::RootType&>(m_impl.m_root),
+                                const_cast<IntrusiveRedBlackTreeNode*>(node), CompareImpl);
     }
 
-    IntrusiveRedBlackTreeNode* NFindImpl(const IntrusiveRedBlackTreeNode* node) const {
-        return RB_NFIND(const_cast<ImplType::RootType*>(&impl.root),
-                        const_cast<IntrusiveRedBlackTreeNode*>(node), CompareImpl);
+    constexpr IntrusiveRedBlackTreeNode* NFindImpl(IntrusiveRedBlackTreeNode const* node) const {
+        return freebsd::RB_NFIND(const_cast<ImplType::RootType&>(m_impl.m_root),
+                                 const_cast<IntrusiveRedBlackTreeNode*>(node), CompareImpl);
     }
 
-    IntrusiveRedBlackTreeNode* FindLightImpl(const_light_pointer lelm) const {
-        return RB_FIND_LIGHT(const_cast<ImplType::RootType*>(&impl.root),
-                             static_cast<const void*>(lelm), LightCompareImpl);
+    constexpr IntrusiveRedBlackTreeNode* FindKeyImpl(const_key_reference key) const {
+        return freebsd::RB_FIND_KEY(const_cast<ImplType::RootType&>(m_impl.m_root), key,
+                                    CompareKeyImpl);
     }
 
-    IntrusiveRedBlackTreeNode* NFindLightImpl(const_light_pointer lelm) const {
-        return RB_NFIND_LIGHT(const_cast<ImplType::RootType*>(&impl.root),
-                              static_cast<const void*>(lelm), LightCompareImpl);
+    constexpr IntrusiveRedBlackTreeNode* NFindKeyImpl(const_key_reference key) const {
+        return freebsd::RB_NFIND_KEY(const_cast<ImplType::RootType&>(m_impl.m_root), key,
+                                     CompareKeyImpl);
+    }
+
+    constexpr IntrusiveRedBlackTreeNode* FindExistingImpl(
+        IntrusiveRedBlackTreeNode const* node) const {
+        return freebsd::RB_FIND_EXISTING(const_cast<ImplType::RootType&>(m_impl.m_root),
+                                         const_cast<IntrusiveRedBlackTreeNode*>(node), CompareImpl);
+    }
+
+    constexpr IntrusiveRedBlackTreeNode* FindExistingKeyImpl(const_key_reference key) const {
+        return freebsd::RB_FIND_EXISTING_KEY(const_cast<ImplType::RootType&>(m_impl.m_root), key,
+                                             CompareKeyImpl);
     }
 
 public:
     constexpr IntrusiveRedBlackTree() = default;
 
     // Iterator accessors.
-    iterator begin() {
-        return iterator(this->impl.begin());
+    constexpr iterator begin() {
+        return iterator(m_impl.begin());
     }
 
-    const_iterator begin() const {
-        return const_iterator(this->impl.begin());
+    constexpr const_iterator begin() const {
+        return const_iterator(m_impl.begin());
     }
 
-    iterator end() {
-        return iterator(this->impl.end());
+    constexpr iterator end() {
+        return iterator(m_impl.end());
     }
 
-    const_iterator end() const {
-        return const_iterator(this->impl.end());
+    constexpr const_iterator end() const {
+        return const_iterator(m_impl.end());
     }
 
-    const_iterator cbegin() const {
+    constexpr const_iterator cbegin() const {
         return this->begin();
     }
 
-    const_iterator cend() const {
+    constexpr const_iterator cend() const {
         return this->end();
     }
 
-    iterator iterator_to(reference ref) {
-        return iterator(this->impl.iterator_to(*Traits::GetNode(std::addressof(ref))));
+    constexpr iterator iterator_to(reference ref) {
+        return iterator(m_impl.iterator_to(*Traits::GetNode(std::addressof(ref))));
     }
 
-    const_iterator iterator_to(const_reference ref) const {
-        return const_iterator(this->impl.iterator_to(*Traits::GetNode(std::addressof(ref))));
+    constexpr const_iterator iterator_to(const_reference ref) const {
+        return const_iterator(m_impl.iterator_to(*Traits::GetNode(std::addressof(ref))));
     }
 
     // Content management.
-    bool empty() const {
-        return this->impl.empty();
+    constexpr bool empty() const {
+        return m_impl.empty();
     }
 
-    reference back() {
-        return *Traits::GetParent(std::addressof(this->impl.back()));
+    constexpr reference back() {
+        return *Traits::GetParent(std::addressof(m_impl.back()));
     }
 
-    const_reference back() const {
-        return *Traits::GetParent(std::addressof(this->impl.back()));
+    constexpr const_reference back() const {
+        return *Traits::GetParent(std::addressof(m_impl.back()));
     }
 
-    reference front() {
-        return *Traits::GetParent(std::addressof(this->impl.front()));
+    constexpr reference front() {
+        return *Traits::GetParent(std::addressof(m_impl.front()));
     }
 
-    const_reference front() const {
-        return *Traits::GetParent(std::addressof(this->impl.front()));
+    constexpr const_reference front() const {
+        return *Traits::GetParent(std::addressof(m_impl.front()));
     }
 
-    iterator erase(iterator it) {
-        return iterator(this->impl.erase(it.GetImplIterator()));
+    constexpr iterator erase(iterator it) {
+        return iterator(m_impl.erase(it.GetImplIterator()));
     }
 
-    iterator insert(reference ref) {
+    constexpr iterator insert(reference ref) {
         ImplType::pointer node = Traits::GetNode(std::addressof(ref));
         this->InsertImpl(node);
         return iterator(node);
     }
 
-    iterator find(const_reference ref) const {
+    constexpr iterator find(const_reference ref) const {
         return iterator(this->FindImpl(Traits::GetNode(std::addressof(ref))));
     }
 
-    iterator nfind(const_reference ref) const {
+    constexpr iterator nfind(const_reference ref) const {
         return iterator(this->NFindImpl(Traits::GetNode(std::addressof(ref))));
     }
 
-    iterator find_light(const_light_reference ref) const {
-        return iterator(this->FindLightImpl(std::addressof(ref)));
+    constexpr iterator find_key(const_key_reference ref) const {
+        return iterator(this->FindKeyImpl(ref));
+    }
+
+    constexpr iterator nfind_key(const_key_reference ref) const {
+        return iterator(this->NFindKeyImpl(ref));
+    }
+
+    constexpr iterator find_existing(const_reference ref) const {
+        return iterator(this->FindExistingImpl(Traits::GetNode(std::addressof(ref))));
     }
 
-    iterator nfind_light(const_light_reference ref) const {
-        return iterator(this->NFindLightImpl(std::addressof(ref)));
+    constexpr iterator find_existing_key(const_key_reference ref) const {
+        return iterator(this->FindExistingKeyImpl(ref));
     }
 };
 
-template <auto T, class Derived = impl::GetParentType<T>>
+template <auto T, class Derived = Common::impl::GetParentType<T>>
 class IntrusiveRedBlackTreeMemberTraits;
 
 template <class Parent, IntrusiveRedBlackTreeNode Parent::*Member, class Derived>
@@ -498,19 +526,16 @@ private:
         return std::addressof(parent->*Member);
     }
 
-    static constexpr Derived* GetParent(IntrusiveRedBlackTreeNode* node) {
-        return GetParentPointer<Member, Derived>(node);
+    static Derived* GetParent(IntrusiveRedBlackTreeNode* node) {
+        return Common::GetParentPointer<Member, Derived>(node);
     }
 
-    static constexpr Derived const* GetParent(const IntrusiveRedBlackTreeNode* node) {
-        return GetParentPointer<Member, Derived>(node);
+    static Derived const* GetParent(IntrusiveRedBlackTreeNode const* node) {
+        return Common::GetParentPointer<Member, Derived>(node);
     }
-
-private:
-    static constexpr TypedStorage<Derived> DerivedStorage = {};
 };
 
-template <auto T, class Derived = impl::GetParentType<T>>
+template <auto T, class Derived = Common::impl::GetParentType<T>>
 class IntrusiveRedBlackTreeMemberTraitsDeferredAssert;
 
 template <class Parent, IntrusiveRedBlackTreeNode Parent::*Member, class Derived>
@@ -521,11 +546,6 @@ public:
         IntrusiveRedBlackTree<Derived, IntrusiveRedBlackTreeMemberTraitsDeferredAssert, Comparator>;
     using TreeTypeImpl = impl::IntrusiveRedBlackTreeImpl;
 
-    static constexpr bool IsValid() {
-        TypedStorage<Derived> DerivedStorage = {};
-        return GetParent(GetNode(GetPointer(DerivedStorage))) == GetPointer(DerivedStorage);
-    }
-
 private:
     template <class, class, class>
     friend class IntrusiveRedBlackTree;
@@ -540,30 +560,36 @@ private:
         return std::addressof(parent->*Member);
     }
 
-    static constexpr Derived* GetParent(IntrusiveRedBlackTreeNode* node) {
-        return GetParentPointer<Member, Derived>(node);
+    static Derived* GetParent(IntrusiveRedBlackTreeNode* node) {
+        return Common::GetParentPointer<Member, Derived>(node);
     }
 
-    static constexpr Derived const* GetParent(const IntrusiveRedBlackTreeNode* node) {
-        return GetParentPointer<Member, Derived>(node);
+    static Derived const* GetParent(IntrusiveRedBlackTreeNode const* node) {
+        return Common::GetParentPointer<Member, Derived>(node);
     }
 };
 
 template <class Derived>
-class IntrusiveRedBlackTreeBaseNode : public IntrusiveRedBlackTreeNode {
+class alignas(void*) IntrusiveRedBlackTreeBaseNode : public IntrusiveRedBlackTreeNode {
 public:
+    using IntrusiveRedBlackTreeNode::IntrusiveRedBlackTreeNode;
+
     constexpr Derived* GetPrev() {
-        return static_cast<Derived*>(impl::IntrusiveRedBlackTreeImpl::GetPrev(this));
+        return static_cast<Derived*>(static_cast<IntrusiveRedBlackTreeBaseNode*>(
+            impl::IntrusiveRedBlackTreeImpl::GetPrev(this)));
     }
     constexpr const Derived* GetPrev() const {
-        return static_cast<const Derived*>(impl::IntrusiveRedBlackTreeImpl::GetPrev(this));
+        return static_cast<const Derived*>(static_cast<const IntrusiveRedBlackTreeBaseNode*>(
+            impl::IntrusiveRedBlackTreeImpl::GetPrev(this)));
     }
 
     constexpr Derived* GetNext() {
-        return static_cast<Derived*>(impl::IntrusiveRedBlackTreeImpl::GetNext(this));
+        return static_cast<Derived*>(static_cast<IntrusiveRedBlackTreeBaseNode*>(
+            impl::IntrusiveRedBlackTreeImpl::GetNext(this)));
     }
     constexpr const Derived* GetNext() const {
-        return static_cast<const Derived*>(impl::IntrusiveRedBlackTreeImpl::GetNext(this));
+        return static_cast<const Derived*>(static_cast<const IntrusiveRedBlackTreeBaseNode*>(
+            impl::IntrusiveRedBlackTreeImpl::GetNext(this)));
     }
 };
 
@@ -581,19 +607,22 @@ private:
     friend class impl::IntrusiveRedBlackTreeImpl;
 
     static constexpr IntrusiveRedBlackTreeNode* GetNode(Derived* parent) {
-        return static_cast<IntrusiveRedBlackTreeNode*>(parent);
+        return static_cast<IntrusiveRedBlackTreeNode*>(
+            static_cast<IntrusiveRedBlackTreeBaseNode<Derived>*>(parent));
     }
 
     static constexpr IntrusiveRedBlackTreeNode const* GetNode(Derived const* parent) {
-        return static_cast<const IntrusiveRedBlackTreeNode*>(parent);
+        return static_cast<const IntrusiveRedBlackTreeNode*>(
+            static_cast<const IntrusiveRedBlackTreeBaseNode<Derived>*>(parent));
     }
 
     static constexpr Derived* GetParent(IntrusiveRedBlackTreeNode* node) {
-        return static_cast<Derived*>(node);
+        return static_cast<Derived*>(static_cast<IntrusiveRedBlackTreeBaseNode<Derived>*>(node));
     }
 
-    static constexpr Derived const* GetParent(const IntrusiveRedBlackTreeNode* node) {
-        return static_cast<const Derived*>(node);
+    static constexpr Derived const* GetParent(IntrusiveRedBlackTreeNode const* node) {
+        return static_cast<const Derived*>(
+            static_cast<const IntrusiveRedBlackTreeBaseNode<Derived>*>(node));
     }
 };
 
diff --git a/src/common/logging/backend.cpp b/src/common/logging/backend.cpp
index c51c05b28..4a2462ec4 100644
--- a/src/common/logging/backend.cpp
+++ b/src/common/logging/backend.cpp
@@ -218,19 +218,17 @@ private:
     Impl(const std::filesystem::path& file_backend_filename, const Filter& filter_)
         : filter{filter_}, file_backend{file_backend_filename} {}
 
-    ~Impl() {
-        StopBackendThread();
-    }
+    ~Impl() = default;
 
     void StartBackendThread() {
-        backend_thread = std::thread([this] {
+        backend_thread = std::jthread([this](std::stop_token stop_token) {
             Common::SetCurrentThreadName("yuzu:Log");
             Entry entry;
             const auto write_logs = [this, &entry]() {
                 ForEachBackend([&entry](Backend& backend) { backend.Write(entry); });
             };
-            while (!stop.stop_requested()) {
-                entry = message_queue.PopWait(stop.get_token());
+            while (!stop_token.stop_requested()) {
+                entry = message_queue.PopWait(stop_token);
                 if (entry.filename != nullptr) {
                     write_logs();
                 }
@@ -244,11 +242,6 @@ private:
         });
     }
 
-    void StopBackendThread() {
-        stop.request_stop();
-        backend_thread.join();
-    }
-
     Entry CreateEntry(Class log_class, Level log_level, const char* filename, unsigned int line_nr,
                       const char* function, std::string&& message) const {
         using std::chrono::duration_cast;
@@ -283,10 +276,9 @@ private:
     ColorConsoleBackend color_console_backend{};
     FileBackend file_backend;
 
-    std::stop_source stop;
-    std::thread backend_thread;
     MPSCQueue<Entry, true> message_queue{};
     std::chrono::steady_clock::time_point time_origin{std::chrono::steady_clock::now()};
+    std::jthread backend_thread;
 };
 } // namespace
 
diff --git a/src/common/logging/filter.cpp b/src/common/logging/filter.cpp
index b898a652c..4afc1369a 100644
--- a/src/common/logging/filter.cpp
+++ b/src/common/logging/filter.cpp
@@ -108,6 +108,7 @@ bool ParseFilterRule(Filter& instance, Iterator begin, Iterator end) {
     SUB(Service, Migration)                                                                        \
     SUB(Service, Mii)                                                                              \
     SUB(Service, MM)                                                                               \
+    SUB(Service, MNPP)                                                                             \
     SUB(Service, NCM)                                                                              \
     SUB(Service, NFC)                                                                              \
     SUB(Service, NFP)                                                                              \
diff --git a/src/common/logging/types.h b/src/common/logging/types.h
index 9ed0c7ad6..2b6e4daa7 100644
--- a/src/common/logging/types.h
+++ b/src/common/logging/types.h
@@ -76,6 +76,7 @@ enum class Class : u8 {
     Service_Migration, ///< The migration service
     Service_Mii,       ///< The Mii service
     Service_MM,        ///< The MM (Multimedia) service
+    Service_MNPP,      ///< The MNPP service
     Service_NCM,       ///< The NCM service
     Service_NFC,       ///< The NFC (Near-field communication) service
     Service_NFP,       ///< The NFP service
diff --git a/src/common/page_table.cpp b/src/common/page_table.cpp
index 9fffd816f..4817b09f9 100644
--- a/src/common/page_table.cpp
+++ b/src/common/page_table.cpp
@@ -10,11 +10,65 @@ PageTable::PageTable() = default;
 
 PageTable::~PageTable() noexcept = default;
 
-void PageTable::Resize(size_t address_space_width_in_bits, size_t page_size_in_bits) {
-    const size_t num_page_table_entries{1ULL << (address_space_width_in_bits - page_size_in_bits)};
+bool PageTable::BeginTraversal(TraversalEntry& out_entry, TraversalContext& out_context,
+                               u64 address) const {
+    // Setup invalid defaults.
+    out_entry.phys_addr = 0;
+    out_entry.block_size = page_size;
+    out_context.next_page = 0;
+
+    // Validate that we can read the actual entry.
+    const auto page = address / page_size;
+    if (page >= backing_addr.size()) {
+        return false;
+    }
+
+    // Validate that the entry is mapped.
+    const auto phys_addr = backing_addr[page];
+    if (phys_addr == 0) {
+        return false;
+    }
+
+    // Populate the results.
+    out_entry.phys_addr = phys_addr + address;
+    out_context.next_page = page + 1;
+    out_context.next_offset = address + page_size;
+
+    return true;
+}
+
+bool PageTable::ContinueTraversal(TraversalEntry& out_entry, TraversalContext& context) const {
+    // Setup invalid defaults.
+    out_entry.phys_addr = 0;
+    out_entry.block_size = page_size;
+
+    // Validate that we can read the actual entry.
+    const auto page = context.next_page;
+    if (page >= backing_addr.size()) {
+        return false;
+    }
+
+    // Validate that the entry is mapped.
+    const auto phys_addr = backing_addr[page];
+    if (phys_addr == 0) {
+        return false;
+    }
+
+    // Populate the results.
+    out_entry.phys_addr = phys_addr + context.next_offset;
+    context.next_page = page + 1;
+    context.next_offset += page_size;
+
+    return true;
+}
+
+void PageTable::Resize(std::size_t address_space_width_in_bits, std::size_t page_size_in_bits) {
+    const std::size_t num_page_table_entries{1ULL
+                                             << (address_space_width_in_bits - page_size_in_bits)};
     pointers.resize(num_page_table_entries);
     backing_addr.resize(num_page_table_entries);
     current_address_space_width_in_bits = address_space_width_in_bits;
+    page_size = 1ULL << page_size_in_bits;
 }
 
 } // namespace Common
diff --git a/src/common/page_table.h b/src/common/page_table.h
index 8267e8b4d..82d91e9f3 100644
--- a/src/common/page_table.h
+++ b/src/common/page_table.h
@@ -27,6 +27,16 @@ enum class PageType : u8 {
  * mimics the way a real CPU page table works.
  */
 struct PageTable {
+    struct TraversalEntry {
+        u64 phys_addr{};
+        std::size_t block_size{};
+    };
+
+    struct TraversalContext {
+        u64 next_page{};
+        u64 next_offset{};
+    };
+
     /// Number of bits reserved for attribute tagging.
     /// This can be at most the guaranteed alignment of the pointers in the page table.
     static constexpr int ATTRIBUTE_BITS = 2;
@@ -89,6 +99,10 @@ struct PageTable {
     PageTable(PageTable&&) noexcept = default;
     PageTable& operator=(PageTable&&) noexcept = default;
 
+    bool BeginTraversal(TraversalEntry& out_entry, TraversalContext& out_context,
+                        u64 address) const;
+    bool ContinueTraversal(TraversalEntry& out_entry, TraversalContext& context) const;
+
     /**
      * Resizes the page table to be able to accommodate enough pages within
      * a given address space.
@@ -96,9 +110,9 @@ struct PageTable {
      * @param address_space_width_in_bits The address size width in bits.
      * @param page_size_in_bits           The page size in bits.
      */
-    void Resize(size_t address_space_width_in_bits, size_t page_size_in_bits);
+    void Resize(std::size_t address_space_width_in_bits, std::size_t page_size_in_bits);
 
-    size_t GetAddressSpaceBits() const {
+    std::size_t GetAddressSpaceBits() const {
         return current_address_space_width_in_bits;
     }
 
@@ -110,9 +124,11 @@ struct PageTable {
 
     VirtualBuffer<u64> backing_addr;
 
-    size_t current_address_space_width_in_bits;
+    std::size_t current_address_space_width_in_bits{};
+
+    u8* fastmem_arena{};
 
-    u8* fastmem_arena;
+    std::size_t page_size{};
 };
 
 } // namespace Common
diff --git a/src/common/settings.cpp b/src/common/settings.cpp
index 6964a8273..877e0faa4 100644
--- a/src/common/settings.cpp
+++ b/src/common/settings.cpp
@@ -167,6 +167,7 @@ void RestoreGlobalState(bool is_powered_on) {
 
     // Core
     values.use_multi_core.SetGlobal(true);
+    values.use_extended_memory_layout.SetGlobal(true);
 
     // CPU
     values.cpu_accuracy.SetGlobal(true);
@@ -175,6 +176,7 @@ void RestoreGlobalState(bool is_powered_on) {
     values.cpuopt_unsafe_ignore_standard_fpcr.SetGlobal(true);
     values.cpuopt_unsafe_inaccurate_nan.SetGlobal(true);
     values.cpuopt_unsafe_fastmem_check.SetGlobal(true);
+    values.cpuopt_unsafe_ignore_global_monitor.SetGlobal(true);
 
     // Renderer
     values.renderer_backend.SetGlobal(true);
diff --git a/src/common/settings.h b/src/common/settings.h
index 9bee6e10f..a37d83fb3 100644
--- a/src/common/settings.h
+++ b/src/common/settings.h
@@ -466,6 +466,7 @@ struct Values {
 
     // Core
     Setting<bool> use_multi_core{true, "use_multi_core"};
+    Setting<bool> use_extended_memory_layout{false, "use_extended_memory_layout"};
 
     // Cpu
     RangedSetting<CPUAccuracy> cpu_accuracy{CPUAccuracy::Auto, CPUAccuracy::Auto,
@@ -483,12 +484,15 @@ struct Values {
     BasicSetting<bool> cpuopt_misc_ir{true, "cpuopt_misc_ir"};
     BasicSetting<bool> cpuopt_reduce_misalign_checks{true, "cpuopt_reduce_misalign_checks"};
     BasicSetting<bool> cpuopt_fastmem{true, "cpuopt_fastmem"};
+    BasicSetting<bool> cpuopt_fastmem_exclusives{true, "cpuopt_fastmem_exclusives"};
+    BasicSetting<bool> cpuopt_recompile_exclusives{true, "cpuopt_recompile_exclusives"};
 
     Setting<bool> cpuopt_unsafe_unfuse_fma{true, "cpuopt_unsafe_unfuse_fma"};
     Setting<bool> cpuopt_unsafe_reduce_fp_error{true, "cpuopt_unsafe_reduce_fp_error"};
     Setting<bool> cpuopt_unsafe_ignore_standard_fpcr{true, "cpuopt_unsafe_ignore_standard_fpcr"};
     Setting<bool> cpuopt_unsafe_inaccurate_nan{true, "cpuopt_unsafe_inaccurate_nan"};
     Setting<bool> cpuopt_unsafe_fastmem_check{true, "cpuopt_unsafe_fastmem_check"};
+    Setting<bool> cpuopt_unsafe_ignore_global_monitor{true, "cpuopt_unsafe_ignore_global_monitor"};
 
     // Renderer
     RangedSetting<RendererBackend> renderer_backend{
diff --git a/src/common/telemetry.cpp b/src/common/telemetry.cpp
index 6241d08b3..98c82cd17 100644
--- a/src/common/telemetry.cpp
+++ b/src/common/telemetry.cpp
@@ -55,22 +55,50 @@ void AppendBuildInfo(FieldCollection& fc) {
 
 void AppendCPUInfo(FieldCollection& fc) {
 #ifdef ARCHITECTURE_x86_64
-    fc.AddField(FieldType::UserSystem, "CPU_Model", Common::GetCPUCaps().cpu_string);
-    fc.AddField(FieldType::UserSystem, "CPU_BrandString", Common::GetCPUCaps().brand_string);
-    fc.AddField(FieldType::UserSystem, "CPU_Extension_x64_AES", Common::GetCPUCaps().aes);
-    fc.AddField(FieldType::UserSystem, "CPU_Extension_x64_AVX", Common::GetCPUCaps().avx);
-    fc.AddField(FieldType::UserSystem, "CPU_Extension_x64_AVX2", Common::GetCPUCaps().avx2);
-    fc.AddField(FieldType::UserSystem, "CPU_Extension_x64_AVX512", Common::GetCPUCaps().avx512);
-    fc.AddField(FieldType::UserSystem, "CPU_Extension_x64_BMI1", Common::GetCPUCaps().bmi1);
-    fc.AddField(FieldType::UserSystem, "CPU_Extension_x64_BMI2", Common::GetCPUCaps().bmi2);
-    fc.AddField(FieldType::UserSystem, "CPU_Extension_x64_FMA", Common::GetCPUCaps().fma);
-    fc.AddField(FieldType::UserSystem, "CPU_Extension_x64_FMA4", Common::GetCPUCaps().fma4);
-    fc.AddField(FieldType::UserSystem, "CPU_Extension_x64_SSE", Common::GetCPUCaps().sse);
-    fc.AddField(FieldType::UserSystem, "CPU_Extension_x64_SSE2", Common::GetCPUCaps().sse2);
-    fc.AddField(FieldType::UserSystem, "CPU_Extension_x64_SSE3", Common::GetCPUCaps().sse3);
-    fc.AddField(FieldType::UserSystem, "CPU_Extension_x64_SSSE3", Common::GetCPUCaps().ssse3);
-    fc.AddField(FieldType::UserSystem, "CPU_Extension_x64_SSE41", Common::GetCPUCaps().sse4_1);
-    fc.AddField(FieldType::UserSystem, "CPU_Extension_x64_SSE42", Common::GetCPUCaps().sse4_2);
+
+    const auto& caps = Common::GetCPUCaps();
+    const auto add_field = [&fc](std::string_view field_name, const auto& field_value) {
+        fc.AddField(FieldType::UserSystem, field_name, field_value);
+    };
+    add_field("CPU_Model", caps.cpu_string);
+    add_field("CPU_BrandString", caps.brand_string);
+
+    add_field("CPU_Extension_x64_SSE", caps.sse);
+    add_field("CPU_Extension_x64_SSE2", caps.sse2);
+    add_field("CPU_Extension_x64_SSE3", caps.sse3);
+    add_field("CPU_Extension_x64_SSSE3", caps.ssse3);
+    add_field("CPU_Extension_x64_SSE41", caps.sse4_1);
+    add_field("CPU_Extension_x64_SSE42", caps.sse4_2);
+
+    add_field("CPU_Extension_x64_AVX", caps.avx);
+    add_field("CPU_Extension_x64_AVX_VNNI", caps.avx_vnni);
+    add_field("CPU_Extension_x64_AVX2", caps.avx2);
+
+    // Skylake-X/SP level AVX512, for compatibility with the previous telemetry field
+    add_field("CPU_Extension_x64_AVX512",
+              caps.avx512f && caps.avx512cd && caps.avx512vl && caps.avx512dq && caps.avx512bw);
+
+    add_field("CPU_Extension_x64_AVX512F", caps.avx512f);
+    add_field("CPU_Extension_x64_AVX512CD", caps.avx512cd);
+    add_field("CPU_Extension_x64_AVX512VL", caps.avx512vl);
+    add_field("CPU_Extension_x64_AVX512DQ", caps.avx512dq);
+    add_field("CPU_Extension_x64_AVX512BW", caps.avx512bw);
+    add_field("CPU_Extension_x64_AVX512BITALG", caps.avx512bitalg);
+    add_field("CPU_Extension_x64_AVX512VBMI", caps.avx512vbmi);
+
+    add_field("CPU_Extension_x64_AES", caps.aes);
+    add_field("CPU_Extension_x64_BMI1", caps.bmi1);
+    add_field("CPU_Extension_x64_BMI2", caps.bmi2);
+    add_field("CPU_Extension_x64_F16C", caps.f16c);
+    add_field("CPU_Extension_x64_FMA", caps.fma);
+    add_field("CPU_Extension_x64_FMA4", caps.fma4);
+    add_field("CPU_Extension_x64_GFNI", caps.gfni);
+    add_field("CPU_Extension_x64_INVARIANT_TSC", caps.invariant_tsc);
+    add_field("CPU_Extension_x64_LZCNT", caps.lzcnt);
+    add_field("CPU_Extension_x64_MOVBE", caps.movbe);
+    add_field("CPU_Extension_x64_PCLMULQDQ", caps.pclmulqdq);
+    add_field("CPU_Extension_x64_POPCNT", caps.popcnt);
+    add_field("CPU_Extension_x64_SHA", caps.sha);
 #else
     fc.AddField(FieldType::UserSystem, "CPU_Model", "Other");
 #endif
diff --git a/src/common/telemetry.h b/src/common/telemetry.h
index 49186e848..3524c857e 100644
--- a/src/common/telemetry.h
+++ b/src/common/telemetry.h
@@ -8,6 +8,8 @@
 #include <map>
 #include <memory>
 #include <string>
+#include <string_view>
+#include "common/common_funcs.h"
 #include "common/common_types.h"
 
 namespace Common::Telemetry {
@@ -28,7 +30,7 @@ struct VisitorInterface;
 /**
  * Interface class for telemetry data fields.
  */
-class FieldInterface : NonCopyable {
+class FieldInterface {
 public:
     virtual ~FieldInterface() = default;
 
@@ -52,14 +54,15 @@ public:
 template <typename T>
 class Field : public FieldInterface {
 public:
-    Field(FieldType type_, std::string name_, T value_)
-        : name(std::move(name_)), type(type_), value(std::move(value_)) {}
+    YUZU_NON_COPYABLE(Field);
 
-    Field(const Field&) = default;
-    Field& operator=(const Field&) = default;
+    Field(FieldType type_, std::string_view name_, T value_)
+        : name(name_), type(type_), value(std::move(value_)) {}
 
-    Field(Field&&) = default;
-    Field& operator=(Field&& other) = default;
+    ~Field() override = default;
+
+    Field(Field&&) noexcept = default;
+    Field& operator=(Field&& other) noexcept = default;
 
     void Accept(VisitorInterface& visitor) const override;
 
@@ -98,9 +101,15 @@ private:
 /**
  * Collection of data fields that have been logged.
  */
-class FieldCollection final : NonCopyable {
+class FieldCollection final {
 public:
+    YUZU_NON_COPYABLE(FieldCollection);
+
     FieldCollection() = default;
+    ~FieldCollection() = default;
+
+    FieldCollection(FieldCollection&&) noexcept = default;
+    FieldCollection& operator=(FieldCollection&&) noexcept = default;
 
     /**
      * Accept method for the visitor pattern, visits each field in the collection.
@@ -115,7 +124,7 @@ public:
      * @param value Value for the field to add.
      */
     template <typename T>
-    void AddField(FieldType type, const char* name, T value) {
+    void AddField(FieldType type, std::string_view name, T value) {
         return AddField(std::make_unique<Field<T>>(type, name, std::move(value)));
     }
 
@@ -133,7 +142,7 @@ private:
  * Telemetry fields visitor interface class. A backend to log to a web service should implement
  * this interface.
  */
-struct VisitorInterface : NonCopyable {
+struct VisitorInterface {
     virtual ~VisitorInterface() = default;
 
     virtual void Visit(const Field<bool>& field) = 0;
@@ -160,8 +169,11 @@ struct VisitorInterface : NonCopyable {
  * Empty implementation of VisitorInterface that drops all fields. Used when a functional
  * backend implementation is not available.
  */
-struct NullVisitor : public VisitorInterface {
-    ~NullVisitor() = default;
+struct NullVisitor final : public VisitorInterface {
+    YUZU_NON_COPYABLE(NullVisitor);
+
+    NullVisitor() = default;
+    ~NullVisitor() override = default;
 
     void Visit(const Field<bool>& /*field*/) override {}
     void Visit(const Field<double>& /*field*/) override {}
diff --git a/src/common/tree.h b/src/common/tree.h
index 18faa4a48..28370e343 100644
--- a/src/common/tree.h
+++ b/src/common/tree.h
@@ -43,294 +43,265 @@
  * The maximum height of a red-black tree is 2lg (n+1).
  */
 
-#include "common/assert.h"
+namespace Common::freebsd {
 
-namespace Common {
+enum class RBColor {
+    RB_BLACK = 0,
+    RB_RED = 1,
+};
+
+#pragma pack(push, 4)
 template <typename T>
-class RBHead {
+class RBEntry {
 public:
-    [[nodiscard]] T* Root() {
-        return rbh_root;
-    }
+    constexpr RBEntry() = default;
 
-    [[nodiscard]] const T* Root() const {
-        return rbh_root;
+    [[nodiscard]] constexpr T* Left() {
+        return m_rbe_left;
     }
-
-    void SetRoot(T* root) {
-        rbh_root = root;
+    [[nodiscard]] constexpr const T* Left() const {
+        return m_rbe_left;
     }
 
-    [[nodiscard]] bool IsEmpty() const {
-        return Root() == nullptr;
+    constexpr void SetLeft(T* e) {
+        m_rbe_left = e;
     }
 
-private:
-    T* rbh_root = nullptr;
-};
-
-enum class EntryColor {
-    Black,
-    Red,
-};
-
-template <typename T>
-class RBEntry {
-public:
-    [[nodiscard]] T* Left() {
-        return rbe_left;
+    [[nodiscard]] constexpr T* Right() {
+        return m_rbe_right;
     }
-
-    [[nodiscard]] const T* Left() const {
-        return rbe_left;
+    [[nodiscard]] constexpr const T* Right() const {
+        return m_rbe_right;
     }
 
-    void SetLeft(T* left) {
-        rbe_left = left;
+    constexpr void SetRight(T* e) {
+        m_rbe_right = e;
     }
 
-    [[nodiscard]] T* Right() {
-        return rbe_right;
+    [[nodiscard]] constexpr T* Parent() {
+        return m_rbe_parent;
     }
-
-    [[nodiscard]] const T* Right() const {
-        return rbe_right;
+    [[nodiscard]] constexpr const T* Parent() const {
+        return m_rbe_parent;
     }
 
-    void SetRight(T* right) {
-        rbe_right = right;
+    constexpr void SetParent(T* e) {
+        m_rbe_parent = e;
     }
 
-    [[nodiscard]] T* Parent() {
-        return rbe_parent;
+    [[nodiscard]] constexpr bool IsBlack() const {
+        return m_rbe_color == RBColor::RB_BLACK;
     }
-
-    [[nodiscard]] const T* Parent() const {
-        return rbe_parent;
+    [[nodiscard]] constexpr bool IsRed() const {
+        return m_rbe_color == RBColor::RB_RED;
     }
-
-    void SetParent(T* parent) {
-        rbe_parent = parent;
+    [[nodiscard]] constexpr RBColor Color() const {
+        return m_rbe_color;
     }
 
-    [[nodiscard]] bool IsBlack() const {
-        return rbe_color == EntryColor::Black;
+    constexpr void SetColor(RBColor c) {
+        m_rbe_color = c;
     }
 
-    [[nodiscard]] bool IsRed() const {
-        return rbe_color == EntryColor::Red;
-    }
+private:
+    T* m_rbe_left{};
+    T* m_rbe_right{};
+    T* m_rbe_parent{};
+    RBColor m_rbe_color{RBColor::RB_BLACK};
+};
+#pragma pack(pop)
 
-    [[nodiscard]] EntryColor Color() const {
-        return rbe_color;
-    }
+template <typename T>
+struct CheckRBEntry {
+    static constexpr bool value = false;
+};
+template <typename T>
+struct CheckRBEntry<RBEntry<T>> {
+    static constexpr bool value = true;
+};
 
-    void SetColor(EntryColor color) {
-        rbe_color = color;
-    }
+template <typename T>
+concept IsRBEntry = CheckRBEntry<T>::value;
 
+template <typename T>
+concept HasRBEntry = requires(T& t, const T& ct) {
+    { t.GetRBEntry() } -> std::same_as<RBEntry<T>&>;
+    { ct.GetRBEntry() } -> std::same_as<const RBEntry<T>&>;
+};
+
+template <typename T>
+requires HasRBEntry<T>
+class RBHead {
 private:
-    T* rbe_left = nullptr;
-    T* rbe_right = nullptr;
-    T* rbe_parent = nullptr;
-    EntryColor rbe_color{};
+    T* m_rbh_root = nullptr;
+
+public:
+    [[nodiscard]] constexpr T* Root() {
+        return m_rbh_root;
+    }
+    [[nodiscard]] constexpr const T* Root() const {
+        return m_rbh_root;
+    }
+    constexpr void SetRoot(T* root) {
+        m_rbh_root = root;
+    }
+
+    [[nodiscard]] constexpr bool IsEmpty() const {
+        return this->Root() == nullptr;
+    }
 };
 
-template <typename Node>
-[[nodiscard]] RBEntry<Node>& RB_ENTRY(Node* node) {
-    return node->GetEntry();
+template <typename T>
+requires HasRBEntry<T>
+[[nodiscard]] constexpr RBEntry<T>& RB_ENTRY(T* t) {
+    return t->GetRBEntry();
 }
-
-template <typename Node>
-[[nodiscard]] const RBEntry<Node>& RB_ENTRY(const Node* node) {
-    return node->GetEntry();
+template <typename T>
+requires HasRBEntry<T>
+[[nodiscard]] constexpr const RBEntry<T>& RB_ENTRY(const T* t) {
+    return t->GetRBEntry();
 }
 
-template <typename Node>
-[[nodiscard]] Node* RB_PARENT(Node* node) {
-    return RB_ENTRY(node).Parent();
+template <typename T>
+requires HasRBEntry<T>
+[[nodiscard]] constexpr T* RB_LEFT(T* t) {
+    return RB_ENTRY(t).Left();
 }
-
-template <typename Node>
-[[nodiscard]] const Node* RB_PARENT(const Node* node) {
-    return RB_ENTRY(node).Parent();
+template <typename T>
+requires HasRBEntry<T>
+[[nodiscard]] constexpr const T* RB_LEFT(const T* t) {
+    return RB_ENTRY(t).Left();
 }
 
-template <typename Node>
-void RB_SET_PARENT(Node* node, Node* parent) {
-    return RB_ENTRY(node).SetParent(parent);
+template <typename T>
+requires HasRBEntry<T>
+[[nodiscard]] constexpr T* RB_RIGHT(T* t) {
+    return RB_ENTRY(t).Right();
 }
-
-template <typename Node>
-[[nodiscard]] Node* RB_LEFT(Node* node) {
-    return RB_ENTRY(node).Left();
+template <typename T>
+requires HasRBEntry<T>
+[[nodiscard]] constexpr const T* RB_RIGHT(const T* t) {
+    return RB_ENTRY(t).Right();
 }
 
-template <typename Node>
-[[nodiscard]] const Node* RB_LEFT(const Node* node) {
-    return RB_ENTRY(node).Left();
+template <typename T>
+requires HasRBEntry<T>
+[[nodiscard]] constexpr T* RB_PARENT(T* t) {
+    return RB_ENTRY(t).Parent();
 }
-
-template <typename Node>
-void RB_SET_LEFT(Node* node, Node* left) {
-    return RB_ENTRY(node).SetLeft(left);
+template <typename T>
+requires HasRBEntry<T>
+[[nodiscard]] constexpr const T* RB_PARENT(const T* t) {
+    return RB_ENTRY(t).Parent();
 }
 
-template <typename Node>
-[[nodiscard]] Node* RB_RIGHT(Node* node) {
-    return RB_ENTRY(node).Right();
+template <typename T>
+requires HasRBEntry<T>
+constexpr void RB_SET_LEFT(T* t, T* e) {
+    RB_ENTRY(t).SetLeft(e);
 }
-
-template <typename Node>
-[[nodiscard]] const Node* RB_RIGHT(const Node* node) {
-    return RB_ENTRY(node).Right();
+template <typename T>
+requires HasRBEntry<T>
+constexpr void RB_SET_RIGHT(T* t, T* e) {
+    RB_ENTRY(t).SetRight(e);
 }
-
-template <typename Node>
-void RB_SET_RIGHT(Node* node, Node* right) {
-    return RB_ENTRY(node).SetRight(right);
+template <typename T>
+requires HasRBEntry<T>
+constexpr void RB_SET_PARENT(T* t, T* e) {
+    RB_ENTRY(t).SetParent(e);
 }
 
-template <typename Node>
-[[nodiscard]] bool RB_IS_BLACK(const Node* node) {
-    return RB_ENTRY(node).IsBlack();
+template <typename T>
+requires HasRBEntry<T>
+[[nodiscard]] constexpr bool RB_IS_BLACK(const T* t) {
+    return RB_ENTRY(t).IsBlack();
 }
-
-template <typename Node>
-[[nodiscard]] bool RB_IS_RED(const Node* node) {
-    return RB_ENTRY(node).IsRed();
+template <typename T>
+requires HasRBEntry<T>
+[[nodiscard]] constexpr bool RB_IS_RED(const T* t) {
+    return RB_ENTRY(t).IsRed();
 }
 
-template <typename Node>
-[[nodiscard]] EntryColor RB_COLOR(const Node* node) {
-    return RB_ENTRY(node).Color();
+template <typename T>
+requires HasRBEntry<T>
+[[nodiscard]] constexpr RBColor RB_COLOR(const T* t) {
+    return RB_ENTRY(t).Color();
 }
 
-template <typename Node>
-void RB_SET_COLOR(Node* node, EntryColor color) {
-    return RB_ENTRY(node).SetColor(color);
+template <typename T>
+requires HasRBEntry<T>
+constexpr void RB_SET_COLOR(T* t, RBColor c) {
+    RB_ENTRY(t).SetColor(c);
 }
 
-template <typename Node>
-void RB_SET(Node* node, Node* parent) {
-    auto& entry = RB_ENTRY(node);
-    entry.SetParent(parent);
-    entry.SetLeft(nullptr);
-    entry.SetRight(nullptr);
-    entry.SetColor(EntryColor::Red);
+template <typename T>
+requires HasRBEntry<T>
+constexpr void RB_SET(T* elm, T* parent) {
+    auto& rb_entry = RB_ENTRY(elm);
+    rb_entry.SetParent(parent);
+    rb_entry.SetLeft(nullptr);
+    rb_entry.SetRight(nullptr);
+    rb_entry.SetColor(RBColor::RB_RED);
 }
 
-template <typename Node>
-void RB_SET_BLACKRED(Node* black, Node* red) {
-    RB_SET_COLOR(black, EntryColor::Black);
-    RB_SET_COLOR(red, EntryColor::Red);
+template <typename T>
+requires HasRBEntry<T>
+constexpr void RB_SET_BLACKRED(T* black, T* red) {
+    RB_SET_COLOR(black, RBColor::RB_BLACK);
+    RB_SET_COLOR(red, RBColor::RB_RED);
 }
 
-template <typename Node>
-void RB_ROTATE_LEFT(RBHead<Node>* head, Node* elm, Node*& tmp) {
+template <typename T>
+requires HasRBEntry<T>
+constexpr void RB_ROTATE_LEFT(RBHead<T>& head, T* elm, T*& tmp) {
     tmp = RB_RIGHT(elm);
-    RB_SET_RIGHT(elm, RB_LEFT(tmp));
-    if (RB_RIGHT(elm) != nullptr) {
+    if (RB_SET_RIGHT(elm, RB_LEFT(tmp)); RB_RIGHT(elm) != nullptr) {
         RB_SET_PARENT(RB_LEFT(tmp), elm);
     }
 
-    RB_SET_PARENT(tmp, RB_PARENT(elm));
-    if (RB_PARENT(tmp) != nullptr) {
+    if (RB_SET_PARENT(tmp, RB_PARENT(elm)); RB_PARENT(tmp) != nullptr) {
         if (elm == RB_LEFT(RB_PARENT(elm))) {
             RB_SET_LEFT(RB_PARENT(elm), tmp);
         } else {
             RB_SET_RIGHT(RB_PARENT(elm), tmp);
         }
     } else {
-        head->SetRoot(tmp);
+        head.SetRoot(tmp);
     }
 
     RB_SET_LEFT(tmp, elm);
     RB_SET_PARENT(elm, tmp);
 }
 
-template <typename Node>
-void RB_ROTATE_RIGHT(RBHead<Node>* head, Node* elm, Node*& tmp) {
+template <typename T>
+requires HasRBEntry<T>
+constexpr void RB_ROTATE_RIGHT(RBHead<T>& head, T* elm, T*& tmp) {
     tmp = RB_LEFT(elm);
-    RB_SET_LEFT(elm, RB_RIGHT(tmp));
-    if (RB_LEFT(elm) != nullptr) {
+    if (RB_SET_LEFT(elm, RB_RIGHT(tmp)); RB_LEFT(elm) != nullptr) {
         RB_SET_PARENT(RB_RIGHT(tmp), elm);
     }
 
-    RB_SET_PARENT(tmp, RB_PARENT(elm));
-    if (RB_PARENT(tmp) != nullptr) {
+    if (RB_SET_PARENT(tmp, RB_PARENT(elm)); RB_PARENT(tmp) != nullptr) {
         if (elm == RB_LEFT(RB_PARENT(elm))) {
             RB_SET_LEFT(RB_PARENT(elm), tmp);
         } else {
             RB_SET_RIGHT(RB_PARENT(elm), tmp);
         }
     } else {
-        head->SetRoot(tmp);
+        head.SetRoot(tmp);
     }
 
     RB_SET_RIGHT(tmp, elm);
     RB_SET_PARENT(elm, tmp);
 }
 
-template <typename Node>
-void RB_INSERT_COLOR(RBHead<Node>* head, Node* elm) {
-    Node* parent = nullptr;
-    Node* tmp = nullptr;
-
-    while ((parent = RB_PARENT(elm)) != nullptr && RB_IS_RED(parent)) {
-        Node* gparent = RB_PARENT(parent);
-        if (parent == RB_LEFT(gparent)) {
-            tmp = RB_RIGHT(gparent);
-            if (tmp && RB_IS_RED(tmp)) {
-                RB_SET_COLOR(tmp, EntryColor::Black);
-                RB_SET_BLACKRED(parent, gparent);
-                elm = gparent;
-                continue;
-            }
-
-            if (RB_RIGHT(parent) == elm) {
-                RB_ROTATE_LEFT(head, parent, tmp);
-                tmp = parent;
-                parent = elm;
-                elm = tmp;
-            }
-
-            RB_SET_BLACKRED(parent, gparent);
-            RB_ROTATE_RIGHT(head, gparent, tmp);
-        } else {
-            tmp = RB_LEFT(gparent);
-            if (tmp && RB_IS_RED(tmp)) {
-                RB_SET_COLOR(tmp, EntryColor::Black);
-                RB_SET_BLACKRED(parent, gparent);
-                elm = gparent;
-                continue;
-            }
-
-            if (RB_LEFT(parent) == elm) {
-                RB_ROTATE_RIGHT(head, parent, tmp);
-                tmp = parent;
-                parent = elm;
-                elm = tmp;
-            }
-
-            RB_SET_BLACKRED(parent, gparent);
-            RB_ROTATE_LEFT(head, gparent, tmp);
-        }
-    }
-
-    RB_SET_COLOR(head->Root(), EntryColor::Black);
-}
-
-template <typename Node>
-void RB_REMOVE_COLOR(RBHead<Node>* head, Node* parent, Node* elm) {
-    Node* tmp;
-    while ((elm == nullptr || RB_IS_BLACK(elm)) && elm != head->Root() && parent != nullptr) {
+template <typename T>
+requires HasRBEntry<T>
+constexpr void RB_REMOVE_COLOR(RBHead<T>& head, T* parent, T* elm) {
+    T* tmp;
+    while ((elm == nullptr || RB_IS_BLACK(elm)) && elm != head.Root()) {
         if (RB_LEFT(parent) == elm) {
             tmp = RB_RIGHT(parent);
-            if (!tmp) {
-                ASSERT_MSG(false, "tmp is invalid!");
-                break;
-            }
             if (RB_IS_RED(tmp)) {
                 RB_SET_BLACKRED(tmp, parent);
                 RB_ROTATE_LEFT(head, parent, tmp);
@@ -339,29 +310,29 @@ void RB_REMOVE_COLOR(RBHead<Node>* head, Node* parent, Node* elm) {
 
             if ((RB_LEFT(tmp) == nullptr || RB_IS_BLACK(RB_LEFT(tmp))) &&
                 (RB_RIGHT(tmp) == nullptr || RB_IS_BLACK(RB_RIGHT(tmp)))) {
-                RB_SET_COLOR(tmp, EntryColor::Red);
+                RB_SET_COLOR(tmp, RBColor::RB_RED);
                 elm = parent;
                 parent = RB_PARENT(elm);
             } else {
                 if (RB_RIGHT(tmp) == nullptr || RB_IS_BLACK(RB_RIGHT(tmp))) {
-                    Node* oleft;
+                    T* oleft;
                     if ((oleft = RB_LEFT(tmp)) != nullptr) {
-                        RB_SET_COLOR(oleft, EntryColor::Black);
+                        RB_SET_COLOR(oleft, RBColor::RB_BLACK);
                     }
 
-                    RB_SET_COLOR(tmp, EntryColor::Red);
+                    RB_SET_COLOR(tmp, RBColor::RB_RED);
                     RB_ROTATE_RIGHT(head, tmp, oleft);
                     tmp = RB_RIGHT(parent);
                 }
 
                 RB_SET_COLOR(tmp, RB_COLOR(parent));
-                RB_SET_COLOR(parent, EntryColor::Black);
+                RB_SET_COLOR(parent, RBColor::RB_BLACK);
                 if (RB_RIGHT(tmp)) {
-                    RB_SET_COLOR(RB_RIGHT(tmp), EntryColor::Black);
+                    RB_SET_COLOR(RB_RIGHT(tmp), RBColor::RB_BLACK);
                 }
 
                 RB_ROTATE_LEFT(head, parent, tmp);
-                elm = head->Root();
+                elm = head.Root();
                 break;
             }
         } else {
@@ -372,68 +343,56 @@ void RB_REMOVE_COLOR(RBHead<Node>* head, Node* parent, Node* elm) {
                 tmp = RB_LEFT(parent);
             }
 
-            if (!tmp) {
-                ASSERT_MSG(false, "tmp is invalid!");
-                break;
-            }
-
             if ((RB_LEFT(tmp) == nullptr || RB_IS_BLACK(RB_LEFT(tmp))) &&
                 (RB_RIGHT(tmp) == nullptr || RB_IS_BLACK(RB_RIGHT(tmp)))) {
-                RB_SET_COLOR(tmp, EntryColor::Red);
+                RB_SET_COLOR(tmp, RBColor::RB_RED);
                 elm = parent;
                 parent = RB_PARENT(elm);
             } else {
                 if (RB_LEFT(tmp) == nullptr || RB_IS_BLACK(RB_LEFT(tmp))) {
-                    Node* oright;
+                    T* oright;
                     if ((oright = RB_RIGHT(tmp)) != nullptr) {
-                        RB_SET_COLOR(oright, EntryColor::Black);
+                        RB_SET_COLOR(oright, RBColor::RB_BLACK);
                     }
 
-                    RB_SET_COLOR(tmp, EntryColor::Red);
+                    RB_SET_COLOR(tmp, RBColor::RB_RED);
                     RB_ROTATE_LEFT(head, tmp, oright);
                     tmp = RB_LEFT(parent);
                 }
 
                 RB_SET_COLOR(tmp, RB_COLOR(parent));
-                RB_SET_COLOR(parent, EntryColor::Black);
+                RB_SET_COLOR(parent, RBColor::RB_BLACK);
 
                 if (RB_LEFT(tmp)) {
-                    RB_SET_COLOR(RB_LEFT(tmp), EntryColor::Black);
+                    RB_SET_COLOR(RB_LEFT(tmp), RBColor::RB_BLACK);
                 }
 
                 RB_ROTATE_RIGHT(head, parent, tmp);
-                elm = head->Root();
+                elm = head.Root();
                 break;
             }
         }
     }
 
     if (elm) {
-        RB_SET_COLOR(elm, EntryColor::Black);
+        RB_SET_COLOR(elm, RBColor::RB_BLACK);
     }
 }
 
-template <typename Node>
-Node* RB_REMOVE(RBHead<Node>* head, Node* elm) {
-    Node* child = nullptr;
-    Node* parent = nullptr;
-    Node* old = elm;
-    EntryColor color{};
-
-    const auto finalize = [&] {
-        if (color == EntryColor::Black) {
-            RB_REMOVE_COLOR(head, parent, child);
-        }
-
-        return old;
-    };
+template <typename T>
+requires HasRBEntry<T>
+constexpr T* RB_REMOVE(RBHead<T>& head, T* elm) {
+    T* child = nullptr;
+    T* parent = nullptr;
+    T* old = elm;
+    RBColor color = RBColor::RB_BLACK;
 
     if (RB_LEFT(elm) == nullptr) {
         child = RB_RIGHT(elm);
     } else if (RB_RIGHT(elm) == nullptr) {
         child = RB_LEFT(elm);
     } else {
-        Node* left;
+        T* left;
         elm = RB_RIGHT(elm);
         while ((left = RB_LEFT(elm)) != nullptr) {
             elm = left;
@@ -446,6 +405,7 @@ Node* RB_REMOVE(RBHead<Node>* head, Node* elm) {
         if (child) {
             RB_SET_PARENT(child, parent);
         }
+
         if (parent) {
             if (RB_LEFT(parent) == elm) {
                 RB_SET_LEFT(parent, child);
@@ -453,14 +413,14 @@ Node* RB_REMOVE(RBHead<Node>* head, Node* elm) {
                 RB_SET_RIGHT(parent, child);
             }
         } else {
-            head->SetRoot(child);
+            head.SetRoot(child);
         }
 
         if (RB_PARENT(elm) == old) {
             parent = elm;
         }
 
-        elm->SetEntry(old->GetEntry());
+        elm->SetRBEntry(old->GetRBEntry());
 
         if (RB_PARENT(old)) {
             if (RB_LEFT(RB_PARENT(old)) == old) {
@@ -469,17 +429,24 @@ Node* RB_REMOVE(RBHead<Node>* head, Node* elm) {
                 RB_SET_RIGHT(RB_PARENT(old), elm);
             }
         } else {
-            head->SetRoot(elm);
+            head.SetRoot(elm);
         }
+
         RB_SET_PARENT(RB_LEFT(old), elm);
+
         if (RB_RIGHT(old)) {
             RB_SET_PARENT(RB_RIGHT(old), elm);
         }
+
         if (parent) {
             left = parent;
         }
 
-        return finalize();
+        if (color == RBColor::RB_BLACK) {
+            RB_REMOVE_COLOR(head, parent, child);
+        }
+
+        return old;
     }
 
     parent = RB_PARENT(elm);
@@ -495,17 +462,69 @@ Node* RB_REMOVE(RBHead<Node>* head, Node* elm) {
             RB_SET_RIGHT(parent, child);
         }
     } else {
-        head->SetRoot(child);
+        head.SetRoot(child);
+    }
+
+    if (color == RBColor::RB_BLACK) {
+        RB_REMOVE_COLOR(head, parent, child);
+    }
+
+    return old;
+}
+
+template <typename T>
+requires HasRBEntry<T>
+constexpr void RB_INSERT_COLOR(RBHead<T>& head, T* elm) {
+    T *parent = nullptr, *tmp = nullptr;
+    while ((parent = RB_PARENT(elm)) != nullptr && RB_IS_RED(parent)) {
+        T* gparent = RB_PARENT(parent);
+        if (parent == RB_LEFT(gparent)) {
+            tmp = RB_RIGHT(gparent);
+            if (tmp && RB_IS_RED(tmp)) {
+                RB_SET_COLOR(tmp, RBColor::RB_BLACK);
+                RB_SET_BLACKRED(parent, gparent);
+                elm = gparent;
+                continue;
+            }
+
+            if (RB_RIGHT(parent) == elm) {
+                RB_ROTATE_LEFT(head, parent, tmp);
+                tmp = parent;
+                parent = elm;
+                elm = tmp;
+            }
+
+            RB_SET_BLACKRED(parent, gparent);
+            RB_ROTATE_RIGHT(head, gparent, tmp);
+        } else {
+            tmp = RB_LEFT(gparent);
+            if (tmp && RB_IS_RED(tmp)) {
+                RB_SET_COLOR(tmp, RBColor::RB_BLACK);
+                RB_SET_BLACKRED(parent, gparent);
+                elm = gparent;
+                continue;
+            }
+
+            if (RB_LEFT(parent) == elm) {
+                RB_ROTATE_RIGHT(head, parent, tmp);
+                tmp = parent;
+                parent = elm;
+                elm = tmp;
+            }
+
+            RB_SET_BLACKRED(parent, gparent);
+            RB_ROTATE_LEFT(head, gparent, tmp);
+        }
     }
 
-    return finalize();
+    RB_SET_COLOR(head.Root(), RBColor::RB_BLACK);
 }
 
-// Inserts a node into the RB tree
-template <typename Node, typename CompareFunction>
-Node* RB_INSERT(RBHead<Node>* head, Node* elm, CompareFunction cmp) {
-    Node* parent = nullptr;
-    Node* tmp = head->Root();
+template <typename T, typename Compare>
+requires HasRBEntry<T>
+constexpr T* RB_INSERT(RBHead<T>& head, T* elm, Compare cmp) {
+    T* parent = nullptr;
+    T* tmp = head.Root();
     int comp = 0;
 
     while (tmp) {
@@ -529,17 +548,17 @@ Node* RB_INSERT(RBHead<Node>* head, Node* elm, CompareFunction cmp) {
             RB_SET_RIGHT(parent, elm);
         }
     } else {
-        head->SetRoot(elm);
+        head.SetRoot(elm);
     }
 
     RB_INSERT_COLOR(head, elm);
     return nullptr;
 }
 
-// Finds the node with the same key as elm
-template <typename Node, typename CompareFunction>
-Node* RB_FIND(RBHead<Node>* head, Node* elm, CompareFunction cmp) {
-    Node* tmp = head->Root();
+template <typename T, typename Compare>
+requires HasRBEntry<T>
+constexpr T* RB_FIND(RBHead<T>& head, T* elm, Compare cmp) {
+    T* tmp = head.Root();
 
     while (tmp) {
         const int comp = cmp(elm, tmp);
@@ -555,11 +574,11 @@ Node* RB_FIND(RBHead<Node>* head, Node* elm, CompareFunction cmp) {
     return nullptr;
 }
 
-// Finds the first node greater than or equal to the search key
-template <typename Node, typename CompareFunction>
-Node* RB_NFIND(RBHead<Node>* head, Node* elm, CompareFunction cmp) {
-    Node* tmp = head->Root();
-    Node* res = nullptr;
+template <typename T, typename Compare>
+requires HasRBEntry<T>
+constexpr T* RB_NFIND(RBHead<T>& head, T* elm, Compare cmp) {
+    T* tmp = head.Root();
+    T* res = nullptr;
 
     while (tmp) {
         const int comp = cmp(elm, tmp);
@@ -576,13 +595,13 @@ Node* RB_NFIND(RBHead<Node>* head, Node* elm, CompareFunction cmp) {
     return res;
 }
 
-// Finds the node with the same key as lelm
-template <typename Node, typename CompareFunction>
-Node* RB_FIND_LIGHT(RBHead<Node>* head, const void* lelm, CompareFunction lcmp) {
-    Node* tmp = head->Root();
+template <typename T, typename U, typename Compare>
+requires HasRBEntry<T>
+constexpr T* RB_FIND_KEY(RBHead<T>& head, const U& key, Compare cmp) {
+    T* tmp = head.Root();
 
     while (tmp) {
-        const int comp = lcmp(lelm, tmp);
+        const int comp = cmp(key, tmp);
         if (comp < 0) {
             tmp = RB_LEFT(tmp);
         } else if (comp > 0) {
@@ -595,14 +614,14 @@ Node* RB_FIND_LIGHT(RBHead<Node>* head, const void* lelm, CompareFunction lcmp)
     return nullptr;
 }
 
-// Finds the first node greater than or equal to the search key
-template <typename Node, typename CompareFunction>
-Node* RB_NFIND_LIGHT(RBHead<Node>* head, const void* lelm, CompareFunction lcmp) {
-    Node* tmp = head->Root();
-    Node* res = nullptr;
+template <typename T, typename U, typename Compare>
+requires HasRBEntry<T>
+constexpr T* RB_NFIND_KEY(RBHead<T>& head, const U& key, Compare cmp) {
+    T* tmp = head.Root();
+    T* res = nullptr;
 
     while (tmp) {
-        const int comp = lcmp(lelm, tmp);
+        const int comp = cmp(key, tmp);
         if (comp < 0) {
             res = tmp;
             tmp = RB_LEFT(tmp);
@@ -616,8 +635,43 @@ Node* RB_NFIND_LIGHT(RBHead<Node>* head, const void* lelm, CompareFunction lcmp)
     return res;
 }
 
-template <typename Node>
-Node* RB_NEXT(Node* elm) {
+template <typename T, typename Compare>
+requires HasRBEntry<T>
+constexpr T* RB_FIND_EXISTING(RBHead<T>& head, T* elm, Compare cmp) {
+    T* tmp = head.Root();
+
+    while (true) {
+        const int comp = cmp(elm, tmp);
+        if (comp < 0) {
+            tmp = RB_LEFT(tmp);
+        } else if (comp > 0) {
+            tmp = RB_RIGHT(tmp);
+        } else {
+            return tmp;
+        }
+    }
+}
+
+template <typename T, typename U, typename Compare>
+requires HasRBEntry<T>
+constexpr T* RB_FIND_EXISTING_KEY(RBHead<T>& head, const U& key, Compare cmp) {
+    T* tmp = head.Root();
+
+    while (true) {
+        const int comp = cmp(key, tmp);
+        if (comp < 0) {
+            tmp = RB_LEFT(tmp);
+        } else if (comp > 0) {
+            tmp = RB_RIGHT(tmp);
+        } else {
+            return tmp;
+        }
+    }
+}
+
+template <typename T>
+requires HasRBEntry<T>
+constexpr T* RB_NEXT(T* elm) {
     if (RB_RIGHT(elm)) {
         elm = RB_RIGHT(elm);
         while (RB_LEFT(elm)) {
@@ -636,8 +690,9 @@ Node* RB_NEXT(Node* elm) {
     return elm;
 }
 
-template <typename Node>
-Node* RB_PREV(Node* elm) {
+template <typename T>
+requires HasRBEntry<T>
+constexpr T* RB_PREV(T* elm) {
     if (RB_LEFT(elm)) {
         elm = RB_LEFT(elm);
         while (RB_RIGHT(elm)) {
@@ -656,30 +711,32 @@ Node* RB_PREV(Node* elm) {
     return elm;
 }
 
-template <typename Node>
-Node* RB_MINMAX(RBHead<Node>* head, bool is_min) {
-    Node* tmp = head->Root();
-    Node* parent = nullptr;
+template <typename T>
+requires HasRBEntry<T>
+constexpr T* RB_MIN(RBHead<T>& head) {
+    T* tmp = head.Root();
+    T* parent = nullptr;
 
     while (tmp) {
         parent = tmp;
-        if (is_min) {
-            tmp = RB_LEFT(tmp);
-        } else {
-            tmp = RB_RIGHT(tmp);
-        }
+        tmp = RB_LEFT(tmp);
     }
 
     return parent;
 }
 
-template <typename Node>
-Node* RB_MIN(RBHead<Node>* head) {
-    return RB_MINMAX(head, true);
-}
+template <typename T>
+requires HasRBEntry<T>
+constexpr T* RB_MAX(RBHead<T>& head) {
+    T* tmp = head.Root();
+    T* parent = nullptr;
 
-template <typename Node>
-Node* RB_MAX(RBHead<Node>* head) {
-    return RB_MINMAX(head, false);
+    while (tmp) {
+        parent = tmp;
+        tmp = RB_RIGHT(tmp);
+    }
+
+    return parent;
 }
-} // namespace Common
+
+} // namespace Common::freebsd
diff --git a/src/common/uuid.cpp b/src/common/uuid.cpp
index d7435a6e9..2b6a530e3 100644
--- a/src/common/uuid.cpp
+++ b/src/common/uuid.cpp
@@ -1,23 +1,25 @@
-// Copyright 2018 yuzu Emulator Project
+// Copyright 2022 yuzu Emulator Project
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
+#include <bit>
+#include <optional>
 #include <random>
 
 #include <fmt/format.h>
 
 #include "common/assert.h"
+#include "common/tiny_mt.h"
 #include "common/uuid.h"
 
 namespace Common {
 
 namespace {
 
-bool IsHexDigit(char c) {
-    return (c >= '0' && c <= '9') || (c >= 'a' && c <= 'f') || (c >= 'A' && c <= 'F');
-}
+constexpr size_t RawStringSize = sizeof(UUID) * 2;
+constexpr size_t FormattedStringSize = RawStringSize + 4;
 
-u8 HexCharToByte(char c) {
+std::optional<u8> HexCharToByte(char c) {
     if (c >= '0' && c <= '9') {
         return static_cast<u8>(c - '0');
     }
@@ -28,60 +30,184 @@ u8 HexCharToByte(char c) {
         return static_cast<u8>(c - 'A' + 10);
     }
     ASSERT_MSG(false, "{} is not a hexadecimal digit!", c);
-    return u8{0};
+    return std::nullopt;
 }
 
-} // Anonymous namespace
+std::array<u8, 0x10> ConstructFromRawString(std::string_view raw_string) {
+    std::array<u8, 0x10> uuid;
+
+    for (size_t i = 0; i < RawStringSize; i += 2) {
+        const auto upper = HexCharToByte(raw_string[i]);
+        const auto lower = HexCharToByte(raw_string[i + 1]);
+        if (!upper || !lower) {
+            return {};
+        }
+        uuid[i / 2] = static_cast<u8>((*upper << 4) | *lower);
+    }
+
+    return uuid;
+}
 
-u128 HexStringToU128(std::string_view hex_string) {
-    const size_t length = hex_string.length();
+std::array<u8, 0x10> ConstructFromFormattedString(std::string_view formatted_string) {
+    std::array<u8, 0x10> uuid;
 
-    // Detect "0x" prefix.
-    const bool has_0x_prefix = length > 2 && hex_string[0] == '0' && hex_string[1] == 'x';
-    const size_t offset = has_0x_prefix ? 2 : 0;
+    size_t i = 0;
 
-    // Check length.
-    if (length > 32 + offset) {
-        ASSERT_MSG(false, "hex_string has more than 32 hexadecimal characters!");
-        return INVALID_UUID;
+    // Process the first 8 characters.
+    const auto* str = formatted_string.data();
+
+    for (; i < 4; ++i) {
+        const auto upper = HexCharToByte(*(str++));
+        const auto lower = HexCharToByte(*(str++));
+        if (!upper || !lower) {
+            return {};
+        }
+        uuid[i] = static_cast<u8>((*upper << 4) | *lower);
+    }
+
+    // Process the next 4 characters.
+    ++str;
+
+    for (; i < 6; ++i) {
+        const auto upper = HexCharToByte(*(str++));
+        const auto lower = HexCharToByte(*(str++));
+        if (!upper || !lower) {
+            return {};
+        }
+        uuid[i] = static_cast<u8>((*upper << 4) | *lower);
     }
 
-    u64 lo = 0;
-    u64 hi = 0;
-    for (size_t i = 0; i < length - offset; ++i) {
-        const char c = hex_string[length - 1 - i];
-        if (!IsHexDigit(c)) {
-            ASSERT_MSG(false, "{} is not a hexadecimal digit!", c);
-            return INVALID_UUID;
+    // Process the next 4 characters.
+    ++str;
+
+    for (; i < 8; ++i) {
+        const auto upper = HexCharToByte(*(str++));
+        const auto lower = HexCharToByte(*(str++));
+        if (!upper || !lower) {
+            return {};
         }
-        if (i < 16) {
-            lo |= u64{HexCharToByte(c)} << (i * 4);
+        uuid[i] = static_cast<u8>((*upper << 4) | *lower);
+    }
+
+    // Process the next 4 characters.
+    ++str;
+
+    for (; i < 10; ++i) {
+        const auto upper = HexCharToByte(*(str++));
+        const auto lower = HexCharToByte(*(str++));
+        if (!upper || !lower) {
+            return {};
         }
-        if (i >= 16) {
-            hi |= u64{HexCharToByte(c)} << ((i - 16) * 4);
+        uuid[i] = static_cast<u8>((*upper << 4) | *lower);
+    }
+
+    // Process the last 12 characters.
+    ++str;
+
+    for (; i < 16; ++i) {
+        const auto upper = HexCharToByte(*(str++));
+        const auto lower = HexCharToByte(*(str++));
+        if (!upper || !lower) {
+            return {};
         }
+        uuid[i] = static_cast<u8>((*upper << 4) | *lower);
+    }
+
+    return uuid;
+}
+
+std::array<u8, 0x10> ConstructUUID(std::string_view uuid_string) {
+    const auto length = uuid_string.length();
+
+    if (length == 0) {
+        return {};
+    }
+
+    // Check if the input string contains 32 hexadecimal characters.
+    if (length == RawStringSize) {
+        return ConstructFromRawString(uuid_string);
+    }
+
+    // Check if the input string has the length of a RFC 4122 formatted UUID string.
+    if (length == FormattedStringSize) {
+        return ConstructFromFormattedString(uuid_string);
     }
-    return u128{lo, hi};
+
+    ASSERT_MSG(false, "UUID string has an invalid length of {} characters!", length);
+
+    return {};
+}
+
+} // Anonymous namespace
+
+UUID::UUID(std::string_view uuid_string) : uuid{ConstructUUID(uuid_string)} {}
+
+std::string UUID::RawString() const {
+    return fmt::format("{:02x}{:02x}{:02x}{:02x}{:02x}{:02x}{:02x}{:02x}"
+                       "{:02x}{:02x}{:02x}{:02x}{:02x}{:02x}{:02x}{:02x}",
+                       uuid[0], uuid[1], uuid[2], uuid[3], uuid[4], uuid[5], uuid[6], uuid[7],
+                       uuid[8], uuid[9], uuid[10], uuid[11], uuid[12], uuid[13], uuid[14],
+                       uuid[15]);
+}
+
+std::string UUID::FormattedString() const {
+    return fmt::format("{:02x}{:02x}{:02x}{:02x}"
+                       "-{:02x}{:02x}-{:02x}{:02x}-{:02x}{:02x}-"
+                       "{:02x}{:02x}{:02x}{:02x}{:02x}{:02x}",
+                       uuid[0], uuid[1], uuid[2], uuid[3], uuid[4], uuid[5], uuid[6], uuid[7],
+                       uuid[8], uuid[9], uuid[10], uuid[11], uuid[12], uuid[13], uuid[14],
+                       uuid[15]);
+}
+
+size_t UUID::Hash() const noexcept {
+    u64 upper_hash;
+    u64 lower_hash;
+
+    std::memcpy(&upper_hash, uuid.data(), sizeof(u64));
+    std::memcpy(&lower_hash, uuid.data() + sizeof(u64), sizeof(u64));
+
+    return upper_hash ^ std::rotl(lower_hash, 1);
 }
 
-UUID UUID::Generate() {
+u128 UUID::AsU128() const {
+    u128 uuid_old;
+    std::memcpy(&uuid_old, uuid.data(), sizeof(UUID));
+    return uuid_old;
+}
+
+UUID UUID::MakeRandom() {
     std::random_device device;
-    std::mt19937 gen(device());
-    std::uniform_int_distribution<u64> distribution(1, std::numeric_limits<u64>::max());
-    return UUID{distribution(gen), distribution(gen)};
+
+    return MakeRandomWithSeed(device());
 }
 
-std::string UUID::Format() const {
-    return fmt::format("{:016x}{:016x}", uuid[1], uuid[0]);
+UUID UUID::MakeRandomWithSeed(u32 seed) {
+    // Create and initialize our RNG.
+    TinyMT rng;
+    rng.Initialize(seed);
+
+    UUID uuid;
+
+    // Populate the UUID with random bytes.
+    rng.GenerateRandomBytes(uuid.uuid.data(), sizeof(UUID));
+
+    return uuid;
 }
 
-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]);
+UUID UUID::MakeRandomRFC4122V4() {
+    auto uuid = MakeRandom();
+
+    // According to Proposed Standard RFC 4122 Section 4.4, we must:
+
+    // 1. Set the two most significant bits (bits 6 and 7) of the
+    //    clock_seq_hi_and_reserved to zero and one, respectively.
+    uuid.uuid[8] = 0x80 | (uuid.uuid[8] & 0x3F);
+
+    // 2. Set the four most significant bits (bits 12 through 15) of the
+    //    time_hi_and_version field to the 4-bit version number from Section 4.1.3.
+    uuid.uuid[6] = 0x40 | (uuid.uuid[6] & 0xF);
+
+    return uuid;
 }
 
 } // namespace Common
diff --git a/src/common/uuid.h b/src/common/uuid.h
index 8ea01f8da..fe31e64e6 100644
--- a/src/common/uuid.h
+++ b/src/common/uuid.h
@@ -1,9 +1,11 @@
-// Copyright 2018 yuzu Emulator Project
+// Copyright 2022 yuzu Emulator Project
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
 #pragma once
 
+#include <array>
+#include <functional>
 #include <string>
 #include <string_view>
 
@@ -11,69 +13,119 @@
 
 namespace Common {
 
-constexpr u128 INVALID_UUID{{0, 0}};
-
-/**
- * Converts a hex string to a 128-bit unsigned integer.
- *
- * The hex string can be formatted in lowercase or uppercase, with or without the "0x" prefix.
- *
- * This function will assert and return INVALID_UUID under the following conditions:
- * - If the hex string is more than 32 characters long
- * - If the hex string contains non-hexadecimal characters
- *
- * @param hex_string Hexadecimal string
- *
- * @returns A 128-bit unsigned integer if successfully converted, INVALID_UUID otherwise.
- */
-[[nodiscard]] u128 HexStringToU128(std::string_view hex_string);
-
 struct UUID {
-    // UUIDs which are 0 are considered invalid!
-    u128 uuid;
-    UUID() = default;
-    constexpr explicit UUID(const u128& id) : uuid{id} {}
-    constexpr explicit UUID(const u64 lo, const u64 hi) : uuid{{lo, hi}} {}
-    explicit UUID(std::string_view hex_string) {
-        uuid = HexStringToU128(hex_string);
-    }
-
-    [[nodiscard]] constexpr explicit operator bool() const {
-        return uuid != INVALID_UUID;
-    }
-
-    [[nodiscard]] constexpr bool operator==(const UUID& rhs) const {
-        return uuid == rhs.uuid;
-    }
-
-    [[nodiscard]] constexpr bool operator!=(const UUID& rhs) const {
-        return !operator==(rhs);
-    }
-
-    // TODO(ogniK): Properly generate uuids based on RFC-4122
-    [[nodiscard]] static UUID Generate();
-
-    // Set the UUID to {0,0} to be considered an invalid user
-    constexpr void Invalidate() {
-        uuid = INVALID_UUID;
+    std::array<u8, 0x10> uuid{};
+
+    /// Constructs an invalid UUID.
+    constexpr UUID() = default;
+
+    /// Constructs a UUID from a reference to a 128 bit array.
+    constexpr explicit UUID(const std::array<u8, 16>& uuid_) : uuid{uuid_} {}
+
+    /**
+     * Constructs a UUID from either:
+     * 1. A 32 hexadecimal character string representing the bytes of the UUID
+     * 2. A RFC 4122 formatted UUID string, in the format xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx
+     *
+     * The input string may contain uppercase or lowercase characters, but they must:
+     * 1. Contain valid hexadecimal characters (0-9, a-f, A-F)
+     * 2. Not contain the "0x" hexadecimal prefix
+     *
+     * Should the input string not meet the above requirements,
+     * an assert will be triggered and an invalid UUID is set instead.
+     */
+    explicit UUID(std::string_view uuid_string);
+
+    ~UUID() = default;
+
+    constexpr UUID(const UUID&) noexcept = default;
+    constexpr UUID(UUID&&) noexcept = default;
+
+    constexpr UUID& operator=(const UUID&) noexcept = default;
+    constexpr UUID& operator=(UUID&&) noexcept = default;
+
+    /**
+     * Returns whether the stored UUID is valid or not.
+     *
+     * @returns True if the stored UUID is valid, false otherwise.
+     */
+    constexpr bool IsValid() const {
+        return uuid != std::array<u8, 0x10>{};
     }
 
-    [[nodiscard]] constexpr bool IsInvalid() const {
-        return uuid == INVALID_UUID;
-    }
-    [[nodiscard]] constexpr bool IsValid() const {
-        return !IsInvalid();
+    /**
+     * Returns whether the stored UUID is invalid or not.
+     *
+     * @returns True if the stored UUID is invalid, false otherwise.
+     */
+    constexpr bool IsInvalid() const {
+        return !IsValid();
     }
 
-    // TODO(ogniK): Properly generate a Nintendo ID
-    [[nodiscard]] constexpr u64 GetNintendoID() const {
-        return uuid[0];
+    /**
+     * Returns a 32 hexadecimal character string representing the bytes of the UUID.
+     *
+     * @returns A 32 hexadecimal character string of the UUID.
+     */
+    std::string RawString() const;
+
+    /**
+     * Returns a RFC 4122 formatted UUID string in the format xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx.
+     *
+     * @returns A RFC 4122 formatted UUID string.
+     */
+    std::string FormattedString() const;
+
+    /**
+     * Returns a 64-bit hash of the UUID for use in hash table data structures.
+     *
+     * @returns A 64-bit hash of the UUID.
+     */
+    size_t Hash() const noexcept;
+
+    /// DO NOT USE. Copies the contents of the UUID into a u128.
+    u128 AsU128() const;
+
+    /**
+     * Creates a default UUID "yuzu Default UID".
+     *
+     * @returns A UUID with its bytes set to the ASCII values of "yuzu Default UID".
+     */
+    static constexpr UUID MakeDefault() {
+        return UUID{
+            {'y', 'u', 'z', 'u', ' ', 'D', 'e', 'f', 'a', 'u', 'l', 't', ' ', 'U', 'I', 'D'},
+        };
     }
 
-    [[nodiscard]] std::string Format() const;
-    [[nodiscard]] std::string FormatSwitch() const;
+    /**
+     * Creates a random UUID.
+     *
+     * @returns A random UUID.
+     */
+    static UUID MakeRandom();
+
+    /**
+     * Creates a random UUID with a seed.
+     *
+     * @param seed A seed to initialize the Mersenne-Twister RNG
+     *
+     * @returns A random UUID.
+     */
+    static UUID MakeRandomWithSeed(u32 seed);
+
+    /**
+     * Creates a random UUID. The generated UUID is RFC 4122 Version 4 compliant.
+     *
+     * @returns A random UUID that is RFC 4122 Version 4 compliant.
+     */
+    static UUID MakeRandomRFC4122V4();
+
+    friend constexpr bool operator==(const UUID& lhs, const UUID& rhs) = default;
 };
-static_assert(sizeof(UUID) == 16, "UUID is an invalid size!");
+static_assert(sizeof(UUID) == 0x10, "UUID has incorrect size.");
+
+/// An invalid UUID. This UUID has all its bytes set to 0.
+constexpr UUID InvalidUUID = {};
 
 } // namespace Common
 
@@ -82,7 +134,7 @@ namespace std {
 template <>
 struct hash<Common::UUID> {
     size_t operator()(const Common::UUID& uuid) const noexcept {
-        return uuid.uuid[1] ^ uuid.uuid[0];
+        return uuid.Hash();
     }
 };
 
diff --git a/src/common/x64/cpu_detect.cpp b/src/common/x64/cpu_detect.cpp
index fbeacc7e2..d81edb140 100644
--- a/src/common/x64/cpu_detect.cpp
+++ b/src/common/x64/cpu_detect.cpp
@@ -1,8 +1,12 @@
-// Copyright 2013 Dolphin Emulator Project / 2015 Citra Emulator Project
-// Licensed under GPLv2 or any later version
-// Refer to the license.txt file included.
+// Copyright 2013 Dolphin Emulator Project / 2015 Citra Emulator Project / 2022 Yuzu Emulator
+// Project Licensed under GPLv2 or any later version Refer to the license.txt file included.
 
+#include <array>
 #include <cstring>
+#include <iterator>
+#include <span>
+#include <string_view>
+#include "common/bit_util.h"
 #include "common/common_types.h"
 #include "common/x64/cpu_detect.h"
 
@@ -17,7 +21,7 @@
 // clang-format on
 #endif
 
-static inline void __cpuidex(int info[4], int function_id, int subfunction_id) {
+static inline void __cpuidex(int info[4], u32 function_id, u32 subfunction_id) {
 #if defined(__DragonFly__) || defined(__FreeBSD__)
     // Despite the name, this is just do_cpuid() with ECX as second input.
     cpuid_count((u_int)function_id, (u_int)subfunction_id, (u_int*)info);
@@ -30,7 +34,7 @@ static inline void __cpuidex(int info[4], int function_id, int subfunction_id) {
 #endif
 }
 
-static inline void __cpuid(int info[4], int function_id) {
+static inline void __cpuid(int info[4], u32 function_id) {
     return __cpuidex(info, function_id, 0);
 }
 
@@ -45,6 +49,17 @@ static inline u64 _xgetbv(u32 index) {
 
 namespace Common {
 
+CPUCaps::Manufacturer CPUCaps::ParseManufacturer(std::string_view brand_string) {
+    if (brand_string == "GenuineIntel") {
+        return Manufacturer::Intel;
+    } else if (brand_string == "AuthenticAMD") {
+        return Manufacturer::AMD;
+    } else if (brand_string == "HygonGenuine") {
+        return Manufacturer::Hygon;
+    }
+    return Manufacturer::Unknown;
+}
+
 // Detects the various CPU features
 static CPUCaps Detect() {
     CPUCaps caps = {};
@@ -53,75 +68,74 @@ static CPUCaps Detect() {
     // yuzu at all anyway
 
     int cpu_id[4];
-    memset(caps.brand_string, 0, sizeof(caps.brand_string));
 
-    // Detect CPU's CPUID capabilities and grab CPU string
+    // Detect CPU's CPUID capabilities and grab manufacturer string
     __cpuid(cpu_id, 0x00000000);
-    u32 max_std_fn = cpu_id[0]; // EAX
-
-    std::memcpy(&caps.brand_string[0], &cpu_id[1], sizeof(int));
-    std::memcpy(&caps.brand_string[4], &cpu_id[3], sizeof(int));
-    std::memcpy(&caps.brand_string[8], &cpu_id[2], sizeof(int));
-    if (cpu_id[1] == 0x756e6547 && cpu_id[2] == 0x6c65746e && cpu_id[3] == 0x49656e69)
-        caps.manufacturer = Manufacturer::Intel;
-    else if (cpu_id[1] == 0x68747541 && cpu_id[2] == 0x444d4163 && cpu_id[3] == 0x69746e65)
-        caps.manufacturer = Manufacturer::AMD;
-    else if (cpu_id[1] == 0x6f677948 && cpu_id[2] == 0x656e6975 && cpu_id[3] == 0x6e65476e)
-        caps.manufacturer = Manufacturer::Hygon;
-    else
-        caps.manufacturer = Manufacturer::Unknown;
+    const u32 max_std_fn = cpu_id[0]; // EAX
 
-    __cpuid(cpu_id, 0x80000000);
+    std::memset(caps.brand_string, 0, std::size(caps.brand_string));
+    std::memcpy(&caps.brand_string[0], &cpu_id[1], sizeof(u32));
+    std::memcpy(&caps.brand_string[4], &cpu_id[3], sizeof(u32));
+    std::memcpy(&caps.brand_string[8], &cpu_id[2], sizeof(u32));
+
+    caps.manufacturer = CPUCaps::ParseManufacturer(caps.brand_string);
 
-    u32 max_ex_fn = cpu_id[0];
+    // Set reasonable default cpu string even if brand string not available
+    std::strncpy(caps.cpu_string, caps.brand_string, std::size(caps.brand_string));
 
-    // Set reasonable default brand string even if brand string not available
-    strcpy(caps.cpu_string, caps.brand_string);
+    __cpuid(cpu_id, 0x80000000);
+
+    const u32 max_ex_fn = cpu_id[0];
 
     // Detect family and other miscellaneous features
     if (max_std_fn >= 1) {
         __cpuid(cpu_id, 0x00000001);
-        if ((cpu_id[3] >> 25) & 1)
-            caps.sse = true;
-        if ((cpu_id[3] >> 26) & 1)
-            caps.sse2 = true;
-        if ((cpu_id[2]) & 1)
-            caps.sse3 = true;
-        if ((cpu_id[2] >> 9) & 1)
-            caps.ssse3 = true;
-        if ((cpu_id[2] >> 19) & 1)
-            caps.sse4_1 = true;
-        if ((cpu_id[2] >> 20) & 1)
-            caps.sse4_2 = true;
-        if ((cpu_id[2] >> 25) & 1)
-            caps.aes = true;
+        caps.sse = Common::Bit<25>(cpu_id[3]);
+        caps.sse2 = Common::Bit<26>(cpu_id[3]);
+        caps.sse3 = Common::Bit<0>(cpu_id[2]);
+        caps.pclmulqdq = Common::Bit<1>(cpu_id[2]);
+        caps.ssse3 = Common::Bit<9>(cpu_id[2]);
+        caps.sse4_1 = Common::Bit<19>(cpu_id[2]);
+        caps.sse4_2 = Common::Bit<20>(cpu_id[2]);
+        caps.movbe = Common::Bit<22>(cpu_id[2]);
+        caps.popcnt = Common::Bit<23>(cpu_id[2]);
+        caps.aes = Common::Bit<25>(cpu_id[2]);
+        caps.f16c = Common::Bit<29>(cpu_id[2]);
 
         // AVX support requires 3 separate checks:
         //  - Is the AVX bit set in CPUID?
         //  - Is the XSAVE bit set in CPUID?
         //  - XGETBV result has the XCR bit set.
-        if (((cpu_id[2] >> 28) & 1) && ((cpu_id[2] >> 27) & 1)) {
+        if (Common::Bit<28>(cpu_id[2]) && Common::Bit<27>(cpu_id[2])) {
             if ((_xgetbv(_XCR_XFEATURE_ENABLED_MASK) & 0x6) == 0x6) {
                 caps.avx = true;
-                if ((cpu_id[2] >> 12) & 1)
+                if (Common::Bit<12>(cpu_id[2]))
                     caps.fma = true;
             }
         }
 
         if (max_std_fn >= 7) {
             __cpuidex(cpu_id, 0x00000007, 0x00000000);
-            // Can't enable AVX2 unless the XSAVE/XGETBV checks above passed
-            if ((cpu_id[1] >> 5) & 1)
-                caps.avx2 = caps.avx;
-            if ((cpu_id[1] >> 3) & 1)
-                caps.bmi1 = true;
-            if ((cpu_id[1] >> 8) & 1)
-                caps.bmi2 = true;
-            // Checks for AVX512F, AVX512CD, AVX512VL, AVX512DQ, AVX512BW (Intel Skylake-X/SP)
-            if ((cpu_id[1] >> 16) & 1 && (cpu_id[1] >> 28) & 1 && (cpu_id[1] >> 31) & 1 &&
-                (cpu_id[1] >> 17) & 1 && (cpu_id[1] >> 30) & 1) {
-                caps.avx512 = caps.avx2;
+            // Can't enable AVX{2,512} unless the XSAVE/XGETBV checks above passed
+            if (caps.avx) {
+                caps.avx2 = Common::Bit<5>(cpu_id[1]);
+                caps.avx512f = Common::Bit<16>(cpu_id[1]);
+                caps.avx512dq = Common::Bit<17>(cpu_id[1]);
+                caps.avx512cd = Common::Bit<28>(cpu_id[1]);
+                caps.avx512bw = Common::Bit<30>(cpu_id[1]);
+                caps.avx512vl = Common::Bit<31>(cpu_id[1]);
+                caps.avx512vbmi = Common::Bit<1>(cpu_id[2]);
+                caps.avx512bitalg = Common::Bit<12>(cpu_id[2]);
             }
+
+            caps.bmi1 = Common::Bit<3>(cpu_id[1]);
+            caps.bmi2 = Common::Bit<8>(cpu_id[1]);
+            caps.sha = Common::Bit<29>(cpu_id[1]);
+
+            caps.gfni = Common::Bit<8>(cpu_id[2]);
+
+            __cpuidex(cpu_id, 0x00000007, 0x00000001);
+            caps.avx_vnni = caps.avx && Common::Bit<4>(cpu_id[0]);
         }
     }
 
@@ -138,15 +152,13 @@ static CPUCaps Detect() {
     if (max_ex_fn >= 0x80000001) {
         // Check for more features
         __cpuid(cpu_id, 0x80000001);
-        if ((cpu_id[2] >> 16) & 1)
-            caps.fma4 = true;
+        caps.lzcnt = Common::Bit<5>(cpu_id[2]);
+        caps.fma4 = Common::Bit<16>(cpu_id[2]);
     }
 
     if (max_ex_fn >= 0x80000007) {
         __cpuid(cpu_id, 0x80000007);
-        if (cpu_id[3] & (1 << 8)) {
-            caps.invariant_tsc = true;
-        }
+        caps.invariant_tsc = Common::Bit<8>(cpu_id[3]);
     }
 
     if (max_std_fn >= 0x16) {
diff --git a/src/common/x64/cpu_detect.h b/src/common/x64/cpu_detect.h
index e3b63302e..40c48b132 100644
--- a/src/common/x64/cpu_detect.h
+++ b/src/common/x64/cpu_detect.h
@@ -1,42 +1,65 @@
-// Copyright 2013 Dolphin Emulator Project / 2015 Citra Emulator Project
-// Licensed under GPLv2 or any later version
-// Refer to the license.txt file included.
+// Copyright 2013 Dolphin Emulator Project / 2015 Citra Emulator Project / 2022 Yuzu Emulator
+// Project Project Licensed under GPLv2 or any later version Refer to the license.txt file included.
 
