configure_input: Use Joycons Docked instead of Connected as label

Handles button configuration for all controller layouts and debug pads. Configurable at construction.

.travis/clang-format/build.sh

@@ -1,3 +1,3 @@
 #!/bin/bash -ex
 
-docker run -v $(pwd):/yuzu ubuntu:18.04 /bin/bash -ex /yuzu/.travis/clang-format/docker.sh
+docker run --env-file .travis/common/travis-ci.env -v $(pwd):/yuzu -v "$HOME/.ccache":/root/.ccache citraemu/build-environments:linux-clang-format /bin/bash -ex /yuzu/.travis/clang-format/docker.sh

.travis/clang-format/deps.sh

@@ -1,3 +1,3 @@
 #!/bin/sh -ex
 
-docker pull ubuntu:18.04
+docker pull citraemu/build-environments:linux-clang-format

.travis/clang-format/docker.sh

@@ -1,8 +1,5 @@
 #!/bin/bash -ex
 
-apt-get update
-apt-get install -y clang-format-6.0
-
 # Run clang-format
 cd /yuzu
 ./.travis/clang-format/script.sh

.travis/linux-mingw/docker.sh

@@ -57,3 +57,4 @@ done
 
 pip3 install pefile
 python3 .travis/linux-mingw/scan_dll.py package/*.exe "package/"
+python3 .travis/linux-mingw/scan_dll.py package/imageformats/*.dll "package/"

appveyor.yml

@@ -125,17 +125,6 @@ after_build:
           Copy-Item -path "$CMAKE_SOURCE_DIR/license.txt" -destination $RELEASE_DIST
           Copy-Item -path "$CMAKE_SOURCE_DIR/README.md" -destination $RELEASE_DIST
 
-          # copy all the dll dependencies to the release folder
-          . "./.appveyor/UtilityFunctions.ps1"
-          $DLLSearchPath = "C:\msys64\mingw64\bin;$env:PATH"
-          $MingwDLLs = RecursivelyGetDeps $DLLSearchPath "$RELEASE_DIST\yuzu.exe"
-          $MingwDLLs += RecursivelyGetDeps $DLLSearchPath  "$RELEASE_DIST\yuzu_cmd.exe"
-          Write-Host "Detected the following dependencies:"
-          Write-Host $MingwDLLs
-          foreach ($file in $MingwDLLs) {
-            Copy-Item -path "$file" -force -destination "$RELEASE_DIST"
-          }
-
           # copy the qt windows plugin dll to platforms
           Copy-Item -path "C:/msys64/mingw64/share/qt5/plugins/platforms/qwindows.dll" -force -destination "$RELEASE_DIST/platforms"
 
@@ -145,6 +134,18 @@ after_build:
           # copy the qt jpeg imageformat dll to platforms
           Copy-Item -path "C:/msys64/mingw64/share/qt5/plugins/imageformats/qjpeg.dll" -force -destination "$RELEASE_DIST/imageformats"
 
+          # copy all the dll dependencies to the release folder
+          . "./.appveyor/UtilityFunctions.ps1"
+          $DLLSearchPath = "C:\msys64\mingw64\bin;$env:PATH"
+          $MingwDLLs = RecursivelyGetDeps $DLLSearchPath "$RELEASE_DIST\yuzu.exe"
+          $MingwDLLs += RecursivelyGetDeps $DLLSearchPath  "$RELEASE_DIST\yuzu_cmd.exe"
+          $MingwDLLs += RecursivelyGetDeps $DLLSearchPath  "$RELEASE_DIST\imageformats\qjpeg.dll"
+          Write-Host "Detected the following dependencies:"
+          Write-Host $MingwDLLs
+          foreach ($file in $MingwDLLs) {
+            Copy-Item -path "$file" -force -destination "$RELEASE_DIST"
+          }
+
           7z a -tzip $MINGW_BUILD_ZIP $RELEASE_DIST\*
           7z a $MINGW_SEVENZIP $RELEASE_DIST
         }

externals/cmake-modules/GetGitRevisionDescription.cmake.in

@@ -33,6 +33,10 @@ else()
 endif()
 
 if(NOT HEAD_HASH)
-	file(READ "@GIT_DATA@/head-ref" HEAD_HASH LIMIT 1024)
-	string(STRIP "${HEAD_HASH}" HEAD_HASH)
+	if(EXISTS "@GIT_DATA@/head-ref")
+		file(READ "@GIT_DATA@/head-ref" HEAD_HASH LIMIT 1024)
+		string(STRIP "${HEAD_HASH}" HEAD_HASH)
+	else()
+		set(HEAD_HASH "Unknown")
+	endif()
 endif()

src/audio_core/audio_renderer.h

@@ -143,7 +143,7 @@ struct AuxInfo {
     std::array<u8, 24> output_mix_buffers;
     u32_le mix_buffer_count;
     u32_le sample_rate; // Stored in the aux buffer currently
-    u32_le sampe_count;
+    u32_le sample_count;
     u64_le send_buffer_info;
     u64_le send_buffer_base;
 

src/common/bit_field.h

@@ -117,21 +117,21 @@ private:
     // We don't delete it because we want BitField to be trivially copyable.
     constexpr BitField& operator=(const BitField&) = default;
 
-    // StorageType is T for non-enum types and the underlying type of T if
+    // UnderlyingType is T for non-enum types and the underlying type of T if
     // T is an enumeration. Note that T is wrapped within an enable_if in the
     // former case to workaround compile errors which arise when using
     // std::underlying_type<T>::type directly.
-    using StorageType = typename std::conditional_t<std::is_enum<T>::value, std::underlying_type<T>,
-                                                    std::enable_if<true, T>>::type;
+    using UnderlyingType = typename std::conditional_t<std::is_enum_v<T>, std::underlying_type<T>,
+                                                       std::enable_if<true, T>>::type;
 
-    // Unsigned version of StorageType
-    using StorageTypeU = std::make_unsigned_t<StorageType>;
+    // We store the value as the unsigned type to avoid undefined behaviour on value shifting
+    using StorageType = std::make_unsigned_t<UnderlyingType>;
 
 public:
     /// Constants to allow limited introspection of fields if needed
     static constexpr std::size_t position = Position;
     static constexpr std::size_t bits = Bits;
-    static constexpr StorageType mask = (((StorageTypeU)~0) >> (8 * sizeof(T) - bits)) << position;
+    static constexpr StorageType mask = (((StorageType)~0) >> (8 * sizeof(T) - bits)) << position;
 
     /**
      * Formats a value by masking and shifting it according to the field parameters. A value
@@ -148,11 +148,12 @@ public:
      * union in a constexpr context.
      */
     static constexpr FORCE_INLINE T ExtractValue(const StorageType& storage) {
-        if (std::numeric_limits<T>::is_signed) {
+        if constexpr (std::numeric_limits<UnderlyingType>::is_signed) {
             std::size_t shift = 8 * sizeof(T) - bits;
-            return (T)((storage << (shift - position)) >> shift);
+            return static_cast<T>(static_cast<UnderlyingType>(storage << (shift - position)) >>
+                                  shift);
         } else {
-            return (T)((storage & mask) >> position);
+            return static_cast<T>((storage & mask) >> position);
         }
     }
 

src/common/string_util.cpp

@@ -4,11 +4,10 @@
 
 #include <algorithm>
 #include <cctype>
-#include <cerrno>
 #include <codecvt>
-#include <cstdio>
 #include <cstdlib>
-#include <cstring>
+#include <locale>
+#include <sstream>
 #include "common/common_paths.h"
 #include "common/logging/log.h"
 #include "common/string_util.h"
@@ -33,24 +32,6 @@ std::string ToUpper(std::string str) {
     return str;
 }
 
-// For Debugging. Read out an u8 array.
-std::string ArrayToString(const u8* data, std::size_t size, int line_len, bool spaces) {
-    std::ostringstream oss;
-    oss << std::setfill('0') << std::hex;
-
-    for (int line = 0; size; ++data, --size) {
-        oss << std::setw(2) << (int)*data;
-
-        if (line_len == ++line) {
-            oss << '\n';
-            line = 0;
-        } else if (spaces)
-            oss << ' ';
-    }
-
-    return oss.str();
-}
-
 std::string StringFromBuffer(const std::vector<u8>& data) {
     return std::string(data.begin(), std::find(data.begin(), data.end(), '\0'));
 }
@@ -75,40 +56,6 @@ std::string StripQuotes(const std::string& s) {
         return s;
 }
 
-bool TryParse(const std::string& str, u32* const output) {
-    char* endptr = nullptr;
-
-    // Reset errno to a value other than ERANGE
-    errno = 0;
-
-    unsigned long value = strtoul(str.c_str(), &endptr, 0);
-
-    if (!endptr || *endptr)
-        return false;
-
-    if (errno == ERANGE)
-        return false;
-
-#if ULONG_MAX > UINT_MAX
-    if (value >= 0x100000000ull && value <= 0xFFFFFFFF00000000ull)
-        return false;
-#endif
-
-    *output = static_cast<u32>(value);
-    return true;
-}
-
-bool TryParse(const std::string& str, bool* const output) {
-    if ("1" == str || "true" == ToLower(str))
-        *output = true;
-    else if ("0" == str || "false" == ToLower(str))
-        *output = false;
-    else
-        return false;
-
-    return true;
-}
-
 std::string StringFromBool(bool value) {
     return value ? "True" : "False";
 }

src/common/string_util.h

@@ -5,8 +5,6 @@
 #pragma once
 
 #include <cstddef>
-#include <iomanip>
-#include <sstream>
 #include <string>
 #include <vector>
 #include "common/common_types.h"
@@ -19,44 +17,13 @@ std::string ToLower(std::string str);
 /// Make a string uppercase
 std::string ToUpper(std::string str);
 
-std::string ArrayToString(const u8* data, std::size_t size, int line_len = 20, bool spaces = true);
-
 std::string StringFromBuffer(const std::vector<u8>& data);
 
 std::string StripSpaces(const std::string& s);
 std::string StripQuotes(const std::string& s);
 
-// Thousand separator. Turns 12345678 into 12,345,678
-template <typename I>
-std::string ThousandSeparate(I value, int spaces = 0) {
-    std::ostringstream oss;
-
-// std::locale("") seems to be broken on many platforms
-#if defined _WIN32 || (defined __linux__ && !defined __clang__)
-    oss.imbue(std::locale(""));
-#endif
-    oss << std::setw(spaces) << value;
-
-    return oss.str();
-}
-
 std::string StringFromBool(bool value);
 
-bool TryParse(const std::string& str, bool* output);
-bool TryParse(const std::string& str, u32* output);
-
-template <typename N>
-static bool TryParse(const std::string& str, N* const output) {
-    std::istringstream iss(str);
-
-    N tmp = 0;
-    if (iss >> tmp) {
-        *output = tmp;
-        return true;
-    } else
-        return false;
-}
-
 std::string TabsToSpaces(int tab_size, std::string in);
 
 void SplitString(const std::string& str, char delim, std::vector<std::string>& output);

src/core/file_sys/bis_factory.cpp

@@ -8,8 +8,9 @@
 
 namespace FileSys {
 
-BISFactory::BISFactory(VirtualDir nand_root_, VirtualDir load_root_)
+BISFactory::BISFactory(VirtualDir nand_root_, VirtualDir load_root_, VirtualDir dump_root_)
     : nand_root(std::move(nand_root_)), load_root(std::move(load_root_)),
+      dump_root(std::move(dump_root_)),
       sysnand_cache(std::make_unique<RegisteredCache>(
           GetOrCreateDirectoryRelative(nand_root, "/system/Contents/registered"))),
       usrnand_cache(std::make_unique<RegisteredCache>(
@@ -32,4 +33,10 @@ VirtualDir BISFactory::GetModificationLoadRoot(u64 title_id) const {
     return GetOrCreateDirectoryRelative(load_root, fmt::format("/{:016X}", title_id));
 }
 
+VirtualDir BISFactory::GetModificationDumpRoot(u64 title_id) const {
+    if (title_id == 0)
+        return nullptr;
+    return GetOrCreateDirectoryRelative(dump_root, fmt::format("/{:016X}", title_id));
+}
+
 } // namespace FileSys

src/core/file_sys/bis_factory.h

@@ -17,17 +17,19 @@ class RegisteredCache;
 /// registered caches.
 class BISFactory {
 public:
-    explicit BISFactory(VirtualDir nand_root, VirtualDir load_root);
+    explicit BISFactory(VirtualDir nand_root, VirtualDir load_root, VirtualDir dump_root);
     ~BISFactory();
 
     RegisteredCache* GetSystemNANDContents() const;
     RegisteredCache* GetUserNANDContents() const;
 
     VirtualDir GetModificationLoadRoot(u64 title_id) const;
+    VirtualDir GetModificationDumpRoot(u64 title_id) const;
 
 private:
     VirtualDir nand_root;
     VirtualDir load_root;
+    VirtualDir dump_root;
 
     std::unique_ptr<RegisteredCache> sysnand_cache;
     std::unique_ptr<RegisteredCache> usrnand_cache;

src/core/file_sys/card_image.cpp

@@ -176,7 +176,7 @@ Loader::ResultStatus XCI::AddNCAFromPartition(XCIPartition part) {
     for (const VirtualFile& file : partitions[static_cast<std::size_t>(part)]->GetFiles()) {
         if (file->GetExtension() != "nca")
             continue;
-        auto nca = std::make_shared<NCA>(file);
+        auto nca = std::make_shared<NCA>(file, nullptr, 0, keys);
         // TODO(DarkLordZach): Add proper Rev1+ Support
         if (nca->IsUpdate())
             continue;

src/core/file_sys/card_image.h

@@ -9,6 +9,7 @@
 #include <vector>
 #include "common/common_types.h"
 #include "common/swap.h"
+#include "core/crypto/key_manager.h"
 #include "core/file_sys/vfs.h"
 
 namespace Loader {
@@ -107,5 +108,7 @@ private:
     std::shared_ptr<NSP> secure_partition;
     std::shared_ptr<NCA> program;
     std::vector<std::shared_ptr<NCA>> ncas;
+
+    Core::Crypto::KeyManager keys;
 };
 } // namespace FileSys

src/core/file_sys/content_archive.cpp

@@ -101,8 +101,9 @@ static bool IsValidNCA(const NCAHeader& header) {
     return header.magic == Common::MakeMagic('N', 'C', 'A', '3');
 }
 
-NCA::NCA(VirtualFile file_, VirtualFile bktr_base_romfs_, u64 bktr_base_ivfc_offset)
-    : file(std::move(file_)), bktr_base_romfs(std::move(bktr_base_romfs_)) {
+NCA::NCA(VirtualFile file_, VirtualFile bktr_base_romfs_, u64 bktr_base_ivfc_offset,
+         Core::Crypto::KeyManager keys_)
+    : file(std::move(file_)), bktr_base_romfs(std::move(bktr_base_romfs_)), keys(std::move(keys_)) {
     if (file == nullptr) {
         status = Loader::ResultStatus::ErrorNullFile;
         return;

src/core/file_sys/content_archive.h

@@ -79,7 +79,8 @@ inline bool IsDirectoryExeFS(const std::shared_ptr<VfsDirectory>& pfs) {
 class NCA : public ReadOnlyVfsDirectory {
 public:
     explicit NCA(VirtualFile file, VirtualFile bktr_base_romfs = nullptr,
-                 u64 bktr_base_ivfc_offset = 0);
+                 u64 bktr_base_ivfc_offset = 0,
+                 Core::Crypto::KeyManager keys = Core::Crypto::KeyManager());
     ~NCA() override;
 
     Loader::ResultStatus GetStatus() const;

src/core/file_sys/control_metadata.cpp

@@ -66,4 +66,10 @@ std::string NACP::GetVersionString() const {
     return Common::StringFromFixedZeroTerminatedBuffer(raw->version_string.data(),
                                                        raw->version_string.size());
 }
+
+std::vector<u8> NACP::GetRawBytes() const {
+    std::vector<u8> out(sizeof(RawNACP));
+    std::memcpy(out.data(), raw.get(), sizeof(RawNACP));
+    return out;
+}
 } // namespace FileSys

src/core/file_sys/control_metadata.h

@@ -81,6 +81,7 @@ public:
     u64 GetTitleId() const;
     u64 GetDLCBaseTitleId() const;
     std::string GetVersionString() const;
+    std::vector<u8> GetRawBytes() const;
 
 private:
     std::unique_ptr<RawNACP> raw;

src/core/file_sys/errors.h

@@ -8,25 +8,10 @@
 
 namespace FileSys {
 
-namespace ErrCodes {
-enum {
-    NotFound = 1,
-    TitleNotFound = 1002,
-    SdCardNotFound = 2001,
-    RomFSNotFound = 2520,
-};
-}
-
-constexpr ResultCode ERROR_PATH_NOT_FOUND(ErrorModule::FS, ErrCodes::NotFound);
-
-// TODO(bunnei): Replace these with correct errors for Switch OS
-constexpr ResultCode ERROR_INVALID_PATH(-1);
-constexpr ResultCode ERROR_UNSUPPORTED_OPEN_FLAGS(-1);
-constexpr ResultCode ERROR_INVALID_OPEN_FLAGS(-1);
-constexpr ResultCode ERROR_FILE_NOT_FOUND(-1);
-constexpr ResultCode ERROR_UNEXPECTED_FILE_OR_DIRECTORY(-1);
-constexpr ResultCode ERROR_DIRECTORY_ALREADY_EXISTS(-1);
-constexpr ResultCode ERROR_FILE_ALREADY_EXISTS(-1);
-constexpr ResultCode ERROR_DIRECTORY_NOT_EMPTY(-1);
+constexpr ResultCode ERROR_PATH_NOT_FOUND{ErrorModule::FS, 1};
+constexpr ResultCode ERROR_ENTITY_NOT_FOUND{ErrorModule::FS, 1002};
+constexpr ResultCode ERROR_SD_CARD_NOT_FOUND{ErrorModule::FS, 2001};
+constexpr ResultCode ERROR_INVALID_OFFSET{ErrorModule::FS, 6061};
+constexpr ResultCode ERROR_INVALID_SIZE{ErrorModule::FS, 6062};
 
 } // namespace FileSys

src/core/file_sys/patch_manager.cpp

@@ -19,6 +19,7 @@
 #include "core/file_sys/vfs_vector.h"
 #include "core/hle/service/filesystem/filesystem.h"
 #include "core/loader/loader.h"
+#include "core/settings.h"
 
 namespace FileSys {
 
@@ -119,6 +120,18 @@ std::vector<u8> PatchManager::PatchNSO(const std::vector<u8>& nso) const {
     const auto build_id_raw = Common::HexArrayToString(header.build_id);
     const auto build_id = build_id_raw.substr(0, build_id_raw.find_last_not_of('0') + 1);
 
+    if (Settings::values.dump_nso) {
+        LOG_INFO(Loader, "Dumping NSO for build_id={}, title_id={:016X}", build_id, title_id);
+        const auto dump_dir = Service::FileSystem::GetModificationDumpRoot(title_id);
+        if (dump_dir != nullptr) {
+            const auto nso_dir = GetOrCreateDirectoryRelative(dump_dir, "/nso");
+            const auto file = nso_dir->CreateFile(fmt::format("{}.nso", build_id));
+
+            file->Resize(nso.size());
+            file->WriteBytes(nso);
+        }
+    }
+
     LOG_INFO(Loader, "Patching NSO for build_id={}", build_id);
 
     const auto load_dir = Service::FileSystem::GetModificationLoadRoot(title_id);

src/core/file_sys/registered_cache.cpp

@@ -106,9 +106,12 @@ static ContentRecordType GetCRTypeFromNCAType(NCAContentType type) {
 
 VirtualFile RegisteredCache::OpenFileOrDirectoryConcat(const VirtualDir& dir,
                                                        std::string_view path) const {
-    if (dir->GetFileRelative(path) != nullptr)
-        return dir->GetFileRelative(path);
-    if (dir->GetDirectoryRelative(path) != nullptr) {
+    const auto file = dir->GetFileRelative(path);
+    if (file != nullptr)
+        return file;
+
+    const auto nca_dir = dir->GetDirectoryRelative(path);
+    if (nca_dir != nullptr) {
         const auto nca_dir = dir->GetDirectoryRelative(path);
         VirtualFile file = nullptr;
 
@@ -225,7 +228,7 @@ void RegisteredCache::ProcessFiles(const std::vector<NcaID>& ids) {
 
         if (file == nullptr)
             continue;
-        const auto nca = std::make_shared<NCA>(parser(file, id));
+        const auto nca = std::make_shared<NCA>(parser(file, id), nullptr, 0, keys);
         if (nca->GetStatus() != Loader::ResultStatus::Success ||
             nca->GetType() != NCAContentType::Meta) {
             continue;
@@ -315,7 +318,7 @@ std::unique_ptr<NCA> RegisteredCache::GetEntry(u64 title_id, ContentRecordType t
     const auto raw = GetEntryRaw(title_id, type);
     if (raw == nullptr)
         return nullptr;
-    return std::make_unique<NCA>(raw);
+    return std::make_unique<NCA>(raw, nullptr, 0, keys);
 }
 
 std::unique_ptr<NCA> RegisteredCache::GetEntry(RegisteredCacheEntry entry) const {

src/core/file_sys/registered_cache.h

@@ -12,6 +12,7 @@
 #include <vector>
 #include <boost/container/flat_map.hpp>
 #include "common/common_types.h"
+#include "core/crypto/key_manager.h"
 #include "core/file_sys/vfs.h"
 
 namespace FileSys {
@@ -133,6 +134,8 @@ private:
 
     VirtualDir dir;
     RegisteredCacheParsingFunction parser;
+    Core::Crypto::KeyManager keys;
+
     // maps tid -> NcaID of meta
     boost::container::flat_map<u64, NcaID> meta_id;
     // maps tid -> meta

src/core/file_sys/savedata_factory.cpp

@@ -83,6 +83,24 @@ ResultVal<VirtualDir> SaveDataFactory::Open(SaveDataSpaceId space, SaveDataDescr
     return MakeResult<VirtualDir>(std::move(out));
 }
 
+VirtualDir SaveDataFactory::GetSaveDataSpaceDirectory(SaveDataSpaceId space) const {
+    return dir->GetDirectoryRelative(GetSaveDataSpaceIdPath(space));
+}
+
+std::string SaveDataFactory::GetSaveDataSpaceIdPath(SaveDataSpaceId space) {
+    switch (space) {
+    case SaveDataSpaceId::NandSystem:
+        return "/system/";
+    case SaveDataSpaceId::NandUser:
+        return "/user/";
+    case SaveDataSpaceId::TemporaryStorage:
+        return "/temp/";
+    default:
+        ASSERT_MSG(false, "Unrecognized SaveDataSpaceId: {:02X}", static_cast<u8>(space));
+        return "/unrecognized/"; ///< To prevent corruption when ignoring asserts.
+    }
+}
+
 std::string SaveDataFactory::GetFullPath(SaveDataSpaceId space, SaveDataType type, u64 title_id,
                                          u128 user_id, u64 save_id) {
     // According to switchbrew, if a save is of type SaveData and the title id field is 0, it should
@@ -90,21 +108,7 @@ std::string SaveDataFactory::GetFullPath(SaveDataSpaceId space, SaveDataType typ
     if (type == SaveDataType::SaveData && title_id == 0)
         title_id = Core::CurrentProcess()->GetTitleID();
 
-    std::string out;
-
-    switch (space) {
-    case SaveDataSpaceId::NandSystem:
-        out = "/system/";
-        break;
-    case SaveDataSpaceId::NandUser:
-        out = "/user/";
-        break;
-    case SaveDataSpaceId::TemporaryStorage:
-        out = "/temp/";
-        break;
-    default:
-        ASSERT_MSG(false, "Unrecognized SaveDataSpaceId: {:02X}", static_cast<u8>(space));
-    }
+    std::string out = GetSaveDataSpaceIdPath(space);
 
     switch (type) {
     case SaveDataType::SystemSaveData:

src/core/file_sys/savedata_factory.h

@@ -52,6 +52,9 @@ public:
 
     ResultVal<VirtualDir> Open(SaveDataSpaceId space, SaveDataDescriptor meta);
 
+    VirtualDir GetSaveDataSpaceDirectory(SaveDataSpaceId space) const;
+
+    static std::string GetSaveDataSpaceIdPath(SaveDataSpaceId space);
     static std::string GetFullPath(SaveDataSpaceId space, SaveDataType type, u64 title_id,
                                    u128 user_id, u64 save_id);
 

src/core/file_sys/submission_package.cpp

@@ -252,7 +252,7 @@ void NSP::ReadNCAs(const std::vector<VirtualFile>& files) {
                     continue;
                 }
 
-                auto next_nca = std::make_shared<NCA>(next_file);
+                auto next_nca = std::make_shared<NCA>(next_file, nullptr, 0, keys);
                 if (next_nca->GetType() == NCAContentType::Program)
                     program_status[cnmt.GetTitleID()] = next_nca->GetStatus();
                 if (next_nca->GetStatus() == Loader::ResultStatus::Success ||

src/core/file_sys/submission_package.h

@@ -70,6 +70,8 @@ private:
     std::map<u64, std::map<ContentRecordType, std::shared_ptr<NCA>>> ncas;
     std::vector<VirtualFile> ticket_files;
 
