[vulkan] Correction on unconditional/ harcoded makeview 2D array vs compute shader 3D views

This reverts commit a9c4c8aefd.

Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4157
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
Reviewed-by: MaranBr <maranbr@eden-emu.dev>

src/android/app/src/main/java/org/yuzu/yuzu_emu/features/settings/model/BooleanSetting.kt

@@ -31,6 +31,7 @@ enum class BooleanSetting(override val key: String) : AbstractBooleanSetting {
     RENDERER_REACTIVE_FLUSHING("use_reactive_flushing"),
     ENABLE_BUFFER_HISTORY("enable_buffer_history"),
     USE_OPTIMIZED_VERTEX_BUFFERS("use_optimized_vertex_buffers"),
+    ENABLE_GPU_BUFFER_READBACK("enable_gpu_buffer_readback"),
     SYNC_MEMORY_OPERATIONS("sync_memory_operations"),
     BUFFER_REORDER_DISABLE("disable_buffer_reorder"),
     RENDERER_DEBUG("debug"),
@@ -83,9 +84,10 @@ enum class BooleanSetting(override val key: String) : AbstractBooleanSetting {
     ENABLE_OVERLAY("enable_overlay"),
 
     // GPU Logging
-    GPU_LOGGING_ENABLED("gpu_logging_enabled"),
     GPU_LOG_VULKAN_CALLS("gpu_log_vulkan_calls"),
     GPU_LOG_SHADER_DUMPS("gpu_log_shader_dumps"),
+    DUMP_GUEST_SHADERS("dump_guest_shaders"),
+    DUMP_MACROS("dump_macros"),
     GPU_LOG_MEMORY_TRACKING("gpu_log_memory_tracking"),
     GPU_LOG_DRIVER_DEBUG("gpu_log_driver_debug"),
 

src/android/app/src/main/java/org/yuzu/yuzu_emu/features/settings/model/view/SettingsItem.kt

@@ -806,6 +806,13 @@ abstract class SettingsItem(
                     descriptionId = R.string.enable_buffer_history_description
                 )
             )
+            put(
+                SwitchSetting(
+                    BooleanSetting.ENABLE_GPU_BUFFER_READBACK,
+                    titleId = R.string.enable_gpu_buffer_readback,
+                    descriptionId = R.string.enable_gpu_buffer_readback_description
+                )
+            )
             put(
                 SwitchSetting(
                     BooleanSetting.USE_OPTIMIZED_VERTEX_BUFFERS,
@@ -931,13 +938,6 @@ abstract class SettingsItem(
             )
 
             // GPU Logging settings
-            put(
-                SwitchSetting(
-                    BooleanSetting.GPU_LOGGING_ENABLED,
-                    titleId = R.string.gpu_logging_enabled,
-                    descriptionId = R.string.gpu_logging_enabled_description
-                )
-            )
             put(
                 SingleChoiceSetting(
                     ByteSetting.GPU_LOG_LEVEL,
@@ -954,6 +954,13 @@ abstract class SettingsItem(
                     descriptionId = R.string.gpu_log_vulkan_calls_description
                 )
             )
+            put(
+                SwitchSetting(
+                    BooleanSetting.DUMP_GUEST_SHADERS,
+                    titleId = R.string.dump_guest_shaders,
+                    descriptionId = R.string.dump_guest_shaders_description
+                )
+            )
             put(
                 SwitchSetting(
                     BooleanSetting.GPU_LOG_SHADER_DUMPS,
@@ -961,6 +968,13 @@ abstract class SettingsItem(
                     descriptionId = R.string.gpu_log_shader_dumps_description
                 )
             )
+            put(
+                SwitchSetting(
+                    BooleanSetting.DUMP_MACROS,
+                    titleId = R.string.dump_macros,
+                    descriptionId = R.string.dump_macros_description
+                )
+            )
             put(
                 SwitchSetting(
                     BooleanSetting.GPU_LOG_MEMORY_TRACKING,

src/android/app/src/main/java/org/yuzu/yuzu_emu/features/settings/ui/SettingsFragmentPresenter.kt

@@ -292,6 +292,7 @@ class SettingsFragmentPresenter(
             add(BooleanSetting.RENDERER_FORCE_MAX_CLOCK.key)
             add(BooleanSetting.RENDERER_REACTIVE_FLUSHING.key)
             add(BooleanSetting.ENABLE_BUFFER_HISTORY.key)
+            add(BooleanSetting.ENABLE_GPU_BUFFER_READBACK.key)
             add(BooleanSetting.USE_OPTIMIZED_VERTEX_BUFFERS.key)
 
             add(HeaderSetting(R.string.hacks))
@@ -1288,14 +1289,17 @@ class SettingsFragmentPresenter(
             add(ShortSetting.DEBUG_KNOBS.key)
             add(StringSetting.PROGRAM_ARGS.key)
 
-            add(HeaderSetting(R.string.gpu_logging_header))
-            add(BooleanSetting.GPU_LOGGING_ENABLED.key)
-            add(ByteSetting.GPU_LOG_LEVEL.key)
-            add(BooleanSetting.GPU_LOG_VULKAN_CALLS.key)
-            add(BooleanSetting.GPU_LOG_SHADER_DUMPS.key)
-            add(BooleanSetting.GPU_LOG_MEMORY_TRACKING.key)
-            add(BooleanSetting.GPU_LOG_DRIVER_DEBUG.key)
-            add(IntSetting.GPU_LOG_RING_BUFFER_SIZE.key)
+            if (!NativeConfig.isPerGameConfigLoaded()) {
+                add(HeaderSetting(R.string.gpu_logging_header))
+                add(ByteSetting.GPU_LOG_LEVEL.key)
+                add(BooleanSetting.GPU_LOG_VULKAN_CALLS.key)
+                add(BooleanSetting.DUMP_GUEST_SHADERS.key)
+                add(BooleanSetting.GPU_LOG_SHADER_DUMPS.key)
+                add(BooleanSetting.DUMP_MACROS.key)
+                add(BooleanSetting.GPU_LOG_MEMORY_TRACKING.key)
+                add(BooleanSetting.GPU_LOG_DRIVER_DEBUG.key)
+                add(IntSetting.GPU_LOG_RING_BUFFER_SIZE.key)
+            }
         }
     }
 

src/android/app/src/main/res/values-ar/strings.xml

@@ -569,8 +569,6 @@
 
     <!-- GPU Logging strings -->
     <string name="gpu_logging_header">تسجيل وحدة معالجة الرسومات</string>
-    <string name="gpu_logging_enabled">تمكين تسجيل وحدة معالجة الرسومات</string>
-    <string name="gpu_logging_enabled_description">تسجيل عمليات وحدة معالجة الرسومات في ملف eden_gpu.log لتصحيح أخطاء برامج تشغيل Adreno</string>
     <string name="gpu_log_level">مستوى السجل</string>
     <string name="gpu_log_level_description">مستوى التفاصيل لسجلات وحدة معالجة الرسومات (كلما زاد المستوى، زادت التفاصيل وزادت التكاليف الإضافية)</string>
     <string name="gpu_log_vulkan_calls">تسجيل استدعاءات واجهة برمجة تطبيقات Vulkan</string>

src/android/app/src/main/res/values-es/strings.xml

@@ -563,8 +563,6 @@
 
     <!-- GPU Logging strings -->
     <string name="gpu_logging_header">Registros de la GPU</string>
-    <string name="gpu_logging_enabled">Activar los registros de la GPU</string>
-    <string name="gpu_logging_enabled_description">Registra las operaciones de la GPU en eden_gpu.log para la depuración de los controladores de Adreno</string>
     <string name="gpu_log_level">Nivel de registros</string>
     <string name="gpu_log_level_description">Nivel de detalle de los registros de la GPU (más alto = más detalles, más sobrecarga)</string>
     <string name="gpu_log_vulkan_calls">Registros de llamadas del API de Vulkan</string>

src/android/app/src/main/res/values-fr/strings.xml

@@ -525,7 +525,6 @@
 
     <!-- GPU Logging strings -->
     <string name="gpu_logging_header">Journalisation GPU</string>
-    <string name="gpu_logging_enabled">Activer la journalisation GPU</string>
     <string name="gpu_log_level">Niveau de journalisation</string>
     <string name="gpu_log_vulkan_calls">Journaliser les appels API Vulkan</string>
     <string name="gpu_log_shader_dumps">Extraire les shaders</string>

src/android/app/src/main/res/values-ru/strings.xml

@@ -562,8 +562,6 @@
 
     <!-- GPU Logging strings -->
     <string name="gpu_logging_header">Ведение журнала ГПУ</string>
-    <string name="gpu_logging_enabled">Включить ведение журнала ГПУ</string>
-    <string name="gpu_logging_enabled_description">Записывать операции ГПУ в файл eden_gpu.log для отладки драйверов Adreno</string>
     <string name="gpu_log_level">Уровень журналирования</string>
     <string name="gpu_log_level_description">Уровень детализации логов ГПУ (больше значение = больше деталей, выше нагрузка)</string>
     <string name="gpu_log_vulkan_calls">Записывать вызовы Vulkan API</string>

src/android/app/src/main/res/values-uk/strings.xml

@@ -565,8 +565,6 @@
 
     <!-- GPU Logging strings -->
     <string name="gpu_logging_header">Журналювання ГП</string>
-    <string name="gpu_logging_enabled">Увімкнути журналювання ГП</string>
-    <string name="gpu_logging_enabled_description">Журналювати операції ГП до eden_gpu.log для зневадження драйверів Adreno</string>
     <string name="gpu_log_level">Рівень журналювання</string>
     <string name="gpu_log_level_description">Рівень подробиць у журналі ГП (вищий = більше подробиць, більший вплив на швидкодію)</string>
     <string name="gpu_log_vulkan_calls">Записувати виклики API Vulkan</string>

src/android/app/src/main/res/values-zh-rCN/strings.xml

@@ -559,8 +559,6 @@
 
     <!-- GPU Logging strings -->
     <string name="gpu_logging_header">GPU 日志</string>
-    <string name="gpu_logging_enabled">启用 GPU 日志</string>
-    <string name="gpu_logging_enabled_description">将 GPU 操作记录至 eden_gpu.log 以供调试 Adreno 驱动</string>
     <string name="gpu_log_level">日志等级</string>
     <string name="gpu_log_level_description">GPU 日志的详细级别(数值越高 = 细节越多,开销越大)</string>
     <string name="gpu_log_vulkan_calls">记录 Vulkan API 调用</string>

src/android/app/src/main/res/values/strings.xml

@@ -503,6 +503,8 @@
     <string name="renderer_reactive_flushing_description">Improves rendering accuracy in some games at the cost of performance.</string>
     <string name="enable_buffer_history">Enable buffer history</string>
     <string name="enable_buffer_history_description">Enables access to previous buffer states. This option may improve rendering quality and performance consistency in some games.</string>
+    <string name="enable_gpu_buffer_readback">Enable GPU Buffer Readback</string>
+    <string name="enable_gpu_buffer_readback_description">Preserves GPU-modified buffer data by reading it back before uploads. Some games require this to render certain effects properly. May cause issues if the hardware cannot handle the additional workload.</string>
     <string name="use_optimized_vertex_buffers">Optimized Vertex Buffers</string>
     <string name="use_optimized_vertex_buffers_description">Enables optimized vertex buffer binding for improved performance. Requires Mesa 26.0+ Turnip drivers/ QCOM drivers. Will crash on older Turnip drivers (25.3 and below).</string>
 
@@ -575,14 +577,16 @@
 
     <!-- GPU Logging strings -->
     <string name="gpu_logging_header">GPU Logging</string>
-    <string name="gpu_logging_enabled">Enable GPU Logging</string>
-    <string name="gpu_logging_enabled_description">Log GPU operations to eden_gpu.log for debugging Adreno drivers</string>
     <string name="gpu_log_level">Log Level</string>
     <string name="gpu_log_level_description">Detail level for GPU logs (higher = more detail, more overhead)</string>
     <string name="gpu_log_vulkan_calls">Log Vulkan API Calls</string>
     <string name="gpu_log_vulkan_calls_description">Track all Vulkan API calls in ring buffer</string>
-    <string name="gpu_log_shader_dumps">Dump Shaders</string>
-    <string name="gpu_log_shader_dumps_description">Save compiled shader SPIR-V to files</string>
+    <string name="gpu_log_shader_dumps">Dump SPIR-V Shaders</string>
+    <string name="gpu_log_shader_dumps_description">Save recompiled SPIR-V binaries (.spv) to dump folder. Inspect with spirv-dis/spirv-cross/spirv-val.</string>
+    <string name="dump_guest_shaders">Dump Guest (Maxwell) Shaders</string>
+    <string name="dump_guest_shaders_description">Save Maxwell guest shader bytecode files (*.ash) to dump folder. Inspect with nvdisasm.</string>
+    <string name="dump_macros">Dump Maxwell Macros</string>
+    <string name="dump_macros_description">Save Maxwell macro program files (*.macro) to dump folder. Inspect with envydis.</string>
     <string name="gpu_log_memory_tracking">Track GPU Memory</string>
     <string name="gpu_log_memory_tracking_description">Monitor GPU memory allocations and deallocations</string>
     <string name="gpu_log_driver_debug">Driver Debug Info</string>

src/common/CMakeLists.txt

@@ -50,7 +50,6 @@ add_library(
     elf.h
     error.cpp
     error.h
-    expected.h
     fiber.cpp
     fiber.h
     fixed_point.h