 #pragma once
 
-namespace Common {
+#include <string_view>
+#include "common/common_types.h"
 
-enum class Manufacturer : u32 {
-    Intel = 0,
-    AMD = 1,
-    Hygon = 2,
-    Unknown = 3,
-};
+namespace Common {
 
 /// x86/x64 CPU capabilities that may be detected by this module
 struct CPUCaps {
+
+    enum class Manufacturer : u8 {
+        Unknown = 0,
+        Intel = 1,
+        AMD = 2,
+        Hygon = 3,
+    };
+
+    static Manufacturer ParseManufacturer(std::string_view brand_string);
+
     Manufacturer manufacturer;
-    char cpu_string[0x21];
-    char brand_string[0x41];
-    bool sse;
-    bool sse2;
-    bool sse3;
-    bool ssse3;
-    bool sse4_1;
-    bool sse4_2;
-    bool lzcnt;
-    bool avx;
-    bool avx2;
-    bool avx512;
-    bool bmi1;
-    bool bmi2;
-    bool fma;
-    bool fma4;
-    bool aes;
-    bool invariant_tsc;
+    char brand_string[13];
+
+    char cpu_string[48];
+
     u32 base_frequency;
     u32 max_frequency;
     u32 bus_frequency;
+
+    bool sse : 1;
+    bool sse2 : 1;
+    bool sse3 : 1;
+    bool ssse3 : 1;
+    bool sse4_1 : 1;
+    bool sse4_2 : 1;
+
+    bool avx : 1;
+    bool avx_vnni : 1;
+    bool avx2 : 1;
+    bool avx512f : 1;
+    bool avx512dq : 1;
+    bool avx512cd : 1;
+    bool avx512bw : 1;
+    bool avx512vl : 1;
+    bool avx512vbmi : 1;
+    bool avx512bitalg : 1;
+
+    bool aes : 1;
+    bool bmi1 : 1;
+    bool bmi2 : 1;
+    bool f16c : 1;
+    bool fma : 1;
+    bool fma4 : 1;
+    bool gfni : 1;
+    bool invariant_tsc : 1;
+    bool lzcnt : 1;
+    bool movbe : 1;
+    bool pclmulqdq : 1;
+    bool popcnt : 1;
+    bool sha : 1;
 };
 
 /**
diff --git a/src/core/CMakeLists.txt b/src/core/CMakeLists.txt
index 6e8d11919..1d4e92edb 100644
--- a/src/core/CMakeLists.txt
+++ b/src/core/CMakeLists.txt
@@ -122,6 +122,8 @@ add_library(core STATIC
     frontend/applets/error.h
     frontend/applets/general_frontend.cpp
     frontend/applets/general_frontend.h
+    frontend/applets/mii.cpp
+    frontend/applets/mii.h
     frontend/applets/profile_select.cpp
     frontend/applets/profile_select.h
     frontend/applets/software_keyboard.cpp
@@ -152,6 +154,7 @@ add_library(core STATIC
     hle/api_version.h
     hle/ipc.h
     hle/ipc_helpers.h
+    hle/kernel/board/nintendo/nx/k_memory_layout.h
     hle/kernel/board/nintendo/nx/k_system_control.cpp
     hle/kernel/board/nintendo/nx/k_system_control.h
     hle/kernel/board/nintendo/nx/secure_monitor.h
@@ -164,6 +167,7 @@ add_library(core STATIC
     hle/kernel/hle_ipc.h
     hle/kernel/init/init_slab_setup.cpp
     hle/kernel/init/init_slab_setup.h
+    hle/kernel/initial_process.h
     hle/kernel/k_address_arbiter.cpp
     hle/kernel/k_address_arbiter.h
     hle/kernel/k_address_space_info.cpp
@@ -205,6 +209,8 @@ add_library(core STATIC
     hle/kernel/k_memory_region.h
     hle/kernel/k_memory_region_type.h
     hle/kernel/k_page_bitmap.h
+    hle/kernel/k_page_buffer.cpp
+    hle/kernel/k_page_buffer.h
     hle/kernel/k_page_heap.cpp
     hle/kernel/k_page_heap.h
     hle/kernel/k_page_linked_list.h
@@ -242,6 +248,8 @@ add_library(core STATIC
     hle/kernel/k_system_control.h
     hle/kernel/k_thread.cpp
     hle/kernel/k_thread.h
+    hle/kernel/k_thread_local_page.cpp
+    hle/kernel/k_thread_local_page.h
     hle/kernel/k_thread_queue.cpp
     hle/kernel/k_thread_queue.h
     hle/kernel/k_trace.h
@@ -298,6 +306,8 @@ add_library(core STATIC
     hle/service/am/applets/applet_error.h
     hle/service/am/applets/applet_general_backend.cpp
     hle/service/am/applets/applet_general_backend.h
+    hle/service/am/applets/applet_mii.cpp
+    hle/service/am/applets/applet_mii.h
     hle/service/am/applets/applet_profile_select.cpp
     hle/service/am/applets/applet_profile_select.h
     hle/service/am/applets/applet_software_keyboard.cpp
@@ -467,6 +477,8 @@ add_library(core STATIC
     hle/service/mii/types.h
     hle/service/mm/mm_u.cpp
     hle/service/mm/mm_u.h
+    hle/service/mnpp/mnpp_app.cpp
+    hle/service/mnpp/mnpp_app.h
     hle/service/ncm/ncm.cpp
     hle/service/ncm/ncm.h
     hle/service/nfc/nfc.cpp
diff --git a/src/core/arm/arm_interface.h b/src/core/arm/arm_interface.h
index 689e3ceb5..c60322442 100644
--- a/src/core/arm/arm_interface.h
+++ b/src/core/arm/arm_interface.h
@@ -6,6 +6,7 @@
 
 #include <array>
 #include <vector>
+#include "common/common_funcs.h"
 #include "common/common_types.h"
 #include "core/hardware_properties.h"
 
@@ -24,8 +25,11 @@ class CPUInterruptHandler;
 using CPUInterrupts = std::array<CPUInterruptHandler, Core::Hardware::NUM_CPU_CORES>;
 
 /// Generic ARMv8 CPU interface
-class ARM_Interface : NonCopyable {
+class ARM_Interface {
 public:
+    YUZU_NON_COPYABLE(ARM_Interface);
+    YUZU_NON_MOVEABLE(ARM_Interface);
+
     explicit ARM_Interface(System& system_, CPUInterrupts& interrupt_handlers_,
                            bool uses_wall_clock_)
         : system{system_}, interrupt_handlers{interrupt_handlers_}, uses_wall_clock{
diff --git a/src/core/arm/dynarmic/arm_dynarmic_32.cpp b/src/core/arm/dynarmic/arm_dynarmic_32.cpp
index b0d89c539..c1c843b8f 100644
--- a/src/core/arm/dynarmic/arm_dynarmic_32.cpp
+++ b/src/core/arm/dynarmic/arm_dynarmic_32.cpp
@@ -137,6 +137,8 @@ std::shared_ptr<Dynarmic::A32::Jit> ARM_Dynarmic_32::MakeJit(Common::PageTable*
     config.page_table_pointer_mask_bits = Common::PageTable::ATTRIBUTE_BITS;
     config.detect_misaligned_access_via_page_table = 16 | 32 | 64 | 128;
     config.only_detect_misalignment_via_page_table_on_page_boundary = true;
+    config.fastmem_exclusive_access = true;
+    config.recompile_on_exclusive_fastmem_failure = true;
 
     // Multi-process state
     config.processor_id = core_index;
@@ -146,8 +148,8 @@ std::shared_ptr<Dynarmic::A32::Jit> ARM_Dynarmic_32::MakeJit(Common::PageTable*
     config.wall_clock_cntpct = uses_wall_clock;
 
     // Code cache size
-    config.code_cache_size = 512_MiB;
-    config.far_code_offset = 400_MiB;
+    config.code_cache_size = 128_MiB;
+    config.far_code_offset = 100_MiB;
 
     // Safe optimizations
     if (Settings::values.cpu_debug_mode) {
@@ -178,6 +180,12 @@ std::shared_ptr<Dynarmic::A32::Jit> ARM_Dynarmic_32::MakeJit(Common::PageTable*
         if (!Settings::values.cpuopt_fastmem) {
             config.fastmem_pointer = nullptr;
         }
+        if (!Settings::values.cpuopt_fastmem_exclusives) {
+            config.fastmem_exclusive_access = false;
+        }
+        if (!Settings::values.cpuopt_recompile_exclusives) {
+            config.recompile_on_exclusive_fastmem_failure = false;
+        }
     }
 
     // Unsafe optimizations
@@ -195,6 +203,9 @@ std::shared_ptr<Dynarmic::A32::Jit> ARM_Dynarmic_32::MakeJit(Common::PageTable*
         if (Settings::values.cpuopt_unsafe_inaccurate_nan) {
             config.optimizations |= Dynarmic::OptimizationFlag::Unsafe_InaccurateNaN;
         }
+        if (Settings::values.cpuopt_unsafe_ignore_global_monitor) {
+            config.optimizations |= Dynarmic::OptimizationFlag::Unsafe_IgnoreGlobalMonitor;
+        }
     }
 
     // Curated optimizations
@@ -203,6 +214,7 @@ std::shared_ptr<Dynarmic::A32::Jit> ARM_Dynarmic_32::MakeJit(Common::PageTable*
         config.optimizations |= Dynarmic::OptimizationFlag::Unsafe_UnfuseFMA;
         config.optimizations |= Dynarmic::OptimizationFlag::Unsafe_IgnoreStandardFPCRValue;
         config.optimizations |= Dynarmic::OptimizationFlag::Unsafe_InaccurateNaN;
+        config.optimizations |= Dynarmic::OptimizationFlag::Unsafe_IgnoreGlobalMonitor;
     }
 
     return std::make_unique<Dynarmic::A32::Jit>(config);
diff --git a/src/core/arm/dynarmic/arm_dynarmic_64.cpp b/src/core/arm/dynarmic/arm_dynarmic_64.cpp
index 56836bd05..aa74fce4d 100644
--- a/src/core/arm/dynarmic/arm_dynarmic_64.cpp
+++ b/src/core/arm/dynarmic/arm_dynarmic_64.cpp
@@ -185,6 +185,9 @@ std::shared_ptr<Dynarmic::A64::Jit> ARM_Dynarmic_64::MakeJit(Common::PageTable*
         config.fastmem_pointer = page_table->fastmem_arena;
         config.fastmem_address_space_bits = address_space_bits;
         config.silently_mirror_fastmem = false;
+
+        config.fastmem_exclusive_access = true;
+        config.recompile_on_exclusive_fastmem_failure = true;
     }
 
     // Multi-process state
@@ -205,8 +208,8 @@ std::shared_ptr<Dynarmic::A64::Jit> ARM_Dynarmic_64::MakeJit(Common::PageTable*
     config.wall_clock_cntpct = uses_wall_clock;
 
     // Code cache size
-    config.code_cache_size = 512_MiB;
-    config.far_code_offset = 400_MiB;
+    config.code_cache_size = 128_MiB;
+    config.far_code_offset = 100_MiB;
 
     // Safe optimizations
     if (Settings::values.cpu_debug_mode) {
@@ -237,6 +240,12 @@ std::shared_ptr<Dynarmic::A64::Jit> ARM_Dynarmic_64::MakeJit(Common::PageTable*
         if (!Settings::values.cpuopt_fastmem) {
             config.fastmem_pointer = nullptr;
         }
+        if (!Settings::values.cpuopt_fastmem_exclusives) {
+            config.fastmem_exclusive_access = false;
+        }
+        if (!Settings::values.cpuopt_recompile_exclusives) {
+            config.recompile_on_exclusive_fastmem_failure = false;
+        }
     }
 
     // Unsafe optimizations
@@ -254,6 +263,9 @@ std::shared_ptr<Dynarmic::A64::Jit> ARM_Dynarmic_64::MakeJit(Common::PageTable*
         if (Settings::values.cpuopt_unsafe_fastmem_check) {
             config.fastmem_address_space_bits = 64;
         }
+        if (Settings::values.cpuopt_unsafe_ignore_global_monitor) {
+            config.optimizations |= Dynarmic::OptimizationFlag::Unsafe_IgnoreGlobalMonitor;
+        }
     }
 
     // Curated optimizations
@@ -262,6 +274,7 @@ std::shared_ptr<Dynarmic::A64::Jit> ARM_Dynarmic_64::MakeJit(Common::PageTable*
         config.optimizations |= Dynarmic::OptimizationFlag::Unsafe_UnfuseFMA;
         config.optimizations |= Dynarmic::OptimizationFlag::Unsafe_InaccurateNaN;
         config.fastmem_address_space_bits = 64;
+        config.optimizations |= Dynarmic::OptimizationFlag::Unsafe_IgnoreGlobalMonitor;
     }
 
     return std::make_shared<Dynarmic::A64::Jit>(config);
diff --git a/src/core/arm/dynarmic/arm_exclusive_monitor.cpp b/src/core/arm/dynarmic/arm_exclusive_monitor.cpp
index 397d054a8..ea6b224e0 100644
--- a/src/core/arm/dynarmic/arm_exclusive_monitor.cpp
+++ b/src/core/arm/dynarmic/arm_exclusive_monitor.cpp
@@ -37,8 +37,8 @@ u128 DynarmicExclusiveMonitor::ExclusiveRead128(std::size_t core_index, VAddr ad
     });
 }
 
-void DynarmicExclusiveMonitor::ClearExclusive() {
-    monitor.Clear();
+void DynarmicExclusiveMonitor::ClearExclusive(std::size_t core_index) {
+    monitor.ClearProcessor(core_index);
 }
 
 bool DynarmicExclusiveMonitor::ExclusiveWrite8(std::size_t core_index, VAddr vaddr, u8 value) {
diff --git a/src/core/arm/dynarmic/arm_exclusive_monitor.h b/src/core/arm/dynarmic/arm_exclusive_monitor.h
index 265c4ecef..5a15b43ef 100644
--- a/src/core/arm/dynarmic/arm_exclusive_monitor.h
+++ b/src/core/arm/dynarmic/arm_exclusive_monitor.h
@@ -29,7 +29,7 @@ public:
     u32 ExclusiveRead32(std::size_t core_index, VAddr addr) override;
     u64 ExclusiveRead64(std::size_t core_index, VAddr addr) override;
     u128 ExclusiveRead128(std::size_t core_index, VAddr addr) override;
-    void ClearExclusive() override;
+    void ClearExclusive(std::size_t core_index) override;
 
     bool ExclusiveWrite8(std::size_t core_index, VAddr vaddr, u8 value) override;
     bool ExclusiveWrite16(std::size_t core_index, VAddr vaddr, u16 value) override;
diff --git a/src/core/arm/exclusive_monitor.h b/src/core/arm/exclusive_monitor.h
index 62f6e6023..9914ca3da 100644
--- a/src/core/arm/exclusive_monitor.h
+++ b/src/core/arm/exclusive_monitor.h
@@ -23,7 +23,7 @@ public:
     virtual u32 ExclusiveRead32(std::size_t core_index, VAddr addr) = 0;
     virtual u64 ExclusiveRead64(std::size_t core_index, VAddr addr) = 0;
     virtual u128 ExclusiveRead128(std::size_t core_index, VAddr addr) = 0;
-    virtual void ClearExclusive() = 0;
+    virtual void ClearExclusive(std::size_t core_index) = 0;
 
     virtual bool ExclusiveWrite8(std::size_t core_index, VAddr vaddr, u8 value) = 0;
     virtual bool ExclusiveWrite16(std::size_t core_index, VAddr vaddr, u16 value) = 0;
diff --git a/src/core/core.cpp b/src/core/core.cpp
index 3f9a7f44b..c60a784c3 100644
--- a/src/core/core.cpp
+++ b/src/core/core.cpp
@@ -28,7 +28,9 @@
 #include "core/file_sys/vfs_real.h"
 #include "core/hardware_interrupt_manager.h"
 #include "core/hid/hid_core.h"
+#include "core/hle/kernel/k_memory_manager.h"
 #include "core/hle/kernel/k_process.h"
+#include "core/hle/kernel/k_resource_limit.h"
 #include "core/hle/kernel/k_scheduler.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/kernel/physical_core.h"
@@ -252,9 +254,16 @@ struct System::Impl {
         }
 
         telemetry_session->AddInitialInfo(*app_loader, fs_controller, *content_provider);
+
+        // Create a resource limit for the process.
+        const auto physical_memory_size =
+            kernel.MemoryManager().GetSize(Kernel::KMemoryManager::Pool::Application);
+        auto* resource_limit = Kernel::CreateResourceLimitForProcess(system, physical_memory_size);
+
+        // Create the process.
         auto main_process = Kernel::KProcess::Create(system.Kernel());
         ASSERT(Kernel::KProcess::Initialize(main_process, system, "main",
-                                            Kernel::KProcess::ProcessType::Userland)
+                                            Kernel::KProcess::ProcessType::Userland, resource_limit)
                    .IsSuccess());
         const auto [load_result, load_parameters] = app_loader->Load(*main_process, system);
         if (load_result != Loader::ResultStatus::Success) {
@@ -317,7 +326,9 @@ struct System::Impl {
         is_powered_on = false;
         exit_lock = false;
 
-        gpu_core->NotifyShutdown();
+        if (gpu_core != nullptr) {
+            gpu_core->NotifyShutdown();
+        }
 
         services.reset();
         service_manager.reset();
diff --git a/src/core/device_memory.cpp b/src/core/device_memory.cpp
index f19c0515f..e6bc63086 100644
--- a/src/core/device_memory.cpp
+++ b/src/core/device_memory.cpp
@@ -3,10 +3,13 @@
 // Refer to the license.txt file included.
 
 #include "core/device_memory.h"
+#include "hle/kernel/board/nintendo/nx/k_system_control.h"
 
 namespace Core {
 
-DeviceMemory::DeviceMemory() : buffer{DramMemoryMap::Size, 1ULL << 39} {}
+DeviceMemory::DeviceMemory()
+    : buffer{Kernel::Board::Nintendo::Nx::KSystemControl::Init::GetIntendedMemorySize(),
+             1ULL << 39} {}
 DeviceMemory::~DeviceMemory() = default;
 
 } // namespace Core
diff --git a/src/core/device_memory.h b/src/core/device_memory.h
index c4d17705f..daeb551ea 100644
--- a/src/core/device_memory.h
+++ b/src/core/device_memory.h
@@ -12,12 +12,8 @@ namespace Core {
 namespace DramMemoryMap {
 enum : u64 {
     Base = 0x80000000ULL,
-    Size = 0x100000000ULL,
-    End = Base + Size,
     KernelReserveBase = Base + 0x60000,
     SlabHeapBase = KernelReserveBase + 0x85000,
-    SlapHeapSize = 0xa21000,
-    SlabHeapEnd = SlabHeapBase + SlapHeapSize,
 };
 }; // namespace DramMemoryMap
 
diff --git a/src/core/file_sys/patch_manager.cpp b/src/core/file_sys/patch_manager.cpp
index 7c0950bb0..f19ac4607 100644
--- a/src/core/file_sys/patch_manager.cpp
+++ b/src/core/file_sys/patch_manager.cpp
@@ -128,15 +128,6 @@ VirtualDir PatchManager::PatchExeFS(VirtualDir exefs) const {
     if (exefs == nullptr)
         return exefs;
 
-    if (Settings::values.dump_exefs) {
-        LOG_INFO(Loader, "Dumping ExeFS for title_id={:016X}", title_id);
-        const auto dump_dir = fs_controller.GetModificationDumpRoot(title_id);
-        if (dump_dir != nullptr) {
-            const auto exefs_dir = GetOrCreateDirectoryRelative(dump_dir, "/exefs");
-            VfsRawCopyD(exefs, exefs_dir);
-        }
-    }
-
     const auto& disabled = Settings::values.disabled_addons[title_id];
     const auto update_disabled =
         std::find(disabled.cbegin(), disabled.cend(), "Update") != disabled.cend();
@@ -179,6 +170,15 @@ VirtualDir PatchManager::PatchExeFS(VirtualDir exefs) const {
         }
     }
 
+    if (Settings::values.dump_exefs) {
+        LOG_INFO(Loader, "Dumping ExeFS for title_id={:016X}", title_id);
+        const auto dump_dir = fs_controller.GetModificationDumpRoot(title_id);
+        if (dump_dir != nullptr) {
+            const auto exefs_dir = GetOrCreateDirectoryRelative(dump_dir, "/exefs");
+            VfsRawCopyD(exefs, exefs_dir);
+        }
+    }
+
     return exefs;
 }
 
diff --git a/src/core/file_sys/vfs.h b/src/core/file_sys/vfs.h
index 3e625fad6..1b9365853 100644
--- a/src/core/file_sys/vfs.h
+++ b/src/core/file_sys/vfs.h
@@ -12,6 +12,7 @@
 #include <type_traits>
 #include <vector>
 
+#include "common/common_funcs.h"
 #include "common/common_types.h"
 #include "core/file_sys/vfs_types.h"
 
@@ -29,8 +30,11 @@ enum class VfsEntryType {
 // A class representing an abstract filesystem. A default implementation given the root VirtualDir
 // is provided for convenience, but if the Vfs implementation has any additional state or
 // functionality, they will need to override.
-class VfsFilesystem : NonCopyable {
+class VfsFilesystem {
 public:
+    YUZU_NON_COPYABLE(VfsFilesystem);
+    YUZU_NON_MOVEABLE(VfsFilesystem);
+
     explicit VfsFilesystem(VirtualDir root);
     virtual ~VfsFilesystem();
 
@@ -77,8 +81,12 @@ protected:
 };
 
 // A class representing a file in an abstract filesystem.
-class VfsFile : NonCopyable {
+class VfsFile {
 public:
+    YUZU_NON_COPYABLE(VfsFile);
+    YUZU_NON_MOVEABLE(VfsFile);
+
+    VfsFile() = default;
     virtual ~VfsFile();
 
     // Retrieves the file name.
@@ -176,8 +184,12 @@ public:
 };
 
 // A class representing a directory in an abstract filesystem.
-class VfsDirectory : NonCopyable {
+class VfsDirectory {
 public:
+    YUZU_NON_COPYABLE(VfsDirectory);
+    YUZU_NON_MOVEABLE(VfsDirectory);
+
+    VfsDirectory() = default;
     virtual ~VfsDirectory();
 
     // Retrives the file located at path as if the current directory was root. Returns nullptr if
diff --git a/src/core/frontend/applets/mii.cpp b/src/core/frontend/applets/mii.cpp
new file mode 100644
index 000000000..1c05ff412
--- /dev/null
+++ b/src/core/frontend/applets/mii.cpp
@@ -0,0 +1,19 @@
+// Copyright 2022 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include "common/logging/log.h"
+#include "core/frontend/applets/mii.h"
+
+namespace Core::Frontend {
+
+MiiApplet::~MiiApplet() = default;
+
+void DefaultMiiApplet::ShowMii(
+    const MiiParameters& parameters,
+    const std::function<void(const Core::Frontend::MiiParameters& parameters)> callback) const {
+    LOG_INFO(Service_HID, "(STUBBED) called");
+    callback(parameters);
+}
+
+} // namespace Core::Frontend
diff --git a/src/core/frontend/applets/mii.h b/src/core/frontend/applets/mii.h
new file mode 100644
index 000000000..1fc40a9c6
--- /dev/null
+++ b/src/core/frontend/applets/mii.h
@@ -0,0 +1,35 @@
+// Copyright 2022 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include <functional>
+
+#include "core/hle/result.h"
+#include "core/hle/service/mii/mii_manager.h"
+
+namespace Core::Frontend {
+
+struct MiiParameters {
+    bool is_editable;
+    Service::Mii::MiiInfo mii_data{};
+};
+
+class MiiApplet {
+public:
+    virtual ~MiiApplet();
+
+    virtual void ShowMii(const MiiParameters& parameters,
+                         const std::function<void(const Core::Frontend::MiiParameters& parameters)>
+                             callback) const = 0;
+};
+
+class DefaultMiiApplet final : public MiiApplet {
+public:
+    void ShowMii(const MiiParameters& parameters,
+                 const std::function<void(const Core::Frontend::MiiParameters& parameters)>
+                     callback) const override;
+};
+
+} // namespace Core::Frontend
diff --git a/src/core/frontend/applets/profile_select.cpp b/src/core/frontend/applets/profile_select.cpp
index 3e4f90be2..4c58c310f 100644
--- a/src/core/frontend/applets/profile_select.cpp
+++ b/src/core/frontend/applets/profile_select.cpp
@@ -13,8 +13,7 @@ ProfileSelectApplet::~ProfileSelectApplet() = default;
 void DefaultProfileSelectApplet::SelectProfile(
     std::function<void(std::optional<Common::UUID>)> callback) const {
     Service::Account::ProfileManager manager;
-    callback(manager.GetUser(Settings::values.current_user.GetValue())
-                 .value_or(Common::UUID{Common::INVALID_UUID}));
+    callback(manager.GetUser(Settings::values.current_user.GetValue()).value_or(Common::UUID{}));
     LOG_INFO(Service_ACC, "called, selecting current user instead of prompting...");
 }
 
diff --git a/src/core/frontend/emu_window.h b/src/core/frontend/emu_window.h
index e413a520a..b3bffecb2 100644
--- a/src/core/frontend/emu_window.h
+++ b/src/core/frontend/emu_window.h
@@ -42,11 +42,20 @@ public:
             context.MakeCurrent();
         }
         ~Scoped() {
-            context.DoneCurrent();
+            if (active) {
+                context.DoneCurrent();
+            }
+        }
+
+        /// In the event that context was destroyed before the Scoped is destroyed, this provides a
+        /// mechanism to prevent calling a destroyed object's method during the deconstructor
+        void Cancel() {
+            active = false;
         }
 
     private:
         GraphicsContext& context;
+        bool active{true};
     };
 
     /// Calls MakeCurrent on the context and calls DoneCurrent when the scope for the returned value
diff --git a/src/core/hid/emulated_console.h b/src/core/hid/emulated_console.h
index 707419102..5eb170823 100644
--- a/src/core/hid/emulated_console.h
+++ b/src/core/hid/emulated_console.h
@@ -10,6 +10,7 @@
 #include <mutex>
 #include <unordered_map>
 
+#include "common/common_funcs.h"
 #include "common/common_types.h"
 #include "common/input.h"
 #include "common/param_package.h"
diff --git a/src/core/hid/emulated_controller.cpp b/src/core/hid/emulated_controller.cpp
index a7cdf45e6..7e05666d6 100644
--- a/src/core/hid/emulated_controller.cpp
+++ b/src/core/hid/emulated_controller.cpp
@@ -269,7 +269,8 @@ void EmulatedController::ReloadInput() {
     }
 
     // Use a common UUID for TAS
-    const auto tas_uuid = Common::UUID{0x0, 0x7A5};
+    static constexpr Common::UUID TAS_UUID = Common::UUID{
+        {0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x7, 0xA5, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0}};
 
     // Register TAS devices. No need to force update
     for (std::size_t index = 0; index < tas_button_devices.size(); ++index) {
@@ -278,8 +279,8 @@ void EmulatedController::ReloadInput() {
         }
         tas_button_devices[index]->SetCallback({
             .on_change =
-                [this, index, tas_uuid](const Common::Input::CallbackStatus& callback) {
-                    SetButton(callback, index, tas_uuid);
+                [this, index](const Common::Input::CallbackStatus& callback) {
+                    SetButton(callback, index, TAS_UUID);
                 },
         });
     }
@@ -290,8 +291,8 @@ void EmulatedController::ReloadInput() {
         }
         tas_stick_devices[index]->SetCallback({
             .on_change =
-                [this, index, tas_uuid](const Common::Input::CallbackStatus& callback) {
-                    SetStick(callback, index, tas_uuid);
+                [this, index](const Common::Input::CallbackStatus& callback) {
+                    SetStick(callback, index, TAS_UUID);
                 },
         });
     }
@@ -884,6 +885,12 @@ bool EmulatedController::TestVibration(std::size_t device_index) {
     return SetVibration(device_index, DEFAULT_VIBRATION_VALUE);
 }
 
+bool EmulatedController::SetPollingMode(Common::Input::PollingMode polling_mode) {
+    LOG_INFO(Service_HID, "Set polling mode {}", polling_mode);
+    auto& output_device = output_devices[static_cast<std::size_t>(DeviceIndex::Right)];
+    return output_device->SetPollingMode(polling_mode) == Common::Input::PollingError::None;
+}
+
 void EmulatedController::SetLedPattern() {
     for (auto& device : output_devices) {
         if (!device) {
diff --git a/src/core/hid/emulated_controller.h b/src/core/hid/emulated_controller.h
index a63a83cce..aa52f9572 100644
--- a/src/core/hid/emulated_controller.h
+++ b/src/core/hid/emulated_controller.h
@@ -13,8 +13,6 @@
 #include "common/common_types.h"
 #include "common/input.h"
 #include "common/param_package.h"
-#include "common/point.h"
-#include "common/quaternion.h"
 #include "common/settings.h"
 #include "common/vector_math.h"
 #include "core/hid/hid_types.h"
@@ -301,16 +299,23 @@ public:
 
     /**
      * Sends a specific vibration to the output device
-     * @return returns true if vibration had no errors
+     * @return true if vibration had no errors
      */
     bool SetVibration(std::size_t device_index, VibrationValue vibration);
 
     /**
      * Sends a small vibration to the output device
-     * @return returns true if SetVibration was successfull
+     * @return true if SetVibration was successfull
      */
     bool TestVibration(std::size_t device_index);
 
+    /**
+     * Sets the desired data to be polled from a controller
+     * @param polling_mode type of input desired buttons, gyro, nfc, ir, etc.
+     * @return true if SetPollingMode was successfull
+     */
+    bool SetPollingMode(Common::Input::PollingMode polling_mode);
+
     /// Returns the led pattern corresponding to this emulated controller
     LedPattern GetLedPattern() const;
 
diff --git a/src/core/hid/hid_core.h b/src/core/hid/hid_core.h
index 837f7de49..717f605e7 100644
--- a/src/core/hid/hid_core.h
+++ b/src/core/hid/hid_core.h
@@ -6,6 +6,7 @@
 
 #include <memory>
 
+#include "common/common_funcs.h"
 #include "core/hid/hid_types.h"
 
 namespace Core::HID {
diff --git a/src/core/hid/input_converter.cpp b/src/core/hid/input_converter.cpp
index 860aab400..cd41607a7 100644
--- a/src/core/hid/input_converter.cpp
+++ b/src/core/hid/input_converter.cpp
@@ -28,7 +28,7 @@ Common::Input::BatteryStatus TransformToBattery(const Common::Input::CallbackSta
         if (value > 0.8f) {
             battery = Common::Input::BatteryLevel::Full;
         }
-        if (value >= 1.0f) {
+        if (value >= 0.95f) {
             battery = Common::Input::BatteryLevel::Charging;
         }
         break;
diff --git a/src/core/hle/ipc_helpers.h b/src/core/hle/ipc_helpers.h
index cf204f570..3c4e45fcd 100644
--- a/src/core/hle/ipc_helpers.h
+++ b/src/core/hle/ipc_helpers.h
@@ -385,7 +385,7 @@ public:
     T PopRaw();
 
     template <class T>
-    std::shared_ptr<T> PopIpcInterface() {
+    std::weak_ptr<T> PopIpcInterface() {
         ASSERT(context->Session()->IsDomain());
         ASSERT(context->GetDomainMessageHeader().input_object_count > 0);
         return context->GetDomainHandler<T>(Pop<u32>() - 1);
@@ -404,6 +404,11 @@ inline s32 RequestParser::Pop() {
     return static_cast<s32>(Pop<u32>());
 }
 
+// Ignore the -Wclass-memaccess warning on memcpy for non-trivially default constructible objects.
+#if defined(__GNUC__)
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wclass-memaccess"
+#endif
 template <typename T>
 void RequestParser::PopRaw(T& value) {
     static_assert(std::is_trivially_copyable_v<T>,
@@ -411,6 +416,9 @@ void RequestParser::PopRaw(T& value) {
     std::memcpy(&value, cmdbuf + index, sizeof(T));
     index += (sizeof(T) + 3) / 4; // round up to word length
 }
+#if defined(__GNUC__)
+#pragma GCC diagnostic pop
+#endif
 
 template <typename T>
 T RequestParser::PopRaw() {
diff --git a/src/core/hle/kernel/board/nintendo/nx/k_memory_layout.h b/src/core/hle/kernel/board/nintendo/nx/k_memory_layout.h
new file mode 100644
index 000000000..01e225088
--- /dev/null
+++ b/src/core/hle/kernel/board/nintendo/nx/k_memory_layout.h
@@ -0,0 +1,13 @@
+// Copyright 2022 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include "common/common_types.h"
+
+namespace Kernel {
+
+constexpr inline PAddr MainMemoryAddress = 0x80000000;
+
+} // namespace Kernel
diff --git a/src/core/hle/kernel/board/nintendo/nx/k_system_control.cpp b/src/core/hle/kernel/board/nintendo/nx/k_system_control.cpp
index 6f335c251..8027bec00 100644
--- a/src/core/hle/kernel/board/nintendo/nx/k_system_control.cpp
+++ b/src/core/hle/kernel/board/nintendo/nx/k_system_control.cpp
@@ -5,6 +5,7 @@
 #include <random>
 
 #include "common/literals.h"
+#include "common/settings.h"
 
 #include "core/hle/kernel/board/nintendo/nx/k_system_control.h"
 #include "core/hle/kernel/board/nintendo/nx/secure_monitor.h"
@@ -28,33 +29,20 @@ namespace {
 
 using namespace Common::Literals;
 
-u32 GetMemoryModeForInit() {
-    return 0x01;
-}
-
 u32 GetMemorySizeForInit() {
-    return 0;
+    return Settings::values.use_extended_memory_layout ? Smc::MemorySize_6GB : Smc::MemorySize_4GB;
 }
 
 Smc::MemoryArrangement GetMemoryArrangeForInit() {
-    switch (GetMemoryModeForInit() & 0x3F) {
-    case 0x01:
-    default:
-        return Smc::MemoryArrangement_4GB;
-    case 0x02:
-        return Smc::MemoryArrangement_4GBForAppletDev;
-    case 0x03:
-        return Smc::MemoryArrangement_4GBForSystemDev;
-    case 0x11:
-        return Smc::MemoryArrangement_6GB;
-    case 0x12:
-        return Smc::MemoryArrangement_6GBForAppletDev;
-    case 0x21:
-        return Smc::MemoryArrangement_8GB;
-    }
+    return Settings::values.use_extended_memory_layout ? Smc::MemoryArrangement_6GB
+                                                       : Smc::MemoryArrangement_4GB;
 }
 } // namespace
 
+size_t KSystemControl::Init::GetRealMemorySize() {
+    return GetIntendedMemorySize();
+}
+
 // Initialization.
 size_t KSystemControl::Init::GetIntendedMemorySize() {
     switch (GetMemorySizeForInit()) {
@@ -69,7 +57,13 @@ size_t KSystemControl::Init::GetIntendedMemorySize() {
 }
 
 PAddr KSystemControl::Init::GetKernelPhysicalBaseAddress(u64 base_address) {
-    return base_address;
+    const size_t real_dram_size = KSystemControl::Init::GetRealMemorySize();
+    const size_t intended_dram_size = KSystemControl::Init::GetIntendedMemorySize();
+    if (intended_dram_size * 2 < real_dram_size) {
+        return base_address;
+    } else {
+        return base_address + ((real_dram_size - intended_dram_size) / 2);
+    }
 }
 
 bool KSystemControl::Init::ShouldIncreaseThreadResourceLimit() {
diff --git a/src/core/hle/kernel/board/nintendo/nx/k_system_control.h b/src/core/hle/kernel/board/nintendo/nx/k_system_control.h
index 52f230ced..df2a17f2a 100644
--- a/src/core/hle/kernel/board/nintendo/nx/k_system_control.h
+++ b/src/core/hle/kernel/board/nintendo/nx/k_system_control.h
@@ -13,6 +13,7 @@ public:
     class Init {
     public:
         // Initialization.
+        static std::size_t GetRealMemorySize();
         static std::size_t GetIntendedMemorySize();
         static PAddr GetKernelPhysicalBaseAddress(u64 base_address);
         static bool ShouldIncreaseThreadResourceLimit();
diff --git a/src/core/hle/kernel/hle_ipc.cpp b/src/core/hle/kernel/hle_ipc.cpp
index e19544c54..9f2175f82 100644
--- a/src/core/hle/kernel/hle_ipc.cpp
+++ b/src/core/hle/kernel/hle_ipc.cpp
@@ -45,7 +45,7 @@ bool SessionRequestManager::HasSessionRequestHandler(const HLERequestContext& co
             LOG_CRITICAL(IPC, "object_id {} is too big!", object_id);
             return false;
         }
-        return DomainHandler(object_id - 1) != nullptr;
+        return DomainHandler(object_id - 1).lock() != nullptr;
     } else {
         return session_handler != nullptr;
     }
@@ -53,9 +53,6 @@ bool SessionRequestManager::HasSessionRequestHandler(const HLERequestContext& co
 
 void SessionRequestHandler::ClientConnected(KServerSession* session) {
     session->ClientConnected(shared_from_this());
-
-    // Ensure our server session is tracked globally.
-    kernel.RegisterServerSession(session);
 }
 
 void SessionRequestHandler::ClientDisconnected(KServerSession* session) {
diff --git a/src/core/hle/kernel/hle_ipc.h b/src/core/hle/kernel/hle_ipc.h
index 754b41ff6..670cc741c 100644
--- a/src/core/hle/kernel/hle_ipc.h
+++ b/src/core/hle/kernel/hle_ipc.h
@@ -94,6 +94,7 @@ protected:
     std::weak_ptr<ServiceThread> service_thread;
 };
 
+using SessionRequestHandlerWeakPtr = std::weak_ptr<SessionRequestHandler>;
 using SessionRequestHandlerPtr = std::shared_ptr<SessionRequestHandler>;
 
 /**
@@ -139,7 +140,7 @@ public:
         }
     }
 
-    SessionRequestHandlerPtr DomainHandler(std::size_t index) const {
+    SessionRequestHandlerWeakPtr DomainHandler(std::size_t index) const {
         ASSERT_MSG(index < DomainHandlerCount(), "Unexpected handler index {}", index);
         return domain_handlers.at(index);
     }
@@ -328,10 +329,10 @@ public:
 
     template <typename T>
     std::shared_ptr<T> GetDomainHandler(std::size_t index) const {
-        return std::static_pointer_cast<T>(manager->DomainHandler(index));
+        return std::static_pointer_cast<T>(manager.lock()->DomainHandler(index).lock());
     }
 
-    void SetSessionRequestManager(std::shared_ptr<SessionRequestManager> manager_) {
+    void SetSessionRequestManager(std::weak_ptr<SessionRequestManager> manager_) {
         manager = std::move(manager_);
     }
 
@@ -374,7 +375,7 @@ private:
     u32 handles_offset{};
     u32 domain_offset{};
 
-    std::shared_ptr<SessionRequestManager> manager;
+    std::weak_ptr<SessionRequestManager> manager;
 
     KernelCore& kernel;
     Core::Memory::Memory& memory;
diff --git a/src/core/hle/kernel/init/init_slab_setup.cpp b/src/core/hle/kernel/init/init_slab_setup.cpp
index 36fc0944a..b0f773ee0 100644
--- a/src/core/hle/kernel/init/init_slab_setup.cpp
+++ b/src/core/hle/kernel/init/init_slab_setup.cpp
@@ -7,19 +7,23 @@
 #include "common/common_funcs.h"
 #include "common/common_types.h"
 #include "core/core.h"
+#include "core/device_memory.h"
 #include "core/hardware_properties.h"
 #include "core/hle/kernel/init/init_slab_setup.h"
 #include "core/hle/kernel/k_code_memory.h"
 #include "core/hle/kernel/k_event.h"
 #include "core/hle/kernel/k_memory_layout.h"
 #include "core/hle/kernel/k_memory_manager.h"
+#include "core/hle/kernel/k_page_buffer.h"
 #include "core/hle/kernel/k_port.h"
 #include "core/hle/kernel/k_process.h"
 #include "core/hle/kernel/k_resource_limit.h"
 #include "core/hle/kernel/k_session.h"
 #include "core/hle/kernel/k_shared_memory.h"
+#include "core/hle/kernel/k_shared_memory_info.h"
 #include "core/hle/kernel/k_system_control.h"
 #include "core/hle/kernel/k_thread.h"
+#include "core/hle/kernel/k_thread_local_page.h"
 #include "core/hle/kernel/k_transfer_memory.h"
 
 namespace Kernel::Init {
@@ -32,9 +36,13 @@ namespace Kernel::Init {
     HANDLER(KEvent, (SLAB_COUNT(KEvent)), ##__VA_ARGS__)                                           \
     HANDLER(KPort, (SLAB_COUNT(KPort)), ##__VA_ARGS__)                                             \
     HANDLER(KSharedMemory, (SLAB_COUNT(KSharedMemory)), ##__VA_ARGS__)                             \
+    HANDLER(KSharedMemoryInfo, (SLAB_COUNT(KSharedMemory) * 8), ##__VA_ARGS__)                     \
     HANDLER(KTransferMemory, (SLAB_COUNT(KTransferMemory)), ##__VA_ARGS__)                         \
     HANDLER(KCodeMemory, (SLAB_COUNT(KCodeMemory)), ##__VA_ARGS__)                                 \
     HANDLER(KSession, (SLAB_COUNT(KSession)), ##__VA_ARGS__)                                       \
+    HANDLER(KThreadLocalPage,                                                                      \
+            (SLAB_COUNT(KProcess) + (SLAB_COUNT(KProcess) + SLAB_COUNT(KThread)) / 8),             \
+            ##__VA_ARGS__)                                                                         \
     HANDLER(KResourceLimit, (SLAB_COUNT(KResourceLimit)), ##__VA_ARGS__)
 
 namespace {
@@ -50,38 +58,46 @@ enum KSlabType : u32 {
 // Constexpr counts.
 constexpr size_t SlabCountKProcess = 80;
 constexpr size_t SlabCountKThread = 800;
-constexpr size_t SlabCountKEvent = 700;
+constexpr size_t SlabCountKEvent = 900;
 constexpr size_t SlabCountKInterruptEvent = 100;
-constexpr size_t SlabCountKPort = 256 + 0x20; // Extra 0x20 ports over Nintendo for homebrew.
+constexpr size_t SlabCountKPort = 384;
 constexpr size_t SlabCountKSharedMemory = 80;
 constexpr size_t SlabCountKTransferMemory = 200;
 constexpr size_t SlabCountKCodeMemory = 10;
 constexpr size_t SlabCountKDeviceAddressSpace = 300;
-constexpr size_t SlabCountKSession = 933;
+constexpr size_t SlabCountKSession = 1133;
 constexpr size_t SlabCountKLightSession = 100;
 constexpr size_t SlabCountKObjectName = 7;
 constexpr size_t SlabCountKResourceLimit = 5;
 constexpr size_t SlabCountKDebug = Core::Hardware::NUM_CPU_CORES;
-constexpr size_t SlabCountKAlpha = 1;
-constexpr size_t SlabCountKBeta = 6;
+constexpr size_t SlabCountKIoPool = 1;
+constexpr size_t SlabCountKIoRegion = 6;
 
 constexpr size_t SlabCountExtraKThread = 160;
 
+/// Helper function to translate from the slab virtual address to the reserved location in physical
+/// memory.
+static PAddr TranslateSlabAddrToPhysical(KMemoryLayout& memory_layout, VAddr slab_addr) {
+    slab_addr -= memory_layout.GetSlabRegionAddress();
+    return slab_addr + Core::DramMemoryMap::SlabHeapBase;
+}
+
 template <typename T>
 VAddr InitializeSlabHeap(Core::System& system, KMemoryLayout& memory_layout, VAddr address,
                          size_t num_objects) {
-    // TODO(bunnei): This is just a place holder. We should initialize the appropriate KSlabHeap for
-    // kernel object type T with the backing kernel memory pointer once we emulate kernel memory.
 
     const size_t size = Common::AlignUp(sizeof(T) * num_objects, alignof(void*));
     VAddr start = Common::AlignUp(address, alignof(T));
 
-    // This is intentionally empty. Once KSlabHeap is fully implemented, we can replace this with
-    // the pointer to emulated memory to pass along. Until then, KSlabHeap will just allocate/free
-    // host memory.
-    void* backing_kernel_memory{};
+    // This should use the virtual memory address passed in, but currently, we do not setup the
+    // kernel virtual memory layout. Instead, we simply map these at a region of physical memory
+    // that we reserve for the slab heaps.
+    // TODO(bunnei): Fix this once we support the kernel virtual memory layout.
 
     if (size > 0) {
+        void* backing_kernel_memory{
+            system.DeviceMemory().GetPointer(TranslateSlabAddrToPhysical(memory_layout, start))};
+
         const KMemoryRegion* region = memory_layout.FindVirtual(start + size - 1);
         ASSERT(region != nullptr);
         ASSERT(region->IsDerivedFrom(KMemoryRegionType_KernelSlab));
@@ -91,6 +107,12 @@ VAddr InitializeSlabHeap(Core::System& system, KMemoryLayout& memory_layout, VAd
     return start + size;
 }
 
+size_t CalculateSlabHeapGapSize() {
+    constexpr size_t KernelSlabHeapGapSize = 2_MiB - 296_KiB;
+    static_assert(KernelSlabHeapGapSize <= KernelSlabHeapGapsSizeMax);
+    return KernelSlabHeapGapSize;
+}
+
 } // namespace
 
 KSlabResourceCounts KSlabResourceCounts::CreateDefault() {
@@ -109,8 +131,8 @@ KSlabResourceCounts KSlabResourceCounts::CreateDefault() {
         .num_KObjectName = SlabCountKObjectName,
         .num_KResourceLimit = SlabCountKResourceLimit,
         .num_KDebug = SlabCountKDebug,
-        .num_KAlpha = SlabCountKAlpha,
-        .num_KBeta = SlabCountKBeta,
+        .num_KIoPool = SlabCountKIoPool,
+        .num_KIoRegion = SlabCountKIoRegion,
     };
 }
 
@@ -136,11 +158,34 @@ size_t CalculateTotalSlabHeapSize(const KernelCore& kernel) {
 #undef ADD_SLAB_SIZE
 
     // Add the reserved size.
-    size += KernelSlabHeapGapsSize;
+    size += CalculateSlabHeapGapSize();
 
     return size;
 }
 
+void InitializeKPageBufferSlabHeap(Core::System& system) {
+    auto& kernel = system.Kernel();
+
+    const auto& counts = kernel.SlabResourceCounts();
+    const size_t num_pages =
+        counts.num_KProcess + counts.num_KThread + (counts.num_KProcess + counts.num_KThread) / 8;
+    const size_t slab_size = num_pages * PageSize;
+
+    // Reserve memory from the system resource limit.
+    ASSERT(kernel.GetSystemResourceLimit()->Reserve(LimitableResource::PhysicalMemory, slab_size));
+
+    // Allocate memory for the slab.
+    constexpr auto AllocateOption = KMemoryManager::EncodeOption(
+        KMemoryManager::Pool::System, KMemoryManager::Direction::FromFront);
+    const PAddr slab_address =
+        kernel.MemoryManager().AllocateAndOpenContinuous(num_pages, 1, AllocateOption);
+    ASSERT(slab_address != 0);
+
+    // Initialize the slabheap.
+    KPageBuffer::InitializeSlabHeap(kernel, system.DeviceMemory().GetPointer(slab_address),
+                                    slab_size);
+}
+
 void InitializeSlabHeaps(Core::System& system, KMemoryLayout& memory_layout) {
     auto& kernel = system.Kernel();
 
@@ -160,13 +205,13 @@ void InitializeSlabHeaps(Core::System& system, KMemoryLayout& memory_layout) {
     }
 
     // Create an array to represent the gaps between the slabs.
-    const size_t total_gap_size = KernelSlabHeapGapsSize;
+    const size_t total_gap_size = CalculateSlabHeapGapSize();
     std::array<size_t, slab_types.size()> slab_gaps;
-    for (size_t i = 0; i < slab_gaps.size(); i++) {
+    for (auto& slab_gap : slab_gaps) {
         // Note: This is an off-by-one error from Nintendo's intention, because GenerateRandomRange
         // is inclusive. However, Nintendo also has the off-by-one error, and it's "harmless", so we
         // will include it ourselves.
-        slab_gaps[i] = KSystemControl::GenerateRandomRange(0, total_gap_size);
+        slab_gap = KSystemControl::GenerateRandomRange(0, total_gap_size);
     }
 
     // Sort the array, so that we can treat differences between values as offsets to the starts of
@@ -177,13 +222,21 @@ void InitializeSlabHeaps(Core::System& system, KMemoryLayout& memory_layout) {
         }
     }
 
-    for (size_t i = 0; i < slab_types.size(); i++) {
+    // Track the gaps, so that we can free them to the unused slab tree.
+    VAddr gap_start = address;
+    size_t gap_size = 0;
+
+    for (size_t i = 0; i < slab_gaps.size(); i++) {
         // Add the random gap to the address.
-        address += (i == 0) ? slab_gaps[0] : slab_gaps[i] - slab_gaps[i - 1];
+        const auto cur_gap = (i == 0) ? slab_gaps[0] : slab_gaps[i] - slab_gaps[i - 1];
+        address += cur_gap;
+        gap_size += cur_gap;
 
 #define INITIALIZE_SLAB_HEAP(NAME, COUNT, ...)                                                     \
     case KSlabType_##NAME:                                                                         \
-        address = InitializeSlabHeap<NAME>(system, memory_layout, address, COUNT);                 \
+        if (COUNT > 0) {                                                                           \
+            address = InitializeSlabHeap<NAME>(system, memory_layout, address, COUNT);             \
+        }                                                                                          \
         break;
 
         // Initialize the slabheap.
@@ -192,7 +245,13 @@ void InitializeSlabHeaps(Core::System& system, KMemoryLayout& memory_layout) {
             FOREACH_SLAB_TYPE(INITIALIZE_SLAB_HEAP)
             // If we somehow get an invalid type, abort.
         default:
-            UNREACHABLE();
+            UNREACHABLE_MSG("Unknown slab type: {}", slab_types[i]);
+        }
+
+        // If we've hit the end of a gap, free it.
+        if (gap_start + gap_size != address) {
+            gap_start = address;
+            gap_size = 0;
         }
     }
 }
diff --git a/src/core/hle/kernel/init/init_slab_setup.h b/src/core/hle/kernel/init/init_slab_setup.h
index a8f7e0918..f54b67d02 100644
--- a/src/core/hle/kernel/init/init_slab_setup.h
+++ b/src/core/hle/kernel/init/init_slab_setup.h
@@ -32,12 +32,13 @@ struct KSlabResourceCounts {
     size_t num_KObjectName;
     size_t num_KResourceLimit;
     size_t num_KDebug;
-    size_t num_KAlpha;
-    size_t num_KBeta;
+    size_t num_KIoPool;
+    size_t num_KIoRegion;
 };
 
 void InitializeSlabResourceCounts(KernelCore& kernel);
 size_t CalculateTotalSlabHeapSize(const KernelCore& kernel);
+void InitializeKPageBufferSlabHeap(Core::System& system);
 void InitializeSlabHeaps(Core::System& system, KMemoryLayout& memory_layout);
 
 } // namespace Kernel::Init
diff --git a/src/core/hle/kernel/initial_process.h b/src/core/hle/kernel/initial_process.h
new file mode 100644
index 000000000..25b27909c
--- /dev/null
+++ b/src/core/hle/kernel/initial_process.h
@@ -0,0 +1,23 @@
+// Copyright 2022 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include "common/common_types.h"
+#include "common/literals.h"
+#include "core/hle/kernel/board/nintendo/nx/k_memory_layout.h"
+#include "core/hle/kernel/board/nintendo/nx/k_system_control.h"
+
+namespace Kernel {
+
+using namespace Common::Literals;
+
+constexpr std::size_t InitialProcessBinarySizeMax = 12_MiB;
+
+static inline PAddr GetInitialProcessBinaryPhysicalAddress() {
+    return Kernel::Board::Nintendo::Nx::KSystemControl::Init::GetKernelPhysicalBaseAddress(
+        MainMemoryAddress);
+}
+
+} // namespace Kernel
diff --git a/src/core/hle/kernel/k_address_arbiter.cpp b/src/core/hle/kernel/k_address_arbiter.cpp
index 783c69858..8cdd0490f 100644
--- a/src/core/hle/kernel/k_address_arbiter.cpp
+++ b/src/core/hle/kernel/k_address_arbiter.cpp
@@ -49,7 +49,7 @@ bool DecrementIfLessThan(Core::System& system, s32* out, VAddr address, s32 valu
         }
     } else {
         // Otherwise, clear our exclusive hold and finish
-        monitor.ClearExclusive();
+        monitor.ClearExclusive(current_core);
     }
 
     // We're done.
@@ -78,7 +78,7 @@ bool UpdateIfEqual(Core::System& system, s32* out, VAddr address, s32 value, s32
         }
     } else {
         // Otherwise, clear our exclusive hold and finish.
-        monitor.ClearExclusive();
+        monitor.ClearExclusive(current_core);
     }
 
     // We're done.
@@ -115,7 +115,7 @@ ResultCode KAddressArbiter::Signal(VAddr addr, s32 count) {
     {
         KScopedSchedulerLock sl(kernel);
 
-        auto it = thread_tree.nfind_light({addr, -1});
+        auto it = thread_tree.nfind_key({addr, -1});
         while ((it != thread_tree.end()) && (count <= 0 || num_waiters < count) &&
                (it->GetAddressArbiterKey() == addr)) {
             // End the thread's wait.
@@ -148,7 +148,7 @@ ResultCode KAddressArbiter::SignalAndIncrementIfEqual(VAddr addr, s32 value, s32
             return ResultInvalidState;
         }
 
-        auto it = thread_tree.nfind_light({addr, -1});
+        auto it = thread_tree.nfind_key({addr, -1});
         while ((it != thread_tree.end()) && (count <= 0 || num_waiters < count) &&
                (it->GetAddressArbiterKey() == addr)) {
             // End the thread's wait.
@@ -171,7 +171,7 @@ ResultCode KAddressArbiter::SignalAndModifyByWaitingCountIfEqual(VAddr addr, s32
     {
         [[maybe_unused]] const KScopedSchedulerLock sl(kernel);
 
-        auto it = thread_tree.nfind_light({addr, -1});
+        auto it = thread_tree.nfind_key({addr, -1});
         // Determine the updated value.
         s32 new_value{};
         if (count <= 0) {
diff --git a/src/core/hle/kernel/k_auto_object.h b/src/core/hle/kernel/k_auto_object.h
index 165b76747..05779f2d5 100644
--- a/src/core/hle/kernel/k_auto_object.h
+++ b/src/core/hle/kernel/k_auto_object.h
@@ -20,8 +20,6 @@ class KernelCore;
 class KProcess;
 
 #define KERNEL_AUTOOBJECT_TRAITS(CLASS, BASE_CLASS)                                                \
-    YUZU_NON_COPYABLE(CLASS);                                                                      \
-    YUZU_NON_MOVEABLE(CLASS);                                                                      \
                                                                                                    \
 private:                                                                                           \
     friend class ::Kernel::KClassTokenGenerator;                                                   \
@@ -32,6 +30,9 @@ private:
     }                                                                                              \
                                                                                                    \
 public:                                                                                            \
+    YUZU_NON_COPYABLE(CLASS);                                                                      \
+    YUZU_NON_MOVEABLE(CLASS);                                                                      \
+                                                                                                   \
     using BaseClass = BASE_CLASS;                                                                  \
     static constexpr TypeObj GetStaticTypeObj() {                                                  \
         constexpr ClassTokenType Token = ClassToken();                                             \
@@ -224,9 +225,9 @@ private:
 
 template <typename T>
 class KScopedAutoObject {
+public:
     YUZU_NON_COPYABLE(KScopedAutoObject);
 
-public:
     constexpr KScopedAutoObject() = default;
 
     constexpr KScopedAutoObject(T* o) : m_obj(o) {
diff --git a/src/core/hle/kernel/k_auto_object_container.h b/src/core/hle/kernel/k_auto_object_container.h
index 4eadfe99d..697cc4289 100644
--- a/src/core/hle/kernel/k_auto_object_container.h
+++ b/src/core/hle/kernel/k_auto_object_container.h
@@ -16,13 +16,12 @@ class KernelCore;
 class KProcess;
 
 class KAutoObjectWithListContainer {
+public:
     YUZU_NON_COPYABLE(KAutoObjectWithListContainer);
     YUZU_NON_MOVEABLE(KAutoObjectWithListContainer);
 
-public:
     using ListType = boost::intrusive::rbtree<KAutoObjectWithList>;
 
-public:
     class ListAccessor : public KScopedLightLock {
     public:
         explicit ListAccessor(KAutoObjectWithListContainer* container)
@@ -48,7 +47,6 @@ public:
 
     friend class ListAccessor;
 
-public:
     KAutoObjectWithListContainer(KernelCore& kernel) : m_lock(kernel), m_object_list() {}
 
     void Initialize() {}
diff --git a/src/core/hle/kernel/k_code_memory.cpp b/src/core/hle/kernel/k_code_memory.cpp
index d69f7ffb7..0b225e8e0 100644
--- a/src/core/hle/kernel/k_code_memory.cpp
+++ b/src/core/hle/kernel/k_code_memory.cpp
@@ -2,6 +2,7 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
+#include "common/alignment.h"
 #include "common/common_types.h"
 #include "core/device_memory.h"
 #include "core/hle/kernel/k_auto_object.h"
@@ -28,8 +29,7 @@ ResultCode KCodeMemory::Initialize(Core::DeviceMemory& device_memory, VAddr addr
     auto& page_table = m_owner->PageTable();
 
     // Construct the page group.
-    KMemoryInfo kBlockInfo = page_table.QueryInfo(addr);
-    m_page_group = KPageLinkedList(kBlockInfo.GetAddress(), kBlockInfo.GetNumPages());
+    m_page_group = KPageLinkedList(addr, Common::DivideUp(size, PageSize));
 
     // Lock the memory.
     R_TRY(page_table.LockForCodeMemory(addr, size))
@@ -143,4 +143,4 @@ ResultCode KCodeMemory::UnmapFromOwner(VAddr address, size_t size) {
     return ResultSuccess;
 }
 
-} // namespace Kernel
\ No newline at end of file
+} // namespace Kernel
diff --git a/src/core/hle/kernel/k_condition_variable.cpp b/src/core/hle/kernel/k_condition_variable.cpp
index aadcc297a..8e2a9593c 100644
--- a/src/core/hle/kernel/k_condition_variable.cpp
+++ b/src/core/hle/kernel/k_condition_variable.cpp
@@ -244,7 +244,7 @@ void KConditionVariable::Signal(u64 cv_key, s32 count) {
     {
         KScopedSchedulerLock sl(kernel);
 
-        auto it = thread_tree.nfind_light({cv_key, -1});
+        auto it = thread_tree.nfind_key({cv_key, -1});
         while ((it != thread_tree.end()) && (count <= 0 || num_waiters < count) &&
                (it->GetConditionVariableKey() == cv_key)) {
             KThread* target_thread = std::addressof(*it);
diff --git a/src/core/hle/kernel/k_event.cpp b/src/core/hle/kernel/k_event.cpp
index 0720efece..2e0e8de80 100644
--- a/src/core/hle/kernel/k_event.cpp
+++ b/src/core/hle/kernel/k_event.cpp
@@ -14,7 +14,7 @@ KEvent::KEvent(KernelCore& kernel_)
 
 KEvent::~KEvent() = default;
 
-void KEvent::Initialize(std::string&& name_) {
+void KEvent::Initialize(std::string&& name_, KProcess* owner_) {
     // Increment reference count.
     // Because reference count is one on creation, this will result
     // in a reference count of two. Thus, when both readable and
@@ -30,10 +30,8 @@ void KEvent::Initialize(std::string&& name_) {
     writable_event.Initialize(this, name_ + ":Writable");
 
     // Set our owner process.
-    owner = kernel.CurrentProcess();
-    if (owner) {
-        owner->Open();
-    }
+    owner = owner_;
+    owner->Open();
 
     // Mark initialized.
     name = std::move(name_);
@@ -47,10 +45,8 @@ void KEvent::Finalize() {
 void KEvent::PostDestroy(uintptr_t arg) {
     // Release the event count resource the owner process holds.
     KProcess* owner = reinterpret_cast<KProcess*>(arg);
-    if (owner) {
-        owner->GetResourceLimit()->Release(LimitableResource::Events, 1);
-        owner->Close();
-    }
+    owner->GetResourceLimit()->Release(LimitableResource::Events, 1);
+    owner->Close();
 }
 
 } // namespace Kernel
diff --git a/src/core/hle/kernel/k_event.h b/src/core/hle/kernel/k_event.h
index 3d3ec99e2..de9732ddf 100644
--- a/src/core/hle/kernel/k_event.h
+++ b/src/core/hle/kernel/k_event.h
@@ -22,7 +22,7 @@ public:
     explicit KEvent(KernelCore& kernel_);
     ~KEvent() override;
 
-    void Initialize(std::string&& name);
+    void Initialize(std::string&& name, KProcess* owner_);
 
     void Finalize() override;
 
diff --git a/src/core/hle/kernel/k_handle_table.cpp b/src/core/hle/kernel/k_handle_table.cpp
index cf95f0852..db7512ee7 100644
--- a/src/core/hle/kernel/k_handle_table.cpp
+++ b/src/core/hle/kernel/k_handle_table.cpp
@@ -63,7 +63,7 @@ bool KHandleTable::Remove(Handle handle) {
     return true;
 }
 
-ResultCode KHandleTable::Add(Handle* out_handle, KAutoObject* obj, u16 type) {
+ResultCode KHandleTable::Add(Handle* out_handle, KAutoObject* obj) {
     KScopedDisableDispatch dd(kernel);
     KScopedSpinLock lk(m_lock);
 
@@ -75,7 +75,7 @@ ResultCode KHandleTable::Add(Handle* out_handle, KAutoObject* obj, u16 type) {
         const auto linear_id = this->AllocateLinearId();
         const auto index = this->AllocateEntry();
 
-        m_entry_infos[index].info = {.linear_id = linear_id, .type = type};
+        m_entry_infos[index].linear_id = linear_id;
         m_objects[index] = obj;
 
         obj->Open();
@@ -116,7 +116,7 @@ void KHandleTable::Unreserve(Handle handle) {
     }
 }
 
-void KHandleTable::Register(Handle handle, KAutoObject* obj, u16 type) {
+void KHandleTable::Register(Handle handle, KAutoObject* obj) {
     KScopedDisableDispatch dd(kernel);
     KScopedSpinLock lk(m_lock);
 
@@ -132,7 +132,7 @@ void KHandleTable::Register(Handle handle, KAutoObject* obj, u16 type) {
         // Set the entry.
         ASSERT(m_objects[index] == nullptr);
 
-        m_entry_infos[index].info = {.linear_id = static_cast<u16>(linear_id), .type = type};
+        m_entry_infos[index].linear_id = static_cast<u16>(linear_id);
         m_objects[index] = obj;
 
         obj->Open();
diff --git a/src/core/hle/kernel/k_handle_table.h b/src/core/hle/kernel/k_handle_table.h
index 4b114ec2f..dd27689b6 100644
--- a/src/core/hle/kernel/k_handle_table.h
+++ b/src/core/hle/kernel/k_handle_table.h
@@ -22,13 +22,12 @@ namespace Kernel {
 class KernelCore;
 
 class KHandleTable {
+public:
     YUZU_NON_COPYABLE(KHandleTable);
     YUZU_NON_MOVEABLE(KHandleTable);
 
-public:
     static constexpr size_t MaxTableSize = 1024;
 
-public:
     explicit KHandleTable(KernelCore& kernel_);
     ~KHandleTable();
 
@@ -43,7 +42,7 @@ public:
         m_free_head_index = -1;
 
         // Free all entries.
-        for (s32 i = 0; i < static_cast<s32>(m_table_size); ++i) {
+        for (s16 i = 0; i < static_cast<s16>(m_table_size); ++i) {
             m_objects[i] = nullptr;
             m_entry_infos[i].next_free_index = i - 1;
             m_free_head_index = i;
@@ -105,17 +104,8 @@ public:
     ResultCode Reserve(Handle* out_handle);
     void Unreserve(Handle handle);
 
-    template <typename T>
-    ResultCode Add(Handle* out_handle, T* obj) {
-        static_assert(std::is_base_of_v<KAutoObject, T>);
-        return this->Add(out_handle, obj, obj->GetTypeObj().GetClassToken());
-    }
-
-    template <typename T>
-    void Register(Handle handle, T* obj) {
-        static_assert(std::is_base_of_v<KAutoObject, T>);
-        return this->Register(handle, obj, obj->GetTypeObj().GetClassToken());
-    }
+    ResultCode Add(Handle* out_handle, KAutoObject* obj);
+    void Register(Handle handle, KAutoObject* obj);
 
     template <typename T>
     bool GetMultipleObjects(T** out, const Handle* handles, size_t num_handles) const {
@@ -161,9 +151,6 @@ public:
     }
 
 private:
-    ResultCode Add(Handle* out_handle, KAutoObject* obj, u16 type);
-    void Register(Handle handle, KAutoObject* obj, u16 type);
-
     s32 AllocateEntry() {
         ASSERT(m_count < m_table_size);
 
@@ -180,7 +167,7 @@ private:
         ASSERT(m_count > 0);
 
         m_objects[index] = nullptr;
-        m_entry_infos[index].next_free_index = m_free_head_index;
+        m_entry_infos[index].next_free_index = static_cast<s16>(m_free_head_index);
 
         m_free_head_index = index;
 
@@ -279,19 +266,13 @@ private:
     }
 
     union EntryInfo {
-        struct {
-            u16 linear_id;
-            u16 type;
-        } info;
-        s32 next_free_index;
+        u16 linear_id;
+        s16 next_free_index;
 
         constexpr u16 GetLinearId() const {
-            return info.linear_id;
-        }
-        constexpr u16 GetType() const {
-            return info.type;
+            return linear_id;
         }
-        constexpr s32 GetNextFreeIndex() const {
+        constexpr s16 GetNextFreeIndex() const {
             return next_free_index;
         }
     };
diff --git a/src/core/hle/kernel/k_memory_layout.h b/src/core/hle/kernel/k_memory_layout.h
index 57ff538cc..0858827b6 100644
--- a/src/core/hle/kernel/k_memory_layout.h
+++ b/src/core/hle/kernel/k_memory_layout.h
@@ -57,11 +57,11 @@ constexpr std::size_t KernelPageTableHeapSize = GetMaximumOverheadSize(MainMemor
 constexpr std::size_t KernelInitialPageHeapSize = 128_KiB;
 
 constexpr std::size_t KernelSlabHeapDataSize = 5_MiB;
-constexpr std::size_t KernelSlabHeapGapsSize = 2_MiB - 64_KiB;
-constexpr std::size_t KernelSlabHeapSize = KernelSlabHeapDataSize + KernelSlabHeapGapsSize;
+constexpr std::size_t KernelSlabHeapGapsSizeMax = 2_MiB - 64_KiB;
+constexpr std::size_t KernelSlabHeapSize = KernelSlabHeapDataSize + KernelSlabHeapGapsSizeMax;
 
 // NOTE: This is calculated from KThread slab counts, assuming KThread size <= 0x860.
-constexpr std::size_t KernelSlabHeapAdditionalSize = 416_KiB;
+constexpr std::size_t KernelSlabHeapAdditionalSize = 0x68000;
 
 constexpr std::size_t KernelResourceSize =
     KernelPageTableHeapSize + KernelInitialPageHeapSize + KernelSlabHeapSize;
@@ -173,6 +173,10 @@ public:
         return Dereference(FindVirtualLinear(address));
     }
 
+    const KMemoryRegion& GetPhysicalLinearRegion(PAddr address) const {
+        return Dereference(FindPhysicalLinear(address));
+    }
+
     const KMemoryRegion* GetPhysicalKernelTraceBufferRegion() const {
         return GetPhysicalMemoryRegionTree().FindFirstDerived(KMemoryRegionType_KernelTraceBuffer);
     }
diff --git a/src/core/hle/kernel/k_memory_manager.cpp b/src/core/hle/kernel/k_memory_manager.cpp
index 1b44541b1..a2f18f643 100644
--- a/src/core/hle/kernel/k_memory_manager.cpp
+++ b/src/core/hle/kernel/k_memory_manager.cpp
@@ -10,189 +10,412 @@
 #include "common/scope_exit.h"
 #include "core/core.h"
 #include "core/device_memory.h"
+#include "core/hle/kernel/initial_process.h"
 #include "core/hle/kernel/k_memory_manager.h"
 #include "core/hle/kernel/k_page_linked_list.h"
+#include "core/hle/kernel/kernel.h"
 #include "core/hle/kernel/svc_results.h"
+#include "core/memory.h"
 
 namespace Kernel {
 
-KMemoryManager::KMemoryManager(Core::System& system_) : system{system_} {}
+namespace {
+
+constexpr KMemoryManager::Pool GetPoolFromMemoryRegionType(u32 type) {
+    if ((type | KMemoryRegionType_DramApplicationPool) == type) {
+        return KMemoryManager::Pool::Application;
+    } else if ((type | KMemoryRegionType_DramAppletPool) == type) {
+        return KMemoryManager::Pool::Applet;
+    } else if ((type | KMemoryRegionType_DramSystemPool) == type) {
+        return KMemoryManager::Pool::System;
+    } else if ((type | KMemoryRegionType_DramSystemNonSecurePool) == type) {
+        return KMemoryManager::Pool::SystemNonSecure;
+    } else {
+        UNREACHABLE_MSG("InvalidMemoryRegionType for conversion to Pool");
+        return {};
+    }
+}
 
-std::size_t KMemoryManager::Impl::Initialize(Pool new_pool, u64 start_address, u64 end_address) {
-    const auto size{end_address - start_address};
+} // namespace
+
+KMemoryManager::KMemoryManager(Core::System& system_)
+    : system{system_}, pool_locks{
+                           KLightLock{system_.Kernel()},
+                           KLightLock{system_.Kernel()},
+                           KLightLock{system_.Kernel()},
+                           KLightLock{system_.Kernel()},
+                       } {}
+
+void KMemoryManager::Initialize(VAddr management_region, size_t management_region_size) {
+
+    // Clear the management region to zero.
+    const VAddr management_region_end = management_region + management_region_size;
+
+    // Reset our manager count.
+    num_managers = 0;
+
+    // Traverse the virtual memory layout tree, initializing each manager as appropriate.
+    while (num_managers != MaxManagerCount) {
+        // Locate the region that should initialize the current manager.
+        PAddr region_address = 0;
+        size_t region_size = 0;
+        Pool region_pool = Pool::Count;
+        for (const auto& it : system.Kernel().MemoryLayout().GetPhysicalMemoryRegionTree()) {
+            // We only care about regions that we need to create managers for.
+            if (!it.IsDerivedFrom(KMemoryRegionType_DramUserPool)) {
+                continue;
+            }
 
-    // Calculate metadata sizes
-    const auto ref_count_size{(size / PageSize) * sizeof(u16)};
-    const auto optimize_map_size{(Common::AlignUp((size / PageSize), 64) / 64) * sizeof(u64)};
-    const auto manager_size{Common::AlignUp(optimize_map_size + ref_count_size, PageSize)};
-    const auto page_heap_size{KPageHeap::CalculateManagementOverheadSize(size)};
-    const auto total_metadata_size{manager_size + page_heap_size};
-    ASSERT(manager_size <= total_metadata_size);
-    ASSERT(Common::IsAligned(total_metadata_size, PageSize));
+            // We want to initialize the managers in order.
+            if (it.GetAttributes() != num_managers) {
+                continue;
+            }
 
-    // Setup region
-    pool = new_pool;
+            const PAddr cur_start = it.GetAddress();
+            const PAddr cur_end = it.GetEndAddress();
+
+            // Validate the region.
+            ASSERT(cur_end != 0);
+            ASSERT(cur_start != 0);
+            ASSERT(it.GetSize() > 0);
+
+            // Update the region's extents.
+            if (region_address == 0) {
+                region_address = cur_start;
+                region_size = it.GetSize();
+                region_pool = GetPoolFromMemoryRegionType(it.GetType());
+            } else {
+                ASSERT(cur_start == region_address + region_size);
+
+                // Update the size.
+                region_size = cur_end - region_address;
+                ASSERT(GetPoolFromMemoryRegionType(it.GetType()) == region_pool);
+            }
+        }
+
+        // If we didn't find a region, we're done.
+        if (region_size == 0) {
+            break;
+        }
 
-    // Initialize the manager's KPageHeap
-    heap.Initialize(start_address, size, page_heap_size);
+        // Initialize a new manager for the region.
+        Impl* manager = std::addressof(managers[num_managers++]);
+        ASSERT(num_managers <= managers.size());
+
+        const size_t cur_size = manager->Initialize(region_address, region_size, management_region,
+                                                    management_region_end, region_pool);
+        management_region += cur_size;
+        ASSERT(management_region <= management_region_end);
+
+        // Insert the manager into the pool list.
+        const auto region_pool_index = static_cast<u32>(region_pool);
+        if (pool_managers_tail[region_pool_index] == nullptr) {
+            pool_managers_head[region_pool_index] = manager;
+        } else {
+            pool_managers_tail[region_pool_index]->SetNext(manager);
+            manager->SetPrev(pool_managers_tail[region_pool_index]);
+        }
+        pool_managers_tail[region_pool_index] = manager;
+    }
 
-    // Free the memory to the heap
-    heap.Free(start_address, size / PageSize);
+    // Free each region to its corresponding heap.
+    size_t reserved_sizes[MaxManagerCount] = {};
+    const PAddr ini_start = GetInitialProcessBinaryPhysicalAddress();
+    const PAddr ini_end = ini_start + InitialProcessBinarySizeMax;
+    const PAddr ini_last = ini_end - 1;
+    for (const auto& it : system.Kernel().MemoryLayout().GetPhysicalMemoryRegionTree()) {
+        if (it.IsDerivedFrom(KMemoryRegionType_DramUserPool)) {
+            // Get the manager for the region.
+            auto index = it.GetAttributes();
+            auto& manager = managers[index];
+
+            const PAddr cur_start = it.GetAddress();
+            const PAddr cur_last = it.GetLastAddress();
+            const PAddr cur_end = it.GetEndAddress();
+
+            if (cur_start <= ini_start && ini_last <= cur_last) {
+                // Free memory before the ini to the heap.
+                if (cur_start != ini_start) {
+                    manager.Free(cur_start, (ini_start - cur_start) / PageSize);
+                }
 
-    // Update the heap's used size
-    heap.UpdateUsedSize();
+                // Open/reserve the ini memory.
+                manager.OpenFirst(ini_start, InitialProcessBinarySizeMax / PageSize);
+                reserved_sizes[it.GetAttributes()] += InitialProcessBinarySizeMax;
 
-    return total_metadata_size;
-}
+                // Free memory after the ini to the heap.
+                if (ini_last != cur_last) {
+                    ASSERT(cur_end != 0);
+                    manager.Free(ini_end, cur_end - ini_end);
+                }
+            } else {
+                // Ensure there's no partial overlap with the ini image.
+                if (cur_start <= ini_last) {
+                    ASSERT(cur_last < ini_start);
+                } else {
+                    // Otherwise, check the region for general validity.
+                    ASSERT(cur_end != 0);
+                }
 
-void KMemoryManager::InitializeManager(Pool pool, u64 start_address, u64 end_address) {
-    ASSERT(pool < Pool::Count);
-    managers[static_cast<std::size_t>(pool)].Initialize(pool, start_address, end_address);
+                // Free the memory to the heap.
+                manager.Free(cur_start, it.GetSize() / PageSize);
+            }
+        }
+    }
+
+    // Update the used size for all managers.
+    for (size_t i = 0; i < num_managers; ++i) {
+        managers[i].SetInitialUsedHeapSize(reserved_sizes[i]);
+    }
 }
 
-VAddr KMemoryManager::AllocateAndOpenContinuous(std::size_t num_pages, std::size_t align_pages,
-                                                u32 option) {
-    // Early return if we're allocating no pages
+PAddr KMemoryManager::AllocateAndOpenContinuous(size_t num_pages, size_t align_pages, u32 option) {
+    // Early return if we're allocating no pages.
     if (num_pages == 0) {
-        return {};
+        return 0;
     }
 
-    // Lock the pool that we're allocating from
+    // Lock the pool that we're allocating from.
     const auto [pool, dir] = DecodeOption(option);
-    const auto pool_index{static_cast<std::size_t>(pool)};
-    std::lock_guard lock{pool_locks[pool_index]};
-
-    // Choose a heap based on our page size request
-    const s32 heap_index{KPageHeap::GetAlignedBlockIndex(num_pages, align_pages)};
-
-    // Loop, trying to iterate from each block
-    // TODO (bunnei): Support multiple managers
-    Impl& chosen_manager{managers[pool_index]};
-    VAddr allocated_block{chosen_manager.AllocateBlock(heap_index, false)};
+    KScopedLightLock lk(pool_locks[static_cast<std::size_t>(pool)]);
+
+    // Choose a heap based on our page size request.
+    const s32 heap_index = KPageHeap::GetAlignedBlockIndex(num_pages, align_pages);
+
+    // Loop, trying to iterate from each block.
+    Impl* chosen_manager = nullptr;
+    PAddr allocated_block = 0;
+    for (chosen_manager = this->GetFirstManager(pool, dir); chosen_manager != nullptr;
+         chosen_manager = this->GetNextManager(chosen_manager, dir)) {
+        allocated_block = chosen_manager->AllocateBlock(heap_index, true);
+        if (allocated_block != 0) {
+            break;
+        }
+    }
 
-    // If we failed to allocate, quit now
-    if (!allocated_block) {
-        return {};
+    // If we failed to allocate, quit now.
+    if (allocated_block == 0) {
+        return 0;
     }
 
-    // If we allocated more than we need, free some
-    const auto allocated_pages{KPageHeap::GetBlockNumPages(heap_index)};
+    // If we allocated more than we need, free some.
+    const size_t allocated_pages = KPageHeap::GetBlockNumPages(heap_index);
     if (allocated_pages > num_pages) {
-        chosen_manager.Free(allocated_block + num_pages * PageSize, allocated_pages - num_pages);
+        chosen_manager->Free(allocated_block + num_pages * PageSize, allocated_pages - num_pages);
     }
 
+    // Open the first reference to the pages.
+    chosen_manager->OpenFirst(allocated_block, num_pages);
+
     return allocated_block;
 }
 
-ResultCode KMemoryManager::Allocate(KPageLinkedList& page_list, std::size_t num_pages, Pool pool,
-                                    Direction dir, u32 heap_fill_value) {
-    ASSERT(page_list.GetNumPages() == 0);
+ResultCode KMemoryManager::AllocatePageGroupImpl(KPageLinkedList* out, size_t num_pages, Pool pool,
+                                                 Direction dir, bool random) {
+    // Choose a heap based on our page size request.
+    const s32 heap_index = KPageHeap::GetBlockIndex(num_pages);
+    R_UNLESS(0 <= heap_index, ResultOutOfMemory);
+
+    // Ensure that we don't leave anything un-freed.
+    auto group_guard = SCOPE_GUARD({
+        for (const auto& it : out->Nodes()) {
+            auto& manager = this->GetManager(system.Kernel().MemoryLayout(), it.GetAddress());
+            const size_t num_pages_to_free =
+                std::min(it.GetNumPages(), (manager.GetEndAddress() - it.GetAddress()) / PageSize);
+            manager.Free(it.GetAddress(), num_pages_to_free);
+        }
+    });
 
-    // Early return if we're allocating no pages
-    if (num_pages == 0) {
-        return ResultSuccess;
-    }
+    // Keep allocating until we've allocated all our pages.
+    for (s32 index = heap_index; index >= 0 && num_pages > 0; index--) {
+        const size_t pages_per_alloc = KPageHeap::GetBlockNumPages(index);
+        for (Impl* cur_manager = this->GetFirstManager(pool, dir); cur_manager != nullptr;
+             cur_manager = this->GetNextManager(cur_manager, dir)) {
+            while (num_pages >= pages_per_alloc) {
+                // Allocate a block.
+                PAddr allocated_block = cur_manager->AllocateBlock(index, random);
+                if (allocated_block == 0) {
+                    break;
+                }
 
-    // Lock the pool that we're allocating from
-    const auto pool_index{static_cast<std::size_t>(pool)};
-    std::lock_guard lock{pool_locks[pool_index]};
+                // Safely add it to our group.
+                {
+                    auto block_guard =
+                        SCOPE_GUARD({ cur_manager->Free(allocated_block, pages_per_alloc); });
+                    R_TRY(out->AddBlock(allocated_block, pages_per_alloc));
+                    block_guard.Cancel();
+                }
 
-    // Choose a heap based on our page size request
-    const s32 heap_index{KPageHeap::GetBlockIndex(num_pages)};
-    if (heap_index < 0) {
-        return ResultOutOfMemory;
+                num_pages -= pages_per_alloc;
+            }
+        }
     }
 
-    // TODO (bunnei): Support multiple managers
-    Impl& chosen_manager{managers[pool_index]};
+    // Only succeed if we allocated as many pages as we wanted.
+    R_UNLESS(num_pages == 0, ResultOutOfMemory);
 
-    // Ensure that we don't leave anything un-freed
-    auto group_guard = detail::ScopeExit([&] {
-        for (const auto& it : page_list.Nodes()) {
-            const auto min_num_pages{std::min<size_t>(
-                it.GetNumPages(), (chosen_manager.GetEndAddress() - it.GetAddress()) / PageSize)};
-            chosen_manager.Free(it.GetAddress(), min_num_pages);
-        }
-    });
+    // We succeeded!
+    group_guard.Cancel();
+    return ResultSuccess;
+}
 
-    // Keep allocating until we've allocated all our pages
-    for (s32 index{heap_index}; index >= 0 && num_pages > 0; index--) {
-        const auto pages_per_alloc{KPageHeap::GetBlockNumPages(index)};
+ResultCode KMemoryManager::AllocateAndOpen(KPageLinkedList* out, size_t num_pages, u32 option) {
+    ASSERT(out != nullptr);
+    ASSERT(out->GetNumPages() == 0);
 
-        while (num_pages >= pages_per_alloc) {
-            // Allocate a block
-            VAddr allocated_block{chosen_manager.AllocateBlock(index, false)};
-            if (!allocated_block) {
-                break;
-            }
+    // Early return if we're allocating no pages.
+    R_SUCCEED_IF(num_pages == 0);
 
-            // Safely add it to our group
-            {
-                auto block_guard = detail::ScopeExit(
-                    [&] { chosen_manager.Free(allocated_block, pages_per_alloc); });
+    // Lock the pool that we're allocating from.
+    const auto [pool, dir] = DecodeOption(option);
+    KScopedLightLock lk(pool_locks[static_cast<size_t>(pool)]);
+
+    // Allocate the page group.
+    R_TRY(this->AllocatePageGroupImpl(out, num_pages, pool, dir, false));
+
+    // Open the first reference to the pages.
+    for (const auto& block : out->Nodes()) {
+        PAddr cur_address = block.GetAddress();
+        size_t remaining_pages = block.GetNumPages();
+        while (remaining_pages > 0) {
+            // Get the manager for the current address.
+            auto& manager = this->GetManager(system.Kernel().MemoryLayout(), cur_address);
+
+            // Process part or all of the block.
+            const size_t cur_pages =
+                std::min(remaining_pages, manager.GetPageOffsetToEnd(cur_address));
+            manager.OpenFirst(cur_address, cur_pages);
+
+            // Advance.
+            cur_address += cur_pages * PageSize;
+            remaining_pages -= cur_pages;
+        }
+    }
 
-                if (const ResultCode result{page_list.AddBlock(allocated_block, pages_per_alloc)};
-                    result.IsError()) {
-                    return result;
-                }
+    return ResultSuccess;
+}
 
-                block_guard.Cancel();
-            }
+ResultCode KMemoryManager::AllocateAndOpenForProcess(KPageLinkedList* out, size_t num_pages,
+                                                     u32 option, u64 process_id, u8 fill_pattern) {
+    ASSERT(out != nullptr);
+    ASSERT(out->GetNumPages() == 0);
 
-            num_pages -= pages_per_alloc;
-        }
-    }
+    // Decode the option.
+    const auto [pool, dir] = DecodeOption(option);
 
-    // Clear allocated memory.
-    for (const auto& it : page_list.Nodes()) {
-        std::memset(system.DeviceMemory().GetPointer(it.GetAddress()), heap_fill_value,
-                    it.GetSize());
+    // Allocate the memory.
+    {
+        // Lock the pool that we're allocating from.
+        KScopedLightLock lk(pool_locks[static_cast<size_t>(pool)]);
+
+        // Allocate the page group.
+        R_TRY(this->AllocatePageGroupImpl(out, num_pages, pool, dir, false));
+
+        // Open the first reference to the pages.
+        for (const auto& block : out->Nodes()) {
+            PAddr cur_address = block.GetAddress();
+            size_t remaining_pages = block.GetNumPages();
+            while (remaining_pages > 0) {
+                // Get the manager for the current address.
+                auto& manager = this->GetManager(system.Kernel().MemoryLayout(), cur_address);
+
+                // Process part or all of the block.
+                const size_t cur_pages =
+                    std::min(remaining_pages, manager.GetPageOffsetToEnd(cur_address));
+                manager.OpenFirst(cur_address, cur_pages);
+
+                // Advance.
+                cur_address += cur_pages * PageSize;
+                remaining_pages -= cur_pages;
+            }
+        }
     }
 
-    // Only succeed if we allocated as many pages as we wanted
-    if (num_pages) {
-        return ResultOutOfMemory;
+    // Set all the allocated memory.
+    for (const auto& block : out->Nodes()) {
+        std::memset(system.DeviceMemory().GetPointer(block.GetAddress()), fill_pattern,
+                    block.GetSize());
     }
 
-    // We succeeded!
-    group_guard.Cancel();
-
     return ResultSuccess;
 }
 
-ResultCode KMemoryManager::Free(KPageLinkedList& page_list, std::size_t num_pages, Pool pool,
-                                Direction dir, u32 heap_fill_value) {
-    // Early return if we're freeing no pages
-    if (!num_pages) {
-        return ResultSuccess;
+void KMemoryManager::Open(PAddr address, size_t num_pages) {
+    // Repeatedly open references until we've done so for all pages.
+    while (num_pages) {
+        auto& manager = this->GetManager(system.Kernel().MemoryLayout(), address);
+        const size_t cur_pages = std::min(num_pages, manager.GetPageOffsetToEnd(address));
+
+        {
+            KScopedLightLock lk(pool_locks[static_cast<size_t>(manager.GetPool())]);
+            manager.Open(address, cur_pages);
+        }
+
+        num_pages -= cur_pages;
+        address += cur_pages * PageSize;
     }
+}
 
-    // Lock the pool that we're freeing from
-    const auto pool_index{static_cast<std::size_t>(pool)};
-    std::lock_guard lock{pool_locks[pool_index]};
+void KMemoryManager::Close(PAddr address, size_t num_pages) {
+    // Repeatedly close references until we've done so for all pages.
+    while (num_pages) {
+        auto& manager = this->GetManager(system.Kernel().MemoryLayout(), address);
+        const size_t cur_pages = std::min(num_pages, manager.GetPageOffsetToEnd(address));
 
-    // TODO (bunnei): Support multiple managers
-    Impl& chosen_manager{managers[pool_index]};
+        {
+            KScopedLightLock lk(pool_locks[static_cast<size_t>(manager.GetPool())]);
+            manager.Close(address, cur_pages);
+        }
 
-    // Free all of the pages
-    for (const auto& it : page_list.Nodes()) {
-        const auto min_num_pages{std::min<size_t>(
-            it.GetNumPages(), (chosen_manager.GetEndAddress() - it.GetAddress()) / PageSize)};
-        chosen_manager.Free(it.GetAddress(), min_num_pages);
+        num_pages -= cur_pages;
+        address += cur_pages * PageSize;
     }
+}
 
-    return ResultSuccess;
+void KMemoryManager::Close(const KPageLinkedList& pg) {
+    for (const auto& node : pg.Nodes()) {
+        Close(node.GetAddress(), node.GetNumPages());
+    }
+}
+void KMemoryManager::Open(const KPageLinkedList& pg) {
+    for (const auto& node : pg.Nodes()) {
+        Open(node.GetAddress(), node.GetNumPages());
+    }
+}
+
+size_t KMemoryManager::Impl::Initialize(PAddr address, size_t size, VAddr management,
+                                        VAddr management_end, Pool p) {
+    // Calculate management sizes.
+    const size_t ref_count_size = (size / PageSize) * sizeof(u16);
+    const size_t optimize_map_size = CalculateOptimizedProcessOverheadSize(size);
+    const size_t manager_size = Common::AlignUp(optimize_map_size + ref_count_size, PageSize);
+    const size_t page_heap_size = KPageHeap::CalculateManagementOverheadSize(size);
+    const size_t total_management_size = manager_size + page_heap_size;
+    ASSERT(manager_size <= total_management_size);
+    ASSERT(management + total_management_size <= management_end);
+    ASSERT(Common::IsAligned(total_management_size, PageSize));
+
+    // Setup region.
+    pool = p;
+    management_region = management;
+    page_reference_counts.resize(
+        Kernel::Board::Nintendo::Nx::KSystemControl::Init::GetIntendedMemorySize() / PageSize);
+    ASSERT(Common::IsAligned(management_region, PageSize));
+
+    // Initialize the manager's KPageHeap.
+    heap.Initialize(address, size, management + manager_size, page_heap_size);
+
+    return total_management_size;
 }
 
-std::size_t KMemoryManager::Impl::CalculateManagementOverheadSize(std::size_t region_size) {
-    const std::size_t ref_count_size = (region_size / PageSize) * sizeof(u16);
-    const std::size_t optimize_map_size =
+size_t KMemoryManager::Impl::CalculateManagementOverheadSize(size_t region_size) {
+    const size_t ref_count_size = (region_size / PageSize) * sizeof(u16);
+    const size_t optimize_map_size =
         (Common::AlignUp((region_size / PageSize), Common::BitSize<u64>()) /
          Common::BitSize<u64>()) *
         sizeof(u64);
-    const std::size_t manager_meta_size =
-        Common::AlignUp(optimize_map_size + ref_count_size, PageSize);
-    const std::size_t page_heap_size = KPageHeap::CalculateManagementOverheadSize(region_size);
+    const size_t manager_meta_size = Common::AlignUp(optimize_map_size + ref_count_size, PageSize);
+    const size_t page_heap_size = KPageHeap::CalculateManagementOverheadSize(region_size);
     return manager_meta_size + page_heap_size;
 }
 
diff --git a/src/core/hle/kernel/k_memory_manager.h b/src/core/hle/kernel/k_memory_manager.h
index abd6c8ace..18775b262 100644
--- a/src/core/hle/kernel/k_memory_manager.h
+++ b/src/core/hle/kernel/k_memory_manager.h
@@ -5,10 +5,12 @@
 #pragma once
 
 #include <array>
-#include <mutex>
 #include <tuple>
 
+#include "common/common_funcs.h"
 #include "common/common_types.h"
+#include "core/hle/kernel/k_light_lock.h"
+#include "core/hle/kernel/k_memory_layout.h"
 #include "core/hle/kernel/k_page_heap.h"
 #include "core/hle/result.h"
 
@@ -20,8 +22,11 @@ namespace Kernel {
 
 class KPageLinkedList;
 
-class KMemoryManager final : NonCopyable {
+class KMemoryManager final {
 public:
+    YUZU_NON_COPYABLE(KMemoryManager);
+    YUZU_NON_MOVEABLE(KMemoryManager);
+
     enum class Pool : u32 {
         Application = 0,
         Applet = 1,
@@ -48,22 +53,33 @@ public:
 
     explicit KMemoryManager(Core::System& system_);
 
-    constexpr std::size_t GetSize(Pool pool) const {
-        return managers[static_cast<std::size_t>(pool)].GetSize();
+    void Initialize(VAddr management_region, size_t management_region_size);
+
+    constexpr size_t GetSize(Pool pool) const {
+        constexpr Direction GetSizeDirection = Direction::FromFront;
+        size_t total = 0;
+        for (auto* manager = this->GetFirstManager(pool, GetSizeDirection); manager != nullptr;
+             manager = this->GetNextManager(manager, GetSizeDirection)) {
+            total += manager->GetSize();
+        }
+        return total;
     }
 
-    void InitializeManager(Pool pool, u64 start_address, u64 end_address);
+    PAddr AllocateAndOpenContinuous(size_t num_pages, size_t align_pages, u32 option);
+    ResultCode AllocateAndOpen(KPageLinkedList* out, size_t num_pages, u32 option);
+    ResultCode AllocateAndOpenForProcess(KPageLinkedList* out, size_t num_pages, u32 option,
+                                         u64 process_id, u8 fill_pattern);
+
+    static constexpr size_t MaxManagerCount = 10;
 
-    VAddr AllocateAndOpenContinuous(size_t num_pages, size_t align_pages, u32 option);
-    ResultCode Allocate(KPageLinkedList& page_list, std::size_t num_pages, Pool pool, Direction dir,
-                        u32 heap_fill_value = 0);
-    ResultCode Free(KPageLinkedList& page_list, std::size_t num_pages, Pool pool, Direction dir,
-                    u32 heap_fill_value = 0);
+    void Close(PAddr address, size_t num_pages);
+    void Close(const KPageLinkedList& pg);
 
-    static constexpr std::size_t MaxManagerCount = 10;
+    void Open(PAddr address, size_t num_pages);
+    void Open(const KPageLinkedList& pg);
 
 public:
-    static std::size_t CalculateManagementOverheadSize(std::size_t region_size) {
+    static size_t CalculateManagementOverheadSize(size_t region_size) {
         return Impl::CalculateManagementOverheadSize(region_size);
     }
 
@@ -88,38 +104,34 @@ public:
     }
 
 private:
-    class Impl final : NonCopyable {
-    private:
-        using RefCount = u16;
-
-    private:
-        KPageHeap heap;
-        Pool pool{};
-
+    class Impl final {
     public:
-        static std::size_t CalculateManagementOverheadSize(std::size_t region_size);
-
-        static constexpr std::size_t CalculateOptimizedProcessOverheadSize(
-            std::size_t region_size) {
-            return (Common::AlignUp((region_size / PageSize), Common::BitSize<u64>()) /
-                    Common::BitSize<u64>()) *
-                   sizeof(u64);
-        }
+        YUZU_NON_COPYABLE(Impl);
+        YUZU_NON_MOVEABLE(Impl);
 
-    public:
         Impl() = default;
+        ~Impl() = default;
 
-        std::size_t Initialize(Pool new_pool, u64 start_address, u64 end_address);
+        size_t Initialize(PAddr address, size_t size, VAddr management, VAddr management_end,
+                          Pool p);
 
         VAddr AllocateBlock(s32 index, bool random) {
             return heap.AllocateBlock(index, random);
         }
 
-        void Free(VAddr addr, std::size_t num_pages) {
+        void Free(VAddr addr, size_t num_pages) {
             heap.Free(addr, num_pages);
         }
 
-        constexpr std::size_t GetSize() const {
+        void SetInitialUsedHeapSize(size_t reserved_size) {
+            heap.SetInitialUsedSize(reserved_size);
+        }
+
+        constexpr Pool GetPool() const {
+            return pool;
+        }
+
+        constexpr size_t GetSize() const {
             return heap.GetSize();
         }
 
@@ -130,12 +142,137 @@ private:
         constexpr VAddr GetEndAddress() const {
             return heap.GetEndAddress();
         }
+
+        constexpr size_t GetPageOffset(PAddr address) const {
+            return heap.GetPageOffset(address);
+        }
+
+        constexpr size_t GetPageOffsetToEnd(PAddr address) const {
+            return heap.GetPageOffsetToEnd(address);
+        }
+
+        constexpr void SetNext(Impl* n) {
+            next = n;
+        }
+
+        constexpr void SetPrev(Impl* n) {
+            prev = n;
+        }
+
+        constexpr Impl* GetNext() const {
+            return next;
+        }
+
+        constexpr Impl* GetPrev() const {
+            return prev;
+        }
+
+        void OpenFirst(PAddr address, size_t num_pages) {
+            size_t index = this->GetPageOffset(address);
+            const size_t end = index + num_pages;
+            while (index < end) {
+                const RefCount ref_count = (++page_reference_counts[index]);
+                ASSERT(ref_count == 1);
+
+                index++;
+            }
+        }
+
+        void Open(PAddr address, size_t num_pages) {
+            size_t index = this->GetPageOffset(address);
+            const size_t end = index + num_pages;
+            while (index < end) {
+                const RefCount ref_count = (++page_reference_counts[index]);
+                ASSERT(ref_count > 1);
+
+                index++;
+            }
+        }
+
+        void Close(PAddr address, size_t num_pages) {
+            size_t index = this->GetPageOffset(address);
+            const size_t end = index + num_pages;
+
+            size_t free_start = 0;
+            size_t free_count = 0;
+            while (index < end) {
+                ASSERT(page_reference_counts[index] > 0);
+                const RefCount ref_count = (--page_reference_counts[index]);
+
+                // Keep track of how many zero refcounts we see in a row, to minimize calls to free.
+                if (ref_count == 0) {
+                    if (free_count > 0) {
+                        free_count++;
+                    } else {
+                        free_start = index;
+                        free_count = 1;
+                    }
+                } else {
+                    if (free_count > 0) {
+                        this->Free(heap.GetAddress() + free_start * PageSize, free_count);
+                        free_count = 0;
+                    }
+                }
+
+                index++;
+            }
+
+            if (free_count > 0) {
+                this->Free(heap.GetAddress() + free_start * PageSize, free_count);
+            }
+        }
+
+        static size_t CalculateManagementOverheadSize(size_t region_size);
+
+        static constexpr size_t CalculateOptimizedProcessOverheadSize(size_t region_size) {
+            return (Common::AlignUp((region_size / PageSize), Common::BitSize<u64>()) /
+                    Common::BitSize<u64>()) *
+                   sizeof(u64);
+        }
+
+    private:
+        using RefCount = u16;
+
+        KPageHeap heap;
+        std::vector<RefCount> page_reference_counts;
+        VAddr management_region{};
+        Pool pool{};
+        Impl* next{};
+        Impl* prev{};
     };
 
 private:
+    Impl& GetManager(const KMemoryLayout& memory_layout, PAddr address) {
+        return managers[memory_layout.GetPhysicalLinearRegion(address).GetAttributes()];
+    }
+
+    const Impl& GetManager(const KMemoryLayout& memory_layout, PAddr address) const {
+        return managers[memory_layout.GetPhysicalLinearRegion(address).GetAttributes()];
+    }
+
+    constexpr Impl* GetFirstManager(Pool pool, Direction dir) const {
+        return dir == Direction::FromBack ? pool_managers_tail[static_cast<size_t>(pool)]
+                                          : pool_managers_head[static_cast<size_t>(pool)];
+    }
+
+    constexpr Impl* GetNextManager(Impl* cur, Direction dir) const {
+        if (dir == Direction::FromBack) {
+            return cur->GetPrev();
+        } else {
+            return cur->GetNext();
+        }
+    }
+
+    ResultCode AllocatePageGroupImpl(KPageLinkedList* out, size_t num_pages, Pool pool,
+                                     Direction dir, bool random);
+
+private:
     Core::System& system;
-    std::array<std::mutex, static_cast<std::size_t>(Pool::Count)> pool_locks;
+    std::array<KLightLock, static_cast<size_t>(Pool::Count)> pool_locks;
+    std::array<Impl*, MaxManagerCount> pool_managers_head{};
+    std::array<Impl*, MaxManagerCount> pool_managers_tail{};
     std::array<Impl, MaxManagerCount> managers;
+    size_t num_managers{};
 };
 