+    Core::Crypto::KeyManager keys;
+
     VirtualFile romfs;
     VirtualDir exefs;
 };

src/core/frontend/input.h

@@ -132,4 +132,11 @@ using MotionDevice = InputDevice<std::tuple<Math::Vec3<float>, Math::Vec3<float>
  */
 using TouchDevice = InputDevice<std::tuple<float, float, bool>>;
 
+/**
+ * A mouse device is an input device that returns a tuple of two floats and four ints.
+ * The first two floats are X and Y device coordinates of the mouse (from 0-1).
+ * The s32s are the mouse wheel.
+ */
+using MouseDevice = InputDevice<std::tuple<float, float, s32, s32>>;
+
 } // namespace Input

src/core/hle/kernel/errors.h

@@ -8,58 +8,28 @@
 
 namespace Kernel {
 
-namespace ErrCodes {
-enum {
-    // Confirmed Switch OS error codes
-    MaxConnectionsReached = 7,
-    InvalidSize = 101,
-    InvalidAddress = 102,
-    HandleTableFull = 105,
-    InvalidMemoryState = 106,
-    InvalidMemoryPermissions = 108,
-    InvalidMemoryRange = 110,
-    InvalidThreadPriority = 112,
-    InvalidProcessorId = 113,
-    InvalidHandle = 114,
-    InvalidPointer = 115,
-    InvalidCombination = 116,
-    Timeout = 117,
-    SynchronizationCanceled = 118,
-    TooLarge = 119,
-    InvalidEnumValue = 120,
-    NoSuchEntry = 121,
-    AlreadyRegistered = 122,
-    SessionClosed = 123,
-    InvalidState = 125,
-    ResourceLimitExceeded = 132,
-};
-}
+// Confirmed Switch kernel error codes
 
-// WARNING: The kernel is quite inconsistent in it's usage of errors code. Make sure to always
-// double check that the code matches before re-using the constant.
-
-constexpr ResultCode ERR_HANDLE_TABLE_FULL(ErrorModule::Kernel, ErrCodes::HandleTableFull);
-constexpr ResultCode ERR_SESSION_CLOSED_BY_REMOTE(ErrorModule::Kernel, ErrCodes::SessionClosed);
-constexpr ResultCode ERR_PORT_NAME_TOO_LONG(ErrorModule::Kernel, ErrCodes::TooLarge);
-constexpr ResultCode ERR_MAX_CONNECTIONS_REACHED(ErrorModule::Kernel,
-                                                 ErrCodes::MaxConnectionsReached);
-constexpr ResultCode ERR_INVALID_ENUM_VALUE(ErrorModule::Kernel, ErrCodes::InvalidEnumValue);
-constexpr ResultCode ERR_INVALID_COMBINATION_KERNEL(ErrorModule::Kernel,
-                                                    ErrCodes::InvalidCombination);
-constexpr ResultCode ERR_INVALID_ADDRESS(ErrorModule::Kernel, ErrCodes::InvalidAddress);
-constexpr ResultCode ERR_INVALID_ADDRESS_STATE(ErrorModule::Kernel, ErrCodes::InvalidMemoryState);
-constexpr ResultCode ERR_INVALID_MEMORY_PERMISSIONS(ErrorModule::Kernel,
-                                                    ErrCodes::InvalidMemoryPermissions);
-constexpr ResultCode ERR_INVALID_MEMORY_RANGE(ErrorModule::Kernel, ErrCodes::InvalidMemoryRange);
-constexpr ResultCode ERR_INVALID_HANDLE(ErrorModule::Kernel, ErrCodes::InvalidHandle);
-constexpr ResultCode ERR_INVALID_PROCESSOR_ID(ErrorModule::Kernel, ErrCodes::InvalidProcessorId);
-constexpr ResultCode ERR_INVALID_SIZE(ErrorModule::Kernel, ErrCodes::InvalidSize);
-constexpr ResultCode ERR_ALREADY_REGISTERED(ErrorModule::Kernel, ErrCodes::AlreadyRegistered);
-constexpr ResultCode ERR_INVALID_STATE(ErrorModule::Kernel, ErrCodes::InvalidState);
-constexpr ResultCode ERR_INVALID_THREAD_PRIORITY(ErrorModule::Kernel,
-                                                 ErrCodes::InvalidThreadPriority);
-constexpr ResultCode ERR_INVALID_POINTER(ErrorModule::Kernel, ErrCodes::InvalidPointer);
-constexpr ResultCode ERR_NOT_FOUND(ErrorModule::Kernel, ErrCodes::NoSuchEntry);
-constexpr ResultCode RESULT_TIMEOUT(ErrorModule::Kernel, ErrCodes::Timeout);
+constexpr ResultCode ERR_MAX_CONNECTIONS_REACHED{ErrorModule::Kernel, 7};
+constexpr ResultCode ERR_INVALID_SIZE{ErrorModule::Kernel, 101};
+constexpr ResultCode ERR_INVALID_ADDRESS{ErrorModule::Kernel, 102};
+constexpr ResultCode ERR_HANDLE_TABLE_FULL{ErrorModule::Kernel, 105};
+constexpr ResultCode ERR_INVALID_ADDRESS_STATE{ErrorModule::Kernel, 106};
+constexpr ResultCode ERR_INVALID_MEMORY_PERMISSIONS{ErrorModule::Kernel, 108};
+constexpr ResultCode ERR_INVALID_MEMORY_RANGE{ErrorModule::Kernel, 110};
+constexpr ResultCode ERR_INVALID_PROCESSOR_ID{ErrorModule::Kernel, 113};
+constexpr ResultCode ERR_INVALID_THREAD_PRIORITY{ErrorModule::Kernel, 112};
+constexpr ResultCode ERR_INVALID_HANDLE{ErrorModule::Kernel, 114};
+constexpr ResultCode ERR_INVALID_POINTER{ErrorModule::Kernel, 115};
+constexpr ResultCode ERR_INVALID_COMBINATION{ErrorModule::Kernel, 116};
+constexpr ResultCode RESULT_TIMEOUT{ErrorModule::Kernel, 117};
+constexpr ResultCode ERR_SYNCHRONIZATION_CANCELED{ErrorModule::Kernel, 118};
+constexpr ResultCode ERR_OUT_OF_RANGE{ErrorModule::Kernel, 119};
+constexpr ResultCode ERR_INVALID_ENUM_VALUE{ErrorModule::Kernel, 120};
+constexpr ResultCode ERR_NOT_FOUND{ErrorModule::Kernel, 121};
+constexpr ResultCode ERR_ALREADY_REGISTERED{ErrorModule::Kernel, 122};
+constexpr ResultCode ERR_SESSION_CLOSED_BY_REMOTE{ErrorModule::Kernel, 123};
+constexpr ResultCode ERR_INVALID_STATE{ErrorModule::Kernel, 125};
+constexpr ResultCode ERR_RESOURCE_LIMIT_EXCEEDED{ErrorModule::Kernel, 132};
 
 } // namespace Kernel

src/core/hle/kernel/process.cpp

@@ -5,11 +5,9 @@
 #include <algorithm>
 #include <memory>
 #include "common/assert.h"
-#include "common/common_funcs.h"
 #include "common/logging/log.h"
 #include "core/core.h"
 #include "core/file_sys/program_metadata.h"
-#include "core/hle/kernel/errors.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/kernel/process.h"
 #include "core/hle/kernel/resource_limit.h"
@@ -17,6 +15,7 @@
 #include "core/hle/kernel/thread.h"
 #include "core/hle/kernel/vm_manager.h"
 #include "core/memory.h"
+#include "core/settings.h"
 
 namespace Kernel {
 
@@ -35,6 +34,11 @@ SharedPtr<Process> Process::Create(KernelCore& kernel, std::string&& name) {
     process->process_id = kernel.CreateNewProcessID();
     process->svc_access_mask.set();
 
+    std::mt19937 rng(Settings::values.rng_seed.value_or(0));
+    std::uniform_int_distribution<u64> distribution;
+    std::generate(process->random_entropy.begin(), process->random_entropy.end(),
+                  [&] { return distribution(rng); });
+
     kernel.AppendNewProcess(process);
     return process;
 }
@@ -241,83 +245,15 @@ void Process::LoadModule(CodeSet module_, VAddr base_addr) {
 }
 
 ResultVal<VAddr> Process::HeapAllocate(VAddr target, u64 size, VMAPermission perms) {
-    if (target < vm_manager.GetHeapRegionBaseAddress() ||
-        target + size > vm_manager.GetHeapRegionEndAddress() || target + size < target) {
-        return ERR_INVALID_ADDRESS;
-    }
-
-    if (heap_memory == nullptr) {
-        // Initialize heap
-        heap_memory = std::make_shared<std::vector<u8>>();
-        heap_start = heap_end = target;
-    } else {
-        vm_manager.UnmapRange(heap_start, heap_end - heap_start);
-    }
-
-    // If necessary, expand backing vector to cover new heap extents.
-    if (target < heap_start) {
-        heap_memory->insert(begin(*heap_memory), heap_start - target, 0);
-        heap_start = target;
-        vm_manager.RefreshMemoryBlockMappings(heap_memory.get());
-    }
-    if (target + size > heap_end) {
-        heap_memory->insert(end(*heap_memory), (target + size) - heap_end, 0);
-        heap_end = target + size;
-        vm_manager.RefreshMemoryBlockMappings(heap_memory.get());
-    }
-    ASSERT(heap_end - heap_start == heap_memory->size());
-
-    CASCADE_RESULT(auto vma, vm_manager.MapMemoryBlock(target, heap_memory, target - heap_start,
-                                                       size, MemoryState::Heap));
-    vm_manager.Reprotect(vma, perms);
-
-    heap_used = size;
-
-    return MakeResult<VAddr>(heap_end - size);
+    return vm_manager.HeapAllocate(target, size, perms);
 }
 
 ResultCode Process::HeapFree(VAddr target, u32 size) {
-    if (target < vm_manager.GetHeapRegionBaseAddress() ||
-        target + size > vm_manager.GetHeapRegionEndAddress() || target + size < target) {
-        return ERR_INVALID_ADDRESS;
-    }
-
-    if (size == 0) {
-        return RESULT_SUCCESS;
-    }
-
-    ResultCode result = vm_manager.UnmapRange(target, size);
-    if (result.IsError())
-        return result;
-
-    heap_used -= size;
-
-    return RESULT_SUCCESS;
+    return vm_manager.HeapFree(target, size);
 }
 
-ResultCode Process::MirrorMemory(VAddr dst_addr, VAddr src_addr, u64 size) {
-    auto vma = vm_manager.FindVMA(src_addr);
-
-    ASSERT_MSG(vma != vm_manager.vma_map.end(), "Invalid memory address");
-    ASSERT_MSG(vma->second.backing_block, "Backing block doesn't exist for address");
-
-    // The returned VMA might be a bigger one encompassing the desired address.
-    auto vma_offset = src_addr - vma->first;
-    ASSERT_MSG(vma_offset + size <= vma->second.size,
-               "Shared memory exceeds bounds of mapped block");
-
-    const std::shared_ptr<std::vector<u8>>& backing_block = vma->second.backing_block;
-    std::size_t backing_block_offset = vma->second.offset + vma_offset;
-
-    CASCADE_RESULT(auto new_vma,
-                   vm_manager.MapMemoryBlock(dst_addr, backing_block, backing_block_offset, size,
-                                             MemoryState::Mapped));
-    // Protect mirror with permissions from old region
-    vm_manager.Reprotect(new_vma, vma->second.permissions);
-    // Remove permissions from old region
-    vm_manager.Reprotect(vma, VMAPermission::None);
-
-    return RESULT_SUCCESS;
+ResultCode Process::MirrorMemory(VAddr dst_addr, VAddr src_addr, u64 size, MemoryState state) {
+    return vm_manager.MirrorMemory(dst_addr, src_addr, size, state);
 }
 
 ResultCode Process::UnmapMemory(VAddr dst_addr, VAddr /*src_addr*/, u64 size) {

src/core/hle/kernel/process.h

@@ -8,6 +8,7 @@
 #include <bitset>
 #include <cstddef>
 #include <memory>
+#include <random>
 #include <string>
 #include <vector>
 #include <boost/container/static_vector.hpp>
@@ -119,6 +120,8 @@ struct CodeSet final {
 
 class Process final : public Object {
 public:
+    static constexpr std::size_t RANDOM_ENTROPY_SIZE = 4;
+
     static SharedPtr<Process> Create(KernelCore& kernel, std::string&& name);
 
     std::string GetTypeName() const override {
@@ -212,6 +215,11 @@ public:
         total_process_running_time_ticks += ticks;
     }
 
+    /// Gets 8 bytes of random data for svcGetInfo RandomEntropy
+    u64 GetRandomEntropy(std::size_t index) const {
+        return random_entropy.at(index);
+    }
+
     /**
      * Loads process-specifics configuration info with metadata provided
      * by an executable.
@@ -251,7 +259,8 @@ public:
     ResultVal<VAddr> HeapAllocate(VAddr target, u64 size, VMAPermission perms);
     ResultCode HeapFree(VAddr target, u32 size);
 
-    ResultCode MirrorMemory(VAddr dst_addr, VAddr src_addr, u64 size);
+    ResultCode MirrorMemory(VAddr dst_addr, VAddr src_addr, u64 size,
+                            MemoryState state = MemoryState::Mapped);
 
     ResultCode UnmapMemory(VAddr dst_addr, VAddr src_addr, u64 size);
 
@@ -292,17 +301,6 @@ private:
     u32 allowed_thread_priority_mask = 0xFFFFFFFF;
     u32 is_virtual_address_memory_enabled = 0;
 
-    // Memory used to back the allocations in the regular heap. A single vector is used to cover
-    // the entire virtual address space extents that bound the allocations, including any holes.
-    // This makes deallocation and reallocation of holes fast and keeps process memory contiguous
-    // in the emulator address space, allowing Memory::GetPointer to be reasonably safe.
-    std::shared_ptr<std::vector<u8>> heap_memory;
-
-    // The left/right bounds of the address space covered by heap_memory.
-    VAddr heap_start = 0;
-    VAddr heap_end = 0;
-    u64 heap_used = 0;
-
     /// The Thread Local Storage area is allocated as processes create threads,
     /// each TLS area is 0x200 bytes, so one page (0x1000) is split up in 8 parts, and each part
     /// holds the TLS for a specific thread. This vector contains which parts are in use for each
@@ -321,6 +319,9 @@ private:
     /// Per-process handle table for storing created object handles in.
     HandleTable handle_table;
 
+    /// Random values for svcGetInfo RandomEntropy
+    std::array<u64, RANDOM_ENTROPY_SIZE> random_entropy;
+
     std::string name;
 };
 

src/core/hle/kernel/svc.cpp

@@ -34,6 +34,7 @@
 #include "core/hle/lock.h"
 #include "core/hle/result.h"
 #include "core/hle/service/service.h"
+#include "core/settings.h"
 
 namespace Kernel {
 namespace {
@@ -122,6 +123,48 @@ static ResultCode SetHeapSize(VAddr* heap_addr, u64 heap_size) {
     return RESULT_SUCCESS;
 }
 
+static ResultCode SetMemoryPermission(VAddr addr, u64 size, u32 prot) {
+    LOG_TRACE(Kernel_SVC, "called, addr=0x{:X}, size=0x{:X}, prot=0x{:X}", addr, size, prot);
+
+    if (!Common::Is4KBAligned(addr)) {
+        return ERR_INVALID_ADDRESS;
+    }
+
+    if (size == 0 || !Common::Is4KBAligned(size)) {
+        return ERR_INVALID_SIZE;
+    }
+
+    if (!IsValidAddressRange(addr, size)) {
+        return ERR_INVALID_ADDRESS_STATE;
+    }
+
+    const auto permission = static_cast<MemoryPermission>(prot);
+    if (permission != MemoryPermission::None && permission != MemoryPermission::Read &&
+        permission != MemoryPermission::ReadWrite) {
+        return ERR_INVALID_MEMORY_PERMISSIONS;
+    }
+
+    auto* const current_process = Core::CurrentProcess();
+    auto& vm_manager = current_process->VMManager();
+
+    if (!IsInsideAddressSpace(vm_manager, addr, size)) {
+        return ERR_INVALID_ADDRESS_STATE;
+    }
+
+    const VMManager::VMAHandle iter = vm_manager.FindVMA(addr);
+    if (iter == vm_manager.vma_map.end()) {
+        return ERR_INVALID_ADDRESS_STATE;
+    }
+
+    LOG_WARNING(Kernel_SVC, "Uniformity check on protected memory is not implemented.");
+    // TODO: Performs a uniformity check to make sure only protected memory is changed (it doesn't
+    // make sense to allow changing permissions on kernel memory itself, etc).
+
+    const auto converted_permissions = SharedMemory::ConvertPermissions(permission);
+
+    return vm_manager.ReprotectRange(addr, size, converted_permissions);
+}
+
 static ResultCode SetMemoryAttribute(VAddr addr, u64 size, u32 state0, u32 state1) {
     LOG_WARNING(Kernel_SVC,
                 "(STUBBED) called, addr=0x{:X}, size=0x{:X}, state0=0x{:X}, state1=0x{:X}", addr,
@@ -171,7 +214,7 @@ static ResultCode ConnectToNamedPort(Handle* out_handle, VAddr port_name_address
     // Read 1 char beyond the max allowed port name to detect names that are too long.
     std::string port_name = Memory::ReadCString(port_name_address, PortNameMaxLength + 1);
     if (port_name.size() > PortNameMaxLength) {
-        return ERR_PORT_NAME_TOO_LONG;
+        return ERR_OUT_OF_RANGE;
     }
 
     LOG_TRACE(Kernel_SVC, "called port_name={}", port_name);
@@ -267,8 +310,9 @@ static ResultCode WaitSynchronization(Handle* index, VAddr handles_address, u64
 
     static constexpr u64 MaxHandles = 0x40;
 
-    if (handle_count > MaxHandles)
-        return ResultCode(ErrorModule::Kernel, ErrCodes::TooLarge);
+    if (handle_count > MaxHandles) {
+        return ERR_OUT_OF_RANGE;
+    }
 
     auto* const thread = GetCurrentThread();
 
@@ -333,8 +377,7 @@ static ResultCode CancelSynchronization(Handle thread_handle) {
     }
 
     ASSERT(thread->GetStatus() == ThreadStatus::WaitSynchAny);
-    thread->SetWaitSynchronizationResult(
-        ResultCode(ErrorModule::Kernel, ErrCodes::SynchronizationCanceled));
+    thread->SetWaitSynchronizationResult(ERR_SYNCHRONIZATION_CANCELED);
     thread->ResumeFromWait();
     return RESULT_SUCCESS;
 }
@@ -558,7 +601,16 @@ static ResultCode GetInfo(u64* result, u64 info_id, u64 handle, u64 info_sub_id)
         *result = 0;
         break;
     case GetInfoType::RandomEntropy:
-        *result = 0;
+        if (handle != 0) {
+            return ERR_INVALID_HANDLE;
+        }
+
+        if (info_sub_id >= Process::RANDOM_ENTROPY_SIZE) {
+            return ERR_INVALID_COMBINATION;
+        }
+
+        *result = current_process->GetRandomEntropy(info_sub_id);
+        return RESULT_SUCCESS;
         break;
     case GetInfoType::ASLRRegionBaseAddr:
         *result = vm_manager.GetASLRRegionBaseAddress();
@@ -591,7 +643,7 @@ static ResultCode GetInfo(u64* result, u64 info_id, u64 handle, u64 info_sub_id)
     case GetInfoType::ThreadTickCount: {
         constexpr u64 num_cpus = 4;
         if (info_sub_id != 0xFFFFFFFFFFFFFFFF && info_sub_id >= num_cpus) {
-            return ERR_INVALID_COMBINATION_KERNEL;
+            return ERR_INVALID_COMBINATION;
         }
 
         const auto thread =
@@ -1129,7 +1181,7 @@ static ResultCode CloseHandle(Handle handle) {
 
 /// Reset an event
 static ResultCode ResetSignal(Handle handle) {
-    LOG_WARNING(Kernel_SVC, "(STUBBED) called handle 0x{:08X}", handle);
+    LOG_DEBUG(Kernel_SVC, "called handle 0x{:08X}", handle);
 
     const auto& handle_table = Core::CurrentProcess()->GetHandleTable();
     auto event = handle_table.Get<Event>(handle);
@@ -1184,7 +1236,7 @@ static ResultCode SetThreadCoreMask(Handle thread_handle, u32 core, u64 mask) {
     }
 
     if (mask == 0) {
-        return ResultCode(ErrorModule::Kernel, ErrCodes::InvalidCombination);
+        return ERR_INVALID_COMBINATION;
     }
 
     /// This value is used to only change the affinity mask without changing the current ideal core.
@@ -1193,12 +1245,12 @@ static ResultCode SetThreadCoreMask(Handle thread_handle, u32 core, u64 mask) {
     if (core == OnlyChangeMask) {
         core = thread->GetIdealCore();
     } else if (core >= Core::NUM_CPU_CORES && core != static_cast<u32>(-1)) {
-        return ResultCode(ErrorModule::Kernel, ErrCodes::InvalidProcessorId);
+        return ERR_INVALID_PROCESSOR_ID;
     }
 
     // Error out if the input core isn't enabled in the input mask.
     if (core < Core::NUM_CPU_CORES && (mask & (1ull << core)) == 0) {
-        return ResultCode(ErrorModule::Kernel, ErrCodes::InvalidCombination);
+        return ERR_INVALID_COMBINATION;
     }
 
     thread->ChangeCore(core, mask);
@@ -1287,7 +1339,7 @@ struct FunctionDef {
 static const FunctionDef SVC_Table[] = {
     {0x00, nullptr, "Unknown"},
     {0x01, SvcWrap<SetHeapSize>, "SetHeapSize"},
-    {0x02, nullptr, "SetMemoryPermission"},
+    {0x02, SvcWrap<SetMemoryPermission>, "SetMemoryPermission"},
     {0x03, SvcWrap<SetMemoryAttribute>, "SetMemoryAttribute"},
     {0x04, SvcWrap<MapMemory>, "MapMemory"},
     {0x05, SvcWrap<UnmapMemory>, "UnmapMemory"},

src/core/hle/kernel/svc_wrap.h

@@ -121,6 +121,11 @@ void SvcWrap() {
     FuncReturn(func(Param(0), Param(1), Param(2)).raw);
 }
 
+template <ResultCode func(u64, u64, u32)>
+void SvcWrap() {
+    FuncReturn(func(Param(0), Param(1), static_cast<u32>(Param(2))).raw);
+}
+
 template <ResultCode func(u32, u64, u64, u32)>
 void SvcWrap() {
     FuncReturn(

src/core/hle/kernel/thread.cpp

@@ -142,36 +142,7 @@ void Thread::ResumeFromWait() {
 
     status = ThreadStatus::Ready;
 
-    std::optional<s32> new_processor_id = GetNextProcessorId(affinity_mask);
-    if (!new_processor_id) {
-        new_processor_id = processor_id;
-    }
-    if (ideal_core != -1 &&
-        Core::System::GetInstance().Scheduler(ideal_core).GetCurrentThread() == nullptr) {
-        new_processor_id = ideal_core;
-    }
-
-    ASSERT(*new_processor_id < 4);
-
-    // Add thread to new core's scheduler
-    auto* next_scheduler = &Core::System::GetInstance().Scheduler(*new_processor_id);
-
-    if (*new_processor_id != processor_id) {
-        // Remove thread from previous core's scheduler
-        scheduler->RemoveThread(this);
-        next_scheduler->AddThread(this, current_priority);
-    }
-
-    processor_id = *new_processor_id;
-
-    // If the thread was ready, unschedule from the previous core and schedule on the new core
-    scheduler->UnscheduleThread(this, current_priority);
-    next_scheduler->ScheduleThread(this, current_priority);
-
-    // Change thread's scheduler
-    scheduler = next_scheduler;
-
-    Core::System::GetInstance().CpuCore(processor_id).PrepareReschedule();
+    ChangeScheduler();
 }
 
 /**
@@ -364,42 +335,45 @@ void Thread::UpdatePriority() {
 void Thread::ChangeCore(u32 core, u64 mask) {
     ideal_core = core;
     affinity_mask = mask;
+    ChangeScheduler();
+}
 
+void Thread::ChangeScheduler() {
     if (status != ThreadStatus::Ready) {
         return;
     }
 
+    auto& system = Core::System::GetInstance();
     std::optional<s32> new_processor_id{GetNextProcessorId(affinity_mask)};
 
     if (!new_processor_id) {
         new_processor_id = processor_id;
     }
-    if (ideal_core != -1 &&
-        Core::System::GetInstance().Scheduler(ideal_core).GetCurrentThread() == nullptr) {
+    if (ideal_core != -1 && system.Scheduler(ideal_core).GetCurrentThread() == nullptr) {
         new_processor_id = ideal_core;
     }
 
     ASSERT(*new_processor_id < 4);
 