src/common/expected.h

@@ -1,986 +0,0 @@
-// SPDX-FileCopyrightText: Copyright 2021 yuzu Emulator Project
-// SPDX-License-Identifier: GPL-2.0-or-later
-
-// This is based on the proposed implementation of std::expected (P0323)
-// https://github.com/TartanLlama/expected/blob/master/include/tl/expected.hpp
-
-#pragma once
-
-#include <type_traits>
-#include <utility>
-
-namespace Common {
-
-template <typename T, typename E>
-class Expected;
-
-template <typename E>
-class Unexpected {
-public:
-    Unexpected() = delete;
-
-    constexpr explicit Unexpected(const E& e) : m_val{e} {}
-
-    constexpr explicit Unexpected(E&& e) : m_val{std::move(e)} {}
-
-    constexpr E& value() & {
-        return m_val;
-    }
-
-    constexpr const E& value() const& {
-        return m_val;
-    }
-
-    constexpr E&& value() && {
-        return std::move(m_val);
-    }
-
-    constexpr const E&& value() const&& {
-        return std::move(m_val);
-    }
-
-private:
-    E m_val;
-};
-
-template <typename E>
-constexpr auto operator<=>(const Unexpected<E>& lhs, const Unexpected<E>& rhs) {
-    return lhs.value() <=> rhs.value();
-}
-
-struct unexpect_t {
-    constexpr explicit unexpect_t() = default;
-};
-
-namespace detail {
-
-struct no_init_t {
-    constexpr explicit no_init_t() = default;
-};
-
-/**
- * This specialization is for when T is not trivially destructible,
- * so the destructor must be called on destruction of `expected'
- * Additionally, this requires E to be trivially destructible
- */
-template <typename T, typename E, bool = std::is_trivially_destructible_v<T>>
-    requires std::is_trivially_destructible_v<E>
-struct expected_storage_base {
-    constexpr expected_storage_base() : m_val{T{}}, m_has_val{true} {}
-
-    constexpr expected_storage_base(no_init_t) : m_has_val{false} {}
-
-    template <typename... Args, std::enable_if_t<std::is_constructible_v<T, Args&&...>>* = nullptr>
-    constexpr expected_storage_base(std::in_place_t, Args&&... args)
-        : m_val{std::forward<Args>(args)...}, m_has_val{true} {}
-
-    template <typename U, typename... Args,
-              std::enable_if_t<std::is_constructible_v<T, std::initializer_list<U>&, Args&&...>>* =
-                  nullptr>
-    constexpr expected_storage_base(std::in_place_t, std::initializer_list<U> il, Args&&... args)
-        : m_val{il, std::forward<Args>(args)...}, m_has_val{true} {}
-
-    template <typename... Args, std::enable_if_t<std::is_constructible_v<E, Args&&...>>* = nullptr>
-    constexpr explicit expected_storage_base(unexpect_t, Args&&... args)
-        : m_unexpect{std::forward<Args>(args)...}, m_has_val{false} {}
-
-    template <typename U, typename... Args,
-              std::enable_if_t<std::is_constructible_v<E, std::initializer_list<U>&, Args&&...>>* =
-                  nullptr>
-    constexpr explicit expected_storage_base(unexpect_t, std::initializer_list<U> il,
-                                             Args&&... args)
-        : m_unexpect{il, std::forward<Args>(args)...}, m_has_val{false} {}
-
-    ~expected_storage_base() {
-        if (m_has_val) {
-            m_val.~T();
-        }
-    }
-
-    union {
-        T m_val;
-        Unexpected<E> m_unexpect;
-    };
-
-    bool m_has_val;
-};
-
-/**
- * This specialization is for when T is trivially destructible,
- * so the destructor of `expected` can be trivial
- * Additionally, this requires E to be trivially destructible
- */
-template <typename T, typename E>
-    requires std::is_trivially_destructible_v<E>
-struct expected_storage_base<T, E, true> {
-    constexpr expected_storage_base() : m_val{T{}}, m_has_val{true} {}
-
-    constexpr expected_storage_base(no_init_t) : m_has_val{false} {}
-
-    template <typename... Args, std::enable_if_t<std::is_constructible_v<T, Args&&...>>* = nullptr>
-    constexpr expected_storage_base(std::in_place_t, Args&&... args)
-        : m_val{std::forward<Args>(args)...}, m_has_val{true} {}
-
-    template <typename U, typename... Args,
-              std::enable_if_t<std::is_constructible_v<T, std::initializer_list<U>&, Args&&...>>* =
-                  nullptr>
-    constexpr expected_storage_base(std::in_place_t, std::initializer_list<U> il, Args&&... args)
-        : m_val{il, std::forward<Args>(args)...}, m_has_val{true} {}
-
-    template <typename... Args, std::enable_if_t<std::is_constructible_v<E, Args&&...>>* = nullptr>
-    constexpr explicit expected_storage_base(unexpect_t, Args&&... args)
-        : m_unexpect{std::forward<Args>(args)...}, m_has_val{false} {}
-
-    template <typename U, typename... Args,
-              std::enable_if_t<std::is_constructible_v<E, std::initializer_list<U>&, Args&&...>>* =
-                  nullptr>
-    constexpr explicit expected_storage_base(unexpect_t, std::initializer_list<U> il,
-                                             Args&&... args)
-        : m_unexpect{il, std::forward<Args>(args)...}, m_has_val{false} {}
-
-    ~expected_storage_base() = default;
-
-    union {
-        T m_val;
-        Unexpected<E> m_unexpect;
-    };
-
-    bool m_has_val;
-};
-
-template <typename T, typename E>
-struct expected_operations_base : expected_storage_base<T, E> {
-    using expected_storage_base<T, E>::expected_storage_base;
-
-    template <typename... Args>
-    void construct(Args&&... args) noexcept {
-        new (std::addressof(this->m_val)) T{std::forward<Args>(args)...};
-        this->m_has_val = true;
-    }
-
-    template <typename Rhs>
-    void construct_with(Rhs&& rhs) noexcept {
-        new (std::addressof(this->m_val)) T{std::forward<Rhs>(rhs).get()};
-        this->m_has_val = true;
-    }
-
-    template <typename... Args>
-    void construct_error(Args&&... args) noexcept {
-        new (std::addressof(this->m_unexpect)) Unexpected<E>{std::forward<Args>(args)...};
-        this->m_has_val = false;
-    }
-
-    void assign(const expected_operations_base& rhs) noexcept {
-        if (!this->m_has_val && rhs.m_has_val) {
-            geterr().~Unexpected<E>();
-            construct(rhs.get());
-        } else {
-            assign_common(rhs);
-        }
-    }
-
-    void assign(expected_operations_base&& rhs) noexcept {
-        if (!this->m_has_val && rhs.m_has_val) {
-            geterr().~Unexpected<E>();
-            construct(std::move(rhs).get());
-        } else {
-            assign_common(rhs);
-        }
-    }
-
-    template <typename Rhs>
-    void assign_common(Rhs&& rhs) {
-        if (this->m_has_val) {
-            if (rhs.m_has_val) {
-                get() = std::forward<Rhs>(rhs).get();
-            } else {
-                destroy_val();
-                construct_error(std::forward<Rhs>(rhs).geterr());
-            }
-        } else {
-            if (!rhs.m_has_val) {
-                geterr() = std::forward<Rhs>(rhs).geterr();
-            }
-        }
-    }
-
-    bool has_value() const {
-        return this->m_has_val;
-    }
-
-    constexpr T& get() & {
-        return this->m_val;
-    }
-
-    constexpr const T& get() const& {
-        return this->m_val;
-    }
-
-    constexpr T&& get() && {
-        return std::move(this->m_val);
-    }
-
-    constexpr const T&& get() const&& {
-        return std::move(this->m_val);
-    }
-
-    constexpr Unexpected<E>& geterr() & {
-        return this->m_unexpect;
-    }
-
-    constexpr const Unexpected<E>& geterr() const& {
-        return this->m_unexpect;
-    }
-
-    constexpr Unexpected<E>&& geterr() && {
-        return std::move(this->m_unexpect);
-    }
-
-    constexpr const Unexpected<E>&& geterr() const&& {
-        return std::move(this->m_unexpect);
-    }
-
-    constexpr void destroy_val() {
-        get().~T();
-    }
-};
-
-/**
- * This manages conditionally having a trivial copy constructor
- * This specialization is for when T is trivially copy constructible
- * Additionally, this requires E to be trivially copy constructible
- */
-template <typename T, typename E, bool = std::is_trivially_copy_constructible_v<T>>
-    requires std::is_trivially_copy_constructible_v<E>
-struct expected_copy_base : expected_operations_base<T, E> {
-    using expected_operations_base<T, E>::expected_operations_base;
-};
-
-/**
- * This specialization is for when T is not trivially copy constructible
- * Additionally, this requires E to be trivially copy constructible
- */
-template <typename T, typename E>
-    requires std::is_trivially_copy_constructible_v<E>
-struct expected_copy_base<T, E, false> : expected_operations_base<T, E> {
-    using expected_operations_base<T, E>::expected_operations_base;
-
-    expected_copy_base() = default;
-
-    expected_copy_base(const expected_copy_base& rhs)
-        : expected_operations_base<T, E>{no_init_t{}} {
-        if (rhs.has_value()) {
-            this->construct_with(rhs);
-        } else {
-            this->construct_error(rhs.geterr());
-        }
-    }
-
-    expected_copy_base(expected_copy_base&&) = default;
-
-    expected_copy_base& operator=(const expected_copy_base&) = default;
-
-    expected_copy_base& operator=(expected_copy_base&&) = default;
-};
-
-/**
- * This manages conditionally having a trivial move constructor
- * This specialization is for when T is trivially move constructible
- * Additionally, this requires E to be trivially move constructible
- */
-template <typename T, typename E, bool = std::is_trivially_move_constructible_v<T>>
-    requires std::is_trivially_move_constructible_v<E>
-struct expected_move_base : expected_copy_base<T, E> {
-    using expected_copy_base<T, E>::expected_copy_base;
-};
-
-/**
- * This specialization is for when T is not trivially move constructible
- * Additionally, this requires E to be trivially move constructible
- */
-template <typename T, typename E>
-    requires std::is_trivially_move_constructible_v<E>
-struct expected_move_base<T, E, false> : expected_copy_base<T, E> {
-    using expected_copy_base<T, E>::expected_copy_base;
-
-    expected_move_base() = default;
-
-    expected_move_base(const expected_move_base&) = default;
-
-    expected_move_base(expected_move_base&& rhs) noexcept(std::is_nothrow_move_constructible_v<T>)
-        : expected_copy_base<T, E>{no_init_t{}} {
-        if (rhs.has_value()) {
-            this->construct_with(std::move(rhs));
-        } else {
-            this->construct_error(std::move(rhs.geterr()));
-        }
-    }
-
-    expected_move_base& operator=(const expected_move_base&) = default;
-
-    expected_move_base& operator=(expected_move_base&&) = default;
-};
-
-/**
- * This manages conditionally having a trivial copy assignment operator
- * This specialization is for when T is trivially copy assignable
- * Additionally, this requires E to be trivially copy assignable
- */
-template <typename T, typename E,
-          bool = std::conjunction_v<std::is_trivially_copy_assignable<T>,
-                                    std::is_trivially_copy_constructible<T>,
-                                    std::is_trivially_destructible<T>>>
-    requires std::conjunction_v<std::is_trivially_copy_assignable<E>,
-                                std::is_trivially_copy_constructible<E>,
-                                std::is_trivially_destructible<E>>
-struct expected_copy_assign_base : expected_move_base<T, E> {
-    using expected_move_base<T, E>::expected_move_base;
-};
-
-/**
- * This specialization is for when T is not trivially copy assignable
- * Additionally, this requires E to be trivially copy assignable
- */
-template <typename T, typename E>
-    requires std::conjunction_v<std::is_trivially_copy_assignable<E>,
-                                std::is_trivially_copy_constructible<E>,
-                                std::is_trivially_destructible<E>>
-struct expected_copy_assign_base<T, E, false> : expected_move_base<T, E> {
-    using expected_move_base<T, E>::expected_move_base;
-
-    expected_copy_assign_base() = default;
-
-    expected_copy_assign_base(const expected_copy_assign_base&) = default;
-
-    expected_copy_assign_base(expected_copy_assign_base&&) = default;
-
-    expected_copy_assign_base& operator=(const expected_copy_assign_base& rhs) {
-        this->assign(rhs);
-        return *this;
-    }
-
-    expected_copy_assign_base& operator=(expected_copy_assign_base&&) = default;
-};
-
-/**
- * This manages conditionally having a trivial move assignment operator
- * This specialization is for when T is trivially move assignable
- * Additionally, this requires E to be trivially move assignable
- */
-template <typename T, typename E,
-          bool = std::conjunction_v<std::is_trivially_move_assignable<T>,
-                                    std::is_trivially_move_constructible<T>,
-                                    std::is_trivially_destructible<T>>>
-    requires std::conjunction_v<std::is_trivially_move_assignable<E>,
-                                std::is_trivially_move_constructible<E>,
-                                std::is_trivially_destructible<E>>
-struct expected_move_assign_base : expected_copy_assign_base<T, E> {
-    using expected_copy_assign_base<T, E>::expected_copy_assign_base;
-};
-
-/**
- * This specialization is for when T is not trivially move assignable
- * Additionally, this requires E to be trivially move assignable
- */
-template <typename T, typename E>
-    requires std::conjunction_v<std::is_trivially_move_assignable<E>,
-                                std::is_trivially_move_constructible<E>,
-                                std::is_trivially_destructible<E>>
-struct expected_move_assign_base<T, E, false> : expected_copy_assign_base<T, E> {
-    using expected_copy_assign_base<T, E>::expected_copy_assign_base;
-
-    expected_move_assign_base() = default;
-
-    expected_move_assign_base(const expected_move_assign_base&) = default;
-
-    expected_move_assign_base(expected_move_assign_base&&) = default;
-
-    expected_move_assign_base& operator=(const expected_move_assign_base&) = default;
-
-    expected_move_assign_base& operator=(expected_move_assign_base&& rhs) noexcept(
-        std::conjunction_v<std::is_nothrow_move_constructible<T>,
-                           std::is_nothrow_move_assignable<T>>) {
-        this->assign(std::move(rhs));
-        return *this;
-    }
-};
-
-/**
- * expected_delete_ctor_base will conditionally delete copy and move constructors
- * depending on whether T is copy/move constructible
- * Additionally, this requires E to be copy/move constructible
- */
-template <typename T, typename E, bool EnableCopy = std::is_copy_constructible_v<T>,
-          bool EnableMove = std::is_move_constructible_v<T>>
-    requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>>
-struct expected_delete_ctor_base {
-    expected_delete_ctor_base() = default;
-    expected_delete_ctor_base(const expected_delete_ctor_base&) = default;
-    expected_delete_ctor_base(expected_delete_ctor_base&&) noexcept = default;
-    expected_delete_ctor_base& operator=(const expected_delete_ctor_base&) = default;
-    expected_delete_ctor_base& operator=(expected_delete_ctor_base&&) noexcept = default;
-};
-
-template <typename T, typename E>
-    requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>>
-struct expected_delete_ctor_base<T, E, true, false> {
-    expected_delete_ctor_base() = default;
-    expected_delete_ctor_base(const expected_delete_ctor_base&) = default;
-    expected_delete_ctor_base(expected_delete_ctor_base&&) noexcept = delete;
-    expected_delete_ctor_base& operator=(const expected_delete_ctor_base&) = default;
-    expected_delete_ctor_base& operator=(expected_delete_ctor_base&&) noexcept = default;
-};
-
-template <typename T, typename E>
-    requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>>
-struct expected_delete_ctor_base<T, E, false, true> {
-    expected_delete_ctor_base() = default;
-    expected_delete_ctor_base(const expected_delete_ctor_base&) = delete;
-    expected_delete_ctor_base(expected_delete_ctor_base&&) noexcept = default;
-    expected_delete_ctor_base& operator=(const expected_delete_ctor_base&) = default;
-    expected_delete_ctor_base& operator=(expected_delete_ctor_base&&) noexcept = default;
-};
-
-template <typename T, typename E>
-    requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>>
-struct expected_delete_ctor_base<T, E, false, false> {
-    expected_delete_ctor_base() = default;
-    expected_delete_ctor_base(const expected_delete_ctor_base&) = delete;
-    expected_delete_ctor_base(expected_delete_ctor_base&&) noexcept = delete;
-    expected_delete_ctor_base& operator=(const expected_delete_ctor_base&) = default;
-    expected_delete_ctor_base& operator=(expected_delete_ctor_base&&) noexcept = default;
-};
-
-/**
- * expected_delete_assign_base will conditionally delete copy and move assignment operators
- * depending on whether T is copy/move constructible + assignable
- * Additionally, this requires E to be copy/move constructible + assignable
- */
-template <
-    typename T, typename E,
-    bool EnableCopy = std::conjunction_v<std::is_copy_constructible<T>, std::is_copy_assignable<T>>,
-    bool EnableMove = std::conjunction_v<std::is_move_constructible<T>, std::is_move_assignable<T>>>
-    requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>,
-                                std::is_copy_assignable<E>, std::is_move_assignable<E>>
-struct expected_delete_assign_base {
-    expected_delete_assign_base() = default;
-    expected_delete_assign_base(const expected_delete_assign_base&) = default;
-    expected_delete_assign_base(expected_delete_assign_base&&) noexcept = default;
-    expected_delete_assign_base& operator=(const expected_delete_assign_base&) = default;
-    expected_delete_assign_base& operator=(expected_delete_assign_base&&) noexcept = default;
-};
-
-template <typename T, typename E>
-    requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>,
-                                std::is_copy_assignable<E>, std::is_move_assignable<E>>
-struct expected_delete_assign_base<T, E, true, false> {
-    expected_delete_assign_base() = default;
-    expected_delete_assign_base(const expected_delete_assign_base&) = default;
-    expected_delete_assign_base(expected_delete_assign_base&&) noexcept = default;
-    expected_delete_assign_base& operator=(const expected_delete_assign_base&) = default;
-    expected_delete_assign_base& operator=(expected_delete_assign_base&&) noexcept = delete;
-};
-
-template <typename T, typename E>
-    requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>,
-                                std::is_copy_assignable<E>, std::is_move_assignable<E>>
-struct expected_delete_assign_base<T, E, false, true> {
-    expected_delete_assign_base() = default;
-    expected_delete_assign_base(const expected_delete_assign_base&) = default;
-    expected_delete_assign_base(expected_delete_assign_base&&) noexcept = default;
-    expected_delete_assign_base& operator=(const expected_delete_assign_base&) = delete;
-    expected_delete_assign_base& operator=(expected_delete_assign_base&&) noexcept = default;
-};
-
-template <typename T, typename E>
-    requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>,
-                                std::is_copy_assignable<E>, std::is_move_assignable<E>>
-struct expected_delete_assign_base<T, E, false, false> {
-    expected_delete_assign_base() = default;
-    expected_delete_assign_base(const expected_delete_assign_base&) = default;
-    expected_delete_assign_base(expected_delete_assign_base&&) noexcept = default;
-    expected_delete_assign_base& operator=(const expected_delete_assign_base&) = delete;
-    expected_delete_assign_base& operator=(expected_delete_assign_base&&) noexcept = delete;
-};
-
-/**
- * This is needed to be able to construct the expected_default_ctor_base which follows,
- * while still conditionally deleting the default constructor.
- */
-struct default_constructor_tag {
-    constexpr explicit default_constructor_tag() = default;
-};
-
-/**
- * expected_default_ctor_base will ensure that expected
- * has a deleted default constructor if T is not default constructible
- * This specialization is for when T is default constructible
- */
-template <typename T, typename E, bool Enable = std::is_default_constructible_v<T>>
-struct expected_default_ctor_base {
-    constexpr expected_default_ctor_base() noexcept = default;
-    constexpr expected_default_ctor_base(expected_default_ctor_base const&) noexcept = default;
-    constexpr expected_default_ctor_base(expected_default_ctor_base&&) noexcept = default;
-    expected_default_ctor_base& operator=(expected_default_ctor_base const&) noexcept = default;
-    expected_default_ctor_base& operator=(expected_default_ctor_base&&) noexcept = default;
-
-    constexpr explicit expected_default_ctor_base(default_constructor_tag) {}
-};
-
-template <typename T, typename E>
-struct expected_default_ctor_base<T, E, false> {
-    constexpr expected_default_ctor_base() noexcept = delete;
-    constexpr expected_default_ctor_base(expected_default_ctor_base const&) noexcept = default;
-    constexpr expected_default_ctor_base(expected_default_ctor_base&&) noexcept = default;
-    expected_default_ctor_base& operator=(expected_default_ctor_base const&) noexcept = default;
-    expected_default_ctor_base& operator=(expected_default_ctor_base&&) noexcept = default;
-
-    constexpr explicit expected_default_ctor_base(default_constructor_tag) {}
-};
-
-template <typename T, typename E, typename U>
-using expected_enable_forward_value =
-    std::enable_if_t<std::is_constructible_v<T, U&&> &&
-                     !std::is_same_v<std::remove_cvref_t<U>, std::in_place_t> &&
-                     !std::is_same_v<Expected<T, E>, std::remove_cvref_t<U>> &&
-                     !std::is_same_v<Unexpected<E>, std::remove_cvref_t<U>>>;
-
-template <typename T, typename E, typename U, typename G, typename UR, typename GR>
-using expected_enable_from_other = std::enable_if_t<
-    std::is_constructible_v<T, UR> && std::is_constructible_v<E, GR> &&
-    !std::is_constructible_v<T, Expected<U, G>&> && !std::is_constructible_v<T, Expected<U, G>&&> &&
-    !std::is_constructible_v<T, const Expected<U, G>&> &&
-    !std::is_constructible_v<T, const Expected<U, G>&&> &&
-    !std::is_convertible_v<Expected<U, G>&, T> && !std::is_convertible_v<Expected<U, G>&&, T> &&
-    !std::is_convertible_v<const Expected<U, G>&, T> &&
-    !std::is_convertible_v<const Expected<U, G>&&, T>>;
-
-} // namespace detail
-
-template <typename T, typename E>
-class Expected : private detail::expected_move_assign_base<T, E>,
-                 private detail::expected_delete_ctor_base<T, E>,
-                 private detail::expected_delete_assign_base<T, E>,
-                 private detail::expected_default_ctor_base<T, E> {
-public:
-    using value_type = T;
-    using error_type = E;
-    using unexpected_type = Unexpected<E>;
-
-    constexpr Expected() = default;
-    constexpr Expected(const Expected&) = default;
-    constexpr Expected(Expected&&) = default;
-    Expected& operator=(const Expected&) = default;
-    Expected& operator=(Expected&&) = default;
-
-    template <typename... Args, std::enable_if_t<std::is_constructible_v<T, Args&&...>>* = nullptr>
-    constexpr Expected(std::in_place_t, Args&&... args)
-        : impl_base{std::in_place, std::forward<Args>(args)...},
-          ctor_base{detail::default_constructor_tag{}} {}
-
-    template <typename U, typename... Args,
-              std::enable_if_t<std::is_constructible_v<T, std::initializer_list<U>&, Args&&...>>* =
-                  nullptr>
-    constexpr Expected(std::in_place_t, std::initializer_list<U> il, Args&&... args)
-        : impl_base{std::in_place, il, std::forward<Args>(args)...},
-          ctor_base{detail::default_constructor_tag{}} {}
-
-    template <typename G = E, std::enable_if_t<std::is_constructible_v<E, const G&>>* = nullptr,
-              std::enable_if_t<!std::is_convertible_v<const G&, E>>* = nullptr>
-    constexpr explicit Expected(const Unexpected<G>& e)
-        : impl_base{unexpect_t{}, e.value()}, ctor_base{detail::default_constructor_tag{}} {}
-
-    template <typename G = E, std::enable_if_t<std::is_constructible_v<E, const G&>>* = nullptr,
-              std::enable_if_t<std::is_convertible_v<const G&, E>>* = nullptr>
-    constexpr Expected(Unexpected<G> const& e)
-        : impl_base{unexpect_t{}, e.value()}, ctor_base{detail::default_constructor_tag{}} {}
-
-    template <typename G = E, std::enable_if_t<std::is_constructible_v<E, G&&>>* = nullptr,
-              std::enable_if_t<!std::is_convertible_v<G&&, E>>* = nullptr>
-    constexpr explicit Expected(Unexpected<G>&& e) noexcept(std::is_nothrow_constructible_v<E, G&&>)
-        : impl_base{unexpect_t{}, std::move(e.value())}, ctor_base{
-                                                             detail::default_constructor_tag{}} {}
-
-    template <typename G = E, std::enable_if_t<std::is_constructible_v<E, G&&>>* = nullptr,
-              std::enable_if_t<std::is_convertible_v<G&&, E>>* = nullptr>
-    constexpr Expected(Unexpected<G>&& e) noexcept(std::is_nothrow_constructible_v<E, G&&>)
-        : impl_base{unexpect_t{}, std::move(e.value())}, ctor_base{
-                                                             detail::default_constructor_tag{}} {}
-
-    template <typename... Args, std::enable_if_t<std::is_constructible_v<E, Args&&...>>* = nullptr>
-    constexpr explicit Expected(unexpect_t, Args&&... args)
-        : impl_base{unexpect_t{}, std::forward<Args>(args)...},
-          ctor_base{detail::default_constructor_tag{}} {}
-
-    template <typename U, typename... Args,
-              std::enable_if_t<std::is_constructible_v<E, std::initializer_list<U>&, Args&&...>>* =
-                  nullptr>
-    constexpr explicit Expected(unexpect_t, std::initializer_list<U> il, Args&&... args)
-        : impl_base{unexpect_t{}, il, std::forward<Args>(args)...},
-          ctor_base{detail::default_constructor_tag{}} {}
-
-    template <typename U, typename G,
-              std::enable_if_t<!(std::is_convertible_v<U const&, T> &&
-                                 std::is_convertible_v<G const&, E>)>* = nullptr,
-              detail::expected_enable_from_other<T, E, U, G, const U&, const G&>* = nullptr>
-    constexpr explicit Expected(const Expected<U, G>& rhs)
-        : ctor_base{detail::default_constructor_tag{}} {
-        if (rhs.has_value()) {
-            this->construct(*rhs);
-        } else {
-            this->construct_error(rhs.error());
-        }
-    }
-
-    template <typename U, typename G,
-              std::enable_if_t<(std::is_convertible_v<U const&, T> &&
-                                std::is_convertible_v<G const&, E>)>* = nullptr,
-              detail::expected_enable_from_other<T, E, U, G, const U&, const G&>* = nullptr>
-    constexpr Expected(const Expected<U, G>& rhs) : ctor_base{detail::default_constructor_tag{}} {
-        if (rhs.has_value()) {
-            this->construct(*rhs);
-        } else {
-            this->construct_error(rhs.error());
-        }
-    }
-
-    template <typename U, typename G,
-              std::enable_if_t<!(std::is_convertible_v<U&&, T> && std::is_convertible_v<G&&, E>)>* =
-                  nullptr,
-              detail::expected_enable_from_other<T, E, U, G, U&&, G&&>* = nullptr>
-    constexpr explicit Expected(Expected<U, G>&& rhs)
-        : ctor_base{detail::default_constructor_tag{}} {
-        if (rhs.has_value()) {
-            this->construct(std::move(*rhs));
-        } else {
-            this->construct_error(std::move(rhs.error()));
-        }
-    }
-
-    template <typename U, typename G,
-              std::enable_if_t<(std::is_convertible_v<U&&, T> && std::is_convertible_v<G&&, E>)>* =
-                  nullptr,
-              detail::expected_enable_from_other<T, E, U, G, U&&, G&&>* = nullptr>
-    constexpr Expected(Expected<U, G>&& rhs) : ctor_base{detail::default_constructor_tag{}} {
-        if (rhs.has_value()) {
-            this->construct(std::move(*rhs));
-        } else {
-            this->construct_error(std::move(rhs.error()));
-        }
-    }
-
-    template <typename U = T, std::enable_if_t<!std::is_convertible_v<U&&, T>>* = nullptr,
-              detail::expected_enable_forward_value<T, E, U>* = nullptr>
-    constexpr explicit Expected(U&& v) : Expected{std::in_place, std::forward<U>(v)} {}
-
-    template <typename U = T, std::enable_if_t<std::is_convertible_v<U&&, T>>* = nullptr,
-              detail::expected_enable_forward_value<T, E, U>* = nullptr>
-    constexpr Expected(U&& v) : Expected{std::in_place, std::forward<U>(v)} {}
-
-    template <typename U = T, typename G = T,
-              std::enable_if_t<std::is_nothrow_constructible_v<T, U&&>>* = nullptr,
-              std::enable_if_t<(
-                  !std::is_same_v<Expected<T, E>, std::remove_cvref_t<U>> &&
-                  !std::conjunction_v<std::is_scalar<T>, std::is_same<T, std::remove_cvref_t<U>>> &&
-                  std::is_constructible_v<T, U> && std::is_assignable_v<G&, U> &&
-                  std::is_nothrow_move_constructible_v<E>)>* = nullptr>
-    Expected& operator=(U&& v) {
-        if (has_value()) {
-            val() = std::forward<U>(v);
-        } else {
-            err().~Unexpected<E>();
-            new (valptr()) T{std::forward<U>(v)};
-            this->m_has_val = true;
-        }
-
-        return *this;
-    }
-
-    template <typename U = T, typename G = T,
-              std::enable_if_t<!std::is_nothrow_constructible_v<T, U&&>>* = nullptr,
-              std::enable_if_t<(
-                  !std::is_same_v<Expected<T, E>, std::remove_cvref_t<U>> &&
-                  !std::conjunction_v<std::is_scalar<T>, std::is_same<T, std::remove_cvref_t<U>>> &&
-                  std::is_constructible_v<T, U> && std::is_assignable_v<G&, U> &&
-                  std::is_nothrow_move_constructible_v<E>)>* = nullptr>
-    Expected& operator=(U&& v) {
-        if (has_value()) {
-            val() = std::forward<U>(v);
-        } else {
-            auto tmp = std::move(err());
-            err().~Unexpected<E>();
-            new (valptr()) T{std::forward<U>(v)};
-            this->m_has_val = true;
-        }
-
-        return *this;
-    }
-
-    template <typename G = E, std::enable_if_t<std::is_nothrow_copy_constructible_v<G> &&
-                                               std::is_assignable_v<G&, G>>* = nullptr>
-    Expected& operator=(const Unexpected<G>& rhs) {
-        if (!has_value()) {
-            err() = rhs;
-        } else {
-            this->destroy_val();
-            new (errptr()) Unexpected<E>{rhs};
-            this->m_has_val = false;
-        }
-
-        return *this;
-    }
-
-    template <typename G = E, std::enable_if_t<std::is_nothrow_move_constructible_v<G> &&
-                                               std::is_move_assignable_v<G>>* = nullptr>
-    Expected& operator=(Unexpected<G>&& rhs) noexcept {
-        if (!has_value()) {
-            err() = std::move(rhs);
-        } else {
-            this->destroy_val();
-            new (errptr()) Unexpected<E>{std::move(rhs)};
-            this->m_has_val = false;
-        }
-
-        return *this;
-    }
-
-    template <typename... Args,
-              std::enable_if_t<std::is_nothrow_constructible_v<T, Args&&...>>* = nullptr>
-    void emplace(Args&&... args) {
-        if (has_value()) {
-            val() = T{std::forward<Args>(args)...};
-        } else {
-            err().~Unexpected<E>();
-            new (valptr()) T{std::forward<Args>(args)...};
-            this->m_has_val = true;
-        }
-    }
-
-    template <typename... Args,
-              std::enable_if_t<!std::is_nothrow_constructible_v<T, Args&&...>>* = nullptr>
-    void emplace(Args&&... args) {
-        if (has_value()) {
-            val() = T{std::forward<Args>(args)...};
-        } else {
-            auto tmp = std::move(err());
-            err().~Unexpected<E>();
-            new (valptr()) T{std::forward<Args>(args)...};
-            this->m_has_val = true;
-        }
-    }
-
-    template <typename U, typename... Args,
-              std::enable_if_t<std::is_nothrow_constructible_v<T, std::initializer_list<U>&,
-                                                               Args&&...>>* = nullptr>
-    void emplace(std::initializer_list<U> il, Args&&... args) {
-        if (has_value()) {
-            T t{il, std::forward<Args>(args)...};
-            val() = std::move(t);
-        } else {
-            err().~Unexpected<E>();
-            new (valptr()) T{il, std::forward<Args>(args)...};
-            this->m_has_val = true;
-        }
-    }
-
-    template <typename U, typename... Args,
-              std::enable_if_t<!std::is_nothrow_constructible_v<T, std::initializer_list<U>&,
-                                                                Args&&...>>* = nullptr>
-    void emplace(std::initializer_list<U> il, Args&&... args) {
-        if (has_value()) {
-            T t{il, std::forward<Args>(args)...};
-            val() = std::move(t);
-        } else {
-            auto tmp = std::move(err());
-            err().~Unexpected<E>();
-            new (valptr()) T{il, std::forward<Args>(args)...};
-            this->m_has_val = true;
-        }
-    }
-
-    constexpr T* operator->() {
-        return valptr();
-    }
-
-    constexpr const T* operator->() const {
-        return valptr();
-    }
-
-    template <typename U = T>
-    constexpr U& operator*() & {
-        return val();
-    }
-
-    template <typename U = T>
-    constexpr const U& operator*() const& {
-        return val();
-    }
-
-    template <typename U = T>
-    constexpr U&& operator*() && {
-        return std::move(val());
-    }
-
-    template <typename U = T>
-    constexpr const U&& operator*() const&& {
-        return std::move(val());
-    }
-
-    constexpr bool has_value() const noexcept {
-        return this->m_has_val;
-    }
-
-    constexpr explicit operator bool() const noexcept {
-        return this->m_has_val;
-    }
-
-    template <typename U = T>
-    constexpr U& value() & {
-        return val();
-    }
-
-    template <typename U = T>
-    constexpr const U& value() const& {
-        return val();
-    }
-
-    template <typename U = T>
-    constexpr U&& value() && {
-        return std::move(val());
-    }
-
-    template <typename U = T>
-    constexpr const U&& value() const&& {
-        return std::move(val());
-    }
-
-    constexpr E& error() & {
-        return err().value();
-    }
-
-    constexpr const E& error() const& {
-        return err().value();
-    }
-
-    constexpr E&& error() && {
-        return std::move(err().value());
-    }
-
-    constexpr const E&& error() const&& {
-        return std::move(err().value());
-    }
-
-    template <typename U>
-    constexpr T value_or(U&& v) const& {
-        static_assert(std::is_copy_constructible_v<T> && std::is_convertible_v<U&&, T>,
-                      "T must be copy-constructible and convertible from U&&");
-        return bool(*this) ? **this : static_cast<T>(std::forward<U>(v));
-    }
-
-    template <typename U>
-    constexpr T value_or(U&& v) && {
-        static_assert(std::is_move_constructible_v<T> && std::is_convertible_v<U&&, T>,
-                      "T must be move-constructible and convertible from U&&");
-        return bool(*this) ? std::move(**this) : static_cast<T>(std::forward<U>(v));
-    }
-
-private:
-    static_assert(!std::is_reference_v<T>, "T must not be a reference");
-    static_assert(!std::is_same_v<T, std::remove_cv_t<std::in_place_t>>,
-                  "T must not be std::in_place_t");
-    static_assert(!std::is_same_v<T, std::remove_cv_t<unexpect_t>>, "T must not be unexpect_t");
-    static_assert(!std::is_same_v<T, std::remove_cv_t<Unexpected<E>>>,
-                  "T must not be Unexpected<E>");
-    static_assert(!std::is_reference_v<E>, "E must not be a reference");
-
-    T* valptr() {
-        return std::addressof(this->m_val);
-    }
-
-    const T* valptr() const {
-        return std::addressof(this->m_val);
-    }
-
-    Unexpected<E>* errptr() {
-        return std::addressof(this->m_unexpect);
-    }
-
-    const Unexpected<E>* errptr() const {
-        return std::addressof(this->m_unexpect);
-    }
-
-    template <typename U = T>
-    constexpr U& val() {
-        return this->m_val;
-    }
-
-    template <typename U = T>
-    constexpr const U& val() const {
-        return this->m_val;
-    }
-
-    constexpr Unexpected<E>& err() {
-        return this->m_unexpect;
-    }
-
-    constexpr const Unexpected<E>& err() const {
-        return this->m_unexpect;
-    }
-
-    using impl_base = detail::expected_move_assign_base<T, E>;
-    using ctor_base = detail::expected_default_ctor_base<T, E>;
-};
-
-template <typename T, typename E, typename U, typename F>
-constexpr bool operator==(const Expected<T, E>& lhs, const Expected<U, F>& rhs) {
-    return (lhs.has_value() != rhs.has_value())
-               ? false
-               : (!lhs.has_value() ? lhs.error() == rhs.error() : *lhs == *rhs);
-}
-
-template <typename T, typename E, typename U, typename F>
-constexpr bool operator!=(const Expected<T, E>& lhs, const Expected<U, F>& rhs) {
-    return !operator==(lhs, rhs);
-}
-
-template <typename T, typename E, typename U>
-constexpr bool operator==(const Expected<T, E>& x, const U& v) {
-    return x.has_value() ? *x == v : false;
-}
-
-template <typename T, typename E, typename U>
-constexpr bool operator==(const U& v, const Expected<T, E>& x) {
-    return x.has_value() ? *x == v : false;
-}
-
-template <typename T, typename E, typename U>
-constexpr bool operator!=(const Expected<T, E>& x, const U& v) {
-    return !operator==(x, v);
-}
-
-template <typename T, typename E, typename U>
-constexpr bool operator!=(const U& v, const Expected<T, E>& x) {
-    return !operator==(v, x);
-}
-
-template <typename T, typename E>
-constexpr bool operator==(const Expected<T, E>& x, const Unexpected<E>& e) {
-    return x.has_value() ? false : x.error() == e.value();
-}
-
-template <typename T, typename E>
-constexpr bool operator==(const Unexpected<E>& e, const Expected<T, E>& x) {
-    return x.has_value() ? false : x.error() == e.value();
-}
-
-template <typename T, typename E>
-constexpr bool operator!=(const Expected<T, E>& x, const Unexpected<E>& e) {
-    return !operator==(x, e);
-}
-
-template <typename T, typename E>
-constexpr bool operator!=(const Unexpected<E>& e, const Expected<T, E>& x) {
-    return !operator==(e, x);
-}
-
-} // namespace Common

src/common/settings.cpp

@@ -54,7 +54,6 @@ SWITCHABLE(CpuBackend, true);
 SWITCHABLE(CpuAccuracy, true);
 SWITCHABLE(FullscreenMode, true);
 SWITCHABLE(GpuAccuracy, true);
-SWITCHABLE(GpuLogLevel, true);
 SWITCHABLE(Language, true);
 SWITCHABLE(MemoryLayout, true);
 SWITCHABLE(NvdecEmulation, false);
@@ -213,6 +212,16 @@ bool IsNceEnabled() {
     return is_nce_enabled;
 }
 
+static u64 current_program_id = 0;
+
+void SetCurrentProgramID(u64 program_id) {
+    current_program_id = program_id;
+}
+
+u64 GetCurrentProgramID() {
+    return current_program_id;
+}
+
 bool IsDockedMode() {
     return values.use_docked_mode.GetValue() == Settings::ConsoleMode::Docked;
 }

src/common/settings.h

@@ -573,6 +573,13 @@ struct Values {
         false,
 #endif
         "rescale_hack", Category::RendererHacks};
+    SwitchableSetting<bool> enable_gpu_buffer_readback{linkage,
+                                                       false,
+                                                       "enable_gpu_buffer_readback",
+                                                       Category::RendererAdvanced,
+                                                       Specialization::Default,
+                                                       true,
+                                                       true};
 
     SwitchableSetting<bool> use_asynchronous_shaders{linkage, false, "use_asynchronous_shaders",
                                                      Category::RendererHacks};
@@ -788,8 +795,8 @@ struct Values {
                                                 false};  // runtime_modifiable_ — startup-only
     Setting<bool> dump_exefs{linkage, false, "dump_exefs", Category::Debugging};
     Setting<bool> dump_nso{linkage, false, "dump_nso", Category::Debugging};
-    Setting<bool> dump_shaders{
-                               linkage, false, "dump_shaders", Category::DebuggingGraphics, Specialization::Default,
+    Setting<bool> dump_guest_shaders{
+                               linkage, false, "dump_guest_shaders", Category::DebuggingGraphics, Specialization::Default,
                                false};
     Setting<bool> dump_macros{
                               linkage, false, "dump_macros", Category::DebuggingGraphics, Specialization::Default, false};
@@ -813,9 +820,8 @@ struct Values {
     Setting<bool> disable_web_applet{linkage, true, "disable_web_applet", Category::Debugging};
 
     // GPU Logging
-    Setting<bool> gpu_logging_enabled{linkage, false, "gpu_logging_enabled", Category::Debugging};
-    SwitchableSetting<GpuLogLevel> gpu_log_level{linkage, GpuLogLevel::Standard, "gpu_log_level",
-                                                   Category::Debugging};
+    Setting<GpuLogLevel> gpu_log_level{linkage, GpuLogLevel::Off, "gpu_log_level",
+                                       Category::Debugging};
     Setting<bool> gpu_log_vulkan_calls{linkage, true, "gpu_log_vulkan_calls", Category::Debugging};
     Setting<bool> gpu_log_shader_dumps{linkage, false, "gpu_log_shader_dumps", Category::Debugging};
     Setting<bool> gpu_log_memory_tracking{linkage, true, "gpu_log_memory_tracking",
@@ -876,6 +882,9 @@ bool IsFastmemEnabled();
 void SetNceEnabled(bool is_64bit);
 bool IsNceEnabled();
 
+void SetCurrentProgramID(u64 program_id);
+u64 GetCurrentProgramID();
+
 bool IsOpenGL();
 
 bool IsDockedMode();

src/core/core.cpp

@@ -326,6 +326,9 @@ struct System::Impl {
 
         LOG_INFO(Core, "Loading {} ({:016X}) ...", name, params.program_id);
 
+        // Expose program id to dump sites and other global readers.
+        Settings::SetCurrentProgramID(params.program_id);
+
         // Track launch time for frontend launches
         LaunchTimestampCache::SaveLaunchTimestamp(params.program_id);
 
@@ -347,7 +350,7 @@ struct System::Impl {
 
         // Register with applet manager
         // All threads are started, begin main process execution, now that we're in the clear
-        applet_manager.CreateAndInsertByFrontendAppletParameters(std::make_unique<Service::Process>(*std::move(process)), params);
+        applet_manager.CreateAndInsertByFrontendAppletParameters(std::move(process), params);
 
         if (Settings::values.gamecard_inserted) {
             if (Settings::values.gamecard_current_game) {

src/core/hle/result.h

@@ -10,7 +10,6 @@
 #include "common/bit_field.h"
 #include "common/common_funcs.h"
 #include "common/common_types.h"
-#include "common/expected.h"
 
 // All the constants in this file come from <https://switchbrew.org/wiki/Error_codes>
 

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

@@ -729,7 +729,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(ResultSuccess);
-        rb.PushIpcInterface(ensure_token_id);
+        rb.PushIpcInterface(ctx, ensure_token_id);
     }
 
     void LoadIdTokenCacheDeprecated(HLERequestContext& ctx) {
@@ -921,7 +921,7 @@ void Module::Interface::GetProfile(HLERequestContext& ctx) {
 
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(ResultSuccess);
-    rb.PushIpcInterface<IProfile>(system, user_id, *profile_manager);
+    rb.PushIpcInterface<IProfile>(ctx, system, user_id, *profile_manager);
 }
 
 void Module::Interface::IsUserRegistrationRequestPermitted(HLERequestContext& ctx) {
@@ -993,7 +993,7 @@ void Module::Interface::GetBaasAccountManagerForApplication(HLERequestContext& c
     LOG_DEBUG(Service_ACC, "called");
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(ResultSuccess);
-    rb.PushIpcInterface<IManagerForApplication>(system, profile_manager);
+    rb.PushIpcInterface<IManagerForApplication>(ctx, system, profile_manager);
 }
 
 void Module::Interface::IsUserAccountSwitchLocked(HLERequestContext& ctx) {
@@ -1089,7 +1089,7 @@ void Module::Interface::GetProfileEditor(HLERequestContext& ctx) {
 
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(ResultSuccess);
-    rb.PushIpcInterface<IProfileEditor>(system, user_id, *profile_manager);
+    rb.PushIpcInterface<IProfileEditor>(ctx, system, user_id, *profile_manager);
 }
 
 void Module::Interface::GetBaasAccountAdministrator(HLERequestContext &ctx) {
@@ -1100,7 +1100,7 @@ void Module::Interface::GetBaasAccountAdministrator(HLERequestContext &ctx) {
 
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(ResultSuccess);
-    rb.PushIpcInterface<IAdministrator>(system, uuid);
+    rb.PushIpcInterface<IAdministrator>(ctx, system, uuid);
 }
 
 void Module::Interface::ListQualifiedUsers(HLERequestContext& ctx) {
@@ -1143,7 +1143,7 @@ void Module::Interface::GetBaasAccountManagerForSystemService(HLERequestContext&
 
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(ResultSuccess);
-    rb.PushIpcInterface<IManagerForSystemService>(system, uuid);
+    rb.PushIpcInterface<IManagerForSystemService>(ctx, system, uuid);
 }
 
 void Module::Interface::StoreSaveDataThumbnailSystem(HLERequestContext& ctx) {

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

@@ -35,7 +35,7 @@ void IAsyncContext::GetSystemEvent(HLERequestContext& ctx) {
 
     IPC::ResponseBuilder rb{ctx, 2, 1};
     rb.Push(ResultSuccess);
-    rb.PushCopyObjects(completion_event->GetReadableEvent());
+    rb.PushCopyObjects(ctx, completion_event->GetReadableEvent());
 }
 
 void IAsyncContext::Cancel(HLERequestContext& ctx) {

src/core/hle/service/am/applet.h

@@ -109,6 +109,10 @@ struct Applet {
     std::list<std::shared_ptr<Applet>> child_applets{};
     bool is_completed{};
 
+    std::shared_ptr<Applet> reserved_applet{};
+    bool unwind_after_reserved{};
+    bool is_winding{};
+
     // Self state
     bool exit_locked{};
     s32 fatal_section_count{};

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

@@ -25,6 +25,13 @@ void AppletStorageChannel::Push(Kernel::KernelCore& kernel, std::shared_ptr<ISto
     m_event.Signal(kernel);
 }
 
+void AppletStorageChannel::Unpop(Kernel::KernelCore& kernel, std::shared_ptr<IStorage> storage) {
+    std::scoped_lock lk{m_lock};
+
+    m_data.emplace_front(std::move(storage));
+    m_event.Signal(kernel);
+}
+
 Result AppletStorageChannel::Pop(Kernel::KernelCore& kernel, std::shared_ptr<IStorage>* out_storage) {
     std::scoped_lock lk{m_lock};
 

src/core/hle/service/am/applet_data_broker.h

@@ -26,6 +26,7 @@ public:
     ~AppletStorageChannel();
 
     void Push(Kernel::KernelCore& kernel, std::shared_ptr<IStorage> storage);
+    void Unpop(Kernel::KernelCore& kernel, std::shared_ptr<IStorage> storage);
     Result Pop(Kernel::KernelCore& kernel, std::shared_ptr<IStorage>* out_storage);
     Kernel::KReadableEvent* GetEvent();
 

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

@@ -267,7 +267,7 @@ void AppletManager::SetWindowSystem(WindowSystem* window_system) {
 
     if (Settings::values.enable_overlay && m_window_system->GetOverlayDisplayApplet() == nullptr) {
         if (auto overlay_process = CreateProcess(m_system, static_cast<u64>(AppletProgramId::OverlayDisplay), 0, 0)) {
-            auto overlay_applet = std::make_shared<Applet>(m_system, std::make_unique<Service::Process>(*std::move(overlay_process)), false);
+            auto overlay_applet = std::make_shared<Applet>(m_system, std::move(overlay_process), false);
             overlay_applet->program_id = static_cast<u64>(AppletProgramId::OverlayDisplay);
             overlay_applet->applet_id = AppletId::OverlayDisplay;
             overlay_applet->type = AppletType::OverlayApplet;

src/core/hle/service/am/frontend/applet_web_browser.cpp

@@ -4,6 +4,7 @@
 // SPDX-FileCopyrightText: Copyright 2020 yuzu Emulator Project
 // SPDX-License-Identifier: GPL-2.0-or-later
 