 } // namespace Kernel
diff --git a/src/core/hle/kernel/k_memory_region.h b/src/core/hle/kernel/k_memory_region.h
index 90ab8fd62..e9bdf4e59 100644
--- a/src/core/hle/kernel/k_memory_region.h
+++ b/src/core/hle/kernel/k_memory_region.h
@@ -5,6 +5,7 @@
 #pragma once
 
 #include "common/assert.h"
+#include "common/common_funcs.h"
 #include "common/common_types.h"
 #include "common/intrusive_red_black_tree.h"
 #include "core/hle/kernel/k_memory_region_type.h"
@@ -13,11 +14,13 @@ namespace Kernel {
 
 class KMemoryRegionAllocator;
 
-class KMemoryRegion final : public Common::IntrusiveRedBlackTreeBaseNode<KMemoryRegion>,
-                            NonCopyable {
+class KMemoryRegion final : public Common::IntrusiveRedBlackTreeBaseNode<KMemoryRegion> {
     friend class KMemoryRegionTree;
 
 public:
+    YUZU_NON_COPYABLE(KMemoryRegion);
+    YUZU_NON_MOVEABLE(KMemoryRegion);
+
     constexpr KMemoryRegion() = default;
     constexpr KMemoryRegion(u64 address_, u64 last_address_)
         : address{address_}, last_address{last_address_} {}
@@ -29,6 +32,8 @@ public:
         : KMemoryRegion(address_, last_address_, std::numeric_limits<u64>::max(), attributes_,
                         type_id_) {}
 
+    ~KMemoryRegion() = default;
+
     static constexpr int Compare(const KMemoryRegion& lhs, const KMemoryRegion& rhs) {
         if (lhs.GetAddress() < rhs.GetAddress()) {
             return -1;
@@ -39,16 +44,6 @@ public:
         }
     }
 
-private:
-    constexpr void Reset(u64 a, u64 la, u64 p, u32 r, u32 t) {
-        address = a;
-        pair_address = p;
-        last_address = la;
-        attributes = r;
-        type_id = t;
-    }
-
-public:
     constexpr u64 GetAddress() const {
         return address;
     }
@@ -108,6 +103,14 @@ public:
     }
 
 private:
+    constexpr void Reset(u64 a, u64 la, u64 p, u32 r, u32 t) {
+        address = a;
+        pair_address = p;
+        last_address = la;
+        attributes = r;
+        type_id = t;
+    }
+
     u64 address{};
     u64 last_address{};
     u64 pair_address{};
@@ -115,8 +118,25 @@ private:
     u32 type_id{};
 };
 
-class KMemoryRegionTree final : NonCopyable {
+class KMemoryRegionTree final {
+private:
+    using TreeType =
+        Common::IntrusiveRedBlackTreeBaseTraits<KMemoryRegion>::TreeType<KMemoryRegion>;
+
 public:
+    YUZU_NON_COPYABLE(KMemoryRegionTree);
+    YUZU_NON_MOVEABLE(KMemoryRegionTree);
+
+    using value_type = TreeType::value_type;
+    using size_type = TreeType::size_type;
+    using difference_type = TreeType::difference_type;
+    using pointer = TreeType::pointer;
+    using const_pointer = TreeType::const_pointer;
+    using reference = TreeType::reference;
+    using const_reference = TreeType::const_reference;
+    using iterator = TreeType::iterator;
+    using const_iterator = TreeType::const_iterator;
+
     struct DerivedRegionExtents {
         const KMemoryRegion* first_region{};
         const KMemoryRegion* last_region{};
@@ -140,29 +160,9 @@ public:
         }
     };
 
-private:
-    using TreeType =
-        Common::IntrusiveRedBlackTreeBaseTraits<KMemoryRegion>::TreeType<KMemoryRegion>;
-
-public:
-    using value_type = TreeType::value_type;
-    using size_type = TreeType::size_type;
-    using difference_type = TreeType::difference_type;
-    using pointer = TreeType::pointer;
-    using const_pointer = TreeType::const_pointer;
-    using reference = TreeType::reference;
-    using const_reference = TreeType::const_reference;
-    using iterator = TreeType::iterator;
-    using const_iterator = TreeType::const_iterator;
-
-private:
-    TreeType m_tree{};
-    KMemoryRegionAllocator& memory_region_allocator;
-
-public:
     explicit KMemoryRegionTree(KMemoryRegionAllocator& memory_region_allocator_);
+    ~KMemoryRegionTree() = default;
 
-public:
     KMemoryRegion* FindModifiable(u64 address) {
         if (auto it = this->find(KMemoryRegion(address, address, 0, 0)); it != this->end()) {
             return std::addressof(*it);
@@ -241,7 +241,6 @@ public:
         return GetDerivedRegionExtents(static_cast<KMemoryRegionType>(type_id));
     }
 
-public:
     void InsertDirectly(u64 address, u64 last_address, u32 attr = 0, u32 type_id = 0);
     bool Insert(u64 address, size_t size, u32 type_id, u32 new_attr = 0, u32 old_attr = 0);
 
@@ -252,7 +251,6 @@ public:
         return this->GetRandomAlignedRegion(size + 2 * guard_size, alignment, type_id) + guard_size;
     }
 
-public:
     // Iterator accessors.
     iterator begin() {
         return m_tree.begin();
@@ -322,13 +320,21 @@ public:
     iterator nfind(const_reference ref) const {
         return m_tree.nfind(ref);
     }
+
+private:
+    TreeType m_tree{};
+    KMemoryRegionAllocator& memory_region_allocator;
 };
 
-class KMemoryRegionAllocator final : NonCopyable {
+class KMemoryRegionAllocator final {
 public:
+    YUZU_NON_COPYABLE(KMemoryRegionAllocator);
+    YUZU_NON_MOVEABLE(KMemoryRegionAllocator);
+
     static constexpr size_t MaxMemoryRegions = 200;
 
     constexpr KMemoryRegionAllocator() = default;
+    constexpr ~KMemoryRegionAllocator() = default;
 
     template <typename... Args>
     KMemoryRegion* Allocate(Args&&... args) {
diff --git a/src/core/hle/kernel/k_memory_region_type.h b/src/core/hle/kernel/k_memory_region_type.h
index a05e66677..0baeddf51 100644
--- a/src/core/hle/kernel/k_memory_region_type.h
+++ b/src/core/hle/kernel/k_memory_region_type.h
@@ -14,7 +14,8 @@
 namespace Kernel {
 
 enum KMemoryRegionType : u32 {
-    KMemoryRegionAttr_CarveoutProtected = 0x04000000,
+    KMemoryRegionAttr_CarveoutProtected = 0x02000000,
+    KMemoryRegionAttr_Uncached = 0x04000000,
     KMemoryRegionAttr_DidKernelMap = 0x08000000,
     KMemoryRegionAttr_ShouldKernelMap = 0x10000000,
     KMemoryRegionAttr_UserReadOnly = 0x20000000,
@@ -239,6 +240,11 @@ static_assert(KMemoryRegionType_VirtualDramHeapBase.GetValue() == 0x1A);
 static_assert(KMemoryRegionType_VirtualDramKernelPtHeap.GetValue() == 0x2A);
 static_assert(KMemoryRegionType_VirtualDramKernelTraceBuffer.GetValue() == 0x4A);
 
+// UNUSED: .DeriveSparse(2, 2, 0);
+constexpr auto KMemoryRegionType_VirtualDramUnknownDebug =
+    KMemoryRegionType_Dram.DeriveSparse(2, 2, 1);
+static_assert(KMemoryRegionType_VirtualDramUnknownDebug.GetValue() == (0x52));
+
 constexpr auto KMemoryRegionType_VirtualDramKernelInitPt =
     KMemoryRegionType_VirtualDramHeapBase.Derive(3, 0);
 constexpr auto KMemoryRegionType_VirtualDramPoolManagement =
@@ -330,6 +336,8 @@ constexpr KMemoryRegionType GetTypeForVirtualLinearMapping(u32 type_id) {
         return KMemoryRegionType_VirtualDramKernelTraceBuffer;
     } else if (KMemoryRegionType_DramKernelPtHeap.IsAncestorOf(type_id)) {
         return KMemoryRegionType_VirtualDramKernelPtHeap;
+    } else if ((type_id | KMemoryRegionAttr_ShouldKernelMap) == type_id) {
+        return KMemoryRegionType_VirtualDramUnknownDebug;
     } else {
         return KMemoryRegionType_Dram;
     }
diff --git a/src/core/hle/kernel/k_page_buffer.cpp b/src/core/hle/kernel/k_page_buffer.cpp
new file mode 100644
index 000000000..f7df4a9a8
--- /dev/null
+++ b/src/core/hle/kernel/k_page_buffer.cpp
@@ -0,0 +1,19 @@
+// Copyright 2022 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include "common/alignment.h"
+#include "common/assert.h"
+#include "core/core.h"
+#include "core/device_memory.h"
+#include "core/hle/kernel/k_page_buffer.h"
+#include "core/hle/kernel/memory_types.h"
+
+namespace Kernel {
+
+KPageBuffer* KPageBuffer::FromPhysicalAddress(Core::System& system, PAddr phys_addr) {
+    ASSERT(Common::IsAligned(phys_addr, PageSize));
+    return reinterpret_cast<KPageBuffer*>(system.DeviceMemory().GetPointer(phys_addr));
+}
+
+} // namespace Kernel
diff --git a/src/core/hle/kernel/k_page_buffer.h b/src/core/hle/kernel/k_page_buffer.h
new file mode 100644
index 000000000..6ff3c1568
--- /dev/null
+++ b/src/core/hle/kernel/k_page_buffer.h
@@ -0,0 +1,29 @@
+// Copyright 2022 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include <array>
+
+#include "common/common_types.h"
+#include "core/device_memory.h"
+#include "core/hle/kernel/memory_types.h"
+#include "core/hle/kernel/slab_helpers.h"
+
+namespace Kernel {
+
+class KPageBuffer final : public KSlabAllocated<KPageBuffer> {
+public:
+    KPageBuffer() = default;
+
+    static KPageBuffer* FromPhysicalAddress(Core::System& system, PAddr phys_addr);
+
+private:
+    [[maybe_unused]] alignas(PageSize) std::array<u8, PageSize> m_buffer{};
+};
+
+static_assert(sizeof(KPageBuffer) == PageSize);
+static_assert(alignof(KPageBuffer) == PageSize);
+
+} // namespace Kernel
diff --git a/src/core/hle/kernel/k_page_heap.cpp b/src/core/hle/kernel/k_page_heap.cpp
index 29d996d62..97a5890a0 100644
--- a/src/core/hle/kernel/k_page_heap.cpp
+++ b/src/core/hle/kernel/k_page_heap.cpp
@@ -7,35 +7,51 @@
 
 namespace Kernel {
 
-void KPageHeap::Initialize(VAddr address, std::size_t size, std::size_t metadata_size) {
-    // Check our assumptions
-    ASSERT(Common::IsAligned((address), PageSize));
+void KPageHeap::Initialize(PAddr address, size_t size, VAddr management_address,
+                           size_t management_size, const size_t* block_shifts,
+                           size_t num_block_shifts) {
+    // Check our assumptions.
+    ASSERT(Common::IsAligned(address, PageSize));
     ASSERT(Common::IsAligned(size, PageSize));
+    ASSERT(0 < num_block_shifts && num_block_shifts <= NumMemoryBlockPageShifts);
+    const VAddr management_end = management_address + management_size;
 
-    // Set our members
-    heap_address = address;
-    heap_size = size;
-
-    // Setup bitmaps
-    metadata.resize(metadata_size / sizeof(u64));
-    u64* cur_bitmap_storage{metadata.data()};
-    for (std::size_t i = 0; i < MemoryBlockPageShifts.size(); i++) {
-        const std::size_t cur_block_shift{MemoryBlockPageShifts[i]};
-        const std::size_t next_block_shift{
-            (i != MemoryBlockPageShifts.size() - 1) ? MemoryBlockPageShifts[i + 1] : 0};
-        cur_bitmap_storage = blocks[i].Initialize(heap_address, heap_size, cur_block_shift,
-                                                  next_block_shift, cur_bitmap_storage);
+    // Set our members.
+    m_heap_address = address;
+    m_heap_size = size;
+    m_num_blocks = num_block_shifts;
+
+    // Setup bitmaps.
+    m_management_data.resize(management_size / sizeof(u64));
+    u64* cur_bitmap_storage{m_management_data.data()};
+    for (size_t i = 0; i < num_block_shifts; i++) {
+        const size_t cur_block_shift = block_shifts[i];
+        const size_t next_block_shift = (i != num_block_shifts - 1) ? block_shifts[i + 1] : 0;
+        cur_bitmap_storage = m_blocks[i].Initialize(m_heap_address, m_heap_size, cur_block_shift,
+                                                    next_block_shift, cur_bitmap_storage);
     }
+
+    // Ensure we didn't overextend our bounds.
+    ASSERT(VAddr(cur_bitmap_storage) <= management_end);
+}
+
+size_t KPageHeap::GetNumFreePages() const {
+    size_t num_free = 0;
+
+    for (size_t i = 0; i < m_num_blocks; i++) {
+        num_free += m_blocks[i].GetNumFreePages();
+    }
+
+    return num_free;
 }
 
-VAddr KPageHeap::AllocateBlock(s32 index, bool random) {
-    const std::size_t needed_size{blocks[index].GetSize()};
+PAddr KPageHeap::AllocateBlock(s32 index, bool random) {
+    const size_t needed_size = m_blocks[index].GetSize();
 
-    for (s32 i{index}; i < static_cast<s32>(MemoryBlockPageShifts.size()); i++) {
-        if (const VAddr addr{blocks[i].PopBlock(random)}; addr) {
-            if (const std::size_t allocated_size{blocks[i].GetSize()};
-                allocated_size > needed_size) {
-                Free(addr + needed_size, (allocated_size - needed_size) / PageSize);
+    for (s32 i = index; i < static_cast<s32>(m_num_blocks); i++) {
+        if (const PAddr addr = m_blocks[i].PopBlock(random); addr != 0) {
+            if (const size_t allocated_size = m_blocks[i].GetSize(); allocated_size > needed_size) {
+                this->Free(addr + needed_size, (allocated_size - needed_size) / PageSize);
             }
             return addr;
         }
@@ -44,34 +60,34 @@ VAddr KPageHeap::AllocateBlock(s32 index, bool random) {
     return 0;
 }
 
-void KPageHeap::FreeBlock(VAddr block, s32 index) {
+void KPageHeap::FreeBlock(PAddr block, s32 index) {
     do {
-        block = blocks[index++].PushBlock(block);
+        block = m_blocks[index++].PushBlock(block);
     } while (block != 0);
 }
 
-void KPageHeap::Free(VAddr addr, std::size_t num_pages) {
-    // Freeing no pages is a no-op
+void KPageHeap::Free(PAddr addr, size_t num_pages) {
+    // Freeing no pages is a no-op.
     if (num_pages == 0) {
         return;
     }
 
-    // Find the largest block size that we can free, and free as many as possible
-    s32 big_index{static_cast<s32>(MemoryBlockPageShifts.size()) - 1};
-    const VAddr start{addr};
-    const VAddr end{(num_pages * PageSize) + addr};
-    VAddr before_start{start};
-    VAddr before_end{start};
-    VAddr after_start{end};
-    VAddr after_end{end};
+    // Find the largest block size that we can free, and free as many as possible.
+    s32 big_index = static_cast<s32>(m_num_blocks) - 1;
+    const PAddr start = addr;
+    const PAddr end = addr + num_pages * PageSize;
+    PAddr before_start = start;
+    PAddr before_end = start;
+    PAddr after_start = end;
+    PAddr after_end = end;
     while (big_index >= 0) {
-        const std::size_t block_size{blocks[big_index].GetSize()};
-        const VAddr big_start{Common::AlignUp((start), block_size)};
-        const VAddr big_end{Common::AlignDown((end), block_size)};
+        const size_t block_size = m_blocks[big_index].GetSize();
+        const PAddr big_start = Common::AlignUp(start, block_size);
+        const PAddr big_end = Common::AlignDown(end, block_size);
         if (big_start < big_end) {
-            // Free as many big blocks as we can
-            for (auto block{big_start}; block < big_end; block += block_size) {
-                FreeBlock(block, big_index);
+            // Free as many big blocks as we can.
+            for (auto block = big_start; block < big_end; block += block_size) {
+                this->FreeBlock(block, big_index);
             }
             before_end = big_start;
             after_start = big_end;
@@ -81,31 +97,31 @@ void KPageHeap::Free(VAddr addr, std::size_t num_pages) {
     }
     ASSERT(big_index >= 0);
 
-    // Free space before the big blocks
-    for (s32 i{big_index - 1}; i >= 0; i--) {
-        const std::size_t block_size{blocks[i].GetSize()};
+    // Free space before the big blocks.
+    for (s32 i = big_index - 1; i >= 0; i--) {
+        const size_t block_size = m_blocks[i].GetSize();
         while (before_start + block_size <= before_end) {
             before_end -= block_size;
-            FreeBlock(before_end, i);
+            this->FreeBlock(before_end, i);
         }
     }
 
-    // Free space after the big blocks
-    for (s32 i{big_index - 1}; i >= 0; i--) {
-        const std::size_t block_size{blocks[i].GetSize()};
+    // Free space after the big blocks.
+    for (s32 i = big_index - 1; i >= 0; i--) {
+        const size_t block_size = m_blocks[i].GetSize();
         while (after_start + block_size <= after_end) {
-            FreeBlock(after_start, i);
+            this->FreeBlock(after_start, i);
             after_start += block_size;
         }
     }
 }
 
-std::size_t KPageHeap::CalculateManagementOverheadSize(std::size_t region_size) {
-    std::size_t overhead_size = 0;
-    for (std::size_t i = 0; i < MemoryBlockPageShifts.size(); i++) {
-        const std::size_t cur_block_shift{MemoryBlockPageShifts[i]};
-        const std::size_t next_block_shift{
-            (i != MemoryBlockPageShifts.size() - 1) ? MemoryBlockPageShifts[i + 1] : 0};
+size_t KPageHeap::CalculateManagementOverheadSize(size_t region_size, const size_t* block_shifts,
+                                                  size_t num_block_shifts) {
+    size_t overhead_size = 0;
+    for (size_t i = 0; i < num_block_shifts; i++) {
+        const size_t cur_block_shift = block_shifts[i];
+        const size_t next_block_shift = (i != num_block_shifts - 1) ? block_shifts[i + 1] : 0;
         overhead_size += KPageHeap::Block::CalculateManagementOverheadSize(
             region_size, cur_block_shift, next_block_shift);
     }
diff --git a/src/core/hle/kernel/k_page_heap.h b/src/core/hle/kernel/k_page_heap.h
index 8d9f30523..60fff766b 100644
--- a/src/core/hle/kernel/k_page_heap.h
+++ b/src/core/hle/kernel/k_page_heap.h
@@ -8,183 +8,210 @@
 #include <vector>
 
 #include "common/alignment.h"
+#include "common/common_funcs.h"
 #include "common/common_types.h"
 #include "core/hle/kernel/k_page_bitmap.h"
 #include "core/hle/kernel/memory_types.h"
 
 namespace Kernel {
 
-class KPageHeap final : NonCopyable {
+class KPageHeap final {
 public:
-    static constexpr s32 GetAlignedBlockIndex(std::size_t num_pages, std::size_t align_pages) {
-        const auto target_pages{std::max(num_pages, align_pages)};
-        for (std::size_t i = 0; i < NumMemoryBlockPageShifts; i++) {
-            if (target_pages <=
-                (static_cast<std::size_t>(1) << MemoryBlockPageShifts[i]) / PageSize) {
+    YUZU_NON_COPYABLE(KPageHeap);
+    YUZU_NON_MOVEABLE(KPageHeap);
+
+    KPageHeap() = default;
+    ~KPageHeap() = default;
+
+    constexpr PAddr GetAddress() const {
+        return m_heap_address;
+    }
+    constexpr size_t GetSize() const {
+        return m_heap_size;
+    }
+    constexpr PAddr GetEndAddress() const {
+        return this->GetAddress() + this->GetSize();
+    }
+    constexpr size_t GetPageOffset(PAddr block) const {
+        return (block - this->GetAddress()) / PageSize;
+    }
+    constexpr size_t GetPageOffsetToEnd(PAddr block) const {
+        return (this->GetEndAddress() - block) / PageSize;
+    }
+
+    void Initialize(PAddr heap_address, size_t heap_size, VAddr management_address,
+                    size_t management_size) {
+        return this->Initialize(heap_address, heap_size, management_address, management_size,
+                                MemoryBlockPageShifts.data(), NumMemoryBlockPageShifts);
+    }
+
+    size_t GetFreeSize() const {
+        return this->GetNumFreePages() * PageSize;
+    }
+
+    void SetInitialUsedSize(size_t reserved_size) {
+        // Check that the reserved size is valid.
+        const size_t free_size = this->GetNumFreePages() * PageSize;
+        ASSERT(m_heap_size >= free_size + reserved_size);
+
+        // Set the initial used size.
+        m_initial_used_size = m_heap_size - free_size - reserved_size;
+    }
+
+    PAddr AllocateBlock(s32 index, bool random);
+    void Free(PAddr addr, size_t num_pages);
+
+    static size_t CalculateManagementOverheadSize(size_t region_size) {
+        return CalculateManagementOverheadSize(region_size, MemoryBlockPageShifts.data(),
+                                               NumMemoryBlockPageShifts);
+    }
+
+    static constexpr s32 GetAlignedBlockIndex(size_t num_pages, size_t align_pages) {
+        const size_t target_pages = std::max(num_pages, align_pages);
+        for (size_t i = 0; i < NumMemoryBlockPageShifts; i++) {
+            if (target_pages <= (size_t(1) << MemoryBlockPageShifts[i]) / PageSize) {
                 return static_cast<s32>(i);
             }
         }
         return -1;
     }
 
-    static constexpr s32 GetBlockIndex(std::size_t num_pages) {
-        for (s32 i{static_cast<s32>(NumMemoryBlockPageShifts) - 1}; i >= 0; i--) {
-            if (num_pages >= (static_cast<std::size_t>(1) << MemoryBlockPageShifts[i]) / PageSize) {
+    static constexpr s32 GetBlockIndex(size_t num_pages) {
+        for (s32 i = static_cast<s32>(NumMemoryBlockPageShifts) - 1; i >= 0; i--) {
+            if (num_pages >= (size_t(1) << MemoryBlockPageShifts[i]) / PageSize) {
                 return i;
             }
         }
         return -1;
     }
 
-    static constexpr std::size_t GetBlockSize(std::size_t index) {
-        return static_cast<std::size_t>(1) << MemoryBlockPageShifts[index];
+    static constexpr size_t GetBlockSize(size_t index) {
+        return size_t(1) << MemoryBlockPageShifts[index];
     }
 
-    static constexpr std::size_t GetBlockNumPages(std::size_t index) {
+    static constexpr size_t GetBlockNumPages(size_t index) {
         return GetBlockSize(index) / PageSize;
     }
 
 private:
-    static constexpr std::size_t NumMemoryBlockPageShifts{7};
-    static constexpr std::array<std::size_t, NumMemoryBlockPageShifts> MemoryBlockPageShifts{
-        0xC, 0x10, 0x15, 0x16, 0x19, 0x1D, 0x1E,
-    };
-
-    class Block final : NonCopyable {
-    private:
-        KPageBitmap bitmap;
-        VAddr heap_address{};
-        uintptr_t end_offset{};
-        std::size_t block_shift{};
-        std::size_t next_block_shift{};
-
+    class Block final {
     public:
+        YUZU_NON_COPYABLE(Block);
+        YUZU_NON_MOVEABLE(Block);
+
         Block() = default;
+        ~Block() = default;
 
-        constexpr std::size_t GetShift() const {
-            return block_shift;
+        constexpr size_t GetShift() const {
+            return m_block_shift;
         }
-        constexpr std::size_t GetNextShift() const {
-            return next_block_shift;
+        constexpr size_t GetNextShift() const {
+            return m_next_block_shift;
         }
-        constexpr std::size_t GetSize() const {
-            return static_cast<std::size_t>(1) << GetShift();
+        constexpr size_t GetSize() const {
+            return u64(1) << this->GetShift();
         }
-        constexpr std::size_t GetNumPages() const {
-            return GetSize() / PageSize;
+        constexpr size_t GetNumPages() const {
+            return this->GetSize() / PageSize;
         }
-        constexpr std::size_t GetNumFreeBlocks() const {
-            return bitmap.GetNumBits();
+        constexpr size_t GetNumFreeBlocks() const {
+            return m_bitmap.GetNumBits();
         }
-        constexpr std::size_t GetNumFreePages() const {
-            return GetNumFreeBlocks() * GetNumPages();
+        constexpr size_t GetNumFreePages() const {
+            return this->GetNumFreeBlocks() * this->GetNumPages();
         }
 
-        u64* Initialize(VAddr addr, std::size_t size, std::size_t bs, std::size_t nbs,
-                        u64* bit_storage) {
-            // Set shifts
-            block_shift = bs;
-            next_block_shift = nbs;
-
-            // Align up the address
-            VAddr end{addr + size};
-            const auto align{(next_block_shift != 0) ? (1ULL << next_block_shift)
-                                                     : (1ULL << block_shift)};
-            addr = Common::AlignDown((addr), align);
-            end = Common::AlignUp((end), align);
-
-            heap_address = addr;
-            end_offset = (end - addr) / (1ULL << block_shift);
-            return bitmap.Initialize(bit_storage, end_offset);
+        u64* Initialize(PAddr addr, size_t size, size_t bs, size_t nbs, u64* bit_storage) {
+            // Set shifts.
+            m_block_shift = bs;
+            m_next_block_shift = nbs;
+
+            // Align up the address.
+            PAddr end = addr + size;
+            const size_t align = (m_next_block_shift != 0) ? (u64(1) << m_next_block_shift)
+                                                           : (u64(1) << m_block_shift);
+            addr = Common::AlignDown(addr, align);
+            end = Common::AlignUp(end, align);
+
+            m_heap_address = addr;
+            m_end_offset = (end - addr) / (u64(1) << m_block_shift);
+            return m_bitmap.Initialize(bit_storage, m_end_offset);
         }
 
-        VAddr PushBlock(VAddr address) {
-            // Set the bit for the free block
-            std::size_t offset{(address - heap_address) >> GetShift()};
-            bitmap.SetBit(offset);
+        PAddr PushBlock(PAddr address) {
+            // Set the bit for the free block.
+            size_t offset = (address - m_heap_address) >> this->GetShift();
+            m_bitmap.SetBit(offset);
 
-            // If we have a next shift, try to clear the blocks below and return the address
-            if (GetNextShift()) {
-                const auto diff{1ULL << (GetNextShift() - GetShift())};
+            // If we have a next shift, try to clear the blocks below this one and return the new
+            // address.
+            if (this->GetNextShift()) {
+                const size_t diff = u64(1) << (this->GetNextShift() - this->GetShift());
                 offset = Common::AlignDown(offset, diff);
-                if (bitmap.ClearRange(offset, diff)) {
-                    return heap_address + (offset << GetShift());
+                if (m_bitmap.ClearRange(offset, diff)) {
+                    return m_heap_address + (offset << this->GetShift());
                 }
             }
 
-            // We couldn't coalesce, or we're already as big as possible
-            return 0;
+            // We couldn't coalesce, or we're already as big as possible.
+            return {};
         }
 
-        VAddr PopBlock(bool random) {
-            // Find a free block
-            const s64 soffset{bitmap.FindFreeBlock(random)};
+        PAddr PopBlock(bool random) {
+            // Find a free block.
+            s64 soffset = m_bitmap.FindFreeBlock(random);
             if (soffset < 0) {
-                return 0;
+                return {};
             }
-            const auto offset{static_cast<std::size_t>(soffset)};
+            const size_t offset = static_cast<size_t>(soffset);
 
-            // Update our tracking and return it
-            bitmap.ClearBit(offset);
-            return heap_address + (offset << GetShift());
+            // Update our tracking and return it.
+            m_bitmap.ClearBit(offset);
+            return m_heap_address + (offset << this->GetShift());
         }
 
     public:
-        static constexpr std::size_t CalculateManagementOverheadSize(std::size_t region_size,
-                                                                     std::size_t cur_block_shift,
-                                                                     std::size_t next_block_shift) {
-            const auto cur_block_size{(1ULL << cur_block_shift)};
-            const auto next_block_size{(1ULL << next_block_shift)};
-            const auto align{(next_block_shift != 0) ? next_block_size : cur_block_size};
+        static constexpr size_t CalculateManagementOverheadSize(size_t region_size,
+                                                                size_t cur_block_shift,
+                                                                size_t next_block_shift) {
+            const size_t cur_block_size = (u64(1) << cur_block_shift);
+            const size_t next_block_size = (u64(1) << next_block_shift);
+            const size_t align = (next_block_shift != 0) ? next_block_size : cur_block_size;
             return KPageBitmap::CalculateManagementOverheadSize(
                 (align * 2 + Common::AlignUp(region_size, align)) / cur_block_size);
         }
-    };
-
-public:
-    KPageHeap() = default;
-
-    constexpr VAddr GetAddress() const {
-        return heap_address;
-    }
-    constexpr std::size_t GetSize() const {
-        return heap_size;
-    }
-    constexpr VAddr GetEndAddress() const {
-        return GetAddress() + GetSize();
-    }
-    constexpr std::size_t GetPageOffset(VAddr block) const {
-        return (block - GetAddress()) / PageSize;
-    }
 
-    void Initialize(VAddr heap_address, std::size_t heap_size, std::size_t metadata_size);
-    VAddr AllocateBlock(s32 index, bool random);
-    void Free(VAddr addr, std::size_t num_pages);
-
-    void UpdateUsedSize() {
-        used_size = heap_size - (GetNumFreePages() * PageSize);
-    }
-
-    static std::size_t CalculateManagementOverheadSize(std::size_t region_size);
+    private:
+        KPageBitmap m_bitmap;
+        PAddr m_heap_address{};
+        uintptr_t m_end_offset{};
+        size_t m_block_shift{};
+        size_t m_next_block_shift{};
+    };
 
 private:
-    constexpr std::size_t GetNumFreePages() const {
-        std::size_t num_free{};
+    void Initialize(PAddr heap_address, size_t heap_size, VAddr management_address,
+                    size_t management_size, const size_t* block_shifts, size_t num_block_shifts);
+    size_t GetNumFreePages() const;
 
-        for (const auto& block : blocks) {
-            num_free += block.GetNumFreePages();
-        }
+    void FreeBlock(PAddr block, s32 index);
 
-        return num_free;
-    }
+    static constexpr size_t NumMemoryBlockPageShifts{7};
+    static constexpr std::array<size_t, NumMemoryBlockPageShifts> MemoryBlockPageShifts{
+        0xC, 0x10, 0x15, 0x16, 0x19, 0x1D, 0x1E,
+    };
 
-    void FreeBlock(VAddr block, s32 index);
+private:
+    static size_t CalculateManagementOverheadSize(size_t region_size, const size_t* block_shifts,
+                                                  size_t num_block_shifts);
 
-    VAddr heap_address{};
-    std::size_t heap_size{};
-    std::size_t used_size{};
-    std::array<Block, NumMemoryBlockPageShifts> blocks{};
-    std::vector<u64> metadata;
+private:
+    PAddr m_heap_address{};
+    size_t m_heap_size{};
+    size_t m_initial_used_size{};
+    size_t m_num_blocks{};
+    std::array<Block, NumMemoryBlockPageShifts> m_blocks{};
+    std::vector<u64> m_management_data;
 };
 
 } // namespace Kernel
diff --git a/src/core/hle/kernel/k_page_table.cpp b/src/core/hle/kernel/k_page_table.cpp
index 2ebbc0819..02d93b12e 100644
--- a/src/core/hle/kernel/k_page_table.cpp
+++ b/src/core/hle/kernel/k_page_table.cpp
@@ -41,27 +41,12 @@ constexpr std::size_t GetAddressSpaceWidthFromType(FileSys::ProgramAddressSpaceT
     }
 }
 
-constexpr u64 GetAddressInRange(const KMemoryInfo& info, VAddr addr) {
-    if (info.GetAddress() < addr) {
-        return addr;
-    }
-    return info.GetAddress();
-}
-
-constexpr std::size_t GetSizeInRange(const KMemoryInfo& info, VAddr start, VAddr end) {
-    std::size_t size{info.GetSize()};
-    if (info.GetAddress() < start) {
-        size -= start - info.GetAddress();
-    }
-    if (info.GetEndAddress() > end) {
-        size -= info.GetEndAddress() - end;
-    }
-    return size;
-}
-
 } // namespace
 
-KPageTable::KPageTable(Core::System& system_) : system{system_} {}
+KPageTable::KPageTable(Core::System& system_)
+    : general_lock{system_.Kernel()}, map_physical_memory_lock{system_.Kernel()}, system{system_} {}
+
+KPageTable::~KPageTable() = default;
 
 ResultCode KPageTable::InitializeForProcess(FileSys::ProgramAddressSpaceType as_type,
                                             bool enable_aslr, VAddr code_addr,
@@ -282,96 +267,228 @@ ResultCode KPageTable::MapProcessCode(VAddr addr, std::size_t num_pages, KMemory
     R_UNLESS(this->CanContain(addr, size, state), ResultInvalidCurrentMemory);
 
     // Lock the table.
-    std::lock_guard lock{page_table_lock};
+    KScopedLightLock lk(general_lock);
 
     // Verify that the destination memory is unmapped.
     R_TRY(this->CheckMemoryState(addr, size, KMemoryState::All, KMemoryState::Free,
                                  KMemoryPermission::None, KMemoryPermission::None,
                                  KMemoryAttribute::None, KMemoryAttribute::None));
+    KPageLinkedList pg;
+    R_TRY(system.Kernel().MemoryManager().AllocateAndOpen(
+        &pg, num_pages,
+        KMemoryManager::EncodeOption(KMemoryManager::Pool::Application, allocation_option)));
 
-    KPageLinkedList page_linked_list;
-    R_TRY(system.Kernel().MemoryManager().Allocate(page_linked_list, num_pages, memory_pool,
-                                                   allocation_option));
-    R_TRY(Operate(addr, num_pages, page_linked_list, OperationType::MapGroup));
+    R_TRY(Operate(addr, num_pages, pg, OperationType::MapGroup));
 
     block_manager->Update(addr, num_pages, state, perm);
 
     return ResultSuccess;
 }
 
-ResultCode KPageTable::MapCodeMemory(VAddr dst_addr, VAddr src_addr, std::size_t size) {
-    std::lock_guard lock{page_table_lock};
-
-    const std::size_t num_pages{size / PageSize};
+ResultCode KPageTable::MapCodeMemory(VAddr dst_address, VAddr src_address, std::size_t size) {
+    // Validate the mapping request.
+    R_UNLESS(this->CanContain(dst_address, size, KMemoryState::AliasCode),
+             ResultInvalidMemoryRegion);
 
-    KMemoryState state{};
-    KMemoryPermission perm{};
-    CASCADE_CODE(CheckMemoryState(&state, &perm, nullptr, nullptr, src_addr, size,
-                                  KMemoryState::All, KMemoryState::Normal, KMemoryPermission::All,
-                                  KMemoryPermission::UserReadWrite, KMemoryAttribute::Mask,
-                                  KMemoryAttribute::None, KMemoryAttribute::IpcAndDeviceMapped));
+    // Lock the table.
+    KScopedLightLock lk(general_lock);
 
-    if (IsRegionMapped(dst_addr, size)) {
-        return ResultInvalidCurrentMemory;
-    }
+    // Verify that the source memory is normal heap.
+    KMemoryState src_state{};
+    KMemoryPermission src_perm{};
+    std::size_t num_src_allocator_blocks{};
+    R_TRY(this->CheckMemoryState(&src_state, &src_perm, nullptr, &num_src_allocator_blocks,
+                                 src_address, size, KMemoryState::All, KMemoryState::Normal,
+                                 KMemoryPermission::All, KMemoryPermission::UserReadWrite,
+                                 KMemoryAttribute::All, KMemoryAttribute::None));
 
-    KPageLinkedList page_linked_list;
-    AddRegionToPages(src_addr, num_pages, page_linked_list);
+    // Verify that the destination memory is unmapped.
+    std::size_t num_dst_allocator_blocks{};
+    R_TRY(this->CheckMemoryState(&num_dst_allocator_blocks, dst_address, size, KMemoryState::All,
+                                 KMemoryState::Free, KMemoryPermission::None,
+                                 KMemoryPermission::None, KMemoryAttribute::None,
+                                 KMemoryAttribute::None));
 
+    // Map the code memory.
     {
-        auto block_guard = detail::ScopeExit(
-            [&] { Operate(src_addr, num_pages, perm, OperationType::ChangePermissions); });
+        // Determine the number of pages being operated on.
+        const std::size_t num_pages = size / PageSize;
 
-        CASCADE_CODE(Operate(src_addr, num_pages, KMemoryPermission::None,
-                             OperationType::ChangePermissions));
-        CASCADE_CODE(MapPages(dst_addr, page_linked_list, KMemoryPermission::None));
+        // Create page groups for the memory being mapped.
+        KPageLinkedList pg;
+        AddRegionToPages(src_address, num_pages, pg);
 
-        block_guard.Cancel();
-    }
+        // Reprotect the source as kernel-read/not mapped.
+        const auto new_perm = static_cast<KMemoryPermission>(KMemoryPermission::KernelRead |
+                                                             KMemoryPermission::NotMapped);
+        R_TRY(Operate(src_address, num_pages, new_perm, OperationType::ChangePermissions));
 
-    block_manager->Update(src_addr, num_pages, state, KMemoryPermission::None,
-                          KMemoryAttribute::Locked);
-    block_manager->Update(dst_addr, num_pages, KMemoryState::AliasCode);
+        // Ensure that we unprotect the source pages on failure.
+        auto unprot_guard = SCOPE_GUARD({
+            ASSERT(this->Operate(src_address, num_pages, src_perm, OperationType::ChangePermissions)
+                       .IsSuccess());
+        });
+
+        // Map the alias pages.
+        R_TRY(MapPages(dst_address, pg, new_perm));
+
+        // We successfully mapped the alias pages, so we don't need to unprotect the src pages on
+        // failure.
+        unprot_guard.Cancel();
+
+        // Apply the memory block updates.
+        block_manager->Update(src_address, num_pages, src_state, new_perm,
+                              KMemoryAttribute::Locked);
+        block_manager->Update(dst_address, num_pages, KMemoryState::AliasCode, new_perm,
+                              KMemoryAttribute::None);
+    }
 
     return ResultSuccess;
 }
 
-ResultCode KPageTable::UnmapCodeMemory(VAddr dst_addr, VAddr src_addr, std::size_t size) {
-    std::lock_guard lock{page_table_lock};
+ResultCode KPageTable::UnmapCodeMemory(VAddr dst_address, VAddr src_address, std::size_t size) {
+    // Validate the mapping request.
+    R_UNLESS(this->CanContain(dst_address, size, KMemoryState::AliasCode),
+             ResultInvalidMemoryRegion);
+
+    // Lock the table.
+    KScopedLightLock lk(general_lock);
 
-    if (!size) {
-        return ResultSuccess;
+    // Verify that the source memory is locked normal heap.
+    std::size_t num_src_allocator_blocks{};
+    R_TRY(this->CheckMemoryState(std::addressof(num_src_allocator_blocks), src_address, size,
+                                 KMemoryState::All, KMemoryState::Normal, KMemoryPermission::None,
+                                 KMemoryPermission::None, KMemoryAttribute::All,
+                                 KMemoryAttribute::Locked));
+
+    // Verify that the destination memory is aliasable code.
+    std::size_t num_dst_allocator_blocks{};
+    R_TRY(this->CheckMemoryStateContiguous(
+        std::addressof(num_dst_allocator_blocks), dst_address, size, KMemoryState::FlagCanCodeAlias,
+        KMemoryState::FlagCanCodeAlias, KMemoryPermission::None, KMemoryPermission::None,
+        KMemoryAttribute::All, KMemoryAttribute::None));
+
+    // Determine whether any pages being unmapped are code.
+    bool any_code_pages = false;
+    {
+        KMemoryBlockManager::const_iterator it = block_manager->FindIterator(dst_address);
+        while (true) {
+            // Get the memory info.
+            const KMemoryInfo info = it->GetMemoryInfo();
+
+            // Check if the memory has code flag.
+            if ((info.GetState() & KMemoryState::FlagCode) != KMemoryState::None) {
+                any_code_pages = true;
+                break;
+            }
+
+            // Check if we're done.
+            if (dst_address + size - 1 <= info.GetLastAddress()) {
+                break;
+            }
+
+            // Advance.
+            ++it;
+        }
     }
 
-    const std::size_t num_pages{size / PageSize};
+    // Ensure that we maintain the instruction cache.
+    bool reprotected_pages = false;
+    SCOPE_EXIT({
+        if (reprotected_pages && any_code_pages) {
+            system.InvalidateCpuInstructionCacheRange(dst_address, size);
+        }
+    });
 
-    CASCADE_CODE(CheckMemoryState(nullptr, nullptr, nullptr, nullptr, src_addr, size,
-                                  KMemoryState::All, KMemoryState::Normal, KMemoryPermission::None,
-                                  KMemoryPermission::None, KMemoryAttribute::Mask,
-                                  KMemoryAttribute::Locked, KMemoryAttribute::IpcAndDeviceMapped));
+    // Unmap.
+    {
+        // Determine the number of pages being operated on.
+        const std::size_t num_pages = size / PageSize;
 
-    KMemoryState state{};
-    CASCADE_CODE(CheckMemoryState(
-        &state, nullptr, nullptr, nullptr, dst_addr, PageSize, KMemoryState::FlagCanCodeAlias,
-        KMemoryState::FlagCanCodeAlias, KMemoryPermission::None, KMemoryPermission::None,
-        KMemoryAttribute::Mask, KMemoryAttribute::None, KMemoryAttribute::IpcAndDeviceMapped));
-    CASCADE_CODE(CheckMemoryState(dst_addr, size, KMemoryState::All, state, KMemoryPermission::None,
-                                  KMemoryPermission::None, KMemoryAttribute::Mask,
-                                  KMemoryAttribute::None));
-    CASCADE_CODE(Operate(dst_addr, num_pages, KMemoryPermission::None, OperationType::Unmap));
+        // Unmap the aliased copy of the pages.
+        R_TRY(Operate(dst_address, num_pages, KMemoryPermission::None, OperationType::Unmap));
 
-    block_manager->Update(dst_addr, num_pages, KMemoryState::Free);
-    block_manager->Update(src_addr, num_pages, KMemoryState::Normal,
-                          KMemoryPermission::UserReadWrite);
+        // Try to set the permissions for the source pages back to what they should be.
+        R_TRY(Operate(src_address, num_pages, KMemoryPermission::UserReadWrite,
+                      OperationType::ChangePermissions));
 
-    system.InvalidateCpuInstructionCacheRange(dst_addr, size);
+        // Apply the memory block updates.
+        block_manager->Update(dst_address, num_pages, KMemoryState::None);
+        block_manager->Update(src_address, num_pages, KMemoryState::Normal,
+                              KMemoryPermission::UserReadWrite);
+
+        // Note that we reprotected pages.
+        reprotected_pages = true;
+    }
 
     return ResultSuccess;
 }
 
+VAddr KPageTable::FindFreeArea(VAddr region_start, std::size_t region_num_pages,
+                               std::size_t num_pages, std::size_t alignment, std::size_t offset,
+                               std::size_t guard_pages) {
+    VAddr address = 0;
+
+    if (num_pages <= region_num_pages) {
+        if (this->IsAslrEnabled()) {
+            // Try to directly find a free area up to 8 times.
+            for (std::size_t i = 0; i < 8; i++) {
+                const std::size_t random_offset =
+                    KSystemControl::GenerateRandomRange(
+                        0, (region_num_pages - num_pages - guard_pages) * PageSize / alignment) *
+                    alignment;
+                const VAddr candidate =
+                    Common::AlignDown((region_start + random_offset), alignment) + offset;
+
+                KMemoryInfo info = this->QueryInfoImpl(candidate);
+
+                if (info.state != KMemoryState::Free) {
+                    continue;
+                }
+                if (region_start > candidate) {
+                    continue;
+                }
+                if (info.GetAddress() + guard_pages * PageSize > candidate) {
+                    continue;
+                }
+
+                const VAddr candidate_end = candidate + (num_pages + guard_pages) * PageSize - 1;
+                if (candidate_end > info.GetLastAddress()) {
+                    continue;
+                }
+                if (candidate_end > region_start + region_num_pages * PageSize - 1) {
+                    continue;
+                }
+
+                address = candidate;
+                break;
+            }
+            // Fall back to finding the first free area with a random offset.
+            if (address == 0) {
+                // NOTE: Nintendo does not account for guard pages here.
+                // This may theoretically cause an offset to be chosen that cannot be mapped. We
+                // will account for guard pages.
+                const std::size_t offset_pages = KSystemControl::GenerateRandomRange(
+                    0, region_num_pages - num_pages - guard_pages);
+                address = block_manager->FindFreeArea(region_start + offset_pages * PageSize,
+                                                      region_num_pages - offset_pages, num_pages,
+                                                      alignment, offset, guard_pages);
+            }
+        }
+
+        // Find the first free area.
+        if (address == 0) {
+            address = block_manager->FindFreeArea(region_start, region_num_pages, num_pages,
+                                                  alignment, offset, guard_pages);
+        }
+    }
+
+    return address;
+}
+
 ResultCode KPageTable::UnmapProcessMemory(VAddr dst_addr, std::size_t size,
                                           KPageTable& src_page_table, VAddr src_addr) {
-    std::lock_guard lock{page_table_lock};
+    KScopedLightLock lk(general_lock);
 
     const std::size_t num_pages{size / PageSize};
 
@@ -397,150 +514,482 @@ ResultCode KPageTable::UnmapProcessMemory(VAddr dst_addr, std::size_t size,
     return ResultSuccess;
 }
 
-ResultCode KPageTable::MapPhysicalMemory(VAddr addr, std::size_t size) {
+ResultCode KPageTable::MapPhysicalMemory(VAddr address, std::size_t size) {
     // Lock the physical memory lock.
-    std::lock_guard phys_lk(map_physical_memory_lock);
+    KScopedLightLock map_phys_mem_lk(map_physical_memory_lock);
 
-    // Lock the table.
-    std::lock_guard lock{page_table_lock};
+    // Calculate the last address for convenience.
+    const VAddr last_address = address + size - 1;
 
-    std::size_t mapped_size{};
-    const VAddr end_addr{addr + size};
+    // Define iteration variables.
+    VAddr cur_address;
+    std::size_t mapped_size;
 
-    block_manager->IterateForRange(addr, end_addr, [&](const KMemoryInfo& info) {
-        if (info.state != KMemoryState::Free) {
-            mapped_size += GetSizeInRange(info, addr, end_addr);
-        }
-    });
-
-    if (mapped_size == size) {
-        return ResultSuccess;
-    }
+    // The entire mapping process can be retried.
+    while (true) {
+        // Check if the memory is already mapped.
+        {
+            // Lock the table.
+            KScopedLightLock lk(general_lock);
+
+            // Iterate over the memory.
+            cur_address = address;
+            mapped_size = 0;
+
+            auto it = block_manager->FindIterator(cur_address);
+            while (true) {
+                // Check that the iterator is valid.
+                ASSERT(it != block_manager->end());
+
+                // Get the memory info.
+                const KMemoryInfo info = it->GetMemoryInfo();
+
+                // Check if we're done.
+                if (last_address <= info.GetLastAddress()) {
+                    if (info.GetState() != KMemoryState::Free) {
+                        mapped_size += (last_address + 1 - cur_address);
+                    }
+                    break;
+                }
+
+                // Track the memory if it's mapped.
+                if (info.GetState() != KMemoryState::Free) {
+                    mapped_size += VAddr(info.GetEndAddress()) - cur_address;
+                }
+
+                // Advance.
+                cur_address = info.GetEndAddress();
+                ++it;
+            }
 
-    const std::size_t remaining_size{size - mapped_size};
-    const std::size_t remaining_pages{remaining_size / PageSize};
+            // If the size mapped is the size requested, we've nothing to do.
+            R_SUCCEED_IF(size == mapped_size);
+        }
 
-    // Reserve the memory from the process resource limit.
-    KScopedResourceReservation memory_reservation(
-        system.Kernel().CurrentProcess()->GetResourceLimit(), LimitableResource::PhysicalMemory,
-        remaining_size);
-    if (!memory_reservation.Succeeded()) {
-        LOG_ERROR(Kernel, "Could not reserve remaining {:X} bytes", remaining_size);
-        return ResultLimitReached;
+        // Allocate and map the memory.
+        {
+            // Reserve the memory from the process resource limit.
+            KScopedResourceReservation memory_reservation(
+                system.Kernel().CurrentProcess()->GetResourceLimit(),
+                LimitableResource::PhysicalMemory, size - mapped_size);
+            R_UNLESS(memory_reservation.Succeeded(), ResultLimitReached);
+
+            // Allocate pages for the new memory.
+            KPageLinkedList pg;
+            R_TRY(system.Kernel().MemoryManager().AllocateAndOpenForProcess(
+                &pg, (size - mapped_size) / PageSize,
+                KMemoryManager::EncodeOption(memory_pool, allocation_option), 0, 0));
+
+            // Map the memory.
+            {
+                // Lock the table.
+                KScopedLightLock lk(general_lock);
+
+                size_t num_allocator_blocks = 0;
+
+                // Verify that nobody has mapped memory since we first checked.
+                {
+                    // Iterate over the memory.
+                    size_t checked_mapped_size = 0;
+                    cur_address = address;
+
+                    auto it = block_manager->FindIterator(cur_address);
+                    while (true) {
+                        // Check that the iterator is valid.
+                        ASSERT(it != block_manager->end());
+
+                        // Get the memory info.
+                        const KMemoryInfo info = it->GetMemoryInfo();
+
+                        const bool is_free = info.GetState() == KMemoryState::Free;
+                        if (is_free) {
+                            if (info.GetAddress() < address) {
+                                ++num_allocator_blocks;
+                            }
+                            if (last_address < info.GetLastAddress()) {
+                                ++num_allocator_blocks;
+                            }
+                        }
+
+                        // Check if we're done.
+                        if (last_address <= info.GetLastAddress()) {
+                            if (!is_free) {
+                                checked_mapped_size += (last_address + 1 - cur_address);
+                            }
+                            break;
+                        }
+
+                        // Track the memory if it's mapped.
+                        if (!is_free) {
+                            checked_mapped_size += VAddr(info.GetEndAddress()) - cur_address;
+                        }
+
+                        // Advance.
+                        cur_address = info.GetEndAddress();
+                        ++it;
+                    }
+
+                    // If the size now isn't what it was before, somebody mapped or unmapped
+                    // concurrently. If this happened, retry.
+                    if (mapped_size != checked_mapped_size) {
+                        continue;
+                    }
+                }
+
+                // Reset the current tracking address, and make sure we clean up on failure.
+                cur_address = address;
+                auto unmap_guard = detail::ScopeExit([&] {
+                    if (cur_address > address) {
+                        const VAddr last_unmap_address = cur_address - 1;
+
+                        // Iterate, unmapping the pages.
+                        cur_address = address;
+
+                        auto it = block_manager->FindIterator(cur_address);
+                        while (true) {
+                            // Check that the iterator is valid.
+                            ASSERT(it != block_manager->end());
+
+                            // Get the memory info.
+                            const KMemoryInfo info = it->GetMemoryInfo();
+
+                            // If the memory state is free, we mapped it and need to unmap it.
+                            if (info.GetState() == KMemoryState::Free) {
+                                // Determine the range to unmap.
+                                const size_t cur_pages =
+                                    std::min(VAddr(info.GetEndAddress()) - cur_address,
+                                             last_unmap_address + 1 - cur_address) /
+                                    PageSize;
+
+                                // Unmap.
+                                ASSERT(Operate(cur_address, cur_pages, KMemoryPermission::None,
+                                               OperationType::Unmap)
+                                           .IsSuccess());
+                            }
+
+                            // Check if we're done.
+                            if (last_unmap_address <= info.GetLastAddress()) {
+                                break;
+                            }
+
+                            // Advance.
+                            cur_address = info.GetEndAddress();
+                            ++it;
+                        }
+                    }
+                });
+
+                // Iterate over the memory.
+                auto pg_it = pg.Nodes().begin();
+                PAddr pg_phys_addr = pg_it->GetAddress();
+                size_t pg_pages = pg_it->GetNumPages();
+
+                auto it = block_manager->FindIterator(cur_address);
+                while (true) {
+                    // Check that the iterator is valid.
+                    ASSERT(it != block_manager->end());
+
+                    // Get the memory info.
+                    const KMemoryInfo info = it->GetMemoryInfo();
+
+                    // If it's unmapped, we need to map it.
+                    if (info.GetState() == KMemoryState::Free) {
+                        // Determine the range to map.
+                        size_t map_pages = std::min(VAddr(info.GetEndAddress()) - cur_address,
+                                                    last_address + 1 - cur_address) /
+                                           PageSize;
+
+                        // While we have pages to map, map them.
+                        while (map_pages > 0) {
+                            // Check if we're at the end of the physical block.
+                            if (pg_pages == 0) {
+                                // Ensure there are more pages to map.
+                                ASSERT(pg_it != pg.Nodes().end());
+
+                                // Advance our physical block.
+                                ++pg_it;
+                                pg_phys_addr = pg_it->GetAddress();
+                                pg_pages = pg_it->GetNumPages();
+                            }
+
+                            // Map whatever we can.
+                            const size_t cur_pages = std::min(pg_pages, map_pages);
+                            R_TRY(Operate(cur_address, cur_pages, KMemoryPermission::UserReadWrite,
+                                          OperationType::Map, pg_phys_addr));
+
+                            // Advance.
+                            cur_address += cur_pages * PageSize;
+                            map_pages -= cur_pages;
+
+                            pg_phys_addr += cur_pages * PageSize;
+                            pg_pages -= cur_pages;
+                        }
+                    }
+
+                    // Check if we're done.
+                    if (last_address <= info.GetLastAddress()) {
+                        break;
+                    }
+
+                    // Advance.
+                    cur_address = info.GetEndAddress();
+                    ++it;
+                }
+
+                // We succeeded, so commit the memory reservation.
+                memory_reservation.Commit();
+
+                // Increase our tracked mapped size.
+                mapped_physical_memory_size += (size - mapped_size);
+
+                // Update the relevant memory blocks.
+                block_manager->Update(address, size / PageSize, KMemoryState::Free,
+                                      KMemoryPermission::None, KMemoryAttribute::None,
+                                      KMemoryState::Normal, KMemoryPermission::UserReadWrite,
+                                      KMemoryAttribute::None);
+
+                // Cancel our guard.
+                unmap_guard.Cancel();
+
+                return ResultSuccess;
+            }
+        }
     }
+}
 
-    KPageLinkedList page_linked_list;
+ResultCode KPageTable::UnmapPhysicalMemory(VAddr address, std::size_t size) {
+    // Lock the physical memory lock.
+    KScopedLightLock map_phys_mem_lk(map_physical_memory_lock);
 
-    CASCADE_CODE(system.Kernel().MemoryManager().Allocate(page_linked_list, remaining_pages,
-                                                          memory_pool, allocation_option));
+    // Lock the table.
+    KScopedLightLock lk(general_lock);
 
-    // We succeeded, so commit the memory reservation.
-    memory_reservation.Commit();
+    // Calculate the last address for convenience.
+    const VAddr last_address = address + size - 1;
 
-    // Map the memory.
-    auto node{page_linked_list.Nodes().begin()};
-    PAddr map_addr{node->GetAddress()};
-    std::size_t src_num_pages{node->GetNumPages()};
-    block_manager->IterateForRange(addr, end_addr, [&](const KMemoryInfo& info) {
-        if (info.state != KMemoryState::Free) {
-            return;
-        }
+    // Define iteration variables.
+    VAddr cur_address = 0;
+    std::size_t mapped_size = 0;
+    std::size_t num_allocator_blocks = 0;
 
-        std::size_t dst_num_pages{GetSizeInRange(info, addr, end_addr) / PageSize};
-        VAddr dst_addr{GetAddressInRange(info, addr)};
+    // Check if the memory is mapped.
+    {
+        // Iterate over the memory.
+        cur_address = address;
+        mapped_size = 0;
+
+        auto it = block_manager->FindIterator(cur_address);
+        while (true) {
+            // Check that the iterator is valid.
+            ASSERT(it != block_manager->end());
+
+            // Get the memory info.
+            const KMemoryInfo info = it->GetMemoryInfo();
+
+            // Verify the memory's state.
+            const bool is_normal = info.GetState() == KMemoryState::Normal &&
+                                   info.GetAttribute() == KMemoryAttribute::None;
+            const bool is_free = info.GetState() == KMemoryState::Free;
+            R_UNLESS(is_normal || is_free, ResultInvalidCurrentMemory);
+
+            if (is_normal) {
+                R_UNLESS(info.GetAttribute() == KMemoryAttribute::None, ResultInvalidCurrentMemory);
+
+                if (info.GetAddress() < address) {
+                    ++num_allocator_blocks;
+                }
+                if (last_address < info.GetLastAddress()) {
+                    ++num_allocator_blocks;
+                }
+            }
 
-        while (dst_num_pages) {
-            if (!src_num_pages) {
-                node = std::next(node);
-                map_addr = node->GetAddress();
-                src_num_pages = node->GetNumPages();
+            // Check if we're done.
+            if (last_address <= info.GetLastAddress()) {
+                if (is_normal) {
+                    mapped_size += (last_address + 1 - cur_address);
+                }
+                break;
             }
 
-            const std::size_t num_pages{std::min(src_num_pages, dst_num_pages)};
-            Operate(dst_addr, num_pages, KMemoryPermission::UserReadWrite, OperationType::Map,
-                    map_addr);
+            // Track the memory if it's mapped.
+            if (is_normal) {
+                mapped_size += VAddr(info.GetEndAddress()) - cur_address;
+            }
 
-            dst_addr += num_pages * PageSize;
-            map_addr += num_pages * PageSize;
-            src_num_pages -= num_pages;
-            dst_num_pages -= num_pages;
+            // Advance.
+            cur_address = info.GetEndAddress();
+            ++it;
         }
-    });
-
-    mapped_physical_memory_size += remaining_size;
 
-    const std::size_t num_pages{size / PageSize};
-    block_manager->Update(addr, num_pages, KMemoryState::Free, KMemoryPermission::None,
-                          KMemoryAttribute::None, KMemoryState::Normal,
-                          KMemoryPermission::UserReadWrite, KMemoryAttribute::None);
+        // If there's nothing mapped, we've nothing to do.
+        R_SUCCEED_IF(mapped_size == 0);
+    }
 
-    return ResultSuccess;
-}
+    // Make a page group for the unmap region.
+    KPageLinkedList pg;
+    {
+        auto& impl = this->PageTableImpl();
+
+        // Begin traversal.
+        Common::PageTable::TraversalContext context;
+        Common::PageTable::TraversalEntry cur_entry = {.phys_addr = 0, .block_size = 0};
+        bool cur_valid = false;
+        Common::PageTable::TraversalEntry next_entry;
+        bool next_valid = false;
+        size_t tot_size = 0;
+
+        cur_address = address;
+        next_valid = impl.BeginTraversal(next_entry, context, cur_address);
+        next_entry.block_size =
+            (next_entry.block_size - (next_entry.phys_addr & (next_entry.block_size - 1)));
+
+        // Iterate, building the group.
+        while (true) {
+            if ((!next_valid && !cur_valid) ||
+                (next_valid && cur_valid &&
+                 next_entry.phys_addr == cur_entry.phys_addr + cur_entry.block_size)) {
+                cur_entry.block_size += next_entry.block_size;
+            } else {
+                if (cur_valid) {
+                    // ASSERT(IsHeapPhysicalAddress(cur_entry.phys_addr));
+                    R_TRY(pg.AddBlock(cur_entry.phys_addr, cur_entry.block_size / PageSize));
+                }
+
+                // Update tracking variables.
+                tot_size += cur_entry.block_size;
+                cur_entry = next_entry;
+                cur_valid = next_valid;
+            }
 
-ResultCode KPageTable::UnmapPhysicalMemory(VAddr addr, std::size_t size) {
-    std::lock_guard lock{page_table_lock};
+            if (cur_entry.block_size + tot_size >= size) {
+                break;
+            }
 
-    const VAddr end_addr{addr + size};
-    ResultCode result{ResultSuccess};
-    std::size_t mapped_size{};
+            next_valid = impl.ContinueTraversal(next_entry, context);
+        }
 
-    // Verify that the region can be unmapped
-    block_manager->IterateForRange(addr, end_addr, [&](const KMemoryInfo& info) {
-        if (info.state == KMemoryState::Normal) {
-            if (info.attribute != KMemoryAttribute::None) {
-                result = ResultInvalidCurrentMemory;
-                return;
+        // Add the last block.
+        if (cur_valid) {
+            // ASSERT(IsHeapPhysicalAddress(cur_entry.phys_addr));
+            R_TRY(pg.AddBlock(cur_entry.phys_addr, (size - tot_size) / PageSize));
+        }
+    }
+    ASSERT(pg.GetNumPages() == mapped_size / PageSize);
+
+    // Reset the current tracking address, and make sure we clean up on failure.
+    cur_address = address;
+    auto remap_guard = detail::ScopeExit([&] {
+        if (cur_address > address) {
+            const VAddr last_map_address = cur_address - 1;
+            cur_address = address;
+
+            // Iterate over the memory we unmapped.
+            auto it = block_manager->FindIterator(cur_address);
+            auto pg_it = pg.Nodes().begin();
+            PAddr pg_phys_addr = pg_it->GetAddress();
+            size_t pg_pages = pg_it->GetNumPages();
+
+            while (true) {
+                // Get the memory info for the pages we unmapped, convert to property.
+                const KMemoryInfo info = it->GetMemoryInfo();
+
+                // If the memory is normal, we unmapped it and need to re-map it.
+                if (info.GetState() == KMemoryState::Normal) {
+                    // Determine the range to map.
+                    size_t map_pages = std::min(VAddr(info.GetEndAddress()) - cur_address,
+                                                last_map_address + 1 - cur_address) /
+                                       PageSize;
+
+                    // While we have pages to map, map them.
+                    while (map_pages > 0) {
+                        // Check if we're at the end of the physical block.
+                        if (pg_pages == 0) {
+                            // Ensure there are more pages to map.
+                            ASSERT(pg_it != pg.Nodes().end());
+
+                            // Advance our physical block.
+                            ++pg_it;
+                            pg_phys_addr = pg_it->GetAddress();
+                            pg_pages = pg_it->GetNumPages();
+                        }
+
+                        // Map whatever we can.
+                        const size_t cur_pages = std::min(pg_pages, map_pages);
+                        ASSERT(this->Operate(cur_address, cur_pages, info.GetPermission(),
+                                             OperationType::Map, pg_phys_addr) == ResultSuccess);
+
+                        // Advance.
+                        cur_address += cur_pages * PageSize;
+                        map_pages -= cur_pages;
+
+                        pg_phys_addr += cur_pages * PageSize;
+                        pg_pages -= cur_pages;
+                    }
+                }
+
+                // Check if we're done.
+                if (last_map_address <= info.GetLastAddress()) {
+                    break;
+                }
+
+                // Advance.
+                ++it;
             }
-            mapped_size += GetSizeInRange(info, addr, end_addr);
-        } else if (info.state != KMemoryState::Free) {
-            result = ResultInvalidCurrentMemory;
         }
     });
 
-    if (result.IsError()) {
-        return result;
-    }
+    // Iterate over the memory, unmapping as we go.
+    auto it = block_manager->FindIterator(cur_address);
+    while (true) {
+        // Check that the iterator is valid.
+        ASSERT(it != block_manager->end());
 
-    if (!mapped_size) {
-        return ResultSuccess;
-    }
+        // Get the memory info.
+        const KMemoryInfo info = it->GetMemoryInfo();
 
-    // Unmap each region within the range
-    KPageLinkedList page_linked_list;
-    block_manager->IterateForRange(addr, end_addr, [&](const KMemoryInfo& info) {
-        if (info.state == KMemoryState::Normal) {
-            const std::size_t block_size{GetSizeInRange(info, addr, end_addr)};
-            const std::size_t block_num_pages{block_size / PageSize};
-            const VAddr block_addr{GetAddressInRange(info, addr)};
-
-            AddRegionToPages(block_addr, block_size / PageSize, page_linked_list);
-
-            if (result = Operate(block_addr, block_num_pages, KMemoryPermission::None,
-                                 OperationType::Unmap);
-                result.IsError()) {
-                return;
-            }
+        // If the memory state is normal, we need to unmap it.
+        if (info.GetState() == KMemoryState::Normal) {
+            // Determine the range to unmap.
+            const size_t cur_pages = std::min(VAddr(info.GetEndAddress()) - cur_address,
+                                              last_address + 1 - cur_address) /
+                                     PageSize;
+
+            // Unmap.
+            R_TRY(Operate(cur_address, cur_pages, KMemoryPermission::None, OperationType::Unmap));
         }
-    });
-    if (result.IsError()) {
-        return result;
-    }
 
-    const std::size_t num_pages{size / PageSize};
-    system.Kernel().MemoryManager().Free(page_linked_list, num_pages, memory_pool,
-                                         allocation_option);
+        // Check if we're done.
+        if (last_address <= info.GetLastAddress()) {
+            break;
+        }
 
-    block_manager->Update(addr, num_pages, KMemoryState::Free);
+        // Advance.
+        cur_address = info.GetEndAddress();
+        ++it;
+    }
 
+    // Release the memory resource.
+    mapped_physical_memory_size -= mapped_size;
     auto process{system.Kernel().CurrentProcess()};
     process->GetResourceLimit()->Release(LimitableResource::PhysicalMemory, mapped_size);
-    mapped_physical_memory_size -= mapped_size;
+
+    // Update memory blocks.
+    block_manager->Update(address, size / PageSize, KMemoryState::Free, KMemoryPermission::None,
+                          KMemoryAttribute::None);
+
+    // TODO(bunnei): This is a workaround until the next set of changes, where we add reference
+    // counting for mapped pages. Until then, we must manually close the reference to the page
+    // group.
+    system.Kernel().MemoryManager().Close(pg);
+
+    // We succeeded.
+    remap_guard.Cancel();
 
     return ResultSuccess;
 }
 
 ResultCode KPageTable::MapMemory(VAddr dst_addr, VAddr src_addr, std::size_t size) {
-    std::lock_guard lock{page_table_lock};
+    KScopedLightLock lk(general_lock);
 
     KMemoryState src_state{};
     CASCADE_CODE(CheckMemoryState(
@@ -579,7 +1028,7 @@ ResultCode KPageTable::MapMemory(VAddr dst_addr, VAddr src_addr, std::size_t siz
 }
 
 ResultCode KPageTable::UnmapMemory(VAddr dst_addr, VAddr src_addr, std::size_t size) {
-    std::lock_guard lock{page_table_lock};
+    KScopedLightLock lk(general_lock);
 
     KMemoryState src_state{};
     CASCADE_CODE(CheckMemoryState(
@@ -622,6 +1071,8 @@ ResultCode KPageTable::UnmapMemory(VAddr dst_addr, VAddr src_addr, std::size_t s
 
 ResultCode KPageTable::MapPages(VAddr addr, const KPageLinkedList& page_linked_list,
                                 KMemoryPermission perm) {
+    ASSERT(this->IsLockedByCurrentThread());
+
     VAddr cur_addr{addr};
 
     for (const auto& node : page_linked_list.Nodes()) {
@@ -650,7 +1101,7 @@ ResultCode KPageTable::MapPages(VAddr address, KPageLinkedList& page_linked_list
     R_UNLESS(this->CanContain(address, size, state), ResultInvalidCurrentMemory);
 
     // Lock the table.
-    std::lock_guard lock{page_table_lock};
+    KScopedLightLock lk(general_lock);
 
     // Check the memory state.
     R_TRY(this->CheckMemoryState(address, size, KMemoryState::All, KMemoryState::Free,
@@ -666,13 +1117,54 @@ ResultCode KPageTable::MapPages(VAddr address, KPageLinkedList& page_linked_list
     return ResultSuccess;
 }
 
+ResultCode KPageTable::MapPages(VAddr* out_addr, std::size_t num_pages, std::size_t alignment,
+                                PAddr phys_addr, bool is_pa_valid, VAddr region_start,
+                                std::size_t region_num_pages, KMemoryState state,
+                                KMemoryPermission perm) {
+    ASSERT(Common::IsAligned(alignment, PageSize) && alignment >= PageSize);
+
+    // Ensure this is a valid map request.
+    R_UNLESS(this->CanContain(region_start, region_num_pages * PageSize, state),
+             ResultInvalidCurrentMemory);
+    R_UNLESS(num_pages < region_num_pages, ResultOutOfMemory);
+
+    // Lock the table.
+    KScopedLightLock lk(general_lock);
+
+    // Find a random address to map at.
+    VAddr addr = this->FindFreeArea(region_start, region_num_pages, num_pages, alignment, 0,
+                                    this->GetNumGuardPages());
+    R_UNLESS(addr != 0, ResultOutOfMemory);
+    ASSERT(Common::IsAligned(addr, alignment));
+    ASSERT(this->CanContain(addr, num_pages * PageSize, state));
+    ASSERT(this->CheckMemoryState(addr, num_pages * PageSize, KMemoryState::All, KMemoryState::Free,
+                                  KMemoryPermission::None, KMemoryPermission::None,
+                                  KMemoryAttribute::None, KMemoryAttribute::None)
+               .IsSuccess());
+
+    // Perform mapping operation.
+    if (is_pa_valid) {
+        R_TRY(this->Operate(addr, num_pages, perm, OperationType::Map, phys_addr));
+    } else {
+        UNIMPLEMENTED();
+    }
+
+    // Update the blocks.
+    block_manager->Update(addr, num_pages, state, perm);
+
+    // We successfully mapped the pages.
+    *out_addr = addr;
+    return ResultSuccess;
+}
+
 ResultCode KPageTable::UnmapPages(VAddr addr, const KPageLinkedList& page_linked_list) {
+    ASSERT(this->IsLockedByCurrentThread());
+
     VAddr cur_addr{addr};
 
     for (const auto& node : page_linked_list.Nodes()) {
-        const std::size_t num_pages{(addr - cur_addr) / PageSize};
-        if (const auto result{
-                Operate(addr, num_pages, KMemoryPermission::None, OperationType::Unmap)};
+        if (const auto result{Operate(cur_addr, node.GetNumPages(), KMemoryPermission::None,
+                                      OperationType::Unmap)};
             result.IsError()) {
             return result;
         }
@@ -691,7 +1183,7 @@ ResultCode KPageTable::UnmapPages(VAddr addr, KPageLinkedList& page_linked_list,
     R_UNLESS(this->Contains(addr, size), ResultInvalidCurrentMemory);
 
     // Lock the table.
-    std::lock_guard lock{page_table_lock};
+    KScopedLightLock lk(general_lock);
 
     // Check the memory state.
     R_TRY(this->CheckMemoryState(addr, size, KMemoryState::All, state, KMemoryPermission::None,
@@ -707,12 +1199,36 @@ ResultCode KPageTable::UnmapPages(VAddr addr, KPageLinkedList& page_linked_list,
     return ResultSuccess;
 }
 
+ResultCode KPageTable::UnmapPages(VAddr address, std::size_t num_pages, KMemoryState state) {
+    // Check that the unmap is in range.
+    const std::size_t size = num_pages * PageSize;
+    R_UNLESS(this->Contains(address, size), ResultInvalidCurrentMemory);
+
+    // Lock the table.
+    KScopedLightLock lk(general_lock);
+
+    // Check the memory state.
+    std::size_t num_allocator_blocks{};
+    R_TRY(this->CheckMemoryState(std::addressof(num_allocator_blocks), address, size,
+                                 KMemoryState::All, state, KMemoryPermission::None,
+                                 KMemoryPermission::None, KMemoryAttribute::All,
+                                 KMemoryAttribute::None));
+
+    // Perform the unmap.
+    R_TRY(Operate(address, num_pages, KMemoryPermission::None, OperationType::Unmap));
+
+    // Update the blocks.
+    block_manager->Update(address, num_pages, KMemoryState::Free, KMemoryPermission::None);
+
+    return ResultSuccess;
+}
+
 ResultCode KPageTable::SetProcessMemoryPermission(VAddr addr, std::size_t size,
                                                   Svc::MemoryPermission svc_perm) {
     const size_t num_pages = size / PageSize;
 
     // Lock the table.
-    std::lock_guard lock{page_table_lock};
+    KScopedLightLock lk(general_lock);
 
     // Verify we can change the memory permission.
     KMemoryState old_state;
@@ -766,7 +1282,7 @@ ResultCode KPageTable::SetProcessMemoryPermission(VAddr addr, std::size_t size,
 }
 
 KMemoryInfo KPageTable::QueryInfoImpl(VAddr addr) {
-    std::lock_guard lock{page_table_lock};
+    KScopedLightLock lk(general_lock);
 
     return block_manager->FindBlock(addr).GetMemoryInfo();
 }
@@ -781,7 +1297,7 @@ KMemoryInfo KPageTable::QueryInfo(VAddr addr) {
 }
 
 ResultCode KPageTable::ReserveTransferMemory(VAddr addr, std::size_t size, KMemoryPermission perm) {
-    std::lock_guard lock{page_table_lock};
+    KScopedLightLock lk(general_lock);
 
     KMemoryState state{};
     KMemoryAttribute attribute{};
@@ -799,7 +1315,7 @@ ResultCode KPageTable::ReserveTransferMemory(VAddr addr, std::size_t size, KMemo
 }
 
 ResultCode KPageTable::ResetTransferMemory(VAddr addr, std::size_t size) {
-    std::lock_guard lock{page_table_lock};
+    KScopedLightLock lk(general_lock);
 
     KMemoryState state{};
 
@@ -818,7 +1334,7 @@ ResultCode KPageTable::SetMemoryPermission(VAddr addr, std::size_t size,
     const size_t num_pages = size / PageSize;
 
     // Lock the table.
-    std::lock_guard lock{page_table_lock};
+    KScopedLightLock lk(general_lock);
 
     // Verify we can change the memory permission.
     KMemoryState old_state;
@@ -847,7 +1363,7 @@ ResultCode KPageTable::SetMemoryAttribute(VAddr addr, std::size_t size, u32 mask
            KMemoryAttribute::SetMask);
 
     // Lock the table.
-    std::lock_guard lock{page_table_lock};
+    KScopedLightLock lk(general_lock);
 
     // Verify we can change the memory attribute.
     KMemoryState old_state;
@@ -878,7 +1394,7 @@ ResultCode KPageTable::SetMemoryAttribute(VAddr addr, std::size_t size, u32 mask
 
 ResultCode KPageTable::SetMaxHeapSize(std::size_t size) {
     // Lock the table.
-    std::lock_guard lock{page_table_lock};
+    KScopedLightLock lk(general_lock);
 
     // Only process page tables are allowed to set heap size.
     ASSERT(!this->IsKernel());
@@ -889,15 +1405,15 @@ ResultCode KPageTable::SetMaxHeapSize(std::size_t size) {
 }
 
 ResultCode KPageTable::SetHeapSize(VAddr* out, std::size_t size) {
-    // Lock the physical memory lock.
-    std::lock_guard phys_lk(map_physical_memory_lock);
+    // Lock the physical memory mutex.
+    KScopedLightLock map_phys_mem_lk(map_physical_memory_lock);
 
     // Try to perform a reduction in heap, instead of an extension.
     VAddr cur_address{};
     std::size_t allocation_size{};
     {
         // Lock the table.
-        std::lock_guard lk(page_table_lock);
+        KScopedLightLock lk(general_lock);
 
         // Validate that setting heap size is possible at all.
         R_UNLESS(!is_kernel, ResultOutOfMemory);
@@ -955,14 +1471,21 @@ ResultCode KPageTable::SetHeapSize(VAddr* out, std::size_t size) {
     R_UNLESS(memory_reservation.Succeeded(), ResultLimitReached);
 
     // Allocate pages for the heap extension.
-    KPageLinkedList page_linked_list;
-    R_TRY(system.Kernel().MemoryManager().Allocate(page_linked_list, allocation_size / PageSize,
-                                                   memory_pool, allocation_option));
+    KPageLinkedList pg;
+    R_TRY(system.Kernel().MemoryManager().AllocateAndOpen(
+        &pg, allocation_size / PageSize,
+        KMemoryManager::EncodeOption(memory_pool, allocation_option)));
+
+    // Clear all the newly allocated pages.
+    for (const auto& it : pg.Nodes()) {
+        std::memset(system.DeviceMemory().GetPointer(it.GetAddress()), heap_fill_value,
+                    it.GetSize());
+    }
 
     // Map the pages.
     {
         // Lock the table.
-        std::lock_guard lk(page_table_lock);
+        KScopedLightLock lk(general_lock);
 
         // Ensure that the heap hasn't changed since we began executing.
         ASSERT(cur_address == current_heap_end);
@@ -976,7 +1499,7 @@ ResultCode KPageTable::SetHeapSize(VAddr* out, std::size_t size) {
 
         // Map the pages.
         const auto num_pages = allocation_size / PageSize;
-        R_TRY(Operate(current_heap_end, num_pages, page_linked_list, OperationType::MapGroup));
+        R_TRY(Operate(current_heap_end, num_pages, pg, OperationType::MapGroup));
 
         // Clear all the newly allocated pages.
         for (std::size_t cur_page = 0; cur_page < num_pages; ++cur_page) {
@@ -1004,7 +1527,7 @@ ResultVal<VAddr> KPageTable::AllocateAndMapMemory(std::size_t needed_num_pages,
                                                   bool is_map_only, VAddr region_start,
                                                   std::size_t region_num_pages, KMemoryState state,
                                                   KMemoryPermission perm, PAddr map_addr) {
-    std::lock_guard lock{page_table_lock};
+    KScopedLightLock lk(general_lock);
 
     if (!CanContain(region_start, region_num_pages * PageSize, state)) {
         return ResultInvalidCurrentMemory;
@@ -1024,8 +1547,9 @@ ResultVal<VAddr> KPageTable::AllocateAndMapMemory(std::size_t needed_num_pages,
         R_TRY(Operate(addr, needed_num_pages, perm, OperationType::Map, map_addr));
     } else {
         KPageLinkedList page_group;
-        R_TRY(system.Kernel().MemoryManager().Allocate(page_group, needed_num_pages, memory_pool,
-                                                       allocation_option));
+        R_TRY(system.Kernel().MemoryManager().AllocateAndOpenForProcess(
+            &page_group, needed_num_pages,
+            KMemoryManager::EncodeOption(memory_pool, allocation_option), 0, 0));
         R_TRY(Operate(addr, needed_num_pages, page_group, OperationType::MapGroup));
     }
 
@@ -1035,7 +1559,7 @@ ResultVal<VAddr> KPageTable::AllocateAndMapMemory(std::size_t needed_num_pages,
 }
 
 ResultCode KPageTable::LockForDeviceAddressSpace(VAddr addr, std::size_t size) {
-    std::lock_guard lock{page_table_lock};
+    KScopedLightLock lk(general_lock);
 
     KMemoryPermission perm{};
     if (const ResultCode result{CheckMemoryState(
@@ -1058,7 +1582,7 @@ ResultCode KPageTable::LockForDeviceAddressSpace(VAddr addr, std::size_t size) {
 }
 
 ResultCode KPageTable::UnlockForDeviceAddressSpace(VAddr addr, std::size_t size) {
-    std::lock_guard lock{page_table_lock};
+    KScopedLightLock lk(general_lock);
 
     KMemoryPermission perm{};
     if (const ResultCode result{CheckMemoryState(
@@ -1081,7 +1605,7 @@ ResultCode KPageTable::UnlockForDeviceAddressSpace(VAddr addr, std::size_t size)
 }
 
 ResultCode KPageTable::LockForCodeMemory(VAddr addr, std::size_t size) {
-    std::lock_guard lock{page_table_lock};
+    KScopedLightLock lk(general_lock);
 
     KMemoryPermission new_perm = KMemoryPermission::NotMapped | KMemoryPermission::KernelReadWrite;
 
@@ -1108,7 +1632,7 @@ ResultCode KPageTable::LockForCodeMemory(VAddr addr, std::size_t size) {
 }
 
 ResultCode KPageTable::UnlockForCodeMemory(VAddr addr, std::size_t size) {
-    std::lock_guard lock{page_table_lock};
+    KScopedLightLock lk(general_lock);
 
     KMemoryPermission new_perm = KMemoryPermission::UserReadWrite;
 
@@ -1232,7 +1756,7 @@ ResultCode KPageTable::Operate(VAddr addr, std::size_t num_pages, KMemoryPermiss
     return ResultSuccess;
 }
 
-constexpr VAddr KPageTable::GetRegionAddress(KMemoryState state) const {
+VAddr KPageTable::GetRegionAddress(KMemoryState state) const {
     switch (state) {
     case KMemoryState::Free:
     case KMemoryState::Kernel:
@@ -1268,7 +1792,7 @@ constexpr VAddr KPageTable::GetRegionAddress(KMemoryState state) const {
     }
 }
 
-constexpr std::size_t KPageTable::GetRegionSize(KMemoryState state) const {
+std::size_t KPageTable::GetRegionSize(KMemoryState state) const {
     switch (state) {
     case KMemoryState::Free:
     case KMemoryState::Kernel:
diff --git a/src/core/hle/kernel/k_page_table.h b/src/core/hle/kernel/k_page_table.h
index 60ae9b9e8..54c6adf8d 100644
--- a/src/core/hle/kernel/k_page_table.h
+++ b/src/core/hle/kernel/k_page_table.h
@@ -5,11 +5,12 @@
 #pragma once
 
 #include <memory>
-#include <mutex>
 
+#include "common/common_funcs.h"
 #include "common/common_types.h"
 #include "common/page_table.h"
 #include "core/file_sys/program_metadata.h"
+#include "core/hle/kernel/k_light_lock.h"
 #include "core/hle/kernel/k_memory_block.h"
 #include "core/hle/kernel/k_memory_manager.h"
 #include "core/hle/result.h"
@@ -22,17 +23,21 @@ namespace Kernel {
 
 class KMemoryBlockManager;
 
-class KPageTable final : NonCopyable {
+class KPageTable final {
 public:
+    YUZU_NON_COPYABLE(KPageTable);
+    YUZU_NON_MOVEABLE(KPageTable);
+
     explicit KPageTable(Core::System& system_);
+    ~KPageTable();
 
     ResultCode InitializeForProcess(FileSys::ProgramAddressSpaceType as_type, bool enable_aslr,
                                     VAddr code_addr, std::size_t code_size,
                                     KMemoryManager::Pool pool);
     ResultCode MapProcessCode(VAddr addr, std::size_t pages_count, KMemoryState state,
                               KMemoryPermission perm);
-    ResultCode MapCodeMemory(VAddr dst_addr, VAddr src_addr, std::size_t size);
-    ResultCode UnmapCodeMemory(VAddr dst_addr, VAddr src_addr, std::size_t size);
+    ResultCode MapCodeMemory(VAddr dst_address, VAddr src_address, std::size_t size);
+    ResultCode UnmapCodeMemory(VAddr dst_address, VAddr src_address, std::size_t size);
     ResultCode UnmapProcessMemory(VAddr dst_addr, std::size_t size, KPageTable& src_page_table,
                                   VAddr src_addr);
     ResultCode MapPhysicalMemory(VAddr addr, std::size_t size);
@@ -41,7 +46,14 @@ public:
     ResultCode UnmapMemory(VAddr dst_addr, VAddr src_addr, std::size_t size);
     ResultCode MapPages(VAddr addr, KPageLinkedList& page_linked_list, KMemoryState state,
                         KMemoryPermission perm);
+    ResultCode MapPages(VAddr* out_addr, std::size_t num_pages, std::size_t alignment,
+                        PAddr phys_addr, KMemoryState state, KMemoryPermission perm) {
+        return this->MapPages(out_addr, num_pages, alignment, phys_addr, true,
+                              this->GetRegionAddress(state), this->GetRegionSize(state) / PageSize,
+                              state, perm);
+    }
     ResultCode UnmapPages(VAddr addr, KPageLinkedList& page_linked_list, KMemoryState state);
+    ResultCode UnmapPages(VAddr address, std::size_t num_pages, KMemoryState state);
     ResultCode SetProcessMemoryPermission(VAddr addr, std::size_t size,
                                           Svc::MemoryPermission svc_perm);
     KMemoryInfo QueryInfo(VAddr addr);
@@ -86,6 +98,9 @@ private:
     ResultCode InitializeMemoryLayout(VAddr start, VAddr end);
     ResultCode MapPages(VAddr addr, const KPageLinkedList& page_linked_list,
                         KMemoryPermission perm);
+    ResultCode MapPages(VAddr* out_addr, std::size_t num_pages, std::size_t alignment,
+                        PAddr phys_addr, bool is_pa_valid, VAddr region_start,
+                        std::size_t region_num_pages, KMemoryState state, KMemoryPermission perm);
     ResultCode UnmapPages(VAddr addr, const KPageLinkedList& page_linked_list);
     bool IsRegionMapped(VAddr address, u64 size);
     bool IsRegionContiguous(VAddr addr, u64 size) const;
@@ -97,8 +112,11 @@ private:
                        OperationType operation);
     ResultCode Operate(VAddr addr, std::size_t num_pages, KMemoryPermission perm,
                        OperationType operation, PAddr map_addr = 0);
-    constexpr VAddr GetRegionAddress(KMemoryState state) const;
-    constexpr std::size_t GetRegionSize(KMemoryState state) const;
+    VAddr GetRegionAddress(KMemoryState state) const;
+    std::size_t GetRegionSize(KMemoryState state) const;
+
+    VAddr FindFreeArea(VAddr region_start, std::size_t region_num_pages, std::size_t num_pages,
+                       std::size_t alignment, std::size_t offset, std::size_t guard_pages);
 
     ResultCode CheckMemoryStateContiguous(std::size_t* out_blocks_needed, VAddr addr,
                                           std::size_t size, KMemoryState state_mask,
@@ -132,7 +150,7 @@ private:
         return CheckMemoryState(nullptr, nullptr, nullptr, out_blocks_needed, addr, size,
                                 state_mask, state, perm_mask, perm, attr_mask, attr, ignore_attr);
     }
-    ResultCode CheckMemoryState(VAddr addr, size_t size, KMemoryState state_mask,
+    ResultCode CheckMemoryState(VAddr addr, std::size_t size, KMemoryState state_mask,
                                 KMemoryState state, KMemoryPermission perm_mask,
                                 KMemoryPermission perm, KMemoryAttribute attr_mask,
                                 KMemoryAttribute attr,
@@ -142,11 +160,12 @@ private:
     }
 
     bool IsLockedByCurrentThread() const {
-        return true;
+        return general_lock.IsLockedByCurrentThread();
     }
 
-    std::recursive_mutex page_table_lock;
-    std::mutex map_physical_memory_lock;
+    mutable KLightLock general_lock;
+    mutable KLightLock map_physical_memory_lock;
+
     std::unique_ptr<KMemoryBlockManager> block_manager;
 
 public:
@@ -204,8 +223,8 @@ public:
     constexpr VAddr GetAliasCodeRegionSize() const {
         return alias_code_region_end - alias_code_region_start;
     }
-    size_t GetNormalMemorySize() {
-        std::lock_guard lk(page_table_lock);
+    std::size_t GetNormalMemorySize() {
+        KScopedLightLock lk(general_lock);
         return GetHeapSize() + mapped_physical_memory_size;
     }
     constexpr std::size_t GetAddressSpaceWidth() const {
@@ -247,7 +266,10 @@ public:
     constexpr bool IsInsideASLRRegion(VAddr address, std::size_t size) const {
         return !IsOutsideASLRRegion(address, size);
     }
-    constexpr PAddr GetPhysicalAddr(VAddr addr) {
+    constexpr std::size_t GetNumGuardPages() const {
+        return IsKernel() ? 1 : 4;
+    }
+    PAddr GetPhysicalAddr(VAddr addr) const {
         const auto backing_addr = page_table_impl.backing_addr[addr >> PageBits];
         ASSERT(backing_addr);
         return backing_addr + addr;
@@ -268,10 +290,6 @@ private:
         return is_aslr_enabled;
     }
 
-    constexpr std::size_t GetNumGuardPages() const {
-        return IsKernel() ? 1 : 4;
-    }
-
     constexpr bool ContainsPages(VAddr addr, std::size_t num_pages) const {
         return (address_space_start <= addr) &&
                (num_pages <= (address_space_end - address_space_start) / PageSize) &&
@@ -303,6 +321,8 @@ private:
     bool is_kernel{};
     bool is_aslr_enabled{};
 
+    u32 heap_fill_value{};
+
     KMemoryManager::Pool memory_pool{KMemoryManager::Pool::Application};
     KMemoryManager::Direction allocation_option{KMemoryManager::Direction::FromFront};
 
diff --git a/src/core/hle/kernel/k_port.cpp b/src/core/hle/kernel/k_port.cpp
index a8ba09c4a..ceb98709f 100644
--- a/src/core/hle/kernel/k_port.cpp
+++ b/src/core/hle/kernel/k_port.cpp
@@ -57,7 +57,12 @@ ResultCode KPort::EnqueueSession(KServerSession* session) {
     R_UNLESS(state == State::Normal, ResultPortClosed);
 
     server.EnqueueSession(session);
-    server.GetSessionRequestHandler()->ClientConnected(server.AcceptSession());
+
+    if (auto session_ptr = server.GetSessionRequestHandler().lock()) {
+        session_ptr->ClientConnected(server.AcceptSession());
+    } else {
+        UNREACHABLE();
+    }
 
     return ResultSuccess;
 }
diff --git a/src/core/hle/kernel/k_process.cpp b/src/core/hle/kernel/k_process.cpp
index 85c506979..b39405496 100644
--- a/src/core/hle/kernel/k_process.cpp
+++ b/src/core/hle/kernel/k_process.cpp
@@ -70,65 +70,12 @@ void SetupMainThread(Core::System& system, KProcess& owner_process, u32 priority
 }
 } // Anonymous namespace
 
-// Represents a page used for thread-local storage.
-//
-// Each TLS page contains slots that may be used by processes and threads.
-// Every process and thread is created with a slot in some arbitrary page
-// (whichever page happens to have an available slot).
-class TLSPage {
-public:
-    static constexpr std::size_t num_slot_entries =
-        Core::Memory::PAGE_SIZE / Core::Memory::TLS_ENTRY_SIZE;
-
-    explicit TLSPage(VAddr address) : base_address{address} {}
-
-    bool HasAvailableSlots() const {
-        return !is_slot_used.all();
-    }
-
-    VAddr GetBaseAddress() const {
-        return base_address;
-    }
-
-    std::optional<VAddr> ReserveSlot() {
-        for (std::size_t i = 0; i < is_slot_used.size(); i++) {
-            if (is_slot_used[i]) {
-                continue;
-            }
-
-            is_slot_used[i] = true;
-            return base_address + (i * Core::Memory::TLS_ENTRY_SIZE);
-        }
-
-        return std::nullopt;
-    }
-
-    void ReleaseSlot(VAddr address) {
-        // Ensure that all given addresses are consistent with how TLS pages
-        // are intended to be used when releasing slots.
-        ASSERT(IsWithinPage(address));
-        ASSERT((address % Core::Memory::TLS_ENTRY_SIZE) == 0);
-
-        const std::size_t index = (address - base_address) / Core::Memory::TLS_ENTRY_SIZE;
-        is_slot_used[index] = false;
-    }
-
-private:
-    bool IsWithinPage(VAddr address) const {
-        return base_address <= address && address < base_address + Core::Memory::PAGE_SIZE;
-    }
-
-    VAddr base_address;
-    std::bitset<num_slot_entries> is_slot_used;
-};
-
 ResultCode KProcess::Initialize(KProcess* process, Core::System& system, std::string process_name,
-                                ProcessType type) {
+                                ProcessType type, KResourceLimit* res_limit) {
     auto& kernel = system.Kernel();
 
     process->name = std::move(process_name);
-
-    process->resource_limit = kernel.GetSystemResourceLimit();
+    process->resource_limit = res_limit;
     process->status = ProcessStatus::Created;
     process->program_id = 0;
     process->process_id = type == ProcessType::KernelInternal ? kernel.CreateNewKernelProcessID()
@@ -143,9 +90,6 @@ ResultCode KProcess::Initialize(KProcess* process, Core::System& system, std::st
 
     kernel.AppendNewProcess(process);
 
-    // Open a reference to the resource limit.
-    process->resource_limit->Open();
-
     // Clear remaining fields.
     process->num_running_threads = 0;
     process->is_signaled = false;
@@ -153,6 +97,9 @@ ResultCode KProcess::Initialize(KProcess* process, Core::System& system, std::st
     process->is_suspended = false;
     process->schedule_count = 0;
 
+    // Open a reference to the resource limit.
+    process->resource_limit->Open();
+
     return ResultSuccess;
 }
 
@@ -405,7 +352,7 @@ ResultCode KProcess::LoadFromMetadata(const FileSys::ProgramMetadata& metadata,
     }
 
     // Create TLS region
-    tls_region_address = CreateTLSRegion();
+    R_TRY(this->CreateThreadLocalRegion(std::addressof(tls_region_address)));
     memory_reservation.Commit();
 
     return handle_table.Initialize(capabilities.GetHandleTableSize());
@@ -445,7 +392,7 @@ void KProcess::PrepareForTermination() {
 
     stop_threads(kernel.System().GlobalSchedulerContext().GetThreadList());
 
-    FreeTLSRegion(tls_region_address);
+    this->DeleteThreadLocalRegion(tls_region_address);
     tls_region_address = 0;
 
     if (resource_limit) {
@@ -457,9 +404,6 @@ void KProcess::PrepareForTermination() {
 }
 
 void KProcess::Finalize() {
-    // Finalize the handle table and close any open handles.
-    handle_table.Finalize();
-
     // Free all shared memory infos.
     {
         auto it = shared_memory_list.begin();
@@ -484,67 +428,110 @@ void KProcess::Finalize() {
         resource_limit = nullptr;
     }
 
+    // Finalize the page table.
+    page_table.reset();
+
     // Perform inherited finalization.
     KAutoObjectWithSlabHeapAndContainer<KProcess, KWorkerTask>::Finalize();
 }
 
-/**
- * Attempts to find a TLS page that contains a free slot for
- * use by a thread.
- *
- * @returns If a page with an available slot is found, then an iterator
- *          pointing to the page is returned. Otherwise the end iterator
- *          is returned instead.
- */
-static auto FindTLSPageWithAvailableSlots(std::vector<TLSPage>& tls_pages) {
-    return std::find_if(tls_pages.begin(), tls_pages.end(),
-                        [](const auto& page) { return page.HasAvailableSlots(); });
-}
+ResultCode KProcess::CreateThreadLocalRegion(VAddr* out) {
+    KThreadLocalPage* tlp = nullptr;
+    VAddr tlr = 0;
 
-VAddr KProcess::CreateTLSRegion() {
-    KScopedSchedulerLock lock(kernel);
-    if (auto tls_page_iter{FindTLSPageWithAvailableSlots(tls_pages)};
-        tls_page_iter != tls_pages.cend()) {
-        return *tls_page_iter->ReserveSlot();
-    }
+    // See if we can get a region from a partially used TLP.
+    {
+        KScopedSchedulerLock sl{kernel};
 
-    Page* const tls_page_ptr{kernel.GetUserSlabHeapPages().Allocate()};
-    ASSERT(tls_page_ptr);
+        if (auto it = partially_used_tlp_tree.begin(); it != partially_used_tlp_tree.end()) {
+            tlr = it->Reserve();
+            ASSERT(tlr != 0);
 
-    const VAddr start{page_table->GetKernelMapRegionStart()};
-    const VAddr size{page_table->GetKernelMapRegionEnd() - start};
-    const PAddr tls_map_addr{kernel.System().DeviceMemory().GetPhysicalAddr(tls_page_ptr)};
-    const VAddr tls_page_addr{page_table
-                                  ->AllocateAndMapMemory(1, PageSize, true, start, size / PageSize,
-                                                         KMemoryState::ThreadLocal,
-                                                         KMemoryPermission::UserReadWrite,
-                                                         tls_map_addr)
-                                  .ValueOr(0)};
+            if (it->IsAllUsed()) {
+                tlp = std::addressof(*it);
+                partially_used_tlp_tree.erase(it);
+                fully_used_tlp_tree.insert(*tlp);
+            }
+
+            *out = tlr;
+            return ResultSuccess;
+        }
+    }
 
-    ASSERT(tls_page_addr);
+    // Allocate a new page.
+    tlp = KThreadLocalPage::Allocate(kernel);
+    R_UNLESS(tlp != nullptr, ResultOutOfMemory);
+    auto tlp_guard = SCOPE_GUARD({ KThreadLocalPage::Free(kernel, tlp); });
 
-    std::memset(tls_page_ptr, 0, PageSize);
-    tls_pages.emplace_back(tls_page_addr);
+    // Initialize the new page.
+    R_TRY(tlp->Initialize(kernel, this));
 
-    const auto reserve_result{tls_pages.back().ReserveSlot()};
-    ASSERT(reserve_result.has_value());
+    // Reserve a TLR.
+    tlr = tlp->Reserve();
+    ASSERT(tlr != 0);
 
-    return *reserve_result;
+    // Insert into our tree.
+    {
+        KScopedSchedulerLock sl{kernel};
+        if (tlp->IsAllUsed()) {
+            fully_used_tlp_tree.insert(*tlp);
+        } else {
+            partially_used_tlp_tree.insert(*tlp);
+        }
+    }
+
+    // We succeeded!
+    tlp_guard.Cancel();
+    *out = tlr;
+    return ResultSuccess;
 }
 
-void KProcess::FreeTLSRegion(VAddr tls_address) {
-    KScopedSchedulerLock lock(kernel);
-    const VAddr aligned_address = Common::AlignDown(tls_address, Core::Memory::PAGE_SIZE);
-    auto iter =
-        std::find_if(tls_pages.begin(), tls_pages.end(), [aligned_address](const auto& page) {
-            return page.GetBaseAddress() == aligned_address;
-        });
+ResultCode KProcess::DeleteThreadLocalRegion(VAddr addr) {
+    KThreadLocalPage* page_to_free = nullptr;
+
+    // Release the region.
+    {
+        KScopedSchedulerLock sl{kernel};
+
+        // Try to find the page in the partially used list.
+        auto it = partially_used_tlp_tree.find_key(Common::AlignDown(addr, PageSize));
+        if (it == partially_used_tlp_tree.end()) {
+            // If we don't find it, it has to be in the fully used list.
+            it = fully_used_tlp_tree.find_key(Common::AlignDown(addr, PageSize));
+            R_UNLESS(it != fully_used_tlp_tree.end(), ResultInvalidAddress);
+
+            // Release the region.
+            it->Release(addr);
+
+            // Move the page out of the fully used list.
+            KThreadLocalPage* tlp = std::addressof(*it);
+            fully_used_tlp_tree.erase(it);
+            if (tlp->IsAllFree()) {
+                page_to_free = tlp;
+            } else {
+                partially_used_tlp_tree.insert(*tlp);
+            }
+        } else {
+            // Release the region.
+            it->Release(addr);
+
+            // Handle the all-free case.
+            KThreadLocalPage* tlp = std::addressof(*it);
+            if (tlp->IsAllFree()) {
+                partially_used_tlp_tree.erase(it);
+                page_to_free = tlp;
+            }
+        }
+    }
+
+    // If we should free the page it was in, do so.
+    if (page_to_free != nullptr) {
+        page_to_free->Finalize();
 
-    // Something has gone very wrong if we're freeing a region
-    // with no actual page available.
-    ASSERT(iter != tls_pages.cend());
+        KThreadLocalPage::Free(kernel, page_to_free);
+    }
 
-    iter->ReleaseSlot(tls_address);
+    return ResultSuccess;
 }
 
 void KProcess::LoadModule(CodeSet code_set, VAddr base_addr) {
diff --git a/src/core/hle/kernel/k_process.h b/src/core/hle/kernel/k_process.h
index 38b446350..5ed0f2d83 100644
--- a/src/core/hle/kernel/k_process.h
+++ b/src/core/hle/kernel/k_process.h
@@ -15,6 +15,7 @@
 #include "core/hle/kernel/k_condition_variable.h"
 #include "core/hle/kernel/k_handle_table.h"
 #include "core/hle/kernel/k_synchronization_object.h"
+#include "core/hle/kernel/k_thread_local_page.h"
 #include "core/hle/kernel/k_worker_task.h"
 #include "core/hle/kernel/process_capability.h"
 #include "core/hle/kernel/slab_helpers.h"
@@ -91,7 +92,7 @@ public:
     static constexpr std::size_t RANDOM_ENTROPY_SIZE = 4;
 
     static ResultCode Initialize(KProcess* process, Core::System& system, std::string process_name,
-                                 ProcessType type);
+                                 ProcessType type, KResourceLimit* res_limit);
 
     /// Gets a reference to the process' page table.
     KPageTable& PageTable() {
@@ -362,10 +363,10 @@ public:
     // Thread-local storage management
 
     // Marks the next available region as used and returns the address of the slot.
-    [[nodiscard]] VAddr CreateTLSRegion();
+    [[nodiscard]] ResultCode CreateThreadLocalRegion(VAddr* out);
 
     // Frees a used TLS slot identified by the given address
-    void FreeTLSRegion(VAddr tls_address);
+    ResultCode DeleteThreadLocalRegion(VAddr addr);
 
 private:
     void PinThread(s32 core_id, KThread* thread) {
@@ -413,13 +414,6 @@ private:
     /// The ideal CPU core for this process, threads are scheduled on this core by default.
     u8 ideal_core = 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
-    /// page as a bitmask.
-    /// This vector will grow as more pages are allocated for new threads.
-    std::vector<TLSPage> tls_pages;
-
     /// Contains the parsed process capability descriptors.
     ProcessCapabilities capabilities;
 
@@ -482,6 +476,12 @@ private:
     KThread* exception_thread{};
 
     KLightLock state_lock;
+
+    using TLPTree =
+        Common::IntrusiveRedBlackTreeBaseTraits<KThreadLocalPage>::TreeType<KThreadLocalPage>;
+    using TLPIterator = TLPTree::iterator;
+    TLPTree fully_used_tlp_tree;
+    TLPTree partially_used_tlp_tree;
 };
 
 } // namespace Kernel
diff --git a/src/core/hle/kernel/k_resource_limit.cpp b/src/core/hle/kernel/k_resource_limit.cpp
index 0c4bba66b..a84977c68 100644
--- a/src/core/hle/kernel/k_resource_limit.cpp
+++ b/src/core/hle/kernel/k_resource_limit.cpp
@@ -3,6 +3,7 @@
 // Refer to the license.txt file included.
 