     // Add thread to new core's scheduler
-    auto* next_scheduler = &Core::System::GetInstance().Scheduler(*new_processor_id);
+    auto& next_scheduler = system.Scheduler(*new_processor_id);
 
     if (*new_processor_id != processor_id) {
         // Remove thread from previous core's scheduler
         scheduler->RemoveThread(this);
-        next_scheduler->AddThread(this, current_priority);
+        next_scheduler.AddThread(this, current_priority);
     }
 
     processor_id = *new_processor_id;
 
     // If the thread was ready, unschedule from the previous core and schedule on the new core
     scheduler->UnscheduleThread(this, current_priority);
-    next_scheduler->ScheduleThread(this, current_priority);
+    next_scheduler.ScheduleThread(this, current_priority);
 
     // Change thread's scheduler
-    scheduler = next_scheduler;
+    scheduler = &next_scheduler;
 
-    Core::System::GetInstance().CpuCore(processor_id).PrepareReschedule();
+    system.CpuCore(processor_id).PrepareReschedule();
 }
 
 bool Thread::AllWaitObjectsReady() {

src/core/hle/kernel/thread.h

@@ -374,6 +374,8 @@ private:
     explicit Thread(KernelCore& kernel);
     ~Thread() override;
 
+    void ChangeScheduler();
+
     Core::ARM_Interface::ThreadContext context{};
 
     u32 thread_id = 0;

src/core/hle/kernel/vm_manager.cpp

@@ -243,6 +243,85 @@ ResultCode VMManager::ReprotectRange(VAddr target, u64 size, VMAPermission new_p
     return RESULT_SUCCESS;
 }
 
+ResultVal<VAddr> VMManager::HeapAllocate(VAddr target, u64 size, VMAPermission perms) {
+    if (target < GetHeapRegionBaseAddress() || target + size > GetHeapRegionEndAddress() ||
+        target + size < target) {
+        return ERR_INVALID_ADDRESS;
+    }
+
+    if (heap_memory == nullptr) {
+        // Initialize heap
+        heap_memory = std::make_shared<std::vector<u8>>();
+        heap_start = heap_end = target;
+    } else {
+        UnmapRange(heap_start, heap_end - heap_start);
+    }
+
+    // If necessary, expand backing vector to cover new heap extents.
+    if (target < heap_start) {
+        heap_memory->insert(begin(*heap_memory), heap_start - target, 0);
+        heap_start = target;
+        RefreshMemoryBlockMappings(heap_memory.get());
+    }
+    if (target + size > heap_end) {
+        heap_memory->insert(end(*heap_memory), (target + size) - heap_end, 0);
+        heap_end = target + size;
+        RefreshMemoryBlockMappings(heap_memory.get());
+    }
+    ASSERT(heap_end - heap_start == heap_memory->size());
+
+    CASCADE_RESULT(auto vma, MapMemoryBlock(target, heap_memory, target - heap_start, size,
+                                            MemoryState::Heap));
+    Reprotect(vma, perms);
+
+    heap_used = size;
+
+    return MakeResult<VAddr>(heap_end - size);
+}
+
+ResultCode VMManager::HeapFree(VAddr target, u64 size) {
+    if (target < GetHeapRegionBaseAddress() || target + size > GetHeapRegionEndAddress() ||
+        target + size < target) {
+        return ERR_INVALID_ADDRESS;
+    }
+
+    if (size == 0) {
+        return RESULT_SUCCESS;
+    }
+
+    const ResultCode result = UnmapRange(target, size);
+    if (result.IsError()) {
+        return result;
+    }
+
+    heap_used -= size;
+    return RESULT_SUCCESS;
+}
+
+ResultCode VMManager::MirrorMemory(VAddr dst_addr, VAddr src_addr, u64 size, MemoryState state) {
+    const auto vma = FindVMA(src_addr);
+
+    ASSERT_MSG(vma != vma_map.end(), "Invalid memory address");
+    ASSERT_MSG(vma->second.backing_block, "Backing block doesn't exist for address");
+
+    // The returned VMA might be a bigger one encompassing the desired address.
+    const auto vma_offset = src_addr - vma->first;
+    ASSERT_MSG(vma_offset + size <= vma->second.size,
+               "Shared memory exceeds bounds of mapped block");
+
+    const std::shared_ptr<std::vector<u8>>& backing_block = vma->second.backing_block;
+    const std::size_t backing_block_offset = vma->second.offset + vma_offset;
+
+    CASCADE_RESULT(auto new_vma,
+                   MapMemoryBlock(dst_addr, backing_block, backing_block_offset, size, state));
+    // Protect mirror with permissions from old region
+    Reprotect(new_vma, vma->second.permissions);
+    // Remove permissions from old region
+    Reprotect(vma, VMAPermission::None);
+
+    return RESULT_SUCCESS;
+}
+
 void VMManager::RefreshMemoryBlockMappings(const std::vector<u8>* block) {
     // If this ever proves to have a noticeable performance impact, allow users of the function to
     // specify a specific range of addresses to limit the scan to.
@@ -495,8 +574,7 @@ u64 VMManager::GetTotalMemoryUsage() const {
 }
 
 u64 VMManager::GetTotalHeapUsage() const {
-    LOG_WARNING(Kernel, "(STUBBED) called");
-    return 0x0;
+    return heap_used;
 }
 
 VAddr VMManager::GetAddressSpaceBaseAddress() const {

src/core/hle/kernel/vm_manager.h

@@ -186,6 +186,12 @@ public:
     /// Changes the permissions of a range of addresses, splitting VMAs as necessary.
     ResultCode ReprotectRange(VAddr target, u64 size, VMAPermission new_perms);
 
+    ResultVal<VAddr> HeapAllocate(VAddr target, u64 size, VMAPermission perms);
+    ResultCode HeapFree(VAddr target, u64 size);
+
+    ResultCode MirrorMemory(VAddr dst_addr, VAddr src_addr, u64 size,
+                            MemoryState state = MemoryState::Mapped);
+
     /**
      * Scans all VMAs and updates the page table range of any that use the given vector as backing
      * memory. This should be called after any operation that causes reallocation of the vector.
@@ -343,5 +349,15 @@ private:
 
     VAddr tls_io_region_base = 0;
     VAddr tls_io_region_end = 0;
+
+    // Memory used to back the allocations in the regular heap. A single vector is used to cover
+    // the entire virtual address space extents that bound the allocations, including any holes.
+    // This makes deallocation and reallocation of holes fast and keeps process memory contiguous
+    // in the emulator address space, allowing Memory::GetPointer to be reasonably safe.
+    std::shared_ptr<std::vector<u8>> heap_memory;
+    // The left/right bounds of the address space covered by heap_memory.
+    VAddr heap_start = 0;
+    VAddr heap_end = 0;
+    u64 heap_used = 0;
 };
 } // namespace Kernel

src/core/hle/result.h

@@ -19,8 +19,6 @@
 enum class ErrorDescription : u32 {
     Success = 0,
     RemoteProcessDead = 301,
-    InvalidOffset = 6061,
-    InvalidLength = 6062,
 };
 
 /**

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

@@ -252,8 +252,10 @@ void Module::Interface::TrySelectUserWithoutInteraction(Kernel::HLERequestContex
         rb.PushRaw<u128>(INVALID_UUID);
         return;
     }
-    auto user_list = profile_manager->GetAllUsers();
-    if (user_list.empty()) {
+
+    const auto user_list = profile_manager->GetAllUsers();
+    if (std::all_of(user_list.begin(), user_list.end(),
+                    [](const auto& user) { return user.uuid == INVALID_UUID; })) {
         rb.Push(ResultCode(-1)); // TODO(ogniK): Find the correct error code
         rb.PushRaw<u128>(INVALID_UUID);
         return;

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

@@ -2,8 +2,11 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
+#include <cstring>
 #include <random>
 
+#include <fmt/format.h>
+
 #include "common/file_util.h"
 #include "core/hle/service/acc/profile_manager.h"
 #include "core/settings.h"
@@ -39,6 +42,19 @@ UUID UUID::Generate() {
     return UUID{distribution(gen), distribution(gen)};
 }
 
+std::string UUID::Format() const {
+    return fmt::format("0x{:016X}{:016X}", uuid[1], uuid[0]);
+}
+
+std::string UUID::FormatSwitch() const {
+    std::array<u8, 16> s{};
+    std::memcpy(s.data(), uuid.data(), sizeof(u128));
+    return fmt::format("{:02x}{:02x}{:02x}{:02x}-{:02x}{:02x}-{:02x}{:02x}-{:02x}{:02x}-{:02x}{"
+                       ":02x}{:02x}{:02x}{:02x}{:02x}",
+                       s[0], s[1], s[2], s[3], s[4], s[5], s[6], s[7], s[8], s[9], s[10], s[11],
+                       s[12], s[13], s[14], s[15]);
+}
+
 ProfileManager::ProfileManager() {
     ParseUserSaveFile();
 
@@ -325,11 +341,12 @@ void ProfileManager::ParseUserSaveFile() {
         return;
     }
 
-    for (std::size_t i = 0; i < MAX_USERS; ++i) {
-        const auto& user = data.users[i];
+    for (const auto& user : data.users) {
+        if (user.uuid == UUID(INVALID_UUID)) {
+            continue;
+        }
 
-        if (user.uuid != UUID(INVALID_UUID))
-            AddUser({user.uuid, user.username, user.timestamp, {}, false});
+        AddUser({user.uuid, user.username, user.timestamp, {}, false});
     }
 
     std::stable_partition(profiles.begin(), profiles.end(),

src/core/hle/service/acc/profile_manager.h

@@ -42,18 +42,9 @@ struct UUID {
     void Invalidate() {
         uuid = INVALID_UUID;
     }
-    std::string Format() const {
-        return fmt::format("0x{:016X}{:016X}", uuid[1], uuid[0]);
-    }
 
-    std::string FormatSwitch() const {
-        std::array<u8, 16> s{};
-        std::memcpy(s.data(), uuid.data(), sizeof(u128));
-        return fmt::format("{:02x}{:02x}{:02x}{:02x}-{:02x}{:02x}-{:02x}{:02x}-{:02x}{:02x}-{:02x}{"
-                           ":02x}{:02x}{:02x}{:02x}{:02x}",
-                           s[0], s[1], s[2], s[3], s[4], s[5], s[6], s[7], s[8], s[9], s[10], s[11],
-                           s[12], s[13], s[14], s[15]);
-    }
+    std::string Format() const;
+    std::string FormatSwitch() const;
 };
 static_assert(sizeof(UUID) == 16, "UUID is an invalid size!");
 

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

@@ -203,8 +203,8 @@ ISelfController::ISelfController(std::shared_ptr<NVFlinger::NVFlinger> nvflinger
 ISelfController::~ISelfController() = default;
 
 void ISelfController::SetFocusHandlingMode(Kernel::HLERequestContext& ctx) {
-    // Takes 3 input u8s with each field located immediately after the previous u8, these are
-    // bool flags. No output.
+    // Takes 3 input u8s with each field located immediately after the previous
+    // u8, these are bool flags. No output.
 
     IPC::RequestParser rp{ctx};
 
@@ -258,8 +258,8 @@ void ISelfController::SetOperationModeChangedNotification(Kernel::HLERequestCont
 }
 
 void ISelfController::SetOutOfFocusSuspendingEnabled(Kernel::HLERequestContext& ctx) {
-    // Takes 3 input u8s with each field located immediately after the previous u8, these are
-    // bool flags. No output.
+    // Takes 3 input u8s with each field located immediately after the previous
+    // u8, these are bool flags. No output.
     IPC::RequestParser rp{ctx};
 
     bool enabled = rp.Pop<bool>();
@@ -302,8 +302,8 @@ void ISelfController::SetScreenShotImageOrientation(Kernel::HLERequestContext& c
 }
 
 void ISelfController::CreateManagedDisplayLayer(Kernel::HLERequestContext& ctx) {
-    // TODO(Subv): Find out how AM determines the display to use, for now just create the layer
-    // in the Default display.
+    // TODO(Subv): Find out how AM determines the display to use, for now just
+    // create the layer in the Default display.
     u64 display_id = nvflinger->OpenDisplay("Default");
     u64 layer_id = nvflinger->CreateLayer(display_id);
 
@@ -338,7 +338,54 @@ void ISelfController::GetIdleTimeDetectionExtension(Kernel::HLERequestContext& c
     LOG_WARNING(Service_AM, "(STUBBED) called");
 }
 
-ICommonStateGetter::ICommonStateGetter() : ServiceFramework("ICommonStateGetter") {
+AppletMessageQueue::AppletMessageQueue() {
+    auto& kernel = Core::System::GetInstance().Kernel();
+    on_new_message = Kernel::Event::Create(kernel, Kernel::ResetType::Sticky,
+                                           "AMMessageQueue:OnMessageRecieved");
+    on_operation_mode_changed = Kernel::Event::Create(kernel, Kernel::ResetType::OneShot,
+                                                      "AMMessageQueue:OperationModeChanged");
+}
+
+AppletMessageQueue::~AppletMessageQueue() = default;
+
+const Kernel::SharedPtr<Kernel::Event>& AppletMessageQueue::GetMesssageRecieveEvent() const {
+    return on_new_message;
+}
+
+const Kernel::SharedPtr<Kernel::Event>& AppletMessageQueue::GetOperationModeChangedEvent() const {
+    return on_operation_mode_changed;
+}
+
+void AppletMessageQueue::PushMessage(AppletMessage msg) {
+    messages.push(msg);
+    on_new_message->Signal();
+}
+
+AppletMessageQueue::AppletMessage AppletMessageQueue::PopMessage() {
+    if (messages.empty()) {
+        on_new_message->Clear();
+        return AppletMessage::NoMessage;
+    }
+    auto msg = messages.front();
+    messages.pop();
+    if (messages.empty()) {
+        on_new_message->Clear();
+    }
+    return msg;
+}
+
+std::size_t AppletMessageQueue::GetMessageCount() const {
+    return messages.size();
+}
+
+void AppletMessageQueue::OperationModeChanged() {
+    PushMessage(AppletMessage::OperationModeChanged);
+    PushMessage(AppletMessage::PerformanceModeChanged);
+    on_operation_mode_changed->Signal();
+}
+
+ICommonStateGetter::ICommonStateGetter(std::shared_ptr<AppletMessageQueue> msg_queue)
+    : ServiceFramework("ICommonStateGetter"), msg_queue(std::move(msg_queue)) {
     // clang-format off
     static const FunctionInfo functions[] = {
         {0, &ICommonStateGetter::GetEventHandle, "GetEventHandle"},
@@ -388,21 +435,19 @@ void ICommonStateGetter::GetBootMode(Kernel::HLERequestContext& ctx) {
 }
 
 void ICommonStateGetter::GetEventHandle(Kernel::HLERequestContext& ctx) {
-    event->Signal();
-
     IPC::ResponseBuilder rb{ctx, 2, 1};
     rb.Push(RESULT_SUCCESS);
-    rb.PushCopyObjects(event);
+    rb.PushCopyObjects(msg_queue->GetMesssageRecieveEvent());
 
-    LOG_WARNING(Service_AM, "(STUBBED) called");
+    LOG_DEBUG(Service_AM, "called");
 }
 
 void ICommonStateGetter::ReceiveMessage(Kernel::HLERequestContext& ctx) {
     IPC::ResponseBuilder rb{ctx, 3};
     rb.Push(RESULT_SUCCESS);
-    rb.Push<u32>(15);
+    rb.PushEnum<AppletMessageQueue::AppletMessage>(msg_queue->PopMessage());
 
-    LOG_WARNING(Service_AM, "(STUBBED) called");
+    LOG_DEBUG(Service_AM, "called");
 }
 
 void ICommonStateGetter::GetCurrentFocusState(Kernel::HLERequestContext& ctx) {
@@ -414,13 +459,11 @@ void ICommonStateGetter::GetCurrentFocusState(Kernel::HLERequestContext& ctx) {
 }
 
 void ICommonStateGetter::GetDefaultDisplayResolutionChangeEvent(Kernel::HLERequestContext& ctx) {
-    event->Signal();
-
     IPC::ResponseBuilder rb{ctx, 2, 1};
     rb.Push(RESULT_SUCCESS);
-    rb.PushCopyObjects(event);
+    rb.PushCopyObjects(msg_queue->GetOperationModeChangedEvent());
 
-    LOG_WARNING(Service_AM, "(STUBBED) called");
+    LOG_DEBUG(Service_AM, "called");
 }
 
 void ICommonStateGetter::GetDefaultDisplayResolution(Kernel::HLERequestContext& ctx) {
@@ -444,7 +487,7 @@ void ICommonStateGetter::GetOperationMode(Kernel::HLERequestContext& ctx) {
     rb.Push(RESULT_SUCCESS);
     rb.Push(static_cast<u8>(use_docked_mode ? OperationMode::Docked : OperationMode::Handheld));
 
-    LOG_WARNING(Service_AM, "(STUBBED) called");
+    LOG_DEBUG(Service_AM, "called");
 }
 
 void ICommonStateGetter::GetPerformanceMode(Kernel::HLERequestContext& ctx) {
@@ -454,7 +497,7 @@ void ICommonStateGetter::GetPerformanceMode(Kernel::HLERequestContext& ctx) {
     rb.Push(static_cast<u32>(use_docked_mode ? APM::PerformanceMode::Docked
                                              : APM::PerformanceMode::Handheld));
 
-    LOG_WARNING(Service_AM, "(STUBBED) called");
+    LOG_DEBUG(Service_AM, "called");
 }
 
 class IStorageAccessor final : public ServiceFramework<IStorageAccessor> {
@@ -690,7 +733,7 @@ IApplicationFunctions::IApplicationFunctions() : ServiceFramework("IApplicationF
         {70, nullptr, "RequestToShutdown"},
         {71, nullptr, "RequestToReboot"},
         {80, nullptr, "ExitAndRequestToShowThanksMessage"},
-        {90, nullptr, "EnableApplicationCrashReport"},
+        {90, &IApplicationFunctions::EnableApplicationCrashReport, "EnableApplicationCrashReport"},
         {100, nullptr, "InitializeApplicationCopyrightFrameBuffer"},
         {101, nullptr, "SetApplicationCopyrightImage"},
         {102, nullptr, "SetApplicationCopyrightVisibility"},
@@ -709,6 +752,12 @@ IApplicationFunctions::IApplicationFunctions() : ServiceFramework("IApplicationF
 
 IApplicationFunctions::~IApplicationFunctions() = default;
 
+void IApplicationFunctions::EnableApplicationCrashReport(Kernel::HLERequestContext& ctx) {
+    IPC::ResponseBuilder rb{ctx, 2};
+    rb.Push(RESULT_SUCCESS);
+    LOG_WARNING(Service_AM, "(STUBBED) called");
+}
+
 void IApplicationFunctions::BeginBlockingHomeButtonShortAndLongPressed(
     Kernel::HLERequestContext& ctx) {
     IPC::ResponseBuilder rb{ctx, 2};
@@ -778,7 +827,8 @@ void IApplicationFunctions::EnsureSaveData(Kernel::HLERequestContext& ctx) {
 
 void IApplicationFunctions::SetTerminateResult(Kernel::HLERequestContext& ctx) {
     // Takes an input u32 Result, no output.
-    // For example, in some cases official apps use this with error 0x2A2 then uses svcBreak.
+    // For example, in some cases official apps use this with error 0x2A2 then
+    // uses svcBreak.
 
     IPC::RequestParser rp{ctx};
     u32 result = rp.Pop<u32>();
@@ -840,8 +890,12 @@ void IApplicationFunctions::GetPseudoDeviceId(Kernel::HLERequestContext& ctx) {
 
 void InstallInterfaces(SM::ServiceManager& service_manager,
                        std::shared_ptr<NVFlinger::NVFlinger> nvflinger) {
-    std::make_shared<AppletAE>(nvflinger)->InstallAsService(service_manager);
-    std::make_shared<AppletOE>(nvflinger)->InstallAsService(service_manager);
+    auto message_queue = std::make_shared<AppletMessageQueue>();
+    message_queue->PushMessage(AppletMessageQueue::AppletMessage::FocusStateChanged); // Needed on
+                                                                                      // game boot
+
+    std::make_shared<AppletAE>(nvflinger, message_queue)->InstallAsService(service_manager);
+    std::make_shared<AppletOE>(nvflinger, message_queue)->InstallAsService(service_manager);
     std::make_shared<IdleSys>()->InstallAsService(service_manager);
     std::make_shared<OMM>()->InstallAsService(service_manager);
     std::make_shared<SPSM>()->InstallAsService(service_manager);

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

@@ -5,6 +5,7 @@
 #pragma once
 
 #include <memory>
+#include <queue>
 #include "core/hle/service/service.h"
 
 namespace Kernel {
@@ -39,6 +40,31 @@ enum SystemLanguage {
     TraditionalChinese = 16,
 };
 
+class AppletMessageQueue {
+public:
+    enum class AppletMessage : u32 {
+        NoMessage = 0,
+        FocusStateChanged = 15,
+        OperationModeChanged = 30,
+        PerformanceModeChanged = 31,
+    };
+
+    AppletMessageQueue();
+    ~AppletMessageQueue();
+
+    const Kernel::SharedPtr<Kernel::Event>& GetMesssageRecieveEvent() const;
+    const Kernel::SharedPtr<Kernel::Event>& GetOperationModeChangedEvent() const;
+    void PushMessage(AppletMessage msg);
+    AppletMessage PopMessage();
+    std::size_t GetMessageCount() const;
+    void OperationModeChanged();
+
+private:
+    std::queue<AppletMessage> messages;
+    Kernel::SharedPtr<Kernel::Event> on_new_message;
+    Kernel::SharedPtr<Kernel::Event> on_operation_mode_changed;
+};
+
 class IWindowController final : public ServiceFramework<IWindowController> {
 public:
     IWindowController();
@@ -102,7 +128,7 @@ private:
 
 class ICommonStateGetter final : public ServiceFramework<ICommonStateGetter> {
 public:
-    ICommonStateGetter();
+    explicit ICommonStateGetter(std::shared_ptr<AppletMessageQueue> msg_queue);
     ~ICommonStateGetter() override;
 
 private:
@@ -126,6 +152,7 @@ private:
     void GetDefaultDisplayResolution(Kernel::HLERequestContext& ctx);
 
     Kernel::SharedPtr<Kernel::Event> event;
+    std::shared_ptr<AppletMessageQueue> msg_queue;
 };
 
 class ILibraryAppletCreator final : public ServiceFramework<ILibraryAppletCreator> {
@@ -158,6 +185,7 @@ private:
     void EndBlockingHomeButtonShortAndLongPressed(Kernel::HLERequestContext& ctx);
     void BeginBlockingHomeButton(Kernel::HLERequestContext& ctx);
     void EndBlockingHomeButton(Kernel::HLERequestContext& ctx);
+    void EnableApplicationCrashReport(Kernel::HLERequestContext& ctx);
 };
 
 class IHomeMenuFunctions final : public ServiceFramework<IHomeMenuFunctions> {

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

@@ -12,8 +12,10 @@ namespace Service::AM {
 
 class ILibraryAppletProxy final : public ServiceFramework<ILibraryAppletProxy> {
 public:
-    explicit ILibraryAppletProxy(std::shared_ptr<NVFlinger::NVFlinger> nvflinger)
-        : ServiceFramework("ILibraryAppletProxy"), nvflinger(std::move(nvflinger)) {
+    explicit ILibraryAppletProxy(std::shared_ptr<NVFlinger::NVFlinger> nvflinger,
+                                 std::shared_ptr<AppletMessageQueue> msg_queue)
+        : ServiceFramework("ILibraryAppletProxy"), nvflinger(std::move(nvflinger)),
+          msg_queue(std::move(msg_queue)) {
         static const FunctionInfo functions[] = {
             {0, &ILibraryAppletProxy::GetCommonStateGetter, "GetCommonStateGetter"},
             {1, &ILibraryAppletProxy::GetSelfController, "GetSelfController"},
@@ -32,7 +34,7 @@ private:
     void GetCommonStateGetter(Kernel::HLERequestContext& ctx) {
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(RESULT_SUCCESS);
-        rb.PushIpcInterface<ICommonStateGetter>();
+        rb.PushIpcInterface<ICommonStateGetter>(msg_queue);
         LOG_DEBUG(Service_AM, "called");
     }
 
@@ -93,12 +95,15 @@ private:
     }
 
     std::shared_ptr<NVFlinger::NVFlinger> nvflinger;
+    std::shared_ptr<AppletMessageQueue> msg_queue;
 };
 
 class ISystemAppletProxy final : public ServiceFramework<ISystemAppletProxy> {
 public:
-    explicit ISystemAppletProxy(std::shared_ptr<NVFlinger::NVFlinger> nvflinger)
-        : ServiceFramework("ISystemAppletProxy"), nvflinger(std::move(nvflinger)) {
+    explicit ISystemAppletProxy(std::shared_ptr<NVFlinger::NVFlinger> nvflinger,
+                                std::shared_ptr<AppletMessageQueue> msg_queue)
+        : ServiceFramework("ISystemAppletProxy"), nvflinger(std::move(nvflinger)),
+          msg_queue(std::move(msg_queue)) {
         static const FunctionInfo functions[] = {
             {0, &ISystemAppletProxy::GetCommonStateGetter, "GetCommonStateGetter"},
             {1, &ISystemAppletProxy::GetSelfController, "GetSelfController"},
@@ -119,7 +124,7 @@ private:
     void GetCommonStateGetter(Kernel::HLERequestContext& ctx) {
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(RESULT_SUCCESS);
-        rb.PushIpcInterface<ICommonStateGetter>();
+        rb.PushIpcInterface<ICommonStateGetter>(msg_queue);
         LOG_DEBUG(Service_AM, "called");
     }
 