+#include "applet_web_browser_types.h"
 #include "common/assert.h"
 #include "common/fs/file.h"
 #include "common/fs/fs.h"
@@ -370,16 +371,13 @@ void WebBrowser::ExtractOfflineRomFS() {
 
 void WebBrowser::WebBrowserExit(WebExitReason exit_reason, std::string last_url) {
     const bool use_tlv_output =
-        (web_arg_header.shim_kind == ShimKind::Share &&
-         web_applet_version >= WebAppletVersion::Version196608) ||
-        (web_arg_header.shim_kind == ShimKind::Web &&
-         web_applet_version >= WebAppletVersion::Version524288) ||
-        (web_arg_header.shim_kind == ShimKind::Lhub);
+        (web_arg_header.shim_kind == ShimKind::Share && web_applet_version >= WebAppletVersion::Version196608)
+        || (web_arg_header.shim_kind == ShimKind::Web && web_applet_version >= WebAppletVersion::Version524288)
+        || (web_arg_header.shim_kind == ShimKind::Lhub);
 
     // https://switchbrew.org/wiki/Internet_Browser#TLVs
     if (use_tlv_output) {
-        LOG_DEBUG(Service_AM, "Using TLV output: exit_reason={}, last_url={}, last_url_size={}",
-                  exit_reason, last_url, last_url.size());
+        LOG_DEBUG(Service_AM, "Using TLV output: exit_reason={}, last_url={}, last_url_size={}", exit_reason, last_url, last_url.size());
 
         // storage size for TLVs is 0x2000 bytes (as per switchbrew documentation)
         constexpr size_t TLV_STORAGE_SIZE = 0x2000;
@@ -600,11 +598,14 @@ void WebBrowser::ExecuteShare() {
 
 void WebBrowser::ExecuteWeb() {
     LOG_INFO(Service_AM, "Opening external URL at {}", external_url);
-
-    frontend.OpenExternalWebPage(external_url,
-                                 [this](WebExitReason exit_reason, std::string last_url) {
-                                     WebBrowserExit(exit_reason, last_url);
-                                 });
+    frontend.OpenExternalWebPage(external_url, [this](WebExitReason exit_reason, std::string last_url) {
+        // Offline and web applets must be explicitly exited from because they respect exit state
+        // Unlike the other web stuffs
+        if (exit_reason == WebExitReason::ExitRequested || exit_reason == WebExitReason::EndButtonPressed)
+            exit_reason = (web_arg_header.shim_kind == ShimKind::Web || web_arg_header.shim_kind == ShimKind::Offline)
+                ? WebExitReason::ExitRequested : WebExitReason::EndButtonPressed;
+        WebBrowserExit(exit_reason, last_url);
+    });
 }
 
 void WebBrowser::ExecuteWifi() {

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

@@ -1,3 +1,6 @@
+// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
+// SPDX-License-Identifier: GPL-3.0-or-later
+
 // SPDX-FileCopyrightText: Copyright 2024 yuzu Emulator Project
 // SPDX-License-Identifier: GPL-2.0-or-later
 
@@ -12,7 +15,10 @@ namespace Service::AM {
 
 HidRegistration::HidRegistration(Core::System& system, Process& process) : m_process(process) {
     m_hid_server = system.ServiceManager().GetService<HID::IHidServer>("hid", true);
+    this->RegisterCurrentProcess();
+}
 
+void HidRegistration::RegisterCurrentProcess() {
     if (m_process.IsInitialized()) {
         m_hid_server->GetResourceManager()->RegisterAppletResourceUserId(m_process.GetProcessId(),
                                                                          true);

src/core/hle/service/am/hid_registration.h

@@ -1,3 +1,6 @@
+// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
+// SPDX-License-Identifier: GPL-3.0-or-later
+
 // SPDX-FileCopyrightText: Copyright 2024 yuzu Emulator Project
 // SPDX-License-Identifier: GPL-2.0-or-later
 
@@ -24,6 +27,7 @@ public:
     explicit HidRegistration(Core::System& system, Process& process);
     ~HidRegistration();
 
+    void RegisterCurrentProcess();
     void EnableAppletToGetInput(bool enable);
 
 private:

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

@@ -171,6 +171,18 @@ void LifecycleManager::SignalSystemEventIfNeeded(Kernel::KernelCore& kernel) {
     }
 }
 
+void LifecycleManager::ResetForRelaunch() {
+    m_unordered_messages.clear();
+
+    m_activity_state = ActivityState::BackgroundVisible;
+    m_requested_focus_state = FocusState{};
+    m_acknowledged_focus_state = FocusState{};
+    m_has_focus_state_changed = true;
+
+    m_suspend_mode = SuspendMode::NoOverride;
+    m_forced_suspend = false;
+}
+
 bool LifecycleManager::PopMessage(Kernel::KernelCore& kernel, AppletMessage* out_message) {
     const auto message = this->PopMessageInOrderOfPriority();
     this->SignalSystemEventIfNeeded(kernel);

src/core/hle/service/am/lifecycle_manager.h

@@ -138,6 +138,8 @@ public:
     void PushUnorderedMessage(Kernel::KernelCore& kernel, AppletMessage message);
     bool PopMessage(Kernel::KernelCore& kernel, AppletMessage* out_message);
 
+    void ResetForRelaunch();
+
 private:
     FocusState GetFocusStateWhileForegroundObscured() const;
     FocusState GetFocusStateWhileBackground(bool is_obscured) const;

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

@@ -40,23 +40,23 @@ namespace {
     }
 }
 
-[[nodiscard]] inline std::optional<Process> CreateProcessImpl(std::unique_ptr<Loader::AppLoader>& out_loader, Loader::ResultStatus& out_load_result, Core::System& system, FileSys::VirtualFile file, u64 program_id, u64 program_index) {
+[[nodiscard]] inline std::unique_ptr<Process> CreateProcessImpl(std::unique_ptr<Loader::AppLoader>& out_loader, Loader::ResultStatus& out_load_result, Core::System& system, FileSys::VirtualFile file, u64 program_id, u64 program_index) {
     // Get the appropriate loader to parse this NCA.
     out_loader = Loader::GetLoader(system, file, program_id, program_index);
     // Ensure we have a loader which can parse the NCA.
     if (out_loader) {
         // Try to load the process.
-        auto process = std::make_optional<Process>(system);
+        auto process = std::make_unique<Process>(system);
         if (process->Initialize(*out_loader, out_load_result)) {
             return process;
         }
     }
-    return std::nullopt;
+    return nullptr;
 }
 
 } // Anonymous namespace
 
-std::optional<Process> CreateProcess(Core::System& system, u64 program_id, u8 minimum_key_generation, u8 maximum_key_generation) {
+std::unique_ptr<Process> CreateProcess(Core::System& system, u64 program_id, u8 minimum_key_generation, u8 maximum_key_generation) {
     // Attempt to load program NCA.
     FileSys::VirtualFile nca_raw{};
 
@@ -66,7 +66,7 @@ std::optional<Process> CreateProcess(Core::System& system, u64 program_id, u8 mi
 
     // Ensure we retrieved a program NCA.
     if (!nca_raw) {
-        return std::nullopt;
+        return nullptr;
     }
 
     // Ensure we have a suitable version.
@@ -76,7 +76,7 @@ std::optional<Process> CreateProcess(Core::System& system, u64 program_id, u8 mi
             (nca.GetKeyGeneration() < minimum_key_generation ||
              nca.GetKeyGeneration() > maximum_key_generation)) {
             LOG_WARNING(Service_LDR, "Skipping program {:016X} with generation {}", program_id, nca.GetKeyGeneration());
-            return std::nullopt;
+            return nullptr;
         }
     }
 
@@ -85,7 +85,7 @@ std::optional<Process> CreateProcess(Core::System& system, u64 program_id, u8 mi
     return CreateProcessImpl(loader, status, system, nca_raw, program_id, 0);
 }
 
-std::optional<Process> CreateApplicationProcess(std::vector<u8>& out_control, std::unique_ptr<Loader::AppLoader>& out_loader, Loader::ResultStatus& out_load_result, Core::System& system, FileSys::VirtualFile file, u64 program_id, u64 program_index) {
+std::unique_ptr<Process> CreateApplicationProcess(std::vector<u8>& out_control, std::unique_ptr<Loader::AppLoader>& out_loader, Loader::ResultStatus& out_load_result, Core::System& system, FileSys::VirtualFile file, u64 program_id, u64 program_index) {
     if (auto process = CreateProcessImpl(out_loader, out_load_result, system, file, program_id, program_index); process) {
         FileSys::NACP nacp;
         if (out_loader->ReadControlData(nacp) == Loader::ResultStatus::Success) {
@@ -110,7 +110,23 @@ std::optional<Process> CreateApplicationProcess(std::vector<u8>& out_control, st
         system.GetARPManager().Register(launch.title_id, launch, out_control);
         return process;
     }
-    return std::nullopt;
+    return nullptr;
+}
+
+bool ReinitializeProcess(Core::System& system, Process& process, u64 program_id) {
+    auto& storage = system.GetContentProviderUnion();
+    const auto nca_raw = storage.GetEntryRaw(program_id, FileSys::ContentRecordType::Program);
+    if (!nca_raw) {
+        return false;
+    }
+
+    auto loader = Loader::GetLoader(system, nca_raw, program_id, 0);
+    if (!loader) {
+        return false;
+    }
+
+    Loader::ResultStatus status{};
+    return process.Initialize(*loader, status);
 }
 
 } // namespace Service::AM

src/core/hle/service/am/process_creation.h

@@ -27,7 +27,9 @@ class Process;
 
 namespace Service::AM {
 
-std::optional<Process> CreateProcess(Core::System& system, u64 program_id, u8 minimum_key_generation, u8 maximum_key_generation);
-std::optional<Process> CreateApplicationProcess(std::vector<u8>& out_control, std::unique_ptr<Loader::AppLoader>& out_loader, Loader::ResultStatus& out_load_result, Core::System& system, FileSys::VirtualFile file, u64 program_id, u64 program_index);
+std::unique_ptr<Process> CreateProcess(Core::System& system, u64 program_id, u8 minimum_key_generation, u8 maximum_key_generation);
+std::unique_ptr<Process> CreateApplicationProcess(std::vector<u8>& out_control, std::unique_ptr<Loader::AppLoader>& out_loader, Loader::ResultStatus& out_load_result, Core::System& system, FileSys::VirtualFile file, u64 program_id, u64 program_index);
+
+bool ReinitializeProcess(Core::System& system, Process& process, u64 program_id);
 
 } // namespace Service::AM

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

@@ -35,9 +35,9 @@ Result CreateGuestApplication(SharedPointer<IApplicationAccessor>* out_applicati
     std::unique_ptr<Loader::AppLoader> loader;
     Loader::ResultStatus result;
     auto process = CreateApplicationProcess(control, loader, result, system, nca_raw, program_id, 0);
-    R_UNLESS(process != std::nullopt, ResultUnknown);
+    R_UNLESS(process != nullptr, ResultUnknown);
 
-    const auto applet = std::make_shared<Applet>(system, std::make_unique<Service::Process>(*std::move(process)), true);
+    const auto applet = std::make_shared<Applet>(system, std::move(process), true);
     applet->program_id = program_id;
     applet->applet_id = AppletId::Application;
     applet->type = AppletType::Application;
@@ -88,9 +88,9 @@ Result IApplicationCreator::CreateSystemApplication(
     std::vector<u8> control;
     std::unique_ptr<Loader::AppLoader> loader;
     auto process = CreateProcess(system, application_id, 1, 22);
-    R_UNLESS(process != std::nullopt, ResultUnknown);
+    R_UNLESS(process != nullptr, ResultUnknown);
 
-    const auto applet = std::make_shared<Applet>(system, std::make_unique<Service::Process>(*std::move(process)), true);
+    const auto applet = std::make_shared<Applet>(system, std::move(process), true);
     applet->program_id = application_id;
     applet->applet_id = AppletId::Starter;
     applet->type = AppletType::LibraryApplet;

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

@@ -75,7 +75,7 @@ ILibraryAppletAccessor::ILibraryAppletAccessor(Core::System& system_,
         {105, D<&ILibraryAppletAccessor::GetPopOutDataEvent>, "GetPopOutDataEvent"},
         {106, D<&ILibraryAppletAccessor::GetPopInteractiveOutDataEvent>, "GetPopInteractiveOutDataEvent"},
         {110, nullptr, "NeedsToExitProcess"},
-        {120, nullptr, "GetLibraryAppletInfo"},
+        {120, D<&ILibraryAppletAccessor::GetLibraryAppletInfo>, "GetLibraryAppletInfo"},
         {150, nullptr, "RequestForAppletToGetForeground"},
         {160, D<&ILibraryAppletAccessor::GetIndirectLayerConsumerHandle>, "GetIndirectLayerConsumerHandle"}, //2.0.0+
         {170, D<&ILibraryAppletAccessor::Unknown170>, "Unknown170"}, //22.0.0+
@@ -218,6 +218,16 @@ Result ILibraryAppletAccessor::GetIndirectLayerConsumerHandle(Out<u64> out_handl
     R_SUCCEED();
 }
 
+Result ILibraryAppletAccessor::GetLibraryAppletInfo(
+    Out<LibraryAppletInfo> out_library_applet_info) {
+    LOG_INFO(Service_AM, "called");
+    *out_library_applet_info = {
+        .applet_id = m_applet->applet_id,
+        .library_applet_mode = m_applet->library_applet_mode,
+    };
+    R_SUCCEED();
+}
+
 Result ILibraryAppletAccessor::Unknown170(OutCopyHandle<Kernel::KReadableEvent> out_event) {
     LOG_WARNING(Service_AM, "(STUBBED) called");
     *out_event = m_applet->unknown_event.GetHandle();

src/core/hle/service/am/service/library_applet_accessor.h

@@ -6,6 +6,7 @@
 
 #pragma once
 
+#include "core/hle/service/am/service/library_applet_self_accessor.h"
 #include "core/hle/service/cmif_types.h"
 #include "core/hle/service/service.h"
 
@@ -21,6 +22,10 @@ public:
                                     std::shared_ptr<Applet> applet);
     ~ILibraryAppletAccessor();
 
+    std::shared_ptr<Applet> GetApplet() const {
+        return m_applet;
+    }
+
 private:
     Result GetAppletStateChangedEvent(OutCopyHandle<Kernel::KReadableEvent> out_event);
     Result IsCompleted(Out<bool> out_is_completed);
@@ -37,6 +42,7 @@ private:
     Result GetPopOutDataEvent(OutCopyHandle<Kernel::KReadableEvent> out_event);
     Result GetPopInteractiveOutDataEvent(OutCopyHandle<Kernel::KReadableEvent> out_event);
     Result GetIndirectLayerConsumerHandle(Out<u64> out_handle);
+    Result GetLibraryAppletInfo(Out<LibraryAppletInfo> out_library_applet_info);
     Result Unknown170(OutCopyHandle<Kernel::KReadableEvent> out_event);
 
     void FrontendExecute();

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

@@ -123,7 +123,7 @@ std::shared_ptr<ILibraryAppletAccessor> CreateGuestApplet(Core::System& system,
 
     auto process = CreateProcess(system, program_id, Firmware1400, Firmware2200);
     if (process) {
-        const auto applet = std::make_shared<Applet>(system, std::make_unique<Service::Process>(*std::move(process)), false);
+        const auto applet = std::make_shared<Applet>(system, std::move(process), false);
         applet->program_id = program_id;
         applet->applet_id = applet_id;
         applet->type = AppletType::LibraryApplet;

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

@@ -233,8 +233,9 @@ Result ILibraryAppletSelfAccessor::ReportVisibleErrorWithErrorContext(
     R_SUCCEED();
 }
 
-Result ILibraryAppletSelfAccessor::UnpopInData() {
-    LOG_WARNING(Service_AM, "(STUBBED) called");
+Result ILibraryAppletSelfAccessor::UnpopInData(SharedPointer<IStorage> storage) {
+    LOG_INFO(Service_AM, "called");
+    m_broker->GetInData().Unpop(system.Kernel(), storage);
     R_SUCCEED();
 }
 

src/core/hle/service/am/service/library_applet_self_accessor.h

@@ -72,7 +72,7 @@ private:
     Result ReportVisibleError(ErrorCode error_code);
     Result ReportVisibleErrorWithErrorContext(
         ErrorCode error_code, InLargeData<ErrorContext, BufferAttr_HipcMapAlias> error_context);
-    Result UnpopInData();
+    Result UnpopInData(SharedPointer<IStorage> storage);
     Result GetMainAppletApplicationDesiredLanguage(Out<u64> out_desired_language);
     Result GetCurrentApplicationId(Out<u64> out_application_id);
     Result GetMainAppletAvailableUsers(Out<bool> out_can_select_any_user, Out<s32> out_users_count,

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

@@ -1,9 +1,11 @@
-// SPDX-FileCopyrightText: Copyright 2025 Eden Emulator Project
+// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
 // SPDX-License-Identifier: GPL-3.0-or-later
 
 // SPDX-FileCopyrightText: Copyright 2024 yuzu Emulator Project
 // SPDX-License-Identifier: GPL-2.0-or-later
 
+#include "core/core.h"
+#include "core/hle/service/am/applet.h"
 #include "core/hle/service/am/frontend/applets.h"
 #include "core/hle/service/am/service/library_applet_accessor.h"
 #include "core/hle/service/am/service/process_winding_controller.h"
@@ -43,6 +45,18 @@ Result IProcessWindingController::OpenCallingLibraryApplet(
     Out<SharedPointer<ILibraryAppletAccessor>> out_calling_library_applet) {
     LOG_INFO(Service_AM, "called");
 
+    std::shared_ptr<Applet> reserved_applet;
+    {
+        std::scoped_lock lk{m_applet->lock};
+        reserved_applet = std::move(m_applet->reserved_applet);
+    }
+
+    if (reserved_applet != nullptr) {
+        *out_calling_library_applet = std::make_shared<ILibraryAppletAccessor>(
+            system, reserved_applet->caller_applet_broker, reserved_applet);
+        R_SUCCEED();
+    }
+
     const auto caller_applet = m_applet->caller_applet.lock();
     if (caller_applet == nullptr) {
         LOG_ERROR(Service_AM, "No caller applet available");
@@ -74,22 +88,82 @@ Result IProcessWindingController::PopContext(Out<SharedPointer<IStorage>> out_co
 }
 
 Result IProcessWindingController::CancelWindingReservation() {
-    LOG_WARNING(Service_AM, "STUBBED");
+    LOG_INFO(Service_AM, "called");
+
+    std::scoped_lock lk{m_applet->lock};
+    m_applet->reserved_applet.reset();
+    m_applet->unwind_after_reserved = false;
+
     R_SUCCEED();
 }
 
 Result IProcessWindingController::WindAndDoReserved() {
-    LOG_WARNING(Service_AM, "STUBBED");
+    LOG_INFO(Service_AM, "called");
+
+    std::shared_ptr<Applet> reserved_applet;
+    {
+        std::scoped_lock lk{m_applet->lock};
+
+        reserved_applet = m_applet->reserved_applet;
+        m_applet->display_layer_manager.SetWindowVisibility(false);
+        m_applet->exit_locked = false;
+        system.SetExitLocked(false);
+    }
+
+    if (reserved_applet) {
+        {
+            std::scoped_lock lk{m_applet->lock};
+            m_applet->is_winding = true;
+        }
+
+        {
+            std::scoped_lock lk{reserved_applet->lock};
+            reserved_applet->window_visible = true;
+            reserved_applet->process->Run();
+        }
+
+        if (reserved_applet->frontend) {
+            reserved_applet->frontend->Initialize();
+            reserved_applet->frontend->Execute();
+        }
+    } else {
+        LOG_WARNING(Service_AM, "called without a reserved applet to start");
+    }
+
+    m_applet->process->Terminate();
+
     R_SUCCEED();
 }
 
-Result IProcessWindingController::ReserveToStartAndWaitAndUnwindThis() {
-    LOG_WARNING(Service_AM, "STUBBED");
+Result IProcessWindingController::ReserveToStartAndWaitAndUnwindThis(
+    SharedPointer<ILibraryAppletAccessor> reserved_applet_accessor) {
+    LOG_INFO(Service_AM, "called");
+
+    if (reserved_applet_accessor == nullptr) {
+        LOG_ERROR(Service_AM, "No applet accessor provided");
+        R_THROW(ResultUnknown);
+    }
+
+    std::scoped_lock lk{m_applet->lock};
+    m_applet->reserved_applet = reserved_applet_accessor->GetApplet();
+    m_applet->unwind_after_reserved = true;
+
     R_SUCCEED();
 }
 
-Result IProcessWindingController::ReserveToStartAndWait() {
-    LOG_WARNING(Service_AM, "STUBBED");
+Result IProcessWindingController::ReserveToStartAndWait(
+    SharedPointer<ILibraryAppletAccessor> reserved_applet_accessor) {
+    LOG_INFO(Service_AM, "called");
+
+    if (reserved_applet_accessor == nullptr) {
+        LOG_ERROR(Service_AM, "No applet accessor provided");
+        R_THROW(ResultUnknown);
+    }
+
+    std::scoped_lock lk{m_applet->lock};
+    m_applet->reserved_applet = reserved_applet_accessor->GetApplet();
+    m_applet->unwind_after_reserved = false;
+
     R_SUCCEED();
 }
 

src/core/hle/service/am/service/process_winding_controller.h

@@ -1,4 +1,4 @@
-// SPDX-FileCopyrightText: Copyright 2025 Eden Emulator Project
+// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
 // SPDX-License-Identifier: GPL-3.0-or-later
 
 // SPDX-FileCopyrightText: Copyright 2024 yuzu Emulator Project
@@ -29,8 +29,9 @@ private:
     Result PopContext(Out<SharedPointer<IStorage>> out_context);
     Result CancelWindingReservation();
     Result WindAndDoReserved();
-    Result ReserveToStartAndWaitAndUnwindThis();
-    Result ReserveToStartAndWait();
+    Result ReserveToStartAndWaitAndUnwindThis(
+        SharedPointer<ILibraryAppletAccessor> reserved_applet_accessor);
+    Result ReserveToStartAndWait(SharedPointer<ILibraryAppletAccessor> reserved_applet_accessor);
 
     const std::shared_ptr<Applet> m_applet;
 };

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

@@ -9,6 +9,7 @@
 #include "core/hle/service/am/applet.h"
 #include "core/hle/service/am/applet_manager.h"
 #include "core/hle/service/am/event_observer.h"
+#include "core/hle/service/am/process_creation.h"
 #include "core/hle/service/am/window_system.h"
 
 namespace Service::AM {
@@ -240,6 +241,37 @@ void WindowSystem::PruneTerminatedAppletsLocked() {
             continue;
         }
 
+        // A winding applet has had its own process killed but is kept alive as a transparent slot
+        // while the reserved applet running in its place finishes (see WindAndDoReserved()).
+        if (applet->is_winding) {
+            if (!applet->child_applets.empty()) {
+                it = std::next(it);
+                continue;
+            }
+
+            const bool unwind = applet->unwind_after_reserved;
+            applet->is_winding = false;
+            applet->unwind_after_reserved = false;
+
+            if (unwind && this->RestartAppletProcessLocked(applet.get())) {
+                const u64 new_aruid = applet->aruid.pid;
+                const auto next = std::next(it);
+
+                if (new_aruid != aruid) {
+                    auto node = m_applets.extract(it);
+                    node.key() = new_aruid;
+                    m_applets.insert(std::move(node));
+                }
+
+                applet->process->Run();
+                m_event_observer->RequestUpdate();
+                it = next;
+                continue;
+            }
+
+            applet->reserved_applet.reset();
+        }
+
         // Terminated, so ensure all child applets are terminated.
         if (!applet->child_applets.empty()) {
             this->TerminateChildAppletsLocked(applet.get());
@@ -301,6 +333,28 @@ void WindowSystem::PruneTerminatedAppletsLocked() {
     }
 }
 
+bool WindowSystem::RestartAppletProcessLocked(Applet* applet) {
+    if (!ReinitializeProcess(m_system, *applet->process, applet->program_id)) {
+        LOG_ERROR(Service_AM, "Failed to restart winding applet_id={}",
+                  static_cast<u32>(applet->applet_id));
+        return false;
+    }
+
+    applet->aruid.pid = applet->process->GetProcessId();
+    applet->is_process_running = false;
+    applet->is_completed = false;
+
+    applet->hid_registration.RegisterCurrentProcess();
+
+    applet->lifecycle_manager.ResetForRelaunch();
+    applet->is_activity_runnable = false;
+
+    applet->launch_reason.flag = 1;
+
+    m_event_observer->TrackAppletProcess(*applet);
+    return true;
+}
+
 bool WindowSystem::LockHomeMenuIntoForegroundLocked() {
     // If the home menu is not locked into foreground, then there's nothing to do.
     if (m_home_menu == nullptr || !m_home_menu_foreground_locked) {
@@ -352,6 +406,9 @@ void WindowSystem::UpdateAppletStateLocked(Applet* applet, bool is_foreground, b
     const bool has_obscuring_child_applets = [&] {
         for (const auto& child_applet : applet->child_applets) {
             std::scoped_lock lk2{child_applet->lock};
+            if (child_applet->is_winding) {
+                return true;
+            }
             const auto mode = child_applet->library_applet_mode;
             if (child_applet->is_process_running && child_applet->window_visible &&
                 (mode == LibraryAppletMode::AllForeground ||

src/core/hle/service/am/window_system.h

@@ -61,6 +61,7 @@ public:
 
 private:
     void PruneTerminatedAppletsLocked();
+    bool RestartAppletProcessLocked(Applet* applet);
     bool LockHomeMenuIntoForegroundLocked();
     void TerminateChildAppletsLocked(Applet* applet);
     void UpdateAppletStateLocked(Applet* applet, bool is_foreground, bool overlay_blocking = false);

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

@@ -82,7 +82,7 @@ void APM::OpenSession(HLERequestContext& ctx) {
 
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(ResultSuccess);
-    rb.PushIpcInterface<ISession>(system, controller);
+    rb.PushIpcInterface<ISession>(ctx, system, controller);
 }
 
 void APM::GetPerformanceMode(HLERequestContext& ctx) {
@@ -125,7 +125,7 @@ void APM_Sys::GetPerformanceEvent(HLERequestContext& ctx) {
 
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(ResultSuccess);
-    rb.PushIpcInterface<ISession>(system, controller);
+    rb.PushIpcInterface<ISession>(ctx, system, controller);
 }
 
 void APM_Sys::SetCpuBoostMode(HLERequestContext& ctx) {

src/core/hle/service/fgm/fgm.cpp

@@ -1,3 +1,6 @@
+// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
+// SPDX-License-Identifier: GPL-3.0-or-later
+
 // SPDX-FileCopyrightText: Copyright 2018 yuzu Emulator Project
 // SPDX-License-Identifier: GPL-2.0-or-later
 
@@ -45,7 +48,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(ResultSuccess);
-        rb.PushIpcInterface<IRequest>(system);
+        rb.PushIpcInterface<IRequest>(ctx, system);
     }
 };
 

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

@@ -184,7 +184,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 2, 1};
         rb.Push(readable_event.Signal(system.Kernel()));
-        rb.PushCopyObjects(readable_event);
+        rb.PushCopyObjects(ctx, readable_event);
     }
 
     void Cancel(HLERequestContext& ctx) {
@@ -400,7 +400,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 2, 1};
         rb.Push(ResultSuccess);
-        rb.PushCopyObjects(notification_event->GetReadableEvent());
+        rb.PushCopyObjects(ctx, notification_event->GetReadableEvent());
     }
 
     void Clear(HLERequestContext& ctx) {
@@ -476,7 +476,7 @@ private:
 void Module::Interface::CreateFriendService(HLERequestContext& ctx) {
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(ResultSuccess);
-    rb.PushIpcInterface<IFriendService>(system);
+    rb.PushIpcInterface<IFriendService>(ctx, system);
     LOG_DEBUG(Service_Friend, "called");
 }
 
@@ -488,12 +488,12 @@ void Module::Interface::CreateNotificationService(HLERequestContext& ctx) {
 
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(ResultSuccess);
-    rb.PushIpcInterface<INotificationService>(system, uuid);
+    rb.PushIpcInterface<INotificationService>(ctx, system, uuid);
 }
 
-Module::Interface::Interface(std::shared_ptr<Module> module_, Core::System& system_,
-                             const char* name)
-    : ServiceFramework{system_, name}, module{std::move(module_)} {}
+Module::Interface::Interface(std::shared_ptr<Module> module_, Core::System& system_, const char* name)
+    : ServiceFramework{system_, name}, module{std::move(module_)}
+{}
 
 Module::Interface::~Interface() = default;
 

src/core/hle/service/glue/arp.cpp

@@ -279,7 +279,7 @@ void ARP_W::AcquireRegistrar(HLERequestContext& ctx) {
 
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(ResultSuccess);
-    rb.PushIpcInterface(registrar);
+    rb.PushIpcInterface(ctx, registrar);
 }
 
 void ARP_W::UnregisterApplicationInstance(HLERequestContext& ctx) {

src/core/hle/service/glue/bgtc.cpp

@@ -28,7 +28,7 @@ void BGTC_T::OpenTaskService(HLERequestContext& ctx) {
 
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(ResultSuccess);
-    rb.PushIpcInterface<ITaskService>(system);
+    rb.PushIpcInterface<ITaskService>(ctx, system);
 }
 
 ITaskService::ITaskService(Core::System& system_) : ServiceFramework{system_, "ITaskService"} {

src/core/hle/service/glue/ectx.cpp

@@ -1,3 +1,6 @@
+// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
+// SPDX-License-Identifier: GPL-3.0-or-later
+
 // SPDX-FileCopyrightText: Copyright 2021 yuzu Emulator Project
 // SPDX-License-Identifier: GPL-2.0-or-later
 
@@ -56,7 +59,7 @@ ECTX_AW::~ECTX_AW() = default;
 void ECTX_AW::CreateContextRegistrar(HLERequestContext& ctx) {
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(ResultSuccess);
-    rb.PushIpcInterface<IContextRegistrar>(std::make_shared<IContextRegistrar>(system));
+    rb.PushIpcInterface<IContextRegistrar>(ctx, system);
 }
 
 } // namespace Service::Glue

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

@@ -726,7 +726,7 @@ void IHidSystemServer::AcquireConnectionTriggerTimeoutEvent(HLERequestContext& c
 
     IPC::ResponseBuilder rb{ctx, 2, 1};
     rb.Push(ResultSuccess);
-    rb.PushCopyObjects(acquire_device_registered_event->GetReadableEvent());
+    rb.PushCopyObjects(ctx, acquire_device_registered_event->GetReadableEvent());
 }
 
 void IHidSystemServer::AcquireDeviceRegisteredEventForControllerSupport(HLERequestContext& ctx) {
@@ -734,7 +734,7 @@ void IHidSystemServer::AcquireDeviceRegisteredEventForControllerSupport(HLEReque
 
     IPC::ResponseBuilder rb{ctx, 2, 1};
     rb.Push(ResultSuccess);
-    rb.PushCopyObjects(acquire_device_registered_event->GetReadableEvent());
+    rb.PushCopyObjects(ctx, acquire_device_registered_event->GetReadableEvent());
 }
 
 void IHidSystemServer::GetRegisteredDevices(HLERequestContext& ctx) {
@@ -759,7 +759,7 @@ void IHidSystemServer::AcquireUniquePadConnectionEventHandle(HLERequestContext&
     LOG_WARNING(Service_HID, "(STUBBED) called");
 
     IPC::ResponseBuilder rb{ctx, 2, 1};
-    rb.PushCopyObjects(unique_pad_connection_event->GetReadableEvent());
+    rb.PushCopyObjects(ctx, unique_pad_connection_event->GetReadableEvent());
     rb.Push(ResultSuccess);
 }
 