 #include "common/assert.h"
+#include "core/core.h"
 #include "core/core_timing.h"
 #include "core/hle/kernel/k_resource_limit.h"
 #include "core/hle/kernel/svc_results.h"
@@ -151,4 +152,22 @@ void KResourceLimit::Release(LimitableResource which, s64 value, s64 hint) {
     }
 }
 
+KResourceLimit* CreateResourceLimitForProcess(Core::System& system, s64 physical_memory_size) {
+    auto* resource_limit = KResourceLimit::Create(system.Kernel());
+    resource_limit->Initialize(&system.CoreTiming());
+
+    // Initialize default resource limit values.
+    // TODO(bunnei): These values are the system defaults, the limits for service processes are
+    // lower. These should use the correct limit values.
+
+    ASSERT(resource_limit->SetLimitValue(LimitableResource::PhysicalMemory, physical_memory_size)
+               .IsSuccess());
+    ASSERT(resource_limit->SetLimitValue(LimitableResource::Threads, 800).IsSuccess());
+    ASSERT(resource_limit->SetLimitValue(LimitableResource::Events, 900).IsSuccess());
+    ASSERT(resource_limit->SetLimitValue(LimitableResource::TransferMemory, 200).IsSuccess());
+    ASSERT(resource_limit->SetLimitValue(LimitableResource::Sessions, 1133).IsSuccess());
+
+    return resource_limit;
+}
+
 } // namespace Kernel
diff --git a/src/core/hle/kernel/k_resource_limit.h b/src/core/hle/kernel/k_resource_limit.h
index fab6005ff..d23d16aa4 100644
--- a/src/core/hle/kernel/k_resource_limit.h
+++ b/src/core/hle/kernel/k_resource_limit.h
@@ -67,4 +67,7 @@ private:
     KLightConditionVariable cond_var;
     const Core::Timing::CoreTiming* core_timing{};
 };
+
+KResourceLimit* CreateResourceLimitForProcess(Core::System& system, s64 physical_memory_size);
+
 } // namespace Kernel
diff --git a/src/core/hle/kernel/k_server_port.h b/src/core/hle/kernel/k_server_port.h
index 6302d5e61..2185736be 100644
--- a/src/core/hle/kernel/k_server_port.h
+++ b/src/core/hle/kernel/k_server_port.h
@@ -30,11 +30,11 @@ public:
 
     /// Whether or not this server port has an HLE handler available.
     bool HasSessionRequestHandler() const {
-        return session_handler != nullptr;
+        return !session_handler.expired();
     }
 
     /// Gets the HLE handler for this port.
-    SessionRequestHandlerPtr GetSessionRequestHandler() const {
+    SessionRequestHandlerWeakPtr GetSessionRequestHandler() const {
         return session_handler;
     }
 
@@ -42,7 +42,7 @@ public:
      * Sets the HLE handler template for the port. ServerSessions crated by connecting to this port
      * will inherit a reference to this handler.
      */
-    void SetSessionHandler(SessionRequestHandlerPtr&& handler) {
+    void SetSessionHandler(SessionRequestHandlerWeakPtr&& handler) {
         session_handler = std::move(handler);
     }
 
@@ -66,7 +66,7 @@ private:
     void CleanupSessions();
 
     SessionList session_list;
-    SessionRequestHandlerPtr session_handler;
+    SessionRequestHandlerWeakPtr session_handler;
     KPort* parent{};
 };
 
diff --git a/src/core/hle/kernel/k_server_session.cpp b/src/core/hle/kernel/k_server_session.cpp
index 4d94eb9cf..30c56ff29 100644
--- a/src/core/hle/kernel/k_server_session.cpp
+++ b/src/core/hle/kernel/k_server_session.cpp
@@ -27,10 +27,7 @@ namespace Kernel {
 
 KServerSession::KServerSession(KernelCore& kernel_) : KSynchronizationObject{kernel_} {}
 
-KServerSession::~KServerSession() {
-    // Ensure that the global list tracking server sessions does not hold on to a reference.
-    kernel.UnregisterServerSession(this);
-}
+KServerSession::~KServerSession() = default;
 
 void KServerSession::Initialize(KSession* parent_session_, std::string&& name_,
                                 std::shared_ptr<SessionRequestManager> manager_) {
@@ -49,6 +46,9 @@ void KServerSession::Destroy() {
     parent->OnServerClosed();
 
     parent->Close();
+
+    // Release host emulation members.
+    manager.reset();
 }
 
 void KServerSession::OnClientClosed() {
@@ -98,7 +98,12 @@ ResultCode KServerSession::HandleDomainSyncRequest(Kernel::HLERequestContext& co
             UNREACHABLE();
             return ResultSuccess; // Ignore error if asserts are off
         }
-        return manager->DomainHandler(object_id - 1)->HandleSyncRequest(*this, context);
+        if (auto strong_ptr = manager->DomainHandler(object_id - 1).lock()) {
+            return strong_ptr->HandleSyncRequest(*this, context);
+        } else {
+            UNREACHABLE();
+            return ResultSuccess;
+        }
 
     case IPC::DomainMessageHeader::CommandType::CloseVirtualHandle: {
         LOG_DEBUG(IPC, "CloseVirtualHandle, object_id=0x{:08X}", object_id);
diff --git a/src/core/hle/kernel/k_slab_heap.h b/src/core/hle/kernel/k_slab_heap.h
index 0ad74b0a0..5690cc757 100644
--- a/src/core/hle/kernel/k_slab_heap.h
+++ b/src/core/hle/kernel/k_slab_heap.h
@@ -7,6 +7,7 @@
 #include <atomic>
 
 #include "common/assert.h"
+#include "common/common_funcs.h"
 #include "common/common_types.h"
 
 namespace Kernel {
@@ -15,35 +16,34 @@ class KernelCore;
 
 namespace impl {
 
-class KSlabHeapImpl final : NonCopyable {
+class KSlabHeapImpl {
+    YUZU_NON_COPYABLE(KSlabHeapImpl);
+    YUZU_NON_MOVEABLE(KSlabHeapImpl);
+
 public:
     struct Node {
         Node* next{};
     };
 
+public:
     constexpr KSlabHeapImpl() = default;
 
-    void Initialize(std::size_t size) {
-        ASSERT(head == nullptr);
-        obj_size = size;
-    }
-
-    constexpr std::size_t GetObjectSize() const {
-        return obj_size;
+    void Initialize() {
+        ASSERT(m_head == nullptr);
     }
 
     Node* GetHead() const {
-        return head;
+        return m_head;
     }
 
     void* Allocate() {
-        Node* ret = head.load();
+        Node* ret = m_head.load();
 
         do {
             if (ret == nullptr) {
                 break;
             }
-        } while (!head.compare_exchange_weak(ret, ret->next));
+        } while (!m_head.compare_exchange_weak(ret, ret->next));
 
         return ret;
     }
@@ -51,166 +51,157 @@ public:
     void Free(void* obj) {
         Node* node = static_cast<Node*>(obj);
 
-        Node* cur_head = head.load();
+        Node* cur_head = m_head.load();
         do {
             node->next = cur_head;
-        } while (!head.compare_exchange_weak(cur_head, node));
+        } while (!m_head.compare_exchange_weak(cur_head, node));
     }
 
 private:
-    std::atomic<Node*> head{};
-    std::size_t obj_size{};
+    std::atomic<Node*> m_head{};
 };
 
 } // namespace impl
 
-class KSlabHeapBase : NonCopyable {
+template <bool SupportDynamicExpansion>
+class KSlabHeapBase : protected impl::KSlabHeapImpl {
+    YUZU_NON_COPYABLE(KSlabHeapBase);
+    YUZU_NON_MOVEABLE(KSlabHeapBase);
+
+private:
+    size_t m_obj_size{};
+    uintptr_t m_peak{};
+    uintptr_t m_start{};
+    uintptr_t m_end{};
+
+private:
+    void UpdatePeakImpl(uintptr_t obj) {
+        static_assert(std::atomic_ref<uintptr_t>::is_always_lock_free);
+        std::atomic_ref<uintptr_t> peak_ref(m_peak);
+
+        const uintptr_t alloc_peak = obj + this->GetObjectSize();
+        uintptr_t cur_peak = m_peak;
+        do {
+            if (alloc_peak <= cur_peak) {
+                break;
+            }
+        } while (!peak_ref.compare_exchange_strong(cur_peak, alloc_peak));
+    }
+
 public:
     constexpr KSlabHeapBase() = default;
 
-    constexpr bool Contains(uintptr_t addr) const {
-        return start <= addr && addr < end;
+    bool Contains(uintptr_t address) const {
+        return m_start <= address && address < m_end;
     }
 
-    constexpr std::size_t GetSlabHeapSize() const {
-        return (end - start) / GetObjectSize();
-    }
+    void Initialize(size_t obj_size, void* memory, size_t memory_size) {
+        // Ensure we don't initialize a slab using null memory.
+        ASSERT(memory != nullptr);
 
-    constexpr std::size_t GetObjectSize() const {
-        return impl.GetObjectSize();
-    }
+        // Set our object size.
+        m_obj_size = obj_size;
 
-    constexpr uintptr_t GetSlabHeapAddress() const {
-        return start;
-    }
+        // Initialize the base allocator.
+        KSlabHeapImpl::Initialize();
+
+        // Set our tracking variables.
+        const size_t num_obj = (memory_size / obj_size);
+        m_start = reinterpret_cast<uintptr_t>(memory);
+        m_end = m_start + num_obj * obj_size;
+        m_peak = m_start;
 
-    std::size_t GetObjectIndexImpl(const void* obj) const {
-        return (reinterpret_cast<uintptr_t>(obj) - start) / GetObjectSize();
+        // Free the objects.
+        u8* cur = reinterpret_cast<u8*>(m_end);
+
+        for (size_t i = 0; i < num_obj; i++) {
+            cur -= obj_size;
+            KSlabHeapImpl::Free(cur);
+        }
     }
 
-    std::size_t GetPeakIndex() const {
-        return GetObjectIndexImpl(reinterpret_cast<const void*>(peak));
+    size_t GetSlabHeapSize() const {
+        return (m_end - m_start) / this->GetObjectSize();
     }
 
-    void* AllocateImpl() {
-        return impl.Allocate();
+    size_t GetObjectSize() const {
+        return m_obj_size;
     }
 
-    void FreeImpl(void* obj) {
-        // Don't allow freeing an object that wasn't allocated from this heap
-        ASSERT(Contains(reinterpret_cast<uintptr_t>(obj)));
+    void* Allocate() {
+        void* obj = KSlabHeapImpl::Allocate();
 
-        impl.Free(obj);
+        return obj;
     }
 
-    void InitializeImpl(std::size_t obj_size, void* memory, std::size_t memory_size) {
-        // Ensure we don't initialize a slab using null memory
-        ASSERT(memory != nullptr);
-
-        // Initialize the base allocator
-        impl.Initialize(obj_size);
+    void Free(void* obj) {
+        // Don't allow freeing an object that wasn't allocated from this heap.
+        const bool contained = this->Contains(reinterpret_cast<uintptr_t>(obj));
+        ASSERT(contained);
+        KSlabHeapImpl::Free(obj);
+    }
 
-        // Set our tracking variables
-        const std::size_t num_obj = (memory_size / obj_size);
-        start = reinterpret_cast<uintptr_t>(memory);
-        end = start + num_obj * obj_size;
-        peak = start;
+    size_t GetObjectIndex(const void* obj) const {
+        if constexpr (SupportDynamicExpansion) {
+            if (!this->Contains(reinterpret_cast<uintptr_t>(obj))) {
+                return std::numeric_limits<size_t>::max();
+            }
+        }
 
-        // Free the objects
-        u8* cur = reinterpret_cast<u8*>(end);
+        return (reinterpret_cast<uintptr_t>(obj) - m_start) / this->GetObjectSize();
+    }
 
-        for (std::size_t i{}; i < num_obj; i++) {
-            cur -= obj_size;
-            impl.Free(cur);
-        }
+    size_t GetPeakIndex() const {
+        return this->GetObjectIndex(reinterpret_cast<const void*>(m_peak));
     }
 
-private:
-    using Impl = impl::KSlabHeapImpl;
+    uintptr_t GetSlabHeapAddress() const {
+        return m_start;
+    }
 
-    Impl impl;
-    uintptr_t peak{};
-    uintptr_t start{};
-    uintptr_t end{};
+    size_t GetNumRemaining() const {
+        // Only calculate the number of remaining objects under debug configuration.
+        return 0;
+    }
 };
 
 template <typename T>
-class KSlabHeap final : public KSlabHeapBase {
-public:
-    enum class AllocationType {
-        Host,
-        Guest,
-    };
+class KSlabHeap final : public KSlabHeapBase<false> {
+private:
+    using BaseHeap = KSlabHeapBase<false>;
 
-    explicit constexpr KSlabHeap(AllocationType allocation_type_ = AllocationType::Host)
-        : KSlabHeapBase(), allocation_type{allocation_type_} {}
+public:
+    constexpr KSlabHeap() = default;
 
-    void Initialize(void* memory, std::size_t memory_size) {
-        if (allocation_type == AllocationType::Guest) {
-            InitializeImpl(sizeof(T), memory, memory_size);
-        }
+    void Initialize(void* memory, size_t memory_size) {
+        BaseHeap::Initialize(sizeof(T), memory, memory_size);
     }
 
     T* Allocate() {
-        switch (allocation_type) {
-        case AllocationType::Host:
-            // Fallback for cases where we do not yet support allocating guest memory from the slab
-            // heap, such as for kernel memory regions.
-            return new T;
-
-        case AllocationType::Guest:
-            T* obj = static_cast<T*>(AllocateImpl());
-            if (obj != nullptr) {
-                new (obj) T();
-            }
-            return obj;
-        }
+        T* obj = static_cast<T*>(BaseHeap::Allocate());
 
-        UNREACHABLE_MSG("Invalid AllocationType {}", allocation_type);
-        return nullptr;
+        if (obj != nullptr) [[likely]] {
+            std::construct_at(obj);
+        }
+        return obj;
     }
 
-    T* AllocateWithKernel(KernelCore& kernel) {
-        switch (allocation_type) {
-        case AllocationType::Host:
-            // Fallback for cases where we do not yet support allocating guest memory from the slab
-            // heap, such as for kernel memory regions.
-            return new T(kernel);
+    T* Allocate(KernelCore& kernel) {
+        T* obj = static_cast<T*>(BaseHeap::Allocate());
 
-        case AllocationType::Guest:
-            T* obj = static_cast<T*>(AllocateImpl());
-            if (obj != nullptr) {
-                new (obj) T(kernel);
-            }
-            return obj;
+        if (obj != nullptr) [[likely]] {
+            std::construct_at(obj, kernel);
         }
-
-        UNREACHABLE_MSG("Invalid AllocationType {}", allocation_type);
-        return nullptr;
+        return obj;
     }
 
     void Free(T* obj) {
-        switch (allocation_type) {
-        case AllocationType::Host:
-            // Fallback for cases where we do not yet support allocating guest memory from the slab
-            // heap, such as for kernel memory regions.
-            delete obj;
-            return;
-
-        case AllocationType::Guest:
-            FreeImpl(obj);
-            return;
-        }
-
-        UNREACHABLE_MSG("Invalid AllocationType {}", allocation_type);
+        BaseHeap::Free(obj);
     }
 
-    constexpr std::size_t GetObjectIndex(const T* obj) const {
-        return GetObjectIndexImpl(obj);
+    size_t GetObjectIndex(const T* obj) const {
+        return BaseHeap::GetObjectIndex(obj);
     }
-
-private:
-    const AllocationType allocation_type;
 };
 
 } // namespace Kernel
diff --git a/src/core/hle/kernel/k_thread.cpp b/src/core/hle/kernel/k_thread.cpp
index de3ffe0c7..ba7f72c6b 100644
--- a/src/core/hle/kernel/k_thread.cpp
+++ b/src/core/hle/kernel/k_thread.cpp
@@ -210,7 +210,7 @@ ResultCode KThread::Initialize(KThreadFunction func, uintptr_t arg, VAddr user_s
     if (owner != nullptr) {
         // Setup the TLS, if needed.
         if (type == ThreadType::User) {
-            tls_address = owner->CreateTLSRegion();
+            R_TRY(owner->CreateThreadLocalRegion(std::addressof(tls_address)));
         }
 
         parent = owner;
@@ -305,7 +305,7 @@ void KThread::Finalize() {
 
     // If the thread has a local region, delete it.
     if (tls_address != 0) {
-        parent->FreeTLSRegion(tls_address);
+        ASSERT(parent->DeleteThreadLocalRegion(tls_address).IsSuccess());
     }
 
     // Release any waiters.
@@ -326,6 +326,9 @@ void KThread::Finalize() {
         }
     }
 
+    // Release host emulation members.
+    host_context.reset();
+
     // Perform inherited finalization.
     KSynchronizationObject::Finalize();
 }
diff --git a/src/core/hle/kernel/k_thread.h b/src/core/hle/kernel/k_thread.h
index d058db62c..f46db7298 100644
--- a/src/core/hle/kernel/k_thread.h
+++ b/src/core/hle/kernel/k_thread.h
@@ -656,7 +656,7 @@ private:
     static_assert(sizeof(SyncObjectBuffer::sync_objects) == sizeof(SyncObjectBuffer::handles));
 
     struct ConditionVariableComparator {
-        struct LightCompareType {
+        struct RedBlackKeyType {
             u64 cv_key{};
             s32 priority{};
 
@@ -672,8 +672,8 @@ private:
         template <typename T>
         requires(
             std::same_as<T, KThread> ||
-            std::same_as<T, LightCompareType>) static constexpr int Compare(const T& lhs,
-                                                                            const KThread& rhs) {
+            std::same_as<T, RedBlackKeyType>) static constexpr int Compare(const T& lhs,
+                                                                           const KThread& rhs) {
             const u64 l_key = lhs.GetConditionVariableKey();
             const u64 r_key = rhs.GetConditionVariableKey();
 
diff --git a/src/core/hle/kernel/k_thread_local_page.cpp b/src/core/hle/kernel/k_thread_local_page.cpp
new file mode 100644
index 000000000..17b233fca
--- /dev/null
+++ b/src/core/hle/kernel/k_thread_local_page.cpp
@@ -0,0 +1,66 @@
+// Copyright 2022 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include "common/scope_exit.h"
+#include "core/core.h"
+#include "core/hle/kernel/k_memory_block.h"
+#include "core/hle/kernel/k_page_table.h"
+#include "core/hle/kernel/k_process.h"
+#include "core/hle/kernel/k_thread_local_page.h"
+#include "core/hle/kernel/kernel.h"
+
+namespace Kernel {
+
+ResultCode KThreadLocalPage::Initialize(KernelCore& kernel, KProcess* process) {
+    // Set that this process owns us.
+    m_owner = process;
+    m_kernel = &kernel;
+
+    // Allocate a new page.
+    KPageBuffer* page_buf = KPageBuffer::Allocate(kernel);
+    R_UNLESS(page_buf != nullptr, ResultOutOfMemory);
+    auto page_buf_guard = SCOPE_GUARD({ KPageBuffer::Free(kernel, page_buf); });
+
+    // Map the address in.
+    const auto phys_addr = kernel.System().DeviceMemory().GetPhysicalAddr(page_buf);
+    R_TRY(m_owner->PageTable().MapPages(std::addressof(m_virt_addr), 1, PageSize, phys_addr,
+                                        KMemoryState::ThreadLocal,
+                                        KMemoryPermission::UserReadWrite));
+
+    // We succeeded.
+    page_buf_guard.Cancel();
+
+    return ResultSuccess;
+}
+
+ResultCode KThreadLocalPage::Finalize() {
+    // Get the physical address of the page.
+    const PAddr phys_addr = m_owner->PageTable().GetPhysicalAddr(m_virt_addr);
+    ASSERT(phys_addr);
+
+    // Unmap the page.
+    R_TRY(m_owner->PageTable().UnmapPages(this->GetAddress(), 1, KMemoryState::ThreadLocal));
+
+    // Free the page.
+    KPageBuffer::Free(*m_kernel, KPageBuffer::FromPhysicalAddress(m_kernel->System(), phys_addr));
+
+    return ResultSuccess;
+}
+
+VAddr KThreadLocalPage::Reserve() {
+    for (size_t i = 0; i < m_is_region_free.size(); i++) {
+        if (m_is_region_free[i]) {
+            m_is_region_free[i] = false;
+            return this->GetRegionAddress(i);
+        }
+    }
+
+    return 0;
+}
+
+void KThreadLocalPage::Release(VAddr addr) {
+    m_is_region_free[this->GetRegionIndex(addr)] = true;
+}
+
+} // namespace Kernel
diff --git a/src/core/hle/kernel/k_thread_local_page.h b/src/core/hle/kernel/k_thread_local_page.h
new file mode 100644
index 000000000..658c67e94
--- /dev/null
+++ b/src/core/hle/kernel/k_thread_local_page.h
@@ -0,0 +1,112 @@
+// Copyright 2022 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include <algorithm>
+#include <array>
+
+#include "common/alignment.h"
+#include "common/assert.h"
+#include "common/common_types.h"
+#include "common/intrusive_red_black_tree.h"
+#include "core/hle/kernel/k_page_buffer.h"
+#include "core/hle/kernel/memory_types.h"
+#include "core/hle/kernel/slab_helpers.h"
+#include "core/hle/result.h"
+
+namespace Kernel {
+
+class KernelCore;
+class KProcess;
+
+class KThreadLocalPage final : public Common::IntrusiveRedBlackTreeBaseNode<KThreadLocalPage>,
+                               public KSlabAllocated<KThreadLocalPage> {
+public:
+    static constexpr size_t RegionsPerPage = PageSize / Svc::ThreadLocalRegionSize;
+    static_assert(RegionsPerPage > 0);
+
+public:
+    constexpr explicit KThreadLocalPage(VAddr addr = {}) : m_virt_addr(addr) {
+        m_is_region_free.fill(true);
+    }
+
+    constexpr VAddr GetAddress() const {
+        return m_virt_addr;
+    }
+
+    ResultCode Initialize(KernelCore& kernel, KProcess* process);
+    ResultCode Finalize();
+
+    VAddr Reserve();
+    void Release(VAddr addr);
+
+    bool IsAllUsed() const {
+        return std::ranges::all_of(m_is_region_free.begin(), m_is_region_free.end(),
+                                   [](bool is_free) { return !is_free; });
+    }
+
+    bool IsAllFree() const {
+        return std::ranges::all_of(m_is_region_free.begin(), m_is_region_free.end(),
+                                   [](bool is_free) { return is_free; });
+    }
+
+    bool IsAnyUsed() const {
+        return !this->IsAllFree();
+    }
+
+    bool IsAnyFree() const {
+        return !this->IsAllUsed();
+    }
+
+public:
+    using RedBlackKeyType = VAddr;
+
+    static constexpr RedBlackKeyType GetRedBlackKey(const RedBlackKeyType& v) {
+        return v;
+    }
+    static constexpr RedBlackKeyType GetRedBlackKey(const KThreadLocalPage& v) {
+        return v.GetAddress();
+    }
+
+    template <typename T>
+    requires(std::same_as<T, KThreadLocalPage> ||
+             std::same_as<T, RedBlackKeyType>) static constexpr int Compare(const T& lhs,
+                                                                            const KThreadLocalPage&
+                                                                                rhs) {
+        const VAddr lval = GetRedBlackKey(lhs);
+        const VAddr rval = GetRedBlackKey(rhs);
+
+        if (lval < rval) {
+            return -1;
+        } else if (lval == rval) {
+            return 0;
+        } else {
+            return 1;
+        }
+    }
+
+private:
+    constexpr VAddr GetRegionAddress(size_t i) const {
+        return this->GetAddress() + i * Svc::ThreadLocalRegionSize;
+    }
+
+    constexpr bool Contains(VAddr addr) const {
+        return this->GetAddress() <= addr && addr < this->GetAddress() + PageSize;
+    }
+
+    constexpr size_t GetRegionIndex(VAddr addr) const {
+        ASSERT(Common::IsAligned(addr, Svc::ThreadLocalRegionSize));
+        ASSERT(this->Contains(addr));
+        return (addr - this->GetAddress()) / Svc::ThreadLocalRegionSize;
+    }
+
+private:
+    VAddr m_virt_addr{};
+    KProcess* m_owner{};
+    KernelCore* m_kernel{};
+    std::array<bool, RegionsPerPage> m_is_region_free{};
+};
+
+} // namespace Kernel
diff --git a/src/core/hle/kernel/kernel.cpp b/src/core/hle/kernel/kernel.cpp
index 49c0714ed..f9828bc43 100644
--- a/src/core/hle/kernel/kernel.cpp
+++ b/src/core/hle/kernel/kernel.cpp
@@ -52,7 +52,7 @@ namespace Kernel {
 
 struct KernelCore::Impl {
     explicit Impl(Core::System& system_, KernelCore& kernel_)
-        : time_manager{system_}, object_list_container{kernel_},
+        : time_manager{system_},
           service_threads_manager{1, "yuzu:ServiceThreadsManager"}, system{system_} {}
 
     void SetMulticore(bool is_multi) {
@@ -60,6 +60,7 @@ struct KernelCore::Impl {
     }
 
     void Initialize(KernelCore& kernel) {
+        global_object_list_container = std::make_unique<KAutoObjectWithListContainer>(kernel);
         global_scheduler_context = std::make_unique<Kernel::GlobalSchedulerContext>(kernel);
         global_handle_table = std::make_unique<Kernel::KHandleTable>(kernel);
         global_handle_table->Initialize(KHandleTable::MaxTableSize);
@@ -70,14 +71,13 @@ struct KernelCore::Impl {
 
         // Derive the initial memory layout from the emulated board
         Init::InitializeSlabResourceCounts(kernel);
-        KMemoryLayout memory_layout;
-        DeriveInitialMemoryLayout(memory_layout);
-        Init::InitializeSlabHeaps(system, memory_layout);
+        DeriveInitialMemoryLayout();
+        Init::InitializeSlabHeaps(system, *memory_layout);
 
         // Initialize kernel memory and resources.
-        InitializeSystemResourceLimit(kernel, system.CoreTiming(), memory_layout);
-        InitializeMemoryLayout(memory_layout);
-        InitializePageSlab();
+        InitializeSystemResourceLimit(kernel, system.CoreTiming());
+        InitializeMemoryLayout();
+        Init::InitializeKPageBufferSlabHeap(system);
         InitializeSchedulers();
         InitializeSuspendThreads();
         InitializePreemption(kernel);
@@ -108,19 +108,6 @@ struct KernelCore::Impl {
         for (auto* server_port : server_ports_) {
             server_port->Close();
         }
-        // Close all open server sessions.
-        std::unordered_set<KServerSession*> server_sessions_;
-        {
-            std::lock_guard lk(server_sessions_lock);
-            server_sessions_ = server_sessions;
-            server_sessions.clear();
-        }
-        for (auto* server_session : server_sessions_) {
-            server_session->Close();
-        }
-
-        // Ensure that the object list container is finalized and properly shutdown.
-        object_list_container.Finalize();
 
         // Ensures all service threads gracefully shutdown.
         ClearServiceThreads();
@@ -195,11 +182,15 @@ struct KernelCore::Impl {
         {
             std::lock_guard lk(registered_objects_lock);
             if (registered_objects.size()) {
-                LOG_WARNING(Kernel, "{} kernel objects were dangling on shutdown!",
-                            registered_objects.size());
+                LOG_DEBUG(Kernel, "{} kernel objects were dangling on shutdown!",
+                          registered_objects.size());
                 registered_objects.clear();
             }
         }
+
+        // Ensure that the object list container is finalized and properly shutdown.
+        global_object_list_container->Finalize();
+        global_object_list_container.reset();
     }
 
     void InitializePhysicalCores() {
@@ -219,12 +210,11 @@ struct KernelCore::Impl {
 
     // Creates the default system resource limit
     void InitializeSystemResourceLimit(KernelCore& kernel,
-                                       const Core::Timing::CoreTiming& core_timing,
-                                       const KMemoryLayout& memory_layout) {
+                                       const Core::Timing::CoreTiming& core_timing) {
         system_resource_limit = KResourceLimit::Create(system.Kernel());
         system_resource_limit->Initialize(&core_timing);
 
-        const auto [total_size, kernel_size] = memory_layout.GetTotalAndKernelMemorySizes();
+        const auto [total_size, kernel_size] = memory_layout->GetTotalAndKernelMemorySizes();
 
         // If setting the default system values fails, then something seriously wrong has occurred.
         ASSERT(system_resource_limit->SetLimitValue(LimitableResource::PhysicalMemory, total_size)
@@ -240,13 +230,6 @@ struct KernelCore::Impl {
         constexpr u64 secure_applet_memory_size{4_MiB};
         ASSERT(system_resource_limit->Reserve(LimitableResource::PhysicalMemory,
                                               secure_applet_memory_size));
-
-        // This memory seems to be reserved on hardware, but is not reserved/used by yuzu.
-        // Likely Horizon OS reserved memory
-        // TODO(ameerj): Derive the memory rather than hardcode it.
-        constexpr u64 unknown_reserved_memory{0x2f896000};
-        ASSERT(system_resource_limit->Reserve(LimitableResource::PhysicalMemory,
-                                              unknown_reserved_memory));
     }
 
     void InitializePreemption(KernelCore& kernel) {
@@ -300,15 +283,16 @@ struct KernelCore::Impl {
 
     // Gets the dummy KThread for the caller, allocating a new one if this is the first time
     KThread* GetHostDummyThread() {
-        auto make_thread = [this]() {
-            KThread* thread = KThread::Create(system.Kernel());
+        auto initialize = [this](KThread* thread) {
             ASSERT(KThread::InitializeDummyThread(thread).IsSuccess());
             thread->SetName(fmt::format("DummyThread:{}", GetHostThreadId()));
             return thread;
         };
 
-        thread_local KThread* saved_thread = make_thread();
-        return saved_thread;
+        thread_local auto raw_thread = KThread(system.Kernel());
+        thread_local auto thread = initialize(&raw_thread);
+
+        return thread;
     }
 
     /// Registers a CPU core thread by allocating a host thread ID for it
@@ -360,16 +344,18 @@ struct KernelCore::Impl {
         return schedulers[thread_id]->GetCurrentThread();
     }
 
-    void DeriveInitialMemoryLayout(KMemoryLayout& memory_layout) {
+    void DeriveInitialMemoryLayout() {
+        memory_layout = std::make_unique<KMemoryLayout>();
+
         // Insert the root region for the virtual memory tree, from which all other regions will
         // derive.
-        memory_layout.GetVirtualMemoryRegionTree().InsertDirectly(
+        memory_layout->GetVirtualMemoryRegionTree().InsertDirectly(
             KernelVirtualAddressSpaceBase,
             KernelVirtualAddressSpaceBase + KernelVirtualAddressSpaceSize - 1);
 
         // Insert the root region for the physical memory tree, from which all other regions will
         // derive.
-        memory_layout.GetPhysicalMemoryRegionTree().InsertDirectly(
+        memory_layout->GetPhysicalMemoryRegionTree().InsertDirectly(
             KernelPhysicalAddressSpaceBase,
             KernelPhysicalAddressSpaceBase + KernelPhysicalAddressSpaceSize - 1);
 
@@ -386,7 +372,7 @@ struct KernelCore::Impl {
         if (!(kernel_region_start + KernelRegionSize - 1 <= KernelVirtualAddressSpaceLast)) {
             kernel_region_size = KernelVirtualAddressSpaceEnd - kernel_region_start;
         }
-        ASSERT(memory_layout.GetVirtualMemoryRegionTree().Insert(
+        ASSERT(memory_layout->GetVirtualMemoryRegionTree().Insert(
             kernel_region_start, kernel_region_size, KMemoryRegionType_Kernel));
 
         // Setup the code region.
@@ -395,11 +381,11 @@ struct KernelCore::Impl {
             Common::AlignDown(code_start_virt_addr, CodeRegionAlign);
         constexpr VAddr code_region_end = Common::AlignUp(code_end_virt_addr, CodeRegionAlign);
         constexpr size_t code_region_size = code_region_end - code_region_start;
-        ASSERT(memory_layout.GetVirtualMemoryRegionTree().Insert(
+        ASSERT(memory_layout->GetVirtualMemoryRegionTree().Insert(
             code_region_start, code_region_size, KMemoryRegionType_KernelCode));
 
         // Setup board-specific device physical regions.
-        Init::SetupDevicePhysicalMemoryRegions(memory_layout);
+        Init::SetupDevicePhysicalMemoryRegions(*memory_layout);
 
         // Determine the amount of space needed for the misc region.
         size_t misc_region_needed_size;
@@ -408,7 +394,7 @@ struct KernelCore::Impl {
             misc_region_needed_size = Core::Hardware::NUM_CPU_CORES * (3 * (PageSize + PageSize));
 
             // Account for each auto-map device.
-            for (const auto& region : memory_layout.GetPhysicalMemoryRegionTree()) {
+            for (const auto& region : memory_layout->GetPhysicalMemoryRegionTree()) {
                 if (region.HasTypeAttribute(KMemoryRegionAttr_ShouldKernelMap)) {
                     // Check that the region is valid.
                     ASSERT(region.GetEndAddress() != 0);
@@ -433,22 +419,22 @@ struct KernelCore::Impl {
 
         // Setup the misc region.
         const VAddr misc_region_start =
-            memory_layout.GetVirtualMemoryRegionTree().GetRandomAlignedRegion(
+            memory_layout->GetVirtualMemoryRegionTree().GetRandomAlignedRegion(
                 misc_region_size, MiscRegionAlign, KMemoryRegionType_Kernel);
-        ASSERT(memory_layout.GetVirtualMemoryRegionTree().Insert(
+        ASSERT(memory_layout->GetVirtualMemoryRegionTree().Insert(
             misc_region_start, misc_region_size, KMemoryRegionType_KernelMisc));
 
         // Setup the stack region.
         constexpr size_t StackRegionSize = 14_MiB;
         constexpr size_t StackRegionAlign = KernelAslrAlignment;
         const VAddr stack_region_start =
-            memory_layout.GetVirtualMemoryRegionTree().GetRandomAlignedRegion(
+            memory_layout->GetVirtualMemoryRegionTree().GetRandomAlignedRegion(
                 StackRegionSize, StackRegionAlign, KMemoryRegionType_Kernel);
-        ASSERT(memory_layout.GetVirtualMemoryRegionTree().Insert(
+        ASSERT(memory_layout->GetVirtualMemoryRegionTree().Insert(
             stack_region_start, StackRegionSize, KMemoryRegionType_KernelStack));
 
         // Determine the size of the resource region.
-        const size_t resource_region_size = memory_layout.GetResourceRegionSizeForInit();
+        const size_t resource_region_size = memory_layout->GetResourceRegionSizeForInit();
 
         // Determine the size of the slab region.
         const size_t slab_region_size =
@@ -465,23 +451,23 @@ struct KernelCore::Impl {
             Common::AlignUp(code_end_phys_addr + slab_region_size, SlabRegionAlign) -
             Common::AlignDown(code_end_phys_addr, SlabRegionAlign);
         const VAddr slab_region_start =
-            memory_layout.GetVirtualMemoryRegionTree().GetRandomAlignedRegion(
+            memory_layout->GetVirtualMemoryRegionTree().GetRandomAlignedRegion(
                 slab_region_needed_size, SlabRegionAlign, KMemoryRegionType_Kernel) +
             (code_end_phys_addr % SlabRegionAlign);
-        ASSERT(memory_layout.GetVirtualMemoryRegionTree().Insert(
+        ASSERT(memory_layout->GetVirtualMemoryRegionTree().Insert(
             slab_region_start, slab_region_size, KMemoryRegionType_KernelSlab));
 
         // Setup the temp region.
         constexpr size_t TempRegionSize = 128_MiB;
         constexpr size_t TempRegionAlign = KernelAslrAlignment;
         const VAddr temp_region_start =
-            memory_layout.GetVirtualMemoryRegionTree().GetRandomAlignedRegion(
+            memory_layout->GetVirtualMemoryRegionTree().GetRandomAlignedRegion(
                 TempRegionSize, TempRegionAlign, KMemoryRegionType_Kernel);
-        ASSERT(memory_layout.GetVirtualMemoryRegionTree().Insert(temp_region_start, TempRegionSize,
-                                                                 KMemoryRegionType_KernelTemp));
+        ASSERT(memory_layout->GetVirtualMemoryRegionTree().Insert(temp_region_start, TempRegionSize,
+                                                                  KMemoryRegionType_KernelTemp));
 
         // Automatically map in devices that have auto-map attributes.
-        for (auto& region : memory_layout.GetPhysicalMemoryRegionTree()) {
+        for (auto& region : memory_layout->GetPhysicalMemoryRegionTree()) {
             // We only care about kernel regions.
             if (!region.IsDerivedFrom(KMemoryRegionType_Kernel)) {
                 continue;
@@ -508,21 +494,21 @@ struct KernelCore::Impl {
             const size_t map_size =
                 Common::AlignUp(region.GetEndAddress(), PageSize) - map_phys_addr;
             const VAddr map_virt_addr =
-                memory_layout.GetVirtualMemoryRegionTree().GetRandomAlignedRegionWithGuard(
+                memory_layout->GetVirtualMemoryRegionTree().GetRandomAlignedRegionWithGuard(
                     map_size, PageSize, KMemoryRegionType_KernelMisc, PageSize);
-            ASSERT(memory_layout.GetVirtualMemoryRegionTree().Insert(
+            ASSERT(memory_layout->GetVirtualMemoryRegionTree().Insert(
                 map_virt_addr, map_size, KMemoryRegionType_KernelMiscMappedDevice));
             region.SetPairAddress(map_virt_addr + region.GetAddress() - map_phys_addr);
         }
 
-        Init::SetupDramPhysicalMemoryRegions(memory_layout);
+        Init::SetupDramPhysicalMemoryRegions(*memory_layout);
 
         // Insert a physical region for the kernel code region.
-        ASSERT(memory_layout.GetPhysicalMemoryRegionTree().Insert(
+        ASSERT(memory_layout->GetPhysicalMemoryRegionTree().Insert(
             code_start_phys_addr, code_region_size, KMemoryRegionType_DramKernelCode));
 
         // Insert a physical region for the kernel slab region.
-        ASSERT(memory_layout.GetPhysicalMemoryRegionTree().Insert(
+        ASSERT(memory_layout->GetPhysicalMemoryRegionTree().Insert(
             slab_start_phys_addr, slab_region_size, KMemoryRegionType_DramKernelSlab));
 
         // Determine size available for kernel page table heaps, requiring > 8 MB.
@@ -531,12 +517,12 @@ struct KernelCore::Impl {
         ASSERT(page_table_heap_size / 4_MiB > 2);
 
         // Insert a physical region for the kernel page table heap region
-        ASSERT(memory_layout.GetPhysicalMemoryRegionTree().Insert(
+        ASSERT(memory_layout->GetPhysicalMemoryRegionTree().Insert(
             slab_end_phys_addr, page_table_heap_size, KMemoryRegionType_DramKernelPtHeap));
 
         // All DRAM regions that we haven't tagged by this point will be mapped under the linear
         // mapping. Tag them.
-        for (auto& region : memory_layout.GetPhysicalMemoryRegionTree()) {
+        for (auto& region : memory_layout->GetPhysicalMemoryRegionTree()) {
             if (region.GetType() == KMemoryRegionType_Dram) {
                 // Check that the region is valid.
                 ASSERT(region.GetEndAddress() != 0);
@@ -548,7 +534,7 @@ struct KernelCore::Impl {
 
         // Get the linear region extents.
         const auto linear_extents =
-            memory_layout.GetPhysicalMemoryRegionTree().GetDerivedRegionExtents(
+            memory_layout->GetPhysicalMemoryRegionTree().GetDerivedRegionExtents(
                 KMemoryRegionAttr_LinearMapped);
         ASSERT(linear_extents.GetEndAddress() != 0);
 
@@ -560,7 +546,7 @@ struct KernelCore::Impl {
             Common::AlignUp(linear_extents.GetEndAddress(), LinearRegionAlign) -
             aligned_linear_phys_start;
         const VAddr linear_region_start =
-            memory_layout.GetVirtualMemoryRegionTree().GetRandomAlignedRegionWithGuard(
+            memory_layout->GetVirtualMemoryRegionTree().GetRandomAlignedRegionWithGuard(
                 linear_region_size, LinearRegionAlign, KMemoryRegionType_None, LinearRegionAlign);
 
         const u64 linear_region_phys_to_virt_diff = linear_region_start - aligned_linear_phys_start;
@@ -569,7 +555,7 @@ struct KernelCore::Impl {
         {
             PAddr cur_phys_addr = 0;
             u64 cur_size = 0;
-            for (auto& region : memory_layout.GetPhysicalMemoryRegionTree()) {
+            for (auto& region : memory_layout->GetPhysicalMemoryRegionTree()) {
                 if (!region.HasTypeAttribute(KMemoryRegionAttr_LinearMapped)) {
                     continue;
                 }
@@ -588,55 +574,49 @@ struct KernelCore::Impl {
 
                 const VAddr region_virt_addr =
                     region.GetAddress() + linear_region_phys_to_virt_diff;
-                ASSERT(memory_layout.GetVirtualMemoryRegionTree().Insert(
+                ASSERT(memory_layout->GetVirtualMemoryRegionTree().Insert(
                     region_virt_addr, region.GetSize(),
                     GetTypeForVirtualLinearMapping(region.GetType())));
                 region.SetPairAddress(region_virt_addr);
 
                 KMemoryRegion* virt_region =
-                    memory_layout.GetVirtualMemoryRegionTree().FindModifiable(region_virt_addr);
+                    memory_layout->GetVirtualMemoryRegionTree().FindModifiable(region_virt_addr);
                 ASSERT(virt_region != nullptr);
                 virt_region->SetPairAddress(region.GetAddress());
             }
         }
 
         // Insert regions for the initial page table region.
-        ASSERT(memory_layout.GetPhysicalMemoryRegionTree().Insert(
+        ASSERT(memory_layout->GetPhysicalMemoryRegionTree().Insert(
             resource_end_phys_addr, KernelPageTableHeapSize, KMemoryRegionType_DramKernelInitPt));
-        ASSERT(memory_layout.GetVirtualMemoryRegionTree().Insert(
+        ASSERT(memory_layout->GetVirtualMemoryRegionTree().Insert(
             resource_end_phys_addr + linear_region_phys_to_virt_diff, KernelPageTableHeapSize,
             KMemoryRegionType_VirtualDramKernelInitPt));
 
         // All linear-mapped DRAM regions that we haven't tagged by this point will be allocated to
         // some pool partition. Tag them.
-        for (auto& region : memory_layout.GetPhysicalMemoryRegionTree()) {
+        for (auto& region : memory_layout->GetPhysicalMemoryRegionTree()) {
             if (region.GetType() == (KMemoryRegionType_Dram | KMemoryRegionAttr_LinearMapped)) {
                 region.SetType(KMemoryRegionType_DramPoolPartition);
             }
         }
 
         // Setup all other memory regions needed to arrange the pool partitions.
-        Init::SetupPoolPartitionMemoryRegions(memory_layout);
+        Init::SetupPoolPartitionMemoryRegions(*memory_layout);
 
         // Cache all linear regions in their own trees for faster access, later.
-        memory_layout.InitializeLinearMemoryRegionTrees(aligned_linear_phys_start,
-                                                        linear_region_start);
+        memory_layout->InitializeLinearMemoryRegionTrees(aligned_linear_phys_start,
+                                                         linear_region_start);
     }
 
-    void InitializeMemoryLayout(const KMemoryLayout& memory_layout) {
-        const auto system_pool = memory_layout.GetKernelSystemPoolRegionPhysicalExtents();
-        const auto applet_pool = memory_layout.GetKernelAppletPoolRegionPhysicalExtents();
-        const auto application_pool = memory_layout.GetKernelApplicationPoolRegionPhysicalExtents();
+    void InitializeMemoryLayout() {
+        const auto system_pool = memory_layout->GetKernelSystemPoolRegionPhysicalExtents();
 
-        // Initialize memory managers
+        // Initialize the memory manager.
         memory_manager = std::make_unique<KMemoryManager>(system);
-        memory_manager->InitializeManager(KMemoryManager::Pool::Application,
-                                          application_pool.GetAddress(),
-                                          application_pool.GetEndAddress());
-        memory_manager->InitializeManager(KMemoryManager::Pool::Applet, applet_pool.GetAddress(),
-                                          applet_pool.GetEndAddress());
-        memory_manager->InitializeManager(KMemoryManager::Pool::System, system_pool.GetAddress(),
-                                          system_pool.GetEndAddress());
+        const auto& management_region = memory_layout->GetPoolManagementRegion();
+        ASSERT(management_region.GetEndAddress() != 0);
+        memory_manager->Initialize(management_region.GetAddress(), management_region.GetSize());
 
         // Setup memory regions for emulated processes
         // TODO(bunnei): These should not be hardcoded regions initialized within the kernel
@@ -673,22 +653,6 @@ struct KernelCore::Impl {
                                     time_phys_addr, time_size, "Time:SharedMemory");
     }
 
-    void InitializePageSlab() {
-        // Allocate slab heaps
-        user_slab_heap_pages =
-            std::make_unique<KSlabHeap<Page>>(KSlabHeap<Page>::AllocationType::Guest);
-
-        // TODO(ameerj): This should be derived, not hardcoded within the kernel
-        constexpr u64 user_slab_heap_size{0x3de000};
-        // Reserve slab heaps
-        ASSERT(
-            system_resource_limit->Reserve(LimitableResource::PhysicalMemory, user_slab_heap_size));
-        // Initialize slab heap
-        user_slab_heap_pages->Initialize(
-            system.DeviceMemory().GetPointer(Core::DramMemoryMap::SlabHeapBase),
-            user_slab_heap_size);
-    }
-
     KClientPort* CreateNamedServicePort(std::string name) {
         auto search = service_interface_factory.find(name);
         if (search == service_interface_factory.end()) {
@@ -726,7 +690,6 @@ struct KernelCore::Impl {
     }
 
     std::mutex server_ports_lock;
-    std::mutex server_sessions_lock;
     std::mutex registered_objects_lock;
     std::mutex registered_in_use_objects_lock;
 
@@ -750,14 +713,13 @@ struct KernelCore::Impl {
     // stores all the objects in place.
     std::unique_ptr<KHandleTable> global_handle_table;
 
-    KAutoObjectWithListContainer object_list_container;
+    std::unique_ptr<KAutoObjectWithListContainer> global_object_list_container;
 
     /// Map of named ports managed by the kernel, which can be retrieved using
     /// the ConnectToPort SVC.
     std::unordered_map<std::string, ServiceInterfaceFactory> service_interface_factory;
     NamedPortTable named_ports;
     std::unordered_set<KServerPort*> server_ports;
-    std::unordered_set<KServerSession*> server_sessions;
     std::unordered_set<KAutoObject*> registered_objects;
     std::unordered_set<KAutoObject*> registered_in_use_objects;
 
@@ -769,7 +731,6 @@ struct KernelCore::Impl {
 
     // Kernel memory management
     std::unique_ptr<KMemoryManager> memory_manager;
-    std::unique_ptr<KSlabHeap<Page>> user_slab_heap_pages;
 
     // Shared memory for services
     Kernel::KSharedMemory* hid_shared_mem{};
@@ -777,6 +738,9 @@ struct KernelCore::Impl {
     Kernel::KSharedMemory* irs_shared_mem{};
     Kernel::KSharedMemory* time_shared_mem{};
 
+    // Memory layout
+    std::unique_ptr<KMemoryLayout> memory_layout;
+
     // Threads used for services
     std::unordered_set<std::shared_ptr<Kernel::ServiceThread>> service_threads;
     Common::ThreadWorker service_threads_manager;
@@ -925,11 +889,11 @@ const Core::ExclusiveMonitor& KernelCore::GetExclusiveMonitor() const {
 }
 
 KAutoObjectWithListContainer& KernelCore::ObjectListContainer() {
-    return impl->object_list_container;
+    return *impl->global_object_list_container;
 }
 
 const KAutoObjectWithListContainer& KernelCore::ObjectListContainer() const {
-    return impl->object_list_container;
+    return *impl->global_object_list_container;
 }
 
 void KernelCore::InvalidateAllInstructionCaches() {
@@ -959,16 +923,6 @@ KClientPort* KernelCore::CreateNamedServicePort(std::string name) {
     return impl->CreateNamedServicePort(std::move(name));
 }
 
-void KernelCore::RegisterServerSession(KServerSession* server_session) {
-    std::lock_guard lk(impl->server_sessions_lock);
-    impl->server_sessions.insert(server_session);
-}
-
-void KernelCore::UnregisterServerSession(KServerSession* server_session) {
-    std::lock_guard lk(impl->server_sessions_lock);
-    impl->server_sessions.erase(server_session);
-}
-
 void KernelCore::RegisterKernelObject(KAutoObject* object) {
     std::lock_guard lk(impl->registered_objects_lock);
     impl->registered_objects.insert(object);
@@ -1041,14 +995,6 @@ const KMemoryManager& KernelCore::MemoryManager() const {
     return *impl->memory_manager;
 }
 
-KSlabHeap<Page>& KernelCore::GetUserSlabHeapPages() {
-    return *impl->user_slab_heap_pages;
-}
-
-const KSlabHeap<Page>& KernelCore::GetUserSlabHeapPages() const {
-    return *impl->user_slab_heap_pages;
-}
-
 Kernel::KSharedMemory& KernelCore::GetHidSharedMem() {
     return *impl->hid_shared_mem;
 }
@@ -1142,6 +1088,10 @@ const KWorkerTaskManager& KernelCore::WorkerTaskManager() const {
     return impl->worker_task_manager;
 }
 
+const KMemoryLayout& KernelCore::MemoryLayout() const {
+    return *impl->memory_layout;
+}
+
 bool KernelCore::IsPhantomModeForSingleCore() const {
     return impl->IsPhantomModeForSingleCore();
 }
diff --git a/src/core/hle/kernel/kernel.h b/src/core/hle/kernel/kernel.h
index 0e04fc3bb..7087bbda6 100644
--- a/src/core/hle/kernel/kernel.h
+++ b/src/core/hle/kernel/kernel.h
@@ -41,7 +41,9 @@ class KClientSession;
 class KEvent;
 class KHandleTable;
 class KLinkedListNode;
+class KMemoryLayout;
 class KMemoryManager;
+class KPageBuffer;
 class KPort;
 class KProcess;
 class KResourceLimit;
@@ -51,6 +53,7 @@ class KSession;
 class KSharedMemory;
 class KSharedMemoryInfo;
 class KThread;
+class KThreadLocalPage;
 class KTransferMemory;
 class KWorkerTaskManager;
 class KWritableEvent;
@@ -193,14 +196,6 @@ public:
     /// Opens a port to a service previously registered with RegisterNamedService.
     KClientPort* CreateNamedServicePort(std::string name);
 
-    /// Registers a server session with the gobal emulation state, to be freed on shutdown. This is
-    /// necessary because we do not emulate processes for HLE sessions.
-    void RegisterServerSession(KServerSession* server_session);
-
-    /// Unregisters a server session previously registered with RegisterServerSession when it was
-    /// destroyed during the current emulation session.
-    void UnregisterServerSession(KServerSession* server_session);
-
     /// Registers all kernel objects with the global emulation state, this is purely for tracking
     /// leaks after emulation has been shutdown.
     void RegisterKernelObject(KAutoObject* object);
@@ -238,12 +233,6 @@ public:
     /// Gets the virtual memory manager for the kernel.
     const KMemoryManager& MemoryManager() const;
 
-    /// Gets the slab heap allocated for user space pages.
-    KSlabHeap<Page>& GetUserSlabHeapPages();
-
-    /// Gets the slab heap allocated for user space pages.
-    const KSlabHeap<Page>& GetUserSlabHeapPages() const;
-
     /// Gets the shared memory object for HID services.
     Kernel::KSharedMemory& GetHidSharedMem();
 
@@ -335,6 +324,10 @@ public:
             return slab_heap_container->writeable_event;
         } else if constexpr (std::is_same_v<T, KCodeMemory>) {
             return slab_heap_container->code_memory;
+        } else if constexpr (std::is_same_v<T, KPageBuffer>) {
+            return slab_heap_container->page_buffer;
+        } else if constexpr (std::is_same_v<T, KThreadLocalPage>) {
+            return slab_heap_container->thread_local_page;
         }
     }
 
@@ -350,6 +343,9 @@ public:
     /// Gets the current worker task manager, used for dispatching KThread/KProcess tasks.
     const KWorkerTaskManager& WorkerTaskManager() const;
 
+    /// Gets the memory layout.
+    const KMemoryLayout& MemoryLayout() const;
+
 private:
     friend class KProcess;
     friend class KThread;
@@ -393,6 +389,8 @@ private:
         KSlabHeap<KTransferMemory> transfer_memory;
         KSlabHeap<KWritableEvent> writeable_event;
         KSlabHeap<KCodeMemory> code_memory;
+        KSlabHeap<KPageBuffer> page_buffer;
+        KSlabHeap<KThreadLocalPage> thread_local_page;
     };
 
     std::unique_ptr<SlabHeapContainer> slab_heap_container;
diff --git a/src/core/hle/kernel/service_thread.cpp b/src/core/hle/kernel/service_thread.cpp
index 4eb3a5988..52d25b837 100644
--- a/src/core/hle/kernel/service_thread.cpp
+++ b/src/core/hle/kernel/service_thread.cpp
@@ -49,12 +49,9 @@ ServiceThread::Impl::Impl(KernelCore& kernel, std::size_t num_threads, const std
                 return;
             }
 
+            // Allocate a dummy guest thread for this host thread.
             kernel.RegisterHostThread();
 
-            // Ensure the dummy thread allocated for this host thread is closed on exit.
-            auto* dummy_thread = kernel.GetCurrentEmuThread();
-            SCOPE_EXIT({ dummy_thread->Close(); });
-
             while (true) {
                 std::function<void()> task;
 
diff --git a/src/core/hle/kernel/slab_helpers.h b/src/core/hle/kernel/slab_helpers.h
index f1c11256e..dc1e48fc9 100644
--- a/src/core/hle/kernel/slab_helpers.h
+++ b/src/core/hle/kernel/slab_helpers.h
@@ -59,7 +59,7 @@ class KAutoObjectWithSlabHeapAndContainer : public Base {
 
 private:
     static Derived* Allocate(KernelCore& kernel) {
-        return kernel.SlabHeap<Derived>().AllocateWithKernel(kernel);
+        return kernel.SlabHeap<Derived>().Allocate(kernel);
     }
 
     static void Free(KernelCore& kernel, Derived* obj) {
diff --git a/src/core/hle/kernel/svc.cpp b/src/core/hle/kernel/svc.cpp
index 4f7aebf3f..839171e85 100644
--- a/src/core/hle/kernel/svc.cpp
+++ b/src/core/hle/kernel/svc.cpp
@@ -396,7 +396,7 @@ static ResultCode GetProcessId(Core::System& system, u64* out_process_id, Handle
     // Get the process id.
     *out_process_id = process->GetId();
 
-    return ResultInvalidHandle;
+    return ResultSuccess;
 }
 
 static ResultCode GetProcessId32(Core::System& system, u32* out_process_id_low,
@@ -645,6 +645,10 @@ static void OutputDebugString(Core::System& system, VAddr address, u64 len) {
     LOG_DEBUG(Debug_Emulated, "{}", str);
 }
 
+static void OutputDebugString32(Core::System& system, u32 address, u32 len) {
+    OutputDebugString(system, address, len);
+}
+
 /// Gets system/memory information for the current process
 static ResultCode GetInfo(Core::System& system, u64* result, u64 info_id, Handle handle,
                           u64 info_sub_id) {
@@ -1404,7 +1408,7 @@ static ResultCode UnmapProcessMemory(Core::System& system, VAddr dst_address, Ha
 }
 
 static ResultCode CreateCodeMemory(Core::System& system, Handle* out, VAddr address, size_t size) {
-    LOG_TRACE(Kernel_SVC, "called, handle_out=0x{:X}, address=0x{:X}, size=0x{:X}",
+    LOG_TRACE(Kernel_SVC, "called, handle_out={}, address=0x{:X}, size=0x{:X}",
               static_cast<void*>(out), address, size);
     // Get kernel instance.
     auto& kernel = system.Kernel();
@@ -1438,6 +1442,10 @@ static ResultCode CreateCodeMemory(Core::System& system, Handle* out, VAddr addr
     return ResultSuccess;
 }
 
+static ResultCode CreateCodeMemory32(Core::System& system, Handle* out, u32 address, u32 size) {
+    return CreateCodeMemory(system, out, address, size);
+}
+
 static ResultCode ControlCodeMemory(Core::System& system, Handle code_memory_handle, u32 operation,
                                     VAddr address, size_t size, Svc::MemoryPermission perm) {
 
@@ -1517,6 +1525,12 @@ static ResultCode ControlCodeMemory(Core::System& system, Handle code_memory_han
     return ResultSuccess;
 }
 
+static ResultCode ControlCodeMemory32(Core::System& system, Handle code_memory_handle,
+                                      u32 operation, u64 address, u64 size,
+                                      Svc::MemoryPermission perm) {
+    return ControlCodeMemory(system, code_memory_handle, operation, address, size, perm);
+}
+
 static ResultCode QueryProcessMemory(Core::System& system, VAddr memory_info_address,
                                      VAddr page_info_address, Handle process_handle,
                                      VAddr address) {
@@ -2318,7 +2332,7 @@ static ResultCode CreateEvent(Core::System& system, Handle* out_write, Handle* o
     R_UNLESS(event != nullptr, ResultOutOfResource);
 
     // Initialize the event.
-    event->Initialize("CreateEvent");
+    event->Initialize("CreateEvent", kernel.CurrentProcess());
 
     // Commit the thread reservation.
     event_reservation.Commit();
@@ -2559,9 +2573,9 @@ struct FunctionDef {
 } // namespace
 
 static const FunctionDef SVC_Table_32[] = {
-    {0x00, nullptr, "Unknown"},
+    {0x00, nullptr, "Unknown0"},
     {0x01, SvcWrap32<SetHeapSize32>, "SetHeapSize32"},
-    {0x02, nullptr, "Unknown"},
+    {0x02, nullptr, "SetMemoryPermission32"},
     {0x03, SvcWrap32<SetMemoryAttribute32>, "SetMemoryAttribute32"},
     {0x04, SvcWrap32<MapMemory32>, "MapMemory32"},
     {0x05, SvcWrap32<UnmapMemory32>, "UnmapMemory32"},
@@ -2591,97 +2605,97 @@ static const FunctionDef SVC_Table_32[] = {
     {0x1d, SvcWrap32<SignalProcessWideKey32>, "SignalProcessWideKey32"},
     {0x1e, SvcWrap32<GetSystemTick32>, "GetSystemTick32"},
     {0x1f, SvcWrap32<ConnectToNamedPort32>, "ConnectToNamedPort32"},
-    {0x20, nullptr, "Unknown"},
+    {0x20, nullptr, "SendSyncRequestLight32"},
     {0x21, SvcWrap32<SendSyncRequest32>, "SendSyncRequest32"},
     {0x22, nullptr, "SendSyncRequestWithUserBuffer32"},
-    {0x23, nullptr, "Unknown"},
+    {0x23, nullptr, "SendAsyncRequestWithUserBuffer32"},
     {0x24, SvcWrap32<GetProcessId32>, "GetProcessId32"},
     {0x25, SvcWrap32<GetThreadId32>, "GetThreadId32"},
     {0x26, SvcWrap32<Break32>, "Break32"},
-    {0x27, nullptr, "OutputDebugString32"},
-    {0x28, nullptr, "Unknown"},
+    {0x27, SvcWrap32<OutputDebugString32>, "OutputDebugString32"},
+    {0x28, nullptr, "ReturnFromException32"},
     {0x29, SvcWrap32<GetInfo32>, "GetInfo32"},
-    {0x2a, nullptr, "Unknown"},
-    {0x2b, nullptr, "Unknown"},
+    {0x2a, nullptr, "FlushEntireDataCache32"},
+    {0x2b, nullptr, "FlushDataCache32"},
     {0x2c, SvcWrap32<MapPhysicalMemory32>, "MapPhysicalMemory32"},
     {0x2d, SvcWrap32<UnmapPhysicalMemory32>, "UnmapPhysicalMemory32"},
-    {0x2e, nullptr, "Unknown"},
-    {0x2f, nullptr, "Unknown"},
-    {0x30, nullptr, "Unknown"},
-    {0x31, nullptr, "Unknown"},
+    {0x2e, nullptr, "GetDebugFutureThreadInfo32"},
+    {0x2f, nullptr, "GetLastThreadInfo32"},
+    {0x30, nullptr, "GetResourceLimitLimitValue32"},
+    {0x31, nullptr, "GetResourceLimitCurrentValue32"},
     {0x32, SvcWrap32<SetThreadActivity32>, "SetThreadActivity32"},
     {0x33, SvcWrap32<GetThreadContext32>, "GetThreadContext32"},
     {0x34, SvcWrap32<WaitForAddress32>, "WaitForAddress32"},
     {0x35, SvcWrap32<SignalToAddress32>, "SignalToAddress32"},
     {0x36, SvcWrap32<SynchronizePreemptionState>, "SynchronizePreemptionState32"},
-    {0x37, nullptr, "Unknown"},
-    {0x38, nullptr, "Unknown"},
-    {0x39, nullptr, "Unknown"},
-    {0x3a, nullptr, "Unknown"},
-    {0x3b, nullptr, "Unknown"},
-    {0x3c, nullptr, "Unknown"},
-    {0x3d, nullptr, "Unknown"},
-    {0x3e, nullptr, "Unknown"},
-    {0x3f, nullptr, "Unknown"},
+    {0x37, nullptr, "GetResourceLimitPeakValue32"},
+    {0x38, nullptr, "Unknown38"},
+    {0x39, nullptr, "CreateIoPool32"},
+    {0x3a, nullptr, "CreateIoRegion32"},
+    {0x3b, nullptr, "Unknown3b"},
+    {0x3c, nullptr, "KernelDebug32"},
+    {0x3d, nullptr, "ChangeKernelTraceState32"},
+    {0x3e, nullptr, "Unknown3e"},
+    {0x3f, nullptr, "Unknown3f"},
     {0x40, nullptr, "CreateSession32"},
     {0x41, nullptr, "AcceptSession32"},
-    {0x42, nullptr, "Unknown"},
+    {0x42, nullptr, "ReplyAndReceiveLight32"},
     {0x43, nullptr, "ReplyAndReceive32"},
-    {0x44, nullptr, "Unknown"},
+    {0x44, nullptr, "ReplyAndReceiveWithUserBuffer32"},
     {0x45, SvcWrap32<CreateEvent32>, "CreateEvent32"},
-    {0x46, nullptr, "Unknown"},
-    {0x47, nullptr, "Unknown"},
-    {0x48, nullptr, "Unknown"},
-    {0x49, nullptr, "Unknown"},
-    {0x4a, nullptr, "Unknown"},
-    {0x4b, nullptr, "Unknown"},
-    {0x4c, nullptr, "Unknown"},
-    {0x4d, nullptr, "Unknown"},
-    {0x4e, nullptr, "Unknown"},
-    {0x4f, nullptr, "Unknown"},
-    {0x50, nullptr, "Unknown"},
-    {0x51, nullptr, "Unknown"},
-    {0x52, nullptr, "Unknown"},
-    {0x53, nullptr, "Unknown"},
-    {0x54, nullptr, "Unknown"},
-    {0x55, nullptr, "Unknown"},
-    {0x56, nullptr, "Unknown"},
-    {0x57, nullptr, "Unknown"},
-    {0x58, nullptr, "Unknown"},
-    {0x59, nullptr, "Unknown"},
-    {0x5a, nullptr, "Unknown"},
-    {0x5b, nullptr, "Unknown"},
-    {0x5c, nullptr, "Unknown"},
-    {0x5d, nullptr, "Unknown"},
-    {0x5e, nullptr, "Unknown"},
+    {0x46, nullptr, "MapIoRegion32"},
+    {0x47, nullptr, "UnmapIoRegion32"},
+    {0x48, nullptr, "MapPhysicalMemoryUnsafe32"},
+    {0x49, nullptr, "UnmapPhysicalMemoryUnsafe32"},
+    {0x4a, nullptr, "SetUnsafeLimit32"},
+    {0x4b, SvcWrap32<CreateCodeMemory32>, "CreateCodeMemory32"},
+    {0x4c, SvcWrap32<ControlCodeMemory32>, "ControlCodeMemory32"},
+    {0x4d, nullptr, "SleepSystem32"},
+    {0x4e, nullptr, "ReadWriteRegister32"},
+    {0x4f, nullptr, "SetProcessActivity32"},
+    {0x50, nullptr, "CreateSharedMemory32"},
+    {0x51, nullptr, "MapTransferMemory32"},
+    {0x52, nullptr, "UnmapTransferMemory32"},
+    {0x53, nullptr, "CreateInterruptEvent32"},
+    {0x54, nullptr, "QueryPhysicalAddress32"},
+    {0x55, nullptr, "QueryIoMapping32"},
+    {0x56, nullptr, "CreateDeviceAddressSpace32"},
+    {0x57, nullptr, "AttachDeviceAddressSpace32"},
+    {0x58, nullptr, "DetachDeviceAddressSpace32"},
+    {0x59, nullptr, "MapDeviceAddressSpaceByForce32"},
+    {0x5a, nullptr, "MapDeviceAddressSpaceAligned32"},
+    {0x5b, nullptr, "MapDeviceAddressSpace32"},
+    {0x5c, nullptr, "UnmapDeviceAddressSpace32"},
+    {0x5d, nullptr, "InvalidateProcessDataCache32"},
+    {0x5e, nullptr, "StoreProcessDataCache32"},
     {0x5F, SvcWrap32<FlushProcessDataCache32>, "FlushProcessDataCache32"},
-    {0x60, nullptr, "Unknown"},
-    {0x61, nullptr, "Unknown"},
-    {0x62, nullptr, "Unknown"},
-    {0x63, nullptr, "Unknown"},
-    {0x64, nullptr, "Unknown"},
+    {0x60, nullptr, "StoreProcessDataCache32"},
+    {0x61, nullptr, "BreakDebugProcess32"},
+    {0x62, nullptr, "TerminateDebugProcess32"},
+    {0x63, nullptr, "GetDebugEvent32"},
+    {0x64, nullptr, "ContinueDebugEvent32"},
     {0x65, nullptr, "GetProcessList32"},
-    {0x66, nullptr, "Unknown"},
-    {0x67, nullptr, "Unknown"},
-    {0x68, nullptr, "Unknown"},
-    {0x69, nullptr, "Unknown"},
-    {0x6A, nullptr, "Unknown"},
-    {0x6B, nullptr, "Unknown"},
-    {0x6C, nullptr, "Unknown"},
-    {0x6D, nullptr, "Unknown"},
-    {0x6E, nullptr, "Unknown"},
+    {0x66, nullptr, "GetThreadList"},
+    {0x67, nullptr, "GetDebugThreadContext32"},
+    {0x68, nullptr, "SetDebugThreadContext32"},
+    {0x69, nullptr, "QueryDebugProcessMemory32"},
+    {0x6A, nullptr, "ReadDebugProcessMemory32"},
+    {0x6B, nullptr, "WriteDebugProcessMemory32"},
+    {0x6C, nullptr, "SetHardwareBreakPoint32"},
+    {0x6D, nullptr, "GetDebugThreadParam32"},
+    {0x6E, nullptr, "Unknown6E"},
     {0x6f, nullptr, "GetSystemInfo32"},
     {0x70, nullptr, "CreatePort32"},
     {0x71, nullptr, "ManageNamedPort32"},
     {0x72, nullptr, "ConnectToPort32"},
     {0x73, nullptr, "SetProcessMemoryPermission32"},
-    {0x74, nullptr, "Unknown"},
-    {0x75, nullptr, "Unknown"},
-    {0x76, nullptr, "Unknown"},
+    {0x74, nullptr, "MapProcessMemory32"},
+    {0x75, nullptr, "UnmapProcessMemory32"},
+    {0x76, nullptr, "QueryProcessMemory32"},
     {0x77, nullptr, "MapProcessCodeMemory32"},
     {0x78, nullptr, "UnmapProcessCodeMemory32"},
-    {0x79, nullptr, "Unknown"},
-    {0x7A, nullptr, "Unknown"},
+    {0x79, nullptr, "CreateProcess32"},
+    {0x7A, nullptr, "StartProcess32"},
     {0x7B, nullptr, "TerminateProcess32"},
     {0x7C, nullptr, "GetProcessInfo32"},
     {0x7D, nullptr, "CreateResourceLimit32"},
@@ -2754,7 +2768,7 @@ static const FunctionDef SVC_Table_32[] = {
 };
 
 static const FunctionDef SVC_Table_64[] = {
-    {0x00, nullptr, "Unknown"},
+    {0x00, nullptr, "Unknown0"},
     {0x01, SvcWrap64<SetHeapSize>, "SetHeapSize"},
     {0x02, SvcWrap64<SetMemoryPermission>, "SetMemoryPermission"},
     {0x03, SvcWrap64<SetMemoryAttribute>, "SetMemoryAttribute"},
@@ -2809,23 +2823,23 @@ static const FunctionDef SVC_Table_64[] = {
     {0x34, SvcWrap64<WaitForAddress>, "WaitForAddress"},
     {0x35, SvcWrap64<SignalToAddress>, "SignalToAddress"},
     {0x36, SvcWrap64<SynchronizePreemptionState>, "SynchronizePreemptionState"},
-    {0x37, nullptr, "Unknown"},
-    {0x38, nullptr, "Unknown"},
-    {0x39, nullptr, "Unknown"},
-    {0x3A, nullptr, "Unknown"},
-    {0x3B, nullptr, "Unknown"},
+    {0x37, nullptr, "GetResourceLimitPeakValue"},
+    {0x38, nullptr, "Unknown38"},
+    {0x39, nullptr, "CreateIoPool"},
+    {0x3A, nullptr, "CreateIoRegion"},
+    {0x3B, nullptr, "Unknown3B"},
     {0x3C, SvcWrap64<KernelDebug>, "KernelDebug"},
     {0x3D, SvcWrap64<ChangeKernelTraceState>, "ChangeKernelTraceState"},
-    {0x3E, nullptr, "Unknown"},
-    {0x3F, nullptr, "Unknown"},
+    {0x3E, nullptr, "Unknown3e"},
+    {0x3F, nullptr, "Unknown3f"},
     {0x40, nullptr, "CreateSession"},
     {0x41, nullptr, "AcceptSession"},
     {0x42, nullptr, "ReplyAndReceiveLight"},
     {0x43, nullptr, "ReplyAndReceive"},
     {0x44, nullptr, "ReplyAndReceiveWithUserBuffer"},
     {0x45, SvcWrap64<CreateEvent>, "CreateEvent"},
-    {0x46, nullptr, "Unknown"},
-    {0x47, nullptr, "Unknown"},
+    {0x46, nullptr, "MapIoRegion"},
+    {0x47, nullptr, "UnmapIoRegion"},
     {0x48, nullptr, "MapPhysicalMemoryUnsafe"},
     {0x49, nullptr, "UnmapPhysicalMemoryUnsafe"},
     {0x4A, nullptr, "SetUnsafeLimit"},
@@ -2864,7 +2878,7 @@ static const FunctionDef SVC_Table_64[] = {
     {0x6B, nullptr, "WriteDebugProcessMemory"},
     {0x6C, nullptr, "SetHardwareBreakPoint"},
     {0x6D, nullptr, "GetDebugThreadParam"},
-    {0x6E, nullptr, "Unknown"},
+    {0x6E, nullptr, "Unknown6E"},
     {0x6F, nullptr, "GetSystemInfo"},
     {0x70, nullptr, "CreatePort"},
     {0x71, nullptr, "ManageNamedPort"},
diff --git a/src/core/hle/kernel/svc_types.h b/src/core/hle/kernel/svc_types.h
index 365e22e4e..b2e9ec092 100644
--- a/src/core/hle/kernel/svc_types.h
+++ b/src/core/hle/kernel/svc_types.h
@@ -96,4 +96,6 @@ constexpr inline s32 IdealCoreNoUpdate = -3;
 constexpr inline s32 LowestThreadPriority = 63;
 constexpr inline s32 HighestThreadPriority = 0;
 
+constexpr inline size_t ThreadLocalRegionSize = 0x200;
+
 } // namespace Kernel::Svc
diff --git a/src/core/hle/kernel/svc_wrap.h b/src/core/hle/kernel/svc_wrap.h
index a60adfcab..d309f166c 100644
--- a/src/core/hle/kernel/svc_wrap.h
+++ b/src/core/hle/kernel/svc_wrap.h
@@ -669,4 +669,26 @@ void SvcWrap32(Core::System& system) {
     FuncReturn(system, retval);
 }
 
+// Used by CreateCodeMemory32
+template <ResultCode func(Core::System&, Handle*, u32, u32)>
+void SvcWrap32(Core::System& system) {
+    Handle handle = 0;
+
+    const u32 retval = func(system, &handle, Param32(system, 1), Param32(system, 2)).raw;
+
+    system.CurrentArmInterface().SetReg(1, handle);
+    FuncReturn(system, retval);
+}
+
+// Used by ControlCodeMemory32
+template <ResultCode func(Core::System&, Handle, u32, u64, u64, Svc::MemoryPermission)>
+void SvcWrap32(Core::System& system) {
+    const u32 retval =
+        func(system, Param32(system, 0), Param32(system, 1), Param(system, 2), Param(system, 4),
+             static_cast<Svc::MemoryPermission>(Param32(system, 6)))
+            .raw;
+
+    FuncReturn(system, retval);
+}
+
 } // namespace Kernel
diff --git a/src/core/hle/service/acc/acc.cpp b/src/core/hle/service/acc/acc.cpp
index 6e63e057e..e34ef5a78 100644
--- a/src/core/hle/service/acc/acc.cpp
+++ b/src/core/hle/service/acc/acc.cpp
@@ -39,9 +39,9 @@ constexpr ResultCode ERR_FAILED_SAVE_DATA{ErrorModule::Account, 100};
 // Thumbnails are hard coded to be at least this size
 constexpr std::size_t THUMBNAIL_SIZE = 0x24000;
 
-static std::filesystem::path GetImagePath(Common::UUID uuid) {
+static std::filesystem::path GetImagePath(const Common::UUID& uuid) {
     return Common::FS::GetYuzuPath(Common::FS::YuzuPath::NANDDir) /
-           fmt::format("system/save/8000000000000010/su/avators/{}.jpg", uuid.FormatSwitch());
+           fmt::format("system/save/8000000000000010/su/avators/{}.jpg", uuid.FormattedString());
 }
 
 static constexpr u32 SanitizeJPEGSize(std::size_t size) {
@@ -290,7 +290,7 @@ public:
 
 protected:
     void Get(Kernel::HLERequestContext& ctx) {
-        LOG_DEBUG(Service_ACC, "called user_id=0x{}", user_id.Format());
+        LOG_DEBUG(Service_ACC, "called user_id=0x{}", user_id.RawString());
         ProfileBase profile_base{};
         ProfileData data{};
         if (profile_manager.GetProfileBaseAndData(user_id, profile_base, data)) {
@@ -300,21 +300,21 @@ protected:
             rb.PushRaw(profile_base);
         } else {
             LOG_ERROR(Service_ACC, "Failed to get profile base and data for user=0x{}",
-                      user_id.Format());
+                      user_id.RawString());
             IPC::ResponseBuilder rb{ctx, 2};
             rb.Push(ResultUnknown); // TODO(ogniK): Get actual error code
         }
     }
 
     void GetBase(Kernel::HLERequestContext& ctx) {
-        LOG_DEBUG(Service_ACC, "called user_id=0x{}", user_id.Format());
+        LOG_DEBUG(Service_ACC, "called user_id=0x{}", user_id.RawString());
         ProfileBase profile_base{};
         if (profile_manager.GetProfileBase(user_id, profile_base)) {
             IPC::ResponseBuilder rb{ctx, 16};
             rb.Push(ResultSuccess);
             rb.PushRaw(profile_base);
         } else {
-            LOG_ERROR(Service_ACC, "Failed to get profile base for user=0x{}", user_id.Format());
+            LOG_ERROR(Service_ACC, "Failed to get profile base for user=0x{}", user_id.RawString());
             IPC::ResponseBuilder rb{ctx, 2};
             rb.Push(ResultUnknown); // TODO(ogniK): Get actual error code
         }
@@ -373,7 +373,7 @@ protected:
         LOG_DEBUG(Service_ACC, "called, username='{}', timestamp={:016X}, uuid=0x{}",
                   Common::StringFromFixedZeroTerminatedBuffer(
                       reinterpret_cast<const char*>(base.username.data()), base.username.size()),
-                  base.timestamp, base.user_uuid.Format());
+                  base.timestamp, base.user_uuid.RawString());
 
         if (user_data.size() < sizeof(ProfileData)) {
             LOG_ERROR(Service_ACC, "ProfileData buffer too small!");
@@ -406,7 +406,7 @@ protected:
         LOG_DEBUG(Service_ACC, "called, username='{}', timestamp={:016X}, uuid=0x{}",
                   Common::StringFromFixedZeroTerminatedBuffer(
                       reinterpret_cast<const char*>(base.username.data()), base.username.size()),
-                  base.timestamp, base.user_uuid.Format());
+                  base.timestamp, base.user_uuid.RawString());
 
         if (user_data.size() < sizeof(ProfileData)) {
             LOG_ERROR(Service_ACC, "ProfileData buffer too small!");
@@ -435,7 +435,7 @@ protected:
     }
 
     ProfileManager& profile_manager;
-    Common::UUID user_id{Common::INVALID_UUID}; ///< The user id this profile refers to.
+    Common::UUID user_id{}; ///< The user id this profile refers to.
 };
 
 class IProfile final : public IProfileCommon {
@@ -547,7 +547,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 4};
         rb.Push(ResultSuccess);
-        rb.PushRaw<u64>(user_id.GetNintendoID());
+        rb.PushRaw<u64>(user_id.Hash());
     }
 
     void EnsureIdTokenCacheAsync(Kernel::HLERequestContext& ctx) {
@@ -577,7 +577,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 4};
         rb.Push(ResultSuccess);
-        rb.PushRaw<u64>(user_id.GetNintendoID());
+        rb.PushRaw<u64>(user_id.Hash());
     }
 
     void StoreOpenContext(Kernel::HLERequestContext& ctx) {
@@ -587,7 +587,7 @@ private:
     }
 
     std::shared_ptr<EnsureTokenIdCacheAsyncInterface> ensure_token_id{};
-    Common::UUID user_id{Common::INVALID_UUID};
+    Common::UUID user_id{};
 };
 
 // 6.0.0+
@@ -687,7 +687,7 @@ void Module::Interface::GetUserCount(Kernel::HLERequestContext& ctx) {
 void Module::Interface::GetUserExistence(Kernel::HLERequestContext& ctx) {
     IPC::RequestParser rp{ctx};
     Common::UUID user_id = rp.PopRaw<Common::UUID>();
-    LOG_DEBUG(Service_ACC, "called user_id=0x{}", user_id.Format());
+    LOG_DEBUG(Service_ACC, "called user_id=0x{}", user_id.RawString());
 
     IPC::ResponseBuilder rb{ctx, 3};
     rb.Push(ResultSuccess);
@@ -718,7 +718,7 @@ void Module::Interface::GetLastOpenedUser(Kernel::HLERequestContext& ctx) {
 void Module::Interface::GetProfile(Kernel::HLERequestContext& ctx) {
     IPC::RequestParser rp{ctx};
     Common::UUID user_id = rp.PopRaw<Common::UUID>();
-    LOG_DEBUG(Service_ACC, "called user_id=0x{}", user_id.Format());
+    LOG_DEBUG(Service_ACC, "called user_id=0x{}", user_id.RawString());
 
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(ResultSuccess);
@@ -833,7 +833,7 @@ void Module::Interface::GetProfileEditor(Kernel::HLERequestContext& ctx) {
     IPC::RequestParser rp{ctx};
     Common::UUID user_id = rp.PopRaw<Common::UUID>();
 
-    LOG_DEBUG(Service_ACC, "called, user_id=0x{}", user_id.Format());
+    LOG_DEBUG(Service_ACC, "called, user_id=0x{}", user_id.RawString());
 
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(ResultSuccess);
@@ -875,7 +875,7 @@ void Module::Interface::StoreSaveDataThumbnailApplication(Kernel::HLERequestCont
     IPC::RequestParser rp{ctx};
     const auto uuid = rp.PopRaw<Common::UUID>();
 
-    LOG_WARNING(Service_ACC, "(STUBBED) called, uuid=0x{}", uuid.Format());
+    LOG_WARNING(Service_ACC, "(STUBBED) called, uuid=0x{}", uuid.RawString());
 
     // TODO(ogniK): Check if application ID is zero on acc initialize. As we don't have a reliable
     // way of confirming things like the TID, we're going to assume a non zero value for the time
@@ -889,7 +889,7 @@ void Module::Interface::StoreSaveDataThumbnailSystem(Kernel::HLERequestContext&
     const auto uuid = rp.PopRaw<Common::UUID>();
     const auto tid = rp.Pop<u64_le>();
 
-    LOG_WARNING(Service_ACC, "(STUBBED) called, uuid=0x{}, tid={:016X}", uuid.Format(), tid);
+    LOG_WARNING(Service_ACC, "(STUBBED) called, uuid=0x{}, tid={:016X}", uuid.RawString(), tid);
     StoreSaveDataThumbnail(ctx, uuid, tid);
 }
 
@@ -903,7 +903,7 @@ void Module::Interface::StoreSaveDataThumbnail(Kernel::HLERequestContext& ctx,
         return;
     }
 
-    if (!uuid) {
+    if (uuid.IsInvalid()) {
         LOG_ERROR(Service_ACC, "User ID is not valid!");
         rb.Push(ERR_INVALID_USER_ID);
         return;
@@ -927,20 +927,20 @@ void Module::Interface::TrySelectUserWithoutInteraction(Kernel::HLERequestContex
     IPC::ResponseBuilder rb{ctx, 6};
     if (profile_manager->GetUserCount() != 1) {
         rb.Push(ResultSuccess);
-        rb.PushRaw<u128>(Common::INVALID_UUID);
+        rb.PushRaw(Common::InvalidUUID);
         return;
     }
 
     const auto user_list = profile_manager->GetAllUsers();
     if (std::ranges::all_of(user_list, [](const auto& user) { return user.IsInvalid(); })) {
         rb.Push(ResultUnknown); // TODO(ogniK): Find the correct error code
-        rb.PushRaw<u128>(Common::INVALID_UUID);
+        rb.PushRaw(Common::InvalidUUID);
         return;
     }
 
     // Select the first user we have
     rb.Push(ResultSuccess);
-    rb.PushRaw<u128>(profile_manager->GetUser(0)->uuid);
+    rb.PushRaw(profile_manager->GetUser(0)->uuid);
 }
 
 Module::Interface::Interface(std::shared_ptr<Module> module_,
diff --git a/src/core/hle/service/acc/profile_manager.cpp b/src/core/hle/service/acc/profile_manager.cpp
index 568303ced..fba847142 100644
--- a/src/core/hle/service/acc/profile_manager.cpp
+++ b/src/core/hle/service/acc/profile_manager.cpp
@@ -19,8 +19,8 @@ namespace FS = Common::FS;
 using Common::UUID;
 
 struct UserRaw {
-    UUID uuid{Common::INVALID_UUID};
-    UUID uuid2{Common::INVALID_UUID};
+    UUID uuid{};
+    UUID uuid2{};
     u64 timestamp{};
     ProfileUsername username{};
     ProfileData extra_data{};
@@ -45,7 +45,7 @@ ProfileManager::ProfileManager() {
 
     // Create an user if none are present
     if (user_count == 0) {
-        CreateNewUser(UUID::Generate(), "yuzu");
+        CreateNewUser(UUID::MakeRandom(), "yuzu");
     }
 
     auto current =
@@ -101,7 +101,7 @@ ResultCode ProfileManager::CreateNewUser(UUID uuid, const ProfileUsername& usern
     if (user_count == MAX_USERS) {
         return ERROR_TOO_MANY_USERS;
     }
-    if (!uuid) {
+    if (uuid.IsInvalid()) {
         return ERROR_ARGUMENT_IS_NULL;
     }
     if (username[0] == 0x0) {
@@ -145,7 +145,7 @@ std::optional<UUID> ProfileManager::GetUser(std::size_t index) const {
 
 /// Returns a users profile index based on their user id.
 std::optional<std::size_t> ProfileManager::GetUserIndex(const UUID& uuid) const {
-    if (!uuid) {
+    if (uuid.IsInvalid()) {
         return std::nullopt;
     }
 
@@ -250,9 +250,10 @@ UserIDArray ProfileManager::GetOpenUsers() const {
     std::ranges::transform(profiles, output.begin(), [](const ProfileInfo& p) {
         if (p.is_open)
             return p.user_uuid;
-        return UUID{Common::INVALID_UUID};
+        return Common::InvalidUUID;
     });
-    std::stable_partition(output.begin(), output.end(), [](const UUID& uuid) { return uuid; });
+    std::stable_partition(output.begin(), output.end(),
+                          [](const UUID& uuid) { return uuid.IsValid(); });
     return output;
 }
 