@@ -186,31 +191,34 @@ private:
         LOG_DEBUG(Service_AM, "called");
     }
     std::shared_ptr<NVFlinger::NVFlinger> nvflinger;
+    std::shared_ptr<AppletMessageQueue> msg_queue;
 };
 
 void AppletAE::OpenSystemAppletProxy(Kernel::HLERequestContext& ctx) {
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(RESULT_SUCCESS);
-    rb.PushIpcInterface<ISystemAppletProxy>(nvflinger);
+    rb.PushIpcInterface<ISystemAppletProxy>(nvflinger, msg_queue);
     LOG_DEBUG(Service_AM, "called");
 }
 
 void AppletAE::OpenLibraryAppletProxy(Kernel::HLERequestContext& ctx) {
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(RESULT_SUCCESS);
-    rb.PushIpcInterface<ILibraryAppletProxy>(nvflinger);
+    rb.PushIpcInterface<ILibraryAppletProxy>(nvflinger, msg_queue);
     LOG_DEBUG(Service_AM, "called");
 }
 
 void AppletAE::OpenLibraryAppletProxyOld(Kernel::HLERequestContext& ctx) {
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(RESULT_SUCCESS);
-    rb.PushIpcInterface<ILibraryAppletProxy>(nvflinger);
+    rb.PushIpcInterface<ILibraryAppletProxy>(nvflinger, msg_queue);
     LOG_DEBUG(Service_AM, "called");
 }
 
-AppletAE::AppletAE(std::shared_ptr<NVFlinger::NVFlinger> nvflinger)
-    : ServiceFramework("appletAE"), nvflinger(std::move(nvflinger)) {
+AppletAE::AppletAE(std::shared_ptr<NVFlinger::NVFlinger> nvflinger,
+                   std::shared_ptr<AppletMessageQueue> msg_queue)
+    : ServiceFramework("appletAE"), nvflinger(std::move(nvflinger)),
+      msg_queue(std::move(msg_queue)) {
     // clang-format off
     static const FunctionInfo functions[] = {
         {100, &AppletAE::OpenSystemAppletProxy, "OpenSystemAppletProxy"},
@@ -228,4 +236,8 @@ AppletAE::AppletAE(std::shared_ptr<NVFlinger::NVFlinger> nvflinger)
 
 AppletAE::~AppletAE() = default;
 
+const std::shared_ptr<AppletMessageQueue>& AppletAE::GetMessageQueue() const {
+    return msg_queue;
+}
+
 } // namespace Service::AM

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

@@ -17,15 +17,19 @@ namespace AM {
 
 class AppletAE final : public ServiceFramework<AppletAE> {
 public:
-    explicit AppletAE(std::shared_ptr<NVFlinger::NVFlinger> nvflinger);
+    explicit AppletAE(std::shared_ptr<NVFlinger::NVFlinger> nvflinger,
+                      std::shared_ptr<AppletMessageQueue> msg_queue);
     ~AppletAE() override;
 
+    const std::shared_ptr<AppletMessageQueue>& GetMessageQueue() const;
+
 private:
     void OpenSystemAppletProxy(Kernel::HLERequestContext& ctx);
     void OpenLibraryAppletProxy(Kernel::HLERequestContext& ctx);
     void OpenLibraryAppletProxyOld(Kernel::HLERequestContext& ctx);
 
     std::shared_ptr<NVFlinger::NVFlinger> nvflinger;
+    std::shared_ptr<AppletMessageQueue> msg_queue;
 };
 
 } // namespace AM

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

@@ -12,8 +12,10 @@ namespace Service::AM {
 
 class IApplicationProxy final : public ServiceFramework<IApplicationProxy> {
 public:
-    explicit IApplicationProxy(std::shared_ptr<NVFlinger::NVFlinger> nvflinger)
-        : ServiceFramework("IApplicationProxy"), nvflinger(std::move(nvflinger)) {
+    explicit IApplicationProxy(std::shared_ptr<NVFlinger::NVFlinger> nvflinger,
+                               std::shared_ptr<AppletMessageQueue> msg_queue)
+        : ServiceFramework("IApplicationProxy"), nvflinger(std::move(nvflinger)),
+          msg_queue(std::move(msg_queue)) {
         // clang-format off
         static const FunctionInfo functions[] = {
             {0, &IApplicationProxy::GetCommonStateGetter, "GetCommonStateGetter"},
@@ -70,7 +72,7 @@ private:
     void GetCommonStateGetter(Kernel::HLERequestContext& ctx) {
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(RESULT_SUCCESS);
-        rb.PushIpcInterface<ICommonStateGetter>();
+        rb.PushIpcInterface<ICommonStateGetter>(msg_queue);
         LOG_DEBUG(Service_AM, "called");
     }
 
@@ -89,17 +91,20 @@ private:
     }
 
     std::shared_ptr<NVFlinger::NVFlinger> nvflinger;
+    std::shared_ptr<AppletMessageQueue> msg_queue;
 };
 
 void AppletOE::OpenApplicationProxy(Kernel::HLERequestContext& ctx) {
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(RESULT_SUCCESS);
-    rb.PushIpcInterface<IApplicationProxy>(nvflinger);
+    rb.PushIpcInterface<IApplicationProxy>(nvflinger, msg_queue);
     LOG_DEBUG(Service_AM, "called");
 }
 
-AppletOE::AppletOE(std::shared_ptr<NVFlinger::NVFlinger> nvflinger)
-    : ServiceFramework("appletOE"), nvflinger(std::move(nvflinger)) {
+AppletOE::AppletOE(std::shared_ptr<NVFlinger::NVFlinger> nvflinger,
+                   std::shared_ptr<AppletMessageQueue> msg_queue)
+    : ServiceFramework("appletOE"), nvflinger(std::move(nvflinger)),
+      msg_queue(std::move(msg_queue)) {
     static const FunctionInfo functions[] = {
         {0, &AppletOE::OpenApplicationProxy, "OpenApplicationProxy"},
     };
@@ -108,4 +113,8 @@ AppletOE::AppletOE(std::shared_ptr<NVFlinger::NVFlinger> nvflinger)
 
 AppletOE::~AppletOE() = default;
 
+const std::shared_ptr<AppletMessageQueue>& AppletOE::GetMessageQueue() const {
+    return msg_queue;
+}
+
 } // namespace Service::AM

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

@@ -17,13 +17,17 @@ namespace AM {
 
 class AppletOE final : public ServiceFramework<AppletOE> {
 public:
-    explicit AppletOE(std::shared_ptr<NVFlinger::NVFlinger> nvflinger);
+    explicit AppletOE(std::shared_ptr<NVFlinger::NVFlinger> nvflinger,
+                      std::shared_ptr<AppletMessageQueue> msg_queue);
     ~AppletOE() override;
 
+    const std::shared_ptr<AppletMessageQueue>& GetMessageQueue() const;
+
 private:
     void OpenApplicationProxy(Kernel::HLERequestContext& ctx);
 
     std::shared_ptr<NVFlinger::NVFlinger> nvflinger;
+    std::shared_ptr<AppletMessageQueue> msg_queue;
 };
 
 } // namespace AM

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

@@ -28,13 +28,13 @@ public:
             {1, &IAudioRenderer::GetSampleCount, "GetSampleCount"},
             {2, &IAudioRenderer::GetMixBufferCount, "GetMixBufferCount"},
             {3, &IAudioRenderer::GetState, "GetState"},
-            {4, &IAudioRenderer::RequestUpdate, "RequestUpdate"},
+            {4, &IAudioRenderer::RequestUpdateImpl, "RequestUpdate"},
             {5, &IAudioRenderer::Start, "Start"},
             {6, &IAudioRenderer::Stop, "Stop"},
             {7, &IAudioRenderer::QuerySystemEvent, "QuerySystemEvent"},
-            {8, nullptr, "SetRenderingTimeLimit"},
-            {9, nullptr, "GetRenderingTimeLimit"},
-            {10, nullptr, "RequestUpdateAuto"},
+            {8, &IAudioRenderer::SetRenderingTimeLimit, "SetRenderingTimeLimit"},
+            {9, &IAudioRenderer::GetRenderingTimeLimit, "GetRenderingTimeLimit"},
+            {10, &IAudioRenderer::RequestUpdateImpl, "RequestUpdateAuto"},
             {11, nullptr, "ExecuteAudioRendererRendering"},
         };
         // clang-format on
@@ -79,7 +79,7 @@ private:
         LOG_DEBUG(Service_Audio, "called");
     }
 
-    void RequestUpdate(Kernel::HLERequestContext& ctx) {
+    void RequestUpdateImpl(Kernel::HLERequestContext& ctx) {
         ctx.WriteBuffer(renderer->UpdateAudioRenderer(ctx.ReadBuffer()));
         IPC::ResponseBuilder rb{ctx, 2};
         rb.Push(RESULT_SUCCESS);
@@ -110,8 +110,29 @@ private:
         LOG_WARNING(Service_Audio, "(STUBBED) called");
     }
 
+    void SetRenderingTimeLimit(Kernel::HLERequestContext& ctx) {
+        IPC::RequestParser rp{ctx};
+        rendering_time_limit_percent = rp.Pop<u32>();
+        ASSERT(rendering_time_limit_percent >= 0 && rendering_time_limit_percent <= 100);
+
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(RESULT_SUCCESS);
+
+        LOG_DEBUG(Service_Audio, "called. rendering_time_limit_percent={}",
+                  rendering_time_limit_percent);
+    }
+
+    void GetRenderingTimeLimit(Kernel::HLERequestContext& ctx) {
+        LOG_DEBUG(Service_Audio, "called");
+
+        IPC::ResponseBuilder rb{ctx, 3};
+        rb.Push(RESULT_SUCCESS);
+        rb.Push(rendering_time_limit_percent);
+    }
+
     Kernel::SharedPtr<Kernel::Event> system_event;
     std::unique_ptr<AudioCore::AudioRenderer> renderer;
+    u32 rendering_time_limit_percent = 100;
 };
 
 class IAudioDevice final : public ServiceFramework<IAudioDevice> {

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

@@ -77,8 +77,8 @@ private:
         IPC::ResponseBuilder rb{ctx, 6};
         rb.Push(RESULT_SUCCESS);
         rb.Push<u32>(consumed);
-        rb.Push<u64>(performance);
         rb.Push<u32>(sample_count);
+        rb.Push<u64>(performance);
         ctx.WriteBuffer(samples.data(), samples.size() * sizeof(s16));
     }
 

src/core/hle/service/btdrv/btdrv.cpp

@@ -2,12 +2,49 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
+#include "common/logging/log.h"
+#include "core/hle/ipc_helpers.h"
+#include "core/hle/kernel/event.h"
+#include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/service/btdrv/btdrv.h"
 #include "core/hle/service/service.h"
 #include "core/hle/service/sm/sm.h"
 
 namespace Service::BtDrv {
 
+class Bt final : public ServiceFramework<Bt> {
+public:
+    explicit Bt() : ServiceFramework{"bt"} {
+        // clang-format off
+        static const FunctionInfo functions[] = {
+            {0, nullptr, "Unknown0"},
+            {1, nullptr, "Unknown1"},
+            {2, nullptr, "Unknown2"},
+            {3, nullptr, "Unknown3"},
+            {4, nullptr, "Unknown4"},
+            {5, nullptr, "Unknown5"},
+            {6, nullptr, "Unknown6"},
+            {7, nullptr, "Unknown7"},
+            {8, nullptr, "Unknown8"},
+            {9, &Bt::RegisterEvent, "RegisterEvent"},
+        };
+        // clang-format on
+        RegisterHandlers(functions);
+    }
+
+private:
+    void RegisterEvent(Kernel::HLERequestContext& ctx) {
+        auto& kernel = Core::System::GetInstance().Kernel();
+        register_event =
+            Kernel::Event::Create(kernel, Kernel::ResetType::OneShot, "BT:RegisterEvent");
+        IPC::ResponseBuilder rb{ctx, 2, 1};
+        rb.Push(RESULT_SUCCESS);
+        rb.PushCopyObjects(register_event);
+        LOG_WARNING(Service_BTM, "(STUBBED) called");
+    }
+    Kernel::SharedPtr<Kernel::Event> register_event;
+};
+
 class BtDrv final : public ServiceFramework<BtDrv> {
 public:
     explicit BtDrv() : ServiceFramework{"btdrv"} {
@@ -67,6 +104,7 @@ public:
 
 void InstallInterfaces(SM::ServiceManager& sm) {
     std::make_shared<BtDrv>()->InstallAsService(sm);
+    std::make_shared<Bt>()->InstallAsService(sm);
 }
 
 } // namespace Service::BtDrv

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

@@ -6,13 +6,118 @@
 
 #include "common/logging/log.h"
 #include "core/hle/ipc_helpers.h"
+#include "core/hle/kernel/event.h"
 #include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/service/btm/btm.h"
 #include "core/hle/service/service.h"
-#include "core/hle/service/sm/sm.h"
 
 namespace Service::BTM {
 
+class IBtmUserCore final : public ServiceFramework<IBtmUserCore> {
+public:
+    explicit IBtmUserCore() : ServiceFramework{"IBtmUserCore"} {
+        // clang-format off
+        static const FunctionInfo functions[] = {
+            {0, &IBtmUserCore::GetScanEvent, "GetScanEvent"},
+            {1, nullptr, "Unknown1"},
+            {2, nullptr, "Unknown2"},
+            {3, nullptr, "Unknown3"},
+            {4, nullptr, "Unknown4"},
+            {5, nullptr, "Unknown5"},
+            {6, nullptr, "Unknown6"},
+            {7, nullptr, "Unknown7"},
+            {8, nullptr, "Unknown8"},
+            {9, nullptr, "Unknown9"},
+            {10, nullptr, "Unknown10"},
+            {17, &IBtmUserCore::GetConnectionEvent, "GetConnectionEvent"},
+            {18, nullptr, "Unknown18"},
+            {19, nullptr, "Unknown19"},
+            {20, nullptr, "Unknown20"},
+            {21, nullptr, "Unknown21"},
+            {22, nullptr, "Unknown22"},
+            {23, nullptr, "Unknown23"},
+            {24, nullptr, "Unknown24"},
+            {25, nullptr, "Unknown25"},
+            {26, &IBtmUserCore::GetDiscoveryEvent, "AcquireBleServiceDiscoveryEventImpl"},
+            {27, nullptr, "Unknown27"},
+            {28, nullptr, "Unknown28"},
+            {29, nullptr, "Unknown29"},
+            {30, nullptr, "Unknown30"},
+            {31, nullptr, "Unknown31"},
+            {32, nullptr, "Unknown32"},
+            {33, &IBtmUserCore::GetConfigEvent, "GetConfigEvent"},
+            {34, nullptr, "Unknown34"},
+            {35, nullptr, "Unknown35"},
+            {36, nullptr, "Unknown36"},
+            {37, nullptr, "Unknown37"},
+        };
+        // clang-format on
+        RegisterHandlers(functions);
+    }
+
+private:
+    void GetScanEvent(Kernel::HLERequestContext& ctx) {
+        auto& kernel = Core::System::GetInstance().Kernel();
+        scan_event =
+            Kernel::Event::Create(kernel, Kernel::ResetType::OneShot, "IBtmUserCore:ScanEvent");
+        IPC::ResponseBuilder rb{ctx, 2, 1};
+        rb.Push(RESULT_SUCCESS);
+        rb.PushCopyObjects(scan_event);
+        LOG_WARNING(Service_BTM, "(STUBBED) called");
+    }
+    void GetConnectionEvent(Kernel::HLERequestContext& ctx) {
+        auto& kernel = Core::System::GetInstance().Kernel();
+        connection_event = Kernel::Event::Create(kernel, Kernel::ResetType::OneShot,
+                                                 "IBtmUserCore:ConnectionEvent");
+        IPC::ResponseBuilder rb{ctx, 2, 1};
+        rb.Push(RESULT_SUCCESS);
+        rb.PushCopyObjects(connection_event);
+        LOG_WARNING(Service_BTM, "(STUBBED) called");
+    }
+    void GetDiscoveryEvent(Kernel::HLERequestContext& ctx) {
+        auto& kernel = Core::System::GetInstance().Kernel();
+        service_discovery =
+            Kernel::Event::Create(kernel, Kernel::ResetType::OneShot, "IBtmUserCore:Discovery");
+        IPC::ResponseBuilder rb{ctx, 2, 1};
+        rb.Push(RESULT_SUCCESS);
+        rb.PushCopyObjects(service_discovery);
+        LOG_WARNING(Service_BTM, "(STUBBED) called");
+    }
+    void GetConfigEvent(Kernel::HLERequestContext& ctx) {
+        auto& kernel = Core::System::GetInstance().Kernel();
+        config_event =
+            Kernel::Event::Create(kernel, Kernel::ResetType::OneShot, "IBtmUserCore:ConfigEvent");
+        IPC::ResponseBuilder rb{ctx, 2, 1};
+        rb.Push(RESULT_SUCCESS);
+        rb.PushCopyObjects(config_event);
+        LOG_WARNING(Service_BTM, "(STUBBED) called");
+    }
+    Kernel::SharedPtr<Kernel::Event> scan_event;
+    Kernel::SharedPtr<Kernel::Event> connection_event;
+    Kernel::SharedPtr<Kernel::Event> service_discovery;
+    Kernel::SharedPtr<Kernel::Event> config_event;
+};
+
+class BTM_USR final : public ServiceFramework<BTM_USR> {
+public:
+    explicit BTM_USR() : ServiceFramework{"btm:u"} {
+        // clang-format off
+        static const FunctionInfo functions[] = {
+            {0, &BTM_USR::GetCoreImpl, "GetCoreImpl"},
+        };
+        // clang-format on
+        RegisterHandlers(functions);
+    }
+
+private:
+    void GetCoreImpl(Kernel::HLERequestContext& ctx) {
+        IPC::ResponseBuilder rb{ctx, 2, 0, 1};
+        rb.Push(RESULT_SUCCESS);
+        rb.PushIpcInterface<IBtmUserCore>();
+        LOG_DEBUG(Service_BTM, "called");
+    }
+};
+
 class BTM final : public ServiceFramework<BTM> {
 public:
     explicit BTM() : ServiceFramework{"btm"} {
@@ -116,6 +221,7 @@ void InstallInterfaces(SM::ServiceManager& sm) {
     std::make_shared<BTM>()->InstallAsService(sm);
     std::make_shared<BTM_DBG>()->InstallAsService(sm);
     std::make_shared<BTM_SYS>()->InstallAsService(sm);
+    std::make_shared<BTM_USR>()->InstallAsService(sm);
 }
 
 } // namespace Service::BTM

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

@@ -303,25 +303,42 @@ ResultVal<FileSys::VirtualDir> OpenSaveData(FileSys::SaveDataSpaceId space,
               static_cast<u8>(space), save_struct.DebugInfo());
 
     if (save_data_factory == nullptr) {
-        return ResultCode(ErrorModule::FS, FileSys::ErrCodes::TitleNotFound);
+        return FileSys::ERROR_ENTITY_NOT_FOUND;
     }
 
     return save_data_factory->Open(space, save_struct);
 }
 
+ResultVal<FileSys::VirtualDir> OpenSaveDataSpace(FileSys::SaveDataSpaceId space) {
+    LOG_TRACE(Service_FS, "Opening Save Data Space for space_id={:01X}", static_cast<u8>(space));
+
+    if (save_data_factory == nullptr) {
+        return FileSys::ERROR_ENTITY_NOT_FOUND;
+    }
+
+    return MakeResult(save_data_factory->GetSaveDataSpaceDirectory(space));
+}
+
 ResultVal<FileSys::VirtualDir> OpenSDMC() {
     LOG_TRACE(Service_FS, "Opening SDMC");
 
     if (sdmc_factory == nullptr) {
-        return ResultCode(ErrorModule::FS, FileSys::ErrCodes::SdCardNotFound);
+        return FileSys::ERROR_SD_CARD_NOT_FOUND;
     }
 
     return sdmc_factory->Open();
 }
 
-std::unique_ptr<FileSys::RegisteredCacheUnion> GetUnionContents() {
-    return std::make_unique<FileSys::RegisteredCacheUnion>(std::vector<FileSys::RegisteredCache*>{
-        GetSystemNANDContents(), GetUserNANDContents(), GetSDMCContents()});
+std::shared_ptr<FileSys::RegisteredCacheUnion> registered_cache_union;
+
+std::shared_ptr<FileSys::RegisteredCacheUnion> GetUnionContents() {
+    if (registered_cache_union == nullptr) {
+        registered_cache_union =
+            std::make_shared<FileSys::RegisteredCacheUnion>(std::vector<FileSys::RegisteredCache*>{
+                GetSystemNANDContents(), GetUserNANDContents(), GetSDMCContents()});
+    }
+
+    return registered_cache_union;
 }
 
 FileSys::RegisteredCache* GetSystemNANDContents() {
@@ -360,6 +377,15 @@ FileSys::VirtualDir GetModificationLoadRoot(u64 title_id) {
     return bis_factory->GetModificationLoadRoot(title_id);
 }
 
+FileSys::VirtualDir GetModificationDumpRoot(u64 title_id) {
+    LOG_TRACE(Service_FS, "Opening mod dump root for tid={:016X}", title_id);
+
+    if (bis_factory == nullptr)
+        return nullptr;
+
+    return bis_factory->GetModificationDumpRoot(title_id);
+}
+
 void CreateFactories(FileSys::VfsFilesystem& vfs, bool overwrite) {
     if (overwrite) {
         bis_factory = nullptr;
@@ -373,13 +399,21 @@ void CreateFactories(FileSys::VfsFilesystem& vfs, bool overwrite) {
                                           FileSys::Mode::ReadWrite);
     auto load_directory = vfs.OpenDirectory(FileUtil::GetUserPath(FileUtil::UserPath::LoadDir),
                                             FileSys::Mode::ReadWrite);
+    auto dump_directory = vfs.OpenDirectory(FileUtil::GetUserPath(FileUtil::UserPath::DumpDir),
+                                            FileSys::Mode::ReadWrite);
 
-    if (bis_factory == nullptr)
-        bis_factory = std::make_unique<FileSys::BISFactory>(nand_directory, load_directory);
-    if (save_data_factory == nullptr)
+    if (bis_factory == nullptr) {
+        bis_factory =
+            std::make_unique<FileSys::BISFactory>(nand_directory, load_directory, dump_directory);
+    }
+
+    if (save_data_factory == nullptr) {
         save_data_factory = std::make_unique<FileSys::SaveDataFactory>(std::move(nand_directory));
-    if (sdmc_factory == nullptr)
+    }
+
+    if (sdmc_factory == nullptr) {
         sdmc_factory = std::make_unique<FileSys::SDMCFactory>(std::move(sd_directory));
+    }
 }
 
 void InstallInterfaces(SM::ServiceManager& service_manager, FileSys::VfsFilesystem& vfs) {

src/core/hle/service/filesystem/filesystem.h

@@ -45,15 +45,17 @@ ResultVal<FileSys::VirtualFile> OpenRomFS(u64 title_id, FileSys::StorageId stora
                                           FileSys::ContentRecordType type);
 ResultVal<FileSys::VirtualDir> OpenSaveData(FileSys::SaveDataSpaceId space,
                                             FileSys::SaveDataDescriptor save_struct);
+ResultVal<FileSys::VirtualDir> OpenSaveDataSpace(FileSys::SaveDataSpaceId space);
 ResultVal<FileSys::VirtualDir> OpenSDMC();
 
-std::unique_ptr<FileSys::RegisteredCacheUnion> GetUnionContents();
+std::shared_ptr<FileSys::RegisteredCacheUnion> GetUnionContents();
 
 FileSys::RegisteredCache* GetSystemNANDContents();
 FileSys::RegisteredCache* GetUserNANDContents();
 FileSys::RegisteredCache* GetSDMCContents();
 
 FileSys::VirtualDir GetModificationLoadRoot(u64 title_id);
+FileSys::VirtualDir GetModificationDumpRoot(u64 title_id);
 
 // Creates the SaveData, SDMC, and BIS Factories. Should be called once and before any function
 // above is called.