@@ -776,7 +776,7 @@ void IHidSystemServer::AcquireJoyDetachOnBluetoothOffEventHandle(HLERequestConte
 
     IPC::ResponseBuilder rb{ctx, 2, 1};
     rb.Push(ResultSuccess);
-    rb.PushCopyObjects(joy_detach_event->GetReadableEvent());
+    rb.PushCopyObjects(ctx, joy_detach_event->GetReadableEvent());
 }
 
 void IHidSystemServer::IsUsbFullKeyControllerEnabled(HLERequestContext& ctx) {

src/core/hle/service/hle_ipc.h

@@ -31,7 +31,7 @@ class Memory;
 }
 
 namespace IPC {
-class ResponseBuilder;
+struct ResponseBuilder;
 }
 
 namespace Service {
@@ -392,7 +392,7 @@ public:
     }
 
 private:
-    friend class IPC::ResponseBuilder;
+    friend struct IPC::ResponseBuilder;
 
     void ParseCommandBuffer(u32_le* src_cmdbuf, bool incoming);
 

src/core/hle/service/ipc_helpers.h

@@ -24,45 +24,7 @@ namespace IPC {
 
 constexpr Result ResultSessionClosed{ErrorModule::HIPC, 301};
 
-class RequestHelperBase {
-protected:
-    Service::HLERequestContext* context = nullptr;
-    u32* cmdbuf;
-    u32 index = 0;
-
-public:
-    explicit RequestHelperBase(u32* command_buffer) : cmdbuf(command_buffer) {}
-
-    explicit RequestHelperBase(Service::HLERequestContext& ctx)
-        : context(&ctx), cmdbuf(ctx.CommandBuffer()) {}
-
-    void Skip(u32 size_in_words, bool set_to_null) {
-        if (set_to_null) {
-            memset(cmdbuf + index, 0, size_in_words * sizeof(u32));
-        }
-        index += size_in_words;
-    }
-
-    /**
-     * Aligns the current position forward to a 16-byte boundary, padding with zeros.
-     */
-    void AlignWithPadding() {
-        if (index & 3) {
-            Skip(static_cast<u32>(4 - (index & 3)), true);
-        }
-    }
-
-    u32 GetCurrentOffset() const {
-        return index;
-    }
-
-    void SetCurrentOffset(u32 offset) {
-        index = offset;
-    }
-};
-
-class ResponseBuilder : public RequestHelperBase {
-public:
+struct ResponseBuilder {
     /// Flags used for customizing the behavior of ResponseBuilder
     enum class Flags : u32 {
         None = 0,
@@ -71,14 +33,13 @@ public:
         AlwaysMoveHandles = 1,
     };
 
-    explicit ResponseBuilder(Service::HLERequestContext& ctx, u32 normal_params_size_,
-                             u32 num_handles_to_copy_ = 0, u32 num_objects_to_move_ = 0,
-                             Flags flags = Flags::None)
-        : RequestHelperBase(ctx), normal_params_size(normal_params_size_),
-          num_handles_to_copy(num_handles_to_copy_),
-          num_objects_to_move(num_objects_to_move_), kernel{ctx.kernel} {
-
-        memset(cmdbuf, 0, sizeof(u32) * IPC::COMMAND_BUFFER_LENGTH);
+    inline explicit ResponseBuilder(Service::HLERequestContext& ctx, u32 normal_params_size_, u32 num_handles_to_copy_ = 0, u32 num_objects_to_move_ = 0, Flags flags = Flags::None)
+        : cmdbuf(ctx.CommandBuffer())
+        , normal_params_size(normal_params_size_)
+        , num_handles_to_copy(num_handles_to_copy_)
+        , num_objects_to_move(num_objects_to_move_)
+    {
+        std::memset(cmdbuf, 0, sizeof(u32) * IPC::COMMAND_BUFFER_LENGTH);
 
         IPC::CommandHeader header{};
         auto const mgr = ctx.GetManager().get();
@@ -117,9 +78,7 @@ public:
             handle_descriptor_header.num_handles_to_copy.Assign(num_handles_to_copy_);
             handle_descriptor_header.num_handles_to_move.Assign(num_handles_to_move);
             PushRaw(handle_descriptor_header);
-
             ctx.handles_offset = index;
-
             Skip(num_handles_to_copy + num_handles_to_move, true);
         }
 
@@ -131,7 +90,6 @@ public:
                 domain_header.num_objects = num_domain_objects;
                 PushRaw(domain_header);
             }
-
             IPC::DataPayloadHeader data_payload_header{};
             data_payload_header.magic = Common::MakeMagic('S', 'F', 'C', 'O');
             PushRaw(data_payload_header);
@@ -141,34 +99,39 @@ public:
 
         ctx.data_payload_offset = index;
         ctx.write_size += index;
-        ctx.domain_offset = static_cast<u32>(index + raw_data_size / sizeof(u32));
+        ctx.domain_offset = u32(index + raw_data_size / sizeof(u32));
     }
 
-    template <class T>
-    void PushIpcInterface(std::shared_ptr<T> iface) {
-        auto manager{context->GetManager()};
+    inline void Skip(u32 size_in_words, bool set_to_null) {
+        if (set_to_null) std::memset(cmdbuf + index, 0, size_in_words * sizeof(u32));
+        index += size_in_words;
+    }
+    /// @brief Aligns the current position forward to a 16-byte boundary, padding with zeros.
+    inline void AlignWithPadding() { if (index & 3) Skip(u32(4 - (index & 3)), true); }
+    inline u32 GetCurrentOffset() const { return index; }
+    inline void SetCurrentOffset(u32 offset) { index = offset; }
 
+    template <class T> inline void PushIpcInterface(Service::HLERequestContext& ctx, std::shared_ptr<T> iface) {
+        auto manager = ctx.GetManager();
         if (manager->IsDomain()) {
-            context->AddDomainObject(std::move(iface));
+            ctx.AddDomainObject(std::move(iface));
         } else {
-            ASSERT(Kernel::GetCurrentProcess(kernel).GetResourceLimit()->Reserve(kernel, Kernel::LimitableResource::SessionCountMax, 1));
+            ASSERT(Kernel::GetCurrentProcess(ctx.kernel).GetResourceLimit()->Reserve(ctx.kernel, Kernel::LimitableResource::SessionCountMax, 1));
 
-            auto* session = Kernel::KSession::Create(kernel);
-            session->Initialize(kernel, nullptr, 0);
-            Kernel::KSession::Register(kernel, session);
+            auto* session = Kernel::KSession::Create(ctx.kernel);
+            session->Initialize(ctx.kernel, nullptr, 0);
+            Kernel::KSession::Register(ctx.kernel, session);
 
-            auto next_manager = std::make_shared<Service::SessionRequestManager>(
-                kernel, manager->GetServerManager());
+            auto next_manager = std::make_shared<Service::SessionRequestManager>(ctx.kernel, manager->GetServerManager());
             next_manager->SetSessionHandler(iface);
             manager->GetServerManager().RegisterSession(&session->GetServerSession(), next_manager);
 
-            context->AddMoveObject(&session->GetClientSession());
+            ctx.AddMoveObject(&session->GetClientSession());
         }
     }
 
-    template <class T, class... Args>
-    void PushIpcInterface(Args&&... args) {
-        PushIpcInterface<T>(std::make_shared<T>(std::forward<Args>(args)...));
+    template <class T, class... Args> inline void PushIpcInterface(Service::HLERequestContext& ctx, Args&&... args) {
+        PushIpcInterface<T>(ctx, std::make_shared<T>(std::forward<Args>(args)...));
     }
 
     void PushImpl(s8 value);
@@ -184,59 +147,38 @@ public:
     void PushImpl(bool value);
     void PushImpl(Result value);
 
-    template <typename T>
-    void Push(T value) {
+    template <typename T> inline void Push(T value) {
         return PushImpl(value);
     }
 
     template <typename First, typename... Other>
     void Push(const First& first_value, const Other&... other_values);
 
-    /**
-     * Helper function for pushing strongly-typed enumeration values.
-     *
-     * @tparam Enum The enumeration type to be pushed
-     *
-     * @param value The value to push.
-     *
-     * @note The underlying size of the enumeration type is the size of the
-     *       data that gets pushed. e.g. "enum class SomeEnum : u16" will
-     *       push a u16-sized amount of data.
-     */
-    template <typename Enum>
-    void PushEnum(Enum value) {
+    /// @brief Helper function for pushing strongly-typed enumeration values.
+    /// @tparam Enum The enumeration type to be pushed
+    /// @param value The value to push.
+    /// @note The underlying size of the enumeration type is the size of the data that gets pushed.
+    /// e.g. "enum class SomeEnum : u16" will push a u16-sized amount of data.
+    template <typename Enum> inline void PushEnum(Enum value) {
         static_assert(std::is_enum_v<Enum>, "T must be an enum type within a PushEnum call.");
-        static_assert(!std::is_convertible_v<Enum, int>,
-                      "enum type in PushEnum must be a strongly typed enum.");
+        static_assert(!std::is_convertible_v<Enum, int>, "enum type in PushEnum must be a strongly typed enum.");
         Push(static_cast<std::underlying_type_t<Enum>>(value));
     }
 
-    /**
-     * @brief Copies the content of the given trivially copyable class to the buffer as a normal
-     * param
-     * @note: The input class must be correctly packed/padded to fit hardware layout.
-     */
-    template <typename T>
-    void PushRaw(const T& value);
+    /// @brief Copies the content of the given trivially copyable class to the buffer as a normal param
+    /// @note: The input class must be correctly packed/padded to fit hardware layout.
+    template <typename T> void PushRaw(const T& value);
+    template <typename... O> void PushMoveObjects(Service::HLERequestContext& ctx, O*... pointers);
+    template <typename... O> void PushMoveObjects(Service::HLERequestContext& ctx, O&... pointers);
+    template <typename... O> void PushCopyObjects(Service::HLERequestContext& ctx, O*... pointers);
+    template <typename... O> void PushCopyObjects(Service::HLERequestContext& ctx, O&... pointers);
 
-    template <typename... O>
-    void PushMoveObjects(O*... pointers);
-
-    template <typename... O>
-    void PushMoveObjects(O&... pointers);
-
-    template <typename... O>
-    void PushCopyObjects(O*... pointers);
-
-    template <typename... O>
-    void PushCopyObjects(O&... pointers);
-
-private:
+    u32* cmdbuf;
+    u32 index = 0;
     u32 normal_params_size{};
     u32 num_handles_to_copy{};
     u32 num_objects_to_move{}; ///< Domain objects or move handles, context dependent
     u32 data_payload_index{};
-    Kernel::KernelCore& kernel;
 };
 
 /// Push ///
@@ -251,8 +193,7 @@ inline void ResponseBuilder::PushImpl(u32 value) {
 
 template <typename T>
 void ResponseBuilder::PushRaw(const T& value) {
-    static_assert(std::is_trivially_copyable_v<T>,
-                  "It's undefined behavior to use memcpy with non-trivially copyable objects");
+    static_assert(std::is_trivially_copyable_v<T>, "It's undefined behavior to use memcpy with non-trivially copyable objects");
     std::memcpy(cmdbuf + index, &value, sizeof(T));
     index += (sizeof(T) + 3) / 4; // round up to word length
 }
@@ -272,8 +213,8 @@ inline void ResponseBuilder::PushImpl(s16 value) {
 }
 
 inline void ResponseBuilder::PushImpl(s64 value) {
-    PushImpl(static_cast<u32>(value));
-    PushImpl(static_cast<u32>(value >> 32));
+    PushImpl(u32(value));
+    PushImpl(u32(value >> 32));
 }
 
 inline void ResponseBuilder::PushImpl(u8 value) {
@@ -285,8 +226,8 @@ inline void ResponseBuilder::PushImpl(u16 value) {
 }
 
 inline void ResponseBuilder::PushImpl(u64 value) {
-    PushImpl(static_cast<u32>(value));
-    PushImpl(static_cast<u32>(value >> 32));
+    PushImpl(u32(value));
+    PushImpl(u32(value >> 32));
 }
 
 inline void ResponseBuilder::PushImpl(float value) {
@@ -312,90 +253,88 @@ void ResponseBuilder::Push(const First& first_value, const Other&... other_value
 }
 
 template <typename... O>
-inline void ResponseBuilder::PushCopyObjects(O*... pointers) {
+inline void ResponseBuilder::PushCopyObjects(Service::HLERequestContext& ctx, O*... pointers) {
     auto objects = {pointers...};
     for (auto& object : objects) {
-        context->AddCopyObject(object);
+        ctx.AddCopyObject(object);
     }
 }
 
 template <typename... O>
-inline void ResponseBuilder::PushCopyObjects(O&... pointers) {
+inline void ResponseBuilder::PushCopyObjects(Service::HLERequestContext& ctx, O&... pointers) {
     auto objects = {&pointers...};
     for (auto& object : objects) {
-        context->AddCopyObject(object);
+        ctx.AddCopyObject(object);
     }
 }
 
 template <typename... O>
-inline void ResponseBuilder::PushMoveObjects(O*... pointers) {
+inline void ResponseBuilder::PushMoveObjects(Service::HLERequestContext& ctx, O*... pointers) {
     auto objects = {pointers...};
     for (auto& object : objects) {
-        context->AddMoveObject(object);
+        ctx.AddMoveObject(object);
     }
 }
 
 template <typename... O>
-inline void ResponseBuilder::PushMoveObjects(O&... pointers) {
+inline void ResponseBuilder::PushMoveObjects(Service::HLERequestContext& ctx, O&... pointers) {
     auto objects = {&pointers...};
     for (auto& object : objects) {
-        context->AddMoveObject(object);
+        ctx.AddMoveObject(object);
     }
 }
 
-class RequestParser : public RequestHelperBase {
-public:
-    explicit RequestParser(u32* command_buffer) : RequestHelperBase(command_buffer) {}
-
-    explicit RequestParser(Service::HLERequestContext& ctx) : RequestHelperBase(ctx) {
+struct RequestParser {
+    inline explicit RequestParser(u32* command_buffer) : cmdbuf(command_buffer) {}
+    inline explicit RequestParser(Service::HLERequestContext& ctx)
+        : cmdbuf(ctx.CommandBuffer())
+    {
         // TIPC does not have data payload offset
         if (!ctx.IsTipc()) {
             ASSERT_MSG(ctx.GetDataPayloadOffset(), "context is incomplete");
             Skip(ctx.GetDataPayloadOffset(), false);
         }
-
         // Skip the u64 command id, it's already stored in the context
         static constexpr u32 CommandIdSize = 2;
         Skip(CommandIdSize, false);
     }
 
-    template <typename T>
-    T Pop();
+    inline void Skip(u32 size_in_words, bool set_to_null) {
+        if (set_to_null) std::memset(cmdbuf + index, 0, size_in_words * sizeof(u32));
+        index += size_in_words;
+    }
+    /// @brief Aligns the current position forward to a 16-byte boundary, padding with zeros.
+    inline void AlignWithPadding() { if (index & 3) Skip(u32(4 - (index & 3)), true); }
+    inline u32 GetCurrentOffset() const { return index; }
+    inline void SetCurrentOffset(u32 offset) { index = offset; }
 
-    template <typename T>
-    void Pop(T& value);
-
-    template <typename First, typename... Other>
-    void Pop(First& first_value, Other&... other_values);
+    template <typename T> T Pop();
+    template <typename T> void Pop(T& value);
+    template <typename First, typename... Other> void Pop(First& first_value, Other&... other_values);
 
     template <typename T>
     T PopEnum() {
         static_assert(std::is_enum_v<T>, "T must be an enum type within a PopEnum call.");
-        static_assert(!std::is_convertible_v<T, int>,
-                      "enum type in PopEnum must be a strongly typed enum.");
-        return static_cast<T>(Pop<std::underlying_type_t<T>>());
+        static_assert(!std::is_convertible_v<T, int>, "enum type in PopEnum must be a strongly typed enum.");
+        return T(Pop<std::underlying_type_t<T>>());
     }
 
-    /**
-     * @brief Reads the next normal parameters as a struct, by copying it
-     * @note: The output class must be correctly packed/padded to fit hardware layout.
-     */
-    template <typename T>
-    void PopRaw(T& value);
+    /// @brief Reads the next normal parameters as a struct, by copying it
+    /// @note: The output class must be correctly packed/padded to fit hardware layout.
+    template <typename T> void PopRaw(T& value);
 
-    /**
-     * @brief Reads the next normal parameters as a struct, by copying it into a new value
-     * @note: The output class must be correctly packed/padded to fit hardware layout.
-     */
-    template <typename T>
-    T PopRaw();
+    /// @brief Reads the next normal parameters as a struct, by copying it into a new value
+    /// @note: The output class must be correctly packed/padded to fit hardware layout.
+    template <typename T> T PopRaw();
 
-    template <class T>
-    std::weak_ptr<T> PopIpcInterface() {
-        ASSERT(context->GetManager()->IsDomain());
-        ASSERT(context->GetDomainMessageHeader().input_object_count > 0);
-        return context->GetDomainHandler<T>(Pop<u32>() - 1);
+    template <class T> [[nodiscard]] std::weak_ptr<T> PopIpcInterface(Service::HLERequestContext& ctx) {
+        ASSERT(ctx.GetManager()->IsDomain());
+        ASSERT(ctx.GetDomainMessageHeader().input_object_count > 0);
+        return ctx.GetDomainHandler<T>(Pop<u32>() - 1);
     }
+
+    u32* cmdbuf;
+    u32 index = 0;
 };
 
 /// Pop ///
@@ -407,7 +346,7 @@ inline u32 RequestParser::Pop() {
 
 template <>
 inline s32 RequestParser::Pop() {
-    return static_cast<s32>(Pop<u32>());
+    return s32(Pop<u32>());
 }
 
 // Ignore the -Wclass-memaccess warning on memcpy for non-trivially default constructible objects.
@@ -417,8 +356,7 @@ inline s32 RequestParser::Pop() {
 #endif
 template <typename T>
 void RequestParser::PopRaw(T& value) {
-    static_assert(std::is_trivially_copyable_v<T>,
-                  "It's undefined behavior to use memcpy with non-trivially copyable objects");
+    static_assert(std::is_trivially_copyable_v<T>, "It's undefined behavior to use memcpy with non-trivially copyable objects");
     std::memcpy(&value, cmdbuf + index, sizeof(T));
     index += (sizeof(T) + 3) / 4; // round up to word length
 }

src/core/hle/service/lm/lm.cpp

@@ -353,7 +353,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(ResultSuccess);
-        rb.PushIpcInterface<ILogger>(system);
+        rb.PushIpcInterface<ILogger>(ctx, system);
     }
 };
 

src/core/hle/service/nfc/nfc.cpp

@@ -151,7 +151,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(ResultSuccess);
-        rb.PushIpcInterface<IAm>(system);
+        rb.PushIpcInterface<IAm>(ctx, system);
     }
 };
 
@@ -173,7 +173,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(ResultSuccess);
-        rb.PushIpcInterface<MFIUser>(system);
+        rb.PushIpcInterface<MFIUser>(ctx, system);
     }
 };
 
@@ -195,7 +195,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(ResultSuccess);
-        rb.PushIpcInterface<IUser>(system);
+        rb.PushIpcInterface<IUser>(ctx, system);
     }
 };
 
@@ -217,7 +217,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(ResultSuccess);
-        rb.PushIpcInterface<ISystem>(system);
+        rb.PushIpcInterface<ISystem>(ctx, system);
     }
 };
 

src/core/hle/service/nfc/nfc_interface.cpp

@@ -143,7 +143,7 @@ void NfcInterface::AttachAvailabilityChangeEvent(HLERequestContext& ctx) {
 
     IPC::ResponseBuilder rb{ctx, 2, 1};
     rb.Push(ResultSuccess);
-    rb.PushCopyObjects(GetManager()->AttachAvailabilityChangeEvent());
+    rb.PushCopyObjects(ctx, GetManager()->AttachAvailabilityChangeEvent());
 }
 
 void NfcInterface::StartDetection(HLERequestContext& ctx) {
@@ -203,7 +203,7 @@ void NfcInterface::AttachActivateEvent(HLERequestContext& ctx) {
 
     IPC::ResponseBuilder rb{ctx, 2, 1};
     rb.Push(result);
-    rb.PushCopyObjects(out_event);
+    rb.PushCopyObjects(ctx, out_event);
 }
 
 void NfcInterface::AttachDeactivateEvent(HLERequestContext& ctx) {
@@ -217,7 +217,7 @@ void NfcInterface::AttachDeactivateEvent(HLERequestContext& ctx) {
 
     IPC::ResponseBuilder rb{ctx, 2, 1};
     rb.Push(result);
-    rb.PushCopyObjects(out_event);
+    rb.PushCopyObjects(ctx, out_event);
 }
 
 void NfcInterface::SetNfcEnabled(HLERequestContext& ctx) {

src/core/hle/service/nfp/nfp.cpp

@@ -157,7 +157,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(ResultSuccess);
-        rb.PushIpcInterface<IUser>(system);
+        rb.PushIpcInterface<IUser>(ctx, system);
     }
 };
 
@@ -179,7 +179,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(ResultSuccess);
-        rb.PushIpcInterface<ISystem>(system);
+        rb.PushIpcInterface<ISystem>(ctx, system);
     }
 };
 
@@ -201,7 +201,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(ResultSuccess);
-        rb.PushIpcInterface<IDebug>(system);
+        rb.PushIpcInterface<IDebug>(ctx, system);
     }
 };
 

src/core/hle/service/nifm/nifm.cpp

@@ -307,7 +307,7 @@ private:
     void GetSystemEventReadableHandle(HLERequestContext& ctx) {
         IPC::ResponseBuilder rb{ctx, 2, 2};
         rb.Push(ResultSuccess);
-        rb.PushCopyObjects(evt_scan_complete->GetReadableEvent(),
+        rb.PushCopyObjects(ctx, evt_scan_complete->GetReadableEvent(),
                            evt_processing->GetReadableEvent());
     }
 
@@ -452,7 +452,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 2, 2};
         rb.Push(ResultSuccess);
-        rb.PushCopyObjects(event1->GetReadableEvent(), event2->GetReadableEvent());
+        rb.PushCopyObjects(ctx, event1->GetReadableEvent(), event2->GetReadableEvent());
     }
 
     void Cancel(HLERequestContext& ctx) {
@@ -528,7 +528,7 @@ void IGeneralService::CreateScanRequest(HLERequestContext& ctx) {
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
 
     rb.Push(ResultSuccess);
-    rb.PushIpcInterface<IScanRequest>(system);
+    rb.PushIpcInterface<IScanRequest>(ctx, system);
 }
 
 void IGeneralService::CreateRequest(HLERequestContext& ctx) {
@@ -537,7 +537,7 @@ void IGeneralService::CreateRequest(HLERequestContext& ctx) {
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
 
     rb.Push(ResultSuccess);
-    rb.PushIpcInterface<IRequest>(system);
+    rb.PushIpcInterface<IRequest>(ctx, system);
 }
 
 void IGeneralService::GetCurrentNetworkProfile(HLERequestContext& ctx) {
@@ -716,7 +716,7 @@ void IGeneralService::GetNetworkProfile(HLERequestContext& ctx) {
 
     IPC::ResponseBuilder rb{ctx, 2};
     rb.Push(ResultSuccess);
-    rb.PushIpcInterface<INetworkProfile>(system);
+    rb.PushIpcInterface<INetworkProfile>(ctx, system);
 }
 
 void IGeneralService::SetNetworkProfile(HLERequestContext& ctx) {
@@ -869,7 +869,7 @@ void IGeneralService::CreateTemporaryNetworkProfile(HLERequestContext& ctx) {
     IPC::ResponseBuilder rb{ctx, 6, 0, 1};
 
     rb.Push(ResultSuccess);
-    rb.PushIpcInterface<INetworkProfile>(system);
+    rb.PushIpcInterface<INetworkProfile>(ctx, system);
     rb.PushRaw<u128>(uuid);
 }
 
@@ -1124,7 +1124,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(ResultSuccess);
-        rb.PushIpcInterface<IGeneralService>(system);
+        rb.PushIpcInterface<IGeneralService>(ctx, system);
     }
 
     void CreateGeneralService(HLERequestContext& ctx) {
@@ -1132,7 +1132,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(ResultSuccess);
-        rb.PushIpcInterface<IGeneralService>(system);
+        rb.PushIpcInterface<IGeneralService>(ctx, system);
     }
 };
 

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

@@ -53,7 +53,7 @@ private:
         LOG_WARNING(Service_NIM, "(STUBBED) called");
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(ResultSuccess);
-        rb.PushIpcInterface<IShopServiceAsync>(system);
+        rb.PushIpcInterface<IShopServiceAsync>(ctx, system);
     }
 };
 
@@ -75,7 +75,7 @@ private:
         LOG_WARNING(Service_NIM, "(STUBBED) called");
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(ResultSuccess);
-        rb.PushIpcInterface<IShopServiceAccessor>(system);
+        rb.PushIpcInterface<IShopServiceAccessor>(ctx, system);
     }
 };
 
@@ -336,7 +336,7 @@ private:
         LOG_DEBUG(Service_NIM, "(STUBBED) called");
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(ResultSuccess);
-        rb.PushIpcInterface<IShopServiceAccessServer>(system);
+        rb.PushIpcInterface<IShopServiceAccessServer>(ctx, system);
     }
 
     void IsLargeResourceAvailable(HLERequestContext& ctx) {
@@ -356,7 +356,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(ResultSuccess);
-        rb.PushIpcInterface<IShopServiceAccessServer>(system);
+        rb.PushIpcInterface<IShopServiceAccessServer>(ctx, system);
     }
 };
 
@@ -439,7 +439,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 2, 1};
         rb.Push(ResultSuccess);
-        rb.PushCopyObjects(finished_event->GetReadableEvent());
+        rb.PushCopyObjects(ctx, finished_event->GetReadableEvent());
     }
 
     void GetResult(HLERequestContext& ctx) {
@@ -500,7 +500,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(ResultSuccess);
-        rb.PushIpcInterface<IEnsureNetworkClockAvailabilityService>(system);
+        rb.PushIpcInterface<IEnsureNetworkClockAvailabilityService>(ctx, system);
     }
 
     // TODO(ogniK): Do we need these?

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

@@ -359,8 +359,8 @@ void IReadOnlyApplicationControlDataInterface::ListApplicationTitle(HLERequestCo
 
     IPC::ResponseBuilder rb{ctx, 2, 1, 1};
     rb.Push(ResultSuccess);
-    rb.PushCopyObjects(async_value->ReadableEvent());
-    rb.PushIpcInterface(std::move(async_value));
+    rb.PushCopyObjects(ctx, async_value->ReadableEvent());
+    rb.PushIpcInterface(ctx, std::move(async_value));
 }
 
 Result IReadOnlyApplicationControlDataInterface::GetApplicationControlData3(

src/core/hle/service/nvdrv/nvdrv_interface.cpp

@@ -199,7 +199,7 @@ void NVDRV::QueryEvent(HLERequestContext& ctx) {
         IPC::ResponseBuilder rb{ctx, 3, 1};
         rb.Push(ResultSuccess);
         auto& readable_event = event->GetReadableEvent();
-        rb.PushCopyObjects(readable_event);
+        rb.PushCopyObjects(ctx, readable_event);
         rb.PushEnum(NvResult::Success);
     } else {
         LOG_ERROR(Service_NVDRV, "Invalid event request!");

src/core/hle/service/os/process.h

@@ -6,6 +6,8 @@
 
 #pragma once
 
+#include <utility>
+
 #include "common/common_types.h"
 
 namespace Core {
@@ -28,6 +30,15 @@ public:
     inline explicit Process(Core::System& system) noexcept : m_system(system) {}
     inline ~Process() { this->Finalize(); }
 
+    Process(const Process&) = delete;
+    Process& operator=(const Process&) = delete;
+    Process& operator=(Process&&) = delete;
+    inline Process(Process&& other) noexcept
+        : m_system(other.m_system), m_process(std::exchange(other.m_process, nullptr)),
+          m_main_thread_stack_size(std::exchange(other.m_main_thread_stack_size, 0)),
+          m_main_thread_priority(std::exchange(other.m_main_thread_priority, 0)),
+          m_process_started(std::exchange(other.m_process_started, false)) {}
+
     bool Initialize(Loader::AppLoader& loader, Loader::ResultStatus& out_load_result);
     void Finalize();
 

src/core/hle/service/pcv/pcv.cpp

@@ -1,3 +1,6 @@
+// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
+// SPDX-License-Identifier: GPL-3.0-or-later
+
 // SPDX-FileCopyrightText: Copyright 2018 yuzu Emulator Project
 // SPDX-License-Identifier: GPL-2.0-or-later
 
@@ -124,7 +127,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(ResultSuccess);
-        rb.PushIpcInterface<IClkrstSession>(system, device_code);
+        rb.PushIpcInterface<IClkrstSession>(ctx, system, device_code);
     }
 };
 

src/core/hle/service/pm/pm.cpp

@@ -161,7 +161,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 10, 1};
         rb.Push(ResultSuccess);
-        rb.PushCopyObjects(*process);
+        rb.PushCopyObjects(ctx, *process);
         rb.PushRaw(program_location);
         rb.PushRaw(override_status);
     }

src/core/hle/service/psc/time/alarms.cpp

@@ -147,7 +147,7 @@ void IAlarmService::CreateWakeupAlarm(HLERequestContext& ctx) {
 
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(ResultSuccess);
-    rb.PushIpcInterface<ISteadyClockAlarm>(system, m_alarms, AlarmType::WakeupAlarm);
+    rb.PushIpcInterface<ISteadyClockAlarm>(ctx, system, m_alarms, AlarmType::WakeupAlarm);
 }
 
 void IAlarmService::CreateBackgroundTaskAlarm(HLERequestContext& ctx) {
@@ -155,7 +155,7 @@ void IAlarmService::CreateBackgroundTaskAlarm(HLERequestContext& ctx) {
 
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(ResultSuccess);
-    rb.PushIpcInterface<ISteadyClockAlarm>(system, m_alarms, AlarmType::BackgroundTaskAlarm);
+    rb.PushIpcInterface<ISteadyClockAlarm>(ctx, system, m_alarms, AlarmType::BackgroundTaskAlarm);
 }
 
 ISteadyClockAlarm::ISteadyClockAlarm(Core::System& system_, Alarms& alarms, AlarmType type)
@@ -179,7 +179,7 @@ void ISteadyClockAlarm::GetAlarmEvent(HLERequestContext& ctx) {
 
     IPC::ResponseBuilder rb{ctx, 2, 1};
     rb.Push(ResultSuccess);
-    rb.PushCopyObjects(m_alarm.GetEventHandle());
+    rb.PushCopyObjects(ctx, m_alarm.GetEventHandle());
 }
 
 void ISteadyClockAlarm::Enable(HLERequestContext& ctx) {

src/core/hle/service/ptm/psm.cpp

@@ -67,7 +67,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 2, 1};
         rb.Push(ResultSuccess);
-        rb.PushCopyObjects(state_change_event->GetReadableEvent());
+        rb.PushCopyObjects(ctx, state_change_event->GetReadableEvent());
     }
 
     void UnbindStateChangeEvent(HLERequestContext& ctx) {
@@ -186,7 +186,7 @@ void PSM::OpenSession(HLERequestContext& ctx) {
 