@@ -299,7 +300,7 @@ bool ProfileManager::RemoveUser(UUID uuid) {
 
     profiles[*index] = ProfileInfo{};
     std::stable_partition(profiles.begin(), profiles.end(),
-                          [](const ProfileInfo& profile) { return profile.user_uuid; });
+                          [](const ProfileInfo& profile) { return profile.user_uuid.IsValid(); });
     return true;
 }
 
@@ -361,7 +362,7 @@ void ProfileManager::ParseUserSaveFile() {
     }
 
     std::stable_partition(profiles.begin(), profiles.end(),
-                          [](const ProfileInfo& profile) { return profile.user_uuid; });
+                          [](const ProfileInfo& profile) { return profile.user_uuid.IsValid(); });
 }
 
 void ProfileManager::WriteUserSaveFile() {
diff --git a/src/core/hle/service/acc/profile_manager.h b/src/core/hle/service/acc/profile_manager.h
index 71b9d5518..17347f7ef 100644
--- a/src/core/hle/service/acc/profile_manager.h
+++ b/src/core/hle/service/acc/profile_manager.h
@@ -35,7 +35,7 @@ static_assert(sizeof(ProfileData) == 0x80, "ProfileData structure has incorrect
 /// This holds general information about a users profile. This is where we store all the information
 /// based on a specific user
 struct ProfileInfo {
-    Common::UUID user_uuid{Common::INVALID_UUID};
+    Common::UUID user_uuid{};
     ProfileUsername username{};
     u64 creation_time{};
     ProfileData data{}; // TODO(ognik): Work out what this is
@@ -49,7 +49,7 @@ struct ProfileBase {
 
     // Zero out all the fields to make the profile slot considered "Empty"
     void Invalidate() {
-        user_uuid.Invalidate();
+        user_uuid = {};
         timestamp = 0;
         username.fill(0);
     }
@@ -103,7 +103,7 @@ private:
 
     std::array<ProfileInfo, MAX_USERS> profiles{};
     std::size_t user_count{};
-    Common::UUID last_opened_user{Common::INVALID_UUID};
+    Common::UUID last_opened_user{};
 };
 
 }; // namespace Service::Account
diff --git a/src/core/hle/service/am/am.cpp b/src/core/hle/service/am/am.cpp
index e60661fe1..420de3c54 100644
--- a/src/core/hle/service/am/am.cpp
+++ b/src/core/hle/service/am/am.cpp
@@ -55,7 +55,7 @@ constexpr u32 LAUNCH_PARAMETER_ACCOUNT_PRESELECTED_USER_MAGIC = 0xC79497CA;
 struct LaunchParameterAccountPreselectedUser {
     u32_le magic;
     u32_le is_account_selected;
-    u128 current_user;
+    Common::UUID current_user;
     INSERT_PADDING_BYTES(0x70);
 };
 static_assert(sizeof(LaunchParameterAccountPreselectedUser) == 0x88);
@@ -618,7 +618,7 @@ void AppletMessageQueue::PushMessage(AppletMessage msg) {
 AppletMessageQueue::AppletMessage AppletMessageQueue::PopMessage() {
     if (messages.empty()) {
         on_new_message->GetWritableEvent().Clear();
-        return AppletMessage::NoMessage;
+        return AppletMessage::None;
     }
     auto msg = messages.front();
     messages.pop();
@@ -633,7 +633,7 @@ std::size_t AppletMessageQueue::GetMessageCount() const {
 }
 
 void AppletMessageQueue::RequestExit() {
-    PushMessage(AppletMessage::ExitRequested);
+    PushMessage(AppletMessage::Exit);
 }
 
 void AppletMessageQueue::FocusStateChanged() {
@@ -732,7 +732,7 @@ void ICommonStateGetter::ReceiveMessage(Kernel::HLERequestContext& ctx) {
     const auto message = msg_queue->PopMessage();
     IPC::ResponseBuilder rb{ctx, 3};
 
-    if (message == AppletMessageQueue::AppletMessage::NoMessage) {
+    if (message == AppletMessageQueue::AppletMessage::None) {
         LOG_ERROR(Service_AM, "Message queue is empty");
         rb.Push(ERR_NO_MESSAGES);
         rb.PushEnum<AppletMessageQueue::AppletMessage>(message);
@@ -980,7 +980,7 @@ private:
         LOG_DEBUG(Service_AM, "called");
 
         IPC::RequestParser rp{ctx};
-        applet->GetBroker().PushNormalDataFromGame(rp.PopIpcInterface<IStorage>());
+        applet->GetBroker().PushNormalDataFromGame(rp.PopIpcInterface<IStorage>().lock());
 
         IPC::ResponseBuilder rb{ctx, 2};
         rb.Push(ResultSuccess);
@@ -1007,7 +1007,7 @@ private:
         LOG_DEBUG(Service_AM, "called");
 
         IPC::RequestParser rp{ctx};
-        applet->GetBroker().PushInteractiveDataFromGame(rp.PopIpcInterface<IStorage>());
+        applet->GetBroker().PushInteractiveDataFromGame(rp.PopIpcInterface<IStorage>().lock());
 
         ASSERT(applet->IsInitialized());
         applet->ExecuteInteractive();
@@ -1453,8 +1453,8 @@ void IApplicationFunctions::PopLaunchParameter(Kernel::HLERequestContext& ctx) {
 
         Account::ProfileManager profile_manager{};
         const auto uuid = profile_manager.GetUser(static_cast<s32>(Settings::values.current_user));
-        ASSERT(uuid);
-        params.current_user = uuid->uuid;
+        ASSERT(uuid.has_value() && uuid->IsValid());
+        params.current_user = *uuid;
 
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
 
diff --git a/src/core/hle/service/am/am.h b/src/core/hle/service/am/am.h
index 2a578aea5..fdd937b82 100644
--- a/src/core/hle/service/am/am.h
+++ b/src/core/hle/service/am/am.h
@@ -22,6 +22,7 @@ class NVFlinger;
 
 namespace Service::AM {
 
+// This is nn::settings::Language
 enum SystemLanguage {
     Japanese = 0,
     English = 1, // en-US
@@ -41,16 +42,44 @@ enum SystemLanguage {
     // 4.0.0+
     SimplifiedChinese = 15,
     TraditionalChinese = 16,
+    // 10.1.0+
+    BrazilianPortuguese = 17,
 };
 
 class AppletMessageQueue {
 public:
+    // This is nn::am::AppletMessage
     enum class AppletMessage : u32 {
-        NoMessage = 0,
-        ExitRequested = 4,
+        None = 0,
+        ChangeIntoForeground = 1,
+        ChangeIntoBackground = 2,
+        Exit = 4,
+        ApplicationExited = 6,
         FocusStateChanged = 15,
+        Resume = 16,
+        DetectShortPressingHomeButton = 20,
+        DetectLongPressingHomeButton = 21,
+        DetectShortPressingPowerButton = 22,
+        DetectMiddlePressingPowerButton = 23,
+        DetectLongPressingPowerButton = 24,
+        RequestToPrepareSleep = 25,
+        FinishedSleepSequence = 26,
+        SleepRequiredByHighTemperature = 27,
+        SleepRequiredByLowBattery = 28,
+        AutoPowerDown = 29,
         OperationModeChanged = 30,
         PerformanceModeChanged = 31,
+        DetectReceivingCecSystemStandby = 32,
+        SdCardRemoved = 33,
+        LaunchApplicationRequested = 50,
+        RequestToDisplay = 51,
+        ShowApplicationLogo = 55,
+        HideApplicationLogo = 56,
+        ForceHideApplicationLogo = 57,
+        FloatingApplicationDetected = 60,
+        DetectShortPressingCaptureButton = 90,
+        AlbumScreenShotTaken = 92,
+        AlbumRecordingSaved = 93,
     };
 
     explicit AppletMessageQueue(Core::System& system);
@@ -179,11 +208,14 @@ public:
     ~ICommonStateGetter() override;
 
 private:
+    // This is nn::oe::FocusState
     enum class FocusState : u8 {
         InFocus = 1,
         NotInFocus = 2,
+        Background = 3,
     };
 
+    // This is nn::oe::OperationMode
     enum class OperationMode : u8 {
         Handheld = 0,
         Docked = 1,
diff --git a/src/core/hle/service/am/applets/applet_mii.cpp b/src/core/hle/service/am/applets/applet_mii.cpp
new file mode 100644
index 000000000..8c4173737
--- /dev/null
+++ b/src/core/hle/service/am/applets/applet_mii.cpp
@@ -0,0 +1,101 @@
+// Copyright 2022 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include "common/assert.h"
+#include "common/logging/log.h"
+#include "core/core.h"
+#include "core/frontend/applets/mii.h"
+#include "core/hle/service/am/am.h"
+#include "core/hle/service/am/applets/applet_mii.h"
+#include "core/reporter.h"
+
+namespace Service::AM::Applets {
+
+Mii::Mii(Core::System& system_, LibraryAppletMode applet_mode_,
+         const Core::Frontend::MiiApplet& frontend_)
+    : Applet{system_, applet_mode_}, frontend{frontend_}, system{system_} {}
+
+Mii::~Mii() = default;
+
+void Mii::Initialize() {
+    is_complete = false;
+
+    const auto storage = broker.PopNormalDataToApplet();
+    ASSERT(storage != nullptr);
+
+    const auto data = storage->GetData();
+    ASSERT(data.size() == sizeof(MiiAppletInput));
+
+    std::memcpy(&input_data, data.data(), sizeof(MiiAppletInput));
+}
+
+bool Mii::TransactionComplete() const {
+    return is_complete;
+}
+
+ResultCode Mii::GetStatus() const {
+    return ResultSuccess;
+}
+
+void Mii::ExecuteInteractive() {
+    UNREACHABLE_MSG("Unexpected interactive applet data!");
+}
+
+void Mii::Execute() {
+    if (is_complete) {
+        return;
+    }
+
+    const auto callback = [this](const Core::Frontend::MiiParameters& parameters) {
+        DisplayCompleted(parameters);
+    };
+
+    switch (input_data.applet_mode) {
+    case MiiAppletMode::ShowMiiEdit: {
+        Service::Mii::MiiManager manager;
+        Core::Frontend::MiiParameters params{
+            .is_editable = false,
+            .mii_data = input_data.mii_char_info.mii_data,
+        };
+        frontend.ShowMii(params, callback);
+        break;
+    }
+    case MiiAppletMode::EditMii: {
+        Service::Mii::MiiManager manager;
+        Core::Frontend::MiiParameters params{
+            .is_editable = true,
+            .mii_data = input_data.mii_char_info.mii_data,
+        };
+        frontend.ShowMii(params, callback);
+        break;
+    }
+    case MiiAppletMode::CreateMii: {
+        Service::Mii::MiiManager manager;
+        Core::Frontend::MiiParameters params{
+            .is_editable = true,
+            .mii_data = manager.BuildDefault(0),
+        };
+        frontend.ShowMii(params, callback);
+        break;
+    }
+    default:
+        UNIMPLEMENTED_MSG("Unimplemented LibAppletMiiEdit mode={:02X}!", input_data.applet_mode);
+    }
+}
+
+void Mii::DisplayCompleted(const Core::Frontend::MiiParameters& parameters) {
+    is_complete = true;
+
+    std::vector<u8> reply(sizeof(AppletOutputForCharInfoEditing));
+    output_data = {
+        .result = ResultSuccess,
+        .mii_data = parameters.mii_data,
+    };
+
+    std::memcpy(reply.data(), &output_data, sizeof(AppletOutputForCharInfoEditing));
+    broker.PushNormalDataFromApplet(std::make_shared<IStorage>(system, std::move(reply)));
+    broker.SignalStateChanged();
+}
+
+} // namespace Service::AM::Applets
diff --git a/src/core/hle/service/am/applets/applet_mii.h b/src/core/hle/service/am/applets/applet_mii.h
new file mode 100644
index 000000000..42326bfc2
--- /dev/null
+++ b/src/core/hle/service/am/applets/applet_mii.h
@@ -0,0 +1,90 @@
+// Copyright 2022 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include <array>
+
+#include "core/hle/result.h"
+#include "core/hle/service/am/applets/applets.h"
+#include "core/hle/service/mii/mii_manager.h"
+
+namespace Core {
+class System;
+}
+
+namespace Service::AM::Applets {
+
+// This is nn::mii::AppletMode
+enum class MiiAppletMode : u32 {
+    ShowMiiEdit = 0,
+    AppendMii = 1,
+    AppendMiiImage = 2,
+    UpdateMiiImage = 3,
+    CreateMii = 4,
+    EditMii = 5,
+};
+
+struct MiiCharInfo {
+    Service::Mii::MiiInfo mii_data{};
+    INSERT_PADDING_BYTES(0x28);
+};
+static_assert(sizeof(MiiCharInfo) == 0x80, "MiiCharInfo has incorrect size.");
+
+// This is nn::mii::AppletInput
+struct MiiAppletInput {
+    s32 version{};
+    MiiAppletMode applet_mode{};
+    u32 special_mii_key_code{};
+    union {
+        std::array<Common::UUID, 8> valid_uuid;
+        MiiCharInfo mii_char_info;
+    };
+    Common::UUID used_uuid;
+    INSERT_PADDING_BYTES(0x64);
+};
+static_assert(sizeof(MiiAppletInput) == 0x100, "MiiAppletInput has incorrect size.");
+
+// This is nn::mii::AppletOutput
+struct MiiAppletOutput {
+    ResultCode result{ResultSuccess};
+    s32 index{};
+    INSERT_PADDING_BYTES(0x18);
+};
+static_assert(sizeof(MiiAppletOutput) == 0x20, "MiiAppletOutput has incorrect size.");
+
+// This is nn::mii::AppletOutputForCharInfoEditing
+struct AppletOutputForCharInfoEditing {
+    ResultCode result{ResultSuccess};
+    Service::Mii::MiiInfo mii_data{};
+    INSERT_PADDING_BYTES(0x24);
+};
+static_assert(sizeof(AppletOutputForCharInfoEditing) == 0x80,
+              "AppletOutputForCharInfoEditing has incorrect size.");
+
+class Mii final : public Applet {
+public:
+    explicit Mii(Core::System& system_, LibraryAppletMode applet_mode_,
+                 const Core::Frontend::MiiApplet& frontend_);
+    ~Mii() override;
+
+    void Initialize() override;
+
+    bool TransactionComplete() const override;
+    ResultCode GetStatus() const override;
+    void ExecuteInteractive() override;
+    void Execute() override;
+
+    void DisplayCompleted(const Core::Frontend::MiiParameters& parameters);
+
+private:
+    const Core::Frontend::MiiApplet& frontend;
+    MiiAppletInput input_data{};
+    AppletOutputForCharInfoEditing output_data{};
+
+    bool is_complete = false;
+    Core::System& system;
+};
+
+} // namespace Service::AM::Applets
diff --git a/src/core/hle/service/am/applets/applet_profile_select.cpp b/src/core/hle/service/am/applets/applet_profile_select.cpp
index a6e891944..82500e121 100644
--- a/src/core/hle/service/am/applets/applet_profile_select.cpp
+++ b/src/core/hle/service/am/applets/applet_profile_select.cpp
@@ -62,11 +62,11 @@ void ProfileSelect::SelectionComplete(std::optional<Common::UUID> uuid) {
 
     if (uuid.has_value() && uuid->IsValid()) {
         output.result = 0;
-        output.uuid_selected = uuid->uuid;
+        output.uuid_selected = *uuid;
     } else {
         status = ERR_USER_CANCELLED_SELECTION;
         output.result = ERR_USER_CANCELLED_SELECTION.raw;
-        output.uuid_selected = Common::INVALID_UUID;
+        output.uuid_selected = Common::InvalidUUID;
     }
 
     final_data = std::vector<u8>(sizeof(UserSelectionOutput));
diff --git a/src/core/hle/service/am/applets/applet_profile_select.h b/src/core/hle/service/am/applets/applet_profile_select.h
index 8fb76e6c4..852e1e0c0 100644
--- a/src/core/hle/service/am/applets/applet_profile_select.h
+++ b/src/core/hle/service/am/applets/applet_profile_select.h
@@ -27,7 +27,7 @@ static_assert(sizeof(UserSelectionConfig) == 0xA0, "UserSelectionConfig has inco
 
 struct UserSelectionOutput {
     u64 result;
-    u128 uuid_selected;
+    Common::UUID uuid_selected;
 };
 static_assert(sizeof(UserSelectionOutput) == 0x18, "UserSelectionOutput has incorrect size.");
 
diff --git a/src/core/hle/service/am/applets/applets.cpp b/src/core/hle/service/am/applets/applets.cpp
index 134ac1ee2..79e62679d 100644
--- a/src/core/hle/service/am/applets/applets.cpp
+++ b/src/core/hle/service/am/applets/applets.cpp
@@ -9,6 +9,7 @@
 #include "core/frontend/applets/controller.h"
 #include "core/frontend/applets/error.h"
 #include "core/frontend/applets/general_frontend.h"
+#include "core/frontend/applets/mii.h"
 #include "core/frontend/applets/profile_select.h"
 #include "core/frontend/applets/software_keyboard.h"
 #include "core/frontend/applets/web_browser.h"
@@ -19,6 +20,7 @@
 #include "core/hle/service/am/applets/applet_controller.h"
 #include "core/hle/service/am/applets/applet_error.h"
 #include "core/hle/service/am/applets/applet_general_backend.h"
+#include "core/hle/service/am/applets/applet_mii.h"
 #include "core/hle/service/am/applets/applet_profile_select.h"
 #include "core/hle/service/am/applets/applet_software_keyboard.h"
 #include "core/hle/service/am/applets/applet_web_browser.h"
@@ -172,10 +174,11 @@ AppletFrontendSet::AppletFrontendSet() = default;
 
 AppletFrontendSet::AppletFrontendSet(ControllerApplet controller_applet, ErrorApplet error_applet,
                                      ParentalControlsApplet parental_controls_applet,
-                                     PhotoViewer photo_viewer_, ProfileSelect profile_select_,
+                                     MiiApplet mii_applet, PhotoViewer photo_viewer_,
+                                     ProfileSelect profile_select_,
                                      SoftwareKeyboard software_keyboard_, WebBrowser web_browser_)
     : controller{std::move(controller_applet)}, error{std::move(error_applet)},
-      parental_controls{std::move(parental_controls_applet)},
+      parental_controls{std::move(parental_controls_applet)}, mii{std::move(mii_applet)},
       photo_viewer{std::move(photo_viewer_)}, profile_select{std::move(profile_select_)},
       software_keyboard{std::move(software_keyboard_)}, web_browser{std::move(web_browser_)} {}
 
@@ -206,6 +209,10 @@ void AppletManager::SetAppletFrontendSet(AppletFrontendSet set) {
         frontend.parental_controls = std::move(set.parental_controls);
     }
 
+    if (set.mii != nullptr) {
+        frontend.mii = std::move(set.mii);
+    }
+
     if (set.photo_viewer != nullptr) {
         frontend.photo_viewer = std::move(set.photo_viewer);
     }
@@ -243,6 +250,10 @@ void AppletManager::SetDefaultAppletsIfMissing() {
             std::make_unique<Core::Frontend::DefaultParentalControlsApplet>();
     }
 
+    if (frontend.mii == nullptr) {
+        frontend.mii = std::make_unique<Core::Frontend::DefaultMiiApplet>();
+    }
+
     if (frontend.photo_viewer == nullptr) {
         frontend.photo_viewer = std::make_unique<Core::Frontend::DefaultPhotoViewerApplet>();
     }
@@ -277,6 +288,8 @@ std::shared_ptr<Applet> AppletManager::GetApplet(AppletId id, LibraryAppletMode
         return std::make_shared<ProfileSelect>(system, mode, *frontend.profile_select);
     case AppletId::SoftwareKeyboard:
         return std::make_shared<SoftwareKeyboard>(system, mode, *frontend.software_keyboard);
+    case AppletId::MiiEdit:
+        return std::make_shared<Mii>(system, mode, *frontend.mii);
     case AppletId::Web:
     case AppletId::Shop:
     case AppletId::OfflineWeb:
diff --git a/src/core/hle/service/am/applets/applets.h b/src/core/hle/service/am/applets/applets.h
index 15eeb4ee1..0c44aec79 100644
--- a/src/core/hle/service/am/applets/applets.h
+++ b/src/core/hle/service/am/applets/applets.h
@@ -21,6 +21,7 @@ class ControllerApplet;
 class ECommerceApplet;
 class ErrorApplet;
 class ParentalControlsApplet;
+class MiiApplet;
 class PhotoViewerApplet;
 class ProfileSelectApplet;
 class SoftwareKeyboardApplet;
@@ -179,6 +180,7 @@ struct AppletFrontendSet {
     using ControllerApplet = std::unique_ptr<Core::Frontend::ControllerApplet>;
     using ErrorApplet = std::unique_ptr<Core::Frontend::ErrorApplet>;
     using ParentalControlsApplet = std::unique_ptr<Core::Frontend::ParentalControlsApplet>;
+    using MiiApplet = std::unique_ptr<Core::Frontend::MiiApplet>;
     using PhotoViewer = std::unique_ptr<Core::Frontend::PhotoViewerApplet>;
     using ProfileSelect = std::unique_ptr<Core::Frontend::ProfileSelectApplet>;
     using SoftwareKeyboard = std::unique_ptr<Core::Frontend::SoftwareKeyboardApplet>;
@@ -186,9 +188,9 @@ struct AppletFrontendSet {
 
     AppletFrontendSet();
     AppletFrontendSet(ControllerApplet controller_applet, ErrorApplet error_applet,
-                      ParentalControlsApplet parental_controls_applet, PhotoViewer photo_viewer_,
-                      ProfileSelect profile_select_, SoftwareKeyboard software_keyboard_,
-                      WebBrowser web_browser_);
+                      ParentalControlsApplet parental_controls_applet, MiiApplet mii_applet,
+                      PhotoViewer photo_viewer_, ProfileSelect profile_select_,
+                      SoftwareKeyboard software_keyboard_, WebBrowser web_browser_);
     ~AppletFrontendSet();
 
     AppletFrontendSet(const AppletFrontendSet&) = delete;
@@ -200,6 +202,7 @@ struct AppletFrontendSet {
     ControllerApplet controller;
     ErrorApplet error;
     ParentalControlsApplet parental_controls;
+    MiiApplet mii;
     PhotoViewer photo_viewer;
     ProfileSelect profile_select;
     SoftwareKeyboard software_keyboard;
diff --git a/src/core/hle/service/apm/apm_controller.cpp b/src/core/hle/service/apm/apm_controller.cpp
index 98839fe97..187fef2ad 100644
--- a/src/core/hle/service/apm/apm_controller.cpp
+++ b/src/core/hle/service/apm/apm_controller.cpp
@@ -17,8 +17,8 @@ constexpr auto DEFAULT_PERFORMANCE_CONFIGURATION = PerformanceConfiguration::Con
 
 Controller::Controller(Core::Timing::CoreTiming& core_timing_)
     : core_timing{core_timing_}, configs{
-                                     {PerformanceMode::Handheld, DEFAULT_PERFORMANCE_CONFIGURATION},
-                                     {PerformanceMode::Docked, DEFAULT_PERFORMANCE_CONFIGURATION},
+                                     {PerformanceMode::Normal, DEFAULT_PERFORMANCE_CONFIGURATION},
+                                     {PerformanceMode::Boost, DEFAULT_PERFORMANCE_CONFIGURATION},
                                  } {}
 
 Controller::~Controller() = default;
@@ -63,13 +63,13 @@ void Controller::SetFromCpuBoostMode(CpuBoostMode mode) {
         PerformanceConfiguration::Config15,
     }};
 
-    SetPerformanceConfiguration(PerformanceMode::Docked,
+    SetPerformanceConfiguration(PerformanceMode::Boost,
                                 BOOST_MODE_TO_CONFIG_MAP.at(static_cast<u32>(mode)));
 }
 
 PerformanceMode Controller::GetCurrentPerformanceMode() const {
-    return Settings::values.use_docked_mode.GetValue() ? PerformanceMode::Docked
-                                                       : PerformanceMode::Handheld;
+    return Settings::values.use_docked_mode.GetValue() ? PerformanceMode::Boost
+                                                       : PerformanceMode::Normal;
 }
 
 PerformanceConfiguration Controller::GetCurrentPerformanceConfiguration(PerformanceMode mode) {
diff --git a/src/core/hle/service/apm/apm_controller.h b/src/core/hle/service/apm/apm_controller.h
index 8d48e0104..d6fbd2c0c 100644
--- a/src/core/hle/service/apm/apm_controller.h
+++ b/src/core/hle/service/apm/apm_controller.h
@@ -32,15 +32,18 @@ enum class PerformanceConfiguration : u32 {
     Config16 = 0x9222000C,
 };
 
+// This is nn::oe::CpuBoostMode
 enum class CpuBoostMode : u32 {
-    Disabled = 0,
-    Full = 1,    // CPU + GPU -> Config 13, 14, 15, or 16
-    Partial = 2, // GPU Only -> Config 15 or 16
+    Normal = 0,   // Boost mode disabled
+    FastLoad = 1, // CPU + GPU -> Config 13, 14, 15, or 16
+    Partial = 2,  // GPU Only -> Config 15 or 16
 };
 
-enum class PerformanceMode : u8 {
-    Handheld = 0,
-    Docked = 1,
+// This is nn::oe::PerformanceMode
+enum class PerformanceMode : s32 {
+    Invalid = -1,
+    Normal = 0,
+    Boost = 1,
 };
 
 // Class to manage the state and change of the emulated system performance.
diff --git a/src/core/hle/service/friend/friend.cpp b/src/core/hle/service/friend/friend.cpp
index 9f9cea1e0..79cd3acbb 100644
--- a/src/core/hle/service/friend/friend.cpp
+++ b/src/core/hle/service/friend/friend.cpp
@@ -173,7 +173,7 @@ private:
         const auto uuid = rp.PopRaw<Common::UUID>();
 
         LOG_WARNING(Service_Friend, "(STUBBED) called, local_play={}, uuid=0x{}", local_play,
-                    uuid.Format());
+                    uuid.RawString());
 
         IPC::ResponseBuilder rb{ctx, 2};
         rb.Push(ResultSuccess);
@@ -186,7 +186,7 @@ private:
         [[maybe_unused]] const auto filter = rp.PopRaw<SizedFriendFilter>();
         const auto pid = rp.Pop<u64>();
         LOG_WARNING(Service_Friend, "(STUBBED) called, offset={}, uuid=0x{}, pid={}", friend_offset,
-                    uuid.Format(), pid);
+                    uuid.RawString(), pid);
 
         IPC::ResponseBuilder rb{ctx, 3};
         rb.Push(ResultSuccess);
@@ -312,7 +312,7 @@ void Module::Interface::CreateNotificationService(Kernel::HLERequestContext& ctx
     IPC::RequestParser rp{ctx};
     auto uuid = rp.PopRaw<Common::UUID>();
 
-    LOG_DEBUG(Service_Friend, "called, uuid=0x{}", uuid.Format());
+    LOG_DEBUG(Service_Friend, "called, uuid=0x{}", uuid.RawString());
 
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(ResultSuccess);
diff --git a/src/core/hle/service/hid/hid.cpp b/src/core/hle/service/hid/hid.cpp
index a2bf7defb..d9202ea6c 100644
--- a/src/core/hle/service/hid/hid.cpp
+++ b/src/core/hle/service/hid/hid.cpp
@@ -320,7 +320,7 @@ Hid::Hid(Core::System& system_)
         {308, nullptr, "SetSevenSixAxisSensorFusionStrength"},
         {309, nullptr, "GetSevenSixAxisSensorFusionStrength"},
         {310, &Hid::ResetSevenSixAxisSensorTimestamp, "ResetSevenSixAxisSensorTimestamp"},
-        {400, nullptr, "IsUsbFullKeyControllerEnabled"},
+        {400, &Hid::IsUsbFullKeyControllerEnabled, "IsUsbFullKeyControllerEnabled"},
         {401, nullptr, "EnableUsbFullKeyController"},
         {402, nullptr, "IsUsbFullKeyControllerConnected"},
         {403, nullptr, "HasBattery"},
@@ -1673,6 +1673,16 @@ void Hid::ResetSevenSixAxisSensorTimestamp(Kernel::HLERequestContext& ctx) {
     rb.Push(ResultSuccess);
 }
 
+void Hid::IsUsbFullKeyControllerEnabled(Kernel::HLERequestContext& ctx) {
+    IPC::RequestParser rp{ctx};
+
+    LOG_WARNING(Service_HID, "(STUBBED) called");
+
+    IPC::ResponseBuilder rb{ctx, 3};
+    rb.Push(ResultSuccess);
+    rb.Push(false);
+}
+
 void Hid::SetIsPalmaAllConnectable(Kernel::HLERequestContext& ctx) {
     IPC::RequestParser rp{ctx};
     const auto applet_resource_user_id{rp.Pop<u64>()};
diff --git a/src/core/hle/service/hid/hid.h b/src/core/hle/service/hid/hid.h
index d290df161..c281081a7 100644
--- a/src/core/hle/service/hid/hid.h
+++ b/src/core/hle/service/hid/hid.h
@@ -159,6 +159,7 @@ private:
     void InitializeSevenSixAxisSensor(Kernel::HLERequestContext& ctx);
     void FinalizeSevenSixAxisSensor(Kernel::HLERequestContext& ctx);
     void ResetSevenSixAxisSensorTimestamp(Kernel::HLERequestContext& ctx);
+    void IsUsbFullKeyControllerEnabled(Kernel::HLERequestContext& ctx);
     void SetIsPalmaAllConnectable(Kernel::HLERequestContext& ctx);
     void SetPalmaBoostMode(Kernel::HLERequestContext& ctx);
     void SetNpadCommunicationMode(Kernel::HLERequestContext& ctx);
diff --git a/src/core/hle/service/kernel_helpers.cpp b/src/core/hle/service/kernel_helpers.cpp
index 62f4cdfb2..ff0bbb788 100644
--- a/src/core/hle/service/kernel_helpers.cpp
+++ b/src/core/hle/service/kernel_helpers.cpp
@@ -3,7 +3,9 @@
 // Refer to the license.txt file included.
 
 #include "core/core.h"
+#include "core/core_timing.h"
 #include "core/hle/kernel/k_event.h"
+#include "core/hle/kernel/k_memory_manager.h"
 #include "core/hle/kernel/k_process.h"
 #include "core/hle/kernel/k_readable_event.h"
 #include "core/hle/kernel/k_resource_limit.h"
@@ -15,9 +17,11 @@ namespace Service::KernelHelpers {
 
 ServiceContext::ServiceContext(Core::System& system_, std::string name_)
     : kernel(system_.Kernel()) {
+    // Create the process.
     process = Kernel::KProcess::Create(kernel);
     ASSERT(Kernel::KProcess::Initialize(process, system_, std::move(name_),
-                                        Kernel::KProcess::ProcessType::Userland)
+                                        Kernel::KProcess::ProcessType::KernelInternal,
+                                        kernel.GetSystemResourceLimit())
                .IsSuccess());
 }
 
@@ -43,7 +47,7 @@ Kernel::KEvent* ServiceContext::CreateEvent(std::string&& name) {
     }
 
     // Initialize the event.
-    event->Initialize(std::move(name));
+    event->Initialize(std::move(name), process);
 
     // Commit the thread reservation.
     event_reservation.Commit();
diff --git a/src/core/hle/service/ldr/ldr.cpp b/src/core/hle/service/ldr/ldr.cpp
index 9fc7bb1b1..099276420 100644
--- a/src/core/hle/service/ldr/ldr.cpp
+++ b/src/core/hle/service/ldr/ldr.cpp
@@ -288,7 +288,7 @@ public:
     }
 
     bool ValidateRegionForMap(Kernel::KPageTable& page_table, VAddr start, std::size_t size) const {
-        constexpr std::size_t padding_size{4 * Kernel::PageSize};
+        const std::size_t padding_size{page_table.GetNumGuardPages() * Kernel::PageSize};
         const auto start_info{page_table.QueryInfo(start - 1)};
 
         if (start_info.state != Kernel::KMemoryState::Free) {
@@ -308,31 +308,69 @@ public:
         return (start + size + padding_size) <= (end_info.GetAddress() + end_info.GetSize());
     }
 
-    VAddr GetRandomMapRegion(const Kernel::KPageTable& page_table, std::size_t size) const {
-        VAddr addr{};
-        const std::size_t end_pages{(page_table.GetAliasCodeRegionSize() - size) >>
-                                    Kernel::PageBits};
-        do {
-            addr = page_table.GetAliasCodeRegionStart() +
-                   (Kernel::KSystemControl::GenerateRandomRange(0, end_pages) << Kernel::PageBits);
-        } while (!page_table.IsInsideAddressSpace(addr, size) ||
-                 page_table.IsInsideHeapRegion(addr, size) ||
-                 page_table.IsInsideAliasRegion(addr, size));
-        return addr;
+    ResultCode GetAvailableMapRegion(Kernel::KPageTable& page_table, u64 size, VAddr& out_addr) {
+        size = Common::AlignUp(size, Kernel::PageSize);
+        size += page_table.GetNumGuardPages() * Kernel::PageSize * 4;
+
+        const auto is_region_available = [&](VAddr addr) {
+            const auto end_addr = addr + size;
+            while (addr < end_addr) {
+                if (system.Memory().IsValidVirtualAddress(addr)) {
+                    return false;
+                }
+
+                if (!page_table.IsInsideAddressSpace(out_addr, size)) {
+                    return false;
+                }
+
+                if (page_table.IsInsideHeapRegion(out_addr, size)) {
+                    return false;
+                }
+
+                if (page_table.IsInsideAliasRegion(out_addr, size)) {
+                    return false;
+                }
+
+                addr += Kernel::PageSize;
+            }
+            return true;
+        };
+
+        bool succeeded = false;
+        const auto map_region_end =
+            page_table.GetAliasCodeRegionStart() + page_table.GetAliasCodeRegionSize();
+        while (current_map_addr < map_region_end) {
+            if (is_region_available(current_map_addr)) {
+                succeeded = true;
+                break;
+            }
+            current_map_addr += 0x100000;
+        }
+
+        if (!succeeded) {
+            UNREACHABLE_MSG("Out of address space!");
+            return Kernel::ResultOutOfMemory;
+        }
+
+        out_addr = current_map_addr;
+        current_map_addr += size;
+
+        return ResultSuccess;
     }
 
-    ResultVal<VAddr> MapProcessCodeMemory(Kernel::KProcess* process, VAddr baseAddress,
-                                          u64 size) const {
+    ResultVal<VAddr> MapProcessCodeMemory(Kernel::KProcess* process, VAddr base_addr, u64 size) {
+        auto& page_table{process->PageTable()};
+        VAddr addr{};
+
         for (std::size_t retry = 0; retry < MAXIMUM_MAP_RETRIES; retry++) {
-            auto& page_table{process->PageTable()};
-            const VAddr addr{GetRandomMapRegion(page_table, size)};
-            const ResultCode result{page_table.MapCodeMemory(addr, baseAddress, size)};
+            R_TRY(GetAvailableMapRegion(page_table, size, addr));
 
+            const ResultCode result{page_table.MapCodeMemory(addr, base_addr, size)};
             if (result == Kernel::ResultInvalidCurrentMemory) {
                 continue;
             }
 
-            CASCADE_CODE(result);
+            R_TRY(result);
 
             if (ValidateRegionForMap(page_table, addr, size)) {
                 return addr;
@@ -343,7 +381,7 @@ public:
     }
 
     ResultVal<VAddr> MapNro(Kernel::KProcess* process, VAddr nro_addr, std::size_t nro_size,
-                            VAddr bss_addr, std::size_t bss_size, std::size_t size) const {
+                            VAddr bss_addr, std::size_t bss_size, std::size_t size) {
         for (std::size_t retry = 0; retry < MAXIMUM_MAP_RETRIES; retry++) {
             auto& page_table{process->PageTable()};
             VAddr addr{};
@@ -597,6 +635,7 @@ public:
         LOG_WARNING(Service_LDR, "(STUBBED) called");
 
         initialized = true;
+        current_map_addr = system.CurrentProcess()->PageTable().GetAliasCodeRegionStart();
 
         IPC::ResponseBuilder rb{ctx, 2};
         rb.Push(ResultSuccess);
@@ -607,6 +646,7 @@ private:
 
     std::map<VAddr, NROInfo> nro;
     std::map<VAddr, std::vector<SHA256Hash>> nrr;
+    VAddr current_map_addr{};
 
     bool IsValidNROHash(const SHA256Hash& hash) const {
         return std::any_of(nrr.begin(), nrr.end(), [&hash](const auto& p) {
diff --git a/src/core/hle/service/mii/mii_manager.cpp b/src/core/hle/service/mii/mii_manager.cpp
index ca4ed35bb..0a57c3cde 100644
--- a/src/core/hle/service/mii/mii_manager.cpp
+++ b/src/core/hle/service/mii/mii_manager.cpp
@@ -118,16 +118,6 @@ u16 GenerateCrc16(const void* data, std::size_t size) {
     return Common::swap16(static_cast<u16>(crc));
 }
 
-Common::UUID GenerateValidUUID() {
-    auto uuid{Common::UUID::Generate()};
-
-    // Bit 7 must be set, and bit 6 unset for the UUID to be valid
-    uuid.uuid[1] &= 0xFFFFFFFFFFFFFF3FULL;
-    uuid.uuid[1] |= 0x0000000000000080ULL;
-
-    return uuid;
-}
-
 template <typename T>
 T GetRandomValue(T min, T max) {
     std::random_device device;
@@ -383,7 +373,7 @@ MiiStoreData::MiiStoreData() = default;
 MiiStoreData::MiiStoreData(const MiiStoreData::Name& name, const MiiStoreBitFields& bit_fields,
                            const Common::UUID& user_id) {
     data.name = name;
-    data.uuid = GenerateValidUUID();
+    data.uuid = Common::UUID::MakeRandomRFC4122V4();
 
     std::memcpy(data.data.data(), &bit_fields, sizeof(MiiStoreBitFields));
     data_crc = GenerateCrc16(data.data.data(), sizeof(data));
diff --git a/src/core/hle/service/mii/mii_manager.h b/src/core/hle/service/mii/mii_manager.h
index 8e048fc56..6999d15b1 100644
--- a/src/core/hle/service/mii/mii_manager.h
+++ b/src/core/hle/service/mii/mii_manager.h
@@ -202,7 +202,7 @@ struct MiiStoreData {
         static_assert(sizeof(MiiStoreBitFields) == sizeof(data), "data field has incorrect size.");
 
         Name name{};
-        Common::UUID uuid{Common::INVALID_UUID};
+        Common::UUID uuid{};
     } data;
 
     u16 data_crc{};
@@ -326,7 +326,7 @@ public:
     ResultCode GetIndex(const MiiInfo& info, u32& index);
 
 private:
-    const Common::UUID user_id{Common::INVALID_UUID};
+    const Common::UUID user_id{};
     u64 update_counter{};
 };
 
diff --git a/src/core/hle/service/mnpp/mnpp_app.cpp b/src/core/hle/service/mnpp/mnpp_app.cpp
new file mode 100644
index 000000000..53497612f
--- /dev/null
+++ b/src/core/hle/service/mnpp/mnpp_app.cpp
@@ -0,0 +1,45 @@
+// Copyright 2022 yuzu emulator team
+// 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/service/mnpp/mnpp_app.h"
+#include "core/hle/service/sm/sm.h"
+
+namespace Service::MNPP {
+
+class MNPP_APP final : public ServiceFramework<MNPP_APP> {
+public:
+    explicit MNPP_APP(Core::System& system_) : ServiceFramework{system_, "mnpp:app"} {
+        // clang-format off
+        static const FunctionInfo functions[] = {
+            {0, &MNPP_APP::Unknown0, "unknown0"},
+            {1, &MNPP_APP::Unknown1, "unknown1"},
+        };
+        // clang-format on
+
+        RegisterHandlers(functions);
+    }
+
+private:
+    void Unknown0(Kernel::HLERequestContext& ctx) {
+        LOG_WARNING(Service_MNPP, "(STUBBED) called");
+
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(ResultSuccess);
+    }
+
+    void Unknown1(Kernel::HLERequestContext& ctx) {
+        LOG_WARNING(Service_MNPP, "(STUBBED) called");
+
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(ResultSuccess);
+    }
+};
+
+void InstallInterfaces(SM::ServiceManager& service_manager, Core::System& system) {
+    std::make_shared<MNPP_APP>(system)->InstallAsService(service_manager);
+}
+
+} // namespace Service::MNPP
diff --git a/src/core/hle/service/mnpp/mnpp_app.h b/src/core/hle/service/mnpp/mnpp_app.h
new file mode 100644
index 000000000..6bf20b494
--- /dev/null
+++ b/src/core/hle/service/mnpp/mnpp_app.h
@@ -0,0 +1,20 @@
+// Copyright 2022 yuzu emulator team
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+namespace Core {
+class System;
+}
+
+namespace Service::SM {
+class ServiceManager;
+}
+
+namespace Service::MNPP {
+
+/// Registers all MNPP services with the specified service manager.
+void InstallInterfaces(SM::ServiceManager& service_manager, Core::System& system);
+
+} // namespace Service::MNPP
diff --git a/src/core/hle/service/nfp/nfp.cpp b/src/core/hle/service/nfp/nfp.cpp
index 761d0d3c6..513107715 100644
--- a/src/core/hle/service/nfp/nfp.cpp
+++ b/src/core/hle/service/nfp/nfp.cpp
@@ -7,6 +7,9 @@
 
 #include "common/logging/log.h"
 #include "core/core.h"
+#include "core/hid/emulated_controller.h"
+#include "core/hid/hid_core.h"
+#include "core/hid/hid_types.h"
 #include "core/hle/ipc_helpers.h"
 #include "core/hle/kernel/k_event.h"
 #include "core/hle/service/nfp/nfp.h"
@@ -14,343 +17,790 @@
 
 namespace Service::NFP {
 namespace ErrCodes {
-constexpr ResultCode ERR_NO_APPLICATION_AREA(ErrorModule::NFP, 152);
+constexpr ResultCode DeviceNotFound(ErrorModule::NFP, 64);
+constexpr ResultCode WrongDeviceState(ErrorModule::NFP, 73);
+constexpr ResultCode ApplicationAreaIsNotInitialized(ErrorModule::NFP, 128);
+constexpr ResultCode ApplicationAreaExist(ErrorModule::NFP, 168);
 } // namespace ErrCodes
 
-Module::Interface::Interface(std::shared_ptr<Module> module_, Core::System& system_,
-                             const char* name)
-    : ServiceFramework{system_, name}, module{std::move(module_)}, service_context{system_,
-                                                                                   "NFP::IUser"} {
-    nfc_tag_load = service_context.CreateEvent("NFP::IUser:NFCTagDetected");
-}
-
-Module::Interface::~Interface() {
-    service_context.CloseEvent(nfc_tag_load);
-}
-
-class IUser final : public ServiceFramework<IUser> {
-public:
-    explicit IUser(Module::Interface& nfp_interface_, Core::System& system_,
-                   KernelHelpers::ServiceContext& service_context_)
-        : ServiceFramework{system_, "NFP::IUser"}, nfp_interface{nfp_interface_},
-          service_context{service_context_} {
-        static const FunctionInfo functions[] = {
-            {0, &IUser::Initialize, "Initialize"},
-            {1, &IUser::Finalize, "Finalize"},
-            {2, &IUser::ListDevices, "ListDevices"},
-            {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, &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, &IUser::GetApplicationAreaSize, "GetApplicationAreaSize"},
-            {23, &IUser::AttachAvailabilityChangeEvent, "AttachAvailabilityChangeEvent"},
-            {24, nullptr, "RecreateApplicationArea"},
-        };
-        RegisterHandlers(functions);
+constexpr u32 ApplicationAreaSize = 0xD8;
+
+IUser::IUser(Module::Interface& nfp_interface_, Core::System& system_)
+    : ServiceFramework{system_, "NFP::IUser"}, service_context{system_, service_name},
+      nfp_interface{nfp_interface_} {
+    static const FunctionInfo functions[] = {
+        {0, &IUser::Initialize, "Initialize"},
+        {1, &IUser::Finalize, "Finalize"},
+        {2, &IUser::ListDevices, "ListDevices"},
+        {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, &IUser::SetApplicationArea, "SetApplicationArea"},
+        {10, nullptr, "Flush"},
+        {11, nullptr, "Restore"},
+        {12, &IUser::CreateApplicationArea, "CreateApplicationArea"},
+        {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, &IUser::GetApplicationAreaSize, "GetApplicationAreaSize"},
+        {23, &IUser::AttachAvailabilityChangeEvent, "AttachAvailabilityChangeEvent"},
+        {24, nullptr, "RecreateApplicationArea"},
+    };
+    RegisterHandlers(functions);
 
-        deactivate_event = service_context.CreateEvent("NFP::IUser:DeactivateEvent");
-        availability_change_event =
-            service_context.CreateEvent("NFP::IUser:AvailabilityChangeEvent");
-    }
+    availability_change_event = service_context.CreateEvent("IUser:AvailabilityChangeEvent");
+}
 
-    ~IUser() override {
-        service_context.CloseEvent(deactivate_event);
-        service_context.CloseEvent(availability_change_event);
-    }
+void IUser::Initialize(Kernel::HLERequestContext& ctx) {
+    LOG_INFO(Service_NFC, "called");
 
-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
-        std::array<u8, 0x15> padding_1;
-        u32_le protocol;
-        u32_le tag_type;
-        std::array<u8, 0x2c> padding_2;
-    };
-    static_assert(sizeof(TagInfo) == 0x54, "TagInfo is an invalid size");
+    state = State::Initialized;
 
-    enum class State : u32 {
-        NonInitialized = 0,
-        Initialized = 1,
-    };
+    // TODO(german77): Loop through all interfaces
+    nfp_interface.Initialize();
 
-    enum class DeviceState : u32 {
-        Initialized = 0,
-        SearchingForTag = 1,
-        TagFound = 2,
-        TagRemoved = 3,
-        TagNearby = 4,
-        Unknown5 = 5,
-        Finalized = 6
-    };
+    IPC::ResponseBuilder rb{ctx, 2, 0};
+    rb.Push(ResultSuccess);
+}
 
-    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 IUser::Finalize(Kernel::HLERequestContext& ctx) {
+    LOG_INFO(Service_NFP, "called");
 
-    void Initialize(Kernel::HLERequestContext& ctx) {
-        LOG_DEBUG(Service_NFC, "called");
+    state = State::NonInitialized;
 
-        IPC::ResponseBuilder rb{ctx, 2, 0};
-        rb.Push(ResultSuccess);
+    // TODO(german77): Loop through all interfaces
+    nfp_interface.Finalize();
 
-        state = State::Initialized;
-    }
+    IPC::ResponseBuilder rb{ctx, 2};
+    rb.Push(ResultSuccess);
+}
 
-    void GetState(Kernel::HLERequestContext& ctx) {
-        LOG_DEBUG(Service_NFC, "called");
+void IUser::ListDevices(Kernel::HLERequestContext& ctx) {
+    LOG_INFO(Service_NFP, "called");
 
-        IPC::ResponseBuilder rb{ctx, 3, 0};
-        rb.Push(ResultSuccess);
-        rb.PushRaw<u32>(static_cast<u32>(state));
+    std::vector<u64> devices;
+
+    // TODO(german77): Loop through all interfaces
+    devices.push_back(nfp_interface.GetHandle());
+
+    ctx.WriteBuffer(devices);
+
+    IPC::ResponseBuilder rb{ctx, 3};
+    rb.Push(ResultSuccess);
+    rb.Push(static_cast<s32>(devices.size()));
+}
+
+void IUser::StartDetection(Kernel::HLERequestContext& ctx) {
+    IPC::RequestParser rp{ctx};
+    const auto device_handle{rp.Pop<u64>()};
+    const auto nfp_protocol{rp.Pop<s32>()};
+    LOG_INFO(Service_NFP, "called, device_handle={}, nfp_protocol={}", device_handle, nfp_protocol);
+
+    // TODO(german77): Loop through all interfaces
+    if (device_handle == nfp_interface.GetHandle()) {
+        const auto result = nfp_interface.StartDetection(nfp_protocol);
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(result);
+        return;
     }
 
-    void ListDevices(Kernel::HLERequestContext& ctx) {
-        IPC::RequestParser rp{ctx};
-        const u32 array_size = rp.Pop<u32>();
-        LOG_DEBUG(Service_NFP, "called, array_size={}", array_size);
+    LOG_ERROR(Service_NFP, "Handle not found, device_handle={}", device_handle);
 
-        ctx.WriteBuffer(device_handle);
+    IPC::ResponseBuilder rb{ctx, 2};
+    rb.Push(ErrCodes::DeviceNotFound);
+}
 
-        IPC::ResponseBuilder rb{ctx, 3};
-        rb.Push(ResultSuccess);
-        rb.Push<u32>(1);
+void IUser::StopDetection(Kernel::HLERequestContext& ctx) {
+    IPC::RequestParser rp{ctx};
+    const auto device_handle{rp.Pop<u64>()};
+    LOG_INFO(Service_NFP, "called, device_handle={}", device_handle);
+
+    // TODO(german77): Loop through all interfaces
+    if (device_handle == nfp_interface.GetHandle()) {
+        const auto result = nfp_interface.StopDetection();
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(result);
+        return;
     }
 
-    void GetNpadId(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);
+    LOG_ERROR(Service_NFP, "Handle not found, device_handle={}", device_handle);
 
-        IPC::ResponseBuilder rb{ctx, 3};
-        rb.Push(ResultSuccess);
-        rb.Push<u32>(npad_id);
+    IPC::ResponseBuilder rb{ctx, 2};
+    rb.Push(ErrCodes::DeviceNotFound);
+}
+
+void IUser::Mount(Kernel::HLERequestContext& ctx) {
+    IPC::RequestParser rp{ctx};
+    const auto device_handle{rp.Pop<u64>()};
+    const auto model_type{rp.PopEnum<ModelType>()};
+    const auto mount_target{rp.PopEnum<MountTarget>()};
+    LOG_INFO(Service_NFP, "called, device_handle={}, model_type={}, mount_target={}", device_handle,
+             model_type, mount_target);
+
+    // TODO(german77): Loop through all interfaces
+    if (device_handle == nfp_interface.GetHandle()) {
+        const auto result = nfp_interface.Mount();
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(result);
+        return;
     }
 
-    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);
+    LOG_ERROR(Service_NFP, "Handle not found, device_handle={}", device_handle);
 
-        IPC::ResponseBuilder rb{ctx, 2, 1};
-        rb.Push(ResultSuccess);
-        rb.PushCopyObjects(nfp_interface.GetNFCEvent());
-        has_attached_handle = true;
+    IPC::ResponseBuilder rb{ctx, 2};
+    rb.Push(ErrCodes::DeviceNotFound);
+}
+
+void IUser::Unmount(Kernel::HLERequestContext& ctx) {
+    IPC::RequestParser rp{ctx};
+    const auto device_handle{rp.Pop<u64>()};
+    LOG_INFO(Service_NFP, "called, device_handle={}", device_handle);
+
+    // TODO(german77): Loop through all interfaces
+    if (device_handle == nfp_interface.GetHandle()) {
+        const auto result = nfp_interface.Unmount();
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(result);
+        return;
     }
 
-    void AttachDeactivateEvent(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);
+    LOG_ERROR(Service_NFP, "Handle not found, device_handle={}", device_handle);
 
-        IPC::ResponseBuilder rb{ctx, 2, 1};
-        rb.Push(ResultSuccess);
-        rb.PushCopyObjects(deactivate_event->GetReadableEvent());
-    }
-
-    void StopDetection(Kernel::HLERequestContext& ctx) {
-        LOG_DEBUG(Service_NFP, "called");
-
-        switch (device_state) {
-        case DeviceState::TagFound:
-        case DeviceState::TagNearby:
-            deactivate_event->GetWritableEvent().Signal();
-            device_state = DeviceState::Initialized;
-            break;
-        case DeviceState::SearchingForTag:
-        case DeviceState::TagRemoved:
-            device_state = DeviceState::Initialized;
-            break;
-        default:
-            break;
-        }
+    IPC::ResponseBuilder rb{ctx, 2};
+    rb.Push(ErrCodes::DeviceNotFound);
+}
+
+void IUser::OpenApplicationArea(Kernel::HLERequestContext& ctx) {
+    IPC::RequestParser rp{ctx};
+    const auto device_handle{rp.Pop<u64>()};
+    const auto access_id{rp.Pop<u32>()};
+    LOG_WARNING(Service_NFP, "(STUBBED) called, device_handle={}, access_id={}", device_handle,
+                access_id);
+
+    // TODO(german77): Loop through all interfaces
+    if (device_handle == nfp_interface.GetHandle()) {
+        const auto result = nfp_interface.OpenApplicationArea(access_id);
         IPC::ResponseBuilder rb{ctx, 2};
-        rb.Push(ResultSuccess);
+        rb.Push(result);
+        return;
     }
 
-    void GetDeviceState(Kernel::HLERequestContext& ctx) {
-        LOG_DEBUG(Service_NFP, "called");
+    LOG_ERROR(Service_NFP, "Handle not found, device_handle={}", device_handle);
+
+    IPC::ResponseBuilder rb{ctx, 2};
+    rb.Push(ErrCodes::DeviceNotFound);
+}
+
+void IUser::GetApplicationArea(Kernel::HLERequestContext& ctx) {
+    IPC::RequestParser rp{ctx};
+    const auto device_handle{rp.Pop<u64>()};
+    LOG_INFO(Service_NFP, "called, device_handle={}", device_handle);
 
+    // TODO(german77): Loop through all interfaces
+    if (device_handle == nfp_interface.GetHandle()) {
+        std::vector<u8> data{};
+        const auto result = nfp_interface.GetApplicationArea(data);
+        ctx.WriteBuffer(data);
         IPC::ResponseBuilder rb{ctx, 3};
-        rb.Push(ResultSuccess);
-        rb.Push<u32>(static_cast<u32>(device_state));
+        rb.Push(result);
+        rb.Push(static_cast<u32>(data.size()));
+        return;
     }
 
-    void StartDetection(Kernel::HLERequestContext& ctx) {
-        LOG_DEBUG(Service_NFP, "called");
+    LOG_ERROR(Service_NFP, "Handle not found, device_handle={}", device_handle);
 
-        if (device_state == DeviceState::Initialized || device_state == DeviceState::TagRemoved) {
-            device_state = DeviceState::SearchingForTag;
-        }
+    IPC::ResponseBuilder rb{ctx, 2};
+    rb.Push(ErrCodes::DeviceNotFound);
+}
+
+void IUser::SetApplicationArea(Kernel::HLERequestContext& ctx) {
+    IPC::RequestParser rp{ctx};
+    const auto device_handle{rp.Pop<u64>()};
+    const auto data{ctx.ReadBuffer()};
+    LOG_WARNING(Service_NFP, "(STUBBED) called, device_handle={}, data_size={}", device_handle,
+                data.size());
+
+    // TODO(german77): Loop through all interfaces
+    if (device_handle == nfp_interface.GetHandle()) {
+        const auto result = nfp_interface.SetApplicationArea(data);
         IPC::ResponseBuilder rb{ctx, 2};
-        rb.Push(ResultSuccess);
+        rb.Push(result);
+        return;
     }
 
-    void GetTagInfo(Kernel::HLERequestContext& ctx) {
-        LOG_DEBUG(Service_NFP, "called");
+    LOG_ERROR(Service_NFP, "Handle not found, device_handle={}", device_handle);
 
+    IPC::ResponseBuilder rb{ctx, 2};
+    rb.Push(ErrCodes::DeviceNotFound);
+}
+
+void IUser::CreateApplicationArea(Kernel::HLERequestContext& ctx) {
+    IPC::RequestParser rp{ctx};
+    const auto device_handle{rp.Pop<u64>()};
+    const auto access_id{rp.Pop<u32>()};
+    const auto data{ctx.ReadBuffer()};
+    LOG_WARNING(Service_NFP, "(STUBBED) called, device_handle={}, data_size={}, access_id={}",
+                device_handle, access_id, data.size());
+
+    // TODO(german77): Loop through all interfaces
+    if (device_handle == nfp_interface.GetHandle()) {
+        const auto result = nfp_interface.CreateApplicationArea(access_id, data);
         IPC::ResponseBuilder rb{ctx, 2};
-        const auto& amiibo = nfp_interface.GetAmiiboBuffer();
-        const TagInfo tag_info{
-            .uuid = amiibo.uuid,
-            .uuid_length = static_cast<u8>(amiibo.uuid.size()),
-            .padding_1 = {},
-            .protocol = 1, // TODO(ogniK): Figure out actual values
-            .tag_type = 2,
-            .padding_2 = {},
-        };
-        ctx.WriteBuffer(tag_info);
-        rb.Push(ResultSuccess);
+        rb.Push(result);
+        return;
     }
 
-    void Mount(Kernel::HLERequestContext& ctx) {
-        LOG_DEBUG(Service_NFP, "called");
+    LOG_ERROR(Service_NFP, "Handle not found, device_handle={}", device_handle);
+
+    IPC::ResponseBuilder rb{ctx, 2};
+    rb.Push(ErrCodes::DeviceNotFound);
+}
+
+void IUser::GetTagInfo(Kernel::HLERequestContext& ctx) {
+    IPC::RequestParser rp{ctx};
+    const auto device_handle{rp.Pop<u64>()};
+    LOG_INFO(Service_NFP, "called, device_handle={}", device_handle);
 
-        device_state = DeviceState::TagNearby;
+    // TODO(german77): Loop through all interfaces
+    if (device_handle == nfp_interface.GetHandle()) {
+        TagInfo tag_info{};
+        const auto result = nfp_interface.GetTagInfo(tag_info);
+        ctx.WriteBuffer(tag_info);
         IPC::ResponseBuilder rb{ctx, 2};
-        rb.Push(ResultSuccess);
+        rb.Push(result);
+        return;
     }
 
-    void GetModelInfo(Kernel::HLERequestContext& ctx) {
-        LOG_DEBUG(Service_NFP, "called");
+    LOG_ERROR(Service_NFP, "Handle not found, device_handle={}", device_handle);
 
+    IPC::ResponseBuilder rb{ctx, 2};
+    rb.Push(ErrCodes::DeviceNotFound);
+}
+
+void IUser::GetRegisterInfo(Kernel::HLERequestContext& ctx) {
+    IPC::RequestParser rp{ctx};
+    const auto device_handle{rp.Pop<u64>()};
+    LOG_INFO(Service_NFP, "called, device_handle={}", device_handle);
+
+    // TODO(german77): Loop through all interfaces
+    if (device_handle == nfp_interface.GetHandle()) {
+        RegisterInfo register_info{};
+        const auto result = nfp_interface.GetRegisterInfo(register_info);
+        ctx.WriteBuffer(register_info);
         IPC::ResponseBuilder rb{ctx, 2};
-        const auto& amiibo = nfp_interface.GetAmiiboBuffer();
-        ctx.WriteBuffer(amiibo.model_info);
-        rb.Push(ResultSuccess);
+        rb.Push(result);
+        return;
     }
 
-    void Unmount(Kernel::HLERequestContext& ctx) {
-        LOG_DEBUG(Service_NFP, "called");
+    LOG_ERROR(Service_NFP, "Handle not found, device_handle={}", device_handle);
 
-        device_state = DeviceState::TagFound;
+    IPC::ResponseBuilder rb{ctx, 2};
+    rb.Push(ErrCodes::DeviceNotFound);
+}
 
+void IUser::GetCommonInfo(Kernel::HLERequestContext& ctx) {
+    IPC::RequestParser rp{ctx};
+    const auto device_handle{rp.Pop<u64>()};
+    LOG_INFO(Service_NFP, "called, device_handle={}", device_handle);
+
+    // TODO(german77): Loop through all interfaces
+    if (device_handle == nfp_interface.GetHandle()) {
+        CommonInfo common_info{};
+        const auto result = nfp_interface.GetCommonInfo(common_info);
+        ctx.WriteBuffer(common_info);
         IPC::ResponseBuilder rb{ctx, 2};
-        rb.Push(ResultSuccess);
+        rb.Push(result);
+        return;
     }
 
-    void Finalize(Kernel::HLERequestContext& ctx) {
-        LOG_DEBUG(Service_NFP, "called");
+    LOG_ERROR(Service_NFP, "Handle not found, device_handle={}", device_handle);
+
+    IPC::ResponseBuilder rb{ctx, 2};
+    rb.Push(ErrCodes::DeviceNotFound);
+}
 
-        device_state = DeviceState::Finalized;
+void IUser::GetModelInfo(Kernel::HLERequestContext& ctx) {
+    IPC::RequestParser rp{ctx};
+    const auto device_handle{rp.Pop<u64>()};
+    LOG_INFO(Service_NFP, "called, device_handle={}", device_handle);
 
+    // TODO(german77): Loop through all interfaces
+    if (device_handle == nfp_interface.GetHandle()) {
+        ModelInfo model_info{};
+        const auto result = nfp_interface.GetModelInfo(model_info);
+        ctx.WriteBuffer(model_info);
         IPC::ResponseBuilder rb{ctx, 2};
-        rb.Push(ResultSuccess);
+        rb.Push(result);
+        return;
     }
 
-    void AttachAvailabilityChangeEvent(Kernel::HLERequestContext& ctx) {
-        LOG_WARNING(Service_NFP, "(STUBBED) called");
+    LOG_ERROR(Service_NFP, "Handle not found, device_handle={}", device_handle);
+
+    IPC::ResponseBuilder rb{ctx, 2};
+    rb.Push(ErrCodes::DeviceNotFound);
+}
 
+void IUser::AttachActivateEvent(Kernel::HLERequestContext& ctx) {
+    IPC::RequestParser rp{ctx};
+    const auto device_handle{rp.Pop<u64>()};
+    LOG_DEBUG(Service_NFP, "called, device_handle={}", device_handle);
+
+    // TODO(german77): Loop through all interfaces
+    if (device_handle == nfp_interface.GetHandle()) {
         IPC::ResponseBuilder rb{ctx, 2, 1};
         rb.Push(ResultSuccess);
-        rb.PushCopyObjects(availability_change_event->GetReadableEvent());
+        rb.PushCopyObjects(nfp_interface.GetActivateEvent());
+        return;
     }
 
-    void GetRegisterInfo(Kernel::HLERequestContext& ctx) {
-        LOG_WARNING(Service_NFP, "(STUBBED) called");
+    LOG_ERROR(Service_NFP, "Handle not found, device_handle={}", device_handle);
 
-        // TODO(ogniK): Pull Mii and owner data from amiibo
+    IPC::ResponseBuilder rb{ctx, 2};
+    rb.Push(ErrCodes::DeviceNotFound);
+}
 
-        IPC::ResponseBuilder rb{ctx, 2};
+void IUser::AttachDeactivateEvent(Kernel::HLERequestContext& ctx) {
+    IPC::RequestParser rp{ctx};
+    const auto device_handle{rp.Pop<u64>()};
+    LOG_DEBUG(Service_NFP, "called, device_handle={}", device_handle);
+
+    // TODO(german77): Loop through all interfaces
+    if (device_handle == nfp_interface.GetHandle()) {
+        IPC::ResponseBuilder rb{ctx, 2, 1};
         rb.Push(ResultSuccess);
+        rb.PushCopyObjects(nfp_interface.GetDeactivateEvent());
+        return;
     }
 
-    void GetCommonInfo(Kernel::HLERequestContext& ctx) {
-        LOG_WARNING(Service_NFP, "(STUBBED) called");
+    LOG_ERROR(Service_NFP, "Handle not found, device_handle={}", device_handle);
 
-        // TODO(ogniK): Pull common information from amiibo
+    IPC::ResponseBuilder rb{ctx, 2};
+    rb.Push(ErrCodes::DeviceNotFound);
+}
 
-        CommonInfo common_info{};
-        common_info.application_area_size = 0;
-        ctx.WriteBuffer(common_info);
+void IUser::GetState(Kernel::HLERequestContext& ctx) {
+    LOG_DEBUG(Service_NFC, "called");
 
-        IPC::ResponseBuilder rb{ctx, 2};
+    IPC::ResponseBuilder rb{ctx, 3, 0};
+    rb.Push(ResultSuccess);
+    rb.PushEnum(state);
+}
+
+void IUser::GetDeviceState(Kernel::HLERequestContext& ctx) {
+    IPC::RequestParser rp{ctx};
+    const auto device_handle{rp.Pop<u64>()};
+    LOG_DEBUG(Service_NFP, "called, device_handle={}", device_handle);
+
+    // TODO(german77): Loop through all interfaces
+    if (device_handle == nfp_interface.GetHandle()) {
+        IPC::ResponseBuilder rb{ctx, 3};
         rb.Push(ResultSuccess);
+        rb.PushEnum(nfp_interface.GetCurrentState());
+        return;
     }
 
-    void OpenApplicationArea(Kernel::HLERequestContext& ctx) {
-        LOG_WARNING(Service_NFP, "(STUBBED) called");
-        IPC::ResponseBuilder rb{ctx, 2};
-        rb.Push(ErrCodes::ERR_NO_APPLICATION_AREA);
-    }
+    LOG_ERROR(Service_NFP, "Handle not found, device_handle={}", device_handle);
 
-    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, 2};
+    rb.Push(ErrCodes::DeviceNotFound);
+}
+
+void IUser::GetNpadId(Kernel::HLERequestContext& ctx) {
+    IPC::RequestParser rp{ctx};
+    const auto device_handle{rp.Pop<u64>()};
+    LOG_DEBUG(Service_NFP, "called, device_handle={}", device_handle);
+
+    // TODO(german77): Loop through all interfaces
+    if (device_handle == nfp_interface.GetHandle()) {
         IPC::ResponseBuilder rb{ctx, 3};
         rb.Push(ResultSuccess);
-        rb.PushRaw<u32>(0); // This is from the GetCommonInfo stub
+        rb.PushEnum(nfp_interface.GetNpadId());
+        return;
     }
 
-    void GetApplicationArea(Kernel::HLERequestContext& ctx) {
-        LOG_WARNING(Service_NFP, "(STUBBED) called");
+    LOG_ERROR(Service_NFP, "Handle not found, device_handle={}", device_handle);
+
+    IPC::ResponseBuilder rb{ctx, 2};
+    rb.Push(ErrCodes::DeviceNotFound);
+}
 
-        // TODO(ogniK): Pull application area from amiibo
+void IUser::GetApplicationAreaSize(Kernel::HLERequestContext& ctx) {
+    IPC::RequestParser rp{ctx};
+    const auto device_handle{rp.Pop<u64>()};
+    LOG_DEBUG(Service_NFP, "called, device_handle={}", device_handle);
 
+    // TODO(german77): Loop through all interfaces
+    if (device_handle == nfp_interface.GetHandle()) {
         IPC::ResponseBuilder rb{ctx, 3};
         rb.Push(ResultSuccess);
-        rb.PushRaw<u32>(0); // This is from the GetCommonInfo stub
+        rb.Push(ApplicationAreaSize);
+        return;
     }
 
-    Module::Interface& nfp_interface;
-    KernelHelpers::ServiceContext& service_context;
+    LOG_ERROR(Service_NFP, "Handle not found, device_handle={}", device_handle);
+
+    IPC::ResponseBuilder rb{ctx, 2};
+    rb.Push(ErrCodes::DeviceNotFound);
+}
+
+void IUser::AttachAvailabilityChangeEvent(Kernel::HLERequestContext& ctx) {
+    LOG_DEBUG(Service_NFP, "(STUBBED) called");
 
-    bool has_attached_handle{};
-    const u64 device_handle{0}; // Npad device 1
-    const u32 npad_id{0};       // Player 1 controller
-    State state{State::NonInitialized};
-    DeviceState device_state{DeviceState::Initialized};
-    Kernel::KEvent* deactivate_event;
-    Kernel::KEvent* availability_change_event;
-};
+    IPC::ResponseBuilder rb{ctx, 2, 1};
+    rb.Push(ResultSuccess);
+    rb.PushCopyObjects(availability_change_event->GetReadableEvent());
+}
+
+Module::Interface::Interface(std::shared_ptr<Module> module_, Core::System& system_,
+                             const char* name)
+    : ServiceFramework{system_, name}, module{std::move(module_)},
+      npad_id{Core::HID::NpadIdType::Player1}, service_context{system_, service_name} {
+    activate_event = service_context.CreateEvent("IUser:NFPActivateEvent");
+    deactivate_event = service_context.CreateEvent("IUser:NFPDeactivateEvent");
+}
+
+Module::Interface::~Interface() = default;
 
 void Module::Interface::CreateUserInterface(Kernel::HLERequestContext& ctx) {
     LOG_DEBUG(Service_NFP, "called");
 
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(ResultSuccess);
-    rb.PushIpcInterface<IUser>(*this, system, service_context);
+    rb.PushIpcInterface<IUser>(*this, system);
 }
 
 bool Module::Interface::LoadAmiibo(const std::vector<u8>& buffer) {
-    if (buffer.size() < sizeof(AmiiboFile)) {
+    if (device_state != DeviceState::SearchingForTag) {
+        LOG_ERROR(Service_NFP, "Game is not looking for amiibos, current state {}", device_state);
+        return false;
+    }
+
+    constexpr auto tag_size = sizeof(NTAG215File);
+    constexpr auto tag_size_without_password = sizeof(NTAG215File) - sizeof(NTAG215Password);
+
+    std::vector<u8> amiibo_buffer = buffer;
+
+    if (amiibo_buffer.size() < tag_size_without_password) {
+        LOG_ERROR(Service_NFP, "Wrong file size {}", buffer.size());
+        return false;
+    }
+
+    // Ensure it has the correct size
+    if (amiibo_buffer.size() != tag_size) {
+        amiibo_buffer.resize(tag_size, 0);
+    }
+
+    LOG_INFO(Service_NFP, "Amiibo detected");
+    std::memcpy(&tag_data, buffer.data(), tag_size);
+
+    if (!IsAmiiboValid()) {
         return false;
     }
 
-    std::memcpy(&amiibo, buffer.data(), sizeof(amiibo));
-    nfc_tag_load->GetWritableEvent().Signal();
+    // This value can't be dumped from a tag. Generate it
+    tag_data.password.PWD = GetTagPassword(tag_data.uuid);
+
+    device_state = DeviceState::TagFound;
+    activate_event->GetWritableEvent().Signal();
     return true;
 }
 
-Kernel::KReadableEvent& Module::Interface::GetNFCEvent() {
-    return nfc_tag_load->GetReadableEvent();
+void Module::Interface::CloseAmiibo() {
+    LOG_INFO(Service_NFP, "Remove amiibo");
+    device_state = DeviceState::TagRemoved;
+    is_application_area_initialized = false;
+    application_area_id = 0;
+    application_area_data.clear();
+    deactivate_event->GetWritableEvent().Signal();
+}
+
+bool Module::Interface::IsAmiiboValid() const {
+    const auto& amiibo_data = tag_data.user_memory;
+    LOG_DEBUG(Service_NFP, "uuid_lock=0x{0:x}", tag_data.lock_bytes);
+    LOG_DEBUG(Service_NFP, "compability_container=0x{0:x}", tag_data.compability_container);
+    LOG_DEBUG(Service_NFP, "crypto_init=0x{0:x}", amiibo_data.crypto_init);
+    LOG_DEBUG(Service_NFP, "write_count={}", amiibo_data.write_count);
+
+    LOG_DEBUG(Service_NFP, "character_id=0x{0:x}", amiibo_data.model_info.character_id);
+    LOG_DEBUG(Service_NFP, "character_variant={}", amiibo_data.model_info.character_variant);
+    LOG_DEBUG(Service_NFP, "amiibo_type={}", amiibo_data.model_info.amiibo_type);
+    LOG_DEBUG(Service_NFP, "model_number=0x{0:x}", amiibo_data.model_info.model_number);
+    LOG_DEBUG(Service_NFP, "series={}", amiibo_data.model_info.series);
+    LOG_DEBUG(Service_NFP, "fixed_value=0x{0:x}", amiibo_data.model_info.fixed);
+
+    LOG_DEBUG(Service_NFP, "tag_dynamic_lock=0x{0:x}", tag_data.dynamic_lock);
+    LOG_DEBUG(Service_NFP, "tag_CFG0=0x{0:x}", tag_data.CFG0);
+    LOG_DEBUG(Service_NFP, "tag_CFG1=0x{0:x}", tag_data.CFG1);
+
+    // Check against all know constants on an amiibo binary
+    if (tag_data.lock_bytes != 0xE00F) {
+        return false;
+    }
+    if (tag_data.compability_container != 0xEEFF10F1U) {
+        return false;
+    }
+    if ((amiibo_data.crypto_init & 0xFF) != 0xA5) {
+        return false;
+    }
+    if (amiibo_data.model_info.fixed != 0x02) {
+        return false;
+    }
+    if ((tag_data.dynamic_lock & 0xFFFFFF) != 0x0F0001) {
+        return false;
+    }
+    if (tag_data.CFG0 != 0x04000000U) {
+        return false;
+    }
+    if (tag_data.CFG1 != 0x5F) {
+        return false;
+    }
+    return true;
+}
+
+Kernel::KReadableEvent& Module::Interface::GetActivateEvent() const {
+    return activate_event->GetReadableEvent();
+}
+
+Kernel::KReadableEvent& Module::Interface::GetDeactivateEvent() const {
+    return deactivate_event->GetReadableEvent();
+}
+
+void Module::Interface::Initialize() {
+    device_state = DeviceState::Initialized;
+}
+
+void Module::Interface::Finalize() {
+    device_state = DeviceState::Unaviable;
+    is_application_area_initialized = false;
+    application_area_id = 0;
+    application_area_data.clear();
+}
+
+ResultCode Module::Interface::StartDetection(s32 protocol_) {
+    auto npad_device = system.HIDCore().GetEmulatedController(npad_id);
+
+    // TODO(german77): Add callback for when nfc data is available
+
+    if (device_state == DeviceState::Initialized || device_state == DeviceState::TagRemoved) {
+        npad_device->SetPollingMode(Common::Input::PollingMode::NFC);
+        device_state = DeviceState::SearchingForTag;
+        protocol = protocol_;
+        return ResultSuccess;
+    }
+
+    LOG_ERROR(Service_NFP, "Wrong device state {}", device_state);
+    return ErrCodes::WrongDeviceState;
+}
+
+ResultCode Module::Interface::StopDetection() {
+    auto npad_device = system.HIDCore().GetEmulatedController(npad_id);
+    npad_device->SetPollingMode(Common::Input::PollingMode::Active);
+
+    if (device_state == DeviceState::TagFound || device_state == DeviceState::TagMounted) {
+        CloseAmiibo();
+        return ResultSuccess;
+    }
+    if (device_state == DeviceState::SearchingForTag || device_state == DeviceState::TagRemoved) {
+        device_state = DeviceState::Initialized;
+        return ResultSuccess;
+    }
+
+    LOG_ERROR(Service_NFP, "Wrong device state {}", device_state);
+    return ErrCodes::WrongDeviceState;
+}
+
+ResultCode Module::Interface::Mount() {
+    if (device_state == DeviceState::TagFound) {
+        device_state = DeviceState::TagMounted;
+        return ResultSuccess;
+    }
+
+    LOG_ERROR(Service_NFP, "Wrong device state {}", device_state);
+    return ErrCodes::WrongDeviceState;
+}
+
+ResultCode Module::Interface::Unmount() {
+    if (device_state == DeviceState::TagMounted) {
+        is_application_area_initialized = false;
+        application_area_id = 0;
+        application_area_data.clear();
+        device_state = DeviceState::TagFound;
+        return ResultSuccess;
+    }
+
+    LOG_ERROR(Service_NFP, "Wrong device state {}", device_state);
+    return ErrCodes::WrongDeviceState;
+}
+
+ResultCode Module::Interface::GetTagInfo(TagInfo& tag_info) const {
+    if (device_state == DeviceState::TagFound || device_state == DeviceState::TagMounted) {
+        tag_info = {
+            .uuid = tag_data.uuid,
+            .uuid_length = static_cast<u8>(tag_data.uuid.size()),
+            .protocol = protocol,
+            .tag_type = static_cast<u32>(tag_data.user_memory.model_info.amiibo_type),
+        };
+        return ResultSuccess;
+    }
+
+    LOG_ERROR(Service_NFP, "Wrong device state {}", device_state);
+    return ErrCodes::WrongDeviceState;
+}
+
+ResultCode Module::Interface::GetCommonInfo(CommonInfo& common_info) const {
+    if (device_state != DeviceState::TagMounted) {
+        LOG_ERROR(Service_NFP, "Wrong device state {}", device_state);
+        return ErrCodes::WrongDeviceState;
+    }
+
+    // Read this data from the amiibo save file
+    common_info = {
+        .last_write_year = 2022,
+        .last_write_month = 2,
+        .last_write_day = 7,
+        .write_counter = tag_data.user_memory.write_count,
+        .version = 1,
+        .application_area_size = ApplicationAreaSize,
+    };
+    return ResultSuccess;
+}
+
+ResultCode Module::Interface::GetModelInfo(ModelInfo& model_info) const {
+    if (device_state != DeviceState::TagMounted) {
+        LOG_ERROR(Service_NFP, "Wrong device state {}", device_state);
+        return ErrCodes::WrongDeviceState;
+    }
+
+    model_info = tag_data.user_memory.model_info;
+    return ResultSuccess;
+}
+
+ResultCode Module::Interface::GetRegisterInfo(RegisterInfo& register_info) const {
+    if (device_state != DeviceState::TagMounted) {
+        LOG_ERROR(Service_NFP, "Wrong device state {}", device_state);
+        return ErrCodes::WrongDeviceState;
+    }
+
+    Service::Mii::MiiManager manager;
+
+    // Read this data from the amiibo save file
+    register_info = {
+        .mii_char_info = manager.BuildDefault(0),
+        .first_write_year = 2022,
+        .first_write_month = 2,
+        .first_write_day = 7,
+        .amiibo_name = {'Y', 'u', 'z', 'u', 'A', 'm', 'i', 'i', 'b', 'o', 0},
+        .unknown = {},
+    };
+    return ResultSuccess;
+}
+
+ResultCode Module::Interface::OpenApplicationArea(u32 access_id) {
+    if (device_state != DeviceState::TagMounted) {
+        LOG_ERROR(Service_NFP, "Wrong device state {}", device_state);
+        return ErrCodes::WrongDeviceState;
+    }
+    if (AmiiboApplicationDataExist(access_id)) {
+        application_area_data = LoadAmiiboApplicationData(access_id);
+        application_area_id = access_id;
+        is_application_area_initialized = true;
+    }
+    if (!is_application_area_initialized) {
+        LOG_WARNING(Service_NFP, "Application area is not initialized");
+        return ErrCodes::ApplicationAreaIsNotInitialized;
+    }
+    return ResultSuccess;
+}
+
+ResultCode Module::Interface::GetApplicationArea(std::vector<u8>& data) const {
+    if (device_state != DeviceState::TagMounted) {
+        LOG_ERROR(Service_NFP, "Wrong device state {}", device_state);
+        return ErrCodes::WrongDeviceState;
+    }
+    if (!is_application_area_initialized) {
+        LOG_ERROR(Service_NFP, "Application area is not initialized");
+        return ErrCodes::ApplicationAreaIsNotInitialized;
+    }
+
+    data = application_area_data;
+
+    return ResultSuccess;
+}
+
+ResultCode Module::Interface::SetApplicationArea(const std::vector<u8>& data) {
+    if (device_state != DeviceState::TagMounted) {
+        LOG_ERROR(Service_NFP, "Wrong device state {}", device_state);
+        return ErrCodes::WrongDeviceState;
+    }
+    if (!is_application_area_initialized) {
+        LOG_ERROR(Service_NFP, "Application area is not initialized");
+        return ErrCodes::ApplicationAreaIsNotInitialized;
+    }
+    application_area_data = data;
+    SaveAmiiboApplicationData(application_area_id, application_area_data);
+    return ResultSuccess;
+}
+
+ResultCode Module::Interface::CreateApplicationArea(u32 access_id, const std::vector<u8>& data) {
+    if (device_state != DeviceState::TagMounted) {
+        LOG_ERROR(Service_NFP, "Wrong device state {}", device_state);
+        return ErrCodes::WrongDeviceState;
+    }
+    if (AmiiboApplicationDataExist(access_id)) {
+        LOG_ERROR(Service_NFP, "Application area already exist");
+        return ErrCodes::ApplicationAreaExist;
+    }
+    application_area_data = data;
+    application_area_id = access_id;
+    SaveAmiiboApplicationData(application_area_id, application_area_data);
+    return ResultSuccess;
+}
+
+bool Module::Interface::AmiiboApplicationDataExist(u32 access_id) const {
+    // TODO(german77): Check if file exist
+    return false;
+}
+
+std::vector<u8> Module::Interface::LoadAmiiboApplicationData(u32 access_id) const {
+    // TODO(german77): Read file
+    std::vector<u8> data(ApplicationAreaSize);
+    return data;
+}
+
+void Module::Interface::SaveAmiiboApplicationData(u32 access_id,
+                                                  const std::vector<u8>& data) const {
+    // TODO(german77): Save file
+}
+
+u64 Module::Interface::GetHandle() const {
+    // Generate a handle based of the npad id
+    return static_cast<u64>(npad_id);
+}
+
+DeviceState Module::Interface::GetCurrentState() const {
+    return device_state;
+}
+
+Core::HID::NpadIdType Module::Interface::GetNpadId() const {
+    return npad_id;
 }
 
-const Module::Interface::AmiiboFile& Module::Interface::GetAmiiboBuffer() const {
-    return amiibo;
+u32 Module::Interface::GetTagPassword(const TagUuid& uuid) const {
+    // Verifiy that the generated password is correct
+    u32 password = 0xAA ^ (uuid[1] ^ uuid[3]);
+    password &= (0x55 ^ (uuid[2] ^ uuid[4])) << 8;
+    password &= (0xAA ^ (uuid[3] ^ uuid[5])) << 16;
+    password &= (0x55 ^ (uuid[4] ^ uuid[6])) << 24;
+    return password;
 }
 
 void InstallInterfaces(SM::ServiceManager& service_manager, Core::System& system) {
diff --git a/src/core/hle/service/nfp/nfp.h b/src/core/hle/service/nfp/nfp.h
index 95c127efb..022f13b29 100644
--- a/src/core/hle/service/nfp/nfp.h
+++ b/src/core/hle/service/nfp/nfp.h
@@ -7,15 +7,132 @@
 #include <array>
 #include <vector>
 
+#include "common/common_funcs.h"
 #include "core/hle/service/kernel_helpers.h"
+#include "core/hle/service/mii/mii_manager.h"
 #include "core/hle/service/service.h"
 
 namespace Kernel {
 class KEvent;
-}
+class KReadableEvent;
+} // namespace Kernel
+
+namespace Core::HID {
+enum class NpadIdType : u32;
+} // namespace Core::HID
 
 namespace Service::NFP {
 
+enum class ServiceType : u32 {
+    User,
+    Debug,
+    System,
+};
+
+enum class State : u32 {
+    NonInitialized,
+    Initialized,
+};
+
+enum class DeviceState : u32 {
+    Initialized,
+    SearchingForTag,
+    TagFound,
+    TagRemoved,
+    TagMounted,
+    Unaviable,
+    Finalized,
+};
+
+enum class ModelType : u32 {
+    Amiibo,
+};
+
+enum class MountTarget : u32 {
+    Rom,
+    Ram,
+    All,
+};
+
+enum class AmiiboType : u8 {
+    Figure,
+    Card,
+    Yarn,
+};
+
+enum class AmiiboSeries : u8 {
+    SuperSmashBros,
+    SuperMario,
+    ChibiRobo,
+    YoshiWoollyWorld,
+    Splatoon,
+    AnimalCrossing,
+    EightBitMario,
+    Skylanders,
+    Unknown8,
+    TheLegendOfZelda,
+    ShovelKnight,
+    Unknown11,
+    Kiby,
+    Pokemon,
+    MarioSportsSuperstars,
+    MonsterHunter,
+    BoxBoy,
+    Pikmin,
+    FireEmblem,
+    Metroid,
+    Others,
+    MegaMan,
+    Diablo
+};
+
+using TagUuid = std::array<u8, 10>;
+
+struct TagInfo {
+    TagUuid uuid;
+    u8 uuid_length;
+    INSERT_PADDING_BYTES(0x15);
+    s32 protocol;
+    u32 tag_type;
+    INSERT_PADDING_BYTES(0x30);
+};
+static_assert(sizeof(TagInfo) == 0x58, "TagInfo is an invalid size");
+
+struct CommonInfo {
+    u16 last_write_year;
+    u8 last_write_month;
+    u8 last_write_day;
+    u16 write_counter;
+    u16 version;
+    u32 application_area_size;
+    INSERT_PADDING_BYTES(0x34);
+};
+static_assert(sizeof(CommonInfo) == 0x40, "CommonInfo is an invalid size");
+
+struct ModelInfo {
+    u16 character_id;
+    u8 character_variant;
+    AmiiboType amiibo_type;
+    u16 model_number;
+    AmiiboSeries series;
+    u8 fixed;                   // Must be 02
+    INSERT_PADDING_BYTES(0x4);  // Unknown
+    INSERT_PADDING_BYTES(0x20); // Probably a SHA256-(HMAC?) hash
+    INSERT_PADDING_BYTES(0x14); // SHA256-HMAC
+};
+static_assert(sizeof(ModelInfo) == 0x40, "ModelInfo is an invalid size");
+
+struct RegisterInfo {
+    Service::Mii::MiiInfo mii_char_info;
+    u16 first_write_year;
+    u8 first_write_month;
+    u8 first_write_day;
+    std::array<u8, 11> amiibo_name;
+    u8 unknown;
+    INSERT_PADDING_BYTES(0x98);
+};
+static_assert(sizeof(RegisterInfo) == 0x100, "RegisterInfo is an invalid size");
+
 class Module final {
 public:
     class Interface : public ServiceFramework<Interface> {
@@ -24,34 +141,131 @@ public:
                            const char* name);
         ~Interface() override;
 
-        struct ModelInfo {
-            std::array<u8, 0x8> amiibo_identification_block;
-            INSERT_PADDING_BYTES(0x38);
+        struct EncryptedAmiiboFile {
+            u16 crypto_init;             // Must be A5 XX
+            u16 write_count;             // Number of times the amiibo has been written?
+            INSERT_PADDING_BYTES(0x20);  // System crypts
+            INSERT_PADDING_BYTES(0x20);  // SHA256-(HMAC?) hash
+            ModelInfo model_info;        // This struct is bigger than documentation
+            INSERT_PADDING_BYTES(0xC);   // SHA256-HMAC
+            INSERT_PADDING_BYTES(0x114); // section 1 encrypted buffer
+            INSERT_PADDING_BYTES(0x54);  // section 2 encrypted buffer
+        };
+        static_assert(sizeof(EncryptedAmiiboFile) == 0x1F8, "AmiiboFile is an invalid size");
+
+        struct NTAG215Password {
+            u32 PWD;  // Password to allow write access
+            u16 PACK; // Password acknowledge reply
+            u16 RFUI; // Reserved for future use
         };
-        static_assert(sizeof(ModelInfo) == 0x40, "ModelInfo is an invalid size");
+        static_assert(sizeof(NTAG215Password) == 0x8, "NTAG215Password is an invalid size");
 
-        struct AmiiboFile {
-            std::array<u8, 10> uuid;
-            INSERT_PADDING_BYTES(0x4a);
-            ModelInfo model_info;
+        struct NTAG215File {
+            TagUuid uuid;                    // Unique serial number
+            u16 lock_bytes;                  // Set defined pages as read only
+            u32 compability_container;       // Defines available memory
+            EncryptedAmiiboFile user_memory; // Writable data
+            u32 dynamic_lock;                // Dynamic lock
+            u32 CFG0;                        // Defines memory protected by password
+            u32 CFG1;                        // Defines number of verification attempts
+            NTAG215Password password;        // Password data
         };
-        static_assert(sizeof(AmiiboFile) == 0x94, "AmiiboFile is an invalid size");
+        static_assert(sizeof(NTAG215File) == 0x21C, "NTAG215File is an invalid size");
 
         void CreateUserInterface(Kernel::HLERequestContext& ctx);
         bool LoadAmiibo(const std::vector<u8>& buffer);
-        Kernel::KReadableEvent& GetNFCEvent();
-        const AmiiboFile& GetAmiiboBuffer() const;
+        void CloseAmiibo();
+
+        void Initialize();
+        void Finalize();
+
+        ResultCode StartDetection(s32 protocol_);
+        ResultCode StopDetection();
+        ResultCode Mount();
+        ResultCode Unmount();
+
+        ResultCode GetTagInfo(TagInfo& tag_info) const;
+        ResultCode GetCommonInfo(CommonInfo& common_info) const;
+        ResultCode GetModelInfo(ModelInfo& model_info) const;
+        ResultCode GetRegisterInfo(RegisterInfo& register_info) const;
+
+        ResultCode OpenApplicationArea(u32 access_id);
+        ResultCode GetApplicationArea(std::vector<u8>& data) const;
+        ResultCode SetApplicationArea(const std::vector<u8>& data);
+        ResultCode CreateApplicationArea(u32 access_id, const std::vector<u8>& data);
+
+        u64 GetHandle() const;
+        DeviceState GetCurrentState() const;
+        Core::HID::NpadIdType GetNpadId() const;
+
+        Kernel::KReadableEvent& GetActivateEvent() const;
+        Kernel::KReadableEvent& GetDeactivateEvent() const;
 
     protected:
         std::shared_ptr<Module> module;
 
     private:
+        /// Validates that the amiibo file is not corrupted
+        bool IsAmiiboValid() const;
+
+        bool AmiiboApplicationDataExist(u32 access_id) const;
+        std::vector<u8> LoadAmiiboApplicationData(u32 access_id) const;
+        void SaveAmiiboApplicationData(u32 access_id, const std::vector<u8>& data) const;
+
+        /// return password needed to allow write access to protected memory
+        u32 GetTagPassword(const TagUuid& uuid) const;
+
+        const Core::HID::NpadIdType npad_id;
+
+        DeviceState device_state{DeviceState::Unaviable};
         KernelHelpers::ServiceContext service_context;
-        Kernel::KEvent* nfc_tag_load;
-        AmiiboFile amiibo{};
+        Kernel::KEvent* activate_event;
+        Kernel::KEvent* deactivate_event;
+        NTAG215File tag_data{};
+        s32 protocol;
+        bool is_application_area_initialized{};
+        u32 application_area_id;
+        std::vector<u8> application_area_data;
     };
 };
 
+class IUser final : public ServiceFramework<IUser> {
+public:
+    explicit IUser(Module::Interface& nfp_interface_, Core::System& system_);
+
+private:
+    void Initialize(Kernel::HLERequestContext& ctx);
+    void Finalize(Kernel::HLERequestContext& ctx);
+    void ListDevices(Kernel::HLERequestContext& ctx);
+    void StartDetection(Kernel::HLERequestContext& ctx);
+    void StopDetection(Kernel::HLERequestContext& ctx);
+    void Mount(Kernel::HLERequestContext& ctx);
+    void Unmount(Kernel::HLERequestContext& ctx);
+    void OpenApplicationArea(Kernel::HLERequestContext& ctx);
+    void GetApplicationArea(Kernel::HLERequestContext& ctx);
+    void SetApplicationArea(Kernel::HLERequestContext& ctx);
+    void CreateApplicationArea(Kernel::HLERequestContext& ctx);
+    void GetTagInfo(Kernel::HLERequestContext& ctx);
+    void GetRegisterInfo(Kernel::HLERequestContext& ctx);
+    void GetCommonInfo(Kernel::HLERequestContext& ctx);
+    void GetModelInfo(Kernel::HLERequestContext& ctx);
+    void AttachActivateEvent(Kernel::HLERequestContext& ctx);
+    void AttachDeactivateEvent(Kernel::HLERequestContext& ctx);
+    void GetState(Kernel::HLERequestContext& ctx);
+    void GetDeviceState(Kernel::HLERequestContext& ctx);
+    void GetNpadId(Kernel::HLERequestContext& ctx);
+    void GetApplicationAreaSize(Kernel::HLERequestContext& ctx);
+    void AttachAvailabilityChangeEvent(Kernel::HLERequestContext& ctx);
+
+    KernelHelpers::ServiceContext service_context;
+
+    // TODO(german77): We should have a vector of interfaces
+    Module::Interface& nfp_interface;
+
+    State state{State::NonInitialized};
+    Kernel::KEvent* availability_change_event;
+};
+
 void InstallInterfaces(SM::ServiceManager& service_manager, Core::System& system);
 
 } // namespace Service::NFP
diff --git a/src/core/hle/service/ns/pdm_qry.cpp b/src/core/hle/service/ns/pdm_qry.cpp
index e2fab5c3f..36ce46353 100644
--- a/src/core/hle/service/ns/pdm_qry.cpp
+++ b/src/core/hle/service/ns/pdm_qry.cpp
@@ -59,7 +59,7 @@ void PDM_QRY::QueryPlayStatisticsByApplicationIdAndUserAccountId(Kernel::HLERequ
 
     LOG_WARNING(Service_NS,
                 "(STUBBED) called. unknown={}. application_id=0x{:016X}, user_account_uid=0x{}",
-                unknown, application_id, user_account_uid.Format());
+                unknown, application_id, user_account_uid.RawString());
 
     IPC::ResponseBuilder rb{ctx, 12};
     rb.Push(ResultSuccess);
diff --git a/src/core/hle/service/pm/pm.cpp b/src/core/hle/service/pm/pm.cpp
index 277abc17a..057666021 100644
--- a/src/core/hle/service/pm/pm.cpp
+++ b/src/core/hle/service/pm/pm.cpp
@@ -91,6 +91,8 @@ public:
             {4, &DebugMonitor::GetApplicationProcessId, "GetApplicationProcessId"},
             {5, nullptr, "HookToCreateApplicationProgress"},
             {6, nullptr, "ClearHook"},
+            {65000, &DebugMonitor::AtmosphereGetProcessInfo, "AtmosphereGetProcessInfo"},
+            {65001, nullptr, "AtmosphereGetCurrentLimitInfo"},
         };
         // clang-format on
 
@@ -125,6 +127,49 @@ private:
         GetApplicationPidGeneric(ctx, kernel.GetProcessList());
     }
 