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

@@ -11,6 +11,7 @@
 
 #include "common/assert.h"
 #include "common/common_types.h"
+#include "common/hex_util.h"
 #include "common/logging/log.h"
 #include "common/string_util.h"
 #include "core/file_sys/directory.h"
@@ -62,12 +63,12 @@ private:
         // Error checking
         if (length < 0) {
             IPC::ResponseBuilder rb{ctx, 2};
-            rb.Push(ResultCode(ErrorModule::FS, ErrorDescription::InvalidLength));
+            rb.Push(FileSys::ERROR_INVALID_SIZE);
             return;
         }
         if (offset < 0) {
             IPC::ResponseBuilder rb{ctx, 2};
-            rb.Push(ResultCode(ErrorModule::FS, ErrorDescription::InvalidOffset));
+            rb.Push(FileSys::ERROR_INVALID_OFFSET);
             return;
         }
 
@@ -107,12 +108,12 @@ private:
         // Error checking
         if (length < 0) {
             IPC::ResponseBuilder rb{ctx, 2};
-            rb.Push(ResultCode(ErrorModule::FS, ErrorDescription::InvalidLength));
+            rb.Push(FileSys::ERROR_INVALID_SIZE);
             return;
         }
         if (offset < 0) {
             IPC::ResponseBuilder rb{ctx, 2};
-            rb.Push(ResultCode(ErrorModule::FS, ErrorDescription::InvalidOffset));
+            rb.Push(FileSys::ERROR_INVALID_OFFSET);
             return;
         }
 
@@ -138,12 +139,12 @@ private:
         // Error checking
         if (length < 0) {
             IPC::ResponseBuilder rb{ctx, 2};
-            rb.Push(ResultCode(ErrorModule::FS, ErrorDescription::InvalidLength));
+            rb.Push(FileSys::ERROR_INVALID_SIZE);
             return;
         }
         if (offset < 0) {
             IPC::ResponseBuilder rb{ctx, 2};
-            rb.Push(ResultCode(ErrorModule::FS, ErrorDescription::InvalidOffset));
+            rb.Push(FileSys::ERROR_INVALID_OFFSET);
             return;
         }
 
@@ -451,7 +452,147 @@ private:
     VfsDirectoryServiceWrapper backend;
 };
 
+class ISaveDataInfoReader final : public ServiceFramework<ISaveDataInfoReader> {
+public:
+    explicit ISaveDataInfoReader(FileSys::SaveDataSpaceId space)
+        : ServiceFramework("ISaveDataInfoReader") {
+        static const FunctionInfo functions[] = {
+            {0, &ISaveDataInfoReader::ReadSaveDataInfo, "ReadSaveDataInfo"},
+        };
+        RegisterHandlers(functions);
+
+        FindAllSaves(space);
+    }
+
+    void ReadSaveDataInfo(Kernel::HLERequestContext& ctx) {
+        // Calculate how many entries we can fit in the output buffer
+        const u64 count_entries = ctx.GetWriteBufferSize() / sizeof(SaveDataInfo);
+
+        // Cap at total number of entries.
+        const u64 actual_entries = std::min(count_entries, info.size() - next_entry_index);
+
+        // Determine data start and end
+        const auto* begin = reinterpret_cast<u8*>(info.data() + next_entry_index);
+        const auto* end = reinterpret_cast<u8*>(info.data() + next_entry_index + actual_entries);
+        const auto range_size = static_cast<std::size_t>(std::distance(begin, end));
+
+        next_entry_index += actual_entries;
+
+        // Write the data to memory
+        ctx.WriteBuffer(begin, range_size);
+
+        IPC::ResponseBuilder rb{ctx, 3};
+        rb.Push(RESULT_SUCCESS);
+        rb.Push<u32>(static_cast<u32>(actual_entries));
+    }
+
+private:
+    static u64 stoull_be(std::string_view str) {
+        if (str.size() != 16)
+            return 0;
+
+        const auto bytes = Common::HexStringToArray<0x8>(str);
+        u64 out{};
+        std::memcpy(&out, bytes.data(), sizeof(u64));
+
+        return Common::swap64(out);
+    }
+
+    void FindAllSaves(FileSys::SaveDataSpaceId space) {
+        const auto save_root = OpenSaveDataSpace(space);
+        ASSERT(save_root.Succeeded());
+
+        for (const auto& type : (*save_root)->GetSubdirectories()) {
+            if (type->GetName() == "save") {
+                for (const auto& save_id : type->GetSubdirectories()) {
+                    for (const auto& user_id : save_id->GetSubdirectories()) {
+                        const auto save_id_numeric = stoull_be(save_id->GetName());
+                        auto user_id_numeric = Common::HexStringToArray<0x10>(user_id->GetName());
+                        std::reverse(user_id_numeric.begin(), user_id_numeric.end());
+
+                        if (save_id_numeric != 0) {
+                            // System Save Data
+                            info.emplace_back(SaveDataInfo{
+                                0,
+                                space,
+                                FileSys::SaveDataType::SystemSaveData,
+                                {},
+                                user_id_numeric,
+                                save_id_numeric,
+                                0,
+                                user_id->GetSize(),
+                                {},
+                            });
+
+                            continue;
+                        }
+
+                        for (const auto& title_id : user_id->GetSubdirectories()) {
+                            const auto device =
+                                std::all_of(user_id_numeric.begin(), user_id_numeric.end(),
+                                            [](u8 val) { return val == 0; });
+                            info.emplace_back(SaveDataInfo{
+                                0,
+                                space,
+                                device ? FileSys::SaveDataType::DeviceSaveData
+                                       : FileSys::SaveDataType::SaveData,
+                                {},
+                                user_id_numeric,
+                                save_id_numeric,
+                                stoull_be(title_id->GetName()),
+                                title_id->GetSize(),
+                                {},
+                            });
+                        }
+                    }
+                }
+            } else if (space == FileSys::SaveDataSpaceId::TemporaryStorage) {
+                // Temporary Storage
+                for (const auto& user_id : type->GetSubdirectories()) {
+                    for (const auto& title_id : user_id->GetSubdirectories()) {
+                        if (!title_id->GetFiles().empty() ||
+                            !title_id->GetSubdirectories().empty()) {
+                            auto user_id_numeric =
+                                Common::HexStringToArray<0x10>(user_id->GetName());
+                            std::reverse(user_id_numeric.begin(), user_id_numeric.end());
+
+                            info.emplace_back(SaveDataInfo{
+                                0,
+                                space,
+                                FileSys::SaveDataType::TemporaryStorage,
+                                {},
+                                user_id_numeric,
+                                stoull_be(type->GetName()),
+                                stoull_be(title_id->GetName()),
+                                title_id->GetSize(),
+                                {},
+                            });
+                        }
+                    }
+                }
+            }
+        }
+    }
+
+    struct SaveDataInfo {
+        u64_le save_id_unknown;
+        FileSys::SaveDataSpaceId space;
+        FileSys::SaveDataType type;
+        INSERT_PADDING_BYTES(0x6);
+        std::array<u8, 0x10> user_id;
+        u64_le save_id;
+        u64_le title_id;
+        u64_le save_image_size;
+        INSERT_PADDING_BYTES(0x28);
+    };
+    static_assert(sizeof(SaveDataInfo) == 0x60, "SaveDataInfo has incorrect size.");
+
+    std::vector<SaveDataInfo> info;
+    u64 next_entry_index = 0;
+};
+
 FSP_SRV::FSP_SRV() : ServiceFramework("fsp-srv") {
+    // clang-format off
     static const FunctionInfo functions[] = {
         {0, nullptr, "MountContent"},
         {1, &FSP_SRV::Initialize, "Initialize"},
@@ -485,7 +626,7 @@ FSP_SRV::FSP_SRV() : ServiceFramework("fsp-srv") {
         {58, nullptr, "ReadSaveDataFileSystemExtraData"},
         {59, nullptr, "WriteSaveDataFileSystemExtraData"},
         {60, nullptr, "OpenSaveDataInfoReader"},
-        {61, nullptr, "OpenSaveDataInfoReaderBySaveDataSpaceId"},
+        {61, &FSP_SRV::OpenSaveDataInfoReaderBySaveDataSpaceId, "OpenSaveDataInfoReaderBySaveDataSpaceId"},
         {62, nullptr, "OpenCacheStorageList"},
         {64, nullptr, "OpenSaveDataInternalStorageFileSystem"},
         {65, nullptr, "UpdateSaveDataMacForDebug"},
@@ -544,6 +685,7 @@ FSP_SRV::FSP_SRV() : ServiceFramework("fsp-srv") {
         {1009, nullptr, "GetAndClearMemoryReportInfo"},
         {1100, nullptr, "OverrideSaveDataTransferTokenSignVerificationKey"},
     };
+    // clang-format on
     RegisterHandlers(functions);
 }
 
@@ -602,7 +744,7 @@ void FSP_SRV::MountSaveData(Kernel::HLERequestContext& ctx) {
 
     if (dir.Failed()) {
         IPC::ResponseBuilder rb{ctx, 2, 0, 0};
-        rb.Push(ResultCode(ErrorModule::FS, FileSys::ErrCodes::TitleNotFound));
+        rb.Push(FileSys::ERROR_ENTITY_NOT_FOUND);
         return;
     }
 
@@ -618,6 +760,15 @@ void FSP_SRV::OpenReadOnlySaveDataFileSystem(Kernel::HLERequestContext& ctx) {
     MountSaveData(ctx);
 }
 
+void FSP_SRV::OpenSaveDataInfoReaderBySaveDataSpaceId(Kernel::HLERequestContext& ctx) {
+    IPC::RequestParser rp{ctx};
+    const auto space = rp.PopRaw<FileSys::SaveDataSpaceId>();
+
+    IPC::ResponseBuilder rb{ctx, 2, 0, 1};
+    rb.Push(RESULT_SUCCESS);
+    rb.PushIpcInterface<ISaveDataInfoReader>(std::make_shared<ISaveDataInfoReader>(space));
+}
+
 void FSP_SRV::GetGlobalAccessLogMode(Kernel::HLERequestContext& ctx) {
     LOG_WARNING(Service_FS, "(STUBBED) called");
 
@@ -685,7 +836,7 @@ void FSP_SRV::OpenRomStorage(Kernel::HLERequestContext& ctx) {
               static_cast<u8>(storage_id), title_id);
 
     IPC::ResponseBuilder rb{ctx, 2};
-    rb.Push(ResultCode(ErrorModule::FS, FileSys::ErrCodes::TitleNotFound));
+    rb.Push(FileSys::ERROR_ENTITY_NOT_FOUND);
 }
 
 } // namespace Service::FileSystem

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

@@ -25,6 +25,7 @@ private:
     void CreateSaveData(Kernel::HLERequestContext& ctx);
     void MountSaveData(Kernel::HLERequestContext& ctx);
     void OpenReadOnlySaveDataFileSystem(Kernel::HLERequestContext& ctx);
+    void OpenSaveDataInfoReaderBySaveDataSpaceId(Kernel::HLERequestContext& ctx);
     void GetGlobalAccessLogMode(Kernel::HLERequestContext& ctx);
     void OpenDataStorageByCurrentProcess(Kernel::HLERequestContext& ctx);
     void OpenDataStorageByDataId(Kernel::HLERequestContext& ctx);

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

@@ -6,9 +6,14 @@
 #include "common/common_types.h"
 #include "core/core_timing.h"
 #include "core/hle/service/hid/controllers/debug_pad.h"
+#include "core/settings.h"
 
 namespace Service::HID {
 
+constexpr s32 HID_JOYSTICK_MAX = 0x7fff;
+constexpr s32 HID_JOYSTICK_MIN = -0x7fff;
+enum class JoystickId : std::size_t { Joystick_Left, Joystick_Right };
+
 Controller_DebugPad::Controller_DebugPad() = default;
 Controller_DebugPad::~Controller_DebugPad() = default;
 
@@ -33,10 +38,43 @@ void Controller_DebugPad::OnUpdate(u8* data, std::size_t size) {
 
     cur_entry.sampling_number = last_entry.sampling_number + 1;
     cur_entry.sampling_number2 = cur_entry.sampling_number;
-    // TODO(ogniK): Update debug pad states
+    cur_entry.attribute.connected.Assign(1);
+    auto& pad = cur_entry.pad_state;
+
+    using namespace Settings::NativeButton;
+    pad.a.Assign(buttons[A - BUTTON_HID_BEGIN]->GetStatus());
+    pad.b.Assign(buttons[B - BUTTON_HID_BEGIN]->GetStatus());
+    pad.x.Assign(buttons[X - BUTTON_HID_BEGIN]->GetStatus());
+    pad.y.Assign(buttons[Y - BUTTON_HID_BEGIN]->GetStatus());
+    pad.l.Assign(buttons[L - BUTTON_HID_BEGIN]->GetStatus());
+    pad.r.Assign(buttons[R - BUTTON_HID_BEGIN]->GetStatus());
+    pad.zl.Assign(buttons[ZL - BUTTON_HID_BEGIN]->GetStatus());
+    pad.zr.Assign(buttons[ZR - BUTTON_HID_BEGIN]->GetStatus());
+    pad.plus.Assign(buttons[Plus - BUTTON_HID_BEGIN]->GetStatus());
+    pad.minus.Assign(buttons[Minus - BUTTON_HID_BEGIN]->GetStatus());
+    pad.d_left.Assign(buttons[DLeft - BUTTON_HID_BEGIN]->GetStatus());
+    pad.d_up.Assign(buttons[DUp - BUTTON_HID_BEGIN]->GetStatus());
+    pad.d_right.Assign(buttons[DRight - BUTTON_HID_BEGIN]->GetStatus());
+    pad.d_down.Assign(buttons[DDown - BUTTON_HID_BEGIN]->GetStatus());
+
+    const auto [stick_l_x_f, stick_l_y_f] =
+        analogs[static_cast<std::size_t>(JoystickId::Joystick_Left)]->GetStatus();
+    const auto [stick_r_x_f, stick_r_y_f] =
+        analogs[static_cast<std::size_t>(JoystickId::Joystick_Right)]->GetStatus();
+    cur_entry.l_stick.x = static_cast<s32>(stick_l_x_f * HID_JOYSTICK_MAX);
+    cur_entry.l_stick.y = static_cast<s32>(stick_l_y_f * HID_JOYSTICK_MAX);
+    cur_entry.r_stick.x = static_cast<s32>(stick_r_x_f * HID_JOYSTICK_MAX);
+    cur_entry.r_stick.y = static_cast<s32>(stick_r_y_f * HID_JOYSTICK_MAX);
 
     std::memcpy(data, &shared_memory, sizeof(SharedMemory));
 }
 
-void Controller_DebugPad::OnLoadInputDevices() {}
+void Controller_DebugPad::OnLoadInputDevices() {
+    std::transform(Settings::values.debug_pad_buttons.begin(),
+                   Settings::values.debug_pad_buttons.end(), buttons.begin(),
+                   Input::CreateDevice<Input::ButtonDevice>);
+    std::transform(Settings::values.debug_pad_analogs.begin(),
+                   Settings::values.debug_pad_analogs.end(), analogs.begin(),
+                   Input::CreateDevice<Input::AnalogDevice>);
+}
 } // namespace Service::HID

src/core/hle/service/hid/controllers/debug_pad.h

@@ -5,10 +5,13 @@
 #pragma once
 
 #include <array>
+#include "common/bit_field.h"
 #include "common/common_funcs.h"
 #include "common/common_types.h"
 #include "common/swap.h"
+#include "core/frontend/input.h"
 #include "core/hle/service/hid/controllers/controller_base.h"
+#include "core/settings.h"
 
 namespace Service::HID {
 class Controller_DebugPad final : public ControllerBase {
@@ -35,11 +38,40 @@ private:
     };
     static_assert(sizeof(AnalogStick) == 0x8);
 
+    struct PadState {
+        union {
+            u32_le raw{};
+            BitField<0, 1, u32_le> a;
+            BitField<1, 1, u32_le> b;
+            BitField<2, 1, u32_le> x;
+            BitField<3, 1, u32_le> y;
+            BitField<4, 1, u32_le> l;
+            BitField<5, 1, u32_le> r;
+            BitField<6, 1, u32_le> zl;
+            BitField<7, 1, u32_le> zr;
+            BitField<8, 1, u32_le> plus;
+            BitField<9, 1, u32_le> minus;
+            BitField<10, 1, u32_le> d_left;
+            BitField<11, 1, u32_le> d_up;
+            BitField<12, 1, u32_le> d_right;
+            BitField<13, 1, u32_le> d_down;
+        };
+    };
+    static_assert(sizeof(PadState) == 0x4, "PadState is an invalid size");
+
+    struct Attributes {
+        union {
+            u32_le raw{};
+            BitField<0, 1, u32_le> connected;
+        };
+    };
+    static_assert(sizeof(Attributes) == 0x4, "Attributes is an invalid size");
+
     struct PadStates {
         s64_le sampling_number;
         s64_le sampling_number2;
-        u32_le attribute;
-        u32_le button_state;
+        Attributes attribute;
+        PadState pad_state;
         AnalogStick r_stick;
         AnalogStick l_stick;
     };
@@ -52,5 +84,10 @@ private:
     };
     static_assert(sizeof(SharedMemory) == 0x400, "SharedMemory is an invalid size");
     SharedMemory shared_memory{};
+
+    std::array<std::unique_ptr<Input::ButtonDevice>, Settings::NativeButton::NUM_BUTTONS_HID>
+        buttons;
+    std::array<std::unique_ptr<Input::AnalogDevice>, Settings::NativeAnalog::NUM_STICKS_HID>
+        analogs;
 };
 } // namespace Service::HID

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

@@ -6,9 +6,11 @@
 #include "common/common_types.h"
 #include "core/core_timing.h"
 #include "core/hle/service/hid/controllers/keyboard.h"
+#include "core/settings.h"
 
 namespace Service::HID {
 constexpr std::size_t SHARED_MEMORY_OFFSET = 0x3800;
+constexpr u8 KEYS_PER_BYTE = 8;
 
 Controller_Keyboard::Controller_Keyboard() = default;
 Controller_Keyboard::~Controller_Keyboard() = default;
@@ -34,10 +36,24 @@ void Controller_Keyboard::OnUpdate(u8* data, std::size_t size) {
 
     cur_entry.sampling_number = last_entry.sampling_number + 1;
     cur_entry.sampling_number2 = cur_entry.sampling_number;
-    // TODO(ogniK): Update keyboard states
+
+    for (std::size_t i = 0; i < keyboard_keys.size(); ++i) {
+        for (std::size_t k = 0; k < KEYS_PER_BYTE; ++k) {
+            cur_entry.key[i / KEYS_PER_BYTE] |= (keyboard_keys[i]->GetStatus() << k);
+        }
+    }
+
+    for (std::size_t i = 0; i < keyboard_mods.size(); ++i) {
+        cur_entry.modifier |= (keyboard_mods[i]->GetStatus() << i);
+    }
 
     std::memcpy(data + SHARED_MEMORY_OFFSET, &shared_memory, sizeof(SharedMemory));
 }
 
-void Controller_Keyboard::OnLoadInputDevices() {}
+void Controller_Keyboard::OnLoadInputDevices() {
+    std::transform(Settings::values.keyboard_keys.begin(), Settings::values.keyboard_keys.end(),
+                   keyboard_keys.begin(), Input::CreateDevice<Input::ButtonDevice>);
+    std::transform(Settings::values.keyboard_mods.begin(), Settings::values.keyboard_mods.end(),
+                   keyboard_mods.begin(), Input::CreateDevice<Input::ButtonDevice>);
+}
 } // namespace Service::HID

src/core/hle/service/hid/controllers/keyboard.h

@@ -8,7 +8,9 @@
 #include "common/common_funcs.h"
 #include "common/common_types.h"
 #include "common/swap.h"
+#include "core/frontend/input.h"
 #include "core/hle/service/hid/controllers/controller_base.h"
+#include "core/settings.h"
 
 namespace Service::HID {
 class Controller_Keyboard final : public ControllerBase {
@@ -46,5 +48,10 @@ private:
     };
     static_assert(sizeof(SharedMemory) == 0x400, "SharedMemory is an invalid size");
     SharedMemory shared_memory{};
+
+    std::array<std::unique_ptr<Input::ButtonDevice>, Settings::NativeKeyboard::NumKeyboardKeys>
+        keyboard_keys;
+    std::array<std::unique_ptr<Input::ButtonDevice>, Settings::NativeKeyboard::NumKeyboardMods>
+        keyboard_mods;
 };
 } // namespace Service::HID

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

@@ -5,6 +5,7 @@
 #include <cstring>
 #include "common/common_types.h"
 #include "core/core_timing.h"
+#include "core/frontend/emu_window.h"
 #include "core/hle/service/hid/controllers/mouse.h"
 
 namespace Service::HID {
@@ -14,7 +15,6 @@ Controller_Mouse::Controller_Mouse() = default;
 Controller_Mouse::~Controller_Mouse() = default;
 
 void Controller_Mouse::OnInit() {}
-
 void Controller_Mouse::OnRelease() {}
 
 void Controller_Mouse::OnUpdate(u8* data, std::size_t size) {
@@ -34,10 +34,29 @@ void Controller_Mouse::OnUpdate(u8* data, std::size_t size) {
 
     cur_entry.sampling_number = last_entry.sampling_number + 1;
     cur_entry.sampling_number2 = cur_entry.sampling_number;
-    // TODO(ogniK): Update mouse states
+
+    if (Settings::values.mouse_enabled) {
+        const auto [px, py, sx, sy] = mouse_device->GetStatus();
+        const auto x = static_cast<s32>(px * Layout::ScreenUndocked::Width);
+        const auto y = static_cast<s32>(py * Layout::ScreenUndocked::Height);
+        cur_entry.x = x;
+        cur_entry.y = y;
+        cur_entry.delta_x = x - last_entry.x;
+        cur_entry.delta_y = y - last_entry.y;
+        cur_entry.mouse_wheel_x = sx;
+        cur_entry.mouse_wheel_y = sy;
+
+        for (std::size_t i = 0; i < mouse_button_devices.size(); ++i) {
+            cur_entry.button |= (mouse_button_devices[i]->GetStatus() << i);
+        }
+    }
 
     std::memcpy(data + SHARED_MEMORY_OFFSET, &shared_memory, sizeof(SharedMemory));
 }
 
-void Controller_Mouse::OnLoadInputDevices() {}
+void Controller_Mouse::OnLoadInputDevices() {
+    mouse_device = Input::CreateDevice<Input::MouseDevice>(Settings::values.mouse_device);
+    std::transform(Settings::values.mouse_buttons.begin(), Settings::values.mouse_buttons.end(),
+                   mouse_button_devices.begin(), Input::CreateDevice<Input::ButtonDevice>);
+}
 } // namespace Service::HID

src/core/hle/service/hid/controllers/mouse.h

@@ -7,7 +7,9 @@
 #include <array>
 #include "common/common_types.h"
 #include "common/swap.h"
+#include "core/frontend/input.h"
 #include "core/hle/service/hid/controllers/controller_base.h"
+#include "core/settings.h"
 
 namespace Service::HID {
 class Controller_Mouse final : public ControllerBase {
@@ -35,7 +37,8 @@ private:
         s32_le y;
         s32_le delta_x;
         s32_le delta_y;
-        s32_le mouse_wheel;
+        s32_le mouse_wheel_x;
+        s32_le mouse_wheel_y;
         s32_le button;
         s32_le attribute;
     };
@@ -46,5 +49,9 @@ private:
         std::array<MouseState, 17> mouse_states;
     };
     SharedMemory shared_memory{};
+
+    std::unique_ptr<Input::MouseDevice> mouse_device;
+    std::array<std::unique_ptr<Input::ButtonDevice>, Settings::NativeMouseButton::NumMouseButtons>
+        mouse_button_devices;
 };
 } // namespace Service::HID

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

@@ -17,22 +17,13 @@
 #include "core/settings.h"
 
 namespace Service::HID {
-
-constexpr u32 JOYCON_BODY_NEON_RED = 0xFF3C28;
-constexpr u32 JOYCON_BUTTONS_NEON_RED = 0x1E0A0A;
-constexpr u32 JOYCON_BODY_NEON_BLUE = 0x0AB9E6;
-constexpr u32 JOYCON_BUTTONS_NEON_BLUE = 0x001E1E;
 constexpr s32 HID_JOYSTICK_MAX = 0x7fff;
 constexpr s32 HID_JOYSTICK_MIN = -0x7fff;
 constexpr std::size_t NPAD_OFFSET = 0x9A00;
 constexpr u32 BATTERY_FULL = 2;
-constexpr u32 NPAD_HANDHELD = 32;
-constexpr u32 NPAD_UNKNOWN = 16; // TODO(ogniK): What is this?
 constexpr u32 MAX_NPAD_ID = 7;
-constexpr Controller_NPad::NPadControllerType PREFERRED_CONTROLLER =
-    Controller_NPad::NPadControllerType::JoyDual;
 constexpr std::array<u32, 10> npad_id_list{
-    0, 1, 2, 3, 4, 5, 6, 7, 32, 16,
+    0, 1, 2, 3, 4, 5, 6, 7, NPAD_HANDHELD, NPAD_UNKNOWN,
 };
 
 enum class JoystickId : std::size_t {
@@ -40,6 +31,66 @@ enum class JoystickId : std::size_t {
     Joystick_Right,
 };
 