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(ResultSuccess);
-    rb.PushIpcInterface<IPsmSession>(system);
+    rb.PushIpcInterface<IPsmSession>(ctx, system);
 }
 
 void PSM::GetBatteryVoltageState(HLERequestContext& ctx) {

src/core/hle/service/ptm/ts.cpp

@@ -1,3 +1,6 @@
+// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
+// SPDX-License-Identifier: GPL-3.0-or-later
+
 // SPDX-FileCopyrightText: Copyright 2022 yuzu Emulator Project
 // SPDX-License-Identifier: GPL-3.0-or-later
 
@@ -82,7 +85,7 @@ void TS::OpenSession(HLERequestContext& ctx) {
 
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(ResultSuccess);
-    rb.PushIpcInterface<ISession>(system);
+    rb.PushIpcInterface<ISession>(ctx, system);
 }
 
 } // namespace Service::PTM

src/core/hle/service/sm/sm.cpp

@@ -140,7 +140,7 @@ void SM::GetServiceCmif(HLERequestContext& ctx) {
     if (result == ResultSuccess) {
         IPC::ResponseBuilder rb{ctx, 2, 0, 1, IPC::ResponseBuilder::Flags::AlwaysMoveHandles};
         rb.Push(result);
-        rb.PushMoveObjects(client_session);
+        rb.PushMoveObjects(ctx, client_session);
     } else {
         IPC::ResponseBuilder rb{ctx, 2};
         rb.Push(result);
@@ -157,7 +157,7 @@ void SM::GetServiceTipc(HLERequestContext& ctx) {
 
     IPC::ResponseBuilder rb{ctx, 2, 0, 1, IPC::ResponseBuilder::Flags::AlwaysMoveHandles};
     rb.Push(result);
-    rb.PushMoveObjects(result == ResultSuccess ? client_session : nullptr);
+    rb.PushMoveObjects(ctx, result == ResultSuccess ? client_session : nullptr);
 }
 
 static std::string PopServiceName(IPC::RequestParser& rp) {
@@ -230,8 +230,7 @@ void SM::RegisterServiceImpl(HLERequestContext& ctx, std::string name, u32 max_s
               max_session_count, is_light);
 
     Kernel::KServerPort* server_port{};
-    if (const auto result = service_manager.RegisterService(std::addressof(server_port), name,
-                                                            max_session_count, nullptr);
+    if (const auto result = service_manager.RegisterService(std::addressof(server_port), name, max_session_count, nullptr);
         result.IsError()) {
         LOG_ERROR(Service_SM, "failed to register service with error_code={:08X}", result.raw);
         IPC::ResponseBuilder rb{ctx, 2};
@@ -241,7 +240,7 @@ void SM::RegisterServiceImpl(HLERequestContext& ctx, std::string name, u32 max_s
 
     IPC::ResponseBuilder rb{ctx, 2, 0, 1, IPC::ResponseBuilder::Flags::AlwaysMoveHandles};
     rb.Push(ResultSuccess);
-    rb.PushMoveObjects(server_port);
+    rb.PushMoveObjects(ctx, server_port);
 }
 
 void SM::UnregisterService(HLERequestContext& ctx) {

src/core/hle/service/sm/sm_controller.cpp

@@ -60,7 +60,7 @@ void Controller::CloneCurrentObject(HLERequestContext& ctx) {
     // We succeeded.
     IPC::ResponseBuilder rb{ctx, 2, 0, 1, IPC::ResponseBuilder::Flags::AlwaysMoveHandles};
     rb.Push(ResultSuccess);
-    rb.PushMoveObjects(session->GetClientSession());
+    rb.PushMoveObjects(ctx, session->GetClientSession());
 }
 
 void Controller::CloneCurrentObjectEx(HLERequestContext& ctx) {

src/core/hle/service/sockets/bsd.cpp

@@ -24,9 +24,6 @@
 #include "network/network.h"
 #include <common/settings.h>
 
-using Common::Expected;
-using Common::Unexpected;
-
 namespace Service::Sockets {
 
 namespace {
@@ -464,13 +461,22 @@ void BSD::DuplicateSocket(HLERequestContext& ctx) {
     IPC::RequestParser rp{ctx};
     auto input = rp.PopRaw<InputParameters>();
 
-    Expected<s32, Errno> res = DuplicateSocketImpl(input.fd);
     IPC::ResponseBuilder rb{ctx, 4};
     rb.Push(ResultSuccess);
-    rb.PushRaw(OutputParameters{
-        .ret = res.value_or(0),
-        .bsd_errno = res ? Errno::SUCCESS : res.error(),
-    });
+
+    auto const res_v = DuplicateSocketImpl(input.fd);
+    if (auto* res = std::get_if<s32>(&res_v)) {
+        rb.PushRaw(OutputParameters{
+            .ret = *res,
+            .bsd_errno = Errno::SUCCESS,
+        });
+    } else {
+        auto* err = std::get_if<Errno>(&res_v);
+        rb.PushRaw(OutputParameters{
+            .ret = 0,
+            .bsd_errno = *err,
+        });
+    }
 }
 
 void BSD::EventFd(HLERequestContext& ctx) {
@@ -977,15 +983,15 @@ Errno BSD::CloseImpl(s32 fd) {
     return bsd_errno;
 }
 
-Expected<s32, Errno> BSD::DuplicateSocketImpl(s32 fd) {
+std::variant<s32, Errno> BSD::DuplicateSocketImpl(s32 fd) {
     if (!IsFileDescriptorValid(fd)) {
-        return Unexpected(Errno::BADF);
+        return Errno::BADF;
     }
 
     const s32 new_fd = FindFreeFileDescriptorHandle();
     if (new_fd < 0) {
         LOG_ERROR(Service, "No more file descriptors available");
-        return Unexpected(Errno::MFILE);
+        return Errno::MFILE;
     }
 
     file_descriptors[new_fd] = FileDescriptor{

src/core/hle/service/sockets/bsd.h

@@ -1,4 +1,4 @@
-// SPDX-FileCopyrightText: Copyright 2025 Eden Emulator Project
+// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
 // SPDX-License-Identifier: GPL-3.0-or-later
 // SPDX-FileCopyrightText: Copyright 2018 yuzu Emulator Project
 // SPDX-License-Identifier: GPL-2.0-or-later
@@ -8,9 +8,9 @@
 #include <memory>
 #include <span>
 #include <vector>
+#include <variant>
 
 #include "common/common_types.h"
-#include "common/expected.h"
 #include "common/socket_types.h"
 #include "core/hle/service/service.h"
 #include "core/hle/service/sockets/sockets.h"
@@ -35,7 +35,7 @@ public:
     // These methods are called from SSL; the first two are also called from
     // this class for the corresponding IPC methods.
     // On the real device, the SSL service makes IPC calls to this service.
-    Common::Expected<s32, Errno> DuplicateSocketImpl(s32 fd);
+    std::variant<s32, Errno> DuplicateSocketImpl(s32 fd);
     Errno CloseImpl(s32 fd);
     std::optional<std::shared_ptr<Network::SocketBase>> GetSocket(s32 fd);
 

src/core/hle/service/sockets/sfdnsres.cpp

@@ -88,11 +88,12 @@ static const constexpr std::array blockedDomains = {
     "sun.hac.lp1.d4c.nintendo.net",
     "phoenix-api.wbagora.com", //hogwarts legacy
     "battle.net",
-    "microsoft.com", //minecraft dungeons + other games
+    "microsoft.com", // Minecraft dungeons + other games
     "mojang.com",
     "xboxlive.com",
     "api.epicgames.dev", // marvel cosmic invasion +?
-    "minecraftservices.com"
+    "minecraftservices.com",
+    "508223012e5a5ff19f30a391b2bdadc0.my.2k.com", // Civilization 5
 };
 
 static bool IsBlockedHost(const std::string& host) {
@@ -207,16 +208,15 @@ static std::pair<u32, GetAddrInfoError> GetHostByNameRequestImpl(HLERequestConte
         return {0, GetAddrInfoError::AGAIN};
     }
 
-    auto res = Network::GetAddressInfo(host, /*service*/ std::nullopt);
-    if (!res.has_value()) {
-        return {0, Translate(res.error())};
+    auto res_v = Network::GetAddressInfo(host, /*service*/ std::nullopt);
+    if (auto* res = std::get_if<std::vector<Network::AddrInfo>>(&res_v)) {
+        const std::vector<u8> data = SerializeAddrInfoAsHostEnt(*res, host);
+        const u32 data_size = u32(data.size());
+        ctx.WriteBuffer(data, 0);
+        return {data_size, GetAddrInfoError::SUCCESS};
     }
-
-    const std::vector<u8> data = SerializeAddrInfoAsHostEnt(res.value(), host);
-    const u32 data_size = static_cast<u32>(data.size());
-    ctx.WriteBuffer(data, 0);
-
-    return {data_size, GetAddrInfoError::SUCCESS};
+    auto* err = std::get_if<Network::GetAddrInfoError>(&res_v);
+    return {0, Translate(*err)};
 }
 
 void SFDNSRES::GetHostByNameRequest(HLERequestContext& ctx) {
@@ -332,16 +332,15 @@ static std::pair<u32, GetAddrInfoError> GetAddrInfoRequestImpl(HLERequestContext
 
     // Serialized hints are also passed in a buffer, but are ignored for now.
 
-    auto res = Network::GetAddressInfo(host, service);
-    if (!res.has_value()) {
-        return {0, Translate(res.error())};
+    auto res_v = Network::GetAddressInfo(host, service);
+    if (auto* res = std::get_if<std::vector<Network::AddrInfo>>(&res_v)) {
+        const std::vector<u8> data = SerializeAddrInfo(*res, host);
+        const u32 data_size = u32(data.size());
+        ctx.WriteBuffer(data, 0);
+        return {data_size, GetAddrInfoError::SUCCESS};
     }
-
-    const std::vector<u8> data = SerializeAddrInfo(res.value(), host);
-    const u32 data_size = static_cast<u32>(data.size());
-    ctx.WriteBuffer(data, 0);
-
-    return {data_size, GetAddrInfoError::SUCCESS};
+    auto* err = std::get_if<Network::GetAddrInfoError>(&res_v);
+    return {0, Translate(*err)};
 }
 
 void SFDNSRES::GetAddrInfoRequest(HLERequestContext& ctx) {

src/core/hle/service/ssl/ssl.cpp

@@ -160,29 +160,26 @@ private:
         auto bsd = system.ServiceManager().GetService<Service::Sockets::BSD>("bsd:u");
         ASSERT_OR_EXECUTE(bsd, { return ResultInternalError; });
 
-        auto res = bsd->DuplicateSocketImpl(fd);
-        if (!res.has_value()) {
-            LOG_ERROR(Service_SSL, "Failed to duplicate socket with fd {}", fd);
-            return ResultInvalidSocket;
+        auto const res_v = bsd->DuplicateSocketImpl(fd);
+        if (auto *res = std::get_if<s32>(&res_v)) {
+            const s32 duplicated_fd = *res;
+            if (do_not_close_socket) {
+                *out_fd = duplicated_fd;
+            } else {
+                *out_fd = -1;
+                fd_to_close = duplicated_fd;
+            }
+            std::optional<std::shared_ptr<Network::SocketBase>> sock = bsd->GetSocket(duplicated_fd);
+            if (!sock.has_value()) {
+                LOG_ERROR(Service_SSL, "invalid socket fd {} after duplication", duplicated_fd);
+                return ResultInvalidSocket;
+            }
+            socket = std::move(*sock);
+            backend->SetSocket(socket);
+            return ResultSuccess;
         }
-
-        const s32 duplicated_fd = *res;
-
-        if (do_not_close_socket) {
-            *out_fd = duplicated_fd;
-        } else {
-            *out_fd = -1;
-            fd_to_close = duplicated_fd;
-        }
-
-        std::optional<std::shared_ptr<Network::SocketBase>> sock = bsd->GetSocket(duplicated_fd);
-        if (!sock.has_value()) {
-            LOG_ERROR(Service_SSL, "invalid socket fd {} after duplication", duplicated_fd);
-            return ResultInvalidSocket;
-        }
-        socket = std::move(*sock);
-        backend->SetSocket(socket);
-        return ResultSuccess;
+        LOG_ERROR(Service_SSL, "Failed to duplicate socket with fd {}", fd);
+        return ResultInvalidSocket;
     }
 
     Result SetHostNameImpl(const std::string& hostname) {
@@ -546,8 +543,7 @@ private:
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(res);
         if (res == ResultSuccess) {
-            rb.PushIpcInterface<ISslConnection>(system, ssl_version, shared_data,
-                                                std::move(backend));
+            rb.PushIpcInterface<ISslConnection>(ctx, system, ssl_version, shared_data, std::move(backend));
         }
     }
 
@@ -627,12 +623,11 @@ private:
         IPC::RequestParser rp{ctx};
         const auto parameters = rp.PopRaw<Parameters>();
 
-        LOG_WARNING(Service_SSL, "(STUBBED) called, api_version={}, pid_placeholder={}",
-                    parameters.ssl_version.api_version, parameters.pid_placeholder);
+        LOG_WARNING(Service_SSL, "(STUBBED) called, api_version={}, pid_placeholder={}", parameters.ssl_version.api_version, parameters.pid_placeholder);
 
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(ResultSuccess);
-        rb.PushIpcInterface<ISslContext>(system, parameters.ssl_version);
+        rb.PushIpcInterface<ISslContext>(ctx, system, parameters.ssl_version);
     }
 
     void SetInterfaceVersion(HLERequestContext& ctx) {

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

@@ -155,7 +155,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(ResultSuccess);
-        rb.PushIpcInterface<IPdSession>(system);
+        rb.PushIpcInterface<IPdSession>(ctx, system);
     }
 };
 
@@ -199,7 +199,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(ResultSuccess);
-        rb.PushIpcInterface<IPdCradleSession>(system);
+        rb.PushIpcInterface<IPdCradleSession>(ctx, system);
     }
 };
 

src/core/internal_network/network.cpp

@@ -28,7 +28,6 @@
 
 #include "common/assert.h"
 #include "common/common_types.h"
-#include "common/expected.h"
 #include "common/logging.h"
 #include "common/settings.h"
 #include "core/internal_network/network.h"
@@ -733,15 +732,14 @@ u32 IPv4AddressToInteger(IPv4Address ip_addr) {
            static_cast<u32>(ip_addr[2]) << 8 | static_cast<u32>(ip_addr[3]);
 }
 
-Common::Expected<std::vector<AddrInfo>, GetAddrInfoError> GetAddressInfo(
+std::variant<std::vector<AddrInfo>, GetAddrInfoError> GetAddressInfo(
     const std::string& host, const std::optional<std::string>& service) {
     addrinfo hints{};
     hints.ai_family = AF_INET; // Switch only supports IPv4.
     addrinfo* addrinfo;
-    s32 gai_err = getaddrinfo(host.c_str(), service.has_value() ? service->c_str() : nullptr,
-                              &hints, &addrinfo);
+    s32 gai_err = getaddrinfo(host.c_str(), service.has_value() ? service->c_str() : nullptr, &hints, &addrinfo);
     if (gai_err != 0) {
-        return Common::Unexpected(TranslateGetAddrInfoErrorFromNative(gai_err));
+        return TranslateGetAddrInfoErrorFromNative(gai_err);
     }
     std::vector<AddrInfo> ret;
     for (auto* current = addrinfo; current; current = current->ai_next) {

src/core/internal_network/network.h

@@ -9,6 +9,7 @@
 #include <array>
 #include <optional>
 #include <vector>
+#include <variant>
 
 #include "common/common_funcs.h"
 #include "common/common_types.h"
@@ -124,7 +125,6 @@ std::string IPv4AddressToString(IPv4Address ip_addr);
 u32 IPv4AddressToInteger(IPv4Address ip_addr);
 
 // named to avoid name collision with Windows macro
-Common::Expected<std::vector<AddrInfo>, GetAddrInfoError> GetAddressInfo(
-    const std::string& host, const std::optional<std::string>& service);
+std::variant<std::vector<AddrInfo>, GetAddrInfoError> GetAddressInfo(const std::string& host, const std::optional<std::string>& service);
 
 } // namespace Network

src/hid_core/resources/npad/npad.cpp

@@ -50,8 +50,8 @@ NPad::NPad(Core::HID::HIDCore& hid_core_, KernelHelpers::ServiceContext& service
             auto& controller = controller_data[aruid_index][i];
             controller.device = hid_core.GetEmulatedControllerByIndex(i);
             Core::HID::ControllerUpdateCallback engine_callback{
-                .on_change = [this, i](Core::HID::ControllerTriggerType type) {
-                    ControllerUpdate(hid_core.kernel, type, i);
+                .on_change = [this, i, kernel = &hid_core.kernel](Core::HID::ControllerTriggerType type) {
+                    ControllerUpdate(*kernel, type, i);
                 },
                 .is_npad_service = true,
             };

src/qt_common/config/shared_translation.cpp

@@ -229,6 +229,8 @@ std::unique_ptr<TranslationMap> InitializeTranslations(QObject* parent) {
     INSERT(Settings, dma_accuracy, tr("DMA Accuracy:"),
            tr("Controls the DMA precision accuracy. Safe precision fixes issues in some games but "
               "may degrade performance."));
+    INSERT(Settings, enable_gpu_buffer_readback, tr("Enable GPU buffer readback"),
+           tr("Preserves GPU-modified buffer data by reading it back before uploads.\nSome games require this to render certain effects properly.\nMay cause issues if the hardware cannot handle the additional workload."));
     INSERT(Settings, use_asynchronous_shaders, tr("Enable asynchronous shader compilation"),
            tr("May reduce shader stutter."));
     INSERT(Settings, fast_gpu_time, tr("Fast GPU Time"),

src/shader_recompiler/frontend/maxwell/translate_program.cpp

@@ -236,8 +236,11 @@ void LowerGeometryPassthrough(const IR::Program& program, const HostTranslateInf
 
 IR::Program TranslateProgram(ObjectPool<IR::Inst>& inst_pool, ObjectPool<IR::Block>& block_pool,
                              Environment& env, Flow::CFG& cfg, const HostTranslateInfo& host_info) {
+    HostTranslateInfo normalized_host_info{host_info};
+    normalized_host_info.ApplyDescriptorLimitPolicy();
+
     IR::Program program;
-    program.syntax_list = BuildASL(inst_pool, block_pool, env, cfg, host_info);
+    program.syntax_list = BuildASL(inst_pool, block_pool, env, cfg, normalized_host_info);
     program.blocks = GenerateBlocks(program.syntax_list);
     program.post_order_blocks = PostOrder(program.syntax_list.front());
     program.stage = env.ShaderStage();
@@ -260,9 +263,9 @@ IR::Program TranslateProgram(ObjectPool<IR::Inst>& inst_pool, ObjectPool<IR::Blo
                 program.info.passthrough.mask[i] = ((mask[i / 32] >> (i % 32)) & 1) == 0;
             }
 
-            if (!host_info.support_geometry_shader_passthrough) {
+            if (!normalized_host_info.support_geometry_shader_passthrough) {
                 program.output_vertices = GetOutputTopologyVertices(program.output_topology);
-                LowerGeometryPassthrough(program, host_info);
+                LowerGeometryPassthrough(program, normalized_host_info);
             }
         }
         break;
@@ -277,16 +280,16 @@ IR::Program TranslateProgram(ObjectPool<IR::Inst>& inst_pool, ObjectPool<IR::Blo
     RemoveUnreachableBlocks(program);
 
     // Replace instructions before the SSA rewrite
-    if (!host_info.support_float64) {
+    if (!normalized_host_info.support_float64) {
         Optimization::LowerFp64ToFp32(program);
     }
-    if (!host_info.support_float16) {
+    if (!normalized_host_info.support_float16) {
         Optimization::LowerFp16ToFp32(program);
     }
-    if (!host_info.support_int64) {
+    if (!normalized_host_info.support_int64) {
         Optimization::LowerInt64ToInt32(program);
     }
-    if (!host_info.support_conditional_barrier) {
+    if (!normalized_host_info.support_conditional_barrier) {
         Optimization::ConditionalBarrierPass(program);
     }
     Optimization::SsaRewritePass(program);
@@ -295,8 +298,8 @@ IR::Program TranslateProgram(ObjectPool<IR::Inst>& inst_pool, ObjectPool<IR::Blo
 
     Optimization::PositionPass(env, program);
 
-    Optimization::GlobalMemoryToStorageBufferPass(program, host_info);
-    Optimization::TexturePass(env, program, host_info);
+    Optimization::GlobalMemoryToStorageBufferPass(program, normalized_host_info);
+    Optimization::TexturePass(env, program, normalized_host_info);
 
     if (Settings::values.resolution_info.active || Settings::values.rescale_hack.GetValue()) {
         Optimization::RescalingPass(program);
@@ -306,7 +309,7 @@ IR::Program TranslateProgram(ObjectPool<IR::Inst>& inst_pool, ObjectPool<IR::Blo
         Optimization::VerificationPass(program);
     }
     Optimization::CollectShaderInfoPass(env, program);
-    Optimization::LayerPass(program, host_info);
+    Optimization::LayerPass(program, normalized_host_info);
     Optimization::VendorWorkaroundPass(program);
 
     CollectInterpolationInfo(env, program);

src/shader_recompiler/host_translate_info.h

@@ -6,6 +6,8 @@
 
 #pragma once
 
+#include "common/common_types.h"
+
 namespace Shader {
 
 // Try to keep entries here to a minimum
@@ -13,20 +15,52 @@ namespace Shader {
 
 /// Misc information about the host
 struct HostTranslateInfo {
+    static constexpr u32 DEFAULT_DESCRIPTOR_LIMIT = 1024;
+
+    u64 min_ssbo_alignment{};            ///< Minimum alignment supported by the device for SSBOs
+    u32 max_per_stage_descriptor_sampled_images{}; ///< maximum sampled descriptors per stage
+    u32 max_per_stage_resources{};                 ///< maximum resources per stage
+    u32 max_descriptor_set_samplers{};
+    u32 max_descriptor_set_uniform_buffers{};
+    u32 max_descriptor_set_uniform_buffers_dynamic{};
+    u32 max_descriptor_set_storage_buffers{};
+    u32 max_descriptor_set_storage_buffers_dynamic{};
+    u32 max_descriptor_set_sampled_images{};
+    u32 max_descriptor_set_storage_images{};
+    u32 max_descriptor_set_input_attachements{};
     bool support_float64{};      ///< True when the device supports 64-bit floats
     bool support_float16{};      ///< True when the device supports 16-bit floats
     bool support_int64{};        ///< True when the device supports 64-bit integers
     bool needs_demote_reorder{}; ///< True when the device needs DemoteToHelperInvocation reordered
     bool support_snorm_render_buffer{};  ///< True when the device supports SNORM render buffers
     bool support_viewport_index_layer{}; ///< True when the device supports gl_Layer in VS
-    u32 min_ssbo_alignment{};            ///< Minimum alignment supported by the device for SSBOs
-    u32 max_per_stage_descriptor_sampled_images{1024}; ///< maximum sampled descriptors per stage
-    u32 max_per_stage_resources{4096};                 ///< maximum resources per stage
-    u32 max_descriptor_set_sampled_images{1024};        ///< maximum sampled descriptors per set
     bool support_geometry_shader_passthrough{}; ///< True when the device supports geometry
                                                 ///< passthrough shaders
     bool support_conditional_barrier{}; ///< True when the device supports barriers in conditional
                                         ///< control flow
+
+    void ApplyDescriptorLimitPolicy() noexcept {
+        if (min_ssbo_alignment == 0) {
+            min_ssbo_alignment = 1;
+        }
+        ApplyDescriptorLimitFallback(max_per_stage_descriptor_sampled_images);
+        ApplyDescriptorLimitFallback(max_per_stage_resources);
+        ApplyDescriptorLimitFallback(max_descriptor_set_samplers);
+        ApplyDescriptorLimitFallback(max_descriptor_set_uniform_buffers);
+        ApplyDescriptorLimitFallback(max_descriptor_set_uniform_buffers_dynamic);
+        ApplyDescriptorLimitFallback(max_descriptor_set_storage_buffers);
+        ApplyDescriptorLimitFallback(max_descriptor_set_storage_buffers_dynamic);
+        ApplyDescriptorLimitFallback(max_descriptor_set_sampled_images);
+        ApplyDescriptorLimitFallback(max_descriptor_set_storage_images);
+        ApplyDescriptorLimitFallback(max_descriptor_set_input_attachements);
+    }
+
+private:
+    static void ApplyDescriptorLimitFallback(u32& limit) noexcept {
+        if (limit == 0) {
+            limit = DEFAULT_DESCRIPTOR_LIMIT;
+        }
+    }
 };
 
 } // namespace Shader

src/shader_recompiler/ir_opt/global_memory_to_storage_buffer_pass.cpp

@@ -1,3 +1,6 @@
+// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
+// SPDX-License-Identifier: GPL-3.0-or-later
+
 // SPDX-FileCopyrightText: Copyright 2021 yuzu Emulator Project
 // SPDX-License-Identifier: GPL-2.0-or-later
 
@@ -545,7 +548,7 @@ void GlobalMemoryToStorageBufferPass(IR::Program& program, const HostTranslateIn
         IR::Block* const block{storage_inst.block};
         IR::Inst* const inst{storage_inst.inst};
         const IR::U32 offset{
-            StorageOffset(*block, *inst, storage_buffer, host_info.min_ssbo_alignment)};
+            StorageOffset(*block, *inst, storage_buffer, u32(host_info.min_ssbo_alignment))};
         Replace(*block, *inst, index, offset);
     }
 }

src/shader_recompiler/ir_opt/texture_pass.cpp

@@ -32,71 +32,62 @@ struct TextureInst {
 using TextureInstVector = boost::container::small_vector<TextureInst, 24>;
 
 constexpr u32 DESCRIPTOR_SIZE = 8;
-constexpr u32 DESCRIPTOR_SIZE_SHIFT = static_cast<u32>(std::countr_zero(DESCRIPTOR_SIZE));
-constexpr u32 DYNAMIC_DESCRIPTOR_CBUF_BYTES = 16 * 1024;
-constexpr u32 MAX_DYNAMIC_DESCRIPTOR_COUNT = 1024;
+constexpr u32 DESCRIPTOR_SIZE_SHIFT = u32(std::countr_zero(DESCRIPTOR_SIZE));
+constexpr u32 DESCRIPTOR_MAX_COUNT = 1024;
 
 u32 DynamicDescriptorSizeShift(const IR::U32& dynamic_offset) {
-    const IR::Inst* const inst{dynamic_offset.InstRecursive()};
-    if (!inst || inst->GetOpcode() != IR::Opcode::ShiftLeftLogical32) {
+    const IR::Inst* const inst = dynamic_offset.InstRecursive();
+    if (!inst || inst->GetOpcode() != IR::Opcode::ShiftLeftLogical32)
         return DESCRIPTOR_SIZE_SHIFT;
-    }
-    const IR::Value shift{inst->Arg(1)};
-    if (!shift.IsImmediate()) {
+    const IR::Value shift = inst->Arg(1);
+    if (!shift.IsImmediate())
         return DESCRIPTOR_SIZE_SHIFT;
-    }
-    const u32 size_shift{shift.U32()};
-    return size_shift >= DESCRIPTOR_SIZE_SHIFT && size_shift < 31 ? size_shift
-                                                                  : DESCRIPTOR_SIZE_SHIFT;
+    const u32 size_shift = shift.U32();
+    return size_shift >= DESCRIPTOR_SIZE_SHIFT && size_shift < 31 ? size_shift : DESCRIPTOR_SIZE_SHIFT;
 }
 
-u32 DynamicDescriptorCount(u32 base_offset, u32 size_shift) {
-    if (size_shift >= 31 || base_offset >= DYNAMIC_DESCRIPTOR_CBUF_BYTES) {
+u32 DynamicDescriptorCount(u32 base_offset, u32 size_shift, u32 max_descriptors) {
+    auto const descriptor_limit = (std::max)(1U, max_descriptors);
+    auto const max_cbuf_bytes = 16 * descriptor_limit;
+    if (size_shift >= 31 || base_offset >= max_cbuf_bytes)
         return 1;
-    }
-    const u32 stride{1U << size_shift};
-    const u32 available{DYNAMIC_DESCRIPTOR_CBUF_BYTES - base_offset};
-    if (available < DESCRIPTOR_SIZE) {
+    auto const stride = 1U << size_shift;
+    auto const available = max_cbuf_bytes - base_offset;
+    if (available < DESCRIPTOR_SIZE)
         return 1;
-    }
-    const u32 available_count{1U + (available - DESCRIPTOR_SIZE) / stride};
-    return std::min(MAX_DYNAMIC_DESCRIPTOR_COUNT, available_count);
+    auto const available_count = 1U + (available - DESCRIPTOR_SIZE) / stride;
+    return std::min(descriptor_limit, available_count);
 }
 
 u32 SaturatingSub(u32 lhs, u32 rhs) {
     return lhs > rhs ? lhs - rhs : 0;
 }
 
-template <typename Descriptors>
-u32 StaticDescriptorCount(const Descriptors& descriptors) {
-    u32 count{};
-    for (const auto& desc : descriptors) {
-        if (desc.count <= 1) {
-            count += desc.count;
-        }
-    }
-    return count;
+template <typename T>
+[[nodiscard]] u32 StaticDescriptorCount(T const& descriptors) noexcept {
+    return std::accumulate(descriptors.cbegin(), descriptors.cend(), 0U, [](auto const& acc, auto const& e) {
+        return acc + (e.count <= 1 ? e.count : 0);
+    });
 }
 
-u32 DynamicSampledTextureCap(const Info& info, const HostTranslateInfo& host_info,
-                             u32 dynamic_arrays) {
-    if (dynamic_arrays == 0) {
-        return MAX_DYNAMIC_DESCRIPTOR_COUNT;
+u32 DynamicSampledTextureCap(const Info& info, const HostTranslateInfo& host_info, u32 dynamic_arrays) {
+    auto const sampled_limit = (std::max)(1U, std::min(host_info.max_per_stage_descriptor_sampled_images,
+                                                       host_info.max_descriptor_set_sampled_images));
+    auto const resource_limit = (std::max)(1U, host_info.max_per_stage_resources);
+    if (dynamic_arrays > 0) {
+        auto const sampled_static_count = StaticDescriptorCount(info.texture_buffer_descriptors) + StaticDescriptorCount(info.texture_descriptors);
+        auto const resource_static_count =
+            NumDescriptors(info.constant_buffer_descriptors)
+            + NumDescriptors(info.storage_buffers_descriptors)
+            + sampled_static_count + NumDescriptors(info.image_buffer_descriptors)
+            + NumDescriptors(info.image_descriptors);
+        auto const sampled_budget = SaturatingSub(sampled_limit, sampled_static_count);
+        auto const resource_budget = SaturatingSub(resource_limit, resource_static_count);
+        auto const sampled_cap = sampled_budget / dynamic_arrays;
+        auto const resource_cap = resource_budget / dynamic_arrays;
+        return (std::max)(1U, (std::min)(sampled_cap, resource_cap));
     }
-    const u32 sampled_static_count{StaticDescriptorCount(info.texture_buffer_descriptors) +
-                                   StaticDescriptorCount(info.texture_descriptors)};
-    const u32 resource_static_count{
-        NumDescriptors(info.constant_buffer_descriptors) +
-        NumDescriptors(info.storage_buffers_descriptors) + sampled_static_count +
-        NumDescriptors(info.image_buffer_descriptors) + NumDescriptors(info.image_descriptors)};
-    const u32 sampled_limit{std::min(host_info.max_per_stage_descriptor_sampled_images,
-                                     host_info.max_descriptor_set_sampled_images)};
-    const u32 sampled_budget{SaturatingSub(sampled_limit, sampled_static_count)};
-    const u32 resource_budget{SaturatingSub(host_info.max_per_stage_resources,
-                                            resource_static_count)};
-    const u32 sampled_cap{sampled_budget / dynamic_arrays};
-    const u32 resource_cap{resource_budget / dynamic_arrays};
-    return std::max(1U, std::min({MAX_DYNAMIC_DESCRIPTOR_COUNT, sampled_cap, resource_cap}));
+    return (std::min)({DESCRIPTOR_MAX_COUNT, sampled_limit, resource_limit});
 }
 