+    void AtmosphereGetProcessInfo(Kernel::HLERequestContext& ctx) {
+        // https://github.com/Atmosphere-NX/Atmosphere/blob/master/stratosphere/pm/source/impl/pm_process_manager.cpp#L614
+        // This implementation is incomplete; only a handle to the process is returned.
+        IPC::RequestParser rp{ctx};
+        const auto pid = rp.PopRaw<u64>();
+
+        LOG_WARNING(Service_PM, "(Partial Implementation) called, pid={:016X}", pid);
+
+        const auto process = SearchProcessList(kernel.GetProcessList(), [pid](const auto& proc) {
+            return proc->GetProcessID() == pid;
+        });
+
+        if (!process.has_value()) {
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ResultProcessNotFound);
+            return;
+        }
+
+        struct ProgramLocation {
+            u64 program_id;
+            u8 storage_id;
+        };
+        static_assert(sizeof(ProgramLocation) == 0x10, "ProgramLocation has an invalid size");
+
+        struct OverrideStatus {
+            u64 keys_held;
+            u64 flags;
+        };
+        static_assert(sizeof(OverrideStatus) == 0x10, "OverrideStatus has an invalid size");
+
+        OverrideStatus override_status{};
+        ProgramLocation program_location{
+            .program_id = (*process)->GetProgramID(),
+            .storage_id = 0,
+        };
+
+        IPC::ResponseBuilder rb{ctx, 10, 1};
+        rb.Push(ResultSuccess);
+        rb.PushCopyObjects(*process);
+        rb.PushRaw(program_location);
+        rb.PushRaw(override_status);
+    }
+
     const Kernel::KernelCore& kernel;
 };
 
diff --git a/src/core/hle/service/service.cpp b/src/core/hle/service/service.cpp
index f54e6fe56..eb1138313 100644
--- a/src/core/hle/service/service.cpp
+++ b/src/core/hle/service/service.cpp
@@ -39,6 +39,7 @@
 #include "core/hle/service/mig/mig.h"
 #include "core/hle/service/mii/mii.h"
 #include "core/hle/service/mm/mm_u.h"
+#include "core/hle/service/mnpp/mnpp_app.h"
 #include "core/hle/service/ncm/ncm.h"
 #include "core/hle/service/nfc/nfc.h"
 #include "core/hle/service/nfp/nfp.h"
@@ -265,6 +266,7 @@ Services::Services(std::shared_ptr<SM::ServiceManager>& sm, Core::System& system
     Migration::InstallInterfaces(*sm, system);
     Mii::InstallInterfaces(*sm, system);
     MM::InstallInterfaces(*sm, system);
+    MNPP::InstallInterfaces(*sm, system);
     NCM::InstallInterfaces(*sm, system);
     NFC::InstallInterfaces(*sm, system);
     NFP::InstallInterfaces(*sm, system);
diff --git a/src/core/hle/service/sm/sm.cpp b/src/core/hle/service/sm/sm.cpp
index eaa172595..695a1faa6 100644
--- a/src/core/hle/service/sm/sm.cpp
+++ b/src/core/hle/service/sm/sm.cpp
@@ -81,6 +81,8 @@ ResultVal<Kernel::KPort*> ServiceManager::GetServicePort(const std::string& name
     }
 
     auto* port = Kernel::KPort::Create(kernel);
+    SCOPE_EXIT({ port->Close(); });
+
     port->Initialize(ServerSessionCountMax, false, name);
     auto handler = it->second;
     port->GetServerPort().SetSessionHandler(std::move(handler));
diff --git a/src/core/hle/service/sockets/bsd.cpp b/src/core/hle/service/sockets/bsd.cpp
index f83272633..3dbac5a23 100644
--- a/src/core/hle/service/sockets/bsd.cpp
+++ b/src/core/hle/service/sockets/bsd.cpp
@@ -569,9 +569,9 @@ std::pair<s32, Errno> BSD::AcceptImpl(s32 fd, std::vector<u8>& write_buffer) {
     new_descriptor.socket = std::move(result.socket);
     new_descriptor.is_connection_based = descriptor.is_connection_based;
 
-    ASSERT(write_buffer.size() == sizeof(SockAddrIn));
     const SockAddrIn guest_addr_in = Translate(result.sockaddr_in);
-    std::memcpy(write_buffer.data(), &guest_addr_in, sizeof(guest_addr_in));
+    const size_t length = std::min(sizeof(guest_addr_in), write_buffer.size());
+    std::memcpy(write_buffer.data(), &guest_addr_in, length);
 
     return {new_fd, Errno::SUCCESS};
 }
diff --git a/src/core/hle/service/time/clock_types.h b/src/core/hle/service/time/clock_types.h
index 392e16863..d0cacb80c 100644
--- a/src/core/hle/service/time/clock_types.h
+++ b/src/core/hle/service/time/clock_types.h
@@ -36,7 +36,7 @@ struct SteadyClockTimePoint {
     }
 
     static SteadyClockTimePoint GetRandom() {
-        return {0, Common::UUID::Generate()};
+        return {0, Common::UUID::MakeRandom()};
     }
 };
 static_assert(sizeof(SteadyClockTimePoint) == 0x18, "SteadyClockTimePoint is incorrect size");
diff --git a/src/core/hle/service/time/steady_clock_core.h b/src/core/hle/service/time/steady_clock_core.h
index d80a2385f..5ee2c0e0a 100644
--- a/src/core/hle/service/time/steady_clock_core.h
+++ b/src/core/hle/service/time/steady_clock_core.h
@@ -49,7 +49,7 @@ public:
     }
 
 private:
-    Common::UUID clock_source_id{Common::UUID::Generate()};
+    Common::UUID clock_source_id{Common::UUID::MakeRandom()};
     bool is_initialized{};
 };
 
diff --git a/src/core/hle/service/time/time_manager.cpp b/src/core/hle/service/time/time_manager.cpp
index c1e4e6cce..00f1ae8cf 100644
--- a/src/core/hle/service/time/time_manager.cpp
+++ b/src/core/hle/service/time/time_manager.cpp
@@ -45,7 +45,7 @@ struct TimeManager::Impl final {
           time_zone_content_manager{system} {
 
         const auto system_time{Clock::TimeSpanType::FromSeconds(GetExternalRtcValue())};
-        SetupStandardSteadyClock(system, Common::UUID::Generate(), system_time, {}, {});
+        SetupStandardSteadyClock(system, Common::UUID::MakeRandom(), system_time, {}, {});
         SetupStandardLocalSystemClock(system, {}, system_time.ToSeconds());
 
         Clock::SystemClockContext clock_context{};
diff --git a/src/core/hle/service/vi/display/vi_display.h b/src/core/hle/service/vi/display/vi_display.h
index 0979fc421..329f4ba86 100644
--- a/src/core/hle/service/vi/display/vi_display.h
+++ b/src/core/hle/service/vi/display/vi_display.h
@@ -28,10 +28,10 @@ class Layer;
 
 /// Represents a single display type
 class Display {
+public:
     YUZU_NON_COPYABLE(Display);
     YUZU_NON_MOVEABLE(Display);
 
-public:
     /// Constructs a display with a given unique ID and name.
     ///
     /// @param id The unique ID for this display.
diff --git a/src/core/loader/loader.h b/src/core/loader/loader.h
index 7b1bac3f7..8b6b3b68f 100644
--- a/src/core/loader/loader.h
+++ b/src/core/loader/loader.h
@@ -11,6 +11,7 @@
 #include <utility>
 #include <vector>
 
+#include "common/common_funcs.h"
 #include "common/common_types.h"
 #include "core/file_sys/control_metadata.h"
 #include "core/file_sys/vfs.h"
@@ -139,8 +140,11 @@ std::string GetResultStatusString(ResultStatus status);
 std::ostream& operator<<(std::ostream& os, ResultStatus status);
 
 /// Interface for loading an application
-class AppLoader : NonCopyable {
+class AppLoader {
 public:
+    YUZU_NON_COPYABLE(AppLoader);
+    YUZU_NON_MOVEABLE(AppLoader);
+
     struct LoadParameters {
         s32 main_thread_priority;
         u64 main_thread_stack_size;
diff --git a/src/core/memory.cpp b/src/core/memory.cpp
index 88d6ec908..28d30eee2 100644
--- a/src/core/memory.cpp
+++ b/src/core/memory.cpp
@@ -39,8 +39,7 @@ struct Memory::Impl {
     void MapMemoryRegion(Common::PageTable& page_table, VAddr base, u64 size, PAddr target) {
         ASSERT_MSG((size & PAGE_MASK) == 0, "non-page aligned size: {:016X}", size);
         ASSERT_MSG((base & PAGE_MASK) == 0, "non-page aligned base: {:016X}", base);
-        ASSERT_MSG(target >= DramMemoryMap::Base && target < DramMemoryMap::End,
-                   "Out of bounds target: {:016X}", target);
+        ASSERT_MSG(target >= DramMemoryMap::Base, "Out of bounds target: {:016X}", target);
         MapPages(page_table, base / PAGE_SIZE, size / PAGE_SIZE, target, Common::PageType::Memory);
 
         if (Settings::IsFastmemEnabled()) {
diff --git a/src/input_common/drivers/gc_adapter.cpp b/src/input_common/drivers/gc_adapter.cpp
index 7ab4540a8..155caae42 100644
--- a/src/input_common/drivers/gc_adapter.cpp
+++ b/src/input_common/drivers/gc_adapter.cpp
@@ -248,7 +248,7 @@ bool GCAdapter::Setup() {
         std::size_t port = 0;
         for (GCController& pad : pads) {
             pad.identifier = {
-                .guid = Common::UUID{Common::INVALID_UUID},
+                .guid = Common::UUID{},
                 .port = port++,
                 .pad = 0,
             };
diff --git a/src/input_common/drivers/keyboard.cpp b/src/input_common/drivers/keyboard.cpp
index 4c1e5bbec..59e3d9cc0 100644
--- a/src/input_common/drivers/keyboard.cpp
+++ b/src/input_common/drivers/keyboard.cpp
@@ -9,17 +9,17 @@
 namespace InputCommon {
 
 constexpr PadIdentifier key_identifier = {
-    .guid = Common::UUID{Common::INVALID_UUID},
+    .guid = Common::UUID{},
     .port = 0,
     .pad = 0,
 };
 constexpr PadIdentifier keyboard_key_identifier = {
-    .guid = Common::UUID{Common::INVALID_UUID},
+    .guid = Common::UUID{},
     .port = 1,
     .pad = 0,
 };
 constexpr PadIdentifier keyboard_modifier_identifier = {
-    .guid = Common::UUID{Common::INVALID_UUID},
+    .guid = Common::UUID{},
     .port = 1,
     .pad = 1,
 };
diff --git a/src/input_common/drivers/mouse.cpp b/src/input_common/drivers/mouse.cpp
index d8ae7f0c1..3c9a4e747 100644
--- a/src/input_common/drivers/mouse.cpp
+++ b/src/input_common/drivers/mouse.cpp
@@ -20,7 +20,7 @@ constexpr int motion_wheel_y = 4;
 constexpr int touch_axis_x = 10;
 constexpr int touch_axis_y = 11;
 constexpr PadIdentifier identifier = {
-    .guid = Common::UUID{Common::INVALID_UUID},
+    .guid = Common::UUID{},
     .port = 0,
     .pad = 0,
 };
diff --git a/src/input_common/drivers/sdl_driver.cpp b/src/input_common/drivers/sdl_driver.cpp
index 577bf5c31..c17ea305e 100644
--- a/src/input_common/drivers/sdl_driver.cpp
+++ b/src/input_common/drivers/sdl_driver.cpp
@@ -175,23 +175,23 @@ public:
         return false;
     }
 
-    BatteryLevel GetBatteryLevel() {
+    Common::Input::BatteryLevel GetBatteryLevel() {
         const auto level = SDL_JoystickCurrentPowerLevel(sdl_joystick.get());
         switch (level) {
         case SDL_JOYSTICK_POWER_EMPTY:
-            return BatteryLevel::Empty;
+            return Common::Input::BatteryLevel::Empty;
         case SDL_JOYSTICK_POWER_LOW:
-            return BatteryLevel::Critical;
+            return Common::Input::BatteryLevel::Low;
         case SDL_JOYSTICK_POWER_MEDIUM:
-            return BatteryLevel::Low;
+            return Common::Input::BatteryLevel::Medium;
         case SDL_JOYSTICK_POWER_FULL:
-            return BatteryLevel::Medium;
         case SDL_JOYSTICK_POWER_MAX:
-            return BatteryLevel::Full;
-        case SDL_JOYSTICK_POWER_UNKNOWN:
+            return Common::Input::BatteryLevel::Full;
         case SDL_JOYSTICK_POWER_WIRED:
+            return Common::Input::BatteryLevel::Charging;
+        case SDL_JOYSTICK_POWER_UNKNOWN:
         default:
-            return BatteryLevel::Charging;
+            return Common::Input::BatteryLevel::None;
         }
     }
 
@@ -352,6 +352,8 @@ void SDLDriver::HandleGameControllerEvent(const SDL_Event& event) {
         if (const auto joystick = GetSDLJoystickBySDLID(event.jbutton.which)) {
             const PadIdentifier identifier = joystick->GetPadIdentifier();
             SetButton(identifier, event.jbutton.button, true);
+            // Battery doesn't trigger an event so just update every button press
+            SetBattery(identifier, joystick->GetBatteryLevel());
         }
         break;
     }
@@ -503,7 +505,7 @@ std::vector<Common::ParamPackage> SDLDriver::GetInputDevices() const {
 Common::Input::VibrationError SDLDriver::SetRumble(
     const PadIdentifier& identifier, const Common::Input::VibrationStatus& vibration) {
     const auto joystick =
-        GetSDLJoystickByGUID(identifier.guid.Format(), static_cast<int>(identifier.port));
+        GetSDLJoystickByGUID(identifier.guid.RawString(), static_cast<int>(identifier.port));
     const auto process_amplitude_exp = [](f32 amplitude, f32 factor) {
         return (amplitude + std::pow(amplitude, factor)) * 0.5f * 0xFFFF;
     };
@@ -600,7 +602,7 @@ Common::ParamPackage SDLDriver::BuildParamPackageForAnalog(PadIdentifier identif
     Common::ParamPackage params;
     params.Set("engine", GetEngineName());
     params.Set("port", static_cast<int>(identifier.port));
-    params.Set("guid", identifier.guid.Format());
+    params.Set("guid", identifier.guid.RawString());
     params.Set("axis_x", axis_x);
     params.Set("axis_y", axis_y);
     params.Set("offset_x", offset_x);
@@ -812,7 +814,7 @@ AnalogMapping SDLDriver::GetAnalogMappingForDevice(const Common::ParamPackage& p
         PreSetAxis(identifier, binding_left_x.value.axis);
         PreSetAxis(identifier, binding_left_y.value.axis);
         const auto left_offset_x = -GetAxis(identifier, binding_left_x.value.axis);
-        const auto left_offset_y = -GetAxis(identifier, binding_left_y.value.axis);
+        const auto left_offset_y = GetAxis(identifier, binding_left_y.value.axis);
         mapping.insert_or_assign(Settings::NativeAnalog::LStick,
                                  BuildParamPackageForAnalog(identifier, binding_left_x.value.axis,
                                                             binding_left_y.value.axis,
@@ -823,7 +825,7 @@ AnalogMapping SDLDriver::GetAnalogMappingForDevice(const Common::ParamPackage& p
         PreSetAxis(identifier, binding_left_x.value.axis);
         PreSetAxis(identifier, binding_left_y.value.axis);
         const auto left_offset_x = -GetAxis(identifier, binding_left_x.value.axis);
-        const auto left_offset_y = -GetAxis(identifier, binding_left_y.value.axis);
+        const auto left_offset_y = GetAxis(identifier, binding_left_y.value.axis);
         mapping.insert_or_assign(Settings::NativeAnalog::LStick,
                                  BuildParamPackageForAnalog(identifier, binding_left_x.value.axis,
                                                             binding_left_y.value.axis,
@@ -838,7 +840,7 @@ AnalogMapping SDLDriver::GetAnalogMappingForDevice(const Common::ParamPackage& p
     PreSetAxis(identifier, binding_right_x.value.axis);
     PreSetAxis(identifier, binding_right_y.value.axis);
     const auto right_offset_x = -GetAxis(identifier, binding_right_x.value.axis);
-    const auto right_offset_y = -GetAxis(identifier, binding_right_y.value.axis);
+    const auto right_offset_y = GetAxis(identifier, binding_right_y.value.axis);
     mapping.insert_or_assign(Settings::NativeAnalog::RStick,
                              BuildParamPackageForAnalog(identifier, binding_right_x.value.axis,
                                                         binding_right_y.value.axis, right_offset_x,
diff --git a/src/input_common/drivers/touch_screen.cpp b/src/input_common/drivers/touch_screen.cpp
index 880781825..30c727df4 100644
--- a/src/input_common/drivers/touch_screen.cpp
+++ b/src/input_common/drivers/touch_screen.cpp
@@ -8,7 +8,7 @@
 namespace InputCommon {
 
 constexpr PadIdentifier identifier = {
-    .guid = Common::UUID{Common::INVALID_UUID},
+    .guid = Common::UUID{},
     .port = 0,
     .pad = 0,
 };
diff --git a/src/input_common/drivers/udp_client.cpp b/src/input_common/drivers/udp_client.cpp
index 333173e3d..9780ead10 100644
--- a/src/input_common/drivers/udp_client.cpp
+++ b/src/input_common/drivers/udp_client.cpp
@@ -192,22 +192,22 @@ std::size_t UDPClient::GetClientNumber(std::string_view host, u16 port) const {
     return MAX_UDP_CLIENTS;
 }
 
-BatteryLevel UDPClient::GetBatteryLevel(Response::Battery battery) const {
+Common::Input::BatteryLevel UDPClient::GetBatteryLevel(Response::Battery battery) const {
     switch (battery) {
     case Response::Battery::Dying:
-        return BatteryLevel::Empty;
+        return Common::Input::BatteryLevel::Empty;
     case Response::Battery::Low:
-        return BatteryLevel::Critical;
+        return Common::Input::BatteryLevel::Critical;
     case Response::Battery::Medium:
-        return BatteryLevel::Low;
+        return Common::Input::BatteryLevel::Low;
     case Response::Battery::High:
-        return BatteryLevel::Medium;
+        return Common::Input::BatteryLevel::Medium;
     case Response::Battery::Full:
     case Response::Battery::Charged:
-        return BatteryLevel::Full;
+        return Common::Input::BatteryLevel::Full;
     case Response::Battery::Charging:
     default:
-        return BatteryLevel::Charging;
+        return Common::Input::BatteryLevel::Charging;
     }
 }
 
@@ -353,7 +353,7 @@ PadIdentifier UDPClient::GetPadIdentifier(std::size_t pad_index) const {
 
 Common::UUID UDPClient::GetHostUUID(const std::string& host) const {
     const auto ip = boost::asio::ip::make_address_v4(host);
-    const auto hex_host = fmt::format("{:06x}", ip.to_uint());
+    const auto hex_host = fmt::format("00000000-0000-0000-0000-0000{:06x}", ip.to_uint());
     return Common::UUID{hex_host};
 }
 
@@ -385,7 +385,7 @@ std::vector<Common::ParamPackage> UDPClient::GetInputDevices() const {
             Common::ParamPackage identifier{};
             identifier.Set("engine", GetEngineName());
             identifier.Set("display", fmt::format("UDP Controller {}", pad_identifier.pad));
-            identifier.Set("guid", pad_identifier.guid.Format());
+            identifier.Set("guid", pad_identifier.guid.RawString());
             identifier.Set("port", static_cast<int>(pad_identifier.port));
             identifier.Set("pad", static_cast<int>(pad_identifier.pad));
             devices.emplace_back(identifier);
diff --git a/src/input_common/drivers/udp_client.h b/src/input_common/drivers/udp_client.h
index e9c178139..c7cc7d846 100644
--- a/src/input_common/drivers/udp_client.h
+++ b/src/input_common/drivers/udp_client.h
@@ -126,7 +126,7 @@ private:
     struct ClientConnection {
         ClientConnection();
         ~ClientConnection();
-        Common::UUID uuid{"7F000001"};
+        Common::UUID uuid{"00000000-0000-0000-0000-00007F000001"};
         std::string host{"127.0.0.1"};
         u16 port{26760};
         s8 active{-1};
@@ -141,7 +141,7 @@ private:
     std::size_t GetClientNumber(std::string_view host, u16 port) const;
 
     // Translates UDP battery level to input engine battery level
-    BatteryLevel GetBatteryLevel(Response::Battery battery) const;
+    Common::Input::BatteryLevel GetBatteryLevel(Response::Battery battery) const;
 
     void OnVersion(Response::Version);
     void OnPortInfo(Response::PortInfo);
diff --git a/src/input_common/helpers/stick_from_buttons.cpp b/src/input_common/helpers/stick_from_buttons.cpp
index e23394f5f..31e6f62ab 100644
--- a/src/input_common/helpers/stick_from_buttons.cpp
+++ b/src/input_common/helpers/stick_from_buttons.cpp
@@ -167,12 +167,34 @@ public:
     }
 
     void UpdateModButtonStatus(const Common::Input::CallbackStatus& button_callback) {
-        modifier_status = button_callback.button_status.value;
+        const auto& new_status = button_callback.button_status;
+        const bool new_button_value = new_status.inverted ? !new_status.value : new_status.value;
+        modifier_status.toggle = new_status.toggle;
+
+        // Update button status with current
+        if (!modifier_status.toggle) {
+            modifier_status.locked = false;
+            if (modifier_status.value != new_button_value) {
+                modifier_status.value = new_button_value;
+            }
+        } else {
+            // Toggle button and lock status
+            if (new_button_value && !modifier_status.locked) {
+                modifier_status.locked = true;
+                modifier_status.value = !modifier_status.value;
+            }
+
+            // Unlock button ready for next press
+            if (!new_button_value && modifier_status.locked) {
+                modifier_status.locked = false;
+            }
+        }
+
         UpdateStatus();
     }
 
     void UpdateStatus() {
-        const float coef = modifier_status ? modifier_scale : 1.0f;
+        const float coef = modifier_status.value ? modifier_scale : 1.0f;
 
         bool r = right_status;
         bool l = left_status;
@@ -266,7 +288,7 @@ public:
         if (down_status) {
             --y;
         }
-        const float coef = modifier_status ? modifier_scale : 1.0f;
+        const float coef = modifier_status.value ? modifier_scale : 1.0f;
         status.x.raw_value = static_cast<float>(x) * coef * (y == 0 ? 1.0f : SQRT_HALF);
         status.y.raw_value = static_cast<float>(y) * coef * (x == 0 ? 1.0f : SQRT_HALF);
         return status;
@@ -287,9 +309,9 @@ private:
     bool down_status{};
     bool left_status{};
     bool right_status{};
-    bool modifier_status{};
     float last_x_axis_value{};
     float last_y_axis_value{};
+    Common::Input::ButtonStatus modifier_status{};
     const Common::Input::AnalogProperties properties{0.0f, 1.0f, 0.5f, 0.0f, false};
     std::chrono::time_point<std::chrono::steady_clock> last_update;
 };
diff --git a/src/input_common/helpers/touch_from_buttons.cpp b/src/input_common/helpers/touch_from_buttons.cpp
index ece1e3b32..f1b57d03a 100644
--- a/src/input_common/helpers/touch_from_buttons.cpp
+++ b/src/input_common/helpers/touch_from_buttons.cpp
@@ -4,7 +4,6 @@
 
 #include <algorithm>
 #include "common/settings.h"
-#include "core/frontend/framebuffer_layout.h"
 #include "input_common/helpers/touch_from_buttons.h"
 
 namespace InputCommon {
diff --git a/src/input_common/input_engine.cpp b/src/input_common/input_engine.cpp
index 0508b408d..7adf7e3d7 100644
--- a/src/input_common/input_engine.cpp
+++ b/src/input_common/input_engine.cpp
@@ -70,7 +70,7 @@ void InputEngine::SetAxis(const PadIdentifier& identifier, int axis, f32 value)
     TriggerOnAxisChange(identifier, axis, value);
 }
 
-void InputEngine::SetBattery(const PadIdentifier& identifier, BatteryLevel value) {
+void InputEngine::SetBattery(const PadIdentifier& identifier, Common::Input::BatteryLevel value) {
     {
         std::lock_guard lock{mutex};
         ControllerData& controller = controller_list.at(identifier);
@@ -96,7 +96,7 @@ bool InputEngine::GetButton(const PadIdentifier& identifier, int button) const {
     std::lock_guard lock{mutex};
     const auto controller_iter = controller_list.find(identifier);
     if (controller_iter == controller_list.cend()) {
-        LOG_ERROR(Input, "Invalid identifier guid={}, pad={}, port={}", identifier.guid.Format(),
+        LOG_ERROR(Input, "Invalid identifier guid={}, pad={}, port={}", identifier.guid.RawString(),
                   identifier.pad, identifier.port);
         return false;
     }
@@ -113,7 +113,7 @@ bool InputEngine::GetHatButton(const PadIdentifier& identifier, int button, u8 d
     std::lock_guard lock{mutex};
     const auto controller_iter = controller_list.find(identifier);
     if (controller_iter == controller_list.cend()) {
-        LOG_ERROR(Input, "Invalid identifier guid={}, pad={}, port={}", identifier.guid.Format(),
+        LOG_ERROR(Input, "Invalid identifier guid={}, pad={}, port={}", identifier.guid.RawString(),
                   identifier.pad, identifier.port);
         return false;
     }
@@ -130,7 +130,7 @@ f32 InputEngine::GetAxis(const PadIdentifier& identifier, int axis) const {
     std::lock_guard lock{mutex};
     const auto controller_iter = controller_list.find(identifier);
     if (controller_iter == controller_list.cend()) {
-        LOG_ERROR(Input, "Invalid identifier guid={}, pad={}, port={}", identifier.guid.Format(),
+        LOG_ERROR(Input, "Invalid identifier guid={}, pad={}, port={}", identifier.guid.RawString(),
                   identifier.pad, identifier.port);
         return 0.0f;
     }
@@ -143,13 +143,13 @@ f32 InputEngine::GetAxis(const PadIdentifier& identifier, int axis) const {
     return axis_iter->second;
 }
 
-BatteryLevel InputEngine::GetBattery(const PadIdentifier& identifier) const {
+Common::Input::BatteryLevel InputEngine::GetBattery(const PadIdentifier& identifier) const {
     std::lock_guard lock{mutex};
     const auto controller_iter = controller_list.find(identifier);
     if (controller_iter == controller_list.cend()) {
-        LOG_ERROR(Input, "Invalid identifier guid={}, pad={}, port={}", identifier.guid.Format(),
+        LOG_ERROR(Input, "Invalid identifier guid={}, pad={}, port={}", identifier.guid.RawString(),
                   identifier.pad, identifier.port);
-        return BatteryLevel::Charging;
+        return Common::Input::BatteryLevel::Charging;
     }
     const ControllerData& controller = controller_iter->second;
     return controller.battery;
@@ -159,7 +159,7 @@ BasicMotion InputEngine::GetMotion(const PadIdentifier& identifier, int motion)
     std::lock_guard lock{mutex};
     const auto controller_iter = controller_list.find(identifier);
     if (controller_iter == controller_list.cend()) {
-        LOG_ERROR(Input, "Invalid identifier guid={}, pad={}, port={}", identifier.guid.Format(),
+        LOG_ERROR(Input, "Invalid identifier guid={}, pad={}, port={}", identifier.guid.RawString(),
                   identifier.pad, identifier.port);
         return {};
     }
@@ -270,7 +270,7 @@ void InputEngine::TriggerOnAxisChange(const PadIdentifier& identifier, int axis,
 }
 
 void InputEngine::TriggerOnBatteryChange(const PadIdentifier& identifier,
-                                         [[maybe_unused]] BatteryLevel value) {
+                                         [[maybe_unused]] Common::Input::BatteryLevel value) {
     std::lock_guard lock{mutex_callback};
     for (const auto& poller_pair : callback_list) {
         const InputIdentifier& poller = poller_pair.second;
diff --git a/src/input_common/input_engine.h b/src/input_common/input_engine.h
index fe2faee5a..f44e0799b 100644
--- a/src/input_common/input_engine.h
+++ b/src/input_common/input_engine.h
@@ -16,7 +16,7 @@
 
 // Pad Identifier of data source
 struct PadIdentifier {
-    Common::UUID guid{Common::INVALID_UUID};
+    Common::UUID guid{};
     std::size_t port{};
     std::size_t pad{};
 
@@ -34,16 +34,6 @@ struct BasicMotion {
     u64 delta_timestamp{};
 };
 
-// Stages of a battery charge
-enum class BatteryLevel {
-    Empty,
-    Critical,
-    Low,
-    Medium,
-    Full,
-    Charging,
-};
-
 // Types of input that are stored in the engine
 enum class EngineInputType {
     None,
@@ -59,7 +49,7 @@ namespace std {
 template <>
 struct hash<PadIdentifier> {
     size_t operator()(const PadIdentifier& pad_id) const noexcept {
-        u64 hash_value = pad_id.guid.uuid[1] ^ pad_id.guid.uuid[0];
+        u64 hash_value = pad_id.guid.Hash();
         hash_value ^= (static_cast<u64>(pad_id.port) << 32);
         hash_value ^= static_cast<u64>(pad_id.pad);
         return static_cast<size_t>(hash_value);
@@ -178,7 +168,7 @@ public:
     bool GetButton(const PadIdentifier& identifier, int button) const;
     bool GetHatButton(const PadIdentifier& identifier, int button, u8 direction) const;
     f32 GetAxis(const PadIdentifier& identifier, int axis) const;
-    BatteryLevel GetBattery(const PadIdentifier& identifier) const;
+    Common::Input::BatteryLevel GetBattery(const PadIdentifier& identifier) const;
     BasicMotion GetMotion(const PadIdentifier& identifier, int motion) const;
 
     int SetCallback(InputIdentifier input_identifier);
@@ -189,7 +179,7 @@ protected:
     void SetButton(const PadIdentifier& identifier, int button, bool value);
     void SetHatButton(const PadIdentifier& identifier, int button, u8 value);
     void SetAxis(const PadIdentifier& identifier, int axis, f32 value);
-    void SetBattery(const PadIdentifier& identifier, BatteryLevel value);
+    void SetBattery(const PadIdentifier& identifier, Common::Input::BatteryLevel value);
     void SetMotion(const PadIdentifier& identifier, int motion, const BasicMotion& value);
 
     virtual std::string GetHatButtonName([[maybe_unused]] u8 direction_value) const {
@@ -202,13 +192,13 @@ private:
         std::unordered_map<int, u8> hat_buttons;
         std::unordered_map<int, float> axes;
         std::unordered_map<int, BasicMotion> motions;
-        BatteryLevel battery{};
+        Common::Input::BatteryLevel battery{};
     };
 
     void TriggerOnButtonChange(const PadIdentifier& identifier, int button, bool value);
     void TriggerOnHatButtonChange(const PadIdentifier& identifier, int button, u8 value);
     void TriggerOnAxisChange(const PadIdentifier& identifier, int axis, f32 value);
-    void TriggerOnBatteryChange(const PadIdentifier& identifier, BatteryLevel value);
+    void TriggerOnBatteryChange(const PadIdentifier& identifier, Common::Input::BatteryLevel value);
     void TriggerOnMotionChange(const PadIdentifier& identifier, int motion,
                                const BasicMotion& value);
 
diff --git a/src/input_common/input_mapping.cpp b/src/input_common/input_mapping.cpp
index a7a6ad8c2..fb78093b8 100644
--- a/src/input_common/input_mapping.cpp
+++ b/src/input_common/input_mapping.cpp
@@ -57,7 +57,7 @@ void MappingFactory::RegisterButton(const MappingData& data) {
     Common::ParamPackage new_input;
     new_input.Set("engine", data.engine);
     if (data.pad.guid.IsValid()) {
-        new_input.Set("guid", data.pad.guid.Format());
+        new_input.Set("guid", data.pad.guid.RawString());
     }
     new_input.Set("port", static_cast<int>(data.pad.port));
     new_input.Set("pad", static_cast<int>(data.pad.pad));
@@ -93,7 +93,7 @@ void MappingFactory::RegisterStick(const MappingData& data) {
     Common::ParamPackage new_input;
     new_input.Set("engine", data.engine);
     if (data.pad.guid.IsValid()) {
-        new_input.Set("guid", data.pad.guid.Format());
+        new_input.Set("guid", data.pad.guid.RawString());
     }
     new_input.Set("port", static_cast<int>(data.pad.port));
     new_input.Set("pad", static_cast<int>(data.pad.pad));
@@ -138,7 +138,7 @@ void MappingFactory::RegisterMotion(const MappingData& data) {
     Common::ParamPackage new_input;
     new_input.Set("engine", data.engine);
     if (data.pad.guid.IsValid()) {
-        new_input.Set("guid", data.pad.guid.Format());
+        new_input.Set("guid", data.pad.guid.RawString());
     }
     new_input.Set("port", static_cast<int>(data.pad.port));
     new_input.Set("pad", static_cast<int>(data.pad.pad));
diff --git a/src/input_common/input_poller.cpp b/src/input_common/input_poller.cpp
index 2f3c0735a..82b585ff2 100644
--- a/src/input_common/input_poller.cpp
+++ b/src/input_common/input_poller.cpp
@@ -181,7 +181,7 @@ public:
             .raw_value = input_engine->GetAxis(identifier, axis_y),
             .properties = properties_y,
         };
-        // This is a workaround too keep compatibility with old yuzu versions. Vertical axis is
+        // This is a workaround to keep compatibility with old yuzu versions. Vertical axis is
         // inverted on SDL compared to Nintendo
         if (invert_axis_y) {
             status.y.raw_value = -status.y.raw_value;
@@ -470,7 +470,7 @@ public:
     }
 
     Common::Input::BatteryStatus GetStatus() const {
-        return static_cast<Common::Input::BatteryLevel>(input_engine->GetBattery(identifier));
+        return input_engine->GetBattery(identifier);
     }
 
     void ForceUpdate() override {
diff --git a/src/shader_recompiler/backend/glasm/emit_glasm_instructions.h b/src/shader_recompiler/backend/glasm/emit_glasm_instructions.h
index b48007856..5efbe4e6f 100644
--- a/src/shader_recompiler/backend/glasm/emit_glasm_instructions.h
+++ b/src/shader_recompiler/backend/glasm/emit_glasm_instructions.h
@@ -372,6 +372,8 @@ void EmitSharedAtomicExchange32(EmitContext& ctx, IR::Inst& inst, ScalarU32 poin
                                 ScalarU32 value);
 void EmitSharedAtomicExchange64(EmitContext& ctx, IR::Inst& inst, ScalarU32 pointer_offset,
                                 Register value);
+void EmitSharedAtomicExchange32x2(EmitContext& ctx, IR::Inst& inst, ScalarU32 pointer_offset,
+                                  Register value);
 void EmitStorageAtomicIAdd32(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
                              ScalarU32 offset, ScalarU32 value);
 void EmitStorageAtomicSMin32(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
@@ -412,6 +414,24 @@ void EmitStorageAtomicXor64(EmitContext& ctx, IR::Inst& inst, const IR::Value& b
                             ScalarU32 offset, Register value);
 void EmitStorageAtomicExchange64(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
                                  ScalarU32 offset, Register value);
+void EmitStorageAtomicIAdd32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                               ScalarU32 offset, Register value);
+void EmitStorageAtomicSMin32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                               ScalarU32 offset, Register value);
+void EmitStorageAtomicUMin32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                               ScalarU32 offset, Register value);
+void EmitStorageAtomicSMax32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                               ScalarU32 offset, Register value);
+void EmitStorageAtomicUMax32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                               ScalarU32 offset, Register value);
+void EmitStorageAtomicAnd32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                              ScalarU32 offset, Register value);
+void EmitStorageAtomicOr32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                             ScalarU32 offset, Register value);
+void EmitStorageAtomicXor32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                              ScalarU32 offset, Register value);
+void EmitStorageAtomicExchange32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                                   ScalarU32 offset, Register value);
 void EmitStorageAtomicAddF32(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
                              ScalarU32 offset, ScalarF32 value);
 void EmitStorageAtomicAddF16x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
@@ -448,6 +468,17 @@ void EmitGlobalAtomicAnd64(EmitContext& ctx);
 void EmitGlobalAtomicOr64(EmitContext& ctx);
 void EmitGlobalAtomicXor64(EmitContext& ctx);
 void EmitGlobalAtomicExchange64(EmitContext& ctx);
+void EmitGlobalAtomicIAdd32x2(EmitContext& ctx);
+void EmitGlobalAtomicSMin32x2(EmitContext& ctx);
+void EmitGlobalAtomicUMin32x2(EmitContext& ctx);
+void EmitGlobalAtomicSMax32x2(EmitContext& ctx);
+void EmitGlobalAtomicUMax32x2(EmitContext& ctx);
+void EmitGlobalAtomicInc32x2(EmitContext& ctx);
+void EmitGlobalAtomicDec32x2(EmitContext& ctx);
+void EmitGlobalAtomicAnd32x2(EmitContext& ctx);
+void EmitGlobalAtomicOr32x2(EmitContext& ctx);
+void EmitGlobalAtomicXor32x2(EmitContext& ctx);
+void EmitGlobalAtomicExchange32x2(EmitContext& ctx);
 void EmitGlobalAtomicAddF32(EmitContext& ctx);
 void EmitGlobalAtomicAddF16x2(EmitContext& ctx);
 void EmitGlobalAtomicAddF32x2(EmitContext& ctx);
diff --git a/src/shader_recompiler/backend/glasm/emit_glasm_memory.cpp b/src/shader_recompiler/backend/glasm/emit_glasm_memory.cpp
index f135b67f5..f0fd94a28 100644
--- a/src/shader_recompiler/backend/glasm/emit_glasm_memory.cpp
+++ b/src/shader_recompiler/backend/glasm/emit_glasm_memory.cpp
@@ -311,6 +311,13 @@ void EmitSharedAtomicExchange64(EmitContext& ctx, IR::Inst& inst, ScalarU32 poin
     ctx.LongAdd("ATOMS.EXCH.U64 {}.x,{},shared_mem[{}];", inst, value, pointer_offset);
 }
 
+void EmitSharedAtomicExchange32x2([[maybe_unused]] EmitContext& ctx,
+                                  [[maybe_unused]] IR::Inst& inst,
+                                  [[maybe_unused]] ScalarU32 pointer_offset,
+                                  [[maybe_unused]] Register value) {
+    throw NotImplementedException("GLASM instruction");
+}
+
 void EmitStorageAtomicIAdd32(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
                              ScalarU32 offset, ScalarU32 value) {
     Atom(ctx, inst, binding, offset, value, "ADD", "U32");
@@ -411,6 +418,62 @@ void EmitStorageAtomicExchange64(EmitContext& ctx, IR::Inst& inst, const IR::Val
     Atom(ctx, inst, binding, offset, value, "EXCH", "U64");
 }
 
+void EmitStorageAtomicIAdd32x2([[maybe_unused]] EmitContext& ctx, [[maybe_unused]] IR::Inst& inst,
+                               [[maybe_unused]] const IR::Value& binding,
+                               [[maybe_unused]] ScalarU32 offset, [[maybe_unused]] Register value) {
+    throw NotImplementedException("GLASM instruction");
+}
+
+void EmitStorageAtomicSMin32x2([[maybe_unused]] EmitContext& ctx, [[maybe_unused]] IR::Inst& inst,
+                               [[maybe_unused]] const IR::Value& binding,
+                               [[maybe_unused]] ScalarU32 offset, [[maybe_unused]] Register value) {
+    throw NotImplementedException("GLASM instruction");
+}
+
+void EmitStorageAtomicUMin32x2([[maybe_unused]] EmitContext& ctx, [[maybe_unused]] IR::Inst& inst,
+                               [[maybe_unused]] const IR::Value& binding,
+                               [[maybe_unused]] ScalarU32 offset, [[maybe_unused]] Register value) {
+    throw NotImplementedException("GLASM instruction");
+}
+
+void EmitStorageAtomicSMax32x2([[maybe_unused]] EmitContext& ctx, [[maybe_unused]] IR::Inst& inst,
+                               [[maybe_unused]] const IR::Value& binding,
+                               [[maybe_unused]] ScalarU32 offset, [[maybe_unused]] Register value) {
+    throw NotImplementedException("GLASM instruction");
+}
+
+void EmitStorageAtomicUMax32x2([[maybe_unused]] EmitContext& ctx, [[maybe_unused]] IR::Inst& inst,
+                               [[maybe_unused]] const IR::Value& binding,
+                               [[maybe_unused]] ScalarU32 offset, [[maybe_unused]] Register value) {
+    throw NotImplementedException("GLASM instruction");
+}
+
+void EmitStorageAtomicAnd32x2([[maybe_unused]] EmitContext& ctx, [[maybe_unused]] IR::Inst& inst,
+                              [[maybe_unused]] const IR::Value& binding,
+                              [[maybe_unused]] ScalarU32 offset, [[maybe_unused]] Register value) {
+    throw NotImplementedException("GLASM instruction");
+}
+
+void EmitStorageAtomicOr32x2([[maybe_unused]] EmitContext& ctx, [[maybe_unused]] IR::Inst& inst,
+                             [[maybe_unused]] const IR::Value& binding,
+                             [[maybe_unused]] ScalarU32 offset, [[maybe_unused]] Register value) {
+    throw NotImplementedException("GLASM instruction");
+}
+
+void EmitStorageAtomicXor32x2([[maybe_unused]] EmitContext& ctx, [[maybe_unused]] IR::Inst& inst,
+                              [[maybe_unused]] const IR::Value& binding,
+                              [[maybe_unused]] ScalarU32 offset, [[maybe_unused]] Register value) {
+    throw NotImplementedException("GLASM instruction");
+}
+
+void EmitStorageAtomicExchange32x2([[maybe_unused]] EmitContext& ctx,
+                                   [[maybe_unused]] IR::Inst& inst,
+                                   [[maybe_unused]] const IR::Value& binding,
+                                   [[maybe_unused]] ScalarU32 offset,
+                                   [[maybe_unused]] Register value) {
+    throw NotImplementedException("GLASM instruction");
+}
+
 void EmitStorageAtomicAddF32(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
                              ScalarU32 offset, ScalarF32 value) {
     Atom(ctx, inst, binding, offset, value, "ADD", "F32");
@@ -537,6 +600,50 @@ void EmitGlobalAtomicExchange64(EmitContext&) {
     throw NotImplementedException("GLASM instruction");
 }
 
+void EmitGlobalAtomicIAdd32x2(EmitContext&) {
+    throw NotImplementedException("GLASM instruction");
+}
+
+void EmitGlobalAtomicSMin32x2(EmitContext&) {
+    throw NotImplementedException("GLASM instruction");
+}
+
+void EmitGlobalAtomicUMin32x2(EmitContext&) {
+    throw NotImplementedException("GLASM instruction");
+}
+
+void EmitGlobalAtomicSMax32x2(EmitContext&) {
+    throw NotImplementedException("GLASM instruction");
+}
+
+void EmitGlobalAtomicUMax32x2(EmitContext&) {
+    throw NotImplementedException("GLASM instruction");
+}
+
+void EmitGlobalAtomicInc32x2(EmitContext&) {
+    throw NotImplementedException("GLASM instruction");
+}
+
+void EmitGlobalAtomicDec32x2(EmitContext&) {
+    throw NotImplementedException("GLASM instruction");
+}
+
+void EmitGlobalAtomicAnd32x2(EmitContext&) {
+    throw NotImplementedException("GLASM instruction");
+}
+
+void EmitGlobalAtomicOr32x2(EmitContext&) {
+    throw NotImplementedException("GLASM instruction");
+}
+
+void EmitGlobalAtomicXor32x2(EmitContext&) {
+    throw NotImplementedException("GLASM instruction");
+}
+
+void EmitGlobalAtomicExchange32x2(EmitContext&) {
+    throw NotImplementedException("GLASM instruction");
+}
+
 void EmitGlobalAtomicAddF32(EmitContext&) {
     throw NotImplementedException("GLASM instruction");
 }
diff --git a/src/shader_recompiler/backend/glsl/emit_glsl_atomic.cpp b/src/shader_recompiler/backend/glsl/emit_glsl_atomic.cpp
index dc377b053..a409a7ab3 100644
--- a/src/shader_recompiler/backend/glsl/emit_glsl_atomic.cpp
+++ b/src/shader_recompiler/backend/glsl/emit_glsl_atomic.cpp
@@ -105,6 +105,13 @@ void EmitSharedAtomicExchange64(EmitContext& ctx, IR::Inst& inst, std::string_vi
             pointer_offset, value, pointer_offset, value);
 }
 
+void EmitSharedAtomicExchange32x2(EmitContext& ctx, IR::Inst& inst, std::string_view pointer_offset,
+                                  std::string_view value) {
+    LOG_WARNING(Shader_GLSL, "Int64 atomics not supported, fallback to non-atomic");
+    ctx.AddU32x2("{}=uvec2(smem[{}>>2],smem[({}+4)>>2]);", inst, pointer_offset, pointer_offset);
+    ctx.Add("smem[{}>>2]={}.x;smem[({}+4)>>2]={}.y;", pointer_offset, value, pointer_offset, value);
+}
+
 void EmitStorageAtomicIAdd32(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
                              const IR::Value& offset, std::string_view value) {
     ctx.AddU32("{}=atomicAdd({}_ssbo{}[{}>>2],{});", inst, ctx.stage_name, binding.U32(),
@@ -265,6 +272,97 @@ void EmitStorageAtomicExchange64(EmitContext& ctx, IR::Inst& inst, const IR::Val
                ctx.stage_name, binding.U32(), ctx.var_alloc.Consume(offset), value);
 }
 
+void EmitStorageAtomicIAdd32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                               const IR::Value& offset, std::string_view value) {
+    LOG_WARNING(Shader_GLSL, "Int64 atomics not supported, fallback to non-atomic");
+    ctx.AddU32x2("{}=uvec2({}_ssbo{}[{}>>2],{}_ssbo{}[({}>>2)+1]);", inst, ctx.stage_name,
+                 binding.U32(), ctx.var_alloc.Consume(offset), ctx.stage_name, binding.U32(),
+                 ctx.var_alloc.Consume(offset));
+    ctx.Add("{}_ssbo{}[{}>>2]+={}.x;{}_ssbo{}[({}>>2)+1]+={}.y;", ctx.stage_name, binding.U32(),
+            ctx.var_alloc.Consume(offset), value, ctx.stage_name, binding.U32(),
+            ctx.var_alloc.Consume(offset), value);
+}
+
+void EmitStorageAtomicSMin32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                               const IR::Value& offset, std::string_view value) {
+    LOG_WARNING(Shader_GLSL, "Int64 atomics not supported, fallback to non-atomic");
+    ctx.AddU32x2("{}=ivec2({}_ssbo{}[{}>>2],{}_ssbo{}[({}>>2)+1]);", inst, ctx.stage_name,
+                 binding.U32(), ctx.var_alloc.Consume(offset), ctx.stage_name, binding.U32(),
+                 ctx.var_alloc.Consume(offset));
+    ctx.Add("for(int "
+            "i=0;i<2;++i){{{}_ssbo{}[({}>>2)+i]=uint(min(int({}_ssbo{}[({}>>2)+i]),int({}[i])));}}",
+            ctx.stage_name, binding.U32(), ctx.var_alloc.Consume(offset), ctx.stage_name,
+            binding.U32(), ctx.var_alloc.Consume(offset), value);
+}
+
+void EmitStorageAtomicUMin32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                               const IR::Value& offset, std::string_view value) {
+    LOG_WARNING(Shader_GLSL, "Int64 atomics not supported, fallback to non-atomic");
+    ctx.AddU32x2("{}=uvec2({}_ssbo{}[{}>>2],{}_ssbo{}[({}>>2)+1]);", inst, ctx.stage_name,
+                 binding.U32(), ctx.var_alloc.Consume(offset), ctx.stage_name, binding.U32(),
+                 ctx.var_alloc.Consume(offset));
+    ctx.Add("for(int i=0;i<2;++i){{ "
+            "{}_ssbo{}[({}>>2)+i]=min({}_ssbo{}[({}>>2)+i],{}[i]);}}",
+            ctx.stage_name, binding.U32(), ctx.var_alloc.Consume(offset), ctx.stage_name,
+            binding.U32(), ctx.var_alloc.Consume(offset), value);
+}
+
+void EmitStorageAtomicSMax32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                               const IR::Value& offset, std::string_view value) {
+    LOG_WARNING(Shader_GLSL, "Int64 atomics not supported, fallback to non-atomic");
+    ctx.AddU32x2("{}=ivec2({}_ssbo{}[{}>>2],{}_ssbo{}[({}>>2)+1]);", inst, ctx.stage_name,
+                 binding.U32(), ctx.var_alloc.Consume(offset), ctx.stage_name, binding.U32(),
+                 ctx.var_alloc.Consume(offset));
+    ctx.Add("for(int "
+            "i=0;i<2;++i){{{}_ssbo{}[({}>>2)+i]=uint(max(int({}_ssbo{}[({}>>2)+i]),int({}[i])));}}",
+            ctx.stage_name, binding.U32(), ctx.var_alloc.Consume(offset), ctx.stage_name,
+            binding.U32(), ctx.var_alloc.Consume(offset), value);
+}
+
+void EmitStorageAtomicUMax32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                               const IR::Value& offset, std::string_view value) {
+    LOG_WARNING(Shader_GLSL, "Int64 atomics not supported, fallback to non-atomic");
+    ctx.AddU32x2("{}=uvec2({}_ssbo{}[{}>>2],{}_ssbo{}[({}>>2)+1]);", inst, ctx.stage_name,
+                 binding.U32(), ctx.var_alloc.Consume(offset), ctx.stage_name, binding.U32(),
+                 ctx.var_alloc.Consume(offset));
+    ctx.Add("for(int i=0;i<2;++i){{{}_ssbo{}[({}>>2)+i]=max({}_ssbo{}[({}>>2)+i],{}[i]);}}",
+            ctx.stage_name, binding.U32(), ctx.var_alloc.Consume(offset), ctx.stage_name,
+            binding.U32(), ctx.var_alloc.Consume(offset), value);
+}
+
+void EmitStorageAtomicAnd32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                              const IR::Value& offset, std::string_view value) {
+    LOG_WARNING(Shader_GLSL, "Int64 atomics not supported, fallback to 32x2");
+    ctx.AddU32x2("{}=uvec2(atomicAnd({}_ssbo{}[{}>>2],{}.x),atomicAnd({}_ssbo{}[({}>>2)+1],{}.y));",
+                 inst, ctx.stage_name, binding.U32(), ctx.var_alloc.Consume(offset), value,
+                 ctx.stage_name, binding.U32(), ctx.var_alloc.Consume(offset), value);
+}
+
+void EmitStorageAtomicOr32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                             const IR::Value& offset, std::string_view value) {
+    LOG_WARNING(Shader_GLSL, "Int64 atomics not supported, fallback to 32x2");
+    ctx.AddU32x2("{}=uvec2(atomicOr({}_ssbo{}[{}>>2],{}.x),atomicOr({}_ssbo{}[({}>>2)+1],{}.y));",
+                 inst, ctx.stage_name, binding.U32(), ctx.var_alloc.Consume(offset), value,
+                 ctx.stage_name, binding.U32(), ctx.var_alloc.Consume(offset), value);
+}
+
+void EmitStorageAtomicXor32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                              const IR::Value& offset, std::string_view value) {
+    LOG_WARNING(Shader_GLSL, "Int64 atomics not supported, fallback to 32x2");
+    ctx.AddU32x2("{}=uvec2(atomicXor({}_ssbo{}[{}>>2],{}.x),atomicXor({}_ssbo{}[({}>>2)+1],{}.y));",
+                 inst, ctx.stage_name, binding.U32(), ctx.var_alloc.Consume(offset), value,
+                 ctx.stage_name, binding.U32(), ctx.var_alloc.Consume(offset), value);
+}
+
+void EmitStorageAtomicExchange32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                                   const IR::Value& offset, std::string_view value) {
+    LOG_WARNING(Shader_GLSL, "Int64 atomics not supported, fallback to 32x2");
+    ctx.AddU32x2("{}=uvec2(atomicExchange({}_ssbo{}[{}>>2],{}.x),atomicExchange({}_ssbo{}[({}>>2)+"
+                 "1],{}.y));",
+                 inst, ctx.stage_name, binding.U32(), ctx.var_alloc.Consume(offset), value,
+                 ctx.stage_name, binding.U32(), ctx.var_alloc.Consume(offset), value);
+}
+
 void EmitStorageAtomicAddF32(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
                              const IR::Value& offset, std::string_view value) {
     SsboCasFunctionF32(ctx, inst, binding, offset, value, "CasFloatAdd");
@@ -388,6 +486,50 @@ void EmitGlobalAtomicExchange64(EmitContext&) {
     throw NotImplementedException("GLSL Instrucion");
 }
 
+void EmitGlobalAtomicIAdd32x2(EmitContext&) {
+    throw NotImplementedException("GLSL Instrucion");
+}
+
+void EmitGlobalAtomicSMin32x2(EmitContext&) {
+    throw NotImplementedException("GLSL Instrucion");
+}
+
+void EmitGlobalAtomicUMin32x2(EmitContext&) {
+    throw NotImplementedException("GLSL Instrucion");
+}
+
+void EmitGlobalAtomicSMax32x2(EmitContext&) {
+    throw NotImplementedException("GLSL Instrucion");
+}
+
+void EmitGlobalAtomicUMax32x2(EmitContext&) {
+    throw NotImplementedException("GLSL Instrucion");
+}
+
+void EmitGlobalAtomicInc32x2(EmitContext&) {
+    throw NotImplementedException("GLSL Instrucion");
+}
+
+void EmitGlobalAtomicDec32x2(EmitContext&) {
+    throw NotImplementedException("GLSL Instrucion");
+}
+
+void EmitGlobalAtomicAnd32x2(EmitContext&) {
+    throw NotImplementedException("GLSL Instrucion");
+}
+
+void EmitGlobalAtomicOr32x2(EmitContext&) {
+    throw NotImplementedException("GLSL Instrucion");
+}
+
+void EmitGlobalAtomicXor32x2(EmitContext&) {
+    throw NotImplementedException("GLSL Instrucion");
+}
+
+void EmitGlobalAtomicExchange32x2(EmitContext&) {
+    throw NotImplementedException("GLSL Instrucion");
+}
+
 void EmitGlobalAtomicAddF32(EmitContext&) {
     throw NotImplementedException("GLSL Instrucion");
 }
diff --git a/src/shader_recompiler/backend/glsl/emit_glsl_instructions.h b/src/shader_recompiler/backend/glsl/emit_glsl_instructions.h
index 6cabbc717..704baddc9 100644
--- a/src/shader_recompiler/backend/glsl/emit_glsl_instructions.h
+++ b/src/shader_recompiler/backend/glsl/emit_glsl_instructions.h
@@ -442,6 +442,8 @@ void EmitSharedAtomicExchange32(EmitContext& ctx, IR::Inst& inst, std::string_vi
                                 std::string_view value);
 void EmitSharedAtomicExchange64(EmitContext& ctx, IR::Inst& inst, std::string_view pointer_offset,
                                 std::string_view value);
+void EmitSharedAtomicExchange32x2(EmitContext& ctx, IR::Inst& inst, std::string_view pointer_offset,
+                                  std::string_view value);
 void EmitStorageAtomicIAdd32(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
                              const IR::Value& offset, std::string_view value);
 void EmitStorageAtomicSMin32(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
@@ -482,6 +484,24 @@ void EmitStorageAtomicXor64(EmitContext& ctx, IR::Inst& inst, const IR::Value& b
                             const IR::Value& offset, std::string_view value);
 void EmitStorageAtomicExchange64(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
                                  const IR::Value& offset, std::string_view value);
+void EmitStorageAtomicIAdd32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                               const IR::Value& offset, std::string_view value);
+void EmitStorageAtomicSMin32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                               const IR::Value& offset, std::string_view value);
+void EmitStorageAtomicUMin32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                               const IR::Value& offset, std::string_view value);
+void EmitStorageAtomicSMax32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                               const IR::Value& offset, std::string_view value);
+void EmitStorageAtomicUMax32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                               const IR::Value& offset, std::string_view value);
+void EmitStorageAtomicAnd32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                              const IR::Value& offset, std::string_view value);
+void EmitStorageAtomicOr32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                             const IR::Value& offset, std::string_view value);
+void EmitStorageAtomicXor32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                              const IR::Value& offset, std::string_view value);
+void EmitStorageAtomicExchange32x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
+                                   const IR::Value& offset, std::string_view value);
 void EmitStorageAtomicAddF32(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
                              const IR::Value& offset, std::string_view value);
 void EmitStorageAtomicAddF16x2(EmitContext& ctx, IR::Inst& inst, const IR::Value& binding,
@@ -518,6 +538,17 @@ void EmitGlobalAtomicAnd64(EmitContext& ctx);
 void EmitGlobalAtomicOr64(EmitContext& ctx);
 void EmitGlobalAtomicXor64(EmitContext& ctx);
 void EmitGlobalAtomicExchange64(EmitContext& ctx);
+void EmitGlobalAtomicIAdd32x2(EmitContext& ctx);
+void EmitGlobalAtomicSMin32x2(EmitContext& ctx);
+void EmitGlobalAtomicUMin32x2(EmitContext& ctx);
+void EmitGlobalAtomicSMax32x2(EmitContext& ctx);
+void EmitGlobalAtomicUMax32x2(EmitContext& ctx);
+void EmitGlobalAtomicInc32x2(EmitContext& ctx);
+void EmitGlobalAtomicDec32x2(EmitContext& ctx);
+void EmitGlobalAtomicAnd32x2(EmitContext& ctx);
+void EmitGlobalAtomicOr32x2(EmitContext& ctx);
+void EmitGlobalAtomicXor32x2(EmitContext& ctx);
+void EmitGlobalAtomicExchange32x2(EmitContext& ctx);
 void EmitGlobalAtomicAddF32(EmitContext& ctx);
 void EmitGlobalAtomicAddF16x2(EmitContext& ctx);
 void EmitGlobalAtomicAddF32x2(EmitContext& ctx);
diff --git a/src/shader_recompiler/backend/spirv/emit_spirv.h b/src/shader_recompiler/backend/spirv/emit_spirv.h
index b412957c7..2b360e073 100644
--- a/src/shader_recompiler/backend/spirv/emit_spirv.h
+++ b/src/shader_recompiler/backend/spirv/emit_spirv.h
@@ -22,7 +22,7 @@ constexpr u32 NUM_TEXTURE_AND_IMAGE_SCALING_WORDS =
 struct RescalingLayout {
     alignas(16) std::array<u32, NUM_TEXTURE_SCALING_WORDS> rescaling_textures;
     alignas(16) std::array<u32, NUM_IMAGE_SCALING_WORDS> rescaling_images;
-    alignas(16) u32 down_factor;
+    u32 down_factor;
 };
 constexpr u32 RESCALING_LAYOUT_WORDS_OFFSET = offsetof(RescalingLayout, rescaling_textures);
 constexpr u32 RESCALING_LAYOUT_DOWN_FACTOR_OFFSET = offsetof(RescalingLayout, down_factor);
diff --git a/src/shader_recompiler/backend/spirv/emit_spirv_atomic.cpp b/src/shader_recompiler/backend/spirv/emit_spirv_atomic.cpp
index 46ba52a25..d3cbb14a9 100644
--- a/src/shader_recompiler/backend/spirv/emit_spirv_atomic.cpp
+++ b/src/shader_recompiler/backend/spirv/emit_spirv_atomic.cpp
@@ -82,6 +82,17 @@ Id StorageAtomicU64(EmitContext& ctx, const IR::Value& binding, const IR::Value&
     ctx.OpStore(pointer, ctx.OpBitcast(ctx.U32[2], result));
     return original_value;
 }
+
+Id StorageAtomicU32x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset, Id value,
+                      Id (Sirit::Module::*non_atomic_func)(Id, Id, Id)) {
+    LOG_WARNING(Shader_SPIRV, "Int64 atomics not supported, fallback to non-atomic");
+    const Id pointer{StoragePointer(ctx, ctx.storage_types.U32x2, &StorageDefinitions::U32x2,
+                                    binding, offset, sizeof(u32[2]))};
+    const Id original_value{ctx.OpLoad(ctx.U32[2], pointer)};
+    const Id result{(ctx.*non_atomic_func)(ctx.U32[2], value, original_value)};
+    ctx.OpStore(pointer, result);
+    return original_value;
+}
 } // Anonymous namespace
 
 Id EmitSharedAtomicIAdd32(EmitContext& ctx, Id offset, Id value) {
@@ -141,7 +152,7 @@ Id EmitSharedAtomicExchange64(EmitContext& ctx, Id offset, Id value) {
         const auto [scope, semantics]{AtomicArgs(ctx)};
         return ctx.OpAtomicExchange(ctx.U64, pointer, scope, semantics, value);
     }
-    LOG_ERROR(Shader_SPIRV, "Int64 atomics not supported, fallback to non-atomic");
+    LOG_WARNING(Shader_SPIRV, "Int64 atomics not supported, fallback to non-atomic");
     const Id pointer_1{SharedPointer(ctx, offset, 0)};
     const Id pointer_2{SharedPointer(ctx, offset, 1)};
     const Id value_1{ctx.OpLoad(ctx.U32[1], pointer_1)};
@@ -152,6 +163,18 @@ Id EmitSharedAtomicExchange64(EmitContext& ctx, Id offset, Id value) {
     return ctx.OpBitcast(ctx.U64, ctx.OpCompositeConstruct(ctx.U32[2], value_1, value_2));
 }
 
+Id EmitSharedAtomicExchange32x2(EmitContext& ctx, Id offset, Id value) {
+    LOG_WARNING(Shader_SPIRV, "Int64 atomics not supported, fallback to non-atomic");
+    const Id pointer_1{SharedPointer(ctx, offset, 0)};
+    const Id pointer_2{SharedPointer(ctx, offset, 1)};
+    const Id value_1{ctx.OpLoad(ctx.U32[1], pointer_1)};
+    const Id value_2{ctx.OpLoad(ctx.U32[1], pointer_2)};
+    const Id new_vector{ctx.OpBitcast(ctx.U32[2], value)};
+    ctx.OpStore(pointer_1, ctx.OpCompositeExtract(ctx.U32[1], new_vector, 0U));
+    ctx.OpStore(pointer_2, ctx.OpCompositeExtract(ctx.U32[1], new_vector, 1U));
+    return ctx.OpCompositeConstruct(ctx.U32[2], value_1, value_2);
+}
+
 Id EmitStorageAtomicIAdd32(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
                            Id value) {
     return StorageAtomicU32(ctx, binding, offset, value, &Sirit::Module::OpAtomicIAdd);
@@ -275,6 +298,56 @@ Id EmitStorageAtomicExchange64(EmitContext& ctx, const IR::Value& binding, const
     return original;
 }
 
+Id EmitStorageAtomicIAdd32x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
+                             Id value) {
+    return StorageAtomicU32x2(ctx, binding, offset, value, &Sirit::Module::OpIAdd);
+}
+
+Id EmitStorageAtomicSMin32x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
+                             Id value) {
+    return StorageAtomicU32x2(ctx, binding, offset, value, &Sirit::Module::OpSMin);
+}
+
+Id EmitStorageAtomicUMin32x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
+                             Id value) {
+    return StorageAtomicU32x2(ctx, binding, offset, value, &Sirit::Module::OpUMin);
+}
+
+Id EmitStorageAtomicSMax32x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
+                             Id value) {
+    return StorageAtomicU32x2(ctx, binding, offset, value, &Sirit::Module::OpSMax);
+}
+
+Id EmitStorageAtomicUMax32x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
+                             Id value) {
+    return StorageAtomicU32x2(ctx, binding, offset, value, &Sirit::Module::OpUMax);
+}
+
+Id EmitStorageAtomicAnd32x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
+                            Id value) {
+    return StorageAtomicU32x2(ctx, binding, offset, value, &Sirit::Module::OpBitwiseAnd);
+}
+
+Id EmitStorageAtomicOr32x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
+                           Id value) {
+    return StorageAtomicU32x2(ctx, binding, offset, value, &Sirit::Module::OpBitwiseOr);
+}
+
+Id EmitStorageAtomicXor32x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
+                            Id value) {
+    return StorageAtomicU32x2(ctx, binding, offset, value, &Sirit::Module::OpBitwiseXor);
+}
+
+Id EmitStorageAtomicExchange32x2(EmitContext& ctx, const IR::Value& binding,
+                                 const IR::Value& offset, Id value) {
+    LOG_WARNING(Shader_SPIRV, "Int64 atomics not supported, fallback to non-atomic");
+    const Id pointer{StoragePointer(ctx, ctx.storage_types.U32x2, &StorageDefinitions::U32x2,
+                                    binding, offset, sizeof(u32[2]))};
+    const Id original{ctx.OpLoad(ctx.U32[2], pointer)};
+    ctx.OpStore(pointer, value);
+    return original;
+}
+
 Id EmitStorageAtomicAddF32(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
                            Id value) {
     const Id ssbo{ctx.ssbos[binding.U32()].U32};
@@ -418,6 +491,50 @@ Id EmitGlobalAtomicExchange64(EmitContext&) {
     throw NotImplementedException("SPIR-V Instruction");
 }
 
+Id EmitGlobalAtomicIAdd32x2(EmitContext&) {
+    throw NotImplementedException("SPIR-V Instruction");
+}
+
+Id EmitGlobalAtomicSMin32x2(EmitContext&) {
+    throw NotImplementedException("SPIR-V Instruction");
+}
+
+Id EmitGlobalAtomicUMin32x2(EmitContext&) {
+    throw NotImplementedException("SPIR-V Instruction");
+}
+
+Id EmitGlobalAtomicSMax32x2(EmitContext&) {
+    throw NotImplementedException("SPIR-V Instruction");
+}
+
+Id EmitGlobalAtomicUMax32x2(EmitContext&) {
+    throw NotImplementedException("SPIR-V Instruction");
+}
+
+Id EmitGlobalAtomicInc32x2(EmitContext&) {
+    throw NotImplementedException("SPIR-V Instruction");
+}
+
+Id EmitGlobalAtomicDec32x2(EmitContext&) {
+    throw NotImplementedException("SPIR-V Instruction");
+}
+
+Id EmitGlobalAtomicAnd32x2(EmitContext&) {
+    throw NotImplementedException("SPIR-V Instruction");
+}
+
+Id EmitGlobalAtomicOr32x2(EmitContext&) {
+    throw NotImplementedException("SPIR-V Instruction");
+}
+
+Id EmitGlobalAtomicXor32x2(EmitContext&) {
+    throw NotImplementedException("SPIR-V Instruction");
+}
+
+Id EmitGlobalAtomicExchange32x2(EmitContext&) {
+    throw NotImplementedException("SPIR-V Instruction");
+}
+
 Id EmitGlobalAtomicAddF32(EmitContext&) {
     throw NotImplementedException("SPIR-V Instruction");
 }
diff --git a/src/shader_recompiler/backend/spirv/emit_spirv_context_get_set.cpp b/src/shader_recompiler/backend/spirv/emit_spirv_context_get_set.cpp
index 8ea730c80..80b4bbd27 100644
--- a/src/shader_recompiler/backend/spirv/emit_spirv_context_get_set.cpp
+++ b/src/shader_recompiler/backend/spirv/emit_spirv_context_get_set.cpp
@@ -123,34 +123,36 @@ std::optional<OutAttr> OutputAttrPointer(EmitContext& ctx, IR::Attribute attr) {
 }
 
 Id GetCbuf(EmitContext& ctx, Id result_type, Id UniformDefinitions::*member_ptr, u32 element_size,
-           const IR::Value& binding, const IR::Value& offset) {
+           const IR::Value& binding, const IR::Value& offset, const Id indirect_func) {
+    Id buffer_offset;
+    const Id uniform_type{ctx.uniform_types.*member_ptr};
+    if (offset.IsImmediate()) {
+        // Hardware been proved to read the aligned offset (e.g. LDC.U32 at 6 will read offset 4)
+        const Id imm_offset{ctx.Const(offset.U32() / element_size)};
+        buffer_offset = imm_offset;
+    } else if (element_size > 1) {
+        const u32 log2_element_size{static_cast<u32>(std::countr_zero(element_size))};
+        const Id shift{ctx.Const(log2_element_size)};
+        buffer_offset = ctx.OpShiftRightArithmetic(ctx.U32[1], ctx.Def(offset), shift);
+    } else {
+        buffer_offset = ctx.Def(offset);
+    }
     if (!binding.IsImmediate()) {
-        throw NotImplementedException("Constant buffer indexing");
+        return ctx.OpFunctionCall(result_type, indirect_func, ctx.Def(binding), buffer_offset);
     }
     const Id cbuf{ctx.cbufs[binding.U32()].*member_ptr};
-    const Id uniform_type{ctx.uniform_types.*member_ptr};
-    if (!offset.IsImmediate()) {
-        Id index{ctx.Def(offset)};
-        if (element_size > 1) {
-            const u32 log2_element_size{static_cast<u32>(std::countr_zero(element_size))};
-            const Id shift{ctx.Const(log2_element_size)};
-            index = ctx.OpShiftRightArithmetic(ctx.U32[1], ctx.Def(offset), shift);
-        }
-        const Id access_chain{ctx.OpAccessChain(uniform_type, cbuf, ctx.u32_zero_value, index)};
-        return ctx.OpLoad(result_type, access_chain);
-    }
-    // Hardware been proved to read the aligned offset (e.g. LDC.U32 at 6 will read offset 4)
-    const Id imm_offset{ctx.Const(offset.U32() / element_size)};
-    const Id access_chain{ctx.OpAccessChain(uniform_type, cbuf, ctx.u32_zero_value, imm_offset)};
+    const Id access_chain{ctx.OpAccessChain(uniform_type, cbuf, ctx.u32_zero_value, buffer_offset)};
     return ctx.OpLoad(result_type, access_chain);
 }
 
 Id GetCbufU32(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset) {
-    return GetCbuf(ctx, ctx.U32[1], &UniformDefinitions::U32, sizeof(u32), binding, offset);
+    return GetCbuf(ctx, ctx.U32[1], &UniformDefinitions::U32, sizeof(u32), binding, offset,
+                   ctx.load_const_func_u32);
 }
 
 Id GetCbufU32x4(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset) {
-    return GetCbuf(ctx, ctx.U32[4], &UniformDefinitions::U32x4, sizeof(u32[4]), binding, offset);
+    return GetCbuf(ctx, ctx.U32[4], &UniformDefinitions::U32x4, sizeof(u32[4]), binding, offset,
+                   ctx.load_const_func_u32x4);
 }
 
 Id GetCbufElement(EmitContext& ctx, Id vector, const IR::Value& offset, u32 index_offset) {
@@ -201,7 +203,8 @@ void EmitGetIndirectBranchVariable(EmitContext&) {
 
 Id EmitGetCbufU8(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset) {
     if (ctx.profile.support_descriptor_aliasing && ctx.profile.support_int8) {
-        const Id load{GetCbuf(ctx, ctx.U8, &UniformDefinitions::U8, sizeof(u8), binding, offset)};
+        const Id load{GetCbuf(ctx, ctx.U8, &UniformDefinitions::U8, sizeof(u8), binding, offset,
+                              ctx.load_const_func_u8)};
         return ctx.OpUConvert(ctx.U32[1], load);
     }
     Id element{};
@@ -217,7 +220,8 @@ Id EmitGetCbufU8(EmitContext& ctx, const IR::Value& binding, const IR::Value& of
 
 Id EmitGetCbufS8(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset) {
     if (ctx.profile.support_descriptor_aliasing && ctx.profile.support_int8) {
-        const Id load{GetCbuf(ctx, ctx.S8, &UniformDefinitions::S8, sizeof(s8), binding, offset)};
+        const Id load{GetCbuf(ctx, ctx.S8, &UniformDefinitions::S8, sizeof(s8), binding, offset,
+                              ctx.load_const_func_u8)};
         return ctx.OpSConvert(ctx.U32[1], load);
     }
     Id element{};
@@ -233,8 +237,8 @@ Id EmitGetCbufS8(EmitContext& ctx, const IR::Value& binding, const IR::Value& of
 
 Id EmitGetCbufU16(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset) {
     if (ctx.profile.support_descriptor_aliasing && ctx.profile.support_int16) {
-        const Id load{
-            GetCbuf(ctx, ctx.U16, &UniformDefinitions::U16, sizeof(u16), binding, offset)};
+        const Id load{GetCbuf(ctx, ctx.U16, &UniformDefinitions::U16, sizeof(u16), binding, offset,
+                              ctx.load_const_func_u16)};
         return ctx.OpUConvert(ctx.U32[1], load);
     }
     Id element{};
@@ -250,8 +254,8 @@ Id EmitGetCbufU16(EmitContext& ctx, const IR::Value& binding, const IR::Value& o
 
 Id EmitGetCbufS16(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset) {
     if (ctx.profile.support_descriptor_aliasing && ctx.profile.support_int16) {
-        const Id load{
-            GetCbuf(ctx, ctx.S16, &UniformDefinitions::S16, sizeof(s16), binding, offset)};
+        const Id load{GetCbuf(ctx, ctx.S16, &UniformDefinitions::S16, sizeof(s16), binding, offset,
+                              ctx.load_const_func_u16)};
         return ctx.OpSConvert(ctx.U32[1], load);
     }
     Id element{};
@@ -276,7 +280,8 @@ Id EmitGetCbufU32(EmitContext& ctx, const IR::Value& binding, const IR::Value& o
 
 Id EmitGetCbufF32(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset) {
     if (ctx.profile.support_descriptor_aliasing) {
-        return GetCbuf(ctx, ctx.F32[1], &UniformDefinitions::F32, sizeof(f32), binding, offset);
+        return GetCbuf(ctx, ctx.F32[1], &UniformDefinitions::F32, sizeof(f32), binding, offset,
+                       ctx.load_const_func_f32);
     } else {
         const Id vector{GetCbufU32x4(ctx, binding, offset)};
         return ctx.OpBitcast(ctx.F32[1], GetCbufElement(ctx, vector, offset, 0u));
@@ -285,8 +290,8 @@ Id EmitGetCbufF32(EmitContext& ctx, const IR::Value& binding, const IR::Value& o
 
 Id EmitGetCbufU32x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset) {
     if (ctx.profile.support_descriptor_aliasing) {
-        return GetCbuf(ctx, ctx.U32[2], &UniformDefinitions::U32x2, sizeof(u32[2]), binding,
-                       offset);
+        return GetCbuf(ctx, ctx.U32[2], &UniformDefinitions::U32x2, sizeof(u32[2]), binding, offset,
+                       ctx.load_const_func_u32x2);
     } else {
         const Id vector{GetCbufU32x4(ctx, binding, offset)};
         return ctx.OpCompositeConstruct(ctx.U32[2], GetCbufElement(ctx, vector, offset, 0u),
diff --git a/src/shader_recompiler/backend/spirv/emit_spirv_instructions.h b/src/shader_recompiler/backend/spirv/emit_spirv_instructions.h
index 887112deb..f263b41b0 100644
--- a/src/shader_recompiler/backend/spirv/emit_spirv_instructions.h
+++ b/src/shader_recompiler/backend/spirv/emit_spirv_instructions.h
@@ -335,6 +335,7 @@ Id EmitSharedAtomicOr32(EmitContext& ctx, Id pointer_offset, Id value);
 Id EmitSharedAtomicXor32(EmitContext& ctx, Id pointer_offset, Id value);
 Id EmitSharedAtomicExchange32(EmitContext& ctx, Id pointer_offset, Id value);
 Id EmitSharedAtomicExchange64(EmitContext& ctx, Id pointer_offset, Id value);
+Id EmitSharedAtomicExchange32x2(EmitContext& ctx, Id pointer_offset, Id value);
 Id EmitStorageAtomicIAdd32(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
                            Id value);
 Id EmitStorageAtomicSMin32(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
@@ -375,6 +376,24 @@ Id EmitStorageAtomicXor64(EmitContext& ctx, const IR::Value& binding, const IR::
                           Id value);
 Id EmitStorageAtomicExchange64(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
                                Id value);
+Id EmitStorageAtomicIAdd32x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
+                             Id value);
+Id EmitStorageAtomicSMin32x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
+                             Id value);
+Id EmitStorageAtomicUMin32x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
+                             Id value);
+Id EmitStorageAtomicSMax32x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
+                             Id value);
+Id EmitStorageAtomicUMax32x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
+                             Id value);
+Id EmitStorageAtomicAnd32x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
+                            Id value);
+Id EmitStorageAtomicOr32x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
+                           Id value);
+Id EmitStorageAtomicXor32x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
+                            Id value);
+Id EmitStorageAtomicExchange32x2(EmitContext& ctx, const IR::Value& binding,
+                                 const IR::Value& offset, Id value);
 Id EmitStorageAtomicAddF32(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
                            Id value);
 Id EmitStorageAtomicAddF16x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
@@ -411,6 +430,17 @@ Id EmitGlobalAtomicAnd64(EmitContext& ctx);
 Id EmitGlobalAtomicOr64(EmitContext& ctx);
 Id EmitGlobalAtomicXor64(EmitContext& ctx);
 Id EmitGlobalAtomicExchange64(EmitContext& ctx);
+Id EmitGlobalAtomicIAdd32x2(EmitContext& ctx);
+Id EmitGlobalAtomicSMin32x2(EmitContext& ctx);
+Id EmitGlobalAtomicUMin32x2(EmitContext& ctx);
+Id EmitGlobalAtomicSMax32x2(EmitContext& ctx);
+Id EmitGlobalAtomicUMax32x2(EmitContext& ctx);
+Id EmitGlobalAtomicInc32x2(EmitContext& ctx);
+Id EmitGlobalAtomicDec32x2(EmitContext& ctx);
+Id EmitGlobalAtomicAnd32x2(EmitContext& ctx);
+Id EmitGlobalAtomicOr32x2(EmitContext& ctx);
+Id EmitGlobalAtomicXor32x2(EmitContext& ctx);
+Id EmitGlobalAtomicExchange32x2(EmitContext& ctx);
 Id EmitGlobalAtomicAddF32(EmitContext& ctx);
 Id EmitGlobalAtomicAddF16x2(EmitContext& ctx);
 Id EmitGlobalAtomicAddF32x2(EmitContext& ctx);
diff --git a/src/shader_recompiler/backend/spirv/spirv_emit_context.cpp b/src/shader_recompiler/backend/spirv/spirv_emit_context.cpp
index cd90c084a..aa5b6c9b7 100644
--- a/src/shader_recompiler/backend/spirv/spirv_emit_context.cpp
+++ b/src/shader_recompiler/backend/spirv/spirv_emit_context.cpp
@@ -464,6 +464,7 @@ EmitContext::EmitContext(const Profile& profile_, const RuntimeInfo& runtime_inf
     DefineSharedMemory(program);
     DefineSharedMemoryFunctions(program);
     DefineConstantBuffers(program.info, uniform_binding);
+    DefineConstantBufferIndirectFunctions(program.info);
     DefineStorageBuffers(program.info, storage_binding);
     DefineTextureBuffers(program.info, texture_binding);
     DefineImageBuffers(program.info, image_binding);
@@ -993,7 +994,7 @@ void EmitContext::DefineConstantBuffers(const Info& info, u32& binding) {
         }
         return;
     }
-    IR::Type types{info.used_constant_buffer_types};
+    IR::Type types{info.used_constant_buffer_types | info.used_indirect_cbuf_types};
     if (True(types & IR::Type::U8)) {
         if (profile.support_int8) {
             DefineConstBuffers(*this, info, &UniformDefinitions::U8, binding, U8, 'u', sizeof(u8));
@@ -1027,6 +1028,62 @@ void EmitContext::DefineConstantBuffers(const Info& info, u32& binding) {
     binding += static_cast<u32>(info.constant_buffer_descriptors.size());
 }
 
+void EmitContext::DefineConstantBufferIndirectFunctions(const Info& info) {
+    if (!info.uses_cbuf_indirect) {
+        return;
+    }
+    const auto make_accessor{[&](Id buffer_type, Id UniformDefinitions::*member_ptr) {
+        const Id func_type{TypeFunction(buffer_type, U32[1], U32[1])};
+        const Id func{OpFunction(buffer_type, spv::FunctionControlMask::MaskNone, func_type)};
+        const Id binding{OpFunctionParameter(U32[1])};
+        const Id offset{OpFunctionParameter(U32[1])};
+
+        AddLabel();
+
+        const Id merge_label{OpLabel()};
+        const Id uniform_type{uniform_types.*member_ptr};
+
+        std::array<Id, Info::MAX_CBUFS> buf_labels;
+        std::array<Sirit::Literal, Info::MAX_CBUFS> buf_literals;
+        for (u32 i = 0; i < Info::MAX_CBUFS; i++) {
+            buf_labels[i] = OpLabel();
+            buf_literals[i] = Sirit::Literal{i};
+        }
+        OpSelectionMerge(merge_label, spv::SelectionControlMask::MaskNone);
+        OpSwitch(binding, buf_labels[0], buf_literals, buf_labels);
+        for (u32 i = 0; i < Info::MAX_CBUFS; i++) {
+            AddLabel(buf_labels[i]);
+            const Id cbuf{cbufs[i].*member_ptr};
+            const Id access_chain{OpAccessChain(uniform_type, cbuf, u32_zero_value, offset)};
+            const Id result{OpLoad(buffer_type, access_chain)};
+            OpReturnValue(result);
+        }
+        AddLabel(merge_label);
+        OpUnreachable();
+        OpFunctionEnd();
+        return func;
+    }};
+    IR::Type types{info.used_indirect_cbuf_types};
+    if (True(types & IR::Type::U8)) {
+        load_const_func_u8 = make_accessor(U8, &UniformDefinitions::U8);
+    }
+    if (True(types & IR::Type::U16)) {
+        load_const_func_u16 = make_accessor(U16, &UniformDefinitions::U16);
+    }
+    if (True(types & IR::Type::F32)) {
+        load_const_func_f32 = make_accessor(F32[1], &UniformDefinitions::F32);
+    }
+    if (True(types & IR::Type::U32)) {
+        load_const_func_u32 = make_accessor(U32[1], &UniformDefinitions::U32);
+    }
+    if (True(types & IR::Type::U32x2)) {
+        load_const_func_u32x2 = make_accessor(U32[2], &UniformDefinitions::U32x2);
+    }
+    if (True(types & IR::Type::U32x4)) {
+        load_const_func_u32x4 = make_accessor(U32[4], &UniformDefinitions::U32x4);
+    }
+}
+
 void EmitContext::DefineStorageBuffers(const Info& info, u32& binding) {
     if (info.storage_buffers_descriptors.empty()) {
         return;
diff --git a/src/shader_recompiler/backend/spirv/spirv_emit_context.h b/src/shader_recompiler/backend/spirv/spirv_emit_context.h
index f87138f7e..906a1dc2c 100644
--- a/src/shader_recompiler/backend/spirv/spirv_emit_context.h
+++ b/src/shader_recompiler/backend/spirv/spirv_emit_context.h
@@ -294,6 +294,13 @@ public:
 
     std::vector<Id> interfaces;
 
+    Id load_const_func_u8{};
+    Id load_const_func_u16{};
+    Id load_const_func_u32{};
+    Id load_const_func_f32{};
+    Id load_const_func_u32x2{};
+    Id load_const_func_u32x4{};
+
 private:
     void DefineCommonTypes(const Info& info);
     void DefineCommonConstants();
@@ -302,6 +309,7 @@ private:
     void DefineSharedMemory(const IR::Program& program);
     void DefineSharedMemoryFunctions(const IR::Program& program);
     void DefineConstantBuffers(const Info& info, u32& binding);
+    void DefineConstantBufferIndirectFunctions(const Info& info);
     void DefineStorageBuffers(const Info& info, u32& binding);
     void DefineTextureBuffers(const Info& info, u32& binding);
     void DefineImageBuffers(const Info& info, u32& binding);
diff --git a/src/shader_recompiler/frontend/ir/microinstruction.cpp b/src/shader_recompiler/frontend/ir/microinstruction.cpp
index 97e2bf6af..631446cf7 100644
--- a/src/shader_recompiler/frontend/ir/microinstruction.cpp
+++ b/src/shader_recompiler/frontend/ir/microinstruction.cpp
@@ -118,6 +118,7 @@ bool Inst::MayHaveSideEffects() const noexcept {
     case Opcode::SharedAtomicXor32:
     case Opcode::SharedAtomicExchange32:
     case Opcode::SharedAtomicExchange64:
+    case Opcode::SharedAtomicExchange32x2:
     case Opcode::GlobalAtomicIAdd32:
     case Opcode::GlobalAtomicSMin32:
     case Opcode::GlobalAtomicUMin32:
@@ -138,6 +139,15 @@ bool Inst::MayHaveSideEffects() const noexcept {
     case Opcode::GlobalAtomicOr64:
     case Opcode::GlobalAtomicXor64:
     case Opcode::GlobalAtomicExchange64:
+    case Opcode::GlobalAtomicIAdd32x2:
+    case Opcode::GlobalAtomicSMin32x2:
+    case Opcode::GlobalAtomicUMin32x2:
+    case Opcode::GlobalAtomicSMax32x2:
+    case Opcode::GlobalAtomicUMax32x2:
+    case Opcode::GlobalAtomicAnd32x2:
+    case Opcode::GlobalAtomicOr32x2:
+    case Opcode::GlobalAtomicXor32x2:
+    case Opcode::GlobalAtomicExchange32x2:
     case Opcode::GlobalAtomicAddF32:
     case Opcode::GlobalAtomicAddF16x2:
     case Opcode::GlobalAtomicAddF32x2:
@@ -165,6 +175,15 @@ bool Inst::MayHaveSideEffects() const noexcept {
     case Opcode::StorageAtomicOr64:
     case Opcode::StorageAtomicXor64:
     case Opcode::StorageAtomicExchange64:
+    case Opcode::StorageAtomicIAdd32x2:
+    case Opcode::StorageAtomicSMin32x2:
+    case Opcode::StorageAtomicUMin32x2:
+    case Opcode::StorageAtomicSMax32x2:
+    case Opcode::StorageAtomicUMax32x2:
+    case Opcode::StorageAtomicAnd32x2:
+    case Opcode::StorageAtomicOr32x2:
+    case Opcode::StorageAtomicXor32x2:
+    case Opcode::StorageAtomicExchange32x2:
     case Opcode::StorageAtomicAddF32:
     case Opcode::StorageAtomicAddF16x2:
     case Opcode::StorageAtomicAddF32x2:
diff --git a/src/shader_recompiler/frontend/ir/opcodes.inc b/src/shader_recompiler/frontend/ir/opcodes.inc
index b94ce7406..efb6bfac3 100644
--- a/src/shader_recompiler/frontend/ir/opcodes.inc
+++ b/src/shader_recompiler/frontend/ir/opcodes.inc
@@ -341,6 +341,7 @@ OPCODE(SharedAtomicOr32,                                    U32,            U32,
 OPCODE(SharedAtomicXor32,                                   U32,            U32,            U32,                                                            )
 OPCODE(SharedAtomicExchange32,                              U32,            U32,            U32,                                                            )
 OPCODE(SharedAtomicExchange64,                              U64,            U32,            U64,                                                            )
+OPCODE(SharedAtomicExchange32x2,                            U32x2,          U32,            U32x2,                                                          )
 