+static Controller_NPad::NPadControllerType MapSettingsTypeToNPad(Settings::ControllerType type) {
+    switch (type) {
+    case Settings::ControllerType::ProController:
+        return Controller_NPad::NPadControllerType::ProController;
+    case Settings::ControllerType::DualJoycon:
+        return Controller_NPad::NPadControllerType::JoyDual;
+    case Settings::ControllerType::LeftJoycon:
+        return Controller_NPad::NPadControllerType::JoyLeft;
+    case Settings::ControllerType::RightJoycon:
+        return Controller_NPad::NPadControllerType::JoyRight;
+    default:
+        UNREACHABLE();
+        return Controller_NPad::NPadControllerType::JoyDual;
+    }
+}
+
+std::size_t Controller_NPad::NPadIdToIndex(u32 npad_id) {
+    switch (npad_id) {
+    case 0:
+    case 1:
+    case 2:
+    case 3:
+    case 4:
+    case 5:
+    case 6:
+    case 7:
+        return npad_id;
+    case 8:
+    case NPAD_HANDHELD:
+        return 8;
+    case 9:
+    case NPAD_UNKNOWN:
+        return 9;
+    default:
+        UNIMPLEMENTED_MSG("Unknown npad id {}", npad_id);
+        return 0;
+    }
+}
+
+u32 Controller_NPad::IndexToNPad(std::size_t index) {
+    switch (index) {
+    case 0:
+    case 1:
+    case 2:
+    case 3:
+    case 4:
+    case 5:
+    case 6:
+    case 7:
+        return static_cast<u32>(index);
+    case 8:
+        return NPAD_HANDHELD;
+    case 9:
+        return NPAD_UNKNOWN;
+    default:
+        UNIMPLEMENTED_MSG("Unknown npad index {}", index);
+        return 0;
+    };
+}
+
 Controller_NPad::Controller_NPad() = default;
 Controller_NPad::~Controller_NPad() = default;
 
@@ -56,22 +107,32 @@ void Controller_NPad::InitNewlyAddedControler(std::size_t controller_idx) {
         controller.joy_styles.handheld.Assign(1);
         controller.device_type.handheld.Assign(1);
         controller.pad_assignment = NPadAssignments::Dual;
+        controller.properties.is_vertical.Assign(1);
+        controller.properties.use_plus.Assign(1);
+        controller.properties.use_minus.Assign(1);
         break;
     case NPadControllerType::JoyDual:
         controller.joy_styles.joycon_dual.Assign(1);
         controller.device_type.joycon_left.Assign(1);
         controller.device_type.joycon_right.Assign(1);
+        controller.properties.is_vertical.Assign(1);
+        controller.properties.use_plus.Assign(1);
+        controller.properties.use_minus.Assign(1);
         controller.pad_assignment = NPadAssignments::Dual;
         break;
     case NPadControllerType::JoyLeft:
         controller.joy_styles.joycon_left.Assign(1);
         controller.device_type.joycon_left.Assign(1);
-        controller.pad_assignment = NPadAssignments::Dual;
+        controller.properties.is_horizontal.Assign(1);
+        controller.properties.use_minus.Assign(1);
+        controller.pad_assignment = NPadAssignments::Single;
         break;
     case NPadControllerType::JoyRight:
         controller.joy_styles.joycon_right.Assign(1);
         controller.device_type.joycon_right.Assign(1);
-        controller.pad_assignment = NPadAssignments::Dual;
+        controller.properties.is_horizontal.Assign(1);
+        controller.properties.use_plus.Assign(1);
+        controller.pad_assignment = NPadAssignments::Single;
         break;
     case NPadControllerType::Pokeball:
         controller.joy_styles.pokeball.Assign(1);
@@ -81,6 +142,9 @@ void Controller_NPad::InitNewlyAddedControler(std::size_t controller_idx) {
     case NPadControllerType::ProController:
         controller.joy_styles.pro_controller.Assign(1);
         controller.device_type.pro_controller.Assign(1);
+        controller.properties.is_vertical.Assign(1);
+        controller.properties.use_plus.Assign(1);
+        controller.properties.use_minus.Assign(1);
         controller.pad_assignment = NPadAssignments::Single;
         break;
     }
@@ -90,14 +154,12 @@ void Controller_NPad::InitNewlyAddedControler(std::size_t controller_idx) {
     controller.single_color.button_color = 0;
 
     controller.dual_color_error = ColorReadError::ReadOk;
-    controller.left_color.body_color = JOYCON_BODY_NEON_BLUE;
-    controller.left_color.button_color = JOYCON_BUTTONS_NEON_BLUE;
-    controller.right_color.body_color = JOYCON_BODY_NEON_RED;
-    controller.right_color.button_color = JOYCON_BUTTONS_NEON_RED;
+    controller.left_color.body_color = Settings::values.players[controller_idx].body_color_left;
+    controller.left_color.button_color = Settings::values.players[controller_idx].button_color_left;
+    controller.right_color.body_color = Settings::values.players[controller_idx].body_color_right;
+    controller.right_color.button_color =
+        Settings::values.players[controller_idx].button_color_right;
 
-    controller.properties.is_vertical.Assign(1); // TODO(ogniK): Swap joycons orientations
-    controller.properties.use_plus.Assign(1);
-    controller.properties.use_minus.Assign(1);
     controller.battery_level[0] = BATTERY_FULL;
     controller.battery_level[1] = BATTERY_FULL;
     controller.battery_level[2] = BATTERY_FULL;
@@ -121,26 +183,109 @@ void Controller_NPad::OnInit() {
         style.pro_controller.Assign(1);
         style.pokeball.Assign(1);
     }
+
+    std::transform(
+        Settings::values.players.begin(), Settings::values.players.end(),
+        connected_controllers.begin(), [](const Settings::PlayerInput& player) {
+            return ControllerHolder{MapSettingsTypeToNPad(player.type), player.connected};
+        });
+
+    std::stable_partition(connected_controllers.begin(), connected_controllers.begin() + 8,
+                          [](const ControllerHolder& holder) { return holder.is_connected; });
+
+    // Account for handheld
+    if (connected_controllers[8].is_connected)
+        connected_controllers[8].type = NPadControllerType::Handheld;
+
+    supported_npad_id_types.resize(npad_id_list.size());
+    std::memcpy(supported_npad_id_types.data(), npad_id_list.data(),
+                npad_id_list.size() * sizeof(u32));
+
+    // Add a default dual joycon controller if none are present.
     if (std::none_of(connected_controllers.begin(), connected_controllers.end(),
                      [](const ControllerHolder& controller) { return controller.is_connected; })) {
         supported_npad_id_types.resize(npad_id_list.size());
         std::memcpy(supported_npad_id_types.data(), npad_id_list.data(),
                     npad_id_list.size() * sizeof(u32));
-        AddNewController(PREFERRED_CONTROLLER);
+        AddNewController(NPadControllerType::JoyDual);
+    }
+
+    for (std::size_t i = 0; i < connected_controllers.size(); ++i) {
+        const auto& controller = connected_controllers[i];
+        if (controller.is_connected) {
+            AddNewControllerAt(controller.type, IndexToNPad(i));
+        }
     }
 }
 
 void Controller_NPad::OnLoadInputDevices() {
-    std::transform(Settings::values.buttons.begin() + Settings::NativeButton::BUTTON_HID_BEGIN,
-                   Settings::values.buttons.begin() + Settings::NativeButton::BUTTON_HID_END,
-                   buttons.begin(), Input::CreateDevice<Input::ButtonDevice>);
-    std::transform(Settings::values.analogs.begin() + Settings::NativeAnalog::STICK_HID_BEGIN,
-                   Settings::values.analogs.begin() + Settings::NativeAnalog::STICK_HID_END,
-                   sticks.begin(), Input::CreateDevice<Input::AnalogDevice>);
+    const auto& players = Settings::values.players;
+    for (std::size_t i = 0; i < players.size(); ++i) {
+        std::transform(players[i].buttons.begin() + Settings::NativeButton::BUTTON_HID_BEGIN,
+                       players[i].buttons.begin() + Settings::NativeButton::BUTTON_HID_END,
+                       buttons[i].begin(), Input::CreateDevice<Input::ButtonDevice>);
+        std::transform(players[i].analogs.begin() + Settings::NativeAnalog::STICK_HID_BEGIN,
+                       players[i].analogs.begin() + Settings::NativeAnalog::STICK_HID_END,
+                       sticks[i].begin(), Input::CreateDevice<Input::AnalogDevice>);
+    }
 }
 
 void Controller_NPad::OnRelease() {}
 
+void Controller_NPad::RequestPadStateUpdate(u32 npad_id) {
+    const auto controller_idx = NPadIdToIndex(npad_id);
+    const auto controller_type = connected_controllers[controller_idx].type;
+    if (!connected_controllers[controller_idx].is_connected) {
+        return;
+    }
+    auto& pad_state = npad_pad_states[controller_idx].pad_states;
+    auto& lstick_entry = npad_pad_states[controller_idx].l_stick;
+    auto& rstick_entry = npad_pad_states[controller_idx].r_stick;
+    const auto& button_state = buttons[controller_idx];
+    const auto& analog_state = sticks[controller_idx];
+
+    using namespace Settings::NativeButton;
+    pad_state.a.Assign(button_state[A - BUTTON_HID_BEGIN]->GetStatus());
+    pad_state.b.Assign(button_state[B - BUTTON_HID_BEGIN]->GetStatus());
+    pad_state.x.Assign(button_state[X - BUTTON_HID_BEGIN]->GetStatus());
+    pad_state.y.Assign(button_state[Y - BUTTON_HID_BEGIN]->GetStatus());
+    pad_state.l_stick.Assign(button_state[LStick - BUTTON_HID_BEGIN]->GetStatus());
+    pad_state.r_stick.Assign(button_state[RStick - BUTTON_HID_BEGIN]->GetStatus());
+    pad_state.l.Assign(button_state[L - BUTTON_HID_BEGIN]->GetStatus());
+    pad_state.r.Assign(button_state[R - BUTTON_HID_BEGIN]->GetStatus());
+    pad_state.zl.Assign(button_state[ZL - BUTTON_HID_BEGIN]->GetStatus());
+    pad_state.zr.Assign(button_state[ZR - BUTTON_HID_BEGIN]->GetStatus());
+    pad_state.plus.Assign(button_state[Plus - BUTTON_HID_BEGIN]->GetStatus());
+    pad_state.minus.Assign(button_state[Minus - BUTTON_HID_BEGIN]->GetStatus());
+
+    pad_state.d_left.Assign(button_state[DLeft - BUTTON_HID_BEGIN]->GetStatus());
+    pad_state.d_up.Assign(button_state[DUp - BUTTON_HID_BEGIN]->GetStatus());
+    pad_state.d_right.Assign(button_state[DRight - BUTTON_HID_BEGIN]->GetStatus());
+    pad_state.d_down.Assign(button_state[DDown - BUTTON_HID_BEGIN]->GetStatus());
+
+    pad_state.l_stick_left.Assign(button_state[LStick_Left - BUTTON_HID_BEGIN]->GetStatus());
+    pad_state.l_stick_up.Assign(button_state[LStick_Up - BUTTON_HID_BEGIN]->GetStatus());
+    pad_state.l_stick_right.Assign(button_state[LStick_Right - BUTTON_HID_BEGIN]->GetStatus());
+    pad_state.l_stick_down.Assign(button_state[LStick_Down - BUTTON_HID_BEGIN]->GetStatus());
+
+    pad_state.r_stick_left.Assign(button_state[RStick_Left - BUTTON_HID_BEGIN]->GetStatus());
+    pad_state.r_stick_up.Assign(button_state[RStick_Up - BUTTON_HID_BEGIN]->GetStatus());
+    pad_state.r_stick_right.Assign(button_state[RStick_Right - BUTTON_HID_BEGIN]->GetStatus());
+    pad_state.r_stick_down.Assign(button_state[RStick_Down - BUTTON_HID_BEGIN]->GetStatus());
+
+    pad_state.left_sl.Assign(button_state[SL - BUTTON_HID_BEGIN]->GetStatus());
+    pad_state.left_sr.Assign(button_state[SR - BUTTON_HID_BEGIN]->GetStatus());
+
+    const auto [stick_l_x_f, stick_l_y_f] =
+        analog_state[static_cast<std::size_t>(JoystickId::Joystick_Left)]->GetStatus();
+    const auto [stick_r_x_f, stick_r_y_f] =
+        analog_state[static_cast<std::size_t>(JoystickId::Joystick_Right)]->GetStatus();
+    lstick_entry.x = static_cast<s32>(stick_l_x_f * HID_JOYSTICK_MAX);
+    lstick_entry.y = static_cast<s32>(stick_l_y_f * HID_JOYSTICK_MAX);
+    rstick_entry.x = static_cast<s32>(stick_r_x_f * HID_JOYSTICK_MAX);
+    rstick_entry.y = static_cast<s32>(stick_r_y_f * HID_JOYSTICK_MAX);
+}
+
 void Controller_NPad::OnUpdate(u8* data, std::size_t data_len) {
     if (!IsControllerActivated())
         return;
@@ -176,97 +321,9 @@ void Controller_NPad::OnUpdate(u8* data, std::size_t data_len) {
         if (controller_type == NPadControllerType::None || !connected_controllers[i].is_connected) {
             continue;
         }
-
-        // Pad states
-        ControllerPadState pad_state{};
-        using namespace Settings::NativeButton;
-        pad_state.a.Assign(buttons[A - BUTTON_HID_BEGIN]->GetStatus());
-        pad_state.b.Assign(buttons[B - BUTTON_HID_BEGIN]->GetStatus());
-        pad_state.x.Assign(buttons[X - BUTTON_HID_BEGIN]->GetStatus());
-        pad_state.y.Assign(buttons[Y - BUTTON_HID_BEGIN]->GetStatus());
-        pad_state.l_stick.Assign(buttons[LStick - BUTTON_HID_BEGIN]->GetStatus());
-        pad_state.r_stick.Assign(buttons[RStick - BUTTON_HID_BEGIN]->GetStatus());
-        pad_state.l.Assign(buttons[L - BUTTON_HID_BEGIN]->GetStatus());
-        pad_state.r.Assign(buttons[R - BUTTON_HID_BEGIN]->GetStatus());
-        pad_state.zl.Assign(buttons[ZL - BUTTON_HID_BEGIN]->GetStatus());
-        pad_state.zr.Assign(buttons[ZR - BUTTON_HID_BEGIN]->GetStatus());
-        pad_state.plus.Assign(buttons[Plus - BUTTON_HID_BEGIN]->GetStatus());
-        pad_state.minus.Assign(buttons[Minus - BUTTON_HID_BEGIN]->GetStatus());
-
-        pad_state.d_left.Assign(buttons[DLeft - BUTTON_HID_BEGIN]->GetStatus());
-        pad_state.d_up.Assign(buttons[DUp - BUTTON_HID_BEGIN]->GetStatus());
-        pad_state.d_right.Assign(buttons[DRight - BUTTON_HID_BEGIN]->GetStatus());
-        pad_state.d_down.Assign(buttons[DDown - BUTTON_HID_BEGIN]->GetStatus());
-
-        pad_state.l_stick_left.Assign(buttons[LStick_Left - BUTTON_HID_BEGIN]->GetStatus());
-        pad_state.l_stick_up.Assign(buttons[LStick_Up - BUTTON_HID_BEGIN]->GetStatus());
-        pad_state.l_stick_right.Assign(buttons[LStick_Right - BUTTON_HID_BEGIN]->GetStatus());
-        pad_state.l_stick_down.Assign(buttons[LStick_Down - BUTTON_HID_BEGIN]->GetStatus());
-
-        pad_state.r_stick_left.Assign(buttons[RStick_Left - BUTTON_HID_BEGIN]->GetStatus());
-        pad_state.r_stick_up.Assign(buttons[RStick_Up - BUTTON_HID_BEGIN]->GetStatus());
-        pad_state.r_stick_right.Assign(buttons[RStick_Right - BUTTON_HID_BEGIN]->GetStatus());
-        pad_state.r_stick_down.Assign(buttons[RStick_Down - BUTTON_HID_BEGIN]->GetStatus());
-
-        pad_state.sl.Assign(buttons[SL - BUTTON_HID_BEGIN]->GetStatus());
-        pad_state.sr.Assign(buttons[SR - BUTTON_HID_BEGIN]->GetStatus());
-
-        AnalogPosition lstick_entry{};
-        AnalogPosition rstick_entry{};
-
-        const auto [stick_l_x_f, stick_l_y_f] =
-            sticks[static_cast<std::size_t>(JoystickId::Joystick_Left)]->GetStatus();
-        const auto [stick_r_x_f, stick_r_y_f] =
-            sticks[static_cast<std::size_t>(JoystickId::Joystick_Right)]->GetStatus();
-        lstick_entry.x = static_cast<s32>(stick_l_x_f * HID_JOYSTICK_MAX);
-        lstick_entry.y = static_cast<s32>(stick_l_y_f * HID_JOYSTICK_MAX);
-        rstick_entry.x = static_cast<s32>(stick_r_x_f * HID_JOYSTICK_MAX);
-        rstick_entry.y = static_cast<s32>(stick_r_y_f * HID_JOYSTICK_MAX);
-
-        if (controller_type == NPadControllerType::JoyLeft ||
-            controller_type == NPadControllerType::JoyRight) {
-            if (npad.properties.is_horizontal) {
-                ControllerPadState state{};
-                AnalogPosition temp_lstick_entry{};
-                AnalogPosition temp_rstick_entry{};
-                if (controller_type == NPadControllerType::JoyLeft) {
-                    state.d_down.Assign(pad_state.d_left.Value());
-                    state.d_left.Assign(pad_state.d_up.Value());
-                    state.d_right.Assign(pad_state.d_down.Value());
-                    state.d_up.Assign(pad_state.d_right.Value());
-                    state.l.Assign(pad_state.l.Value() | pad_state.sl.Value());
-                    state.r.Assign(pad_state.r.Value() | pad_state.sr.Value());
-
-                    state.zl.Assign(pad_state.zl.Value());
-                    state.plus.Assign(pad_state.minus.Value());
-
-                    temp_lstick_entry = lstick_entry;
-                    temp_rstick_entry = rstick_entry;
-                    std::swap(temp_lstick_entry.x, temp_lstick_entry.y);
-                    std::swap(temp_rstick_entry.x, temp_rstick_entry.y);
-                    temp_lstick_entry.y *= -1;
-                } else if (controller_type == NPadControllerType::JoyRight) {
-                    state.x.Assign(pad_state.a.Value());
-                    state.a.Assign(pad_state.b.Value());
-                    state.b.Assign(pad_state.y.Value());
-                    state.y.Assign(pad_state.b.Value());
-
-                    state.l.Assign(pad_state.l.Value() | pad_state.sl.Value());
-                    state.r.Assign(pad_state.r.Value() | pad_state.sr.Value());
-                    state.zr.Assign(pad_state.zr.Value());
-                    state.plus.Assign(pad_state.plus.Value());
-
-                    temp_lstick_entry = lstick_entry;
-                    temp_rstick_entry = rstick_entry;
-                    std::swap(temp_lstick_entry.x, temp_lstick_entry.y);
-                    std::swap(temp_rstick_entry.x, temp_rstick_entry.y);
-                    temp_rstick_entry.x *= -1;
-                }
-                pad_state.raw = state.raw;
-                lstick_entry = temp_lstick_entry;
-                rstick_entry = temp_rstick_entry;
-            }
-        }
+        const u32 npad_index = static_cast<u32>(i);
+        RequestPadStateUpdate(npad_index);
+        auto& pad_state = npad_pad_states[npad_index];
 
         auto& main_controller =
             npad.main_controller_states.npad[npad.main_controller_states.common.last_entry_index];
@@ -281,20 +338,64 @@ void Controller_NPad::OnUpdate(u8* data, std::size_t data_len) {
         auto& libnx_entry = npad.libnx.npad[npad.libnx.common.last_entry_index];
 
         if (hold_type == NpadHoldType::Horizontal) {
-            // TODO(ogniK): Remap buttons for different orientations
+            ControllerPadState state{};
+            AnalogPosition temp_lstick_entry{};
+            AnalogPosition temp_rstick_entry{};
+            if (controller_type == NPadControllerType::JoyLeft) {
+                state.d_down.Assign(pad_state.pad_states.d_left.Value());
+                state.d_left.Assign(pad_state.pad_states.d_up.Value());
+                state.d_right.Assign(pad_state.pad_states.d_down.Value());
+                state.d_up.Assign(pad_state.pad_states.d_right.Value());
+                state.l.Assign(pad_state.pad_states.l.Value() |
+                               pad_state.pad_states.left_sl.Value());
+                state.r.Assign(pad_state.pad_states.r.Value() |
+                               pad_state.pad_states.left_sr.Value());
+
+                state.zl.Assign(pad_state.pad_states.zl.Value());
+                state.plus.Assign(pad_state.pad_states.minus.Value());
+
+                temp_lstick_entry = pad_state.l_stick;
+                temp_rstick_entry = pad_state.r_stick;
+                std::swap(temp_lstick_entry.x, temp_lstick_entry.y);
+                std::swap(temp_rstick_entry.x, temp_rstick_entry.y);
+                temp_lstick_entry.y *= -1;
+            } else if (controller_type == NPadControllerType::JoyRight) {
+                state.x.Assign(pad_state.pad_states.a.Value());
+                state.a.Assign(pad_state.pad_states.b.Value());
+                state.b.Assign(pad_state.pad_states.y.Value());
+                state.y.Assign(pad_state.pad_states.b.Value());
+
+                state.l.Assign(pad_state.pad_states.l.Value() |
+                               pad_state.pad_states.right_sl.Value());
+                state.r.Assign(pad_state.pad_states.r.Value() |
+                               pad_state.pad_states.right_sr.Value());
+                state.zr.Assign(pad_state.pad_states.zr.Value());
+                state.plus.Assign(pad_state.pad_states.plus.Value());
+
+                temp_lstick_entry = pad_state.l_stick;
+                temp_rstick_entry = pad_state.r_stick;
+                std::swap(temp_lstick_entry.x, temp_lstick_entry.y);
+                std::swap(temp_rstick_entry.x, temp_rstick_entry.y);
+                temp_rstick_entry.x *= -1;
+            }
+            pad_state.pad_states.raw = state.raw;
+            pad_state.l_stick = temp_lstick_entry;
+            pad_state.r_stick = temp_rstick_entry;
         }
+
         libnx_entry.connection_status.raw = 0;
 
         switch (controller_type) {
         case NPadControllerType::Handheld:
             handheld_entry.connection_status.raw = 0;
-            handheld_entry.connection_status.IsConnected.Assign(1);
-            if (!Settings::values.use_docked_mode) {
-                handheld_entry.connection_status.IsWired.Assign(1);
-            }
-            handheld_entry.pad_states.raw = pad_state.raw;
-            handheld_entry.l_stick = lstick_entry;
-            handheld_entry.r_stick = rstick_entry;
+            handheld_entry.connection_status.IsWired.Assign(1);
+            handheld_entry.connection_status.IsLeftJoyConnected.Assign(1);
+            handheld_entry.connection_status.IsRightJoyConnected.Assign(1);
+            handheld_entry.connection_status.IsLeftJoyWired.Assign(1);
+            handheld_entry.connection_status.IsRightJoyWired.Assign(1);
+            handheld_entry.pad.pad_states.raw = pad_state.pad_states.raw;
+            handheld_entry.pad.l_stick = pad_state.l_stick;
+            handheld_entry.pad.r_stick = pad_state.r_stick;
             break;
         case NPadControllerType::JoyDual:
             dual_entry.connection_status.raw = 0;
@@ -307,24 +408,25 @@ void Controller_NPad::OnUpdate(u8* data, std::size_t data_len) {
             libnx_entry.connection_status.IsRightJoyConnected.Assign(1);
             libnx_entry.connection_status.IsConnected.Assign(1);
 
-            dual_entry.pad_states.raw = pad_state.raw;
-            dual_entry.l_stick = lstick_entry;
-            dual_entry.r_stick = rstick_entry;
+            dual_entry.pad.pad_states.raw = pad_state.pad_states.raw;
+            dual_entry.pad.l_stick = pad_state.l_stick;
+            dual_entry.pad.r_stick = pad_state.r_stick;
+            break;
         case NPadControllerType::JoyLeft:
             left_entry.connection_status.raw = 0;
 
             left_entry.connection_status.IsConnected.Assign(1);
-            left_entry.pad_states.raw = pad_state.raw;
-            left_entry.l_stick = lstick_entry;
-            left_entry.r_stick = rstick_entry;
+            left_entry.pad.pad_states.raw = pad_state.pad_states.raw;
+            left_entry.pad.l_stick = pad_state.l_stick;
+            left_entry.pad.r_stick = pad_state.r_stick;
             break;
         case NPadControllerType::JoyRight:
             right_entry.connection_status.raw = 0;
 
             right_entry.connection_status.IsConnected.Assign(1);
-            right_entry.pad_states.raw = pad_state.raw;
-            right_entry.l_stick = lstick_entry;
-            right_entry.r_stick = rstick_entry;
+            right_entry.pad.pad_states.raw = pad_state.pad_states.raw;
+            right_entry.pad.l_stick = pad_state.l_stick;
+            right_entry.pad.r_stick = pad_state.r_stick;
             break;
         case NPadControllerType::Pokeball:
             pokeball_entry.connection_status.raw = 0;
@@ -332,30 +434,30 @@ void Controller_NPad::OnUpdate(u8* data, std::size_t data_len) {
             pokeball_entry.connection_status.IsConnected.Assign(1);
             pokeball_entry.connection_status.IsWired.Assign(1);
 