 IR::Opcode IndexedInstruction(const IR::Inst& inst) {
@@ -304,21 +295,23 @@ static inline bool IsTexturePixelFormatIntegerCached(Environment& env,
 }
 
 
-std::optional<ConstBufferAddr> Track(const IR::Value& value, Environment& env);
-static inline std::optional<ConstBufferAddr> TrackCached(const IR::Value& v, Environment& env) {
+std::optional<ConstBufferAddr> Track(const IR::Value& value, Environment& env, const HostTranslateInfo& host_info);
+static inline std::optional<ConstBufferAddr> TrackCached(const IR::Value& v, Environment& env, const HostTranslateInfo& host_info) {
     if (const IR::Inst* key = v.InstRecursive()) {
         if (auto it = env.track_cache.find(key); it != env.track_cache.end()) return it->second;
-        auto found = Track(v, env);
+        auto found = Track(v, env, host_info);
         if (found) env.track_cache.emplace(key, *found);
         return found;
     }
-    return Track(v, env);
+    return Track(v, env, host_info);
 }
 
-std::optional<ConstBufferAddr> TryGetConstBuffer(const IR::Inst* inst, Environment& env);
+std::optional<ConstBufferAddr> TryGetConstBuffer(const IR::Inst* inst, Environment& env, const HostTranslateInfo& host_info);
 
-std::optional<ConstBufferAddr> Track(const IR::Value& value, Environment& env) {
-    return IR::BreadthFirstSearch(value, [&env](const IR::Inst* inst) { return TryGetConstBuffer(inst, env); });
+std::optional<ConstBufferAddr> Track(const IR::Value& value, Environment& env, const HostTranslateInfo& host_info) {
+    return IR::BreadthFirstSearch(value, [&env, &host_info](const IR::Inst* inst) {
+        return TryGetConstBuffer(inst, env, host_info);
+    });
 }
 
 std::optional<u32> TryGetConstant(IR::Value& value, Environment& env) {
@@ -342,13 +335,13 @@ std::optional<u32> TryGetConstant(IR::Value& value, Environment& env) {
     return ReadCbufCached(env, index_number, offset_number);
 }
 
-std::optional<ConstBufferAddr> TryGetConstBuffer(const IR::Inst* inst, Environment& env) {
+std::optional<ConstBufferAddr> TryGetConstBuffer(const IR::Inst* inst, Environment& env, const HostTranslateInfo& host_info) {
     switch (inst->GetOpcode()) {
     default:
         return std::nullopt;
     case IR::Opcode::BitwiseOr32: {
-        std::optional lhs{TrackCached(inst->Arg(0), env)};
-        std::optional rhs{TrackCached(inst->Arg(1), env)};
+        std::optional lhs{TrackCached(inst->Arg(0), env, host_info)};
+        std::optional rhs{TrackCached(inst->Arg(1), env, host_info)};
         if (!lhs || !rhs) {
             return std::nullopt;
         }
@@ -378,12 +371,11 @@ std::optional<ConstBufferAddr> TryGetConstBuffer(const IR::Inst* inst, Environme
         if (!shift.IsImmediate()) {
             return std::nullopt;
         }
-        std::optional lhs{TrackCached(inst->Arg(0), env)};
+        std::optional lhs{TrackCached(inst->Arg(0), env, host_info)};
         if (lhs) {
             lhs->shift_left = shift.U32();
         }
         return lhs;
-        break;
     }
     case IR::Opcode::BitwiseAnd32: {
         IR::Value op1{inst->Arg(0)};
@@ -407,7 +399,7 @@ std::optional<ConstBufferAddr> TryGetConstBuffer(const IR::Inst* inst, Environme
                 return std::nullopt;
             } while (false);
         }
-        std::optional lhs{TrackCached(op1, env)};
+        std::optional lhs{TrackCached(op1, env, host_info)};
         if (lhs) {
             lhs->shift_left = static_cast<u32>(std::countr_zero(op2.U32()));
         }
@@ -453,7 +445,10 @@ std::optional<ConstBufferAddr> TryGetConstBuffer(const IR::Inst* inst, Environme
     } else {
         return std::nullopt;
     }
-    const u32 size_shift{DynamicDescriptorSizeShift(dynamic_offset)};
+    auto const size_shift = DynamicDescriptorSizeShift(dynamic_offset);
+    auto const sampled_limit = (std::max)(1U, (std::min)(host_info.max_per_stage_descriptor_sampled_images,
+                                                         host_info.max_descriptor_set_sampled_images));
+    auto const resource_limit = (std::max)(1U, host_info.max_per_stage_resources);
     return ConstBufferAddr{
         .index = index.U32(),
         .offset = base_offset,
@@ -462,15 +457,15 @@ std::optional<ConstBufferAddr> TryGetConstBuffer(const IR::Inst* inst, Environme
         .secondary_offset = 0,
         .secondary_shift_left = 0,
         .dynamic_offset = dynamic_offset,
-        .count = DynamicDescriptorCount(base_offset, size_shift),
+        .count = DynamicDescriptorCount(base_offset, size_shift, (std::min)({DESCRIPTOR_MAX_COUNT, sampled_limit, resource_limit})),
         .has_secondary = false,
     };
 }
 
-TextureInst MakeInst(Environment& env, IR::Block* block, IR::Inst& inst) {
+TextureInst MakeInst(Environment& env, IR::Block* block, IR::Inst& inst, const HostTranslateInfo& host_info) {
     ConstBufferAddr addr;
     if (IsBindless(inst)) {
-        const std::optional<ConstBufferAddr> track_addr{TrackCached(inst.Arg(0), env)};
+        const std::optional<ConstBufferAddr> track_addr{TrackCached(inst.Arg(0), env, host_info)};
 
         if (!track_addr) {
             throw NotImplementedException("Failed to track bindless texture constant buffer");
@@ -506,15 +501,15 @@ u32 GetTextureHandle(Environment& env, const ConstBufferAddr& cbuf) {
     return lhs_raw | rhs_raw;
 }
 
-    [[maybe_unused]]TextureType ReadTextureType(Environment& env, const ConstBufferAddr& cbuf) {
+[[maybe_unused]] TextureType ReadTextureType(Environment& env, const ConstBufferAddr& cbuf) {
     return env.ReadTextureType(GetTextureHandle(env, cbuf));
 }
 
-    [[maybe_unused]]TexturePixelFormat ReadTexturePixelFormat(Environment& env, const ConstBufferAddr& cbuf) {
+[[maybe_unused]] TexturePixelFormat ReadTexturePixelFormat(Environment& env, const ConstBufferAddr& cbuf) {
     return env.ReadTexturePixelFormat(GetTextureHandle(env, cbuf));
 }
 
-    [[maybe_unused]]bool IsTexturePixelFormatInteger(Environment& env, const ConstBufferAddr& cbuf) {
+[[maybe_unused]] bool IsTexturePixelFormatInteger(Environment& env, const ConstBufferAddr& cbuf) {
     return env.IsTexturePixelFormatInteger(GetTextureHandle(env, cbuf));
 }
 
@@ -675,7 +670,7 @@ void TexturePass(Environment& env, IR::Program& program, const HostTranslateInfo
             if (!IsTextureInstruction(inst)) {
                 continue;
             }
-            to_replace.push_back(MakeInst(env, block, inst));
+            to_replace.push_back(MakeInst(env, block, inst, host_info));
         }
     }
     // Sort instructions to visit textures by constant buffer index, then by offset
@@ -689,8 +684,7 @@ void TexturePass(Environment& env, IR::Program& program, const HostTranslateInfo
         program.info.texture_descriptors,
         program.info.image_descriptors,
     };
-    const u32 sampled_dynamic_cap{
-        DynamicSampledTextureCap(program.info, host_info, DynamicSampledTextureArrayCount(to_replace))};
+    const u32 sampled_dynamic_cap = DynamicSampledTextureCap(program.info, host_info, DynamicSampledTextureArrayCount(to_replace));
     for (TextureInst& texture_inst : to_replace) {
         // TODO: Handle arrays
         IR::Inst* const inst{texture_inst.inst};

src/shader_recompiler/profile.h

@@ -92,7 +92,6 @@ struct Profile {
     bool has_broken_robust{};
 
     u64 min_ssbo_alignment{};
-
     u32 max_user_clip_distances{};
 };
 

src/tests/common/host_memory.cpp

@@ -1,3 +1,6 @@
+// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
+// SPDX-License-Identifier: GPL-3.0-or-later
+
 // SPDX-FileCopyrightText: Copyright 2021 yuzu Emulator Project
 // SPDX-License-Identifier: GPL-2.0-or-later
 
@@ -15,12 +18,19 @@ static constexpr auto PERMS = Common::MemoryPermission::ReadWrite;
 static constexpr auto HEAP = false;
 
 TEST_CASE("HostMemory: Initialize and deinitialize", "[common]") {
-    { HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE); }
-    { HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE); }
+    {
+        HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
+        REQUIRE(mem.BackingBasePointer() != nullptr);
+    }
+    {
+        HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
+        REQUIRE(mem.BackingBasePointer() != nullptr);
+    }
 }
 
 TEST_CASE("HostMemory: Simple map", "[common]") {
     HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
+    REQUIRE(mem.BackingBasePointer() != nullptr);
     mem.Map(0x5000, 0x8000, 0x1000, PERMS, HEAP);
 
     volatile u8* const data = mem.VirtualBasePointer() + 0x5000;
@@ -30,6 +40,7 @@ TEST_CASE("HostMemory: Simple map", "[common]") {
 
 TEST_CASE("HostMemory: Simple mirror map", "[common]") {
     HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
+    REQUIRE(mem.BackingBasePointer() != nullptr);
     mem.Map(0x5000, 0x3000, 0x2000, PERMS, HEAP);
     mem.Map(0x8000, 0x4000, 0x1000, PERMS, HEAP);
 
@@ -41,6 +52,7 @@ TEST_CASE("HostMemory: Simple mirror map", "[common]") {
 
 TEST_CASE("HostMemory: Simple unmap", "[common]") {
     HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
+    REQUIRE(mem.BackingBasePointer() != nullptr);
     mem.Map(0x5000, 0x3000, 0x2000, PERMS, HEAP);
 
     volatile u8* const data = mem.VirtualBasePointer() + 0x5000;
@@ -52,6 +64,7 @@ TEST_CASE("HostMemory: Simple unmap", "[common]") {
 
 TEST_CASE("HostMemory: Simple unmap and remap", "[common]") {
     HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
+    REQUIRE(mem.BackingBasePointer() != nullptr);
     mem.Map(0x5000, 0x3000, 0x2000, PERMS, HEAP);
 
     volatile u8* const data = mem.VirtualBasePointer() + 0x5000;
@@ -69,6 +82,7 @@ TEST_CASE("HostMemory: Simple unmap and remap", "[common]") {
 
 TEST_CASE("HostMemory: Nieche allocation", "[common]") {
     HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
+    REQUIRE(mem.BackingBasePointer() != nullptr);
     mem.Map(0x0000, 0, 0x20000, PERMS, HEAP);
     mem.Unmap(0x0000, 0x4000, HEAP);
     mem.Map(0x1000, 0, 0x2000, PERMS, HEAP);
@@ -78,6 +92,7 @@ TEST_CASE("HostMemory: Nieche allocation", "[common]") {
 
 TEST_CASE("HostMemory: Full unmap", "[common]") {
     HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
+    REQUIRE(mem.BackingBasePointer() != nullptr);
     mem.Map(0x8000, 0, 0x4000, PERMS, HEAP);
     mem.Unmap(0x8000, 0x4000, HEAP);
     mem.Map(0x6000, 0, 0x16000, PERMS, HEAP);
@@ -85,6 +100,7 @@ TEST_CASE("HostMemory: Full unmap", "[common]") {
 
 TEST_CASE("HostMemory: Right out of bounds unmap", "[common]") {
     HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
+    REQUIRE(mem.BackingBasePointer() != nullptr);
     mem.Map(0x0000, 0, 0x4000, PERMS, HEAP);
     mem.Unmap(0x2000, 0x4000, HEAP);
     mem.Map(0x2000, 0x80000, 0x4000, PERMS, HEAP);
@@ -92,6 +108,7 @@ TEST_CASE("HostMemory: Right out of bounds unmap", "[common]") {
 
 TEST_CASE("HostMemory: Left out of bounds unmap", "[common]") {
     HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
+
     mem.Map(0x8000, 0, 0x4000, PERMS, HEAP);
     mem.Unmap(0x6000, 0x4000, HEAP);
     mem.Map(0x8000, 0, 0x2000, PERMS, HEAP);
@@ -99,6 +116,7 @@ TEST_CASE("HostMemory: Left out of bounds unmap", "[common]") {
 
 TEST_CASE("HostMemory: Multiple placeholder unmap", "[common]") {
     HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
+    REQUIRE(mem.BackingBasePointer() != nullptr);
     mem.Map(0x0000, 0, 0x4000, PERMS, HEAP);
     mem.Map(0x4000, 0, 0x1b000, PERMS, HEAP);
     mem.Unmap(0x3000, 0x1c000, HEAP);
@@ -107,6 +125,7 @@ TEST_CASE("HostMemory: Multiple placeholder unmap", "[common]") {
 
 TEST_CASE("HostMemory: Unmap between placeholders", "[common]") {
     HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
+    REQUIRE(mem.BackingBasePointer() != nullptr);
     mem.Map(0x0000, 0, 0x4000, PERMS, HEAP);
     mem.Map(0x4000, 0, 0x4000, PERMS, HEAP);
     mem.Unmap(0x2000, 0x4000, HEAP);
@@ -115,6 +134,7 @@ TEST_CASE("HostMemory: Unmap between placeholders", "[common]") {
 
 TEST_CASE("HostMemory: Unmap to origin", "[common]") {
     HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
+    REQUIRE(mem.BackingBasePointer() != nullptr);
     mem.Map(0x4000, 0, 0x4000, PERMS, HEAP);
     mem.Map(0x8000, 0, 0x4000, PERMS, HEAP);
     mem.Unmap(0x4000, 0x4000, HEAP);
@@ -124,6 +144,7 @@ TEST_CASE("HostMemory: Unmap to origin", "[common]") {
 
 TEST_CASE("HostMemory: Unmap to right", "[common]") {
     HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
+    REQUIRE(mem.BackingBasePointer() != nullptr);
     mem.Map(0x4000, 0, 0x4000, PERMS, HEAP);
     mem.Map(0x8000, 0, 0x4000, PERMS, HEAP);
     mem.Unmap(0x8000, 0x4000, HEAP);
@@ -132,6 +153,7 @@ TEST_CASE("HostMemory: Unmap to right", "[common]") {
 
 TEST_CASE("HostMemory: Partial right unmap check bindings", "[common]") {
     HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
+    REQUIRE(mem.BackingBasePointer() != nullptr);
     mem.Map(0x4000, 0x10000, 0x4000, PERMS, HEAP);
 
     volatile u8* const ptr = mem.VirtualBasePointer() + 0x4000;
@@ -144,6 +166,7 @@ TEST_CASE("HostMemory: Partial right unmap check bindings", "[common]") {
 
 TEST_CASE("HostMemory: Partial left unmap check bindings", "[common]") {
     HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
+    REQUIRE(mem.BackingBasePointer() != nullptr);
     mem.Map(0x4000, 0x10000, 0x4000, PERMS, HEAP);
 
     volatile u8* const ptr = mem.VirtualBasePointer() + 0x4000;
@@ -158,6 +181,7 @@ TEST_CASE("HostMemory: Partial left unmap check bindings", "[common]") {
 
 TEST_CASE("HostMemory: Partial middle unmap check bindings", "[common]") {
     HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
+    REQUIRE(mem.BackingBasePointer() != nullptr);
     mem.Map(0x4000, 0x10000, 0x4000, PERMS, HEAP);
 
     volatile u8* const ptr = mem.VirtualBasePointer() + 0x4000;
@@ -172,6 +196,7 @@ TEST_CASE("HostMemory: Partial middle unmap check bindings", "[common]") {
 
 TEST_CASE("HostMemory: Partial sparse middle unmap and check bindings", "[common]") {
     HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
+    REQUIRE(mem.BackingBasePointer() != nullptr);
     mem.Map(0x4000, 0x10000, 0x2000, PERMS, HEAP);
     mem.Map(0x6000, 0x20000, 0x2000, PERMS, HEAP);
 

src/video_core/buffer_cache/buffer_cache.h

@@ -1634,15 +1634,17 @@ bool BufferCache<P>::SynchronizeBuffer(Buffer& buffer, DAddr device_addr, u32 si
     if (total_size_bytes == 0) {
         return true;
     }
-    u64 min_offset = (std::numeric_limits<u64>::max)();
-    u64 max_offset = 0;
-    for (const auto& copy : upload_copies) {
-        min_offset = (std::min)(min_offset, copy.dst_offset);
-        max_offset = (std::max)(max_offset, copy.dst_offset + copy.size);
+    if (Settings::values.enable_gpu_buffer_readback.GetValue()) {
+        u64 min_offset = (std::numeric_limits<u64>::max)();
+        u64 max_offset = 0;
+        for (const auto& copy : upload_copies) {
+            min_offset = (std::min)(min_offset, copy.dst_offset);
+            max_offset = (std::max)(max_offset, copy.dst_offset + copy.size);
+        }
+        const DAddr sync_addr = buffer.CpuAddr() + min_offset;
+        const u64 sync_size = max_offset - min_offset;
+        DownloadBufferMemory(buffer, sync_addr, sync_size);
     }
-    const DAddr sync_addr = buffer.CpuAddr() + min_offset;
-    const u64 sync_size = max_offset - min_offset;
-    DownloadBufferMemory(buffer, sync_addr, sync_size);
     const std::span<BufferCopy> copies_span(upload_copies.data(), upload_copies.size());
     UploadMemory(buffer, total_size_bytes, largest_copy, copies_span);
     any_buffer_uploaded = true;

src/video_core/engines/draw_manager.cpp

@@ -4,6 +4,8 @@
 // SPDX-FileCopyrightText: Copyright 2022 yuzu Emulator Project
 // SPDX-License-Identifier: GPL-2.0-or-later
 
+#include <cstring>
+
 #include "common/settings.h"
 #include "video_core/dirty_flags.h"
 #include "video_core/engines/maxwell_3d.h"
@@ -130,6 +132,44 @@ void Maxwell3D::DrawManager::SetInlineIndexBuffer(Maxwell3D& maxwell3d, u32 inde
     draw_state.draw_mode = DrawMode::InlineIndex;
 }
 
+void Maxwell3D::DrawManager::SetInlineIndexBuffer(Maxwell3D& maxwell3d, u32 method,
+                                                  const u32* base_start, u32 amount) {
+    auto& index_buffer = draw_state.inline_index_draw_indexes;
+    switch (method) {
+    case MAXWELL3D_REG_INDEX(draw_inline_index): {
+        const auto* const bytes = reinterpret_cast<const u8*>(base_start);
+        index_buffer.insert(index_buffer.end(), bytes, bytes + size_t(amount) * sizeof(u32));
+        break;
+    }
+    case MAXWELL3D_REG_INDEX(inline_index_2x16.even): {
+        const size_t offset = index_buffer.size();
+        index_buffer.resize(offset + size_t(amount) * 2 * sizeof(u32));
+        u8* dst = index_buffer.data() + offset;
+        for (u32 i = 0; i < amount; ++i) {
+            const u32 word = base_start[i];
+            const u32 indexes[2]{word & 0xFFFF, word >> 16};
+            std::memcpy(dst, indexes, sizeof(indexes));
+            dst += sizeof(indexes);
+        }
+        break;
+    }
+    case MAXWELL3D_REG_INDEX(inline_index_4x8.index0): {
+        const size_t offset = index_buffer.size();
+        index_buffer.resize(offset + size_t(amount) * 4 * sizeof(u32));
+        u8* dst = index_buffer.data() + offset;
+        for (u32 i = 0; i < amount; ++i) {
+            const u32 word = base_start[i];
+            const u32 indexes[4]{word & 0xFF, (word >> 8) & 0xFF, (word >> 16) & 0xFF,
+                                 word >> 24};
+            std::memcpy(dst, indexes, sizeof(indexes));
+            dst += sizeof(indexes);
+        }
+        break;
+    }
+    }
+    draw_state.draw_mode = DrawMode::InlineIndex;
+}
+
 void Maxwell3D::DrawManager::DrawBegin(Maxwell3D& maxwell3d) {
     auto reset_instance_count = maxwell3d.regs.draw.instance_id == Maxwell3D::Regs::Draw::InstanceId::First;
     auto increment_instance_count = maxwell3d.regs.draw.instance_id == Maxwell3D::Regs::Draw::InstanceId::Subsequent;

src/video_core/engines/maxwell_3d.cpp

@@ -436,6 +436,14 @@ void Maxwell3D::CallMultiMethod(Core::System& system, u32 method, const u32* bas
         upload_state.ProcessData(base_start, amount);
         return;
     }
+    case MAXWELL3D_REG_INDEX(draw_inline_index):
+    case MAXWELL3D_REG_INDEX(inline_index_2x16.even):
+    case MAXWELL3D_REG_INDEX(inline_index_4x8.index0):
+        if (shadow_state.shadow_ram_control != Regs::ShadowRamControl::Replay) {
+            ProcessInlineIndexMultiData(method, base_start, amount);
+            break;
+        }
+        [[fallthrough]];
     default:
         for (u32 i = 0; i < amount; i++) {
             CallMethod(system, method, base_start[i], methods_pending - i <= 1);
@@ -622,6 +630,15 @@ void Maxwell3D::ProcessCBData(u32 value) {
     ProcessCBMultiData(&value, 1);
 }
 
+void Maxwell3D::ProcessInlineIndexMultiData(u32 method, const u32* start_base, u32 amount) {
+    if (amount == 0) {
+        return;
+    }
+    const u32 argument = ProcessShadowRam(method, start_base[amount - 1]);
+    ProcessDirtyRegisters(method, argument);
+    draw_manager.SetInlineIndexBuffer(*this, method, start_base, amount);
+}
+
 Texture::TICEntry Maxwell3D::GetTICEntry(u32 tic_index) const {
     const GPUVAddr tic_address_gpu{regs.tex_header.Address() +
                                    tic_index * sizeof(Texture::TICEntry)};

src/video_core/engines/maxwell_3d.h

@@ -3077,6 +3077,7 @@ public:
         void DrawArrayIndirect(Maxwell3D& maxwell3d, Maxwell3D::Regs::PrimitiveTopology topology);
         void DrawIndexedIndirect(Maxwell3D& maxwell3d, Maxwell3D::Regs::PrimitiveTopology topology, u32 index_first, u32 index_count);
         void SetInlineIndexBuffer(Maxwell3D& maxwell3d, u32 index);
+        void SetInlineIndexBuffer(Maxwell3D& maxwell3d, u32 method, const u32* base_start, u32 amount);
         void DrawBegin(Maxwell3D& maxwell3d);
         void DrawEnd(Maxwell3D& maxwell3d, u32 instance_count = 1, bool force_draw = false);
         void DrawIndexSmall(Maxwell3D& maxwell3d, u32 argument);
@@ -3193,6 +3194,8 @@ public:
     void ProcessCBData(u32 value);
     void ProcessCBMultiData(const u32* start_base, u32 amount);
 
+    void ProcessInlineIndexMultiData(u32 method, const u32* start_base, u32 amount);
+
 private:
     void InitializeRegisterDefaults();
 

src/video_core/gpu_logging/gpu_logging.cpp

@@ -4,6 +4,7 @@
 #include "video_core/gpu_logging/gpu_logging.h"
 
 #include <fmt/format.h>
+#include <mutex>
 #include <thread>
 
 #include "common/fs/file.h"
@@ -280,13 +281,12 @@ void GPULogger::LogMemoryDeallocation(uintptr_t memory) {
 }
 
 void GPULogger::LogShaderCompilation(const std::string& shader_name,
-                                     const std::string& shader_info,
-                                     std::span<const u32> spirv_code) {
+                                     const std::string& shader_info) {
     if (!initialized || current_level == LogLevel::Off) {
         return;
     }
 
-    if (!dump_shaders && current_level < LogLevel::Verbose) {
+    if (current_level < LogLevel::Verbose) {
         return;
     }
 
@@ -294,38 +294,36 @@ void GPULogger::LogShaderCompilation(const std::string& shader_name,
         std::chrono::steady_clock::now().time_since_epoch());
 
     const auto log_entry = fmt::format("[{}] [Shader] Compiled: {} ({})\n",
-                                      FormatTimestamp(timestamp), shader_name, shader_info);
+                                       FormatTimestamp(timestamp), shader_name, shader_info);
     WriteToLog(log_entry);
+}
 
-    // Dump SPIR-V binary if enabled and we have data
-    if (dump_shaders && !spirv_code.empty()) {
-        using namespace Common::FS;
-        const auto& log_dir = GetEdenPath(EdenPath::LogDir);
-        const auto shaders_dir = log_dir / "shaders";
+bool IsActive() noexcept {
+    return Settings::values.gpu_log_level.GetValue() != Settings::GpuLogLevel::Off;
+}
 
-        // Create directory on first dump
-        if (!shader_dump_dir_created) {
-            [[maybe_unused]] const bool created = CreateDir(shaders_dir);
-            shader_dump_dir_created = true;
-        }
-
-        // Write SPIR-V binary file
-        const auto shader_path = shaders_dir / fmt::format("{}.spv", shader_name);
-        auto shader_file = std::make_unique<Common::FS::IOFile>(
-            shader_path, FileAccessMode::Write, FileType::BinaryFile);
-
-        if (shader_file->IsOpen()) {
-            const size_t bytes_to_write = spirv_code.size() * sizeof(u32);
-            static_cast<void>(shader_file->WriteSpan(spirv_code));
-            shader_file->Close();
-
-            const auto dump_log = fmt::format("[{}] [Shader] Dumped SPIR-V: {} ({} bytes)\n",
-                FormatTimestamp(timestamp), shader_path.string(), bytes_to_write);
-            WriteToLog(dump_log);
-        } else {
-            LOG_WARNING(Render_Vulkan, "[GPU Logging] Failed to dump shader: {}", shader_path.string());
-        }
+void DumpSpirvShader(u64 shader_hash, std::span<const u32> spirv_code) {
+    if (spirv_code.empty()) {
+        return;
     }
+
+    using namespace Common::FS;
+    const auto& dump_dir = GetEdenPath(EdenPath::DumpDir);
+
+    // Ensure DumpDir exists once. CreateDir is idempotent, so guarded to skip the syscall.
+    static std::once_flag dump_dir_flag;
+    std::call_once(dump_dir_flag, [&dump_dir]() {
+        [[maybe_unused]] const bool created = CreateDir(dump_dir);
+    });
+
+    const auto shader_path = dump_dir / fmt::format("{:016x}_{:016x}.spv",
+                                                    Settings::GetCurrentProgramID(), shader_hash);
+    Common::FS::IOFile shader_file(shader_path, FileAccessMode::Write, FileType::BinaryFile);
+    if (!shader_file.IsOpen()) {
+        LOG_WARNING(Render_Vulkan, "[Shader Dump] Failed to open {}", shader_path.string());
+        return;
+    }
+    static_cast<void>(shader_file.WriteSpan(spirv_code));
 }
 
 void GPULogger::LogPipelineStateChange(const std::string& state_info) {
@@ -657,10 +655,6 @@ void GPULogger::EnableVulkanCallTracking(bool enabled) {
     track_vulkan_calls = enabled;
 }
 
-void GPULogger::EnableShaderDumps(bool enabled) {
-    dump_shaders = enabled;
-}
-
 void GPULogger::EnableMemoryTracking(bool enabled) {
     track_memory = enabled;
 }

src/video_core/gpu_logging/gpu_logging.h

@@ -87,8 +87,7 @@ public:
     void LogVulkanCall(const std::string& call_name, const std::string& params, int result);
     void LogMemoryAllocation(uintptr_t memory, u64 size, u32 memory_flags);
     void LogMemoryDeallocation(uintptr_t memory);
-    void LogShaderCompilation(const std::string& shader_name, const std::string& shader_info,
-                              std::span<const u32> spirv_code = {});
+    void LogShaderCompilation(const std::string& shader_name, const std::string& shader_info);
     void LogPipelineStateChange(const std::string& state_info);
     void LogDriverDebugInfo(const std::string& debug_info);
 
@@ -121,7 +120,6 @@ public:
     // Settings
     void SetLogLevel(LogLevel level);
     void EnableVulkanCallTracking(bool enabled);
-    void EnableShaderDumps(bool enabled);
     void EnableMemoryTracking(bool enabled);
     void EnableDriverDebugInfo(bool enabled);
     void SetRingBufferSize(size_t entries);
@@ -171,7 +169,6 @@ private:
 
     // Feature flags
     bool track_vulkan_calls = true;
-    bool dump_shaders = false;
     bool track_memory = false;
     bool capture_driver_debug = false;
 
@@ -179,15 +176,16 @@ private:
     std::set<std::string> used_extensions;
     mutable std::mutex extension_mutex;
 
-    // Shader dump directory (created on demand)
-    bool shader_dump_dir_created = false;
-
     // Stored state for crash dumps
     std::string stored_driver_debug_info;
     std::string stored_pipeline_state;
     mutable std::mutex state_mutex;
 };
 
+[[nodiscard]] bool IsActive() noexcept;
+
+void DumpSpirvShader(u64 shader_hash, std::span<const u32> spirv_code);
+
 // Helper to get stage name from index
 inline const char* GetShaderStageName(size_t stage_index) {
     static constexpr std::array<const char*, 5> stage_names{

src/video_core/host_shaders/block_linear_unswizzle_2d.comp

@@ -11,9 +11,8 @@
 #define BEGIN_PUSH_CONSTANTS layout(push_constant) uniform PushConstants {
 #define END_PUSH_CONSTANTS };
 #define UNIFORM(n)
-#define BINDING_SWIZZLE_BUFFER 0
-#define BINDING_INPUT_BUFFER 1
-#define BINDING_OUTPUT_IMAGE 2
+#define BINDING_INPUT_BUFFER 0
+#define BINDING_OUTPUT_IMAGE 1
 