 OPCODE(GlobalAtomicIAdd32,                                  U32,            U64,            U32,                                                            )
 OPCODE(GlobalAtomicSMin32,                                  U32,            U64,            U32,                                                            )
@@ -362,6 +363,15 @@ OPCODE(GlobalAtomicAnd64,                                   U64,            U64,
 OPCODE(GlobalAtomicOr64,                                    U64,            U64,            U64,                                                            )
 OPCODE(GlobalAtomicXor64,                                   U64,            U64,            U64,                                                            )
 OPCODE(GlobalAtomicExchange64,                              U64,            U64,            U64,                                                            )
+OPCODE(GlobalAtomicIAdd32x2,                                U32x2,          U32x2,          U32x2,                                                          )
+OPCODE(GlobalAtomicSMin32x2,                                U32x2,          U32x2,          U32x2,                                                          )
+OPCODE(GlobalAtomicUMin32x2,                                U32x2,          U32x2,          U32x2,                                                          )
+OPCODE(GlobalAtomicSMax32x2,                                U32x2,          U32x2,          U32x2,                                                          )
+OPCODE(GlobalAtomicUMax32x2,                                U32x2,          U32x2,          U32x2,                                                          )
+OPCODE(GlobalAtomicAnd32x2,                                 U32x2,          U32x2,          U32x2,                                                          )
+OPCODE(GlobalAtomicOr32x2,                                  U32x2,          U32x2,          U32x2,                                                          )
+OPCODE(GlobalAtomicXor32x2,                                 U32x2,          U32x2,          U32x2,                                                          )
+OPCODE(GlobalAtomicExchange32x2,                            U32x2,          U32x2,          U32x2,                                                          )
 OPCODE(GlobalAtomicAddF32,                                  F32,            U64,            F32,                                                            )
 OPCODE(GlobalAtomicAddF16x2,                                U32,            U64,            F16x2,                                                          )
 OPCODE(GlobalAtomicAddF32x2,                                U32,            U64,            F32x2,                                                          )
@@ -390,6 +400,15 @@ OPCODE(StorageAtomicAnd64,                                  U64,            U32,
 OPCODE(StorageAtomicOr64,                                   U64,            U32,            U32,            U64,                                            )
 OPCODE(StorageAtomicXor64,                                  U64,            U32,            U32,            U64,                                            )
 OPCODE(StorageAtomicExchange64,                             U64,            U32,            U32,            U64,                                            )
+OPCODE(StorageAtomicIAdd32x2,                               U32x2,          U32,            U32,            U32x2,                                          )
+OPCODE(StorageAtomicSMin32x2,                               U32x2,          U32,            U32,            U32x2,                                          )
+OPCODE(StorageAtomicUMin32x2,                               U32x2,          U32,            U32,            U32x2,                                          )
+OPCODE(StorageAtomicSMax32x2,                               U32x2,          U32,            U32,            U32x2,                                          )
+OPCODE(StorageAtomicUMax32x2,                               U32x2,          U32,            U32,            U32x2,                                          )
+OPCODE(StorageAtomicAnd32x2,                                U32x2,          U32,            U32,            U32x2,                                          )
+OPCODE(StorageAtomicOr32x2,                                 U32x2,          U32,            U32,            U32x2,                                          )
+OPCODE(StorageAtomicXor32x2,                                U32x2,          U32,            U32,            U32x2,                                          )
+OPCODE(StorageAtomicExchange32x2,                           U32x2,          U32,            U32,            U32x2,                                          )
 OPCODE(StorageAtomicAddF32,                                 F32,            U32,            U32,            F32,                                            )
 OPCODE(StorageAtomicAddF16x2,                               U32,            U32,            U32,            F16x2,                                          )
 OPCODE(StorageAtomicAddF32x2,                               U32,            U32,            U32,            F32x2,                                          )
diff --git a/src/shader_recompiler/frontend/maxwell/translate/impl/load_constant.cpp b/src/shader_recompiler/frontend/maxwell/translate/impl/load_constant.cpp
index 2300088e3..8007a4d46 100644
--- a/src/shader_recompiler/frontend/maxwell/translate/impl/load_constant.cpp
+++ b/src/shader_recompiler/frontend/maxwell/translate/impl/load_constant.cpp
@@ -11,10 +11,20 @@ namespace Shader::Maxwell {
 using namespace LDC;
 namespace {
 std::pair<IR::U32, IR::U32> Slot(IR::IREmitter& ir, Mode mode, const IR::U32& imm_index,
-                                 const IR::U32& reg, const IR::U32& imm) {
+                                 const IR::U32& reg, const IR::U32& imm_offset) {
     switch (mode) {
     case Mode::Default:
-        return {imm_index, ir.IAdd(reg, imm)};
+        return {imm_index, ir.IAdd(reg, imm_offset)};
+    case Mode::IS: {
+        // Segmented addressing mode
+        // Ra+imm_offset points into a flat mapping of const buffer
+        // address space
+        const IR::U32 address{ir.IAdd(reg, imm_offset)};
+        const IR::U32 index{ir.BitFieldExtract(address, ir.Imm32(16), ir.Imm32(16))};
+        const IR::U32 offset{ir.BitFieldExtract(address, ir.Imm32(0), ir.Imm32(16))};
+
+        return {ir.IAdd(index, imm_index), offset};
+    }
     default:
         break;
     }
diff --git a/src/shader_recompiler/frontend/maxwell/translate/impl/logic_operation_three_input.cpp b/src/shader_recompiler/frontend/maxwell/translate/impl/logic_operation_three_input.cpp
index e0fe47912..f3c7ceb57 100644
--- a/src/shader_recompiler/frontend/maxwell/translate/impl/logic_operation_three_input.cpp
+++ b/src/shader_recompiler/frontend/maxwell/translate/impl/logic_operation_three_input.cpp
@@ -13,59 +13,535 @@ namespace {
 // Emulate GPU's LOP3.LUT (three-input logic op with 8-bit truth table)
 IR::U32 ApplyLUT(IR::IREmitter& ir, const IR::U32& a, const IR::U32& b, const IR::U32& c,
                  u64 ttbl) {
-    IR::U32 r{ir.Imm32(0)};
-    const IR::U32 not_a{ir.BitwiseNot(a)};
-    const IR::U32 not_b{ir.BitwiseNot(b)};
-    const IR::U32 not_c{ir.BitwiseNot(c)};
-    if (ttbl & 0x01) {
-        // r |= ~a & ~b & ~c;
-        const auto lhs{ir.BitwiseAnd(not_a, not_b)};
-        const auto rhs{ir.BitwiseAnd(lhs, not_c)};
-        r = ir.BitwiseOr(r, rhs);
+    switch (ttbl) {
+        // generated code, do not edit manually
+    case 0:
+        return ir.Imm32(0);
+    case 1:
+        return ir.BitwiseNot(ir.BitwiseOr(a, ir.BitwiseOr(b, c)));
+    case 2:
+        return ir.BitwiseAnd(c, ir.BitwiseNot(ir.BitwiseOr(a, b)));
+    case 3:
+        return ir.BitwiseNot(ir.BitwiseOr(a, b));
+    case 4:
+        return ir.BitwiseAnd(b, ir.BitwiseNot(ir.BitwiseOr(a, c)));
+    case 5:
+        return ir.BitwiseNot(ir.BitwiseOr(a, c));
+    case 6:
+        return ir.BitwiseAnd(ir.BitwiseNot(a), ir.BitwiseXor(b, c));
+    case 7:
+        return ir.BitwiseNot(ir.BitwiseOr(a, ir.BitwiseAnd(b, c)));
+    case 8:
+        return ir.BitwiseAnd(ir.BitwiseAnd(b, c), ir.BitwiseNot(a));
+    case 9:
+        return ir.BitwiseNot(ir.BitwiseOr(a, ir.BitwiseXor(b, c)));
+    case 10:
+        return ir.BitwiseAnd(c, ir.BitwiseNot(a));
+    case 11:
+        return ir.BitwiseAnd(ir.BitwiseNot(a), ir.BitwiseOr(c, ir.BitwiseNot(b)));
+    case 12:
+        return ir.BitwiseAnd(b, ir.BitwiseNot(a));
+    case 13:
+        return ir.BitwiseAnd(ir.BitwiseNot(a), ir.BitwiseOr(b, ir.BitwiseNot(c)));
+    case 14:
+        return ir.BitwiseAnd(ir.BitwiseNot(a), ir.BitwiseOr(b, c));
+    case 15:
+        return ir.BitwiseNot(a);
+    case 16:
+        return ir.BitwiseAnd(a, ir.BitwiseNot(ir.BitwiseOr(b, c)));
+    case 17:
+        return ir.BitwiseNot(ir.BitwiseOr(b, c));
+    case 18:
+        return ir.BitwiseAnd(ir.BitwiseNot(b), ir.BitwiseXor(a, c));
+    case 19:
+        return ir.BitwiseNot(ir.BitwiseOr(b, ir.BitwiseAnd(a, c)));
+    case 20:
+        return ir.BitwiseAnd(ir.BitwiseNot(c), ir.BitwiseXor(a, b));
+    case 21:
+        return ir.BitwiseNot(ir.BitwiseOr(c, ir.BitwiseAnd(a, b)));
+    case 22:
+        return ir.BitwiseXor(ir.BitwiseOr(a, b), ir.BitwiseOr(c, ir.BitwiseAnd(a, b)));
+    case 23:
+        return ir.BitwiseXor(ir.BitwiseAnd(ir.BitwiseXor(a, b), ir.BitwiseXor(a, c)),
+                             ir.BitwiseNot(a));
+    case 24:
+        return ir.BitwiseAnd(ir.BitwiseXor(a, b), ir.BitwiseXor(a, c));
+    case 25:
+        return ir.BitwiseNot(ir.BitwiseOr(ir.BitwiseAnd(a, b), ir.BitwiseXor(b, c)));
+    case 26:
+        return ir.BitwiseAnd(ir.BitwiseOr(c, ir.BitwiseNot(b)), ir.BitwiseXor(a, c));
+    case 27:
+        return ir.BitwiseXor(ir.BitwiseOr(a, ir.BitwiseNot(c)), ir.BitwiseOr(b, c));
+    case 28:
+        return ir.BitwiseAnd(ir.BitwiseOr(b, ir.BitwiseNot(c)), ir.BitwiseXor(a, b));
+    case 29:
+        return ir.BitwiseXor(ir.BitwiseOr(a, ir.BitwiseNot(b)), ir.BitwiseOr(b, c));
+    case 30:
+        return ir.BitwiseXor(a, ir.BitwiseOr(b, c));
+    case 31:
+        return ir.BitwiseNot(ir.BitwiseAnd(a, ir.BitwiseOr(b, c)));
+    case 32:
+        return ir.BitwiseAnd(ir.BitwiseAnd(a, c), ir.BitwiseNot(b));
+    case 33:
+        return ir.BitwiseNot(ir.BitwiseOr(b, ir.BitwiseXor(a, c)));
+    case 34:
+        return ir.BitwiseAnd(c, ir.BitwiseNot(b));
+    case 35:
+        return ir.BitwiseAnd(ir.BitwiseNot(b), ir.BitwiseOr(c, ir.BitwiseNot(a)));
+    case 36:
+        return ir.BitwiseAnd(ir.BitwiseXor(a, b), ir.BitwiseXor(b, c));
+    case 37:
+        return ir.BitwiseNot(ir.BitwiseOr(ir.BitwiseAnd(a, b), ir.BitwiseXor(a, c)));
+    case 38:
+        return ir.BitwiseAnd(ir.BitwiseOr(c, ir.BitwiseNot(a)), ir.BitwiseXor(b, c));
+    case 39:
+        return ir.BitwiseXor(ir.BitwiseOr(a, c), ir.BitwiseOr(b, ir.BitwiseNot(c)));
+    case 40:
+        return ir.BitwiseAnd(c, ir.BitwiseXor(a, b));
+    case 41:
+        return ir.BitwiseXor(ir.BitwiseOr(a, b),
+                             ir.BitwiseOr(ir.BitwiseAnd(a, b), ir.BitwiseNot(c)));
+    case 42:
+        return ir.BitwiseAnd(c, ir.BitwiseNot(ir.BitwiseAnd(a, b)));
+    case 43:
+        return ir.BitwiseXor(ir.BitwiseOr(a, ir.BitwiseNot(c)),
+                             ir.BitwiseOr(b, ir.BitwiseXor(a, c)));
+    case 44:
+        return ir.BitwiseAnd(ir.BitwiseOr(b, c), ir.BitwiseXor(a, b));
+    case 45:
+        return ir.BitwiseXor(a, ir.BitwiseOr(b, ir.BitwiseNot(c)));
+    case 46:
+        return ir.BitwiseXor(ir.BitwiseAnd(a, b), ir.BitwiseOr(b, c));
+    case 47:
+        return ir.BitwiseOr(ir.BitwiseAnd(c, ir.BitwiseNot(b)), ir.BitwiseNot(a));
+    case 48:
+        return ir.BitwiseAnd(a, ir.BitwiseNot(b));
+    case 49:
+        return ir.BitwiseAnd(ir.BitwiseNot(b), ir.BitwiseOr(a, ir.BitwiseNot(c)));
+    case 50:
+        return ir.BitwiseAnd(ir.BitwiseNot(b), ir.BitwiseOr(a, c));
+    case 51:
+        return ir.BitwiseNot(b);
+    case 52:
+        return ir.BitwiseAnd(ir.BitwiseOr(a, ir.BitwiseNot(c)), ir.BitwiseXor(a, b));
+    case 53:
+        return ir.BitwiseXor(ir.BitwiseOr(a, c), ir.BitwiseOr(b, ir.BitwiseNot(a)));
+    case 54:
+        return ir.BitwiseXor(b, ir.BitwiseOr(a, c));
+    case 55:
+        return ir.BitwiseNot(ir.BitwiseAnd(b, ir.BitwiseOr(a, c)));
+    case 56:
+        return ir.BitwiseAnd(ir.BitwiseOr(a, c), ir.BitwiseXor(a, b));
+    case 57:
+        return ir.BitwiseXor(b, ir.BitwiseOr(a, ir.BitwiseNot(c)));
+    case 58:
+        return ir.BitwiseXor(ir.BitwiseAnd(a, b), ir.BitwiseOr(a, c));
+    case 59:
+        return ir.BitwiseOr(ir.BitwiseAnd(c, ir.BitwiseNot(a)), ir.BitwiseNot(b));
+    case 60:
+        return ir.BitwiseXor(a, b);
+    case 61:
+        return ir.BitwiseOr(ir.BitwiseNot(ir.BitwiseOr(a, c)), ir.BitwiseXor(a, b));
+    case 62:
+        return ir.BitwiseOr(ir.BitwiseAnd(c, ir.BitwiseNot(a)), ir.BitwiseXor(a, b));
+    case 63:
+        return ir.BitwiseNot(ir.BitwiseAnd(a, b));
+    case 64:
+        return ir.BitwiseAnd(ir.BitwiseAnd(a, b), ir.BitwiseNot(c));
+    case 65:
+        return ir.BitwiseNot(ir.BitwiseOr(c, ir.BitwiseXor(a, b)));
+    case 66:
+        return ir.BitwiseAnd(ir.BitwiseXor(a, c), ir.BitwiseXor(b, c));
+    case 67:
+        return ir.BitwiseNot(ir.BitwiseOr(ir.BitwiseAnd(a, c), ir.BitwiseXor(a, b)));
+    case 68:
+        return ir.BitwiseAnd(b, ir.BitwiseNot(c));
+    case 69:
+        return ir.BitwiseAnd(ir.BitwiseNot(c), ir.BitwiseOr(b, ir.BitwiseNot(a)));
+    case 70:
+        return ir.BitwiseAnd(ir.BitwiseOr(b, ir.BitwiseNot(a)), ir.BitwiseXor(b, c));
+    case 71:
+        return ir.BitwiseXor(ir.BitwiseOr(a, b), ir.BitwiseOr(c, ir.BitwiseNot(b)));
+    case 72:
+        return ir.BitwiseAnd(b, ir.BitwiseXor(a, c));
+    case 73:
+        return ir.BitwiseXor(ir.BitwiseOr(a, c),
+                             ir.BitwiseOr(ir.BitwiseAnd(a, c), ir.BitwiseNot(b)));
+    case 74:
+        return ir.BitwiseAnd(ir.BitwiseOr(b, c), ir.BitwiseXor(a, c));
+    case 75:
+        return ir.BitwiseXor(a, ir.BitwiseOr(c, ir.BitwiseNot(b)));
+    case 76:
+        return ir.BitwiseAnd(b, ir.BitwiseNot(ir.BitwiseAnd(a, c)));
+    case 77:
+        return ir.BitwiseXor(ir.BitwiseOr(a, ir.BitwiseNot(b)),
+                             ir.BitwiseOr(c, ir.BitwiseXor(a, b)));
+    case 78:
+        return ir.BitwiseXor(ir.BitwiseAnd(a, c), ir.BitwiseOr(b, c));
+    case 79:
+        return ir.BitwiseOr(ir.BitwiseAnd(b, ir.BitwiseNot(c)), ir.BitwiseNot(a));
+    case 80:
+        return ir.BitwiseAnd(a, ir.BitwiseNot(c));
+    case 81:
+        return ir.BitwiseAnd(ir.BitwiseNot(c), ir.BitwiseOr(a, ir.BitwiseNot(b)));
+    case 82:
+        return ir.BitwiseAnd(ir.BitwiseOr(a, ir.BitwiseNot(b)), ir.BitwiseXor(a, c));
+    case 83:
+        return ir.BitwiseXor(ir.BitwiseOr(a, b), ir.BitwiseOr(c, ir.BitwiseNot(a)));
+    case 84:
+        return ir.BitwiseAnd(ir.BitwiseNot(c), ir.BitwiseOr(a, b));
+    case 85:
+        return ir.BitwiseNot(c);
+    case 86:
+        return ir.BitwiseXor(c, ir.BitwiseOr(a, b));
+    case 87:
+        return ir.BitwiseNot(ir.BitwiseAnd(c, ir.BitwiseOr(a, b)));
+    case 88:
+        return ir.BitwiseAnd(ir.BitwiseOr(a, b), ir.BitwiseXor(a, c));
+    case 89:
+        return ir.BitwiseXor(c, ir.BitwiseOr(a, ir.BitwiseNot(b)));
+    case 90:
+        return ir.BitwiseXor(a, c);
+    case 91:
+        return ir.BitwiseOr(ir.BitwiseNot(ir.BitwiseOr(a, b)), ir.BitwiseXor(a, c));
+    case 92:
+        return ir.BitwiseXor(ir.BitwiseAnd(a, c), ir.BitwiseOr(a, b));
+    case 93:
+        return ir.BitwiseOr(ir.BitwiseAnd(b, ir.BitwiseNot(a)), ir.BitwiseNot(c));
+    case 94:
+        return ir.BitwiseOr(ir.BitwiseAnd(b, ir.BitwiseNot(a)), ir.BitwiseXor(a, c));
+    case 95:
+        return ir.BitwiseNot(ir.BitwiseAnd(a, c));
+    case 96:
+        return ir.BitwiseAnd(a, ir.BitwiseXor(b, c));
+    case 97:
+        return ir.BitwiseXor(ir.BitwiseOr(b, c),
+                             ir.BitwiseOr(ir.BitwiseAnd(b, c), ir.BitwiseNot(a)));
+    case 98:
+        return ir.BitwiseAnd(ir.BitwiseOr(a, c), ir.BitwiseXor(b, c));
+    case 99:
+        return ir.BitwiseXor(b, ir.BitwiseOr(c, ir.BitwiseNot(a)));
+    case 100:
+        return ir.BitwiseAnd(ir.BitwiseOr(a, b), ir.BitwiseXor(b, c));
+    case 101:
+        return ir.BitwiseXor(c, ir.BitwiseOr(b, ir.BitwiseNot(a)));
+    case 102:
+        return ir.BitwiseXor(b, c);
+    case 103:
+        return ir.BitwiseOr(ir.BitwiseNot(ir.BitwiseOr(a, b)), ir.BitwiseXor(b, c));
+    case 104:
+        return ir.BitwiseAnd(ir.BitwiseOr(a, b), ir.BitwiseXor(c, ir.BitwiseAnd(a, b)));
+    case 105:
+        return ir.BitwiseXor(ir.BitwiseNot(a), ir.BitwiseXor(b, c));
+    case 106:
+        return ir.BitwiseXor(c, ir.BitwiseAnd(a, b));
+    case 107:
+        return ir.BitwiseXor(ir.BitwiseAnd(c, ir.BitwiseOr(a, b)),
+                             ir.BitwiseXor(a, ir.BitwiseNot(b)));
+    case 108:
+        return ir.BitwiseXor(b, ir.BitwiseAnd(a, c));
+    case 109:
+        return ir.BitwiseXor(ir.BitwiseAnd(b, ir.BitwiseOr(a, c)),
+                             ir.BitwiseXor(a, ir.BitwiseNot(c)));
+    case 110:
+        return ir.BitwiseOr(ir.BitwiseAnd(b, ir.BitwiseNot(a)), ir.BitwiseXor(b, c));
+    case 111:
+        return ir.BitwiseOr(ir.BitwiseNot(a), ir.BitwiseXor(b, c));
+    case 112:
+        return ir.BitwiseAnd(a, ir.BitwiseNot(ir.BitwiseAnd(b, c)));
+    case 113:
+        return ir.BitwiseXor(ir.BitwiseOr(b, ir.BitwiseNot(a)),
+                             ir.BitwiseOr(c, ir.BitwiseXor(a, b)));
+    case 114:
+        return ir.BitwiseXor(ir.BitwiseAnd(b, c), ir.BitwiseOr(a, c));
+    case 115:
+        return ir.BitwiseOr(ir.BitwiseAnd(a, ir.BitwiseNot(c)), ir.BitwiseNot(b));
+    case 116:
+        return ir.BitwiseXor(ir.BitwiseAnd(b, c), ir.BitwiseOr(a, b));
+    case 117:
+        return ir.BitwiseOr(ir.BitwiseAnd(a, ir.BitwiseNot(b)), ir.BitwiseNot(c));
+    case 118:
+        return ir.BitwiseOr(ir.BitwiseAnd(a, ir.BitwiseNot(b)), ir.BitwiseXor(b, c));
+    case 119:
+        return ir.BitwiseNot(ir.BitwiseAnd(b, c));
+    case 120:
+        return ir.BitwiseXor(a, ir.BitwiseAnd(b, c));
+    case 121:
+        return ir.BitwiseXor(ir.BitwiseAnd(a, ir.BitwiseOr(b, c)),
+                             ir.BitwiseXor(b, ir.BitwiseNot(c)));
+    case 122:
+        return ir.BitwiseOr(ir.BitwiseAnd(a, ir.BitwiseNot(b)), ir.BitwiseXor(a, c));
+    case 123:
+        return ir.BitwiseOr(ir.BitwiseNot(b), ir.BitwiseXor(a, c));
+    case 124:
+        return ir.BitwiseOr(ir.BitwiseAnd(a, ir.BitwiseNot(c)), ir.BitwiseXor(a, b));
+    case 125:
+        return ir.BitwiseOr(ir.BitwiseNot(c), ir.BitwiseXor(a, b));
+    case 126:
+        return ir.BitwiseOr(ir.BitwiseXor(a, b), ir.BitwiseXor(a, c));
+    case 127:
+        return ir.BitwiseNot(ir.BitwiseAnd(a, ir.BitwiseAnd(b, c)));
+    case 128:
+        return ir.BitwiseAnd(a, ir.BitwiseAnd(b, c));
+    case 129:
+        return ir.BitwiseNot(ir.BitwiseOr(ir.BitwiseXor(a, b), ir.BitwiseXor(a, c)));
+    case 130:
+        return ir.BitwiseAnd(c, ir.BitwiseXor(a, ir.BitwiseNot(b)));
+    case 131:
+        return ir.BitwiseAnd(ir.BitwiseOr(c, ir.BitwiseNot(a)), ir.BitwiseXor(a, ir.BitwiseNot(b)));
+    case 132:
+        return ir.BitwiseAnd(b, ir.BitwiseXor(a, ir.BitwiseNot(c)));
+    case 133:
+        return ir.BitwiseAnd(ir.BitwiseOr(b, ir.BitwiseNot(a)), ir.BitwiseXor(a, ir.BitwiseNot(c)));
+    case 134:
+        return ir.BitwiseAnd(ir.BitwiseOr(b, c), ir.BitwiseXor(a, ir.BitwiseXor(b, c)));
+    case 135:
+        return ir.BitwiseXor(ir.BitwiseAnd(b, c), ir.BitwiseNot(a));
+    case 136:
+        return ir.BitwiseAnd(b, c);
+    case 137:
+        return ir.BitwiseAnd(ir.BitwiseOr(b, ir.BitwiseNot(a)), ir.BitwiseXor(b, ir.BitwiseNot(c)));
+    case 138:
+        return ir.BitwiseAnd(c, ir.BitwiseOr(b, ir.BitwiseNot(a)));
+    case 139:
+        return ir.BitwiseOr(ir.BitwiseAnd(b, c), ir.BitwiseNot(ir.BitwiseOr(a, b)));
+    case 140:
+        return ir.BitwiseAnd(b, ir.BitwiseOr(c, ir.BitwiseNot(a)));
+    case 141:
+        return ir.BitwiseOr(ir.BitwiseAnd(b, c), ir.BitwiseNot(ir.BitwiseOr(a, c)));
+    case 142:
+        return ir.BitwiseXor(a, ir.BitwiseOr(ir.BitwiseXor(a, b), ir.BitwiseXor(a, c)));
+    case 143:
+        return ir.BitwiseOr(ir.BitwiseAnd(b, c), ir.BitwiseNot(a));
+    case 144:
+        return ir.BitwiseAnd(a, ir.BitwiseXor(b, ir.BitwiseNot(c)));
+    case 145:
+        return ir.BitwiseAnd(ir.BitwiseOr(a, ir.BitwiseNot(b)), ir.BitwiseXor(b, ir.BitwiseNot(c)));
+    case 146:
+        return ir.BitwiseAnd(ir.BitwiseOr(a, c), ir.BitwiseXor(a, ir.BitwiseXor(b, c)));
+    case 147:
+        return ir.BitwiseXor(ir.BitwiseAnd(a, c), ir.BitwiseNot(b));
+    case 148:
+        return ir.BitwiseAnd(ir.BitwiseOr(a, b), ir.BitwiseXor(a, ir.BitwiseXor(b, c)));
+    case 149:
+        return ir.BitwiseXor(ir.BitwiseAnd(a, b), ir.BitwiseNot(c));
+    case 150:
+        return ir.BitwiseXor(a, ir.BitwiseXor(b, c));
+    case 151:
+        return ir.BitwiseOr(ir.BitwiseNot(ir.BitwiseOr(a, b)),
+                            ir.BitwiseXor(a, ir.BitwiseXor(b, c)));
+    case 152:
+        return ir.BitwiseAnd(ir.BitwiseOr(a, b), ir.BitwiseXor(b, ir.BitwiseNot(c)));
+    case 153:
+        return ir.BitwiseXor(b, ir.BitwiseNot(c));
+    case 154:
+        return ir.BitwiseXor(c, ir.BitwiseAnd(a, ir.BitwiseNot(b)));
+    case 155:
+        return ir.BitwiseNot(ir.BitwiseAnd(ir.BitwiseOr(a, b), ir.BitwiseXor(b, c)));
+    case 156:
+        return ir.BitwiseXor(b, ir.BitwiseAnd(a, ir.BitwiseNot(c)));
+    case 157:
+        return ir.BitwiseNot(ir.BitwiseAnd(ir.BitwiseOr(a, c), ir.BitwiseXor(b, c)));
+    case 158:
+        return ir.BitwiseOr(ir.BitwiseAnd(b, c), ir.BitwiseXor(a, ir.BitwiseOr(b, c)));
+    case 159:
+        return ir.BitwiseNot(ir.BitwiseAnd(a, ir.BitwiseXor(b, c)));
+    case 160:
+        return ir.BitwiseAnd(a, c);
+    case 161:
+        return ir.BitwiseAnd(ir.BitwiseOr(a, ir.BitwiseNot(b)), ir.BitwiseXor(a, ir.BitwiseNot(c)));
+    case 162:
+        return ir.BitwiseAnd(c, ir.BitwiseOr(a, ir.BitwiseNot(b)));
+    case 163:
+        return ir.BitwiseOr(ir.BitwiseAnd(a, c), ir.BitwiseNot(ir.BitwiseOr(a, b)));
+    case 164:
+        return ir.BitwiseAnd(ir.BitwiseOr(a, b), ir.BitwiseXor(a, ir.BitwiseNot(c)));
+    case 165:
+        return ir.BitwiseXor(a, ir.BitwiseNot(c));
+    case 166:
+        return ir.BitwiseXor(c, ir.BitwiseAnd(b, ir.BitwiseNot(a)));
+    case 167:
+        return ir.BitwiseNot(ir.BitwiseAnd(ir.BitwiseOr(a, b), ir.BitwiseXor(a, c)));
+    case 168:
+        return ir.BitwiseAnd(c, ir.BitwiseOr(a, b));
+    case 169:
+        return ir.BitwiseXor(ir.BitwiseNot(c), ir.BitwiseOr(a, b));
+    case 170:
+        return c;
+    case 171:
+        return ir.BitwiseOr(c, ir.BitwiseNot(ir.BitwiseOr(a, b)));
+    case 172:
+        return ir.BitwiseAnd(ir.BitwiseOr(a, b), ir.BitwiseOr(c, ir.BitwiseNot(a)));
+    case 173:
+        return ir.BitwiseOr(ir.BitwiseAnd(b, c), ir.BitwiseXor(a, ir.BitwiseNot(c)));
+    case 174:
+        return ir.BitwiseOr(c, ir.BitwiseAnd(b, ir.BitwiseNot(a)));
+    case 175:
+        return ir.BitwiseOr(c, ir.BitwiseNot(a));
+    case 176:
+        return ir.BitwiseAnd(a, ir.BitwiseOr(c, ir.BitwiseNot(b)));
+    case 177:
+        return ir.BitwiseOr(ir.BitwiseAnd(a, c), ir.BitwiseNot(ir.BitwiseOr(b, c)));
+    case 178:
+        return ir.BitwiseXor(b, ir.BitwiseOr(ir.BitwiseXor(a, b), ir.BitwiseXor(a, c)));
+    case 179:
+        return ir.BitwiseOr(ir.BitwiseAnd(a, c), ir.BitwiseNot(b));
+    case 180:
+        return ir.BitwiseXor(a, ir.BitwiseAnd(b, ir.BitwiseNot(c)));
+    case 181:
+        return ir.BitwiseNot(ir.BitwiseAnd(ir.BitwiseOr(b, c), ir.BitwiseXor(a, c)));
+    case 182:
+        return ir.BitwiseOr(ir.BitwiseAnd(a, c), ir.BitwiseXor(b, ir.BitwiseOr(a, c)));
+    case 183:
+        return ir.BitwiseNot(ir.BitwiseAnd(b, ir.BitwiseXor(a, c)));
+    case 184:
+        return ir.BitwiseAnd(ir.BitwiseOr(a, b), ir.BitwiseOr(c, ir.BitwiseNot(b)));
+    case 185:
+        return ir.BitwiseOr(ir.BitwiseAnd(a, c), ir.BitwiseXor(b, ir.BitwiseNot(c)));
+    case 186:
+        return ir.BitwiseOr(c, ir.BitwiseAnd(a, ir.BitwiseNot(b)));
+    case 187:
+        return ir.BitwiseOr(c, ir.BitwiseNot(b));
+    case 188:
+        return ir.BitwiseOr(ir.BitwiseAnd(a, c), ir.BitwiseXor(a, b));
+    case 189:
+        return ir.BitwiseOr(ir.BitwiseXor(a, b), ir.BitwiseXor(a, ir.BitwiseNot(c)));
+    case 190:
+        return ir.BitwiseOr(c, ir.BitwiseXor(a, b));
+    case 191:
+        return ir.BitwiseOr(c, ir.BitwiseNot(ir.BitwiseAnd(a, b)));
+    case 192:
+        return ir.BitwiseAnd(a, b);
+    case 193:
+        return ir.BitwiseAnd(ir.BitwiseOr(a, ir.BitwiseNot(c)), ir.BitwiseXor(a, ir.BitwiseNot(b)));
+    case 194:
+        return ir.BitwiseAnd(ir.BitwiseOr(a, c), ir.BitwiseXor(a, ir.BitwiseNot(b)));
+    case 195:
+        return ir.BitwiseXor(a, ir.BitwiseNot(b));
+    case 196:
+        return ir.BitwiseAnd(b, ir.BitwiseOr(a, ir.BitwiseNot(c)));
+    case 197:
+        return ir.BitwiseOr(ir.BitwiseAnd(a, b), ir.BitwiseNot(ir.BitwiseOr(a, c)));
+    case 198:
+        return ir.BitwiseXor(b, ir.BitwiseAnd(c, ir.BitwiseNot(a)));
+    case 199:
+        return ir.BitwiseNot(ir.BitwiseAnd(ir.BitwiseOr(a, c), ir.BitwiseXor(a, b)));
+    case 200:
+        return ir.BitwiseAnd(b, ir.BitwiseOr(a, c));
+    case 201:
+        return ir.BitwiseXor(ir.BitwiseNot(b), ir.BitwiseOr(a, c));
+    case 202:
+        return ir.BitwiseAnd(ir.BitwiseOr(a, c), ir.BitwiseOr(b, ir.BitwiseNot(a)));
+    case 203:
+        return ir.BitwiseOr(ir.BitwiseAnd(b, c), ir.BitwiseXor(a, ir.BitwiseNot(b)));
+    case 204:
+        return b;
+    case 205:
+        return ir.BitwiseOr(b, ir.BitwiseNot(ir.BitwiseOr(a, c)));
+    case 206:
+        return ir.BitwiseOr(b, ir.BitwiseAnd(c, ir.BitwiseNot(a)));
+    case 207:
+        return ir.BitwiseOr(b, ir.BitwiseNot(a));
+    case 208:
+        return ir.BitwiseAnd(a, ir.BitwiseOr(b, ir.BitwiseNot(c)));
+    case 209:
+        return ir.BitwiseOr(ir.BitwiseAnd(a, b), ir.BitwiseNot(ir.BitwiseOr(b, c)));
+    case 210:
+        return ir.BitwiseXor(a, ir.BitwiseAnd(c, ir.BitwiseNot(b)));
+    case 211:
+        return ir.BitwiseNot(ir.BitwiseAnd(ir.BitwiseOr(b, c), ir.BitwiseXor(a, b)));
+    case 212:
+        return ir.BitwiseXor(c, ir.BitwiseOr(ir.BitwiseXor(a, b), ir.BitwiseXor(a, c)));
+    case 213:
+        return ir.BitwiseOr(ir.BitwiseAnd(a, b), ir.BitwiseNot(c));
+    case 214:
+        return ir.BitwiseOr(ir.BitwiseAnd(a, b), ir.BitwiseXor(c, ir.BitwiseOr(a, b)));
+    case 215:
+        return ir.BitwiseNot(ir.BitwiseAnd(c, ir.BitwiseXor(a, b)));
+    case 216:
+        return ir.BitwiseAnd(ir.BitwiseOr(a, c), ir.BitwiseOr(b, ir.BitwiseNot(c)));
+    case 217:
+        return ir.BitwiseOr(ir.BitwiseAnd(a, b), ir.BitwiseXor(b, ir.BitwiseNot(c)));
+    case 218:
+        return ir.BitwiseOr(ir.BitwiseAnd(a, b), ir.BitwiseXor(a, c));
+    case 219:
+        return ir.BitwiseOr(ir.BitwiseXor(a, c), ir.BitwiseXor(a, ir.BitwiseNot(b)));
+    case 220:
+        return ir.BitwiseOr(b, ir.BitwiseAnd(a, ir.BitwiseNot(c)));
+    case 221:
+        return ir.BitwiseOr(b, ir.BitwiseNot(c));
+    case 222:
+        return ir.BitwiseOr(b, ir.BitwiseXor(a, c));
+    case 223:
+        return ir.BitwiseOr(b, ir.BitwiseNot(ir.BitwiseAnd(a, c)));
+    case 224:
+        return ir.BitwiseAnd(a, ir.BitwiseOr(b, c));
+    case 225:
+        return ir.BitwiseXor(ir.BitwiseNot(a), ir.BitwiseOr(b, c));
+    case 226:
+        return ir.BitwiseAnd(ir.BitwiseOr(a, ir.BitwiseNot(b)), ir.BitwiseOr(b, c));
+    case 227:
+        return ir.BitwiseOr(ir.BitwiseAnd(a, c), ir.BitwiseXor(a, ir.BitwiseNot(b)));
+    case 228:
+        return ir.BitwiseAnd(ir.BitwiseOr(a, ir.BitwiseNot(c)), ir.BitwiseOr(b, c));
+    case 229:
+        return ir.BitwiseOr(ir.BitwiseAnd(a, b), ir.BitwiseXor(a, ir.BitwiseNot(c)));
+    case 230:
+        return ir.BitwiseOr(ir.BitwiseAnd(a, b), ir.BitwiseXor(b, c));
+    case 231:
+        return ir.BitwiseOr(ir.BitwiseXor(a, ir.BitwiseNot(b)), ir.BitwiseXor(b, c));
+    case 232:
+        return ir.BitwiseAnd(ir.BitwiseOr(a, b), ir.BitwiseOr(c, ir.BitwiseAnd(a, b)));
+    case 233:
+        return ir.BitwiseOr(ir.BitwiseAnd(a, b),
+                            ir.BitwiseXor(ir.BitwiseNot(c), ir.BitwiseOr(a, b)));
+    case 234:
+        return ir.BitwiseOr(c, ir.BitwiseAnd(a, b));
+    case 235:
+        return ir.BitwiseOr(c, ir.BitwiseXor(a, ir.BitwiseNot(b)));
+    case 236:
+        return ir.BitwiseOr(b, ir.BitwiseAnd(a, c));
+    case 237:
+        return ir.BitwiseOr(b, ir.BitwiseXor(a, ir.BitwiseNot(c)));
+    case 238:
+        return ir.BitwiseOr(b, c);
+    case 239:
+        return ir.BitwiseOr(ir.BitwiseNot(a), ir.BitwiseOr(b, c));
+    case 240:
+        return a;
+    case 241:
+        return ir.BitwiseOr(a, ir.BitwiseNot(ir.BitwiseOr(b, c)));
+    case 242:
+        return ir.BitwiseOr(a, ir.BitwiseAnd(c, ir.BitwiseNot(b)));
+    case 243:
+        return ir.BitwiseOr(a, ir.BitwiseNot(b));
+    case 244:
+        return ir.BitwiseOr(a, ir.BitwiseAnd(b, ir.BitwiseNot(c)));
+    case 245:
+        return ir.BitwiseOr(a, ir.BitwiseNot(c));
+    case 246:
+        return ir.BitwiseOr(a, ir.BitwiseXor(b, c));
+    case 247:
+        return ir.BitwiseOr(a, ir.BitwiseNot(ir.BitwiseAnd(b, c)));
+    case 248:
+        return ir.BitwiseOr(a, ir.BitwiseAnd(b, c));
+    case 249:
+        return ir.BitwiseOr(a, ir.BitwiseXor(b, ir.BitwiseNot(c)));
+    case 250:
+        return ir.BitwiseOr(a, c);
+    case 251:
+        return ir.BitwiseOr(ir.BitwiseNot(b), ir.BitwiseOr(a, c));
+    case 252:
+        return ir.BitwiseOr(a, b);
+    case 253:
+        return ir.BitwiseOr(ir.BitwiseNot(c), ir.BitwiseOr(a, b));
+    case 254:
+        return ir.BitwiseOr(a, ir.BitwiseOr(b, c));
+    case 255:
+        return ir.Imm32(0xFFFFFFFF);
+        // end of generated code
     }
-    if (ttbl & 0x02) {
-        // r |= ~a & ~b & c;
-        const auto lhs{ir.BitwiseAnd(not_a, not_b)};
-        const auto rhs{ir.BitwiseAnd(lhs, c)};
-        r = ir.BitwiseOr(r, rhs);
-    }
-    if (ttbl & 0x04) {
-        // r |= ~a & b & ~c;
-        const auto lhs{ir.BitwiseAnd(not_a, b)};
-        const auto rhs{ir.BitwiseAnd(lhs, not_c)};
-        r = ir.BitwiseOr(r, rhs);
-    }
-    if (ttbl & 0x08) {
-        // r |= ~a & b & c;
-        const auto lhs{ir.BitwiseAnd(not_a, b)};
-        const auto rhs{ir.BitwiseAnd(lhs, c)};
-        r = ir.BitwiseOr(r, rhs);
-    }
-    if (ttbl & 0x10) {
-        // r |= a & ~b & ~c;
-        const auto lhs{ir.BitwiseAnd(a, not_b)};
-        const auto rhs{ir.BitwiseAnd(lhs, not_c)};
-        r = ir.BitwiseOr(r, rhs);
-    }
-    if (ttbl & 0x20) {
-        // r |= a & ~b & c;
-        const auto lhs{ir.BitwiseAnd(a, not_b)};
-        const auto rhs{ir.BitwiseAnd(lhs, c)};
-        r = ir.BitwiseOr(r, rhs);
-    }
-    if (ttbl & 0x40) {
-        // r |= a & b & ~c;
-        const auto lhs{ir.BitwiseAnd(a, b)};
-        const auto rhs{ir.BitwiseAnd(lhs, not_c)};
-        r = ir.BitwiseOr(r, rhs);
-    }
-    if (ttbl & 0x80) {
-        // r |= a & b & c;
-        const auto lhs{ir.BitwiseAnd(a, b)};
-        const auto rhs{ir.BitwiseAnd(lhs, c)};
-        r = ir.BitwiseOr(r, rhs);
-    }
-    return r;
+    throw NotImplementedException("LOP3 with out of range ttbl");
 }
 
 IR::U32 LOP3(TranslatorVisitor& v, u64 insn, const IR::U32& op_b, const IR::U32& op_c, u64 lut) {
diff --git a/src/shader_recompiler/frontend/maxwell/translate/impl/logic_operation_three_input_lut3.py b/src/shader_recompiler/frontend/maxwell/translate/impl/logic_operation_three_input_lut3.py
new file mode 100644
index 000000000..8f547c266
--- /dev/null
+++ b/src/shader_recompiler/frontend/maxwell/translate/impl/logic_operation_three_input_lut3.py
@@ -0,0 +1,92 @@
+# Copyright © 2022 degasus <markus@selfnet.de>
+# This work is free. You can redistribute it and/or modify it under the
+# terms of the Do What The Fuck You Want To Public License, Version 2,
+# as published by Sam Hocevar. See http://www.wtfpl.net/ for more details.
+
+from itertools import product
+
+# The primitive instructions
+OPS = {
+    'ir.BitwiseAnd({}, {})' : (2, 1, lambda a,b: a&b),
+    'ir.BitwiseOr({}, {})' : (2, 1, lambda a,b: a|b),
+    'ir.BitwiseXor({}, {})' : (2, 1, lambda a,b: a^b),
+    'ir.BitwiseNot({})' : (1, 0.1, lambda a: (~a) & 255), # Only tiny cost, as this can often inlined in other instructions
+}
+
+# Our database of combination of instructions
+optimized_calls = {}
+def cmp(lhs, rhs):
+    if lhs is None: # new entry
+        return True
+    if lhs[3] > rhs[3]: # costs
+        return True
+    if lhs[3] < rhs[3]: # costs
+        return False
+    if len(lhs[0]) > len(rhs[0]): # string len
+        return True
+    if len(lhs[0]) < len(rhs[0]): # string len
+        return False
+    if lhs[0] > rhs[0]: # string sorting
+        return True
+    if lhs[0] < rhs[0]: # string sorting
+        return False
+    assert lhs == rhs, "redundant instruction, bug in brute force"
+    return False
+def register(imm, instruction, count, latency):
+    # Use the sum of instruction count and latency as costs to evaluate which combination is best
+    costs = count + latency
+
+    old = optimized_calls.get(imm, None)
+    new = (instruction, count, latency, costs)
+
+    # Update if new or better
+    if cmp(old, new):
+        optimized_calls[imm] = new
+        return True
+
+    return False
+
+# Constants: 0, 1 (for free)
+register(0, 'ir.Imm32(0)', 0, 0)
+register(255, 'ir.Imm32(0xFFFFFFFF)', 0, 0)
+
+# Inputs: a, b, c (for free)
+ta = 0xF0
+tb = 0xCC
+tc = 0xAA
+inputs = {
+    ta : 'a',
+    tb : 'b',
+    tc : 'c',
+}
+for imm, instruction in inputs.items():
+    register(imm, instruction, 0, 0)
+    register((~imm) & 255, 'ir.BitwiseNot({})'.format(instruction), 0.099, 0.099) # slightly cheaper NEG on inputs
+
+# Try to combine two values from the db with an instruction.
+# If it is better than the old method, update it.
+while True:
+    registered = 0
+    calls_copy = optimized_calls.copy()
+    for OP, (argc, cost, f) in OPS.items():
+        for args in product(calls_copy.items(), repeat=argc):
+            # unpack(transponse) the arrays
+            imm = [arg[0] for arg in args]
+            value = [arg[1][0] for arg in args]
+            count = [arg[1][1] for arg in args]
+            latency = [arg[1][2] for arg in args]
+
+            registered += register(
+                f(*imm),
+                OP.format(*value),
+                sum(count) + cost,
+                max(latency) + cost)
+    if registered == 0:
+        # No update at all? So terminate
+        break
+
+# Hacky output. Please improve me to output valid C++ instead.
+s = """    case {imm}:
+        return {op};"""
+for imm in range(256):
+    print(s.format(imm=imm, op=optimized_calls[imm][0]))
diff --git a/src/shader_recompiler/frontend/maxwell/translate_program.cpp b/src/shader_recompiler/frontend/maxwell/translate_program.cpp
index 248ad3ced..b22725584 100644
--- a/src/shader_recompiler/frontend/maxwell/translate_program.cpp
+++ b/src/shader_recompiler/frontend/maxwell/translate_program.cpp
@@ -212,11 +212,11 @@ IR::Program TranslateProgram(ObjectPool<IR::Inst>& inst_pool, ObjectPool<IR::Blo
     }
     Optimization::SsaRewritePass(program);
 
+    Optimization::ConstantPropagationPass(program);
+
     Optimization::GlobalMemoryToStorageBufferPass(program);
     Optimization::TexturePass(env, program);
 
-    Optimization::ConstantPropagationPass(program);
-
     if (Settings::values.resolution_info.active) {
         Optimization::RescalingPass(program);
     }
diff --git a/src/shader_recompiler/ir_opt/collect_shader_info_pass.cpp b/src/shader_recompiler/ir_opt/collect_shader_info_pass.cpp
index b6a20f904..0b2c60842 100644
--- a/src/shader_recompiler/ir_opt/collect_shader_info_pass.cpp
+++ b/src/shader_recompiler/ir_opt/collect_shader_info_pass.cpp
@@ -29,6 +29,46 @@ void AddConstantBufferDescriptor(Info& info, u32 index, u32 count) {
                  });
 }
 
+void AddRegisterIndexedLdc(Info& info) {
+    info.uses_cbuf_indirect = true;
+
+    // The shader can use any possible constant buffer
+    info.constant_buffer_mask = (1 << Info::MAX_CBUFS) - 1;
+
+    auto& cbufs{info.constant_buffer_descriptors};
+    cbufs.clear();
+    for (u32 i = 0; i < Info::MAX_CBUFS; i++) {
+        cbufs.push_back(ConstantBufferDescriptor{.index = i, .count = 1});
+
+        // The shader can use any possible access size
+        info.constant_buffer_used_sizes[i] = 0x10'000;
+    }
+}
+
+u32 GetElementSize(IR::Type& used_type, Shader::IR::Opcode opcode) {
+    switch (opcode) {
+    case IR::Opcode::GetCbufU8:
+    case IR::Opcode::GetCbufS8:
+        used_type |= IR::Type::U8;
+        return 1;
+    case IR::Opcode::GetCbufU16:
+    case IR::Opcode::GetCbufS16:
+        used_type |= IR::Type::U16;
+        return 2;
+    case IR::Opcode::GetCbufU32:
+        used_type |= IR::Type::U32;
+        return 4;
+    case IR::Opcode::GetCbufF32:
+        used_type |= IR::Type::F32;
+        return 4;
+    case IR::Opcode::GetCbufU32x2:
+        used_type |= IR::Type::U32x2;
+        return 8;
+    default:
+        throw InvalidArgument("Invalid opcode {}", opcode);
+    }
+}
+
 void GetPatch(Info& info, IR::Patch patch) {
     if (!IR::IsGeneric(patch)) {
         throw NotImplementedException("Reading non-generic patch {}", patch);
@@ -360,6 +400,15 @@ void VisitUsages(Info& info, IR::Inst& inst) {
     case IR::Opcode::GlobalAtomicOr64:
     case IR::Opcode::GlobalAtomicXor64:
     case IR::Opcode::GlobalAtomicExchange64:
+    case IR::Opcode::GlobalAtomicIAdd32x2:
+    case IR::Opcode::GlobalAtomicSMin32x2:
+    case IR::Opcode::GlobalAtomicUMin32x2:
+    case IR::Opcode::GlobalAtomicSMax32x2:
+    case IR::Opcode::GlobalAtomicUMax32x2:
+    case IR::Opcode::GlobalAtomicAnd32x2:
+    case IR::Opcode::GlobalAtomicOr32x2:
+    case IR::Opcode::GlobalAtomicXor32x2:
+    case IR::Opcode::GlobalAtomicExchange32x2:
     case IR::Opcode::GlobalAtomicAddF32:
     case IR::Opcode::GlobalAtomicAddF16x2:
     case IR::Opcode::GlobalAtomicAddF32x2:
@@ -454,42 +503,18 @@ void VisitUsages(Info& info, IR::Inst& inst) {
     case IR::Opcode::GetCbufU32x2: {
         const IR::Value index{inst.Arg(0)};
         const IR::Value offset{inst.Arg(1)};
-        if (!index.IsImmediate()) {
-            throw NotImplementedException("Constant buffer with non-immediate index");
-        }
-        AddConstantBufferDescriptor(info, index.U32(), 1);
-        u32 element_size{};
-        switch (inst.GetOpcode()) {
-        case IR::Opcode::GetCbufU8:
-        case IR::Opcode::GetCbufS8:
-            info.used_constant_buffer_types |= IR::Type::U8;
-            element_size = 1;
-            break;
-        case IR::Opcode::GetCbufU16:
-        case IR::Opcode::GetCbufS16:
-            info.used_constant_buffer_types |= IR::Type::U16;
-            element_size = 2;
-            break;
-        case IR::Opcode::GetCbufU32:
-            info.used_constant_buffer_types |= IR::Type::U32;
-            element_size = 4;
-            break;
-        case IR::Opcode::GetCbufF32:
-            info.used_constant_buffer_types |= IR::Type::F32;
-            element_size = 4;
-            break;
-        case IR::Opcode::GetCbufU32x2:
-            info.used_constant_buffer_types |= IR::Type::U32x2;
-            element_size = 8;
-            break;
-        default:
-            break;
-        }
-        u32& size{info.constant_buffer_used_sizes[index.U32()]};
-        if (offset.IsImmediate()) {
-            size = Common::AlignUp(std::max(size, offset.U32() + element_size), 16u);
+        if (index.IsImmediate()) {
+            AddConstantBufferDescriptor(info, index.U32(), 1);
+            u32 element_size = GetElementSize(info.used_constant_buffer_types, inst.GetOpcode());
+            u32& size{info.constant_buffer_used_sizes[index.U32()]};
+            if (offset.IsImmediate()) {
+                size = Common::AlignUp(std::max(size, offset.U32() + element_size), 16u);
+            } else {
+                size = 0x10'000;
+            }
         } else {
-            size = 0x10'000;
+            AddRegisterIndexedLdc(info);
+            GetElementSize(info.used_indirect_cbuf_types, inst.GetOpcode());
         }
         break;
     }
@@ -597,6 +622,15 @@ void VisitUsages(Info& info, IR::Inst& inst) {
         break;
     case IR::Opcode::LoadStorage64:
     case IR::Opcode::WriteStorage64:
+    case IR::Opcode::StorageAtomicIAdd32x2:
+    case IR::Opcode::StorageAtomicSMin32x2:
+    case IR::Opcode::StorageAtomicUMin32x2:
+    case IR::Opcode::StorageAtomicSMax32x2:
+    case IR::Opcode::StorageAtomicUMax32x2:
+    case IR::Opcode::StorageAtomicAnd32x2:
+    case IR::Opcode::StorageAtomicOr32x2:
+    case IR::Opcode::StorageAtomicXor32x2:
+    case IR::Opcode::StorageAtomicExchange32x2:
         info.used_storage_buffer_types |= IR::Type::U32x2;
         break;
     case IR::Opcode::LoadStorage128:
diff --git a/src/shader_recompiler/ir_opt/global_memory_to_storage_buffer_pass.cpp b/src/shader_recompiler/ir_opt/global_memory_to_storage_buffer_pass.cpp
index 4197b0095..ddf497e32 100644
--- a/src/shader_recompiler/ir_opt/global_memory_to_storage_buffer_pass.cpp
+++ b/src/shader_recompiler/ir_opt/global_memory_to_storage_buffer_pass.cpp
@@ -92,6 +92,15 @@ bool IsGlobalMemory(const IR::Inst& inst) {
     case IR::Opcode::GlobalAtomicOr64:
     case IR::Opcode::GlobalAtomicXor64:
     case IR::Opcode::GlobalAtomicExchange64:
+    case IR::Opcode::GlobalAtomicIAdd32x2:
+    case IR::Opcode::GlobalAtomicSMin32x2:
+    case IR::Opcode::GlobalAtomicUMin32x2:
+    case IR::Opcode::GlobalAtomicSMax32x2:
+    case IR::Opcode::GlobalAtomicUMax32x2:
+    case IR::Opcode::GlobalAtomicAnd32x2:
+    case IR::Opcode::GlobalAtomicOr32x2:
+    case IR::Opcode::GlobalAtomicXor32x2:
+    case IR::Opcode::GlobalAtomicExchange32x2:
     case IR::Opcode::GlobalAtomicAddF32:
     case IR::Opcode::GlobalAtomicAddF16x2:
     case IR::Opcode::GlobalAtomicAddF32x2:
@@ -135,6 +144,15 @@ bool IsGlobalMemoryWrite(const IR::Inst& inst) {
     case IR::Opcode::GlobalAtomicOr64:
     case IR::Opcode::GlobalAtomicXor64:
     case IR::Opcode::GlobalAtomicExchange64:
+    case IR::Opcode::GlobalAtomicIAdd32x2:
+    case IR::Opcode::GlobalAtomicSMin32x2:
+    case IR::Opcode::GlobalAtomicUMin32x2:
+    case IR::Opcode::GlobalAtomicSMax32x2:
+    case IR::Opcode::GlobalAtomicUMax32x2:
+    case IR::Opcode::GlobalAtomicAnd32x2:
+    case IR::Opcode::GlobalAtomicOr32x2:
+    case IR::Opcode::GlobalAtomicXor32x2:
+    case IR::Opcode::GlobalAtomicExchange32x2:
     case IR::Opcode::GlobalAtomicAddF32:
     case IR::Opcode::GlobalAtomicAddF16x2:
     case IR::Opcode::GlobalAtomicAddF32x2:
@@ -199,6 +217,8 @@ IR::Opcode GlobalToStorage(IR::Opcode opcode) {
         return IR::Opcode::StorageAtomicOr32;
     case IR::Opcode::GlobalAtomicXor32:
         return IR::Opcode::StorageAtomicXor32;
+    case IR::Opcode::GlobalAtomicExchange32:
+        return IR::Opcode::StorageAtomicExchange32;
     case IR::Opcode::GlobalAtomicIAdd64:
         return IR::Opcode::StorageAtomicIAdd64;
     case IR::Opcode::GlobalAtomicSMin64:
@@ -215,10 +235,26 @@ IR::Opcode GlobalToStorage(IR::Opcode opcode) {
         return IR::Opcode::StorageAtomicOr64;
     case IR::Opcode::GlobalAtomicXor64:
         return IR::Opcode::StorageAtomicXor64;
-    case IR::Opcode::GlobalAtomicExchange32:
-        return IR::Opcode::StorageAtomicExchange32;
     case IR::Opcode::GlobalAtomicExchange64:
         return IR::Opcode::StorageAtomicExchange64;
+    case IR::Opcode::GlobalAtomicIAdd32x2:
+        return IR::Opcode::StorageAtomicIAdd32x2;
+    case IR::Opcode::GlobalAtomicSMin32x2:
+        return IR::Opcode::StorageAtomicSMin32x2;
+    case IR::Opcode::GlobalAtomicUMin32x2:
+        return IR::Opcode::StorageAtomicUMin32x2;
+    case IR::Opcode::GlobalAtomicSMax32x2:
+        return IR::Opcode::StorageAtomicSMax32x2;
+    case IR::Opcode::GlobalAtomicUMax32x2:
+        return IR::Opcode::StorageAtomicUMax32x2;
+    case IR::Opcode::GlobalAtomicAnd32x2:
+        return IR::Opcode::StorageAtomicAnd32x2;
+    case IR::Opcode::GlobalAtomicOr32x2:
+        return IR::Opcode::StorageAtomicOr32x2;
+    case IR::Opcode::GlobalAtomicXor32x2:
+        return IR::Opcode::StorageAtomicXor32x2;
+    case IR::Opcode::GlobalAtomicExchange32x2:
+        return IR::Opcode::StorageAtomicExchange32x2;
     case IR::Opcode::GlobalAtomicAddF32:
         return IR::Opcode::StorageAtomicAddF32;
     case IR::Opcode::GlobalAtomicAddF16x2:
@@ -298,7 +334,8 @@ std::optional<LowAddrInfo> TrackLowAddress(IR::Inst* inst) {
 /// Tries to track the storage buffer address used by a global memory instruction
 std::optional<StorageBufferAddr> Track(const IR::Value& value, const Bias* bias) {
     const auto pred{[bias](const IR::Inst* inst) -> std::optional<StorageBufferAddr> {
-        if (inst->GetOpcode() != IR::Opcode::GetCbufU32) {
+        if (inst->GetOpcode() != IR::Opcode::GetCbufU32 &&
+            inst->GetOpcode() != IR::Opcode::GetCbufU32x2) {
             return std::nullopt;
         }
         const IR::Value index{inst->Arg(0)};
@@ -454,6 +491,15 @@ void Replace(IR::Block& block, IR::Inst& inst, const IR::U32& storage_index,
     case IR::Opcode::GlobalAtomicOr64:
     case IR::Opcode::GlobalAtomicXor64:
     case IR::Opcode::GlobalAtomicExchange64:
+    case IR::Opcode::GlobalAtomicIAdd32x2:
+    case IR::Opcode::GlobalAtomicSMin32x2:
+    case IR::Opcode::GlobalAtomicUMin32x2:
+    case IR::Opcode::GlobalAtomicSMax32x2:
+    case IR::Opcode::GlobalAtomicUMax32x2:
+    case IR::Opcode::GlobalAtomicAnd32x2:
+    case IR::Opcode::GlobalAtomicOr32x2:
+    case IR::Opcode::GlobalAtomicXor32x2:
+    case IR::Opcode::GlobalAtomicExchange32x2:
     case IR::Opcode::GlobalAtomicAddF32:
     case IR::Opcode::GlobalAtomicAddF16x2:
     case IR::Opcode::GlobalAtomicAddF32x2:
diff --git a/src/shader_recompiler/ir_opt/lower_int64_to_int32.cpp b/src/shader_recompiler/ir_opt/lower_int64_to_int32.cpp
index e80d3d1d9..c2654cd9b 100644
--- a/src/shader_recompiler/ir_opt/lower_int64_to_int32.cpp
+++ b/src/shader_recompiler/ir_opt/lower_int64_to_int32.cpp
@@ -199,6 +199,26 @@ void Lower(IR::Block& block, IR::Inst& inst) {
         return ShiftRightLogical64To32(block, inst);
     case IR::Opcode::ShiftRightArithmetic64:
         return ShiftRightArithmetic64To32(block, inst);
+    case IR::Opcode::SharedAtomicExchange64:
+        return inst.ReplaceOpcode(IR::Opcode::SharedAtomicExchange32x2);
+    case IR::Opcode::GlobalAtomicIAdd64:
+        return inst.ReplaceOpcode(IR::Opcode::GlobalAtomicIAdd32x2);
+    case IR::Opcode::GlobalAtomicSMin64:
+        return inst.ReplaceOpcode(IR::Opcode::GlobalAtomicSMin32x2);
+    case IR::Opcode::GlobalAtomicUMin64:
+        return inst.ReplaceOpcode(IR::Opcode::GlobalAtomicUMin32x2);
+    case IR::Opcode::GlobalAtomicSMax64:
+        return inst.ReplaceOpcode(IR::Opcode::GlobalAtomicSMax32x2);
+    case IR::Opcode::GlobalAtomicUMax64:
+        return inst.ReplaceOpcode(IR::Opcode::GlobalAtomicUMax32x2);
+    case IR::Opcode::GlobalAtomicAnd64:
+        return inst.ReplaceOpcode(IR::Opcode::GlobalAtomicAnd32x2);
+    case IR::Opcode::GlobalAtomicOr64:
+        return inst.ReplaceOpcode(IR::Opcode::GlobalAtomicOr32x2);
+    case IR::Opcode::GlobalAtomicXor64:
+        return inst.ReplaceOpcode(IR::Opcode::GlobalAtomicXor32x2);
+    case IR::Opcode::GlobalAtomicExchange64:
+        return inst.ReplaceOpcode(IR::Opcode::GlobalAtomicExchange32x2);
     default:
         break;
     }
diff --git a/src/shader_recompiler/ir_opt/rescaling_pass.cpp b/src/shader_recompiler/ir_opt/rescaling_pass.cpp
index c28500dd1..496d4667e 100644
--- a/src/shader_recompiler/ir_opt/rescaling_pass.cpp
+++ b/src/shader_recompiler/ir_opt/rescaling_pass.cpp
@@ -183,6 +183,31 @@ void ScaleIntegerComposite(IR::IREmitter& ir, IR::Inst& inst, const IR::U1& is_s
     }
 }
 
+void ScaleIntegerOffsetComposite(IR::IREmitter& ir, IR::Inst& inst, const IR::U1& is_scaled,
+                                 size_t index) {
+    const IR::Value composite{inst.Arg(index)};
+    if (composite.IsEmpty()) {
+        return;
+    }
+    const auto info{inst.Flags<IR::TextureInstInfo>()};
+    const IR::U32 x{Scale(ir, is_scaled, IR::U32{ir.CompositeExtract(composite, 0)})};
+    const IR::U32 y{Scale(ir, is_scaled, IR::U32{ir.CompositeExtract(composite, 1)})};
+    switch (info.type) {
+    case TextureType::ColorArray2D:
+    case TextureType::Color2D:
+        inst.SetArg(index, ir.CompositeConstruct(x, y));
+        break;
+    case TextureType::Color1D:
+    case TextureType::ColorArray1D:
+    case TextureType::Color3D:
+    case TextureType::ColorCube:
+    case TextureType::ColorArrayCube:
+    case TextureType::Buffer:
+        // Nothing to patch here
+        break;
+    }
+}
+
 void SubScaleCoord(IR::IREmitter& ir, IR::Inst& inst, const IR::U1& is_scaled) {
     const auto info{inst.Flags<IR::TextureInstInfo>()};
     const IR::Value coord{inst.Arg(1)};
@@ -220,7 +245,7 @@ void SubScaleImageFetch(IR::Block& block, IR::Inst& inst) {
     const IR::U1 is_scaled{ir.IsTextureScaled(ir.Imm32(info.descriptor_index))};
     SubScaleCoord(ir, inst, is_scaled);
     // Scale ImageFetch offset
-    ScaleIntegerComposite(ir, inst, is_scaled, 2);
+    ScaleIntegerOffsetComposite(ir, inst, is_scaled, 2);
 }
 
 void SubScaleImageRead(IR::Block& block, IR::Inst& inst) {
@@ -242,7 +267,7 @@ void PatchImageFetch(IR::Block& block, IR::Inst& inst) {
     const IR::U1 is_scaled{ir.IsTextureScaled(ir.Imm32(info.descriptor_index))};
     ScaleIntegerComposite(ir, inst, is_scaled, 1);
     // Scale ImageFetch offset
-    ScaleIntegerComposite(ir, inst, is_scaled, 2);
+    ScaleIntegerOffsetComposite(ir, inst, is_scaled, 2);
 }
 
 void PatchImageRead(IR::Block& block, IR::Inst& inst) {
diff --git a/src/shader_recompiler/shader_info.h b/src/shader_recompiler/shader_info.h
index 9f375c30e..9d36bd9eb 100644
--- a/src/shader_recompiler/shader_info.h
+++ b/src/shader_recompiler/shader_info.h
@@ -173,9 +173,11 @@ struct Info {
     bool uses_atomic_image_u32{};
     bool uses_shadow_lod{};
     bool uses_rescaling_uniform{};
+    bool uses_cbuf_indirect{};
 
     IR::Type used_constant_buffer_types{};
     IR::Type used_storage_buffer_types{};
+    IR::Type used_indirect_cbuf_types{};
 
     u32 constant_buffer_mask{};
     std::array<u32, MAX_CBUFS> constant_buffer_used_sizes{};
diff --git a/src/video_core/engines/maxwell_3d.cpp b/src/video_core/engines/maxwell_3d.cpp
index 5d6d217bb..54a902f56 100644
--- a/src/video_core/engines/maxwell_3d.cpp
+++ b/src/video_core/engines/maxwell_3d.cpp
@@ -7,6 +7,7 @@
 #include "common/assert.h"
 #include "core/core.h"
 #include "core/core_timing.h"
+#include "video_core/dirty_flags.h"
 #include "video_core/engines/maxwell_3d.h"
 #include "video_core/gpu.h"
 #include "video_core/memory_manager.h"
@@ -195,7 +196,7 @@ void Maxwell3D::ProcessMethodCall(u32 method, u32 argument, u32 nonshadow_argume
     case MAXWELL3D_REG_INDEX(const_buffer.cb_data) + 13:
     case MAXWELL3D_REG_INDEX(const_buffer.cb_data) + 14:
     case MAXWELL3D_REG_INDEX(const_buffer.cb_data) + 15:
-        return StartCBData(method);
+        return ProcessCBData(argument);
     case MAXWELL3D_REG_INDEX(cb_bind[0]):
         return ProcessCBBind(0);
     case MAXWELL3D_REG_INDEX(cb_bind[1]):
@@ -208,6 +209,14 @@ void Maxwell3D::ProcessMethodCall(u32 method, u32 argument, u32 nonshadow_argume
         return ProcessCBBind(4);
     case MAXWELL3D_REG_INDEX(draw.vertex_end_gl):
         return DrawArrays();
+    case MAXWELL3D_REG_INDEX(small_index):
+        regs.index_array.count = regs.small_index.count;
+        regs.index_array.first = regs.small_index.first;
+        dirty.flags[VideoCommon::Dirty::IndexBuffer] = true;
+        return DrawArrays();
+    case MAXWELL3D_REG_INDEX(topology_override):
+        use_topology_override = true;
+        return;
     case MAXWELL3D_REG_INDEX(clear_buffers):
         return ProcessClearBuffers();
     case MAXWELL3D_REG_INDEX(query.query_get):
@@ -248,14 +257,6 @@ void Maxwell3D::CallMacroMethod(u32 method, const std::vector<u32>& parameters)
 }
 
 void Maxwell3D::CallMethod(u32 method, u32 method_argument, bool is_last_call) {
-    if (method == cb_data_state.current) {
-        regs.reg_array[method] = method_argument;
-        ProcessCBData(method_argument);
-        return;
-    } else if (cb_data_state.current != null_cb_data) {
-        FinishCBData();
-    }
-
     // It is an error to write to a register other than the current macro's ARG register before it
     // has finished execution.
     if (executing_macro != 0) {
@@ -302,7 +303,7 @@ void Maxwell3D::CallMultiMethod(u32 method, const u32* base_start, u32 amount,
     case MAXWELL3D_REG_INDEX(const_buffer.cb_data) + 13:
     case MAXWELL3D_REG_INDEX(const_buffer.cb_data) + 14:
     case MAXWELL3D_REG_INDEX(const_buffer.cb_data) + 15:
-        ProcessCBMultiData(method, base_start, amount);
+        ProcessCBMultiData(base_start, amount);
         break;
     default:
         for (std::size_t i = 0; i < amount; i++) {
@@ -360,6 +361,35 @@ void Maxwell3D::CallMethodFromMME(u32 method, u32 method_argument) {
     }
 }
 
+void Maxwell3D::ProcessTopologyOverride() {
+    using PrimitiveTopology = Maxwell3D::Regs::PrimitiveTopology;
+    using PrimitiveTopologyOverride = Maxwell3D::Regs::PrimitiveTopologyOverride;
+
+    PrimitiveTopology topology{};
+
+    switch (regs.topology_override) {
+    case PrimitiveTopologyOverride::None:
+        topology = regs.draw.topology;
+        break;
+    case PrimitiveTopologyOverride::Points:
+        topology = PrimitiveTopology::Points;
+        break;
+    case PrimitiveTopologyOverride::Lines:
+        topology = PrimitiveTopology::Lines;
+        break;
+    case PrimitiveTopologyOverride::LineStrip:
+        topology = PrimitiveTopology::LineStrip;
+        break;
+    default:
+        topology = static_cast<PrimitiveTopology>(regs.topology_override);
+        break;
+    }
+
+    if (use_topology_override) {
+        regs.draw.topology.Assign(topology);
+    }
+}
+
 void Maxwell3D::FlushMMEInlineDraw() {
     LOG_TRACE(HW_GPU, "called, topology={}, count={}", regs.draw.topology.Value(),
               regs.vertex_buffer.count);
@@ -370,6 +400,8 @@ void Maxwell3D::FlushMMEInlineDraw() {
     ASSERT_MSG(!regs.draw.instance_next || !regs.draw.instance_cont,
                "Illegal combination of instancing parameters");
 
+    ProcessTopologyOverride();
+
     const bool is_indexed = mme_draw.current_mode == MMEDrawMode::Indexed;
     if (ShouldExecute()) {
         rasterizer->Draw(is_indexed, true);
@@ -529,6 +561,8 @@ void Maxwell3D::DrawArrays() {
     ASSERT_MSG(!regs.draw.instance_next || !regs.draw.instance_cont,
                "Illegal combination of instancing parameters");
 
+    ProcessTopologyOverride();
+
     if (regs.draw.instance_next) {
         // Increment the current instance *before* drawing.
         state.current_instance += 1;
@@ -587,46 +621,7 @@ void Maxwell3D::ProcessCBBind(size_t stage_index) {
     rasterizer->BindGraphicsUniformBuffer(stage_index, bind_data.index, gpu_addr, size);
 }
 
-void Maxwell3D::ProcessCBData(u32 value) {
-    const u32 id = cb_data_state.id;
-    cb_data_state.buffer[id][cb_data_state.counter] = value;
-    // Increment the current buffer position.
-    regs.const_buffer.cb_pos = regs.const_buffer.cb_pos + 4;
-    cb_data_state.counter++;
-}
-
-void Maxwell3D::StartCBData(u32 method) {
-    constexpr u32 first_cb_data = MAXWELL3D_REG_INDEX(const_buffer.cb_data);
-    cb_data_state.start_pos = regs.const_buffer.cb_pos;
-    cb_data_state.id = method - first_cb_data;
-    cb_data_state.current = method;
-    cb_data_state.counter = 0;
-    ProcessCBData(regs.const_buffer.cb_data[cb_data_state.id]);
-}
-
-void Maxwell3D::ProcessCBMultiData(u32 method, const u32* start_base, u32 amount) {
-    if (cb_data_state.current != method) {
-        if (cb_data_state.current != null_cb_data) {
-            FinishCBData();
-        }
-        constexpr u32 first_cb_data = MAXWELL3D_REG_INDEX(const_buffer.cb_data);
-        cb_data_state.start_pos = regs.const_buffer.cb_pos;
-        cb_data_state.id = method - first_cb_data;
-        cb_data_state.current = method;
-        cb_data_state.counter = 0;
-    }
-    const std::size_t id = cb_data_state.id;
-    const std::size_t size = amount;
-    std::size_t i = 0;
-    for (; i < size; i++) {
-        cb_data_state.buffer[id][cb_data_state.counter] = start_base[i];
-        cb_data_state.counter++;
-    }
-    // Increment the current buffer position.
-    regs.const_buffer.cb_pos = regs.const_buffer.cb_pos + 4 * amount;
-}
-
-void Maxwell3D::FinishCBData() {
+void Maxwell3D::ProcessCBMultiData(const u32* start_base, u32 amount) {
     // Write the input value to the current const buffer at the current position.
     const GPUVAddr buffer_address = regs.const_buffer.BufferAddress();
     ASSERT(buffer_address != 0);
@@ -634,14 +629,16 @@ void Maxwell3D::FinishCBData() {
     // Don't allow writing past the end of the buffer.
     ASSERT(regs.const_buffer.cb_pos <= regs.const_buffer.cb_size);
 
-    const GPUVAddr address{buffer_address + cb_data_state.start_pos};
-    const std::size_t size = regs.const_buffer.cb_pos - cb_data_state.start_pos;
+    const GPUVAddr address{buffer_address + regs.const_buffer.cb_pos};
+    const size_t copy_size = amount * sizeof(u32);
+    memory_manager.WriteBlock(address, start_base, copy_size);
 
-    const u32 id = cb_data_state.id;
-    memory_manager.WriteBlock(address, cb_data_state.buffer[id].data(), size);
+    // Increment the current buffer position.
+    regs.const_buffer.cb_pos += static_cast<u32>(copy_size);
+}
 
-    cb_data_state.id = null_cb_data;
-    cb_data_state.current = null_cb_data;
+void Maxwell3D::ProcessCBData(u32 value) {
+    ProcessCBMultiData(&value, 1);
 }
 
 Texture::TICEntry Maxwell3D::GetTICEntry(u32 tic_index) const {
diff --git a/src/video_core/engines/maxwell_3d.h b/src/video_core/engines/maxwell_3d.h
index dc9df6c8b..357a74c70 100644
--- a/src/video_core/engines/maxwell_3d.h
+++ b/src/video_core/engines/maxwell_3d.h
@@ -367,6 +367,22 @@ public:
             Patches = 0xe,
         };
 
+        // Constants as from NVC0_3D_UNK1970_D3D
+        // https://gitlab.freedesktop.org/mesa/mesa/-/blob/main/src/gallium/drivers/nouveau/nvc0/nvc0_3d.xml.h#L1598
+        enum class PrimitiveTopologyOverride : u32 {
+            None = 0x0,
+            Points = 0x1,
+            Lines = 0x2,
+            LineStrip = 0x3,
+            Triangles = 0x4,
+            TriangleStrip = 0x5,
+            LinesAdjacency = 0xa,
+            LineStripAdjacency = 0xb,
+            TrianglesAdjacency = 0xc,
+            TriangleStripAdjacency = 0xd,
+            Patches = 0xe,
+        };
+
         enum class IndexFormat : u32 {
             UnsignedByte = 0x0,
             UnsignedShort = 0x1,
@@ -1200,7 +1216,12 @@ public:
                     }
                 } index_array;
 
-                INSERT_PADDING_WORDS_NOINIT(0x7);
+                union {
+                    BitField<0, 16, u32> first;
+                    BitField<16, 16, u32> count;
+                } small_index;
+
+                INSERT_PADDING_WORDS_NOINIT(0x6);
 
                 INSERT_PADDING_WORDS_NOINIT(0x1F);
 
@@ -1244,7 +1265,11 @@ public:
                     BitField<11, 1, u32> depth_clamp_disabled;
                 } view_volume_clip_control;
 
-                INSERT_PADDING_WORDS_NOINIT(0x1F);
+                INSERT_PADDING_WORDS_NOINIT(0xC);
+
+                PrimitiveTopologyOverride topology_override;
+
+                INSERT_PADDING_WORDS_NOINIT(0x12);
 
                 u32 depth_bounds_enable;
 
@@ -1520,10 +1545,8 @@ private:
     void ProcessSyncPoint();
 
     /// Handles a write to the CB_DATA[i] register.
-    void StartCBData(u32 method);
     void ProcessCBData(u32 value);
-    void ProcessCBMultiData(u32 method, const u32* start_base, u32 amount);
-    void FinishCBData();
+    void ProcessCBMultiData(const u32* start_base, u32 amount);
 
     /// Handles a write to the CB_BIND register.
     void ProcessCBBind(size_t stage_index);
@@ -1531,6 +1554,9 @@ private:
     /// Handles a write to the VERTEX_END_GL register, triggering a draw.
     void DrawArrays();
 
+    /// Handles use of topology overrides (e.g., to avoid using a topology assigned from a macro)
+    void ProcessTopologyOverride();
+
     // Handles a instance drawcall from MME
     void StepInstance(MMEDrawMode expected_mode, u32 count);
 
@@ -1555,20 +1581,10 @@ private:
     /// Interpreter for the macro codes uploaded to the GPU.
     std::unique_ptr<MacroEngine> macro_engine;
 
-    static constexpr u32 null_cb_data = 0xFFFFFFFF;
-    struct CBDataState {
-        static constexpr size_t inline_size = 0x4000;
-        std::array<std::array<u32, inline_size>, 16> buffer;
-        u32 current{null_cb_data};
-        u32 id{null_cb_data};
-        u32 start_pos{};
-        u32 counter{};
-    };
-    CBDataState cb_data_state;
-
     Upload::State upload_state;
 
     bool execute_on{true};
+    bool use_topology_override{false};
 };
 
 #define ASSERT_REG_POSITION(field_name, position)                                                  \
@@ -1685,6 +1701,7 @@ ASSERT_REG_POSITION(draw, 0x585);
 ASSERT_REG_POSITION(primitive_restart, 0x591);
 ASSERT_REG_POSITION(provoking_vertex_last, 0x5A1);
 ASSERT_REG_POSITION(index_array, 0x5F2);
+ASSERT_REG_POSITION(small_index, 0x5F9);
 ASSERT_REG_POSITION(polygon_offset_clamp, 0x61F);
 ASSERT_REG_POSITION(instanced_arrays, 0x620);
 ASSERT_REG_POSITION(vp_point_size, 0x644);
@@ -1694,6 +1711,7 @@ ASSERT_REG_POSITION(cull_face, 0x648);
 ASSERT_REG_POSITION(pixel_center_integer, 0x649);
 ASSERT_REG_POSITION(viewport_transform_enabled, 0x64B);
 ASSERT_REG_POSITION(view_volume_clip_control, 0x64F);
+ASSERT_REG_POSITION(topology_override, 0x65C);
 ASSERT_REG_POSITION(depth_bounds_enable, 0x66F);
 ASSERT_REG_POSITION(logic_op, 0x671);
 ASSERT_REG_POSITION(clear_buffers, 0x674);
diff --git a/src/video_core/engines/maxwell_dma.cpp b/src/video_core/engines/maxwell_dma.cpp
index 67388d980..1fc1358bc 100644
--- a/src/video_core/engines/maxwell_dma.cpp
+++ b/src/video_core/engines/maxwell_dma.cpp
@@ -53,7 +53,6 @@ void MaxwellDMA::Launch() {
 
     // TODO(Subv): Perform more research and implement all features of this engine.
     const LaunchDMA& launch = regs.launch_dma;
-    ASSERT(launch.semaphore_type == LaunchDMA::SemaphoreType::NONE);
     ASSERT(launch.interrupt_type == LaunchDMA::InterruptType::NONE);
     ASSERT(launch.data_transfer_type == LaunchDMA::DataTransferType::NON_PIPELINED);
     ASSERT(regs.dst_params.origin.x == 0);
@@ -79,6 +78,7 @@ void MaxwellDMA::Launch() {
             CopyPitchToBlockLinear();
         }
     }
+    ReleaseSemaphore();
 }
 
 void MaxwellDMA::CopyPitchToPitch() {
@@ -244,4 +244,22 @@ void MaxwellDMA::FastCopyBlockLinearToPitch() {
     memory_manager.WriteBlock(regs.offset_out, write_buffer.data(), dst_size);
 }
 
+void MaxwellDMA::ReleaseSemaphore() {
+    const auto type = regs.launch_dma.semaphore_type;
+    const GPUVAddr address = regs.semaphore.address;
+    switch (type) {
+    case LaunchDMA::SemaphoreType::NONE:
+        break;
+    case LaunchDMA::SemaphoreType::RELEASE_ONE_WORD_SEMAPHORE:
+        memory_manager.Write<u32>(address, regs.semaphore.payload);
+        break;
+    case LaunchDMA::SemaphoreType::RELEASE_FOUR_WORD_SEMAPHORE:
+        memory_manager.Write<u64>(address, static_cast<u64>(regs.semaphore.payload));
+        memory_manager.Write<u64>(address + 8, system.GPU().GetTicks());
+        break;
+    default:
+        UNREACHABLE_MSG("Unknown semaphore type: {}", static_cast<u32>(type.Value()));
+    }
+}
+
 } // namespace Tegra::Engines
diff --git a/src/video_core/engines/maxwell_dma.h b/src/video_core/engines/maxwell_dma.h
index a04514425..2692cac8a 100644
--- a/src/video_core/engines/maxwell_dma.h
+++ b/src/video_core/engines/maxwell_dma.h
@@ -224,6 +224,8 @@ private:
 
     void FastCopyBlockLinearToPitch();
 
+    void ReleaseSemaphore();
+
     Core::System& system;
 
     MemoryManager& memory_manager;
diff --git a/src/video_core/fence_manager.h b/src/video_core/fence_manager.h
index 34dc6c596..f80d62c80 100644
--- a/src/video_core/fence_manager.h
+++ b/src/video_core/fence_manager.h
@@ -8,8 +8,6 @@
 #include <queue>
 
 #include "common/common_types.h"
-#include "common/settings.h"
-#include "core/core.h"
 #include "video_core/delayed_destruction_ring.h"
 #include "video_core/gpu.h"
 #include "video_core/memory_manager.h"
diff --git a/src/video_core/host_shaders/CMakeLists.txt b/src/video_core/host_shaders/CMakeLists.txt
index fd3e41434..af05d47d1 100644
--- a/src/video_core/host_shaders/CMakeLists.txt
+++ b/src/video_core/host_shaders/CMakeLists.txt
@@ -14,6 +14,7 @@ set(SHADER_FILES
     convert_d24s8_to_abgr8.frag
     convert_depth_to_float.frag
     convert_float_to_depth.frag
+    convert_s8d24_to_abgr8.frag
     full_screen_triangle.vert
     fxaa.frag
     fxaa.vert
diff --git a/src/video_core/host_shaders/convert_s8d24_to_abgr8.frag b/src/video_core/host_shaders/convert_s8d24_to_abgr8.frag
new file mode 100644
index 000000000..c8a1683b8
--- /dev/null
+++ b/src/video_core/host_shaders/convert_s8d24_to_abgr8.frag
@@ -0,0 +1,23 @@
+// Copyright 2022 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#version 450
+
+layout(binding = 0) uniform sampler2D depth_tex;
+layout(binding = 1) uniform isampler2D stencil_tex;
+
+layout(location = 0) out vec4 color;
+
+void main() {
+    ivec2 coord = ivec2(gl_FragCoord.xy);
+    uint depth = uint(textureLod(depth_tex, coord, 0).r * (exp2(24.0) - 1.0f));
+    uint stencil = uint(textureLod(stencil_tex, coord, 0).r);
+
+    highp uint depth_val =
+        uint(textureLod(depth_tex, coord, 0).r * (exp2(32.0) - 1.0));
+    lowp uint stencil_val = textureLod(stencil_tex, coord, 0).r;
+    highp uvec4 components =
+        uvec4((uvec3(depth_val) >> uvec3(24u, 16u, 8u)) & 0x000000FFu, stencil_val);
+    color.rgba = vec4(components) / (exp2(8.0) - 1.0);
+}
diff --git a/src/video_core/query_cache.h b/src/video_core/query_cache.h
index 392f82eb7..0173b54d8 100644
--- a/src/video_core/query_cache.h
+++ b/src/video_core/query_cache.h
@@ -18,7 +18,6 @@
 
 #include "common/assert.h"
 #include "common/settings.h"
-#include "core/core.h"
 #include "video_core/engines/maxwell_3d.h"
 #include "video_core/gpu.h"
 #include "video_core/memory_manager.h"
diff --git a/src/video_core/renderer_base.h b/src/video_core/renderer_base.h
index bb204454e..c5f974080 100644
--- a/src/video_core/renderer_base.h
+++ b/src/video_core/renderer_base.h
@@ -5,9 +5,10 @@
 #pragma once
 
 #include <atomic>
+#include <functional>
 #include <memory>
-#include <optional>
 
+#include "common/common_funcs.h"
 #include "common/common_types.h"
 #include "core/frontend/emu_window.h"
 #include "video_core/gpu.h"
@@ -28,8 +29,11 @@ struct RendererSettings {
     Layout::FramebufferLayout screenshot_framebuffer_layout;
 };
 
-class RendererBase : NonCopyable {
+class RendererBase {
 public:
+    YUZU_NON_COPYABLE(RendererBase);
+    YUZU_NON_MOVEABLE(RendererBase);
+
     explicit RendererBase(Core::Frontend::EmuWindow& window,
                           std::unique_ptr<Core::Frontend::GraphicsContext> context);
     virtual ~RendererBase();
diff --git a/src/video_core/renderer_opengl/gl_fence_manager.cpp b/src/video_core/renderer_opengl/gl_fence_manager.cpp
index 151290101..293ad7d59 100644
--- a/src/video_core/renderer_opengl/gl_fence_manager.cpp
+++ b/src/video_core/renderer_opengl/gl_fence_manager.cpp
@@ -31,9 +31,8 @@ bool GLInnerFence::IsSignaled() const {
         return true;
     }
     ASSERT(sync_object.handle != 0);
-    GLsizei length;
     GLint sync_status;
-    glGetSynciv(sync_object.handle, GL_SYNC_STATUS, sizeof(GLint), &length, &sync_status);
+    glGetSynciv(sync_object.handle, GL_SYNC_STATUS, 1, nullptr, &sync_status);
     return sync_status == GL_SIGNALED;
 }
 
diff --git a/src/video_core/renderer_opengl/gl_graphics_pipeline.cpp b/src/video_core/renderer_opengl/gl_graphics_pipeline.cpp
index f8495896c..9e6732abd 100644
--- a/src/video_core/renderer_opengl/gl_graphics_pipeline.cpp
+++ b/src/video_core/renderer_opengl/gl_graphics_pipeline.cpp
@@ -243,10 +243,6 @@ GraphicsPipeline::GraphicsPipeline(
             case Settings::ShaderBackend::GLASM:
                 if (!sources[stage].empty()) {
                     assembly_programs[stage] = CompileProgram(sources[stage], AssemblyStage(stage));
-                    if (in_parallel) {
-                        // Make sure program is built before continuing when building in parallel
-                        glGetString(GL_PROGRAM_ERROR_STRING_NV);
-                    }
                 }
                 break;
             case Settings::ShaderBackend::SPIRV:
@@ -256,20 +252,18 @@ GraphicsPipeline::GraphicsPipeline(
                 break;
             }
         }
-        if (in_parallel && backend != Settings::ShaderBackend::GLASM) {
-            // Make sure programs have built if we are building shaders in parallel
-            for (OGLProgram& program : source_programs) {
-                if (program.handle != 0) {
-                    GLint status{};
-                    glGetProgramiv(program.handle, GL_LINK_STATUS, &status);
-                }
-            }
+        if (in_parallel) {
+            std::lock_guard lock{built_mutex};
+            built_fence.Create();
+            // Flush this context to ensure compilation commands and fence are in the GPU pipe.
+            glFlush();
+            built_condvar.notify_one();
+        } else {
+            is_built = true;
         }
         if (shader_notify) {
             shader_notify->MarkShaderComplete();
         }
-        is_built = true;
-        built_condvar.notify_one();
     }};
     if (thread_worker) {
         thread_worker->QueueWork(std::move(func));
@@ -440,7 +434,7 @@ void GraphicsPipeline::ConfigureImpl(bool is_indexed) {
     buffer_cache.UpdateGraphicsBuffers(is_indexed);
     buffer_cache.BindHostGeometryBuffers(is_indexed);
 
-    if (!is_built.load(std::memory_order::relaxed)) {
+    if (!IsBuilt()) {
         WaitForBuild();
     }
     const bool use_assembly{assembly_programs[0].handle != 0};
@@ -585,8 +579,26 @@ void GraphicsPipeline::GenerateTransformFeedbackState() {
 }
 
 void GraphicsPipeline::WaitForBuild() {
-    std::unique_lock lock{built_mutex};
-    built_condvar.wait(lock, [this] { return is_built.load(std::memory_order::relaxed); });
+    if (built_fence.handle == 0) {
+        std::unique_lock lock{built_mutex};
+        built_condvar.wait(lock, [this] { return built_fence.handle != 0; });
+    }
+    ASSERT(glClientWaitSync(built_fence.handle, 0, GL_TIMEOUT_IGNORED) != GL_WAIT_FAILED);
+    is_built = true;
+}
+
+bool GraphicsPipeline::IsBuilt() noexcept {
+    if (is_built) {
+        return true;
+    }
+    if (built_fence.handle == 0) {
+        return false;
+    }
+    // Timeout of zero means this is non-blocking
+    const auto sync_status = glClientWaitSync(built_fence.handle, 0, 0);
+    ASSERT(sync_status != GL_WAIT_FAILED);
+    is_built = sync_status != GL_TIMEOUT_EXPIRED;
+    return is_built;
 }
 
 } // namespace OpenGL
diff --git a/src/video_core/renderer_opengl/gl_graphics_pipeline.h b/src/video_core/renderer_opengl/gl_graphics_pipeline.h
index 4e28d9a42..311d49f3f 100644
--- a/src/video_core/renderer_opengl/gl_graphics_pipeline.h
+++ b/src/video_core/renderer_opengl/gl_graphics_pipeline.h
@@ -100,9 +100,7 @@ public:
         return writes_global_memory;
     }
 
-    [[nodiscard]] bool IsBuilt() const noexcept {
-        return is_built.load(std::memory_order::relaxed);
-    }
+    [[nodiscard]] bool IsBuilt() noexcept;
 
     template <typename Spec>
     static auto MakeConfigureSpecFunc() {
@@ -154,7 +152,8 @@ private:
 
     std::mutex built_mutex;
     std::condition_variable built_condvar;
-    std::atomic_bool is_built{false};
+    OGLSync built_fence{};
+    bool is_built{false};
 };
 
 } // 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 b2d5bfd3b..84e07f8bd 100644
--- a/src/video_core/renderer_opengl/gl_resource_manager.h
+++ b/src/video_core/renderer_opengl/gl_resource_manager.h
@@ -7,12 +7,14 @@
 #include <string_view>
 #include <utility>
 #include <glad/glad.h>
-#include "common/common_types.h"
+#include "common/common_funcs.h"
 
 namespace OpenGL {
 
-class OGLRenderbuffer : private NonCopyable {
+class OGLRenderbuffer final {
 public:
+    YUZU_NON_COPYABLE(OGLRenderbuffer);
+
     OGLRenderbuffer() = default;
 
     OGLRenderbuffer(OGLRenderbuffer&& o) noexcept : handle(std::exchange(o.handle, 0)) {}
@@ -36,8 +38,10 @@ public:
     GLuint handle = 0;
 };
 
-class OGLTexture : private NonCopyable {
+class OGLTexture final {
 public:
+    YUZU_NON_COPYABLE(OGLTexture);
+
     OGLTexture() = default;
 
     OGLTexture(OGLTexture&& o) noexcept : handle(std::exchange(o.handle, 0)) {}
@@ -61,8 +65,10 @@ public:
     GLuint handle = 0;
 };
 
-class OGLTextureView : private NonCopyable {
+class OGLTextureView final {
 public:
+    YUZU_NON_COPYABLE(OGLTextureView);
+
     OGLTextureView() = default;
 
     OGLTextureView(OGLTextureView&& o) noexcept : handle(std::exchange(o.handle, 0)) {}
@@ -86,8 +92,10 @@ public:
     GLuint handle = 0;
 };
 
-class OGLSampler : private NonCopyable {
+class OGLSampler final {
 public:
+    YUZU_NON_COPYABLE(OGLSampler);
+
     OGLSampler() = default;
 
     OGLSampler(OGLSampler&& o) noexcept : handle(std::exchange(o.handle, 0)) {}
@@ -111,8 +119,10 @@ public:
     GLuint handle = 0;
 };
 
-class OGLShader : private NonCopyable {
+class OGLShader final {
 public:
+    YUZU_NON_COPYABLE(OGLShader);
+
     OGLShader() = default;
 
     OGLShader(OGLShader&& o) noexcept : handle(std::exchange(o.handle, 0)) {}
@@ -132,8 +142,10 @@ public:
     GLuint handle = 0;
 };
 
-class OGLProgram : private NonCopyable {
+class OGLProgram final {
 public:
+    YUZU_NON_COPYABLE(OGLProgram);
+
     OGLProgram() = default;
 
     OGLProgram(OGLProgram&& o) noexcept : handle(std::exchange(o.handle, 0)) {}
@@ -154,8 +166,10 @@ public:
     GLuint handle = 0;
 };
 
-class OGLAssemblyProgram : private NonCopyable {
+class OGLAssemblyProgram final {
 public:
+    YUZU_NON_COPYABLE(OGLAssemblyProgram);
+
     OGLAssemblyProgram() = default;
 
     OGLAssemblyProgram(OGLAssemblyProgram&& o) noexcept : handle(std::exchange(o.handle, 0)) {}
@@ -176,8 +190,10 @@ public:
     GLuint handle = 0;
 };
 
-class OGLPipeline : private NonCopyable {
+class OGLPipeline final {
 public:
+    YUZU_NON_COPYABLE(OGLPipeline);
+
     OGLPipeline() = default;
     OGLPipeline(OGLPipeline&& o) noexcept : handle{std::exchange<GLuint>(o.handle, 0)} {}
 
@@ -198,8 +214,10 @@ public:
     GLuint handle = 0;
 };
 
-class OGLBuffer : private NonCopyable {
+class OGLBuffer final {
 public:
+    YUZU_NON_COPYABLE(OGLBuffer);
+
     OGLBuffer() = default;
 
     OGLBuffer(OGLBuffer&& o) noexcept : handle(std::exchange(o.handle, 0)) {}
@@ -223,8 +241,10 @@ public:
     GLuint handle = 0;
 };
 
-class OGLSync : private NonCopyable {
+class OGLSync final {
 public:
+    YUZU_NON_COPYABLE(OGLSync);
+
     OGLSync() = default;
 
     OGLSync(OGLSync&& o) noexcept : handle(std::exchange(o.handle, nullptr)) {}
@@ -247,8 +267,10 @@ public:
     GLsync handle = 0;
 };
 
-class OGLFramebuffer : private NonCopyable {
+class OGLFramebuffer final {
 public:
+    YUZU_NON_COPYABLE(OGLFramebuffer);
+
     OGLFramebuffer() = default;
 
     OGLFramebuffer(OGLFramebuffer&& o) noexcept : handle(std::exchange(o.handle, 0)) {}
@@ -272,8 +294,10 @@ public:
     GLuint handle = 0;
 };
 
-class OGLQuery : private NonCopyable {
+class OGLQuery final {
 public:
+    YUZU_NON_COPYABLE(OGLQuery);
+
     OGLQuery() = default;
 
     OGLQuery(OGLQuery&& o) noexcept : handle(std::exchange(o.handle, 0)) {}
diff --git a/src/video_core/renderer_opengl/gl_state_tracker.h b/src/video_core/renderer_opengl/gl_state_tracker.h
index 5864c7c07..550ed6d36 100644
--- a/src/video_core/renderer_opengl/gl_state_tracker.h
+++ b/src/video_core/renderer_opengl/gl_state_tracker.h
@@ -9,7 +9,6 @@
 #include <glad/glad.h>
 
 #include "common/common_types.h"
-#include "core/core.h"
 #include "video_core/dirty_flags.h"
 #include "video_core/engines/maxwell_3d.h"
 
diff --git a/src/video_core/renderer_opengl/maxwell_to_gl.h b/src/video_core/renderer_opengl/maxwell_to_gl.h
index daba42ed9..db5bf1d30 100644
--- a/src/video_core/renderer_opengl/maxwell_to_gl.h
+++ b/src/video_core/renderer_opengl/maxwell_to_gl.h
@@ -184,6 +184,8 @@ inline GLenum VertexFormat(Maxwell::VertexAttribute attrib) {
         case Maxwell::VertexAttribute::Size::Size_32_32_32:
         case Maxwell::VertexAttribute::Size::Size_32_32_32_32:
             return GL_FLOAT;
+        case Maxwell::VertexAttribute::Size::Size_11_11_10:
+            return GL_UNSIGNED_INT_10F_11F_11F_REV;
         default:
             break;
         }
diff --git a/src/video_core/renderer_vulkan/blit_image.cpp b/src/video_core/renderer_vulkan/blit_image.cpp
index 2c3914459..ec03cca38 100644
--- a/src/video_core/renderer_vulkan/blit_image.cpp
+++ b/src/video_core/renderer_vulkan/blit_image.cpp
@@ -9,6 +9,7 @@
 #include "video_core/host_shaders/convert_d24s8_to_abgr8_frag_spv.h"
 #include "video_core/host_shaders/convert_depth_to_float_frag_spv.h"
 #include "video_core/host_shaders/convert_float_to_depth_frag_spv.h"
+#include "video_core/host_shaders/convert_s8d24_to_abgr8_frag_spv.h"
 #include "video_core/host_shaders/full_screen_triangle_vert_spv.h"
 #include "video_core/host_shaders/vulkan_blit_color_float_frag_spv.h"
 #include "video_core/host_shaders/vulkan_blit_depth_stencil_frag_spv.h"
@@ -370,6 +371,7 @@ BlitImageHelper::BlitImageHelper(const Device& device_, VKScheduler& scheduler_,
       convert_float_to_depth_frag(BuildShader(device, CONVERT_FLOAT_TO_DEPTH_FRAG_SPV)),
       convert_abgr8_to_d24s8_frag(BuildShader(device, CONVERT_ABGR8_TO_D24S8_FRAG_SPV)),
       convert_d24s8_to_abgr8_frag(BuildShader(device, CONVERT_D24S8_TO_ABGR8_FRAG_SPV)),
+      convert_s8d24_to_abgr8_frag(BuildShader(device, CONVERT_S8D24_TO_ABGR8_FRAG_SPV)),
       linear_sampler(device.GetLogical().CreateSampler(SAMPLER_CREATE_INFO<VK_FILTER_LINEAR>)),
       nearest_sampler(device.GetLogical().CreateSampler(SAMPLER_CREATE_INFO<VK_FILTER_NEAREST>)) {
     if (device.IsExtShaderStencilExportSupported()) {
@@ -474,6 +476,13 @@ void BlitImageHelper::ConvertD24S8ToABGR8(const Framebuffer* dst_framebuffer,
     ConvertDepthStencil(*convert_d24s8_to_abgr8_pipeline, dst_framebuffer, src_image_view);
 }
 
+void BlitImageHelper::ConvertS8D24ToABGR8(const Framebuffer* dst_framebuffer,
+                                          ImageView& src_image_view) {
+    ConvertPipelineColorTargetEx(convert_s8d24_to_abgr8_pipeline, dst_framebuffer->RenderPass(),
+                                 convert_s8d24_to_abgr8_frag);
+    ConvertDepthStencil(*convert_s8d24_to_abgr8_pipeline, dst_framebuffer, src_image_view);
+}
+
 void BlitImageHelper::Convert(VkPipeline pipeline, const Framebuffer* dst_framebuffer,
                               const ImageView& src_image_view) {
     const VkPipelineLayout layout = *one_texture_pipeline_layout;
diff --git a/src/video_core/renderer_vulkan/blit_image.h b/src/video_core/renderer_vulkan/blit_image.h
index 85e7dca5b..1a3944179 100644
--- a/src/video_core/renderer_vulkan/blit_image.h
+++ b/src/video_core/renderer_vulkan/blit_image.h
@@ -56,6 +56,8 @@ public:
 
     void ConvertD24S8ToABGR8(const Framebuffer* dst_framebuffer, ImageView& src_image_view);
 
+    void ConvertS8D24ToABGR8(const Framebuffer* dst_framebuffer, ImageView& src_image_view);
+
 private:
     void Convert(VkPipeline pipeline, const Framebuffer* dst_framebuffer,
                  const ImageView& src_image_view);
@@ -99,6 +101,7 @@ private:
     vk::ShaderModule convert_float_to_depth_frag;
     vk::ShaderModule convert_abgr8_to_d24s8_frag;
     vk::ShaderModule convert_d24s8_to_abgr8_frag;
+    vk::ShaderModule convert_s8d24_to_abgr8_frag;
     vk::Sampler linear_sampler;
     vk::Sampler nearest_sampler;
 
@@ -112,6 +115,7 @@ private:
     vk::Pipeline convert_r16_to_d16_pipeline;
     vk::Pipeline convert_abgr8_to_d24s8_pipeline;
     vk::Pipeline convert_d24s8_to_abgr8_pipeline;
+    vk::Pipeline convert_s8d24_to_abgr8_pipeline;
 };
 