-            pokeball_entry.pad_states.raw = pad_state.raw;
-            pokeball_entry.l_stick = lstick_entry;
-            pokeball_entry.r_stick = rstick_entry;
+            pokeball_entry.pad.pad_states.raw = pad_state.pad_states.raw;
+            pokeball_entry.pad.l_stick = pad_state.l_stick;
+            pokeball_entry.pad.r_stick = pad_state.r_stick;
             break;
         case NPadControllerType::ProController:
             main_controller.connection_status.raw = 0;
 
             main_controller.connection_status.IsConnected.Assign(1);
             main_controller.connection_status.IsWired.Assign(1);
-            main_controller.pad_states.raw = pad_state.raw;
-            main_controller.l_stick = lstick_entry;
-            main_controller.r_stick = rstick_entry;
+            main_controller.pad.pad_states.raw = pad_state.pad_states.raw;
+            main_controller.pad.l_stick = pad_state.l_stick;
+            main_controller.pad.r_stick = pad_state.r_stick;
             break;
         }
 
         // LibNX exclusively uses this section, so we always update it since LibNX doesn't activate
         // any controllers.
-        libnx_entry.pad_states.raw = pad_state.raw;
-        libnx_entry.l_stick = lstick_entry;
-        libnx_entry.r_stick = rstick_entry;
+        libnx_entry.pad.pad_states.raw = pad_state.pad_states.raw;
+        libnx_entry.pad.l_stick = pad_state.l_stick;
+        libnx_entry.pad.r_stick = pad_state.r_stick;
     }
     std::memcpy(data + NPAD_OFFSET, shared_memory_entries.data(),
                 shared_memory_entries.size() * sizeof(NPadEntry));
-} // namespace Service::HID
+}
 
 void Controller_NPad::SetSupportedStyleSet(NPadType style_set) {
     style.raw = style_set.raw;
@@ -370,14 +472,29 @@ void Controller_NPad::SetSupportedNPadIdTypes(u8* data, std::size_t length) {
     supported_npad_id_types.clear();
     supported_npad_id_types.resize(length / sizeof(u32));
     std::memcpy(supported_npad_id_types.data(), data, length);
+    bool had_controller_update = false;
     for (std::size_t i = 0; i < connected_controllers.size(); i++) {
         auto& controller = connected_controllers[i];
         if (!controller.is_connected) {
             continue;
         }
-        if (!IsControllerSupported(PREFERRED_CONTROLLER)) {
-            controller.type = DecideBestController(PREFERRED_CONTROLLER);
-            InitNewlyAddedControler(i);
+        const auto requested_controller =
+            i <= MAX_NPAD_ID ? MapSettingsTypeToNPad(Settings::values.players[i].type)
+                             : NPadControllerType::Handheld;
+        if (!IsControllerSupported(requested_controller)) {
+            const auto is_handheld = requested_controller == NPadControllerType::Handheld;
+            if (is_handheld) {
+                controller.type = NPadControllerType::None;
+                controller.is_connected = false;
+                AddNewController(requested_controller);
+            } else {
+                controller.type = requested_controller;
+                InitNewlyAddedControler(i);
+            }
+            had_controller_update = true;
+        }
+        if (had_controller_update) {
+            styleset_changed_event->Signal();
         }
     }
 }
@@ -411,15 +528,7 @@ void Controller_NPad::VibrateController(const std::vector<u32>& controller_ids,
         return;
     }
     for (std::size_t i = 0; i < controller_ids.size(); i++) {
-        std::size_t controller_pos = i;
-        // Handheld controller conversion
-        if (controller_pos == NPAD_HANDHELD) {
-            controller_pos = 8;
-        }
-        // Unknown controller conversion
-        if (controller_pos == NPAD_UNKNOWN) {
-            controller_pos = 9;
-        }
+        std::size_t controller_pos = NPadIdToIndex(static_cast<u32>(i));
         if (connected_controllers[controller_pos].is_connected) {
             // TODO(ogniK): Vibrate the physical controller
         }
@@ -438,7 +547,9 @@ Kernel::SharedPtr<Kernel::Event> Controller_NPad::GetStyleSetChangedEvent() cons
 Controller_NPad::Vibration Controller_NPad::GetLastVibration() const {
     return last_processed_vibration;
 }
+
 void Controller_NPad::AddNewController(NPadControllerType controller) {
+    controller = DecideBestController(controller);
     if (controller == NPadControllerType::Handheld) {
         connected_controllers[8] = {controller, true};
         InitNewlyAddedControler(8);
@@ -456,16 +567,54 @@ void Controller_NPad::AddNewController(NPadControllerType controller) {
     InitNewlyAddedControler(controller_id);
 }
 
-void Controller_NPad::ConnectNPad(u32 npad_id) {
-    if (npad_id >= connected_controllers.size())
+void Controller_NPad::AddNewControllerAt(NPadControllerType controller, u32 npad_id) {
+    controller = DecideBestController(controller);
+    if (controller == NPadControllerType::Handheld) {
+        connected_controllers[NPadIdToIndex(NPAD_HANDHELD)] = {controller, true};
+        InitNewlyAddedControler(NPadIdToIndex(NPAD_HANDHELD));
         return;
-    connected_controllers[npad_id].is_connected = true;
+    }
+
+    connected_controllers[npad_id] = {controller, true};
+    InitNewlyAddedControler(npad_id);
+}
+
+void Controller_NPad::ConnectNPad(u32 npad_id) {
+    connected_controllers[NPadIdToIndex(npad_id)].is_connected = true;
 }
 
 void Controller_NPad::DisconnectNPad(u32 npad_id) {
-    if (npad_id >= connected_controllers.size())
-        return;
-    connected_controllers[npad_id].is_connected = false;
+    connected_controllers[NPadIdToIndex(npad_id)].is_connected = false;
+}
+
+bool Controller_NPad::IsControllerSupported(NPadControllerType controller) {
+    if (controller == NPadControllerType::Handheld) {
+        // Handheld is not even a supported type, lets stop here
+        if (std::find(supported_npad_id_types.begin(), supported_npad_id_types.end(),
+                      NPAD_HANDHELD) == supported_npad_id_types.end()) {
+            return false;
+        }
+        // Handheld should not be supported in docked mode
+        if (Settings::values.use_docked_mode) {
+            return false;
+        }
+    }
+    switch (controller) {
+    case NPadControllerType::ProController:
+        return style.pro_controller;
+    case NPadControllerType::Handheld:
+        return style.handheld;
+    case NPadControllerType::JoyDual:
+        return style.joycon_dual;
+    case NPadControllerType::JoyLeft:
+        return style.joycon_left;
+    case NPadControllerType::JoyRight:
+        return style.joycon_right;
+    case NPadControllerType::Pokeball:
+        return style.pokeball;
+    default:
+        return false;
+    }
 }
 
 Controller_NPad::LedPattern Controller_NPad::GetLedPattern(u32 npad_id) {
@@ -499,6 +648,36 @@ void Controller_NPad::SetVibrationEnabled(bool can_vibrate) {
     can_controllers_vibrate = can_vibrate;
 }
 
+void Controller_NPad::ClearAllConnectedControllers() {
+    for (auto& controller : connected_controllers) {
+        if (controller.is_connected && controller.type != NPadControllerType::None) {
+            controller.type = NPadControllerType::None;
+            controller.is_connected = false;
+        }
+    }
+}
+void Controller_NPad::DisconnectAllConnectedControllers() {
+    std::for_each(connected_controllers.begin(), connected_controllers.end(),
+                  [](ControllerHolder& controller) { controller.is_connected = false; });
+}
+
+void Controller_NPad::ConnectAllDisconnectedControllers() {
+    std::for_each(connected_controllers.begin(), connected_controllers.end(),
+                  [](ControllerHolder& controller) {
+                      if (controller.type != NPadControllerType::None && !controller.is_connected) {
+                          controller.is_connected = false;
+                      }
+                  });
+}
+
+void Controller_NPad::ClearAllControllers() {
+    std::for_each(connected_controllers.begin(), connected_controllers.end(),
+                  [](ControllerHolder& controller) {
+                      controller.type = NPadControllerType::None;
+                      controller.is_connected = false;
+                  });
+}
+
 bool Controller_NPad::IsControllerSupported(NPadControllerType controller) const {
     const bool support_handheld =
         std::find(supported_npad_id_types.begin(), supported_npad_id_types.end(), NPAD_HANDHELD) !=

src/core/hle/service/hid/controllers/npad.h

@@ -5,13 +5,18 @@
 #pragma once
 
 #include <array>
+#include "common/bit_field.h"
 #include "common/common_types.h"
 #include "core/frontend/input.h"
+#include "core/hle/kernel/event.h"
 #include "core/hle/service/hid/controllers/controller_base.h"
 #include "core/settings.h"
 
 namespace Service::HID {
 
+constexpr u32 NPAD_HANDHELD = 32;
+constexpr u32 NPAD_UNKNOWN = 16; // TODO(ogniK): What is this?
+
 class Controller_NPad final : public ControllerBase {
 public:
     Controller_NPad();
@@ -107,11 +112,19 @@ public:
     Vibration GetLastVibration() const;
 
     void AddNewController(NPadControllerType controller);
+    void AddNewControllerAt(NPadControllerType controller, u32 npad_id);
 
     void ConnectNPad(u32 npad_id);
     void DisconnectNPad(u32 npad_id);
     LedPattern GetLedPattern(u32 npad_id);
     void SetVibrationEnabled(bool can_vibrate);
+    void ClearAllConnectedControllers();
+    void DisconnectAllConnectedControllers();
+    void ConnectAllDisconnectedControllers();
+    void ClearAllControllers();
+
+    static std::size_t NPadIdToIndex(u32 npad_id);
+    static u32 IndexToNPad(std::size_t index);
 
 private:
     struct CommonHeader {
@@ -164,8 +177,11 @@ private:
             BitField<23, 1, u64_le> r_stick_down;
 
             // Not always active?
-            BitField<24, 1, u64_le> sl;
-            BitField<25, 1, u64_le> sr;
+            BitField<24, 1, u64_le> left_sl;
+            BitField<25, 1, u64_le> left_sr;
+
+            BitField<26, 1, u64_le> right_sl;
+            BitField<27, 1, u64_le> right_sr;
         };
     };
     static_assert(sizeof(ControllerPadState) == 8, "ControllerPadState is an invalid size");
@@ -189,12 +205,17 @@ private:
     };
     static_assert(sizeof(ConnectionState) == 4, "ConnectionState is an invalid size");
 
-    struct GenericStates {
-        s64_le timestamp;
-        s64_le timestamp2;
+    struct ControllerPad {
         ControllerPadState pad_states;
         AnalogPosition l_stick;
         AnalogPosition r_stick;
+    };
+    static_assert(sizeof(ControllerPad) == 0x18, "ControllerPad is an invalid size");
+
+    struct GenericStates {
+        s64_le timestamp;
+        s64_le timestamp2;
+        ControllerPad pad;
         ConnectionState connection_status;
     };
     static_assert(sizeof(GenericStates) == 0x30, "NPadGenericStates is an invalid size");
@@ -266,15 +287,20 @@ private:
     static_assert(sizeof(NPadEntry) == 0x5000, "NPadEntry is an invalid size");
 
     struct ControllerHolder {
-        Controller_NPad::NPadControllerType type;
+        NPadControllerType type;
         bool is_connected;
     };
 
     NPadType style{};
     std::array<NPadEntry, 10> shared_memory_entries{};
-    std::array<std::unique_ptr<Input::ButtonDevice>, Settings::NativeButton::NUM_BUTTONS_HID>
+    std::array<
+        std::array<std::unique_ptr<Input::ButtonDevice>, Settings::NativeButton::NUM_BUTTONS_HID>,
+        10>
         buttons;
-    std::array<std::unique_ptr<Input::AnalogDevice>, Settings::NativeAnalog::NUM_STICKS_HID> sticks;
+    std::array<
+        std::array<std::unique_ptr<Input::AnalogDevice>, Settings::NativeAnalog::NUM_STICKS_HID>,
+        10>
+        sticks;
     std::vector<u32> supported_npad_id_types{};
     NpadHoldType hold_type{NpadHoldType::Vertical};
     Kernel::SharedPtr<Kernel::Event> styleset_changed_event;
@@ -285,5 +311,8 @@ private:
     void InitNewlyAddedControler(std::size_t controller_idx);
     bool IsControllerSupported(NPadControllerType controller) const;
     NPadControllerType DecideBestController(NPadControllerType priority) const;
+    void RequestPadStateUpdate(u32 npad_id);
+    std::array<ControllerPad, 10> npad_pad_states{};
+    bool IsControllerSupported(NPadControllerType controller);
 };
 } // namespace Service::HID

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

@@ -41,16 +41,17 @@ void Controller_Touchscreen::OnUpdate(u8* data, std::size_t size) {
 
     const auto [x, y, pressed] = touch_device->GetStatus();
     auto& touch_entry = cur_entry.states[0];
-    if (pressed) {
+    touch_entry.attribute.raw = 0;
+    if (pressed && Settings::values.touchscreen.enabled) {
         touch_entry.x = static_cast<u16>(x * Layout::ScreenUndocked::Width);
         touch_entry.y = static_cast<u16>(y * Layout::ScreenUndocked::Height);
-        touch_entry.diameter_x = 15;
-        touch_entry.diameter_y = 15;
-        touch_entry.rotation_angle = 0;
+        touch_entry.diameter_x = Settings::values.touchscreen.diameter_x;
+        touch_entry.diameter_y = Settings::values.touchscreen.diameter_y;
+        touch_entry.rotation_angle = Settings::values.touchscreen.rotation_angle;
         const u64 tick = CoreTiming::GetTicks();
         touch_entry.delta_time = tick - last_touch;
         last_touch = tick;
-        touch_entry.finger = 0;
+        touch_entry.finger = Settings::values.touchscreen.finger;
         cur_entry.entry_count = 1;
     } else {
         cur_entry.entry_count = 0;
@@ -60,6 +61,6 @@ void Controller_Touchscreen::OnUpdate(u8* data, std::size_t size) {
 }
 
 void Controller_Touchscreen::OnLoadInputDevices() {
-    touch_device = Input::CreateDevice<Input::TouchDevice>(Settings::values.touch_device);
+    touch_device = Input::CreateDevice<Input::TouchDevice>(Settings::values.touchscreen.device);
 }
 } // namespace Service::HID

src/core/hle/service/hid/controllers/touchscreen.h

@@ -4,6 +4,7 @@
 
 #pragma once
 
+#include "common/bit_field.h"
 #include "common/common_funcs.h"
 #include "common/common_types.h"
 #include "common/swap.h"
@@ -29,9 +30,18 @@ public:
     void OnLoadInputDevices() override;
 
 private:
+    struct Attributes {
+        union {
+            u32 raw{};
+            BitField<0, 1, u32_le> start_touch;
+            BitField<1, 1, u32_le> end_touch;
+        };
+    };
+    static_assert(sizeof(Attributes) == 0x4, "Attributes is an invalid size");
+
     struct TouchState {
         u64_le delta_time;
-        u32_le attribute;
+        Attributes attribute;
         u32_le finger;
         u32_le x;
         u32_le y;

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

@@ -34,8 +34,8 @@
 namespace Service::HID {
 
 // Updating period for each HID device.
-// TODO(shinyquagsire23): These need better values.
-constexpr u64 pad_update_ticks = CoreTiming::BASE_CLOCK_RATE / 100;
+// TODO(ogniK): Find actual polling rate of hid
+constexpr u64 pad_update_ticks = CoreTiming::BASE_CLOCK_RATE / 66;
 constexpr u64 accelerometer_update_ticks = CoreTiming::BASE_CLOCK_RATE / 100;
 constexpr u64 gyroscope_update_ticks = CoreTiming::BASE_CLOCK_RATE / 100;
 constexpr std::size_t SHARED_MEMORY_SIZE = 0x40000;
@@ -286,10 +286,10 @@ public:
             {519, nullptr, "GetPalmaOperationResult"},
             {520, nullptr, "ReadPalmaPlayLog"},
             {521, nullptr, "ResetPalmaPlayLog"},
-            {522, nullptr, "SetIsPalmaAllConnectable"},
+            {522, &Hid::SetIsPalmaAllConnectable, "SetIsPalmaAllConnectable"},
             {523, nullptr, "SetIsPalmaPairedConnectable"},
             {524, nullptr, "PairPalma"},
-            {525, nullptr, "SetPalmaBoostMode"},
+            {525, &Hid::SetPalmaBoostMode, "SetPalmaBoostMode"},
             {1000, nullptr, "SetNpadCommunicationMode"},
             {1001, nullptr, "GetNpadCommunicationMode"},
         };
@@ -596,6 +596,18 @@ private:
         rb.Push(RESULT_SUCCESS);
         LOG_WARNING(Service_HID, "(STUBBED) called");
     }
+
+    void SetIsPalmaAllConnectable(Kernel::HLERequestContext& ctx) {
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(RESULT_SUCCESS);
+        LOG_WARNING(Service_HID, "(STUBBED) called");
+    }
+
+    void SetPalmaBoostMode(Kernel::HLERequestContext& ctx) {
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(RESULT_SUCCESS);
+        LOG_WARNING(Service_HID, "(STUBBED) called");
+    }
 };
 
 class HidDbg final : public ServiceFramework<HidDbg> {

src/core/hle/service/ldr/ldr.cpp

@@ -4,7 +4,10 @@
 
 #include <memory>
 #include <fmt/format.h>
+#include <mbedtls/sha256.h>
 
+#include "common/alignment.h"
+#include "common/hex_util.h"
 #include "core/hle/ipc_helpers.h"
 #include "core/hle/kernel/process.h"
 #include "core/hle/service/ldr/ldr.h"
@@ -13,6 +16,38 @@
 
 namespace Service::LDR {
 
+namespace ErrCodes {
+enum {
+    InvalidMemoryState = 51,
+    InvalidNRO = 52,
+    InvalidNRR = 53,
+    MissingNRRHash = 54,
+    MaximumNRO = 55,
+    MaximumNRR = 56,
+    AlreadyLoaded = 57,
+    InvalidAlignment = 81,
+    InvalidSize = 82,
+    InvalidNROAddress = 84,
+    InvalidNRRAddress = 85,
+    NotInitialized = 87,
+};
+}
+
+constexpr ResultCode ERROR_INVALID_MEMORY_STATE(ErrorModule::Loader, ErrCodes::InvalidMemoryState);
+constexpr ResultCode ERROR_INVALID_NRO(ErrorModule::Loader, ErrCodes::InvalidNRO);
+constexpr ResultCode ERROR_INVALID_NRR(ErrorModule::Loader, ErrCodes::InvalidNRR);
+constexpr ResultCode ERROR_MISSING_NRR_HASH(ErrorModule::Loader, ErrCodes::MissingNRRHash);
+constexpr ResultCode ERROR_MAXIMUM_NRO(ErrorModule::Loader, ErrCodes::MaximumNRO);
+constexpr ResultCode ERROR_MAXIMUM_NRR(ErrorModule::Loader, ErrCodes::MaximumNRR);
+constexpr ResultCode ERROR_ALREADY_LOADED(ErrorModule::Loader, ErrCodes::AlreadyLoaded);
+constexpr ResultCode ERROR_INVALID_ALIGNMENT(ErrorModule::Loader, ErrCodes::InvalidAlignment);
+constexpr ResultCode ERROR_INVALID_SIZE(ErrorModule::Loader, ErrCodes::InvalidSize);
+constexpr ResultCode ERROR_INVALID_NRO_ADDRESS(ErrorModule::Loader, ErrCodes::InvalidNROAddress);
+constexpr ResultCode ERROR_INVALID_NRR_ADDRESS(ErrorModule::Loader, ErrCodes::InvalidNRRAddress);
+constexpr ResultCode ERROR_NOT_INITIALIZED(ErrorModule::Loader, ErrCodes::NotInitialized);
+
+constexpr u64 MAXIMUM_LOADED_RO = 0x40;
+
 class DebugMonitor final : public ServiceFramework<DebugMonitor> {
 public:
     explicit DebugMonitor() : ServiceFramework{"ldr:dmnt"} {
@@ -64,9 +99,9 @@ public:
         // clang-format off
         static const FunctionInfo functions[] = {
             {0, &RelocatableObject::LoadNro, "LoadNro"},
-            {1, nullptr, "UnloadNro"},
+            {1, &RelocatableObject::UnloadNro, "UnloadNro"},
             {2, &RelocatableObject::LoadNrr, "LoadNrr"},
-            {3, nullptr, "UnloadNrr"},
+            {3, &RelocatableObject::UnloadNrr, "UnloadNrr"},
             {4, &RelocatableObject::Initialize, "Initialize"},
         };
         // clang-format on
@@ -75,9 +110,123 @@ public:
     }
 
     void LoadNrr(Kernel::HLERequestContext& ctx) {
+        IPC::RequestParser rp{ctx};
+        rp.Skip(2, false);
+        const VAddr nrr_addr{rp.Pop<VAddr>()};
+        const u64 nrr_size{rp.Pop<u64>()};
+
+        if (!initialized) {
+            LOG_ERROR(Service_LDR, "LDR:RO not initialized before use!");
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_NOT_INITIALIZED);
+            return;
+        }
+
+        if (nrr.size() >= MAXIMUM_LOADED_RO) {
+            LOG_ERROR(Service_LDR, "Loading new NRR would exceed the maximum number of loaded NRRs "
+                                   "(0x40)! Failing...");
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_MAXIMUM_NRR);
+            return;
+        }
+
+        // NRR Address does not fall on 0x1000 byte boundary
+        if (!Common::Is4KBAligned(nrr_addr)) {
+            LOG_ERROR(Service_LDR, "NRR Address has invalid alignment (actual {:016X})!", nrr_addr);
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_ALIGNMENT);
+            return;
+        }
+
+        // NRR Size is zero or causes overflow
+        if (nrr_addr + nrr_size <= nrr_addr || nrr_size == 0 || !Common::Is4KBAligned(nrr_size)) {
+            LOG_ERROR(Service_LDR, "NRR Size is invalid! (nrr_address={:016X}, nrr_size={:016X})",
+                      nrr_addr, nrr_size);
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_SIZE);
+            return;
+        }
+        // Read NRR data from memory
+        std::vector<u8> nrr_data(nrr_size);
+        Memory::ReadBlock(nrr_addr, nrr_data.data(), nrr_size);
+        NRRHeader header;
+        std::memcpy(&header, nrr_data.data(), sizeof(NRRHeader));
+
+        if (header.magic != Common::MakeMagic('N', 'R', 'R', '0')) {
+            LOG_ERROR(Service_LDR, "NRR did not have magic 'NRR0' (actual {:08X})!", header.magic);
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_NRR);
+            return;
+        }
+
+        if (header.size != nrr_size) {
+            LOG_ERROR(Service_LDR,
+                      "NRR header reported size did not match LoadNrr parameter size! "
+                      "(header_size={:016X}, loadnrr_size={:016X})",
+                      header.size, nrr_size);
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_SIZE);
+            return;
+        }
+
+        if (Core::CurrentProcess()->GetTitleID() != header.title_id) {
+            LOG_ERROR(Service_LDR,
+                      "Attempting to load NRR with title ID other than current process. (actual "
+                      "{:016X})!",
+                      header.title_id);
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_NRR);
+            return;
+        }
+
+        std::vector<SHA256Hash> hashes;
+
+        // Copy all hashes in the NRR (specified by hash count/hash offset) into vector.
+        for (std::size_t i = header.hash_offset;
+             i < (header.hash_offset + (header.hash_count * sizeof(SHA256Hash))); i += 8) {
+            SHA256Hash hash;
+            std::memcpy(hash.data(), nrr_data.data() + i, sizeof(SHA256Hash));
+            hashes.emplace_back(hash);
+        }
+
+        nrr.insert_or_assign(nrr_addr, std::move(hashes));
+
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(RESULT_SUCCESS);
+    }
+
+    void UnloadNrr(Kernel::HLERequestContext& ctx) {
+        if (!initialized) {
+            LOG_ERROR(Service_LDR, "LDR:RO not initialized before use!");
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_NOT_INITIALIZED);
+            return;
+        }
+
+        IPC::RequestParser rp{ctx};
+        rp.Skip(2, false);
+        const auto nrr_addr{rp.Pop<VAddr>()};
+
+        if (!Common::Is4KBAligned(nrr_addr)) {
+            LOG_ERROR(Service_LDR, "NRR Address has invalid alignment (actual {:016X})!", nrr_addr);
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_ALIGNMENT);
+            return;
+        }
+
+        const auto iter = nrr.find(nrr_addr);
+        if (iter == nrr.end()) {
+            LOG_ERROR(Service_LDR,
+                      "Attempting to unload NRR which has not been loaded! (addr={:016X})",
+                      nrr_addr);
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_NRR_ADDRESS);
+            return;
+        }
+
+        nrr.erase(iter);
         IPC::ResponseBuilder rb{ctx, 2};
         rb.Push(RESULT_SUCCESS);
-        LOG_WARNING(Service_LDR, "(STUBBED) called");
     }
 
     void LoadNro(Kernel::HLERequestContext& ctx) {
@@ -88,33 +237,253 @@ public:
         const VAddr bss_addr{rp.Pop<VAddr>()};
         const u64 bss_size{rp.Pop<u64>()};
 