 #else // ^^^ Vulkan ^^^ // vvv OpenGL vvv
 
@@ -26,8 +25,7 @@
 #define BEGIN_PUSH_CONSTANTS
 #define END_PUSH_CONSTANTS
 #define UNIFORM(n) layout (location = n) uniform
-#define BINDING_SWIZZLE_BUFFER 0
-#define BINDING_INPUT_BUFFER 1
+#define BINDING_INPUT_BUFFER 0
 #define BINDING_OUTPUT_IMAGE 0
 
 #endif
@@ -43,10 +41,6 @@ UNIFORM(6) uint block_height;
 UNIFORM(7) uint block_height_mask;
 END_PUSH_CONSTANTS
 
-layout(binding = BINDING_SWIZZLE_BUFFER, std430) readonly buffer SwizzleTable {
-    uint swizzle_table[];
-};
-
 #if HAS_EXTENDED_TYPES
 layout(binding = BINDING_INPUT_BUFFER, std430) buffer InputBufferU8 { uint8_t u8data[]; };
 layout(binding = BINDING_INPUT_BUFFER, std430) buffer InputBufferU16 { uint16_t u16data[]; };
@@ -71,9 +65,19 @@ const uint GOB_SIZE_SHIFT = GOB_SIZE_X_SHIFT + GOB_SIZE_Y_SHIFT + GOB_SIZE_Z_SHI
 
 const uvec2 SWIZZLE_MASK = uvec2(GOB_SIZE_X - 1, GOB_SIZE_Y - 1);
 
+uint SwizzleTable(uint pos) {
+    const uint t[8] = uint[](
+        0x12100200, 0x13110301, 0x16140604, 0x17150705,
+        0x1a180a08, 0x1b190b09, 0x1e1c0e0c, 0x1f1d0f0d
+    );
+    const uint i = pos >> 4;
+    const uint h = (t[i / 4] >> ((i % 4) * 8)) & 0xff;
+    return (h << 4) | (pos & 0xf);
+}
+
 uint SwizzleOffset(uvec2 pos) {
     pos = pos & SWIZZLE_MASK;
-    return swizzle_table[pos.y * 64 + pos.x];
+    return SwizzleTable(pos.y * 64 + pos.x);
 }
 
 uvec4 ReadTexel(uint offset) {

src/video_core/host_shaders/block_linear_unswizzle_3d.comp

@@ -11,9 +11,8 @@
 #define BEGIN_PUSH_CONSTANTS layout(push_constant) uniform PushConstants {
 #define END_PUSH_CONSTANTS };
 #define UNIFORM(n)
-#define BINDING_SWIZZLE_BUFFER 0
-#define BINDING_INPUT_BUFFER 1
-#define BINDING_OUTPUT_IMAGE 2
+#define BINDING_INPUT_BUFFER 0
+#define BINDING_OUTPUT_IMAGE 1
 
 #else // ^^^ Vulkan ^^^ // vvv OpenGL vvv
 
@@ -26,8 +25,7 @@
 #define BEGIN_PUSH_CONSTANTS
 #define END_PUSH_CONSTANTS
 #define UNIFORM(n) layout (location = n) uniform
-#define BINDING_SWIZZLE_BUFFER 0
-#define BINDING_INPUT_BUFFER 1
+#define BINDING_INPUT_BUFFER 0
 #define BINDING_OUTPUT_IMAGE 0
 
 #endif
@@ -45,10 +43,6 @@ UNIFORM(8) uint block_depth;
 UNIFORM(9) uint block_depth_mask;
 END_PUSH_CONSTANTS
 
-layout(binding = BINDING_SWIZZLE_BUFFER, std430) readonly buffer SwizzleTable {
-    uint swizzle_table[];
-};
-
 #if HAS_EXTENDED_TYPES
 layout(binding = BINDING_INPUT_BUFFER, std430) buffer InputBufferU8 { uint8_t u8data[]; };
 layout(binding = BINDING_INPUT_BUFFER, std430) buffer InputBufferU16 { uint16_t u16data[]; };
@@ -73,9 +67,19 @@ const uint GOB_SIZE_SHIFT = GOB_SIZE_X_SHIFT + GOB_SIZE_Y_SHIFT + GOB_SIZE_Z_SHI
 
 const uvec2 SWIZZLE_MASK = uvec2(GOB_SIZE_X - 1, GOB_SIZE_Y - 1);
 
+uint SwizzleTable(uint pos) {
+    const uint t[8] = uint[](
+        0x12100200, 0x13110301, 0x16140604, 0x17150705,
+        0x1a180a08, 0x1b190b09, 0x1e1c0e0c, 0x1f1d0f0d
+    );
+    const uint i = pos >> 4;
+    const uint h = (t[i / 4] >> ((i % 4) * 8)) & 0xff;
+    return (h << 4) | (pos & 0xf);
+}
+
 uint SwizzleOffset(uvec2 pos) {
     pos = pos & SWIZZLE_MASK;
-    return swizzle_table[pos.y * 64 + pos.x];
+    return SwizzleTable(pos.y * 64 + pos.x);
 }
 
 uvec4 ReadTexel(uint offset) {

src/video_core/host_shaders/block_linear_unswizzle_3d_bcn.comp

@@ -10,9 +10,8 @@
     #define BEGIN_PUSH_CONSTANTS layout(push_constant) uniform PushConstants {
     #define END_PUSH_CONSTANTS };
     #define UNIFORM(n)
-    #define BINDING_SWIZZLE_BUFFER 0
-    #define BINDING_INPUT_BUFFER   1
-    #define BINDING_OUTPUT_BUFFER  2
+    #define BINDING_INPUT_BUFFER   0
+    #define BINDING_OUTPUT_BUFFER  1
 #else
     #extension GL_NV_gpu_shader5 : enable
     #ifdef GL_NV_gpu_shader5
@@ -23,7 +22,6 @@
     #define BEGIN_PUSH_CONSTANTS
     #define END_PUSH_CONSTANTS
     #define UNIFORM(n) layout(location = n) uniform
-    #define BINDING_SWIZZLE_BUFFER 0
     #define BINDING_INPUT_BUFFER   1
     #define BINDING_OUTPUT_BUFFER  0
 #endif
@@ -66,13 +64,9 @@ END_PUSH_CONSTANTS
 #endif
 
 // --- Buffers ---
-layout(binding = BINDING_SWIZZLE_BUFFER, std430) readonly buffer SwizzleTable {
-    uint swizzle_table[];
-};
-
 #if HAS_EXTENDED_TYPES
-    layout(binding = BINDING_INPUT_BUFFER, std430) buffer InputBufferU8   { uint8_t  u8data[];  };
-    layout(binding = BINDING_INPUT_BUFFER, std430) buffer InputBufferU16  { uint16_t u16data[]; };
+layout(binding = BINDING_INPUT_BUFFER, std430) buffer InputBufferU8   { uint8_t  u8data[];  };
+layout(binding = BINDING_INPUT_BUFFER, std430) buffer InputBufferU16  { uint16_t u16data[]; };
 #endif
 layout(binding = BINDING_INPUT_BUFFER, std430) buffer InputBufferU32  { uint   u32data[];  };
 layout(binding = BINDING_INPUT_BUFFER, std430) buffer InputBufferU64  { uvec2  u64data[];  };
@@ -96,10 +90,20 @@ const uint GOB_SIZE_Z_SHIFT = 0;
 const uint GOB_SIZE_SHIFT   = GOB_SIZE_X_SHIFT + GOB_SIZE_Y_SHIFT + GOB_SIZE_Z_SHIFT;
 const uvec2 SWIZZLE_MASK = uvec2(GOB_SIZE_X - 1u, GOB_SIZE_Y - 1u);
 
+uint SwizzleTable(uint pos) {
+    const uint t[8] = uint[](
+        0x12100200, 0x13110301, 0x16140604, 0x17150705,
+        0x1a180a08, 0x1b190b09, 0x1e1c0e0c, 0x1f1d0f0d
+    );
+    const uint i = pos >> 4;
+    const uint h = (t[i / 4] >> ((i % 4) * 8)) & 0xff;
+    return (h << 4) | (pos & 0xf);
+}
+
 // --- Helpers ---
 uint SwizzleOffset(uvec2 pos) {
     pos &= SWIZZLE_MASK;
-    return swizzle_table[pos.y * 64u + pos.x];
+    return SwizzleTable(pos.y * 64u + pos.x);
 }
 
 uvec4 ReadTexel(uint offset) {

src/video_core/macro.cpp

@@ -1328,22 +1328,14 @@ Macro::Opcode MacroJITx64Impl::GetOpCode() const {
 #endif
 
 static void Dump(u64 hash, std::span<const u32> code, bool decompiled = false) {
-    const auto base_dir{Common::FS::GetEdenPath(Common::FS::EdenPath::DumpDir)};
-    const auto macro_dir{base_dir / "macros"};
-    if (!Common::FS::CreateDir(base_dir) || !Common::FS::CreateDir(macro_dir)) {
-        LOG_ERROR(Common_Filesystem, "Failed to create macro dump directories");
+    const auto dump_dir{Common::FS::GetEdenPath(Common::FS::EdenPath::DumpDir)};
+    if (!Common::FS::CreateDir(dump_dir)) {
+        LOG_ERROR(Common_Filesystem, "Failed to create dump directory");
         return;
     }
-    auto name{macro_dir / fmt::format("{:016x}.macro", hash)};
-
-    if (decompiled) {
-        auto new_name{macro_dir / fmt::format("decompiled_{:016x}.macro", hash)};
-        if (Common::FS::Exists(name)) {
-            (void)Common::FS::RenameFile(name, new_name);
-            return;
-        }
-        name = new_name;
-    }
+    const char* const variant_suffix = decompiled ? "jit" : "raw";
+    const auto name{dump_dir / fmt::format("{:016x}_{:016x}_{}.macro",
+                                           Settings::GetCurrentProgramID(), hash, variant_suffix)};
 
     std::fstream macro_file(name, std::ios::out | std::ios::binary);
     if (!macro_file) {

src/video_core/renderer_opengl/gl_shader_cache.cpp

@@ -245,16 +245,31 @@ ShaderCache::ShaderCache(Tegra::MaxwellDeviceMemoryManager& device_memory_,
               std::min<u32>(device.GetMaxUserClipDistances(), Maxwell::Regs::NumClipDistances),
       },
       host_info{
-          .support_float64 = true,
-          .support_float16 = false,
-          .support_int64 = device.HasShaderInt64(),
-          .needs_demote_reorder = device.IsAmd(),
-          .support_snorm_render_buffer = false,
-          .support_viewport_index_layer = device.HasVertexViewportLayer(),
-          .min_ssbo_alignment = static_cast<u32>(device.GetShaderStorageBufferAlignment()),
-          .support_geometry_shader_passthrough = device.HasGeometryShaderPassthrough(),
-          .support_conditional_barrier = device.SupportsConditionalBarriers(),
+        .min_ssbo_alignment = static_cast<u32>(device.GetShaderStorageBufferAlignment()),
+        .max_per_stage_descriptor_sampled_images =
+            Shader::HostTranslateInfo::DEFAULT_DESCRIPTOR_LIMIT,
+        .max_per_stage_resources = Shader::HostTranslateInfo::DEFAULT_DESCRIPTOR_LIMIT,
+        .max_descriptor_set_samplers = Shader::HostTranslateInfo::DEFAULT_DESCRIPTOR_LIMIT,
+        .max_descriptor_set_uniform_buffers = Shader::HostTranslateInfo::DEFAULT_DESCRIPTOR_LIMIT,
+        .max_descriptor_set_uniform_buffers_dynamic =
+            Shader::HostTranslateInfo::DEFAULT_DESCRIPTOR_LIMIT,
+        .max_descriptor_set_storage_buffers = Shader::HostTranslateInfo::DEFAULT_DESCRIPTOR_LIMIT,
+        .max_descriptor_set_storage_buffers_dynamic =
+            Shader::HostTranslateInfo::DEFAULT_DESCRIPTOR_LIMIT,
+        .max_descriptor_set_sampled_images = Shader::HostTranslateInfo::DEFAULT_DESCRIPTOR_LIMIT,
+        .max_descriptor_set_storage_images = Shader::HostTranslateInfo::DEFAULT_DESCRIPTOR_LIMIT,
+        .max_descriptor_set_input_attachements =
+            Shader::HostTranslateInfo::DEFAULT_DESCRIPTOR_LIMIT,
+        .support_float64 = true,
+        .support_float16 = false,
+        .support_int64 = device.HasShaderInt64(),
+        .needs_demote_reorder = device.IsAmd(),
+        .support_snorm_render_buffer = false,
+        .support_viewport_index_layer = device.HasVertexViewportLayer(),
+        .support_geometry_shader_passthrough = device.HasGeometryShaderPassthrough(),
+        .support_conditional_barrier = device.SupportsConditionalBarriers(),
       } {
+    host_info.ApplyDescriptorLimitPolicy();
     if (use_asynchronous_shaders) {
         workers = CreateWorkers();
     }
@@ -481,7 +496,7 @@ std::unique_ptr<GraphicsPipeline> ShaderCache::CreateGraphicsPipeline(
         const u32 cfg_offset = u32(env.StartAddress() + sizeof(Shader::ProgramHeader));
         Shader::Maxwell::Flow::CFG cfg(env, pools.flow_block, cfg_offset, index == 0);
 
-        if (Settings::values.dump_shaders) {
+        if (Settings::values.dump_guest_shaders) {
             env.Dump(hash, key.unique_hashes[index]);
         }
 
@@ -578,7 +593,7 @@ std::unique_ptr<ComputePipeline> ShaderCache::CreateComputePipeline(
 
     Shader::Maxwell::Flow::CFG cfg{env, pools.flow_block, env.StartAddress()};
 
-    if (Settings::values.dump_shaders) {
+    if (Settings::values.dump_guest_shaders) {
         env.Dump(hash, key.unique_hash);
     }
 

src/video_core/renderer_opengl/util_shaders.cpp

@@ -1,3 +1,6 @@
+// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
+// SPDX-License-Identifier: GPL-3.0-or-later
+
 // SPDX-FileCopyrightText: Copyright 2020 yuzu Emulator Project
 // SPDX-License-Identifier: GPL-2.0-or-later
 
@@ -56,10 +59,8 @@ UtilShaders::UtilShaders(ProgramManager& program_manager_)
       copy_bc4_program(MakeProgram(OPENGL_COPY_BC4_COMP)),
       convert_s8d24_program(MakeProgram(OPENGL_CONVERT_S8D24_COMP)),
       convert_ms_to_nonms_program(MakeProgram(CONVERT_MSAA_TO_NON_MSAA_COMP)),
-      convert_nonms_to_ms_program(MakeProgram(CONVERT_NON_MSAA_TO_MSAA_COMP)) {
-    const auto swizzle_table = Tegra::Texture::MakeSwizzleTable();
-    swizzle_table_buffer.Create();
-    glNamedBufferStorage(swizzle_table_buffer.handle, sizeof(swizzle_table), &swizzle_table, 0);
+      convert_nonms_to_ms_program(MakeProgram(CONVERT_NON_MSAA_TO_MSAA_COMP))
+{
 }
 
 UtilShaders::~UtilShaders() = default;
@@ -116,13 +117,11 @@ void UtilShaders::ASTCDecode(Image& image, const StagingBufferMap& map,
 void UtilShaders::BlockLinearUpload2D(Image& image, const StagingBufferMap& map,
                                       std::span<const SwizzleParameters> swizzles) {
     static constexpr Extent3D WORKGROUP_SIZE{32, 32, 1};
-    static constexpr GLuint BINDING_SWIZZLE_BUFFER = 0;
-    static constexpr GLuint BINDING_INPUT_BUFFER = 1;
+    static constexpr GLuint BINDING_INPUT_BUFFER = 0;
     static constexpr GLuint BINDING_OUTPUT_IMAGE = 0;
 
     program_manager.BindComputeProgram(block_linear_unswizzle_2d_program.handle);
     glFlushMappedNamedBufferRange(map.buffer, map.offset, image.guest_size_bytes);
-    glBindBufferBase(GL_SHADER_STORAGE_BUFFER, BINDING_SWIZZLE_BUFFER, swizzle_table_buffer.handle);
 
     const GLenum store_format = StoreFormat(BytesPerBlock(image.info.format));
     for (const SwizzleParameters& swizzle : swizzles) {
@@ -153,14 +152,11 @@ void UtilShaders::BlockLinearUpload2D(Image& image, const StagingBufferMap& map,
 void UtilShaders::BlockLinearUpload3D(Image& image, const StagingBufferMap& map,
                                       std::span<const SwizzleParameters> swizzles) {
     static constexpr Extent3D WORKGROUP_SIZE{16, 8, 8};
-
-    static constexpr GLuint BINDING_SWIZZLE_BUFFER = 0;
-    static constexpr GLuint BINDING_INPUT_BUFFER = 1;
+    static constexpr GLuint BINDING_INPUT_BUFFER = 0;
     static constexpr GLuint BINDING_OUTPUT_IMAGE = 0;
 
     glFlushMappedNamedBufferRange(map.buffer, map.offset, image.guest_size_bytes);
     program_manager.BindComputeProgram(block_linear_unswizzle_3d_program.handle);
-    glBindBufferBase(GL_SHADER_STORAGE_BUFFER, BINDING_SWIZZLE_BUFFER, swizzle_table_buffer.handle);
 
     const GLenum store_format = StoreFormat(BytesPerBlock(image.info.format));
     for (const SwizzleParameters& swizzle : swizzles) {

src/video_core/renderer_opengl/util_shaders.h

@@ -1,3 +1,6 @@
+// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
+// SPDX-License-Identifier: GPL-3.0-or-later
+
 // SPDX-FileCopyrightText: Copyright 2020 yuzu Emulator Project
 // SPDX-License-Identifier: GPL-2.0-or-later
 
@@ -45,9 +48,6 @@ public:
 
 private:
     ProgramManager& program_manager;
-
-    OGLBuffer swizzle_table_buffer;
-
     OGLProgram astc_decoder_program;
     OGLProgram block_linear_unswizzle_2d_program;
     OGLProgram block_linear_unswizzle_3d_program;

src/video_core/renderer_vulkan/pipeline_helper.h

@@ -22,6 +22,15 @@ namespace Vulkan {
 
 using Shader::Backend::SPIRV::NUM_TEXTURE_AND_IMAGE_SCALING_WORDS;
 
+[[nodiscard]] inline u32 NumDescriptorEntries(const Shader::Info& info) {
+    return Shader::NumDescriptors(info.constant_buffer_descriptors) +
+           Shader::NumDescriptors(info.storage_buffers_descriptors) +
+           Shader::NumDescriptors(info.texture_buffer_descriptors) +
+           Shader::NumDescriptors(info.image_buffer_descriptors) +
+           Shader::NumDescriptors(info.texture_descriptors) +
+           Shader::NumDescriptors(info.image_descriptors);
+}
+
 class DescriptorLayoutBuilder {
 public:
     DescriptorLayoutBuilder(const Device& device_) : device{&device_} {}
@@ -194,7 +203,11 @@ inline void PushImageDescriptors(TextureCache& texture_cache,
             const VideoCommon::ImageViewId image_view_id{(views++)->id};
             const VideoCommon::SamplerId sampler_id{*(samplers++)};
             ImageView& image_view{texture_cache.GetImageView(image_view_id)};
-            const VkImageView vk_image_view{image_view.Handle(desc.type)};
+            VkImageView vk_image_view{image_view.Handle(desc.type)};
+            if (vk_image_view == VK_NULL_HANDLE) {
+                const VkImageView null_image_view{texture_cache.GetImageView(VideoCommon::NULL_IMAGE_VIEW_ID).Handle(desc.type)};
+                if (null_image_view != VK_NULL_HANDLE) vk_image_view = null_image_view;
+            }
             const Sampler& sampler{texture_cache.GetSampler(sampler_id)};
             const bool use_fallback_sampler{sampler.HasAddedAnisotropy() &&
                                             !image_view.SupportsAnisotropy()};

src/video_core/renderer_vulkan/renderer_vulkan.cpp

@@ -169,6 +169,7 @@ try
 }
 
 RendererVulkan::~RendererVulkan() {
+    scheduler.WaitWorker();
     scheduler.RegisterOnSubmit([] {});
     void(device.GetLogical().WaitIdle());
 }

src/video_core/renderer_vulkan/vk_compute_pass.cpp

@@ -25,6 +25,9 @@
 #include "video_core/host_shaders/vulkan_quad_indexed_comp_spv.h"
 #include "video_core/host_shaders/vulkan_uint8_comp_spv.h"
 #include "video_core/host_shaders/block_linear_unswizzle_3d_bcn_comp_spv.h"
+#include "video_core/host_shaders/block_linear_unswizzle_2d_comp_spv.h"
+#include "video_core/host_shaders/block_linear_unswizzle_3d_comp_spv.h"
+#include "video_core/host_shaders/pitch_unswizzle_comp_spv.h"
 #include "video_core/renderer_vulkan/vk_compute_pass.h"
 #include "video_core/renderer_vulkan/vk_descriptor_pool.h"
 #include "video_core/renderer_vulkan/vk_scheduler.h"
@@ -232,6 +235,26 @@ struct QueriesPrefixScanPushConstants {
 struct ConditionalRenderingResolvePushConstants {
     u32 compare_to_zero;
 };
+
+struct BlockLinear3DImagePushConstants {
+    alignas(16) std::array<u32, 3> origin;
+    alignas(16) std::array<s32, 3> destination;
+    u32 bytes_per_block_log2;
+    u32 slice_size;
+    u32 block_size;
+    u32 x_shift;
+    u32 block_height;
+    u32 block_height_mask;
+    u32 block_depth;
+    u32 block_depth_mask;
+};
+
+struct PitchUnswizzlePushConstants {
+    std::array<u32, 2> origin;
+    std::array<s32, 2> destination;
+    u32 bytes_per_block;
+    u32 pitch;
+};
 } // Anonymous namespace
 
 ComputePass::ComputePass(const Device& device_, Scheduler& scheduler, DescriptorPool& descriptor_pool,
@@ -326,7 +349,7 @@ std::pair<VkBuffer, VkDeviceSize> Uint8Pass::Assemble(u32 num_vertices, VkBuffer
     const u32 staging_size = static_cast<u32>(num_vertices * sizeof(u16));
     const auto staging = staging_buffer_pool.Request(staging_size, MemoryUsage::DeviceLocal);
 
-    compute_pass_descriptor_queue.Acquire();
+    compute_pass_descriptor_queue.Acquire(scheduler, 2);
     compute_pass_descriptor_queue.AddBuffer(src_buffer, src_offset, num_vertices);
     compute_pass_descriptor_queue.AddBuffer(staging.buffer, staging.offset, staging_size);
     const void* const descriptor_data{compute_pass_descriptor_queue.UpdateData()};
@@ -384,7 +407,7 @@ std::pair<VkBuffer, VkDeviceSize> QuadIndexedPass::Assemble(
     const std::size_t staging_size = num_tri_vertices * sizeof(u32);
     const auto staging = staging_buffer_pool.Request(staging_size, MemoryUsage::DeviceLocal);
 
-    compute_pass_descriptor_queue.Acquire();
+    compute_pass_descriptor_queue.Acquire(scheduler, 2);
     compute_pass_descriptor_queue.AddBuffer(src_buffer, src_offset, input_size);
     compute_pass_descriptor_queue.AddBuffer(staging.buffer, staging.offset, staging_size);
     const void* const descriptor_data{compute_pass_descriptor_queue.UpdateData()};
@@ -429,7 +452,7 @@ void ConditionalRenderingResolvePass::Resolve(VkBuffer dst_buffer, VkBuffer src_
     }
     const size_t compare_size = compare_to_zero ? 8 : 24;
 
-    compute_pass_descriptor_queue.Acquire();
+    compute_pass_descriptor_queue.Acquire(scheduler, 2);
     compute_pass_descriptor_queue.AddBuffer(src_buffer, src_offset, compare_size);
     compute_pass_descriptor_queue.AddBuffer(dst_buffer, 0, sizeof(u32));
     const void* const descriptor_data{compute_pass_descriptor_queue.UpdateData()};
@@ -498,7 +521,7 @@ void QueriesPrefixScanPass::Run(VkBuffer accumulation_buffer, VkBuffer dst_buffe
         static constexpr size_t DISPATCH_SIZE = 2048U;
         size_t runs_to_do = std::min<size_t>(current_runs, DISPATCH_SIZE);
         current_runs -= runs_to_do;
-        compute_pass_descriptor_queue.Acquire();
+        compute_pass_descriptor_queue.Acquire(scheduler, 3);
         compute_pass_descriptor_queue.AddBuffer(src_buffer, 0, number_of_sums * sizeof(u64));
         compute_pass_descriptor_queue.AddBuffer(dst_buffer, 0, number_of_sums * sizeof(u64));
         compute_pass_descriptor_queue.AddBuffer(accumulation_buffer, 0, sizeof(u64));
@@ -600,7 +623,7 @@ void ASTCDecoderPass::Assemble(Image& image, const StagingBufferRef& map,
         const u32 num_dispatches_y = Common::DivCeil(swizzle.num_tiles.height, 8U);
         const u32 num_dispatches_z = image.info.resources.layers;
 
-        compute_pass_descriptor_queue.Acquire();
+        compute_pass_descriptor_queue.Acquire(scheduler, 2);
         compute_pass_descriptor_queue.AddBuffer(map.buffer, input_offset,
                                                 image.guest_size_bytes - swizzle.buffer_offset);
         compute_pass_descriptor_queue.AddImage(image.StorageImageView(swizzle.level));
@@ -653,71 +676,311 @@ void ASTCDecoderPass::Assemble(Image& image, const StagingBufferRef& map,
     scheduler.Finish();
 }
 
-constexpr u32 BL3D_BINDING_SWIZZLE_TABLE = 0;
-constexpr u32 BL3D_BINDING_INPUT_BUFFER  = 1;
-constexpr u32 BL3D_BINDING_OUTPUT_BUFFER = 2;
+BlockLinearUnswizzleImage2DPass::BlockLinearUnswizzleImage2DPass(
+    const Device& device_, Scheduler& scheduler_, DescriptorPool& descriptor_pool_,
+    StagingBufferPool& staging_buffer_pool_,
+    ComputePassDescriptorQueue& compute_pass_descriptor_queue_)
+    : ComputePass(device_, scheduler_, descriptor_pool_, ASTC_DESCRIPTOR_SET_BINDINGS,
+                  ASTC_PASS_DESCRIPTOR_UPDATE_TEMPLATE_ENTRY, ASTC_BANK_INFO,
+                  COMPUTE_PUSH_CONSTANT_RANGE<sizeof(BlockLinearSwizzle2DParams)>,
+                  BLOCK_LINEAR_UNSWIZZLE_2D_COMP_SPV),
+      scheduler{scheduler_}, staging_buffer_pool{staging_buffer_pool_},
+      compute_pass_descriptor_queue{compute_pass_descriptor_queue_} {}
 
-constexpr std::array<VkDescriptorSetLayoutBinding, 3> BL3D_DESCRIPTOR_SET_BINDINGS{{
-    {
-        .binding = BL3D_BINDING_SWIZZLE_TABLE,
-        .descriptorType = VK_DESCRIPTOR_TYPE_STORAGE_BUFFER, // swizzle_table[]
-        .descriptorCount = 1,
-        .stageFlags = VK_SHADER_STAGE_COMPUTE_BIT,
-        .pImmutableSamplers = nullptr,
-    },
-    {
-        .binding = BL3D_BINDING_INPUT_BUFFER,
-        .descriptorType = VK_DESCRIPTOR_TYPE_STORAGE_BUFFER, // block-linear input
-        .descriptorCount = 1,
-        .stageFlags = VK_SHADER_STAGE_COMPUTE_BIT,
-        .pImmutableSamplers = nullptr,
-    },
-    {
-        .binding = BL3D_BINDING_OUTPUT_BUFFER,
-        .descriptorType = VK_DESCRIPTOR_TYPE_STORAGE_BUFFER,
-        .descriptorCount = 1,
-        .stageFlags = VK_SHADER_STAGE_COMPUTE_BIT,
-        .pImmutableSamplers = nullptr,
-    },
-}};
+BlockLinearUnswizzleImage2DPass::~BlockLinearUnswizzleImage2DPass() = default;
 
-constexpr DescriptorBankInfo BL3D_BANK_INFO{
-    .uniform_buffers = 0,
-    .storage_buffers = 3,
-    .texture_buffers = 0,
-    .image_buffers = 0,
-    .textures = 0,
-    .images = 0,
-    .score = 3,
-};
+void BlockLinearUnswizzleImage2DPass::Unswizzle(
+    Image& image, const StagingBufferRef& map,
+    std::span<const VideoCommon::SwizzleParameters> swizzles) {
+    using namespace VideoCommon::Accelerated;
+    scheduler.RequestOutsideRenderPassOperationContext();
+    const VkPipeline vk_pipeline = *pipeline;
+    const VkImageAspectFlags aspect_mask = image.AspectMask();
+    const VkImage vk_image = image.Handle();
+    const bool is_initialized = image.ExchangeInitialization();
+    scheduler.Record([vk_pipeline, vk_image, aspect_mask,
+                      is_initialized](vk::CommandBuffer cmdbuf) {
+        const VkImageMemoryBarrier image_barrier{
+            .sType = VK_STRUCTURE_TYPE_IMAGE_MEMORY_BARRIER,
+            .pNext = nullptr,
+            .srcAccessMask = static_cast<VkAccessFlags>(is_initialized ? VK_ACCESS_SHADER_WRITE_BIT
+                                                                       : VK_ACCESS_NONE),
+            .dstAccessMask = VK_ACCESS_SHADER_READ_BIT | VK_ACCESS_SHADER_WRITE_BIT,
+            .oldLayout = is_initialized ? VK_IMAGE_LAYOUT_GENERAL : VK_IMAGE_LAYOUT_UNDEFINED,
+            .newLayout = VK_IMAGE_LAYOUT_GENERAL,
+            .srcQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED,
+            .dstQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED,
+            .image = vk_image,
+            .subresourceRange{
+                .aspectMask = aspect_mask,
+                .baseMipLevel = 0,
+                .levelCount = VK_REMAINING_MIP_LEVELS,
+                .baseArrayLayer = 0,
+                .layerCount = VK_REMAINING_ARRAY_LAYERS,
+            },
+        };
+        cmdbuf.PipelineBarrier(is_initialized ? vk::PIPELINE_STAGE_GRAPHICS_COMPUTE_TRANSFER
+                              : VkPipelineStageFlags(VK_PIPELINE_STAGE_TOP_OF_PIPE_BIT),
+                       VK_PIPELINE_STAGE_COMPUTE_SHADER_BIT, 0, image_barrier);
+        cmdbuf.BindPipeline(VK_PIPELINE_BIND_POINT_COMPUTE, vk_pipeline);
+    });
+    for (const VideoCommon::SwizzleParameters& swizzle : swizzles) {
+        const size_t input_offset = swizzle.buffer_offset + map.offset;
+        const u32 num_dispatches_x = Common::DivCeil(swizzle.num_tiles.width, 32U);
+        const u32 num_dispatches_y = Common::DivCeil(swizzle.num_tiles.height, 32U);
+        const u32 num_dispatches_z = image.info.resources.layers;
 