 } // namespace Vulkan
diff --git a/src/video_core/renderer_vulkan/maxwell_to_vk.cpp b/src/video_core/renderer_vulkan/maxwell_to_vk.cpp
index 751e4792b..1c136c410 100644
--- a/src/video_core/renderer_vulkan/maxwell_to_vk.cpp
+++ b/src/video_core/renderer_vulkan/maxwell_to_vk.cpp
@@ -495,6 +495,8 @@ VkFormat VertexFormat(Maxwell::VertexAttribute::Type type, Maxwell::VertexAttrib
             return VK_FORMAT_R32G32B32_SFLOAT;
         case Maxwell::VertexAttribute::Size::Size_32_32_32_32:
             return VK_FORMAT_R32G32B32A32_SFLOAT;
+        case Maxwell::VertexAttribute::Size::Size_11_11_10:
+            return VK_FORMAT_B10G11R11_UFLOAT_PACK32;
         default:
             break;
         }
diff --git a/src/video_core/renderer_vulkan/vk_blit_screen.cpp b/src/video_core/renderer_vulkan/vk_blit_screen.cpp
index c71a1f44d..621a6a071 100644
--- a/src/video_core/renderer_vulkan/vk_blit_screen.cpp
+++ b/src/video_core/renderer_vulkan/vk_blit_screen.cpp
@@ -100,6 +100,8 @@ VkFormat GetFormat(const Tegra::FramebufferConfig& framebuffer) {
         return VK_FORMAT_A8B8G8R8_UNORM_PACK32;
     case Tegra::FramebufferConfig::PixelFormat::RGB565_UNORM:
         return VK_FORMAT_R5G6B5_UNORM_PACK16;
+    case Tegra::FramebufferConfig::PixelFormat::B8G8R8A8_UNORM:
+        return VK_FORMAT_B8G8R8A8_UNORM;
     default:
         UNIMPLEMENTED_MSG("Unknown framebuffer pixel format: {}",
                           static_cast<u32>(framebuffer.pixel_format));
diff --git a/src/video_core/renderer_vulkan/vk_compute_pass.cpp b/src/video_core/renderer_vulkan/vk_compute_pass.cpp
index 3e96c0f60..4d73427b4 100644
--- a/src/video_core/renderer_vulkan/vk_compute_pass.cpp
+++ b/src/video_core/renderer_vulkan/vk_compute_pass.cpp
@@ -2,6 +2,7 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
+#include <array>
 #include <cstring>
 #include <memory>
 #include <optional>
@@ -292,7 +293,7 @@ std::pair<VkBuffer, VkDeviceSize> QuadIndexedPass::Assemble(
             .srcAccessMask = VK_ACCESS_SHADER_WRITE_BIT,
             .dstAccessMask = VK_ACCESS_INDEX_READ_BIT,
         };
-        const std::array push_constants{base_vertex, index_shift};
+        const std::array<u32, 2> push_constants{base_vertex, index_shift};
         const VkDescriptorSet set = descriptor_allocator.Commit();
         device.GetLogical().UpdateDescriptorSet(set, *descriptor_template, descriptor_data);
         cmdbuf.BindPipeline(VK_PIPELINE_BIND_POINT_COMPUTE, *pipeline);
diff --git a/src/video_core/renderer_vulkan/vk_scheduler.cpp b/src/video_core/renderer_vulkan/vk_scheduler.cpp
index 3bfdf41ba..7d9d4f7ba 100644
--- a/src/video_core/renderer_vulkan/vk_scheduler.cpp
+++ b/src/video_core/renderer_vulkan/vk_scheduler.cpp
@@ -140,12 +140,12 @@ bool VKScheduler::UpdateRescaling(bool is_rescaling) {
 void VKScheduler::WorkerThread(std::stop_token stop_token) {
     Common::SetCurrentThreadName("yuzu:VulkanWorker");
     do {
-        if (work_queue.empty()) {
-            wait_cv.notify_all();
-        }
         std::unique_ptr<CommandChunk> work;
         {
             std::unique_lock lock{work_mutex};
+            if (work_queue.empty()) {
+                wait_cv.notify_all();
+            }
             work_cv.wait(lock, stop_token, [this] { return !work_queue.empty(); });
             if (stop_token.stop_requested()) {
                 continue;
diff --git a/src/video_core/renderer_vulkan/vk_scheduler.h b/src/video_core/renderer_vulkan/vk_scheduler.h
index 1b06c9296..e69aa136b 100644
--- a/src/video_core/renderer_vulkan/vk_scheduler.h
+++ b/src/video_core/renderer_vulkan/vk_scheduler.h
@@ -146,6 +146,7 @@ private:
             using FuncType = TypedCommand<T>;
             static_assert(sizeof(FuncType) < sizeof(data), "Lambda is too large");
 
+            recorded_counts++;
             command_offset = Common::AlignUp(command_offset, alignof(FuncType));
             if (command_offset > sizeof(data) - sizeof(FuncType)) {
                 return false;
@@ -167,7 +168,7 @@ private:
         }
 
         bool Empty() const {
-            return command_offset == 0;
+            return recorded_counts == 0;
         }
 
         bool HasSubmit() const {
@@ -178,6 +179,7 @@ private:
         Command* first = nullptr;
         Command* last = nullptr;
 
+        size_t recorded_counts = 0;
         size_t command_offset = 0;
         bool submit = false;
         alignas(std::max_align_t) std::array<u8, 0x8000> data{};
diff --git a/src/video_core/renderer_vulkan/vk_state_tracker.h b/src/video_core/renderer_vulkan/vk_state_tracker.h
index 40a149832..8240c83e1 100644
--- a/src/video_core/renderer_vulkan/vk_state_tracker.h
+++ b/src/video_core/renderer_vulkan/vk_state_tracker.h
@@ -8,7 +8,6 @@
 #include <limits>
 
 #include "common/common_types.h"
-#include "core/core.h"
 #include "video_core/dirty_flags.h"
 #include "video_core/engines/maxwell_3d.h"
 
diff --git a/src/video_core/renderer_vulkan/vk_texture_cache.cpp b/src/video_core/renderer_vulkan/vk_texture_cache.cpp
index 0ba56ff1e..8101eb42c 100644
--- a/src/video_core/renderer_vulkan/vk_texture_cache.cpp
+++ b/src/video_core/renderer_vulkan/vk_texture_cache.cpp
@@ -554,10 +554,12 @@ void CopyBufferToImage(vk::CommandBuffer cmdbuf, VkBuffer src_buffer, VkImage im
     };
 }
 
-[[nodiscard]] bool IsFormatFlipped(PixelFormat format) {
+[[nodiscard]] bool IsFormatFlipped(PixelFormat format, bool emulate_bgr565) {
     switch (format) {
     case PixelFormat::A1B5G5R5_UNORM:
         return true;
+    case PixelFormat::B5G6R5_UNORM:
+        return emulate_bgr565;
     default:
         return false;
     }
@@ -1068,6 +1070,9 @@ void TextureCacheRuntime::ConvertImage(Framebuffer* dst, ImageView& dst_view, Im
         if (src_view.format == PixelFormat::S8_UINT_D24_UNORM) {
             return blit_image_helper.ConvertD24S8ToABGR8(dst, src_view);
         }
+        if (src_view.format == PixelFormat::D24_UNORM_S8_UINT) {
+            return blit_image_helper.ConvertS8D24ToABGR8(dst, src_view);
+        }
         break;
     case PixelFormat::R32_FLOAT:
         if (src_view.format == PixelFormat::D32_FLOAT) {
@@ -1488,7 +1493,7 @@ ImageView::ImageView(TextureCacheRuntime& runtime, const VideoCommon::ImageViewI
     };
     if (!info.IsRenderTarget()) {
         swizzle = info.Swizzle();
-        if (IsFormatFlipped(format)) {
+        if (IsFormatFlipped(format, device->MustEmulateBGR565())) {
             std::ranges::transform(swizzle, swizzle.begin(), SwapBlueRed);
         }
         if ((aspect_mask & (VK_IMAGE_ASPECT_DEPTH_BIT | VK_IMAGE_ASPECT_STENCIL_BIT)) != 0) {
diff --git a/src/video_core/video_core.cpp b/src/video_core/video_core.cpp
index 329bf4def..2f2594585 100644
--- a/src/video_core/video_core.cpp
+++ b/src/video_core/video_core.cpp
@@ -50,6 +50,7 @@ std::unique_ptr<Tegra::GPU> CreateGPU(Core::Frontend::EmuWindow& emu_window, Cor
         gpu->BindRenderer(std::move(renderer));
         return gpu;
     } catch (const std::runtime_error& exception) {
+        scope.Cancel();
         LOG_ERROR(HW_GPU, "Failed to initialize GPU: {}", exception.what());
         return nullptr;
     }
diff --git a/src/video_core/vulkan_common/vulkan_device.cpp b/src/video_core/vulkan_common/vulkan_device.cpp
index 153702c0b..effde73c9 100644
--- a/src/video_core/vulkan_common/vulkan_device.cpp
+++ b/src/video_core/vulkan_common/vulkan_device.cpp
@@ -39,6 +39,11 @@ constexpr std::array DEPTH16_UNORM_STENCIL8_UINT{
     VK_FORMAT_D32_SFLOAT_S8_UINT,
     VK_FORMAT_UNDEFINED,
 };
+
+constexpr std::array B5G6R5_UNORM_PACK16{
+    VK_FORMAT_R5G6B5_UNORM_PACK16,
+    VK_FORMAT_UNDEFINED,
+};
 } // namespace Alternatives
 
 enum class NvidiaArchitecture {
@@ -87,6 +92,8 @@ constexpr const VkFormat* GetFormatAlternatives(VkFormat format) {
         return Alternatives::DEPTH24_UNORM_STENCIL8_UINT.data();
     case VK_FORMAT_D16_UNORM_S8_UINT:
         return Alternatives::DEPTH16_UNORM_STENCIL8_UINT.data();
+    case VK_FORMAT_B5G6R5_UNORM_PACK16:
+        return Alternatives::B5G6R5_UNORM_PACK16.data();
     default:
         return nullptr;
     }
@@ -224,9 +231,14 @@ std::vector<std::string> GetSupportedExtensions(vk::PhysicalDevice physical) {
     return supported_extensions;
 }
 
+bool IsExtensionSupported(std::span<const std::string> supported_extensions,
+                          std::string_view extension) {
+    return std::ranges::find(supported_extensions, extension) != supported_extensions.end();
+}
+
 NvidiaArchitecture GetNvidiaArchitecture(vk::PhysicalDevice physical,
                                          std::span<const std::string> exts) {
-    if (std::ranges::find(exts, VK_KHR_FRAGMENT_SHADING_RATE_EXTENSION_NAME) != exts.end()) {
+    if (IsExtensionSupported(exts, VK_KHR_FRAGMENT_SHADING_RATE_EXTENSION_NAME)) {
         VkPhysicalDeviceFragmentShadingRatePropertiesKHR shading_rate_props{};
         shading_rate_props.sType =
             VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_FRAGMENT_SHADING_RATE_PROPERTIES_KHR;
@@ -239,7 +251,7 @@ NvidiaArchitecture GetNvidiaArchitecture(vk::PhysicalDevice physical,
             return NvidiaArchitecture::AmpereOrNewer;
         }
     }
-    if (std::ranges::find(exts, VK_NV_SHADING_RATE_IMAGE_EXTENSION_NAME) != exts.end()) {
+    if (IsExtensionSupported(exts, VK_NV_SHADING_RATE_IMAGE_EXTENSION_NAME)) {
         return NvidiaArchitecture::Turing;
     }
     return NvidiaArchitecture::VoltaOrOlder;
@@ -604,7 +616,8 @@ Device::Device(VkInstance instance_, vk::PhysicalDevice physical_, VkSurfaceKHR
             break;
         }
     }
-    if (ext_extended_dynamic_state && driver_id == VK_DRIVER_ID_MESA_RADV) {
+    const bool is_radv = driver_id == VK_DRIVER_ID_MESA_RADV;
+    if (ext_extended_dynamic_state && is_radv) {
         // Mask driver version variant
         const u32 version = (properties.driverVersion << 3) >> 3;
         if (version < VK_MAKE_API_VERSION(0, 21, 2, 0)) {
@@ -613,6 +626,17 @@ Device::Device(VkInstance instance_, vk::PhysicalDevice physical_, VkSurfaceKHR
             ext_extended_dynamic_state = false;
         }
     }
+    if (ext_vertex_input_dynamic_state && is_radv) {
+        // TODO(ameerj): Blacklist only offending driver versions
+        // TODO(ameerj): Confirm if RDNA1 is affected
+        const bool is_rdna2 =
+            IsExtensionSupported(supported_extensions, VK_KHR_FRAGMENT_SHADING_RATE_EXTENSION_NAME);
+        if (is_rdna2) {
+            LOG_WARNING(Render_Vulkan,
+                        "RADV has broken VK_EXT_vertex_input_dynamic_state on RDNA2 hardware");
+            ext_vertex_input_dynamic_state = false;
+        }
+    }
     sets_per_pool = 64;
 
     const bool is_amd =
@@ -628,7 +652,7 @@ Device::Device(VkInstance instance_, vk::PhysicalDevice physical_, VkSurfaceKHR
             has_broken_cube_compatibility = true;
         }
     }
-    const bool is_amd_or_radv = is_amd || driver_id == VK_DRIVER_ID_MESA_RADV;
+    const bool is_amd_or_radv = is_amd || is_radv;
     if (ext_sampler_filter_minmax && is_amd_or_radv) {
         // Disable ext_sampler_filter_minmax on AMD GCN4 and lower as it is broken.
         if (!is_float16_supported) {
@@ -639,6 +663,7 @@ Device::Device(VkInstance instance_, vk::PhysicalDevice physical_, VkSurfaceKHR
     }
 
     const bool is_intel_windows = driver_id == VK_DRIVER_ID_INTEL_PROPRIETARY_WINDOWS;
+    const bool is_intel_anv = driver_id == VK_DRIVER_ID_INTEL_OPEN_SOURCE_MESA;
     if (ext_vertex_input_dynamic_state && is_intel_windows) {
         LOG_WARNING(Render_Vulkan, "Blacklisting Intel for VK_EXT_vertex_input_dynamic_state");
         ext_vertex_input_dynamic_state = false;
@@ -652,6 +677,10 @@ Device::Device(VkInstance instance_, vk::PhysicalDevice physical_, VkSurfaceKHR
         LOG_WARNING(Render_Vulkan, "Intel proprietary drivers do not support MSAA image blits");
         cant_blit_msaa = true;
     }
+    if (is_intel_anv) {
+        LOG_WARNING(Render_Vulkan, "ANV driver does not support native BGR format");
+        must_emulate_bgr565 = true;
+    }
 
     supports_d24_depth =
         IsFormatSupported(VK_FORMAT_D24_UNORM_S8_UINT,
diff --git a/src/video_core/vulkan_common/vulkan_device.h b/src/video_core/vulkan_common/vulkan_device.h
index 37d140ebd..34b1add16 100644
--- a/src/video_core/vulkan_common/vulkan_device.h
+++ b/src/video_core/vulkan_common/vulkan_device.h
@@ -354,6 +354,10 @@ public:
         return cant_blit_msaa;
     }
 
+    bool MustEmulateBGR565() const {
+        return must_emulate_bgr565;
+    }
+
 private:
     /// Checks if the physical device is suitable.
     void CheckSuitability(bool requires_swapchain) const;
@@ -448,6 +452,7 @@ private:
     bool has_nsight_graphics{};             ///< Has Nsight Graphics attached
     bool supports_d24_depth{};              ///< Supports D24 depth buffers.
     bool cant_blit_msaa{};                  ///< Does not support MSAA<->MSAA blitting.
+    bool must_emulate_bgr565{};             ///< Emulates BGR565 by swizzling RGB565 format.
 
     // Telemetry parameters
     std::string vendor_name;                       ///< Device's driver name.
diff --git a/src/yuzu/applets/qt_controller.cpp b/src/yuzu/applets/qt_controller.cpp
index 4239c17f5..4104928d1 100644
--- a/src/yuzu/applets/qt_controller.cpp
+++ b/src/yuzu/applets/qt_controller.cpp
@@ -257,7 +257,7 @@ void QtControllerSelectorDialog::LoadConfiguration() {
 }
 
 void QtControllerSelectorDialog::CallConfigureVibrationDialog() {
-    ConfigureVibration dialog(this);
+    ConfigureVibration dialog(this, system.HIDCore());
 
     dialog.setWindowFlags(Qt::Dialog | Qt::CustomizeWindowHint | Qt::WindowTitleHint |
                           Qt::WindowSystemMenuHint);
diff --git a/src/yuzu/applets/qt_profile_select.cpp b/src/yuzu/applets/qt_profile_select.cpp
index 5b32da923..4cd8f7784 100644
--- a/src/yuzu/applets/qt_profile_select.cpp
+++ b/src/yuzu/applets/qt_profile_select.cpp
@@ -23,13 +23,13 @@ QString FormatUserEntryText(const QString& username, Common::UUID uuid) {
     return QtProfileSelectionDialog::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()));
+        .arg(username, QString::fromStdString(uuid.FormattedString()));
 }
 
 QString GetImagePath(Common::UUID uuid) {
     const auto path =
         Common::FS::GetYuzuPath(Common::FS::YuzuPath::NANDDir) /
-        fmt::format("system/save/8000000000000010/su/avators/{}.jpg", uuid.FormatSwitch());
+        fmt::format("system/save/8000000000000010/su/avators/{}.jpg", uuid.FormattedString());
     return QString::fromStdString(Common::FS::PathToUTF8String(path));
 }
 
diff --git a/src/yuzu/configuration/config.cpp b/src/yuzu/configuration/config.cpp
index 2c8c10c50..4b943c6ba 100644
--- a/src/yuzu/configuration/config.cpp
+++ b/src/yuzu/configuration/config.cpp
@@ -65,23 +65,25 @@ const std::array<int, 2> Config::default_stick_mod = {
 // This must be in alphabetical order according to action name as it must have the same order as
 // UISetting::values.shortcuts, which is alphabetically ordered.
 // clang-format off
-const std::array<UISettings::Shortcut, 20> Config::default_hotkeys{{
+const std::array<UISettings::Shortcut, 22> Config::default_hotkeys{{
     {QStringLiteral("Audio Mute/Unmute"),        QStringLiteral("Main Window"), {QStringLiteral("Ctrl+M"),  QStringLiteral("Home+Dpad_Right"), Qt::WindowShortcut}},
     {QStringLiteral("Audio Volume Down"),        QStringLiteral("Main Window"), {QStringLiteral("-"),       QStringLiteral("Home+Dpad_Down"), Qt::ApplicationShortcut}},
     {QStringLiteral("Audio Volume Up"),          QStringLiteral("Main Window"), {QStringLiteral("+"),       QStringLiteral("Home+Dpad_Up"), Qt::ApplicationShortcut}},
     {QStringLiteral("Capture Screenshot"),       QStringLiteral("Main Window"), {QStringLiteral("Ctrl+P"),  QStringLiteral("Screenshot"), Qt::WidgetWithChildrenShortcut}},
+    {QStringLiteral("Change Adapting Filter"),   QStringLiteral("Main Window"), {QStringLiteral("F8"),      QStringLiteral("Home+L"), Qt::ApplicationShortcut}},
     {QStringLiteral("Change Docked Mode"),       QStringLiteral("Main Window"), {QStringLiteral("F10"),     QStringLiteral("Home+X"), Qt::ApplicationShortcut}},
+    {QStringLiteral("Change GPU Accuracy"),      QStringLiteral("Main Window"), {QStringLiteral("F9"),      QStringLiteral("Home+R"), Qt::ApplicationShortcut}},
     {QStringLiteral("Continue/Pause Emulation"), QStringLiteral("Main Window"), {QStringLiteral("F4"),      QStringLiteral("Home+Plus"), Qt::WindowShortcut}},
     {QStringLiteral("Exit Fullscreen"),          QStringLiteral("Main Window"), {QStringLiteral("Esc"),     QStringLiteral(""), Qt::WindowShortcut}},
     {QStringLiteral("Exit yuzu"),                QStringLiteral("Main Window"), {QStringLiteral("Ctrl+Q"),  QStringLiteral("Home+Minus"), Qt::WindowShortcut}},
     {QStringLiteral("Fullscreen"),               QStringLiteral("Main Window"), {QStringLiteral("F11"),     QStringLiteral("Home+B"), Qt::WindowShortcut}},
-    {QStringLiteral("Load Amiibo"),              QStringLiteral("Main Window"), {QStringLiteral("F2"),      QStringLiteral("Home+A"), Qt::WidgetWithChildrenShortcut}},
     {QStringLiteral("Load File"),                QStringLiteral("Main Window"), {QStringLiteral("Ctrl+O"),  QStringLiteral(""), Qt::WidgetWithChildrenShortcut}},
+    {QStringLiteral("Load/Remove Amiibo"),       QStringLiteral("Main Window"), {QStringLiteral("F2"),      QStringLiteral("Home+A"), Qt::WidgetWithChildrenShortcut}},
     {QStringLiteral("Restart Emulation"),        QStringLiteral("Main Window"), {QStringLiteral("F6"),      QStringLiteral(""), Qt::WindowShortcut}},
     {QStringLiteral("Stop Emulation"),           QStringLiteral("Main Window"), {QStringLiteral("F5"),      QStringLiteral(""), Qt::WindowShortcut}},
-    {QStringLiteral("TAS Start/Stop"),           QStringLiteral("Main Window"), {QStringLiteral("Ctrl+F5"), QStringLiteral(""), Qt::ApplicationShortcut}},
-    {QStringLiteral("TAS Reset"),                QStringLiteral("Main Window"), {QStringLiteral("Ctrl+F6"), QStringLiteral(""), Qt::ApplicationShortcut}},
     {QStringLiteral("TAS Record"),               QStringLiteral("Main Window"), {QStringLiteral("Ctrl+F7"), QStringLiteral(""), Qt::ApplicationShortcut}},
+    {QStringLiteral("TAS Reset"),                QStringLiteral("Main Window"), {QStringLiteral("Ctrl+F6"), QStringLiteral(""), Qt::ApplicationShortcut}},
+    {QStringLiteral("TAS Start/Stop"),           QStringLiteral("Main Window"), {QStringLiteral("Ctrl+F5"), QStringLiteral(""), Qt::ApplicationShortcut}},
     {QStringLiteral("Toggle Filter Bar"),        QStringLiteral("Main Window"), {QStringLiteral("Ctrl+F"),  QStringLiteral(""), Qt::WindowShortcut}},
     {QStringLiteral("Toggle Framerate Limit"),   QStringLiteral("Main Window"), {QStringLiteral("Ctrl+U"),  QStringLiteral("Home+Y"), Qt::ApplicationShortcut}},
     {QStringLiteral("Toggle Mouse Panning"),     QStringLiteral("Main Window"), {QStringLiteral("Ctrl+F9"), QStringLiteral(""), Qt::ApplicationShortcut}},
@@ -443,6 +445,7 @@ void Config::ReadCoreValues() {
     qt_config->beginGroup(QStringLiteral("Core"));
 
     ReadGlobalSetting(Settings::values.use_multi_core);
+    ReadGlobalSetting(Settings::values.use_extended_memory_layout);
 
     qt_config->endGroup();
 }
@@ -606,6 +609,7 @@ void Config::ReadCpuValues() {
     ReadGlobalSetting(Settings::values.cpuopt_unsafe_ignore_standard_fpcr);
     ReadGlobalSetting(Settings::values.cpuopt_unsafe_inaccurate_nan);
     ReadGlobalSetting(Settings::values.cpuopt_unsafe_fastmem_check);
+    ReadGlobalSetting(Settings::values.cpuopt_unsafe_ignore_global_monitor);
 
     if (global) {
         ReadBasicSetting(Settings::values.cpu_debug_mode);
@@ -618,6 +622,8 @@ void Config::ReadCpuValues() {
         ReadBasicSetting(Settings::values.cpuopt_misc_ir);
         ReadBasicSetting(Settings::values.cpuopt_reduce_misalign_checks);
         ReadBasicSetting(Settings::values.cpuopt_fastmem);
+        ReadBasicSetting(Settings::values.cpuopt_fastmem_exclusives);
+        ReadBasicSetting(Settings::values.cpuopt_recompile_exclusives);
     }
 
     qt_config->endGroup();
@@ -767,6 +773,7 @@ void Config::ReadUIValues() {
     ReadBasicSetting(UISettings::values.callout_flags);
     ReadBasicSetting(UISettings::values.show_console);
     ReadBasicSetting(UISettings::values.pause_when_in_background);
+    ReadBasicSetting(UISettings::values.mute_when_in_background);
     ReadBasicSetting(UISettings::values.hide_mouse);
 
     qt_config->endGroup();
@@ -1016,6 +1023,7 @@ void Config::SaveCoreValues() {
     qt_config->beginGroup(QStringLiteral("Core"));
 
     WriteGlobalSetting(Settings::values.use_multi_core);
+    WriteGlobalSetting(Settings::values.use_extended_memory_layout);
 
     qt_config->endGroup();
 }
@@ -1134,6 +1142,7 @@ void Config::SaveCpuValues() {
     WriteGlobalSetting(Settings::values.cpuopt_unsafe_ignore_standard_fpcr);
     WriteGlobalSetting(Settings::values.cpuopt_unsafe_inaccurate_nan);
     WriteGlobalSetting(Settings::values.cpuopt_unsafe_fastmem_check);
+    WriteGlobalSetting(Settings::values.cpuopt_unsafe_ignore_global_monitor);
 
     if (global) {
         WriteBasicSetting(Settings::values.cpu_debug_mode);
@@ -1146,6 +1155,8 @@ void Config::SaveCpuValues() {
         WriteBasicSetting(Settings::values.cpuopt_misc_ir);
         WriteBasicSetting(Settings::values.cpuopt_reduce_misalign_checks);
         WriteBasicSetting(Settings::values.cpuopt_fastmem);
+        WriteBasicSetting(Settings::values.cpuopt_fastmem_exclusives);
+        WriteBasicSetting(Settings::values.cpuopt_recompile_exclusives);
     }
 
     qt_config->endGroup();
@@ -1295,6 +1306,7 @@ void Config::SaveUIValues() {
     WriteBasicSetting(UISettings::values.callout_flags);
     WriteBasicSetting(UISettings::values.show_console);
     WriteBasicSetting(UISettings::values.pause_when_in_background);
+    WriteBasicSetting(UISettings::values.mute_when_in_background);
     WriteBasicSetting(UISettings::values.hide_mouse);
 
     qt_config->endGroup();
diff --git a/src/yuzu/configuration/config.h b/src/yuzu/configuration/config.h
index 60b20a62f..ae3e36a11 100644
--- a/src/yuzu/configuration/config.h
+++ b/src/yuzu/configuration/config.h
@@ -46,7 +46,7 @@ public:
         default_mouse_buttons;
     static const std::array<int, Settings::NativeKeyboard::NumKeyboardKeys> default_keyboard_keys;
     static const std::array<int, Settings::NativeKeyboard::NumKeyboardMods> default_keyboard_mods;
-    static const std::array<UISettings::Shortcut, 20> default_hotkeys;
+    static const std::array<UISettings::Shortcut, 22> default_hotkeys;
 
     static constexpr UISettings::Theme default_theme{
 #ifdef _WIN32
diff --git a/src/yuzu/configuration/configure_cpu.cpp b/src/yuzu/configuration/configure_cpu.cpp
index f66cab5d4..bf74ccc7c 100644
--- a/src/yuzu/configuration/configure_cpu.cpp
+++ b/src/yuzu/configuration/configure_cpu.cpp
@@ -36,6 +36,7 @@ void ConfigureCpu::SetConfiguration() {
     ui->cpuopt_unsafe_ignore_standard_fpcr->setEnabled(runtime_lock);
     ui->cpuopt_unsafe_inaccurate_nan->setEnabled(runtime_lock);
     ui->cpuopt_unsafe_fastmem_check->setEnabled(runtime_lock);
+    ui->cpuopt_unsafe_ignore_global_monitor->setEnabled(runtime_lock);
 
     ui->cpuopt_unsafe_unfuse_fma->setChecked(Settings::values.cpuopt_unsafe_unfuse_fma.GetValue());
     ui->cpuopt_unsafe_reduce_fp_error->setChecked(
@@ -46,6 +47,8 @@ void ConfigureCpu::SetConfiguration() {
         Settings::values.cpuopt_unsafe_inaccurate_nan.GetValue());
     ui->cpuopt_unsafe_fastmem_check->setChecked(
         Settings::values.cpuopt_unsafe_fastmem_check.GetValue());
+    ui->cpuopt_unsafe_ignore_global_monitor->setChecked(
+        Settings::values.cpuopt_unsafe_ignore_global_monitor.GetValue());
 
     if (Settings::IsConfiguringGlobal()) {
         ui->accuracy->setCurrentIndex(static_cast<int>(Settings::values.cpu_accuracy.GetValue()));
@@ -82,6 +85,9 @@ void ConfigureCpu::ApplyConfiguration() {
     ConfigurationShared::ApplyPerGameSetting(&Settings::values.cpuopt_unsafe_fastmem_check,
                                              ui->cpuopt_unsafe_fastmem_check,
                                              cpuopt_unsafe_fastmem_check);
+    ConfigurationShared::ApplyPerGameSetting(&Settings::values.cpuopt_unsafe_ignore_global_monitor,
+                                             ui->cpuopt_unsafe_ignore_global_monitor,
+                                             cpuopt_unsafe_ignore_global_monitor);
 }
 
 void ConfigureCpu::changeEvent(QEvent* event) {
@@ -120,4 +126,7 @@ void ConfigureCpu::SetupPerGameUI() {
     ConfigurationShared::SetColoredTristate(ui->cpuopt_unsafe_fastmem_check,
                                             Settings::values.cpuopt_unsafe_fastmem_check,
                                             cpuopt_unsafe_fastmem_check);
+    ConfigurationShared::SetColoredTristate(ui->cpuopt_unsafe_ignore_global_monitor,
+                                            Settings::values.cpuopt_unsafe_ignore_global_monitor,
+                                            cpuopt_unsafe_ignore_global_monitor);
 }
diff --git a/src/yuzu/configuration/configure_cpu.h b/src/yuzu/configuration/configure_cpu.h
index ed9af0e9f..733e38be4 100644
--- a/src/yuzu/configuration/configure_cpu.h
+++ b/src/yuzu/configuration/configure_cpu.h
@@ -45,6 +45,7 @@ private:
     ConfigurationShared::CheckState cpuopt_unsafe_ignore_standard_fpcr;
     ConfigurationShared::CheckState cpuopt_unsafe_inaccurate_nan;
     ConfigurationShared::CheckState cpuopt_unsafe_fastmem_check;
+    ConfigurationShared::CheckState cpuopt_unsafe_ignore_global_monitor;
 
     const Core::System& system;
 };
diff --git a/src/yuzu/configuration/configure_cpu.ui b/src/yuzu/configuration/configure_cpu.ui
index d8064db24..5d80a8c91 100644
--- a/src/yuzu/configuration/configure_cpu.ui
+++ b/src/yuzu/configuration/configure_cpu.ui
@@ -150,6 +150,18 @@
           </property>
          </widget>
         </item>
+        <item>
+         <widget class="QCheckBox" name="cpuopt_unsafe_ignore_global_monitor">
+          <property name="toolTip">
+           <string>
+            &lt;div&gt;This option improves speed by relying only on the semantics of cmpxchg to ensure safety of exclusive access instructions. Please note this may result in deadlocks and other race conditions.&lt;/div&gt;
+           </string>
+          </property>
+          <property name="text">
+           <string>Ignore global monitor</string>
+          </property>
+         </widget>
+        </item>
        </layout>
       </widget>
      </item>
diff --git a/src/yuzu/configuration/configure_cpu_debug.cpp b/src/yuzu/configuration/configure_cpu_debug.cpp
index 05a90963d..616a0be75 100644
--- a/src/yuzu/configuration/configure_cpu_debug.cpp
+++ b/src/yuzu/configuration/configure_cpu_debug.cpp
@@ -44,6 +44,12 @@ void ConfigureCpuDebug::SetConfiguration() {
         Settings::values.cpuopt_reduce_misalign_checks.GetValue());
     ui->cpuopt_fastmem->setEnabled(runtime_lock);
     ui->cpuopt_fastmem->setChecked(Settings::values.cpuopt_fastmem.GetValue());
+    ui->cpuopt_fastmem_exclusives->setEnabled(runtime_lock);
+    ui->cpuopt_fastmem_exclusives->setChecked(
+        Settings::values.cpuopt_fastmem_exclusives.GetValue());
+    ui->cpuopt_recompile_exclusives->setEnabled(runtime_lock);
+    ui->cpuopt_recompile_exclusives->setChecked(
+        Settings::values.cpuopt_recompile_exclusives.GetValue());
 }
 
 void ConfigureCpuDebug::ApplyConfiguration() {
@@ -56,6 +62,8 @@ void ConfigureCpuDebug::ApplyConfiguration() {
     Settings::values.cpuopt_misc_ir = ui->cpuopt_misc_ir->isChecked();
     Settings::values.cpuopt_reduce_misalign_checks = ui->cpuopt_reduce_misalign_checks->isChecked();
     Settings::values.cpuopt_fastmem = ui->cpuopt_fastmem->isChecked();
+    Settings::values.cpuopt_fastmem_exclusives = ui->cpuopt_fastmem_exclusives->isChecked();
+    Settings::values.cpuopt_recompile_exclusives = ui->cpuopt_recompile_exclusives->isChecked();
 }
 
 void ConfigureCpuDebug::changeEvent(QEvent* event) {
diff --git a/src/yuzu/configuration/configure_cpu_debug.ui b/src/yuzu/configuration/configure_cpu_debug.ui
index 6e635bb2f..2bc268810 100644
--- a/src/yuzu/configuration/configure_cpu_debug.ui
+++ b/src/yuzu/configuration/configure_cpu_debug.ui
@@ -144,7 +144,34 @@
            </string>
           </property>
           <property name="text">
-           <string>Enable Host MMU Emulation</string>
+           <string>Enable Host MMU Emulation (general memory instructions)</string>
+          </property>
+         </widget>
+        </item>
+        <item>
+         <widget class="QCheckBox" name="cpuopt_fastmem_exclusives">
+          <property name="toolTip">
+           <string>
+            &lt;div style=&quot;white-space: nowrap&quot;&gt;This optimization speeds up exclusive memory accesses by the guest program.&lt;/div&gt;
+            &lt;div style=&quot;white-space: nowrap&quot;&gt;Enabling it causes guest exclusive memory reads/writes to be done directly into memory and make use of Host's MMU.&lt;/div&gt;
+            &lt;div style=&quot;white-space: nowrap&quot;&gt;Disabling this forces all exclusive memory accesses to use Software MMU Emulation.&lt;/div&gt;
+           </string>
+          </property>
+          <property name="text">
+           <string>Enable Host MMU Emulation (exclusive memory instructions)</string>
+          </property>
+         </widget>
+        </item>
+        <item>
+         <widget class="QCheckBox" name="cpuopt_recompile_exclusives">
+          <property name="toolTip">
+           <string>
+            &lt;div style=&quot;white-space: nowrap&quot;&gt;This optimization speeds up exclusive memory accesses by the guest program.&lt;/div&gt;
+            &lt;div style=&quot;white-space: nowrap&quot;&gt;Enabling it reduces the overhead of fastmem failure of exclusive memory accesses.&lt;/div&gt;
+           </string>
+          </property>
+          <property name="text">
+           <string>Enable recompilation of exclusive memory instructions</string>
           </property>
          </widget>
         </item>
diff --git a/src/yuzu/configuration/configure_general.cpp b/src/yuzu/configuration/configure_general.cpp
index 566879317..08d5444ec 100644
--- a/src/yuzu/configuration/configure_general.cpp
+++ b/src/yuzu/configuration/configure_general.cpp
@@ -42,10 +42,14 @@ void ConfigureGeneral::SetConfiguration() {
 
     ui->use_multi_core->setEnabled(runtime_lock);
     ui->use_multi_core->setChecked(Settings::values.use_multi_core.GetValue());
+    ui->use_extended_memory_layout->setEnabled(runtime_lock);
+    ui->use_extended_memory_layout->setChecked(
+        Settings::values.use_extended_memory_layout.GetValue());
 
     ui->toggle_check_exit->setChecked(UISettings::values.confirm_before_closing.GetValue());
     ui->toggle_user_on_boot->setChecked(UISettings::values.select_user_on_boot.GetValue());
     ui->toggle_background_pause->setChecked(UISettings::values.pause_when_in_background.GetValue());
+    ui->toggle_background_mute->setChecked(UISettings::values.mute_when_in_background.GetValue());
     ui->toggle_hide_mouse->setChecked(UISettings::values.hide_mouse.GetValue());
 
     ui->toggle_speed_limit->setChecked(Settings::values.use_speed_limit.GetValue());
@@ -90,11 +94,15 @@ void ConfigureGeneral::ResetDefaults() {
 void ConfigureGeneral::ApplyConfiguration() {
     ConfigurationShared::ApplyPerGameSetting(&Settings::values.use_multi_core, ui->use_multi_core,
                                              use_multi_core);
+    ConfigurationShared::ApplyPerGameSetting(&Settings::values.use_extended_memory_layout,
+                                             ui->use_extended_memory_layout,
+                                             use_extended_memory_layout);
 
     if (Settings::IsConfiguringGlobal()) {
         UISettings::values.confirm_before_closing = ui->toggle_check_exit->isChecked();
         UISettings::values.select_user_on_boot = ui->toggle_user_on_boot->isChecked();
         UISettings::values.pause_when_in_background = ui->toggle_background_pause->isChecked();
+        UISettings::values.mute_when_in_background = ui->toggle_background_mute->isChecked();
         UISettings::values.hide_mouse = ui->toggle_hide_mouse->isChecked();
 
         Settings::values.fps_cap.SetValue(ui->fps_cap->value());
@@ -158,6 +166,9 @@ void ConfigureGeneral::SetupPerGameUI() {
                                             Settings::values.use_speed_limit, use_speed_limit);
     ConfigurationShared::SetColoredTristate(ui->use_multi_core, Settings::values.use_multi_core,
                                             use_multi_core);
+    ConfigurationShared::SetColoredTristate(ui->use_extended_memory_layout,
+                                            Settings::values.use_extended_memory_layout,
+                                            use_extended_memory_layout);
 
     connect(ui->toggle_speed_limit, &QCheckBox::clicked, ui->speed_limit, [this]() {
         ui->speed_limit->setEnabled(ui->toggle_speed_limit->isChecked() &&
diff --git a/src/yuzu/configuration/configure_general.h b/src/yuzu/configuration/configure_general.h
index 85c1dd4a8..b6f3bb5ed 100644
--- a/src/yuzu/configuration/configure_general.h
+++ b/src/yuzu/configuration/configure_general.h
@@ -48,6 +48,7 @@ private:
 
     ConfigurationShared::CheckState use_speed_limit;
     ConfigurationShared::CheckState use_multi_core;
+    ConfigurationShared::CheckState use_extended_memory_layout;
 
     const Core::System& system;
 };
diff --git a/src/yuzu/configuration/configure_general.ui b/src/yuzu/configuration/configure_general.ui
index 112dc72b3..c6ef2ab70 100644
--- a/src/yuzu/configuration/configure_general.ui
+++ b/src/yuzu/configuration/configure_general.ui
@@ -143,6 +143,13 @@
            </widget>
           </item>
           <item>
+           <widget class="QCheckBox" name="use_extended_memory_layout">
+            <property name="text">
+             <string>Extended memory layout (6GB DRAM)</string>
+            </property>
+           </widget>
+          </item>
+          <item>
            <widget class="QCheckBox" name="toggle_check_exit">
             <property name="text">
              <string>Confirm exit while emulation is running</string>
@@ -164,6 +171,13 @@
            </widget>
           </item>
           <item>
+           <widget class="QCheckBox" name="toggle_background_mute">
+            <property name="text">
+             <string>Mute audio when in background</string>
+            </property>
+           </widget>
+          </item>
+          <item>
            <widget class="QCheckBox" name="toggle_hide_mouse">
             <property name="text">
              <string>Hide mouse on inactivity</string>
diff --git a/src/yuzu/configuration/configure_input.cpp b/src/yuzu/configuration/configure_input.cpp
index d53179dbb..7c5776189 100644
--- a/src/yuzu/configuration/configure_input.cpp
+++ b/src/yuzu/configuration/configure_input.cpp
@@ -164,7 +164,7 @@ void ConfigureInput::Initialize(InputCommon::InputSubsystem* input_subsystem,
             });
 
     connect(ui->vibrationButton, &QPushButton::clicked,
-            [this] { CallConfigureDialog<ConfigureVibration>(*this); });
+            [this, &hid_core] { CallConfigureDialog<ConfigureVibration>(*this, hid_core); });
 
     connect(ui->motionButton, &QPushButton::clicked, [this, input_subsystem] {
         CallConfigureDialog<ConfigureMotionTouch>(*this, input_subsystem);
diff --git a/src/yuzu/configuration/configure_input_player.cpp b/src/yuzu/configuration/configure_input_player.cpp
index cc0534907..0aa4ac3e4 100644
--- a/src/yuzu/configuration/configure_input_player.cpp
+++ b/src/yuzu/configuration/configure_input_player.cpp
@@ -488,6 +488,32 @@ ConfigureInputPlayer::ConfigureInputPlayer(QWidget* parent, std::size_t player_i
                             emulated_controller->SetStickParam(analog_id, {});
                             analog_map_buttons[analog_id][sub_button_id]->setText(tr("[not set]"));
                         });
+                        context_menu.addAction(tr("Center axis"), [&] {
+                            const auto stick_value =
+                                emulated_controller->GetSticksValues()[analog_id];
+                            const float offset_x = stick_value.x.properties.offset;
+                            const float offset_y = stick_value.y.properties.offset;
+                            float raw_value_x = stick_value.x.raw_value;
+                            float raw_value_y = stick_value.y.raw_value;
+                            // See Core::HID::SanitizeStick() to obtain the original raw axis value
+                            if (std::abs(offset_x) < 0.5f) {
+                                if (raw_value_x > 0) {
+                                    raw_value_x *= 1 + offset_x;
+                                } else {
+                                    raw_value_x *= 1 - offset_x;
+                                }
+                            }
+                            if (std::abs(offset_x) < 0.5f) {
+                                if (raw_value_y > 0) {
+                                    raw_value_y *= 1 + offset_y;
+                                } else {
+                                    raw_value_y *= 1 - offset_y;
+                                }
+                            }
+                            param.Set("offset_x", -raw_value_x + offset_x);
+                            param.Set("offset_y", -raw_value_y + offset_y);
+                            emulated_controller->SetStickParam(analog_id, param);
+                        });
                         context_menu.addAction(tr("Invert axis"), [&] {
                             if (sub_button_id == 2 || sub_button_id == 3) {
                                 const bool invert_value = param.Get("invert_x", "+") == "-";
@@ -1306,6 +1332,9 @@ void ConfigureInputPlayer::HandleClick(
     QPushButton* button, std::size_t button_id,
     std::function<void(const Common::ParamPackage&)> new_input_setter,
     InputCommon::Polling::InputType type) {
+    if (timeout_timer->isActive()) {
+        return;
+    }
     if (button == ui->buttonMotionLeft || button == ui->buttonMotionRight) {
         button->setText(tr("Shake!"));
     } else {
diff --git a/src/yuzu/configuration/configure_input_player_widget.cpp b/src/yuzu/configuration/configure_input_player_widget.cpp
index 6630321cb..fb168b2ca 100644
--- a/src/yuzu/configuration/configure_input_player_widget.cpp
+++ b/src/yuzu/configuration/configure_input_player_widget.cpp
@@ -70,7 +70,6 @@ void PlayerControlPreview::UpdateColors() {
         colors.slider_arrow = QColor(14, 15, 18);
         colors.font2 = QColor(255, 255, 255);
         colors.indicator = QColor(170, 238, 255);
-        colors.indicator2 = QColor(100, 255, 100);
         colors.deadzone = QColor(204, 136, 136);
         colors.slider_button = colors.button;
     }
@@ -88,7 +87,6 @@ void PlayerControlPreview::UpdateColors() {
         colors.slider_arrow = QColor(65, 68, 73);
         colors.font2 = QColor(0, 0, 0);
         colors.indicator = QColor(0, 0, 200);
-        colors.indicator2 = QColor(0, 150, 0);
         colors.deadzone = QColor(170, 0, 0);
         colors.slider_button = QColor(153, 149, 149);
     }
@@ -101,6 +99,8 @@ void PlayerControlPreview::UpdateColors() {
     colors.font = QColor(255, 255, 255);
     colors.led_on = QColor(255, 255, 0);
     colors.led_off = QColor(170, 238, 255);
+    colors.indicator2 = QColor(59, 165, 93);
+    colors.charging = QColor(250, 168, 26);
 
     colors.left = colors.primary;
     colors.right = colors.primary;
@@ -357,7 +357,7 @@ void PlayerControlPreview::DrawLeftController(QPainter& p, const QPointF center)
     DrawCircle(p, center + QPoint(26, 71), 5);
 
     // Draw battery
-    DrawBattery(p, center + QPoint(-170, -140),
+    DrawBattery(p, center + QPoint(-160, -140),
                 battery_values[Core::HID::EmulatedDeviceIndex::LeftIndex]);
 }
 
@@ -484,7 +484,7 @@ void PlayerControlPreview::DrawRightController(QPainter& p, const QPointF center
     DrawSymbol(p, center + QPoint(-26, 66), Symbol::House, 5);
 
     // Draw battery
-    DrawBattery(p, center + QPoint(110, -140),
+    DrawBattery(p, center + QPoint(120, -140),
                 battery_values[Core::HID::EmulatedDeviceIndex::RightIndex]);
 }
 
@@ -621,9 +621,9 @@ void PlayerControlPreview::DrawDualController(QPainter& p, const QPointF center)
     DrawSymbol(p, center + QPoint(50, 60), Symbol::House, 4.2f);
 
     // Draw battery
-    DrawBattery(p, center + QPoint(-100, -160),
+    DrawBattery(p, center + QPoint(-200, -10),
                 battery_values[Core::HID::EmulatedDeviceIndex::LeftIndex]);
-    DrawBattery(p, center + QPoint(40, -160),
+    DrawBattery(p, center + QPoint(160, -10),
                 battery_values[Core::HID::EmulatedDeviceIndex::RightIndex]);
 }
 
@@ -694,12 +694,12 @@ void PlayerControlPreview::DrawHandheldController(QPainter& p, const QPointF cen
 
     // ZL and ZR buttons
     p.setPen(colors.outline);
-    DrawTriggerButton(p, center + QPoint(-210, -130), Direction::Left, button_values[ZL]);
-    DrawTriggerButton(p, center + QPoint(210, -130), Direction::Right, button_values[ZR]);
+    DrawTriggerButton(p, center + QPoint(-210, -120), Direction::Left, button_values[ZL]);
+    DrawTriggerButton(p, center + QPoint(210, -120), Direction::Right, button_values[ZR]);
     p.setPen(colors.transparent);
     p.setBrush(colors.font);
-    DrawSymbol(p, center + QPoint(-210, -130), Symbol::ZL, 1.5f);
-    DrawSymbol(p, center + QPoint(210, -130), Symbol::ZR, 1.5f);
+    DrawSymbol(p, center + QPoint(-210, -120), Symbol::ZL, 1.5f);
+    DrawSymbol(p, center + QPoint(210, -120), Symbol::ZR, 1.5f);
 
     // Minus and Plus button
     p.setPen(colors.outline);
@@ -725,9 +725,9 @@ void PlayerControlPreview::DrawHandheldController(QPainter& p, const QPointF cen
     DrawSymbol(p, center + QPoint(161, 37), Symbol::House, 2.75f);
 
     // Draw battery
-    DrawBattery(p, center + QPoint(-200, 110),
+    DrawBattery(p, center + QPoint(-188, 95),
                 battery_values[Core::HID::EmulatedDeviceIndex::LeftIndex]);
-    DrawBattery(p, center + QPoint(130, 110),
+    DrawBattery(p, center + QPoint(150, 95),
                 battery_values[Core::HID::EmulatedDeviceIndex::RightIndex]);
 }
 
@@ -781,12 +781,12 @@ void PlayerControlPreview::DrawProController(QPainter& p, const QPointF center)
 
     // ZL and ZR buttons
     p.setPen(colors.outline);
-    DrawTriggerButton(p, center + QPoint(-210, -130), Direction::Left, button_values[ZL]);
-    DrawTriggerButton(p, center + QPoint(210, -130), Direction::Right, button_values[ZR]);
+    DrawTriggerButton(p, center + QPoint(-210, -120), Direction::Left, button_values[ZL]);
+    DrawTriggerButton(p, center + QPoint(210, -120), Direction::Right, button_values[ZR]);
     p.setPen(colors.transparent);
     p.setBrush(colors.font);
-    DrawSymbol(p, center + QPoint(-210, -130), Symbol::ZL, 1.5f);
-    DrawSymbol(p, center + QPoint(210, -130), Symbol::ZR, 1.5f);
+    DrawSymbol(p, center + QPoint(-210, -120), Symbol::ZL, 1.5f);
+    DrawSymbol(p, center + QPoint(210, -120), Symbol::ZR, 1.5f);
 
     // Minus and Plus buttons
     p.setPen(colors.outline);
@@ -818,7 +818,7 @@ void PlayerControlPreview::DrawProController(QPainter& p, const QPointF center)
     DrawSymbol(p, center + QPoint(29, -56), Symbol::House, 3.9f);
 
     // Draw battery
-    DrawBattery(p, center + QPoint(-30, -160),
+    DrawBattery(p, center + QPoint(-20, -160),
                 battery_values[Core::HID::EmulatedDeviceIndex::LeftIndex]);
 }
 
@@ -875,7 +875,7 @@ void PlayerControlPreview::DrawGCController(QPainter& p, const QPointF center) {
     DrawCircleButton(p, center + QPoint(0, -44), button_values[Plus], 8);
 
     // Draw battery
-    DrawBattery(p, center + QPoint(-30, -165),
+    DrawBattery(p, center + QPoint(-20, 110),
                 battery_values[Core::HID::EmulatedDeviceIndex::LeftIndex]);
 }
 
@@ -1030,6 +1030,10 @@ constexpr std::array<float, 30 * 2> symbol_c = {
     -2.37f, 5.64f,  -0.65f, 6.44f,  1.25f,  6.47f,  3.06f,  5.89f,  4.63f,  4.92f,  4.63f,  6.83f,
 };
 
+constexpr std::array<float, 6 * 2> symbol_charging = {
+    6.5f, -1.0f, 1.0f, -1.0f, 1.0f, -3.0f, -6.5f, 1.0f, -1.0f, 1.0f, -1.0f, 3.0f,
+};
+
 constexpr std::array<float, 12 * 2> house = {
     -1.3f, 0.0f,  -0.93f, 0.0f, -0.93f, 1.15f, 0.93f,  1.15f, 0.93f, 0.0f, 1.3f,  0.0f,
     0.0f,  -1.2f, -1.3f,  0.0f, -0.43f, 0.0f,  -0.43f, .73f,  0.43f, .73f, 0.43f, 0.0f,
@@ -2674,36 +2678,43 @@ void PlayerControlPreview::DrawBattery(QPainter& p, QPointF center,
     if (battery == Common::Input::BatteryLevel::None) {
         return;
     }
-    p.setPen(colors.outline);
+    // Draw outline
+    p.setPen(QPen(colors.button, 5));
+    p.setBrush(colors.transparent);
+    p.drawRoundedRect(center.x(), center.y(), 34, 16, 2, 2);
+
+    p.setPen(QPen(colors.button, 3));
+    p.drawRect(center.x() + 35, center.y() + 4.5f, 4, 7);
+
+    // Draw Battery shape
+    p.setPen(QPen(colors.indicator2, 3));
     p.setBrush(colors.transparent);
-    p.drawRect(center.x(), center.y(), 56, 20);
-    p.drawRect(center.x() + 56, center.y() + 6, 3, 8);
-    p.setBrush(colors.deadzone);
+    p.drawRoundedRect(center.x(), center.y(), 34, 16, 2, 2);
+
+    p.setPen(QPen(colors.indicator2, 1));
+    p.setBrush(colors.indicator2);
+    p.drawRect(center.x() + 35, center.y() + 4.5f, 4, 7);
     switch (battery) {
     case Common::Input::BatteryLevel::Charging:
-        p.setBrush(colors.indicator2);
-        p.drawText(center + QPoint(2, 14), tr("Charging"));
+        p.drawRect(center.x(), center.y(), 34, 16);
+        p.setPen(colors.slider);
+        p.setBrush(colors.charging);
+        DrawSymbol(p, center + QPointF(17.0f, 8.0f), Symbol::Charging, 2.1f);
         break;
     case Common::Input::BatteryLevel::Full:
-        p.drawRect(center.x() + 42, center.y(), 14, 20);
-        p.drawRect(center.x() + 28, center.y(), 14, 20);
-        p.drawRect(center.x() + 14, center.y(), 14, 20);
-        p.drawRect(center.x(), center.y(), 14, 20);
+        p.drawRect(center.x(), center.y(), 34, 16);
         break;
     case Common::Input::BatteryLevel::Medium:
-        p.drawRect(center.x() + 28, center.y(), 14, 20);
-        p.drawRect(center.x() + 14, center.y(), 14, 20);
-        p.drawRect(center.x(), center.y(), 14, 20);
+        p.drawRect(center.x(), center.y(), 25, 16);
         break;
     case Common::Input::BatteryLevel::Low:
-        p.drawRect(center.x() + 14, center.y(), 14, 20);
-        p.drawRect(center.x(), center.y(), 14, 20);
+        p.drawRect(center.x(), center.y(), 17, 16);
         break;
     case Common::Input::BatteryLevel::Critical:
-        p.drawRect(center.x(), center.y(), 14, 20);
+        p.drawRect(center.x(), center.y(), 6, 16);
         break;
     case Common::Input::BatteryLevel::Empty:
-        p.drawRect(center.x(), center.y(), 5, 20);
+        p.drawRect(center.x(), center.y(), 3, 16);
         break;
     default:
         break;
@@ -2724,6 +2735,7 @@ void PlayerControlPreview::DrawSymbol(QPainter& p, const QPointF center, Symbol
     std::array<QPointF, symbol_sl.size() / 2> sl_icon;
     std::array<QPointF, symbol_zr.size() / 2> zr_icon;
     std::array<QPointF, symbol_sr.size() / 2> sr_icon;
+    std::array<QPointF, symbol_charging.size() / 2> charging_icon;
     switch (symbol) {
     case Symbol::House:
         for (std::size_t point = 0; point < house.size() / 2; ++point) {
@@ -2809,6 +2821,13 @@ void PlayerControlPreview::DrawSymbol(QPainter& p, const QPointF center, Symbol
         }
         p.drawPolygon(sr_icon.data(), static_cast<int>(sr_icon.size()));
         break;
+    case Symbol::Charging:
+        for (std::size_t point = 0; point < symbol_charging.size() / 2; ++point) {
+            charging_icon[point] = center + QPointF(symbol_charging[point * 2] * icon_size,
+                                                    symbol_charging[point * 2 + 1] * icon_size);
+        }
+        p.drawPolygon(charging_icon.data(), static_cast<int>(charging_icon.size()));
+        break;
     }
 }
 
diff --git a/src/yuzu/configuration/configure_input_player_widget.h b/src/yuzu/configuration/configure_input_player_widget.h
index 4cd5c3be0..3582ef77a 100644
--- a/src/yuzu/configuration/configure_input_player_widget.h
+++ b/src/yuzu/configuration/configure_input_player_widget.h
@@ -72,6 +72,7 @@ private:
         ZL,
         ZR,
         SR,
+        Charging,
     };
 
     struct ColorMapping {
@@ -94,6 +95,7 @@ private:
         QColor slider_button{};
         QColor slider_arrow{};
         QColor deadzone{};
+        QColor charging{};
     };
 
     void UpdateColors();
diff --git a/src/yuzu/configuration/configure_profile_manager.cpp b/src/yuzu/configuration/configure_profile_manager.cpp
index 78b6374c0..d9f6dee4e 100644
--- a/src/yuzu/configuration/configure_profile_manager.cpp
+++ b/src/yuzu/configuration/configure_profile_manager.cpp
@@ -33,10 +33,10 @@ constexpr std::array<u8, 107> backup_jpeg{
     0x01, 0x01, 0x00, 0x00, 0x3f, 0x00, 0xd2, 0xcf, 0x20, 0xff, 0xd9,
 };
 
-QString GetImagePath(Common::UUID uuid) {
+QString GetImagePath(const Common::UUID& uuid) {
     const auto path =
         Common::FS::GetYuzuPath(Common::FS::YuzuPath::NANDDir) /
-        fmt::format("system/save/8000000000000010/su/avators/{}.jpg", uuid.FormatSwitch());
+        fmt::format("system/save/8000000000000010/su/avators/{}.jpg", uuid.FormattedString());
     return QString::fromStdString(Common::FS::PathToUTF8String(path));
 }
 
@@ -55,10 +55,10 @@ QString FormatUserEntryText(const QString& username, Common::UUID uuid) {
     return ConfigureProfileManager::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()));
+        .arg(username, QString::fromStdString(uuid.FormattedString()));
 }
 
-QPixmap GetIcon(Common::UUID uuid) {
+QPixmap GetIcon(const Common::UUID& uuid) {
     QPixmap icon{GetImagePath(uuid)};
 
     if (!icon) {
@@ -200,7 +200,7 @@ void ConfigureProfileManager::AddUser() {
         return;
     }
 
-    const auto uuid = Common::UUID::Generate();
+    const auto uuid = Common::UUID::MakeRandom();
     profile_manager->CreateNewUser(uuid, username.toStdString());
 
     item_model->appendRow(new QStandardItem{GetIcon(uuid), FormatUserEntryText(username, uuid)});
diff --git a/src/yuzu/configuration/configure_touch_from_button.cpp b/src/yuzu/configuration/configure_touch_from_button.cpp
index bde0a08c4..211a00217 100644
--- a/src/yuzu/configuration/configure_touch_from_button.cpp
+++ b/src/yuzu/configuration/configure_touch_from_button.cpp
@@ -227,6 +227,9 @@ void ConfigureTouchFromButton::RenameMapping() {
 }
 
 void ConfigureTouchFromButton::GetButtonInput(const int row_index, const bool is_new) {
+    if (timeout_timer->isActive()) {
+        return;
+    }
     binding_list_model->item(row_index, 0)->setText(tr("[press key]"));
 
     input_setter = [this, row_index, is_new](const Common::ParamPackage& params,
diff --git a/src/yuzu/configuration/configure_vibration.cpp b/src/yuzu/configuration/configure_vibration.cpp
index adce04b27..779b6401c 100644
--- a/src/yuzu/configuration/configure_vibration.cpp
+++ b/src/yuzu/configuration/configure_vibration.cpp
@@ -9,11 +9,14 @@
 
 #include "common/param_package.h"
 #include "common/settings.h"
+#include "core/hid/emulated_controller.h"
+#include "core/hid/hid_core.h"
+#include "core/hid/hid_types.h"
 #include "ui_configure_vibration.h"
 #include "yuzu/configuration/configure_vibration.h"
 
-ConfigureVibration::ConfigureVibration(QWidget* parent)
-    : QDialog(parent), ui(std::make_unique<Ui::ConfigureVibration>()) {
+ConfigureVibration::ConfigureVibration(QWidget* parent, Core::HID::HIDCore& hid_core_)
+    : QDialog(parent), ui(std::make_unique<Ui::ConfigureVibration>()), hid_core{hid_core_} {
     ui->setupUi(this);
 
     vibration_groupboxes = {
@@ -31,6 +34,13 @@ ConfigureVibration::ConfigureVibration(QWidget* parent)
     const auto& players = Settings::values.players.GetValue();
 
     for (std::size_t i = 0; i < NUM_PLAYERS; ++i) {
+        auto controller = hid_core.GetEmulatedControllerByIndex(i);
+        Core::HID::ControllerUpdateCallback engine_callback{
+            .on_change = [this,
+                          i](Core::HID::ControllerTriggerType type) { VibrateController(type, i); },
+            .is_npad_service = false,
+        };
+        controller_callback_key[i] = controller->SetCallback(engine_callback);
         vibration_groupboxes[i]->setChecked(players[i].vibration_enabled);
         vibration_spinboxes[i]->setValue(players[i].vibration_strength);
     }
@@ -45,7 +55,14 @@ ConfigureVibration::ConfigureVibration(QWidget* parent)
     RetranslateUI();
 }
 
-ConfigureVibration::~ConfigureVibration() = default;
+ConfigureVibration::~ConfigureVibration() {
+    StopVibrations();
+
+    for (std::size_t i = 0; i < NUM_PLAYERS; ++i) {
+        auto controller = hid_core.GetEmulatedControllerByIndex(i);
+        controller->DeleteCallback(controller_callback_key[i]);
+    }
+};
 
 void ConfigureVibration::ApplyConfiguration() {
     auto& players = Settings::values.players.GetValue();
@@ -70,3 +87,54 @@ void ConfigureVibration::changeEvent(QEvent* event) {
 void ConfigureVibration::RetranslateUI() {
     ui->retranslateUi(this);
 }
+
+void ConfigureVibration::VibrateController(Core::HID::ControllerTriggerType type,
+                                           std::size_t player_index) {
+    if (type != Core::HID::ControllerTriggerType::Button) {
+        return;
+    }
+
+    auto& player = Settings::values.players.GetValue()[player_index];
+    auto controller = hid_core.GetEmulatedControllerByIndex(player_index);
+    const int vibration_strenght = vibration_spinboxes[player_index]->value();
+    const auto& buttons = controller->GetButtonsValues();
+
+    bool button_is_pressed = false;
+    for (std::size_t i = 0; i < buttons.size(); ++i) {
+        if (buttons[i].value) {
+            button_is_pressed = true;
+            break;
+        }
+    }
+
+    if (!button_is_pressed) {
+        StopVibrations();
+        return;
+    }
+
+    const int old_vibration_enabled = player.vibration_enabled;
+    const bool old_vibration_strenght = player.vibration_strength;
+    player.vibration_enabled = true;
+    player.vibration_strength = vibration_strenght;
+
+    const Core::HID::VibrationValue vibration{
+        .low_amplitude = 1.0f,
+        .low_frequency = 160.0f,
+        .high_amplitude = 1.0f,
+        .high_frequency = 320.0f,
+    };
+    controller->SetVibration(0, vibration);
+    controller->SetVibration(1, vibration);
+
+    // Restore previous values
+    player.vibration_enabled = old_vibration_enabled;
+    player.vibration_strength = old_vibration_strenght;
+}
+
+void ConfigureVibration::StopVibrations() {
+    for (std::size_t i = 0; i < NUM_PLAYERS; ++i) {
+        auto controller = hid_core.GetEmulatedControllerByIndex(i);
+        controller->SetVibration(0, Core::HID::DEFAULT_VIBRATION_VALUE);
+        controller->SetVibration(1, Core::HID::DEFAULT_VIBRATION_VALUE);
+    }
+}
diff --git a/src/yuzu/configuration/configure_vibration.h b/src/yuzu/configuration/configure_vibration.h
index 37bbc2653..50b8195fa 100644
--- a/src/yuzu/configuration/configure_vibration.h
+++ b/src/yuzu/configuration/configure_vibration.h
@@ -15,11 +15,16 @@ namespace Ui {
 class ConfigureVibration;
 }
 
+namespace Core::HID {
+enum class ControllerTriggerType;
+class HIDCore;
+} // namespace Core::HID
+
 class ConfigureVibration : public QDialog {
     Q_OBJECT
 
 public:
-    explicit ConfigureVibration(QWidget* parent);
+    explicit ConfigureVibration(QWidget* parent, Core::HID::HIDCore& hid_core_);
     ~ConfigureVibration() override;
 
     void ApplyConfiguration();
@@ -27,14 +32,21 @@ public:
 private:
     void changeEvent(QEvent* event) override;
     void RetranslateUI();
+    void VibrateController(Core::HID::ControllerTriggerType type, std::size_t player_index);
+    void StopVibrations();
 
     std::unique_ptr<Ui::ConfigureVibration> ui;
 
     static constexpr std::size_t NUM_PLAYERS = 8;
 
-    // Groupboxes encapsulating the vibration strength spinbox.
+    /// Groupboxes encapsulating the vibration strength spinbox.
     std::array<QGroupBox*, NUM_PLAYERS> vibration_groupboxes;
 
-    // Spinboxes representing the vibration strength percentage.
+    /// Spinboxes representing the vibration strength percentage.
     std::array<QSpinBox*, NUM_PLAYERS> vibration_spinboxes;
+
+    /// Callback index to stop the controllers events
+    std::array<int, NUM_PLAYERS> controller_callback_key;
+
+    Core::HID::HIDCore& hid_core;
 };
diff --git a/src/yuzu/configuration/configure_vibration.ui b/src/yuzu/configuration/configure_vibration.ui
index efdf317a9..447a18eb1 100644
--- a/src/yuzu/configuration/configure_vibration.ui
+++ b/src/yuzu/configuration/configure_vibration.ui
@@ -17,6 +17,13 @@
    <string notr="true"/>
   </property>
   <layout class="QVBoxLayout">
+   <item row="0" column="0" colspan="4">
+    <widget class="QLabel" name="label_1">
+     <property name="text">
+      <string>Press any controller button to vibrate the controller.</string>
+     </property>
+    </widget>
+   </item>
    <item>
     <widget class="QGroupBox" name="vibrationStrengthGroup">
      <property name="title">
diff --git a/src/yuzu/debugger/console.cpp b/src/yuzu/debugger/console.cpp
index f89ea8ea7..4b508b466 100644
--- a/src/yuzu/debugger/console.cpp
+++ b/src/yuzu/debugger/console.cpp
@@ -30,6 +30,7 @@ void ToggleConsole() {
             freopen_s(&temp, "CONIN$", "r", stdin);
             freopen_s(&temp, "CONOUT$", "w", stdout);
             freopen_s(&temp, "CONOUT$", "w", stderr);
+            SetConsoleOutputCP(65001);
             SetColorConsoleBackendEnabled(true);
         }
     } else {
diff --git a/src/yuzu/main.cpp b/src/yuzu/main.cpp
index d9e689d14..06774768d 100644
--- a/src/yuzu/main.cpp
+++ b/src/yuzu/main.cpp
@@ -25,6 +25,7 @@
 #include "core/file_sys/vfs_real.h"
 #include "core/frontend/applets/controller.h"
 #include "core/frontend/applets/general_frontend.h"
+#include "core/frontend/applets/mii.h"
 #include "core/frontend/applets/software_keyboard.h"
 #include "core/hid/emulated_controller.h"
 #include "core/hid/hid_core.h"
@@ -249,9 +250,9 @@ GMainWindow::GMainWindow()
 #ifdef ARCHITECTURE_x86_64
     const auto& caps = Common::GetCPUCaps();
     std::string cpu_string = caps.cpu_string;
-    if (caps.avx || caps.avx2 || caps.avx512) {
+    if (caps.avx || caps.avx2 || caps.avx512f) {
         cpu_string += " | AVX";
-        if (caps.avx512) {
+        if (caps.avx512f) {
             cpu_string += "512";
         } else if (caps.avx2) {
             cpu_string += '2';
@@ -806,21 +807,8 @@ void GMainWindow::InitializeWidgets() {
     filter_status_button = new QPushButton();
     filter_status_button->setObjectName(QStringLiteral("TogglableStatusBarButton"));
     filter_status_button->setFocusPolicy(Qt::NoFocus);
-    connect(filter_status_button, &QPushButton::clicked, [&] {
-        auto filter = Settings::values.scaling_filter.GetValue();
-        if (filter == Settings::ScalingFilter::LastFilter) {
-            filter = Settings::ScalingFilter::NearestNeighbor;
-        } else {
-            filter = static_cast<Settings::ScalingFilter>(static_cast<u32>(filter) + 1);
-        }
-        if (Settings::values.renderer_backend.GetValue() == Settings::RendererBackend::OpenGL &&
-            filter == Settings::ScalingFilter::Fsr) {
-            filter = Settings::ScalingFilter::NearestNeighbor;
-        }
-        Settings::values.scaling_filter.SetValue(filter);
-        filter_status_button->setChecked(true);
-        UpdateFilterText();
-    });
+    connect(filter_status_button, &QPushButton::clicked, this,
+            &GMainWindow::OnToggleAdaptingFilter);
     auto filter = Settings::values.scaling_filter.GetValue();
     if (Settings::values.renderer_backend.GetValue() == Settings::RendererBackend::OpenGL &&
         filter == Settings::ScalingFilter::Fsr) {
@@ -835,25 +823,7 @@ void GMainWindow::InitializeWidgets() {
     dock_status_button = new QPushButton();
     dock_status_button->setObjectName(QStringLiteral("TogglableStatusBarButton"));
     dock_status_button->setFocusPolicy(Qt::NoFocus);
-    connect(dock_status_button, &QPushButton::clicked, [&] {
-        const bool is_docked = Settings::values.use_docked_mode.GetValue();
-        auto* player_1 = system->HIDCore().GetEmulatedController(Core::HID::NpadIdType::Player1);
-        auto* handheld = system->HIDCore().GetEmulatedController(Core::HID::NpadIdType::Handheld);
-
-        if (!is_docked && handheld->IsConnected()) {
-            QMessageBox::warning(this, tr("Invalid config detected"),
-                                 tr("Handheld controller can't be used on docked mode. Pro "
-                                    "controller will be selected."));
-            handheld->Disconnect();
-            player_1->SetNpadStyleIndex(Core::HID::NpadStyleIndex::ProController);
-            player_1->Connect();
-            controller_dialog->refreshConfiguration();
-        }
-
-        Settings::values.use_docked_mode.SetValue(!is_docked);
-        dock_status_button->setChecked(!is_docked);
-        OnDockedModeChanged(is_docked, !is_docked, *system);
-    });
+    connect(dock_status_button, &QPushButton::clicked, this, &GMainWindow::OnToggleDockedMode);
     dock_status_button->setText(tr("DOCK"));
     dock_status_button->setCheckable(true);
     dock_status_button->setChecked(Settings::values.use_docked_mode.GetValue());
@@ -863,22 +833,7 @@ void GMainWindow::InitializeWidgets() {
     gpu_accuracy_button->setObjectName(QStringLiteral("GPUStatusBarButton"));
     gpu_accuracy_button->setCheckable(true);
     gpu_accuracy_button->setFocusPolicy(Qt::NoFocus);
-    connect(gpu_accuracy_button, &QPushButton::clicked, [this] {
-        switch (Settings::values.gpu_accuracy.GetValue()) {
-        case Settings::GPUAccuracy::High: {
-            Settings::values.gpu_accuracy.SetValue(Settings::GPUAccuracy::Normal);
-            break;
-        }
-        case Settings::GPUAccuracy::Normal:
-        case Settings::GPUAccuracy::Extreme:
-        default: {
-            Settings::values.gpu_accuracy.SetValue(Settings::GPUAccuracy::High);
-        }
-        }
-
-        system->ApplySettings();
-        UpdateGPUAccuracyButton();
-    });
+    connect(gpu_accuracy_button, &QPushButton::clicked, this, &GMainWindow::OnToggleGpuAccuracy);
     UpdateGPUAccuracyButton();
     statusBar()->insertPermanentWidget(0, gpu_accuracy_button);
 
@@ -965,6 +920,7 @@ void GMainWindow::LinkActionShortcut(QAction* action, const QString& action_name
     static const QString main_window = QStringLiteral("Main Window");
     action->setShortcut(hotkey_registry.GetKeySequence(main_window, action_name));
     action->setShortcutContext(hotkey_registry.GetShortcutContext(main_window, action_name));
+    action->setAutoRepeat(false);
 
     this->addAction(action);
 
@@ -979,7 +935,7 @@ void GMainWindow::InitializeHotkeys() {
     hotkey_registry.LoadHotkeys();
 
     LinkActionShortcut(ui->action_Load_File, QStringLiteral("Load File"));
-    LinkActionShortcut(ui->action_Load_Amiibo, QStringLiteral("Load Amiibo"));
+    LinkActionShortcut(ui->action_Load_Amiibo, QStringLiteral("Load/Remove Amiibo"));
     LinkActionShortcut(ui->action_Exit, QStringLiteral("Exit yuzu"));
     LinkActionShortcut(ui->action_Restart, QStringLiteral("Restart Emulation"));
     LinkActionShortcut(ui->action_Pause, QStringLiteral("Continue/Pause Emulation"));
@@ -1008,12 +964,10 @@ void GMainWindow::InitializeHotkeys() {
             ToggleFullscreen();
         }
     });
-    connect_shortcut(QStringLiteral("Change Docked Mode"), [&] {
-        Settings::values.use_docked_mode.SetValue(!Settings::values.use_docked_mode.GetValue());
-        OnDockedModeChanged(!Settings::values.use_docked_mode.GetValue(),
-                            Settings::values.use_docked_mode.GetValue(), *system);
-        dock_status_button->setChecked(Settings::values.use_docked_mode.GetValue());
-    });
+    connect_shortcut(QStringLiteral("Change Adapting Filter"),
+                     &GMainWindow::OnToggleAdaptingFilter);
+    connect_shortcut(QStringLiteral("Change Docked Mode"), &GMainWindow::OnToggleDockedMode);
+    connect_shortcut(QStringLiteral("Change GPU Accuracy"), &GMainWindow::OnToggleGpuAccuracy);
     connect_shortcut(QStringLiteral("Audio Mute/Unmute"),
                      [] { Settings::values.audio_muted = !Settings::values.audio_muted; });
     connect_shortcut(QStringLiteral("Audio Volume Down"), [] {
@@ -1051,8 +1005,10 @@ void GMainWindow::SetDefaultUIGeometry() {
 }
 
 void GMainWindow::RestoreUIState() {
+    setWindowFlags(windowFlags() & ~Qt::FramelessWindowHint);
     restoreGeometry(UISettings::values.geometry);
     restoreState(UISettings::values.state);
+    render_window->setWindowFlags(render_window->windowFlags() & ~Qt::FramelessWindowHint);
     render_window->restoreGeometry(UISettings::values.renderwindow_geometry);
 #if MICROPROFILE_ENABLED
     microProfileDialog->restoreGeometry(UISettings::values.microprofile_geometry);
@@ -1079,14 +1035,14 @@ void GMainWindow::RestoreUIState() {
 }
 
 void GMainWindow::OnAppFocusStateChanged(Qt::ApplicationState state) {
-    if (!UISettings::values.pause_when_in_background) {
-        return;
-    }
     if (state != Qt::ApplicationHidden && state != Qt::ApplicationInactive &&
         state != Qt::ApplicationActive) {
         LOG_DEBUG(Frontend, "ApplicationState unusual flag: {} ", state);
     }
-    if (emulation_running) {
+    if (!emulation_running) {
+        return;
+    }
+    if (UISettings::values.pause_when_in_background) {
         if (emu_thread->IsRunning() &&
             (state & (Qt::ApplicationHidden | Qt::ApplicationInactive))) {
             auto_paused = true;
@@ -1096,6 +1052,16 @@ void GMainWindow::OnAppFocusStateChanged(Qt::ApplicationState state) {
             OnStartGame();
         }
     }
+    if (UISettings::values.mute_when_in_background) {
+        if (!Settings::values.audio_muted &&
+            (state & (Qt::ApplicationHidden | Qt::ApplicationInactive))) {
+            Settings::values.audio_muted = true;
+            auto_muted = true;
+        } else if (auto_muted && state == Qt::ApplicationActive) {
+            Settings::values.audio_muted = false;
+            auto_muted = false;
+        }
+    }
 }
 
 void GMainWindow::ConnectWidgetEvents() {
@@ -1320,6 +1286,7 @@ bool GMainWindow::LoadROM(const QString& filename, u64 program_id, std::size_t p
         std::make_unique<QtControllerSelector>(*this), // Controller Selector
         std::make_unique<QtErrorDisplay>(*this),       // Error Display
         nullptr,                                       // Parental Controls
+        nullptr,                                       // Mii editor
         nullptr,                                       // Photo Viewer
         std::make_unique<QtProfileSelector>(*this),    // Profile Selector
         std::make_unique<QtSoftwareKeyboard>(*this),   // Software Keyboard
@@ -1687,7 +1654,7 @@ void GMainWindow::OnGameListOpenFolder(u64 program_id, GameListOpenTarget target
 
             const auto user_save_data_path = FileSys::SaveDataFactory::GetFullPath(
                 *system, FileSys::SaveDataSpaceId::NandUser, FileSys::SaveDataType::SaveData,
-                program_id, user_id->uuid, 0);
+                program_id, user_id->AsU128(), 0);
 
             path = Common::FS::ConcatPathSafe(nand_dir, user_save_data_path);
         } else {
@@ -2865,6 +2832,59 @@ void GMainWindow::OnTasReset() {
     input_subsystem->GetTas()->Reset();
 }
 
+void GMainWindow::OnToggleDockedMode() {
+    const bool is_docked = Settings::values.use_docked_mode.GetValue();
+    auto* player_1 = system->HIDCore().GetEmulatedController(Core::HID::NpadIdType::Player1);
+    auto* handheld = system->HIDCore().GetEmulatedController(Core::HID::NpadIdType::Handheld);
+
+    if (!is_docked && handheld->IsConnected()) {
+        QMessageBox::warning(this, tr("Invalid config detected"),
+                             tr("Handheld controller can't be used on docked mode. Pro "
+                                "controller will be selected."));
+        handheld->Disconnect();
+        player_1->SetNpadStyleIndex(Core::HID::NpadStyleIndex::ProController);
+        player_1->Connect();
+        controller_dialog->refreshConfiguration();
+    }
+
+    Settings::values.use_docked_mode.SetValue(!is_docked);
+    dock_status_button->setChecked(!is_docked);
+    OnDockedModeChanged(is_docked, !is_docked, *system);
+}
+
+void GMainWindow::OnToggleGpuAccuracy() {
+    switch (Settings::values.gpu_accuracy.GetValue()) {
+    case Settings::GPUAccuracy::High: {
+        Settings::values.gpu_accuracy.SetValue(Settings::GPUAccuracy::Normal);
+        break;
+    }
+    case Settings::GPUAccuracy::Normal:
+    case Settings::GPUAccuracy::Extreme:
+    default: {
+        Settings::values.gpu_accuracy.SetValue(Settings::GPUAccuracy::High);
+    }
+    }
+
+    system->ApplySettings();
+    UpdateGPUAccuracyButton();
+}
+
+void GMainWindow::OnToggleAdaptingFilter() {
+    auto filter = Settings::values.scaling_filter.GetValue();
+    if (filter == Settings::ScalingFilter::LastFilter) {
+        filter = Settings::ScalingFilter::NearestNeighbor;
+    } else {
+        filter = static_cast<Settings::ScalingFilter>(static_cast<u32>(filter) + 1);
+    }
+    if (Settings::values.renderer_backend.GetValue() == Settings::RendererBackend::OpenGL &&
+        filter == Settings::ScalingFilter::Fsr) {
+        filter = Settings::ScalingFilter::NearestNeighbor;
+    }
+    Settings::values.scaling_filter.SetValue(filter);
+    filter_status_button->setChecked(true);
+    UpdateFilterText();
+}
+
 void GMainWindow::OnConfigurePerGame() {
     const u64 title_id = system->GetCurrentProcessProgramID();
     OpenPerGameConfiguration(title_id, game_path.toStdString());
@@ -2909,6 +2929,25 @@ void GMainWindow::OnLoadAmiibo() {
         return;
     }
 
+    Service::SM::ServiceManager& sm = system->ServiceManager();
+    auto nfc = sm.GetService<Service::NFP::Module::Interface>("nfp:user");
+    if (nfc == nullptr) {
+        QMessageBox::warning(this, tr("Error"), tr("The current game is not looking for amiibos"));
+        return;
+    }
+    const auto nfc_state = nfc->GetCurrentState();
+    if (nfc_state == Service::NFP::DeviceState::TagFound ||
+        nfc_state == Service::NFP::DeviceState::TagMounted) {
+        nfc->CloseAmiibo();
+        QMessageBox::warning(this, tr("Amiibo"), tr("The current amiibo has been removed"));
+        return;
+    }
+
+    if (nfc_state != Service::NFP::DeviceState::SearchingForTag) {
+        QMessageBox::warning(this, tr("Error"), tr("The current game is not looking for amiibos"));
+        return;
+    }
+
     is_amiibo_file_select_active = true;
     const QString extensions{QStringLiteral("*.bin")};
     const QString file_filter = tr("Amiibo File (%1);; All Files (*.*)").arg(extensions);
@@ -2929,6 +2968,15 @@ void GMainWindow::LoadAmiibo(const QString& filename) {
         return;
     }
 
+    // Remove amiibo if one is connected
+    const auto nfc_state = nfc->GetCurrentState();
+    if (nfc_state == Service::NFP::DeviceState::TagFound ||
+        nfc_state == Service::NFP::DeviceState::TagMounted) {
+        nfc->CloseAmiibo();
+        QMessageBox::warning(this, tr("Amiibo"), tr("The current amiibo has been removed"));
+        return;
+    }
+
     QFile nfc_file{filename};
     if (!nfc_file.open(QIODevice::ReadOnly)) {
         QMessageBox::warning(this, tr("Error opening Amiibo data file"),
diff --git a/src/yuzu/main.h b/src/yuzu/main.h
index ca4ab9af5..6a35b9e3d 100644
--- a/src/yuzu/main.h
+++ b/src/yuzu/main.h
@@ -284,6 +284,9 @@ private slots:
     void OnTasStartStop();
     void OnTasRecord();
     void OnTasReset();
+    void OnToggleDockedMode();
+    void OnToggleGpuAccuracy();
+    void OnToggleAdaptingFilter();
     void OnConfigurePerGame();
     void OnLoadAmiibo();
     void OnOpenYuzuFolder();
@@ -369,6 +372,7 @@ private:
     QString game_path;
 
     bool auto_paused = false;
+    bool auto_muted = false;
     QTimer mouse_hide_timer;
 
     // FS
diff --git a/src/yuzu/main.ui b/src/yuzu/main.ui
index 5719b2ee4..6ab95b9a5 100644
--- a/src/yuzu/main.ui
+++ b/src/yuzu/main.ui
@@ -266,7 +266,7 @@
     <bool>false</bool>
    </property>
    <property name="text">
-    <string>Load &amp;Amiibo...</string>
+    <string>Load/Remove &amp;Amiibo...</string>
    </property>
   </action>
   <action name="action_Report_Compatibility">
diff --git a/src/yuzu/uisettings.h b/src/yuzu/uisettings.h
index f7298ddad..06e8b46da 100644
--- a/src/yuzu/uisettings.h
+++ b/src/yuzu/uisettings.h
@@ -73,6 +73,7 @@ struct Values {
     Settings::BasicSetting<bool> confirm_before_closing{true, "confirmClose"};
     Settings::BasicSetting<bool> first_start{true, "firstStart"};
     Settings::BasicSetting<bool> pause_when_in_background{false, "pauseWhenInBackground"};
+    Settings::BasicSetting<bool> mute_when_in_background{false, "muteWhenInBackground"};
     Settings::BasicSetting<bool> hide_mouse{true, "hideInactiveMouse"};
 
     Settings::BasicSetting<bool> select_user_on_boot{false, "select_user_on_boot"};
diff --git a/src/yuzu_cmd/config.cpp b/src/yuzu_cmd/config.cpp
index 8e9c7d211..131bc2201 100644
--- a/src/yuzu_cmd/config.cpp
+++ b/src/yuzu_cmd/config.cpp
@@ -3,6 +3,7 @@
 // Refer to the license.txt file included.
 
 #include <memory>
+#include <optional>
 #include <sstream>
 
 // Ignore -Wimplicit-fallthrough due to https://github.com/libsdl-org/SDL/issues/4307
@@ -29,11 +30,12 @@
 
 namespace FS = Common::FS;
 
-Config::Config() {
-    // TODO: Don't hardcode the path; let the frontend decide where to put the config files.
-    sdl2_config_loc = FS::GetYuzuPath(FS::YuzuPath::ConfigDir) / "sdl2-config.ini";
-    sdl2_config = std::make_unique<INIReader>(FS::PathToUTF8String(sdl2_config_loc));
+const std::filesystem::path default_config_path =
+    FS::GetYuzuPath(FS::YuzuPath::ConfigDir) / "sdl2-config.ini";
 
+Config::Config(std::optional<std::filesystem::path> config_path)
+    : sdl2_config_loc{config_path.value_or(default_config_path)},
+      sdl2_config{std::make_unique<INIReader>(FS::PathToUTF8String(sdl2_config_loc))} {
     Reload();
 }
 
@@ -66,6 +68,11 @@ static const std::array<int, Settings::NativeButton::NumButtons> default_buttons
     SDL_SCANCODE_M, SDL_SCANCODE_N, SDL_SCANCODE_1, SDL_SCANCODE_2, SDL_SCANCODE_B,
 };
 
+static const std::array<int, Settings::NativeMotion::NumMotions> default_motions = {
+    SDL_SCANCODE_7,
+    SDL_SCANCODE_8,
+};
+
 static const std::array<std::array<int, 5>, Settings::NativeAnalog::NumAnalogs> default_analogs{{
     {
         SDL_SCANCODE_UP,
@@ -102,27 +109,42 @@ void Config::ReadSetting(const std::string& group, Settings::BasicSetting<Type>&
 void Config::ReadValues() {
     // Controls
     for (std::size_t p = 0; p < Settings::values.players.GetValue().size(); ++p) {
+        auto& player = Settings::values.players.GetValue()[p];
+
         const auto group = fmt::format("ControlsP{}", p);
         for (int i = 0; i < Settings::NativeButton::NumButtons; ++i) {
             std::string default_param = InputCommon::GenerateKeyboardParam(default_buttons[i]);
-            Settings::values.players.GetValue()[p].buttons[i] =
+            player.buttons[i] =
                 sdl2_config->Get(group, Settings::NativeButton::mapping[i], default_param);
-            if (Settings::values.players.GetValue()[p].buttons[i].empty())
-                Settings::values.players.GetValue()[p].buttons[i] = default_param;
+            if (player.buttons[i].empty()) {
+                player.buttons[i] = default_param;
+            }
         }
 
         for (int i = 0; i < Settings::NativeAnalog::NumAnalogs; ++i) {
             std::string default_param = InputCommon::GenerateAnalogParamFromKeys(
                 default_analogs[i][0], default_analogs[i][1], default_analogs[i][2],
                 default_analogs[i][3], default_analogs[i][4], 0.5f);
-            Settings::values.players.GetValue()[p].analogs[i] =
+            player.analogs[i] =
                 sdl2_config->Get(group, Settings::NativeAnalog::mapping[i], default_param);
-            if (Settings::values.players.GetValue()[p].analogs[i].empty())
-                Settings::values.players.GetValue()[p].analogs[i] = default_param;
+            if (player.analogs[i].empty()) {
+                player.analogs[i] = default_param;
+            }
+        }
+
+        for (int i = 0; i < Settings::NativeMotion::NumMotions; ++i) {
+            const std::string default_param =
+                InputCommon::GenerateKeyboardParam(default_motions[i]);
+            auto& player_motions = player.motions[i];
+
+            player_motions =
+                sdl2_config->Get(group, Settings::NativeMotion::mapping[i], default_param);
+            if (player_motions.empty()) {
+                player_motions = default_param;
+            }
         }
 
-        Settings::values.players.GetValue()[p].connected =
-            sdl2_config->GetBoolean(group, "connected", false);
+        player.connected = sdl2_config->GetBoolean(group, "connected", false);
     }
 
     ReadSetting("ControlsGeneral", Settings::values.mouse_enabled);
@@ -246,6 +268,7 @@ void Config::ReadValues() {
 
     // Core
     ReadSetting("Core", Settings::values.use_multi_core);
+    ReadSetting("Core", Settings::values.use_extended_memory_layout);
 
     // Cpu
     ReadSetting("Cpu", Settings::values.cpu_accuracy);
@@ -259,11 +282,14 @@ void Config::ReadValues() {
     ReadSetting("Cpu", Settings::values.cpuopt_misc_ir);
     ReadSetting("Cpu", Settings::values.cpuopt_reduce_misalign_checks);
     ReadSetting("Cpu", Settings::values.cpuopt_fastmem);
+    ReadSetting("Cpu", Settings::values.cpuopt_fastmem_exclusives);
+    ReadSetting("Cpu", Settings::values.cpuopt_recompile_exclusives);
     ReadSetting("Cpu", Settings::values.cpuopt_unsafe_unfuse_fma);
     ReadSetting("Cpu", Settings::values.cpuopt_unsafe_reduce_fp_error);
     ReadSetting("Cpu", Settings::values.cpuopt_unsafe_ignore_standard_fpcr);
     ReadSetting("Cpu", Settings::values.cpuopt_unsafe_inaccurate_nan);
     ReadSetting("Cpu", Settings::values.cpuopt_unsafe_fastmem_check);
+    ReadSetting("Cpu", Settings::values.cpuopt_unsafe_ignore_global_monitor);
 
     // Renderer
     ReadSetting("Renderer", Settings::values.renderer_backend);
diff --git a/src/yuzu_cmd/config.h b/src/yuzu_cmd/config.h
index 1ee932be2..f61ba23ec 100644
--- a/src/yuzu_cmd/config.h
+++ b/src/yuzu_cmd/config.h
@@ -6,6 +6,7 @@
 
 #include <filesystem>
 #include <memory>
+#include <optional>
 #include <string>
 
 #include "common/settings.h"
@@ -13,14 +14,14 @@
 class INIReader;
 
 class Config {
-    std::unique_ptr<INIReader> sdl2_config;
     std::filesystem::path sdl2_config_loc;
+    std::unique_ptr<INIReader> sdl2_config;
 
     bool LoadINI(const std::string& default_contents = "", bool retry = true);
     void ReadValues();
 
 public:
-    Config();
+    explicit Config(std::optional<std::filesystem::path> config_path);
     ~Config();
 
     void Reload();
diff --git a/src/yuzu_cmd/default_ini.h b/src/yuzu_cmd/default_ini.h
index 6d613bf7a..34782c378 100644
--- a/src/yuzu_cmd/default_ini.h
+++ b/src/yuzu_cmd/default_ini.h
@@ -124,7 +124,11 @@ keyboard_enabled =
 [Core]
 # Whether to use multi-core for CPU emulation
 # 0: Disabled, 1 (default): Enabled
-use_multi_core=
+use_multi_core =
+
+# Enable extended guest system memory layout (6GB DRAM)
+# 0 (default): Disabled, 1: Enabled
+use_extended_memory_layout =
 
 [Cpu]
 # Adjusts various optimizations.
@@ -174,6 +178,14 @@ cpuopt_reduce_misalign_checks =
 # 0: Disabled, 1 (default): Enabled
 cpuopt_fastmem =
 
+# Enable Host MMU Emulation for exclusive memory instructions (faster guest memory access)
+# 0: Disabled, 1 (default): Enabled
+cpuopt_fastmem_exclusives =
+
+# Enable fallback on failure of fastmem of exclusive memory instructions (faster guest memory access)
+# 0: Disabled, 1 (default): Enabled
+cpuopt_recompile_exclusives =
+
 # Enable unfuse FMA (improve performance on CPUs without FMA)
 # Only enabled if cpu_accuracy is set to Unsafe. Automatically chosen with cpu_accuracy = Auto-select.
 # 0: Disabled, 1 (default): Enabled
@@ -199,6 +211,11 @@ cpuopt_unsafe_inaccurate_nan =
 # 0: Disabled, 1 (default): Enabled
 cpuopt_unsafe_fastmem_check =
 
+# Enable faster exclusive instructions
+# Only enabled if cpu_accuracy is set to Unsafe. Automatically chosen with cpu_accuracy = Auto-select.
+# 0: Disabled, 1 (default): Enabled
+cpuopt_unsafe_ignore_global_monitor =
+
 [Renderer]
 # Which backend API to use.
 # 0 (default): OpenGL, 1: Vulkan
diff --git a/src/yuzu_cmd/yuzu.cpp b/src/yuzu_cmd/yuzu.cpp
index b44ea0cc4..f6d563017 100644
--- a/src/yuzu_cmd/yuzu.cpp
+++ b/src/yuzu_cmd/yuzu.cpp
@@ -66,7 +66,8 @@ static void PrintHelp(const char* argv0) {
                  "-f, --fullscreen      Start in fullscreen mode\n"
                  "-h, --help            Display this help and exit\n"
                  "-v, --version         Output version information and exit\n"
-                 "-p, --program         Pass following string as arguments to executable\n";
+                 "-p, --program         Pass following string as arguments to executable\n"
+                 "-c, --config          Load the specified configuration file\n";
 }
 
 static void PrintVersion() {
@@ -78,7 +79,6 @@ int main(int argc, char** argv) {
     Common::Log::Initialize();
     Common::Log::SetColorConsoleBackendEnabled(true);
     Common::DetachedTasks detached_tasks;
-    Config config;
 
     int option_index = 0;
 #ifdef _WIN32
@@ -91,19 +91,24 @@ int main(int argc, char** argv) {
     }
 #endif
     std::string filepath;
+    std::optional<std::string> config_path;
+    std::string program_args;
 
     bool fullscreen = false;
 
     static struct option long_options[] = {
+        // clang-format off
         {"fullscreen", no_argument, 0, 'f'},
         {"help", no_argument, 0, 'h'},
         {"version", no_argument, 0, 'v'},
         {"program", optional_argument, 0, 'p'},
+        {"config", required_argument, 0, 'c'},
         {0, 0, 0, 0},
+        // clang-format on
     };
 
     while (optind < argc) {
-        int arg = getopt_long(argc, argv, "g:fhvp::", long_options, &option_index);
+        int arg = getopt_long(argc, argv, "g:fhvp::c:", long_options, &option_index);
         if (arg != -1) {
             switch (static_cast<char>(arg)) {
             case 'f':
@@ -117,9 +122,12 @@ int main(int argc, char** argv) {
                 PrintVersion();
                 return 0;
             case 'p':
-                Settings::values.program_args = argv[optind];
+                program_args = argv[optind];
                 ++optind;
                 break;
+            case 'c':
+                config_path = optarg;
+                break;
             }
         } else {
 #ifdef _WIN32
@@ -131,6 +139,12 @@ int main(int argc, char** argv) {
         }
     }
 
+    Config config{config_path};
+
+    if (!program_args.empty()) {
+        Settings::values.program_args = program_args;
+    }
+
 #ifdef _WIN32
     LocalFree(argv_w);
 #endif