-constexpr std::array<VkDescriptorUpdateTemplateEntry, 3>
-    BL3D_DESCRIPTOR_UPDATE_TEMPLATE_ENTRY{{
-        {
-            .dstBinding = BL3D_BINDING_SWIZZLE_TABLE,
-            .dstArrayElement = 0,
-            .descriptorCount = 1,
-            .descriptorType = VK_DESCRIPTOR_TYPE_STORAGE_BUFFER,
-            .offset = BL3D_BINDING_SWIZZLE_TABLE * sizeof(DescriptorUpdateEntry),
-            .stride = sizeof(DescriptorUpdateEntry),
-        },
-        {
-            .dstBinding = BL3D_BINDING_INPUT_BUFFER,
-            .dstArrayElement = 0,
-            .descriptorCount = 1,
-            .descriptorType = VK_DESCRIPTOR_TYPE_STORAGE_BUFFER,
-            .offset = BL3D_BINDING_INPUT_BUFFER * sizeof(DescriptorUpdateEntry),
-            .stride = sizeof(DescriptorUpdateEntry),
-        },
-        {
-            .dstBinding = BL3D_BINDING_OUTPUT_BUFFER,
-            .dstArrayElement = 0,
-            .descriptorCount = 1,
-            .descriptorType = VK_DESCRIPTOR_TYPE_STORAGE_BUFFER,
-            .offset = BL3D_BINDING_OUTPUT_BUFFER * sizeof(DescriptorUpdateEntry),
-            .stride = sizeof(DescriptorUpdateEntry),
-        }
-    }};
+        compute_pass_descriptor_queue.Acquire(scheduler, 2);
+        compute_pass_descriptor_queue.AddBuffer(map.buffer, input_offset,
+                                                image.guest_size_bytes - swizzle.buffer_offset);
+        compute_pass_descriptor_queue.AddImage(image.StorageImageView(swizzle.level));
+        const void* const descriptor_data{compute_pass_descriptor_queue.UpdateData()};
+
+        const auto params = MakeBlockLinearSwizzle2DParams(swizzle, image.info);
+        scheduler.Record([this, num_dispatches_x, num_dispatches_y, num_dispatches_z, params,
+                          descriptor_data](vk::CommandBuffer cmdbuf) {
+            const VkDescriptorSet set = descriptor_allocator.Commit();
+            device.GetLogical().UpdateDescriptorSet(set, *descriptor_template, descriptor_data);
+            cmdbuf.BindDescriptorSets(VK_PIPELINE_BIND_POINT_COMPUTE, *layout, 0, set, {});
+            cmdbuf.PushConstants(*layout, VK_SHADER_STAGE_COMPUTE_BIT, params);
+            cmdbuf.Dispatch(num_dispatches_x, num_dispatches_y, num_dispatches_z);
+        });
+    }
+    scheduler.Record([vk_image, aspect_mask](vk::CommandBuffer cmdbuf) {
+        const VkImageMemoryBarrier image_barrier{
+            .sType = VK_STRUCTURE_TYPE_IMAGE_MEMORY_BARRIER,
+            .pNext = nullptr,
+            .srcAccessMask = VK_ACCESS_SHADER_WRITE_BIT,
+            .dstAccessMask = VK_ACCESS_SHADER_READ_BIT | VK_ACCESS_SHADER_WRITE_BIT |
+                             VK_ACCESS_TRANSFER_READ_BIT | VK_ACCESS_TRANSFER_WRITE_BIT,
+            .oldLayout = VK_IMAGE_LAYOUT_GENERAL,
+            .newLayout = VK_IMAGE_LAYOUT_GENERAL,
+            .srcQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED,
+            .dstQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED,
+            .image = vk_image,
+            .subresourceRange{
+                .aspectMask = aspect_mask,
+                .baseMipLevel = 0,
+                .levelCount = VK_REMAINING_MIP_LEVELS,
+                .baseArrayLayer = 0,
+                .layerCount = VK_REMAINING_ARRAY_LAYERS,
+            },
+        };
+        cmdbuf.PipelineBarrier(VK_PIPELINE_STAGE_COMPUTE_SHADER_BIT,
+                       vk::PIPELINE_STAGE_GRAPHICS_COMPUTE_TRANSFER, 0, image_barrier);
+    });
+    scheduler.Finish();
+}
+
+BlockLinearUnswizzleImage3DPass::BlockLinearUnswizzleImage3DPass(
+    const Device& device_, Scheduler& scheduler_, DescriptorPool& descriptor_pool_,
+    StagingBufferPool& staging_buffer_pool_,
+    ComputePassDescriptorQueue& compute_pass_descriptor_queue_)
+    : ComputePass(device_, scheduler_, descriptor_pool_, ASTC_DESCRIPTOR_SET_BINDINGS,
+                  ASTC_PASS_DESCRIPTOR_UPDATE_TEMPLATE_ENTRY, ASTC_BANK_INFO,
+                  COMPUTE_PUSH_CONSTANT_RANGE<sizeof(BlockLinear3DImagePushConstants)>,
+                  BLOCK_LINEAR_UNSWIZZLE_3D_COMP_SPV),
+      scheduler{scheduler_}, staging_buffer_pool{staging_buffer_pool_},
+      compute_pass_descriptor_queue{compute_pass_descriptor_queue_} {}
+
+BlockLinearUnswizzleImage3DPass::~BlockLinearUnswizzleImage3DPass() = default;
+
+void BlockLinearUnswizzleImage3DPass::Unswizzle(
+    Image& image, const StagingBufferRef& map,
+    std::span<const VideoCommon::SwizzleParameters> swizzles) {
+    using namespace VideoCommon::Accelerated;
+    scheduler.RequestOutsideRenderPassOperationContext();
+    const VkPipeline vk_pipeline = *pipeline;
+    const VkImageAspectFlags aspect_mask = image.AspectMask();
+    const VkImage vk_image = image.Handle();
+    const bool is_initialized = image.ExchangeInitialization();
+    scheduler.Record([vk_pipeline, vk_image, aspect_mask,
+                      is_initialized](vk::CommandBuffer cmdbuf) {
+        const VkImageMemoryBarrier image_barrier{
+            .sType = VK_STRUCTURE_TYPE_IMAGE_MEMORY_BARRIER,
+            .pNext = nullptr,
+            .srcAccessMask = static_cast<VkAccessFlags>(is_initialized ? VK_ACCESS_SHADER_WRITE_BIT
+                                                                       : VK_ACCESS_NONE),
+            .dstAccessMask = VK_ACCESS_SHADER_READ_BIT | VK_ACCESS_SHADER_WRITE_BIT,
+            .oldLayout = is_initialized ? VK_IMAGE_LAYOUT_GENERAL : VK_IMAGE_LAYOUT_UNDEFINED,
+            .newLayout = VK_IMAGE_LAYOUT_GENERAL,
+            .srcQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED,
+            .dstQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED,
+            .image = vk_image,
+            .subresourceRange{
+                .aspectMask = aspect_mask,
+                .baseMipLevel = 0,
+                .levelCount = VK_REMAINING_MIP_LEVELS,
+                .baseArrayLayer = 0,
+                .layerCount = VK_REMAINING_ARRAY_LAYERS,
+            },
+        };
+        cmdbuf.PipelineBarrier(is_initialized ? vk::PIPELINE_STAGE_GRAPHICS_COMPUTE_TRANSFER
+                              : VkPipelineStageFlags(VK_PIPELINE_STAGE_TOP_OF_PIPE_BIT),
+                       VK_PIPELINE_STAGE_COMPUTE_SHADER_BIT, 0, image_barrier);
+        cmdbuf.BindPipeline(VK_PIPELINE_BIND_POINT_COMPUTE, vk_pipeline);
+    });
+    for (const VideoCommon::SwizzleParameters& swizzle : swizzles) {
+        const size_t input_offset = swizzle.buffer_offset + map.offset;
+        const u32 num_dispatches_x = Common::DivCeil(swizzle.num_tiles.width, 16U);
+        const u32 num_dispatches_y = Common::DivCeil(swizzle.num_tiles.height, 8U);
+        const u32 num_dispatches_z = Common::DivCeil(swizzle.num_tiles.depth, 8U);
+
+        compute_pass_descriptor_queue.Acquire(scheduler, 2);
+        compute_pass_descriptor_queue.AddBuffer(map.buffer, input_offset,
+                                                image.guest_size_bytes - swizzle.buffer_offset);
+        compute_pass_descriptor_queue.AddImage(image.StorageImageView(swizzle.level));
+        const void* const descriptor_data{compute_pass_descriptor_queue.UpdateData()};
+
+        const auto p = MakeBlockLinearSwizzle3DParams(swizzle, image.info);
+        const BlockLinear3DImagePushConstants params{
+            .origin = p.origin,
+            .destination = p.destination,
+            .bytes_per_block_log2 = p.bytes_per_block_log2,
+            .slice_size = p.slice_size,
+            .block_size = p.block_size,
+            .x_shift = p.x_shift,
+            .block_height = p.block_height,
+            .block_height_mask = p.block_height_mask,
+            .block_depth = p.block_depth,
+            .block_depth_mask = p.block_depth_mask,
+        };
+        scheduler.Record([this, num_dispatches_x, num_dispatches_y, num_dispatches_z, params,
+                          descriptor_data](vk::CommandBuffer cmdbuf) {
+            const VkDescriptorSet set = descriptor_allocator.Commit();
+            device.GetLogical().UpdateDescriptorSet(set, *descriptor_template, descriptor_data);
+            cmdbuf.BindDescriptorSets(VK_PIPELINE_BIND_POINT_COMPUTE, *layout, 0, set, {});
+            cmdbuf.PushConstants(*layout, VK_SHADER_STAGE_COMPUTE_BIT, params);
+            cmdbuf.Dispatch(num_dispatches_x, num_dispatches_y, num_dispatches_z);
+        });
+    }
+    scheduler.Record([vk_image, aspect_mask](vk::CommandBuffer cmdbuf) {
+        const VkImageMemoryBarrier image_barrier{
+            .sType = VK_STRUCTURE_TYPE_IMAGE_MEMORY_BARRIER,
+            .pNext = nullptr,
+            .srcAccessMask = VK_ACCESS_SHADER_WRITE_BIT,
+            .dstAccessMask = VK_ACCESS_SHADER_READ_BIT | VK_ACCESS_SHADER_WRITE_BIT |
+                             VK_ACCESS_TRANSFER_READ_BIT | VK_ACCESS_TRANSFER_WRITE_BIT,
+            .oldLayout = VK_IMAGE_LAYOUT_GENERAL,
+            .newLayout = VK_IMAGE_LAYOUT_GENERAL,
+            .srcQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED,
+            .dstQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED,
+            .image = vk_image,
+            .subresourceRange{
+                .aspectMask = aspect_mask,
+                .baseMipLevel = 0,
+                .levelCount = VK_REMAINING_MIP_LEVELS,
+                .baseArrayLayer = 0,
+                .layerCount = VK_REMAINING_ARRAY_LAYERS,
+            },
+        };
+        cmdbuf.PipelineBarrier(VK_PIPELINE_STAGE_COMPUTE_SHADER_BIT,
+                       vk::PIPELINE_STAGE_GRAPHICS_COMPUTE_TRANSFER, 0, image_barrier);
+    });
+    scheduler.Finish();
+}
+
+PitchUnswizzlePass::PitchUnswizzlePass(const Device& device_, Scheduler& scheduler_,
+                                       DescriptorPool& descriptor_pool_,
+                                       StagingBufferPool& staging_buffer_pool_,
+                                       ComputePassDescriptorQueue& compute_pass_descriptor_queue_)
+    : ComputePass(device_, scheduler_, descriptor_pool_, ASTC_DESCRIPTOR_SET_BINDINGS,
+                  ASTC_PASS_DESCRIPTOR_UPDATE_TEMPLATE_ENTRY, ASTC_BANK_INFO,
+                  COMPUTE_PUSH_CONSTANT_RANGE<sizeof(PitchUnswizzlePushConstants)>,
+                  PITCH_UNSWIZZLE_COMP_SPV),
+      scheduler{scheduler_}, staging_buffer_pool{staging_buffer_pool_},
+      compute_pass_descriptor_queue{compute_pass_descriptor_queue_} {}
+
+PitchUnswizzlePass::~PitchUnswizzlePass() = default;
+
+void PitchUnswizzlePass::Unswizzle(Image& image, const StagingBufferRef& map,
+                                   std::span<const VideoCommon::SwizzleParameters> swizzles) {
+    scheduler.RequestOutsideRenderPassOperationContext();
+    const VkPipeline vk_pipeline = *pipeline;
+    const VkImageAspectFlags aspect_mask = image.AspectMask();
+    const VkImage vk_image = image.Handle();
+    const bool is_initialized = image.ExchangeInitialization();
+    scheduler.Record([vk_pipeline, vk_image, aspect_mask,
+                      is_initialized](vk::CommandBuffer cmdbuf) {
+        const VkImageMemoryBarrier image_barrier{
+            .sType = VK_STRUCTURE_TYPE_IMAGE_MEMORY_BARRIER,
+            .pNext = nullptr,
+            .srcAccessMask = static_cast<VkAccessFlags>(is_initialized ? VK_ACCESS_SHADER_WRITE_BIT
+                                                                       : VK_ACCESS_NONE),
+            .dstAccessMask = VK_ACCESS_SHADER_READ_BIT | VK_ACCESS_SHADER_WRITE_BIT,
+            .oldLayout = is_initialized ? VK_IMAGE_LAYOUT_GENERAL : VK_IMAGE_LAYOUT_UNDEFINED,
+            .newLayout = VK_IMAGE_LAYOUT_GENERAL,
+            .srcQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED,
+            .dstQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED,
+            .image = vk_image,
+            .subresourceRange{
+                .aspectMask = aspect_mask,
+                .baseMipLevel = 0,
+                .levelCount = VK_REMAINING_MIP_LEVELS,
+                .baseArrayLayer = 0,
+                .layerCount = VK_REMAINING_ARRAY_LAYERS,
+            },
+        };
+        cmdbuf.PipelineBarrier(is_initialized ? vk::PIPELINE_STAGE_GRAPHICS_COMPUTE_TRANSFER
+                              : VkPipelineStageFlags(VK_PIPELINE_STAGE_TOP_OF_PIPE_BIT),
+                       VK_PIPELINE_STAGE_COMPUTE_SHADER_BIT, 0, image_barrier);
+        cmdbuf.BindPipeline(VK_PIPELINE_BIND_POINT_COMPUTE, vk_pipeline);
+    });
+    const u32 bytes_per_block = VideoCore::Surface::BytesPerBlock(image.info.format);
+    for (const VideoCommon::SwizzleParameters& swizzle : swizzles) {
+        const size_t input_offset = swizzle.buffer_offset + map.offset;
+        const u32 num_dispatches_x = Common::DivCeil(swizzle.num_tiles.width, 32U);
+        const u32 num_dispatches_y = Common::DivCeil(swizzle.num_tiles.height, 32U);
+
+        compute_pass_descriptor_queue.Acquire(scheduler, 2);
+        compute_pass_descriptor_queue.AddBuffer(map.buffer, input_offset,
+                                                image.guest_size_bytes - swizzle.buffer_offset);
+        compute_pass_descriptor_queue.AddImage(image.StorageImageView(swizzle.level));
+        const void* const descriptor_data{compute_pass_descriptor_queue.UpdateData()};
+
+        const PitchUnswizzlePushConstants params{
+            .origin = {0, 0},
+            .destination = {0, 0},
+            .bytes_per_block = bytes_per_block,
+            .pitch = image.info.pitch,
+        };
+        scheduler.Record([this, num_dispatches_x, num_dispatches_y, params,
+                          descriptor_data](vk::CommandBuffer cmdbuf) {
+            const VkDescriptorSet set = descriptor_allocator.Commit();
+            device.GetLogical().UpdateDescriptorSet(set, *descriptor_template, descriptor_data);
+            cmdbuf.BindDescriptorSets(VK_PIPELINE_BIND_POINT_COMPUTE, *layout, 0, set, {});
+            cmdbuf.PushConstants(*layout, VK_SHADER_STAGE_COMPUTE_BIT, params);
+            cmdbuf.Dispatch(num_dispatches_x, num_dispatches_y, 1);
+        });
+    }
+    scheduler.Record([vk_image, aspect_mask](vk::CommandBuffer cmdbuf) {
+        const VkImageMemoryBarrier image_barrier{
+            .sType = VK_STRUCTURE_TYPE_IMAGE_MEMORY_BARRIER,
+            .pNext = nullptr,
+            .srcAccessMask = VK_ACCESS_SHADER_WRITE_BIT,
+            .dstAccessMask = VK_ACCESS_SHADER_READ_BIT | VK_ACCESS_SHADER_WRITE_BIT |
+                             VK_ACCESS_TRANSFER_READ_BIT | VK_ACCESS_TRANSFER_WRITE_BIT,
+            .oldLayout = VK_IMAGE_LAYOUT_GENERAL,
+            .newLayout = VK_IMAGE_LAYOUT_GENERAL,
+            .srcQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED,
+            .dstQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED,
+            .image = vk_image,
+            .subresourceRange{
+                .aspectMask = aspect_mask,
+                .baseMipLevel = 0,
+                .levelCount = VK_REMAINING_MIP_LEVELS,
+                .baseArrayLayer = 0,
+                .layerCount = VK_REMAINING_ARRAY_LAYERS,
+            },
+        };
+        cmdbuf.PipelineBarrier(VK_PIPELINE_STAGE_COMPUTE_SHADER_BIT,
+                       vk::PIPELINE_STAGE_GRAPHICS_COMPUTE_TRANSFER, 0, image_barrier);
+    });
+    scheduler.Finish();
+}
+
+constexpr u32 BL3D_BINDING_INPUT_BUFFER  = 0;
+constexpr u32 BL3D_BINDING_OUTPUT_BUFFER = 1;
 
 struct alignas(16) BlockLinearUnswizzle3DPushConstants {
     u32 blocks_dim[3];           // Offset 0
@@ -745,11 +1008,50 @@ BlockLinearUnswizzle3DPass::BlockLinearUnswizzle3DPass(
     DescriptorPool& descriptor_pool_,
     StagingBufferPool& staging_buffer_pool_,
     ComputePassDescriptorQueue& compute_pass_descriptor_queue_)
-    : ComputePass(
-          device_, scheduler_, descriptor_pool_,
-          BL3D_DESCRIPTOR_SET_BINDINGS,
-          BL3D_DESCRIPTOR_UPDATE_TEMPLATE_ENTRY,
-          BL3D_BANK_INFO,
+    : ComputePass(device_, scheduler_, descriptor_pool_,
+        std::array<VkDescriptorSetLayoutBinding, 2>{{
+    {
+        .binding = BL3D_BINDING_INPUT_BUFFER,
+        .descriptorType = VK_DESCRIPTOR_TYPE_STORAGE_BUFFER, // block-linear input
+        .descriptorCount = 1,
+        .stageFlags = VK_SHADER_STAGE_COMPUTE_BIT,
+        .pImmutableSamplers = nullptr,
+    },
+    {
+        .binding = BL3D_BINDING_OUTPUT_BUFFER,
+        .descriptorType = VK_DESCRIPTOR_TYPE_STORAGE_BUFFER,
+        .descriptorCount = 1,
+        .stageFlags = VK_SHADER_STAGE_COMPUTE_BIT,
+        .pImmutableSamplers = nullptr,
+    },
+        }},
+        std::array<VkDescriptorUpdateTemplateEntry, 2>{{
+        {
+            .dstBinding = BL3D_BINDING_INPUT_BUFFER,
+            .dstArrayElement = 0,
+            .descriptorCount = 1,
+            .descriptorType = VK_DESCRIPTOR_TYPE_STORAGE_BUFFER,
+            .offset = BL3D_BINDING_INPUT_BUFFER * sizeof(DescriptorUpdateEntry),
+            .stride = sizeof(DescriptorUpdateEntry),
+        },
+        {
+            .dstBinding = BL3D_BINDING_OUTPUT_BUFFER,
+            .dstArrayElement = 0,
+            .descriptorCount = 1,
+            .descriptorType = VK_DESCRIPTOR_TYPE_STORAGE_BUFFER,
+            .offset = BL3D_BINDING_OUTPUT_BUFFER * sizeof(DescriptorUpdateEntry),
+            .stride = sizeof(DescriptorUpdateEntry),
+        }
+        }},
+        DescriptorBankInfo{
+            .uniform_buffers = 0,
+            .storage_buffers = 2,
+            .texture_buffers = 0,
+            .image_buffers = 0,
+            .textures = 0,
+            .images = 0,
+            .score = 2,
+        },
           COMPUTE_PUSH_CONSTANT_RANGE<sizeof(BlockLinearUnswizzle3DPushConstants)>,
           BLOCK_LINEAR_UNSWIZZLE_3D_BCN_COMP_SPV),
       scheduler{scheduler_},
@@ -821,9 +1123,7 @@ void BlockLinearUnswizzle3DPass::UnswizzleChunk(
     pc.blocks_dim[1] = blocks_y;
     pc.blocks_dim[2] = z_count; // Only process the count
 
-    compute_pass_descriptor_queue.Acquire();
-    compute_pass_descriptor_queue.AddBuffer(*image.runtime->swizzle_table_buffer, 0,
-                                           image.runtime->swizzle_table_size);
+    compute_pass_descriptor_queue.Acquire(scheduler, 3);
     compute_pass_descriptor_queue.AddBuffer(swizzled.buffer,
                                            sw.buffer_offset + swizzled.offset,
                                            image.guest_size_bytes - sw.buffer_offset);
@@ -859,6 +1159,23 @@ void BlockLinearUnswizzle3DPass::UnswizzleChunk(
             return;
         }
 
+        if (!is_first_chunk) {
+            const VkBufferMemoryBarrier reuse_barrier{
+                .sType = VK_STRUCTURE_TYPE_BUFFER_MEMORY_BARRIER,
+                .pNext = nullptr,
+                .srcAccessMask = VK_ACCESS_TRANSFER_READ_BIT,
+                .dstAccessMask = VK_ACCESS_SHADER_WRITE_BIT,
+                .srcQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED,
+                .dstQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED,
+                .buffer = out_buffer,
+                .offset = 0,
+                .size = VK_WHOLE_SIZE,
+            };
+            cmdbuf.PipelineBarrier(VK_PIPELINE_STAGE_TRANSFER_BIT,
+                                   VK_PIPELINE_STAGE_COMPUTE_SHADER_BIT, 0, nullptr,
+                                   reuse_barrier, nullptr);
+        }
+
         device.GetLogical().UpdateDescriptorSet(set, *descriptor_template, descriptor_data);
         cmdbuf.BindPipeline(VK_PIPELINE_BIND_POINT_COMPUTE, *pipeline);
         cmdbuf.BindDescriptorSets(VK_PIPELINE_BIND_POINT_COMPUTE, *layout, 0, set, {});
@@ -989,7 +1306,7 @@ void MSAACopyPass::CopyImage(Image& dst_image, Image& src_image,
         ASSERT(copy.dst_subresource.base_layer == 0);
         ASSERT(copy.dst_subresource.num_layers == 1);
 
-        compute_pass_descriptor_queue.Acquire();
+        compute_pass_descriptor_queue.Acquire(scheduler, 2);
         compute_pass_descriptor_queue.AddImage(
             src_image.StorageImageView(copy.src_subresource.base_level));
         compute_pass_descriptor_queue.AddImage(

src/video_core/renderer_vulkan/vk_compute_pass.h

@@ -25,6 +25,7 @@ struct SwizzleParameters;
 
 namespace Vulkan {
 
+using VideoCommon::Accelerated::BlockLinearSwizzle2DParams;
 using VideoCommon::Accelerated::BlockLinearSwizzle3DParams;
 
 class Device;
@@ -164,6 +165,56 @@ private:
     ComputePassDescriptorQueue& compute_pass_descriptor_queue;
 };
 
+class BlockLinearUnswizzleImage2DPass final : public ComputePass {
+public:
+    explicit BlockLinearUnswizzleImage2DPass(const Device& device_, Scheduler& scheduler_,
+                                             DescriptorPool& descriptor_pool_,
+                                             StagingBufferPool& staging_buffer_pool_,
+                                             ComputePassDescriptorQueue& compute_pass_descriptor_queue_);
+    ~BlockLinearUnswizzleImage2DPass();
+
+    void Unswizzle(Image& image, const StagingBufferRef& map,
+                   std::span<const VideoCommon::SwizzleParameters> swizzles);
+
+private:
+    Scheduler& scheduler;
+    StagingBufferPool& staging_buffer_pool;
+    ComputePassDescriptorQueue& compute_pass_descriptor_queue;
+};
+
+class BlockLinearUnswizzleImage3DPass final : public ComputePass {
+public:
+    explicit BlockLinearUnswizzleImage3DPass(const Device& device_, Scheduler& scheduler_,
+                                             DescriptorPool& descriptor_pool_,
+                                             StagingBufferPool& staging_buffer_pool_,
+                                             ComputePassDescriptorQueue& compute_pass_descriptor_queue_);
+    ~BlockLinearUnswizzleImage3DPass();
+
+    void Unswizzle(Image& image, const StagingBufferRef& map,
+                   std::span<const VideoCommon::SwizzleParameters> swizzles);
+
+private:
+    Scheduler& scheduler;
+    StagingBufferPool& staging_buffer_pool;
+    ComputePassDescriptorQueue& compute_pass_descriptor_queue;
+};
+
+class PitchUnswizzlePass final : public ComputePass {
+public:
+    explicit PitchUnswizzlePass(const Device& device_, Scheduler& scheduler_,
+                                DescriptorPool& descriptor_pool_,
+                                StagingBufferPool& staging_buffer_pool_,
+                                ComputePassDescriptorQueue& compute_pass_descriptor_queue_);
+    ~PitchUnswizzlePass();
+
+    void Unswizzle(Image& image, const StagingBufferRef& map,
+                   std::span<const VideoCommon::SwizzleParameters> swizzles);
+
+private:
+    Scheduler& scheduler;
+    StagingBufferPool& staging_buffer_pool;
+    ComputePassDescriptorQueue& compute_pass_descriptor_queue;
+};
 
 class MSAACopyPass final : public ComputePass {
 public:

src/video_core/renderer_vulkan/vk_compute_pipeline.cpp

@@ -45,6 +45,7 @@ ComputePipeline::ComputePipeline(const Device& device_, Scheduler& scheduler, vk
     }
     std::copy_n(info.constant_buffer_used_sizes.begin(), uniform_buffer_sizes.size(),
                 uniform_buffer_sizes.begin());
+    num_descriptor_entries = NumDescriptorEntries(info);
 
     auto func{[this, &scheduler, &descriptor_pool, shader_notify, pipeline_statistics] {
         DescriptorLayoutBuilder builder{device};
@@ -87,7 +88,7 @@ ComputePipeline::ComputePipeline(const Device& device_, Scheduler& scheduler, vk
         }, *pipeline_cache);
 
         // Log compute pipeline creation
-        if (Settings::values.gpu_logging_enabled.GetValue()) {
+        if (GPU::Logging::IsActive()) {
             GPU::Logging::GPULogger::GetInstance().LogPipelineStateChange(
                 "ComputePipeline created"
             );
@@ -113,7 +114,7 @@ ComputePipeline::ComputePipeline(const Device& device_, Scheduler& scheduler, vk
 void ComputePipeline::Configure(Tegra::Engines::KeplerCompute& kepler_compute,
                                 Tegra::MemoryManager& gpu_memory, Scheduler& scheduler,
                                 BufferCache& buffer_cache, TextureCache& texture_cache) {
-    guest_descriptor_queue.Acquire();
+    guest_descriptor_queue.Acquire(scheduler, num_descriptor_entries);
 
     buffer_cache.SetComputeUniformBufferState(info.constant_buffer_mask, &uniform_buffer_sizes);
     buffer_cache.UnbindComputeStorageBuffers();
@@ -223,7 +224,7 @@ void ComputePipeline::Configure(Tegra::Engines::KeplerCompute& kepler_compute,
     }
 
     // Log compute pipeline binding
-    if (Settings::values.gpu_logging_enabled.GetValue() &&
+    if (GPU::Logging::IsActive() &&
         Settings::values.gpu_log_vulkan_calls.GetValue()) {
         GPU::Logging::GPULogger::GetInstance().LogPipelineBind(true, "compute pipeline");
     }

src/video_core/renderer_vulkan/vk_compute_pipeline.h

@@ -53,6 +53,7 @@ private:
     vk::PipelineCache& pipeline_cache;
     GuestDescriptorQueue& guest_descriptor_queue;
     Shader::Info info;
+    u32 num_descriptor_entries{};
 
     VideoCommon::ComputeUniformBufferSizes uniform_buffer_sizes{};
 

src/video_core/renderer_vulkan/vk_graphics_pipeline.cpp

@@ -268,6 +268,7 @@ GraphicsPipeline::GraphicsPipeline(
         num_textures += Shader::NumDescriptors(info->texture_descriptors);
         num_image_elements += Shader::NumDescriptors(info->texture_descriptors);
         num_image_elements += Shader::NumDescriptors(info->image_descriptors);
+        num_descriptor_entries += NumDescriptorEntries(*info);
     }
     fragment_has_color0_output = stage_infos[NUM_STAGES - 1].stores_frag_color[0];
     auto func{[this, shader_notify, &render_pass_cache, &descriptor_pool, pipeline_statistics] {
@@ -473,7 +474,7 @@ bool GraphicsPipeline::ConfigureImpl(bool is_indexed) {
     buffer_cache.UpdateGraphicsBuffers(is_indexed);
     buffer_cache.BindHostGeometryBuffers(is_indexed);
 
-    guest_descriptor_queue.Acquire();
+    guest_descriptor_queue.Acquire(scheduler, num_descriptor_entries);
 
     RescalingPushConstant rescaling;
     RenderAreaPushConstant render_area;
@@ -532,7 +533,7 @@ void GraphicsPipeline::ConfigureDraw(const RescalingPushConstant& rescaling,
     const bool bind_pipeline{scheduler.UpdateGraphicsPipeline(this)};
 
     // Log graphics pipeline binding
-    if (bind_pipeline && Settings::values.gpu_logging_enabled.GetValue() &&
+    if (bind_pipeline && GPU::Logging::IsActive() &&
         Settings::values.gpu_log_vulkan_calls.GetValue()) {
         const std::string pipeline_info = fmt::format("hash=0x{:016x}", key.Hash());
         GPU::Logging::GPULogger::GetInstance().LogPipelineBind(false, pipeline_info);
@@ -986,7 +987,7 @@ void GraphicsPipeline::MakePipeline(VkRenderPass render_pass) {
     }, *pipeline_cache);
 
     // Log graphics pipeline creation
-    if (Settings::values.gpu_logging_enabled.GetValue()) {
+    if (GPU::Logging::IsActive()) {
         const std::string pipeline_info = fmt::format(
             "GraphicsPipeline created: stages={}, attachments={}",
             shader_stages.size(),

src/video_core/renderer_vulkan/vk_graphics_pipeline.h

@@ -159,6 +159,7 @@ private:
     std::array<Shader::Info, NUM_STAGES> stage_infos;
     std::array<u32, 5> enabled_uniform_buffer_masks{};
     VideoCommon::UniformBufferSizes uniform_buffer_sizes{};
+    u32 num_descriptor_entries{};
     size_t num_image_elements{};
     u32 num_textures{};
     bool fragment_has_color0_output{};