+        if (!initialized) {
+            LOG_ERROR(Service_LDR, "LDR:RO not initialized before use!");
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_NOT_INITIALIZED);
+            return;
+        }
+
+        if (nro.size() >= MAXIMUM_LOADED_RO) {
+            LOG_ERROR(Service_LDR, "Loading new NRO would exceed the maximum number of loaded NROs "
+                                   "(0x40)! Failing...");
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_MAXIMUM_NRO);
+            return;
+        }
+
+        // NRO Address does not fall on 0x1000 byte boundary
+        if (!Common::Is4KBAligned(nro_addr)) {
+            LOG_ERROR(Service_LDR, "NRO Address has invalid alignment (actual {:016X})!", nro_addr);
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_ALIGNMENT);
+            return;
+        }
+
+        // NRO Size or BSS Size is zero or causes overflow
+        const auto nro_size_valid =
+            nro_size != 0 && nro_addr + nro_size > nro_addr && Common::Is4KBAligned(nro_size);
+        const auto bss_size_valid =
+            nro_size + bss_size >= nro_size && (bss_size == 0 || bss_addr + bss_size > bss_addr);
+
+        if (!nro_size_valid || !bss_size_valid) {
+            LOG_ERROR(Service_LDR,
+                      "NRO Size or BSS Size is invalid! (nro_address={:016X}, nro_size={:016X}, "
+                      "bss_address={:016X}, bss_size={:016X})",
+                      nro_addr, nro_size, bss_addr, bss_size);
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_SIZE);
+            return;
+        }
+
         // Read NRO data from memory
         std::vector<u8> nro_data(nro_size);
         Memory::ReadBlock(nro_addr, nro_data.data(), nro_size);
 
-        // Load NRO as new executable module
-        const VAddr addr{*Core::CurrentProcess()->VMManager().FindFreeRegion(nro_size + bss_size)};
-        Loader::AppLoader_NRO::LoadNro(nro_data, fmt::format("nro-{:08x}", addr), addr);
+        SHA256Hash hash{};
+        mbedtls_sha256(nro_data.data(), nro_data.size(), hash.data(), 0);
 
-        // TODO(bunnei): This is an incomplete implementation. It was tested with Super Mario Party.
-        // It is currently missing:
-        // - Signature checks with LoadNRR
-        // - Checking if a module has already been loaded
-        // - Using/validating BSS, etc. params (these are used from NRO header instead)
-        // - Error checking
-        // - ...Probably other things
+        // NRO Hash is already loaded
+        if (std::any_of(nro.begin(), nro.end(), [&hash](const std::pair<VAddr, NROInfo>& info) {
+                return info.second.hash == hash;
+            })) {
+            LOG_ERROR(Service_LDR, "NRO is already loaded!");
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_ALREADY_LOADED);
+            return;
+        }
+
+        // NRO Hash is not in any loaded NRR
+        if (!IsValidNROHash(hash)) {
+            LOG_ERROR(Service_LDR,
+                      "NRO hash is not present in any currently loaded NRRs (hash={})!",
+                      Common::HexArrayToString(hash));
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_MISSING_NRR_HASH);
+            return;
+        }
+
+        NROHeader header;
+        std::memcpy(&header, nro_data.data(), sizeof(NROHeader));
+
+        if (!IsValidNRO(header, nro_size, bss_size)) {
+            LOG_ERROR(Service_LDR, "NRO was invalid!");
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_NRO);
+            return;
+        }
+
+        // Load NRO as new executable module
+        auto* process = Core::CurrentProcess();
+        auto& vm_manager = process->VMManager();
+        auto map_address = vm_manager.FindFreeRegion(nro_size + bss_size);
+
+        if (!map_address.Succeeded() ||
+            *map_address + nro_size + bss_size > vm_manager.GetAddressSpaceEndAddress()) {
+
+            LOG_ERROR(Service_LDR,
+                      "General error while allocation memory or no available memory to allocate!");
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_MEMORY_STATE);
+            return;
+        }
+
+        ASSERT(process->MirrorMemory(*map_address, nro_addr, nro_size,
+                                     Kernel::MemoryState::ModuleCodeStatic) == RESULT_SUCCESS);
+        ASSERT(process->UnmapMemory(nro_addr, 0, nro_size) == RESULT_SUCCESS);
+
+        if (bss_size > 0) {
+            ASSERT(process->MirrorMemory(*map_address + nro_size, bss_addr, bss_size,
+                                         Kernel::MemoryState::ModuleCodeStatic) == RESULT_SUCCESS);
+            ASSERT(process->UnmapMemory(bss_addr, 0, bss_size) == RESULT_SUCCESS);
+        }
+
+        vm_manager.ReprotectRange(*map_address, header.text_size,
+                                  Kernel::VMAPermission::ReadExecute);
+        vm_manager.ReprotectRange(*map_address + header.ro_offset, header.ro_size,
+                                  Kernel::VMAPermission::Read);
+        vm_manager.ReprotectRange(*map_address + header.rw_offset, header.rw_size,
+                                  Kernel::VMAPermission::ReadWrite);
+
+        Core::System::GetInstance().ArmInterface(0).ClearInstructionCache();
+        Core::System::GetInstance().ArmInterface(1).ClearInstructionCache();
+        Core::System::GetInstance().ArmInterface(2).ClearInstructionCache();
+        Core::System::GetInstance().ArmInterface(3).ClearInstructionCache();
+
+        nro.insert_or_assign(*map_address, NROInfo{hash, nro_size + bss_size});
 
         IPC::ResponseBuilder rb{ctx, 4};
         rb.Push(RESULT_SUCCESS);
-        rb.Push(addr);
-        LOG_WARNING(Service_LDR, "(STUBBED) called");
+        rb.Push(*map_address);
+    }
+
+    void UnloadNro(Kernel::HLERequestContext& ctx) {
+        IPC::RequestParser rp{ctx};
+        rp.Skip(2, false);
+        const VAddr mapped_addr{rp.PopRaw<VAddr>()};
+        const VAddr heap_addr{rp.PopRaw<VAddr>()};
+
+        if (!initialized) {
+            LOG_ERROR(Service_LDR, "LDR:RO not initialized before use!");
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_NOT_INITIALIZED);
+            return;
+        }
+
+        if (!Common::Is4KBAligned(mapped_addr) || !Common::Is4KBAligned(heap_addr)) {
+            LOG_ERROR(Service_LDR,
+                      "NRO/BSS Address has invalid alignment (actual nro_addr={:016X}, "
+                      "bss_addr={:016X})!",
+                      mapped_addr, heap_addr);
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_ALIGNMENT);
+            return;
+        }
+
+        const auto iter = nro.find(mapped_addr);
+        if (iter == nro.end()) {
+            LOG_ERROR(Service_LDR,
+                      "The NRO attempting to unmap was not mapped or has an invalid address "
+                      "(actual {:016X})!",
+                      mapped_addr);
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_NRO_ADDRESS);
+            return;
+        }
+
+        auto* process = Core::CurrentProcess();
+        auto& vm_manager = process->VMManager();
+        const auto& nro_size = iter->second.size;
+
+        ASSERT(process->MirrorMemory(heap_addr, mapped_addr, nro_size,
+                                     Kernel::MemoryState::ModuleCodeStatic) == RESULT_SUCCESS);
+        ASSERT(process->UnmapMemory(mapped_addr, 0, nro_size) == RESULT_SUCCESS);
+
+        Core::System::GetInstance().ArmInterface(0).ClearInstructionCache();
+        Core::System::GetInstance().ArmInterface(1).ClearInstructionCache();
+        Core::System::GetInstance().ArmInterface(2).ClearInstructionCache();
+        Core::System::GetInstance().ArmInterface(3).ClearInstructionCache();
+
+        nro.erase(iter);
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(RESULT_SUCCESS);
     }
 
     void Initialize(Kernel::HLERequestContext& ctx) {
+        initialized = true;
+
         IPC::ResponseBuilder rb{ctx, 2};
         rb.Push(RESULT_SUCCESS);
         LOG_WARNING(Service_LDR, "(STUBBED) called");
     }
+
+private:
+    using SHA256Hash = std::array<u8, 0x20>;
+
+    struct NROHeader {
+        u32_le entrypoint_insn;
+        u32_le mod_offset;
+        INSERT_PADDING_WORDS(2);
+        u32_le magic;
+        INSERT_PADDING_WORDS(1);
+        u32_le nro_size;
+        INSERT_PADDING_WORDS(1);
+        u32_le text_offset;
+        u32_le text_size;
+        u32_le ro_offset;
+        u32_le ro_size;
+        u32_le rw_offset;
+        u32_le rw_size;
+        u32_le bss_size;
+        INSERT_PADDING_WORDS(1);
+        std::array<u8, 0x20> build_id;
+        INSERT_PADDING_BYTES(0x20);
+    };
+    static_assert(sizeof(NROHeader) == 0x80, "NROHeader has invalid size.");
+
+    struct NRRHeader {
+        u32_le magic;
+        INSERT_PADDING_BYTES(0x1C);
+        u64_le title_id_mask;
+        u64_le title_id_pattern;
+        std::array<u8, 0x100> modulus;
+        std::array<u8, 0x100> signature_1;
+        std::array<u8, 0x100> signature_2;
+        u64_le title_id;
+        u32_le size;
+        INSERT_PADDING_BYTES(4);
+        u32_le hash_offset;
+        u32_le hash_count;
+        INSERT_PADDING_BYTES(8);
+    };
+    static_assert(sizeof(NRRHeader) == 0x350, "NRRHeader has incorrect size.");
+
+    struct NROInfo {
+        SHA256Hash hash;
+        u64 size;
+    };
+
+    bool initialized = false;
+
+    std::map<VAddr, NROInfo> nro;
+    std::map<VAddr, std::vector<SHA256Hash>> nrr;
+
+    bool IsValidNROHash(const SHA256Hash& hash) {
+        return std::any_of(
+            nrr.begin(), nrr.end(), [&hash](const std::pair<VAddr, std::vector<SHA256Hash>>& p) {
+                return std::find(p.second.begin(), p.second.end(), hash) != p.second.end();
+            });
+    }
+
+    static bool IsValidNRO(const NROHeader& header, u64 nro_size, u64 bss_size) {
+        return header.magic == Common::MakeMagic('N', 'R', 'O', '0') &&
+               header.nro_size == nro_size && header.bss_size == bss_size &&
+               header.ro_offset == header.text_offset + header.text_size &&
+               header.rw_offset == header.ro_offset + header.ro_size &&
+               nro_size == header.rw_offset + header.rw_size &&
+               Common::Is4KBAligned(header.text_size) && Common::Is4KBAligned(header.ro_size) &&
+               Common::Is4KBAligned(header.rw_size);
+    }
 };
 
 void InstallInterfaces(SM::ServiceManager& sm) {

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

@@ -212,7 +212,7 @@ private:
         IPC::ResponseBuilder rb{ctx, 2};
         auto amiibo = nfp_interface.GetAmiiboBuffer();
         TagInfo tag_info{};
-        std::memcpy(tag_info.uuid.data(), amiibo.uuid.data(), sizeof(tag_info.uuid.size()));
+        tag_info.uuid = amiibo.uuid;
         tag_info.uuid_length = static_cast<u8>(tag_info.uuid.size());
 
         tag_info.protocol = 1; // TODO(ogniK): Figure out actual values

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

@@ -2,6 +2,9 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
+#include "common/logging/log.h"
+#include "core/file_sys/control_metadata.h"
+#include "core/file_sys/patch_manager.h"
 #include "core/hle/ipc_helpers.h"
 #include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/service/ns/ns.h"
@@ -118,7 +121,7 @@ public:
             {305, nullptr, "TerminateSystemApplet"},
             {306, nullptr, "LaunchOverlayApplet"},
             {307, nullptr, "TerminateOverlayApplet"},
-            {400, nullptr, "GetApplicationControlData"},
+            {400, &IApplicationManagerInterface::GetApplicationControlData, "GetApplicationControlData"},
             {401, nullptr, "InvalidateAllApplicationControlCache"},
             {402, nullptr, "RequestDownloadApplicationControlData"},
             {403, nullptr, "GetMaxApplicationControlCacheCount"},
@@ -243,6 +246,65 @@ public:
 
         RegisterHandlers(functions);
     }
+
+    void GetApplicationControlData(Kernel::HLERequestContext& ctx) {
+        IPC::RequestParser rp{ctx};
+        const auto flag = rp.PopRaw<u64>();
+        LOG_DEBUG(Service_NS, "called with flag={:016X}", flag);
+
+        const auto title_id = rp.PopRaw<u64>();
+
+        const auto size = ctx.GetWriteBufferSize();
+
+        const FileSys::PatchManager pm{title_id};
+        const auto control = pm.GetControlMetadata();
+
+        std::vector<u8> out;
+
+        if (control.first != nullptr) {
+            if (size < 0x4000) {
+                LOG_ERROR(Service_NS,
+                          "output buffer is too small! (actual={:016X}, expected_min=0x4000)",
+                          size);
+                IPC::ResponseBuilder rb{ctx, 2};
+                // TODO(DarkLordZach): Find a better error code for this.
+                rb.Push(ResultCode(-1));
+                return;
+            }
+
+            out.resize(0x4000);
+            const auto bytes = control.first->GetRawBytes();
+            std::memcpy(out.data(), bytes.data(), bytes.size());
+        } else {
+            LOG_WARNING(Service_NS, "missing NACP data for title_id={:016X}, defaulting to zeros.",
+                        title_id);
+            out.resize(std::min<u64>(0x4000, size));
+        }
+
+        if (control.second != nullptr) {
+            if (size < 0x4000 + control.second->GetSize()) {
+                LOG_ERROR(Service_NS,
+                          "output buffer is too small! (actual={:016X}, expected_min={:016X})",
+                          size, 0x4000 + control.second->GetSize());
+                IPC::ResponseBuilder rb{ctx, 2};
+                // TODO(DarkLordZach): Find a better error code for this.
+                rb.Push(ResultCode(-1));
+                return;
+            }
+
+            out.resize(0x4000 + control.second->GetSize());
+            control.second->Read(out.data() + 0x4000, control.second->GetSize());
+        } else {
+            LOG_WARNING(Service_NS, "missing icon data for title_id={:016X}, defaulting to zeros.",
+                        title_id);
+        }
+
+        ctx.WriteBuffer(out);
+
+        IPC::ResponseBuilder rb{ctx, 3};
+        rb.Push(RESULT_SUCCESS);
+        rb.Push<u32>(static_cast<u32>(out.size()));
+    }
 };
 
 class IApplicationVersionInterface final : public ServiceFramework<IApplicationVersionInterface> {

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

@@ -351,6 +351,14 @@ void PL_U::GetSharedFontInOrderOfPriority(Kernel::HLERequestContext& ctx) {
         font_sizes.push_back(region.size);
     }
 
+    // Resize buffers if game requests smaller size output.
+    font_codes.resize(
+        std::min<std::size_t>(font_codes.size(), ctx.GetWriteBufferSize(0) / sizeof(u32)));
+    font_offsets.resize(
+        std::min<std::size_t>(font_offsets.size(), ctx.GetWriteBufferSize(1) / sizeof(u32)));
+    font_sizes.resize(
+        std::min<std::size_t>(font_sizes.size(), ctx.GetWriteBufferSize(2) / sizeof(u32)));
+
     ctx.WriteBuffer(font_codes, 0);
     ctx.WriteBuffer(font_offsets, 1);
     ctx.WriteBuffer(font_sizes, 2);

src/core/hle/service/service.cpp

@@ -80,8 +80,8 @@ namespace Service {
  * Creates a function string for logging, complete with the name (or header code, depending
  * on what's passed in) the port name, and all the cmd_buff arguments.
  */
-static std::string MakeFunctionString(const char* name, const char* port_name,
-                                      const u32* cmd_buff) {
+[[maybe_unused]] static std::string MakeFunctionString(const char* name, const char* port_name,
+                                                       const u32* cmd_buff) {
     // Number of params == bits 0-5 + bits 6-11
     int num_params = (cmd_buff[0] & 0x3F) + ((cmd_buff[0] >> 6) & 0x3F);
 

src/core/hle/service/spl/module.cpp

@@ -3,18 +3,23 @@
 // Refer to the license.txt file included.
 
 #include <algorithm>
+#include <chrono>
 #include <cstdlib>
+#include <ctime>
+#include <functional>
 #include <vector>
 #include "common/logging/log.h"
 #include "core/hle/ipc_helpers.h"
 #include "core/hle/service/spl/csrng.h"
 #include "core/hle/service/spl/module.h"
 #include "core/hle/service/spl/spl.h"
+#include "core/settings.h"
 
 namespace Service::SPL {
 
 Module::Interface::Interface(std::shared_ptr<Module> module, const char* name)
-    : ServiceFramework(name), module(std::move(module)) {}
+    : ServiceFramework(name), module(std::move(module)),
+      rng(Settings::values.rng_seed.value_or(std::time(nullptr))) {}
 
 Module::Interface::~Interface() = default;
 
@@ -23,8 +28,9 @@ void Module::Interface::GetRandomBytes(Kernel::HLERequestContext& ctx) {
 
     std::size_t size = ctx.GetWriteBufferSize();
 
+    std::uniform_int_distribution<u16> distribution(0, std::numeric_limits<u8>::max());
     std::vector<u8> data(size);
-    std::generate(data.begin(), data.end(), std::rand);
+    std::generate(data.begin(), data.end(), [&] { return static_cast<u8>(distribution(rng)); });
 
     ctx.WriteBuffer(data);
 

src/core/hle/service/spl/module.h

@@ -4,6 +4,7 @@
 
 #pragma once
 
+#include <random>
 #include "core/hle/service/service.h"
 
 namespace Service::SPL {
@@ -19,6 +20,9 @@ public:
 
     protected:
         std::shared_ptr<Module> module;
+
+    private:
+        std::mt19937 rng;
     };
 };
 

src/core/hle/service/time/interface.cpp

@@ -21,9 +21,10 @@ Time::Time(std::shared_ptr<Module> time, const char* name)
         {102, nullptr, "GetStandardUserSystemClockInitialYear"},
         {200, nullptr, "IsStandardNetworkSystemClockAccuracySufficient"},
         {300, nullptr, "CalculateMonotonicSystemClockBaseTimePoint"},
-        {400, nullptr, "GetClockSnapshot"},
+        {400, &Time::GetClockSnapshot, "GetClockSnapshot"},
         {401, nullptr, "GetClockSnapshotFromSystemClockContext"},
-        {500, nullptr, "CalculateStandardUserSystemClockDifferenceByUser"},
+        {500, &Time::CalculateStandardUserSystemClockDifferenceByUser,
+         "CalculateStandardUserSystemClockDifferenceByUser"},
         {501, nullptr, "CalculateSpanBetween"},
     };
     RegisterHandlers(functions);

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

@@ -15,6 +15,44 @@
 
 namespace Service::Time {
 
+static void PosixToCalendar(u64 posix_time, CalendarTime& calendar_time,
+                            CalendarAdditionalInfo& additional_info,
+                            [[maybe_unused]] const TimeZoneRule& /*rule*/) {
+    const std::time_t time(posix_time);
+    const std::tm* tm = std::localtime(&time);
+    if (tm == nullptr) {
+        calendar_time = {};
+        additional_info = {};
+        return;
+    }
+    calendar_time.year = tm->tm_year + 1900;
+    calendar_time.month = tm->tm_mon + 1;
+    calendar_time.day = tm->tm_mday;
+    calendar_time.hour = tm->tm_hour;
+    calendar_time.minute = tm->tm_min;
+    calendar_time.second = tm->tm_sec;
+
+    additional_info.day_of_week = tm->tm_wday;
+    additional_info.day_of_year = tm->tm_yday;
+    std::memcpy(additional_info.name.data(), "UTC", sizeof("UTC"));
+    additional_info.utc_offset = 0;
+}
+
+static u64 CalendarToPosix(const CalendarTime& calendar_time,
+                           [[maybe_unused]] const TimeZoneRule& /*rule*/) {
+    std::tm time{};
+    time.tm_year = calendar_time.year - 1900;
+    time.tm_mon = calendar_time.month - 1;
+    time.tm_mday = calendar_time.day;
+
+    time.tm_hour = calendar_time.hour;
+    time.tm_min = calendar_time.minute;
+    time.tm_sec = calendar_time.second;
+
+    std::time_t epoch_time = std::mktime(&time);
+    return static_cast<u64>(epoch_time);
+}
+
 class ISystemClock final : public ServiceFramework<ISystemClock> {
 public:
     ISystemClock() : ServiceFramework("ISystemClock") {
@@ -80,8 +118,8 @@ public:
             {5, nullptr, "GetTimeZoneRuleVersion"},
             {100, &ITimeZoneService::ToCalendarTime, "ToCalendarTime"},
             {101, &ITimeZoneService::ToCalendarTimeWithMyRule, "ToCalendarTimeWithMyRule"},
-            {201, nullptr, "ToPosixTime"},
-            {202, nullptr, "ToPosixTimeWithMyRule"},
+            {201, &ITimeZoneService::ToPosixTime, "ToPosixTime"},
+            {202, &ITimeZoneService::ToPosixTimeWithMyRule, "ToPosixTimeWithMyRule"},
         };
         RegisterHandlers(functions);
     }
@@ -151,24 +189,29 @@ private:
         rb.PushRaw(additional_info);
     }
 
-    void PosixToCalendar(u64 posix_time, CalendarTime& calendar_time,
-                         CalendarAdditionalInfo& additional_info, const TimeZoneRule& /*rule*/) {
-        std::time_t t(posix_time);
-        std::tm* tm = std::localtime(&t);
-        if (!tm) {
-            return;
-        }
-        calendar_time.year = tm->tm_year + 1900;
-        calendar_time.month = tm->tm_mon + 1;
-        calendar_time.day = tm->tm_mday;
-        calendar_time.hour = tm->tm_hour;
-        calendar_time.minute = tm->tm_min;
-        calendar_time.second = tm->tm_sec;
+    void ToPosixTime(Kernel::HLERequestContext& ctx) {
+        // TODO(ogniK): Figure out how to handle multiple times
+        LOG_WARNING(Service_Time, "(STUBBED) called");
+        IPC::RequestParser rp{ctx};
+        auto calendar_time = rp.PopRaw<CalendarTime>();
+        auto posix_time = CalendarToPosix(calendar_time, {});
 
-        additional_info.day_of_week = tm->tm_wday;
-        additional_info.day_of_year = tm->tm_yday;
-        std::memcpy(additional_info.name.data(), "UTC", sizeof("UTC"));
-        additional_info.utc_offset = 0;
+        IPC::ResponseBuilder rb{ctx, 3};
+        rb.Push(RESULT_SUCCESS);
+        rb.PushRaw<u32>(1); // Amount of times we're returning
+        ctx.WriteBuffer(&posix_time, sizeof(u64));
+    }
+
+    void ToPosixTimeWithMyRule(Kernel::HLERequestContext& ctx) {
+        LOG_WARNING(Service_Time, "(STUBBED) called");
+        IPC::RequestParser rp{ctx};
+        auto calendar_time = rp.PopRaw<CalendarTime>();
+        auto posix_time = CalendarToPosix(calendar_time, {});
+
+        IPC::ResponseBuilder rb{ctx, 3};
+        rb.Push(RESULT_SUCCESS);
+        rb.PushRaw<u32>(1); // Amount of times we're returning
+        ctx.WriteBuffer(&posix_time, sizeof(u64));
     }
 };
 
@@ -207,6 +250,70 @@ void Module::Interface::GetStandardLocalSystemClock(Kernel::HLERequestContext& c
     LOG_DEBUG(Service_Time, "called");
 }
 
+void Module::Interface::GetClockSnapshot(Kernel::HLERequestContext& ctx) {
+    LOG_DEBUG(Service_Time, "called");
+
+    IPC::RequestParser rp{ctx};
+    auto unknown_u8 = rp.PopRaw<u8>();
+
+    ClockSnapshot clock_snapshot{};
+
+    const s64 time_since_epoch{std::chrono::duration_cast<std::chrono::seconds>(
+                                   std::chrono::system_clock::now().time_since_epoch())
+                                   .count()};
+    CalendarTime calendar_time{};
+    const std::time_t time(time_since_epoch);
+    const std::tm* tm = std::localtime(&time);
+    if (tm == nullptr) {
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(ResultCode(-1)); // TODO(ogniK): Find appropriate error code
+        return;
+    }
+    SteadyClockTimePoint steady_clock_time_point{CoreTiming::cyclesToMs(CoreTiming::GetTicks()) /
+                                                 1000};
+
+    LocationName location_name{"UTC"};
+    calendar_time.year = tm->tm_year + 1900;
+    calendar_time.month = tm->tm_mon + 1;
+    calendar_time.day = tm->tm_mday;
+    calendar_time.hour = tm->tm_hour;
+    calendar_time.minute = tm->tm_min;
+    calendar_time.second = tm->tm_sec;
+    clock_snapshot.system_posix_time = time_since_epoch;
+    clock_snapshot.network_posix_time = time_since_epoch;
+    clock_snapshot.system_calendar_time = calendar_time;
+    clock_snapshot.network_calendar_time = calendar_time;
+
+    CalendarAdditionalInfo additional_info{};
+    PosixToCalendar(time_since_epoch, calendar_time, additional_info, {});
+
+    clock_snapshot.system_calendar_info = additional_info;
+    clock_snapshot.network_calendar_info = additional_info;
+
+    clock_snapshot.steady_clock_timepoint = steady_clock_time_point;
+    clock_snapshot.location_name = location_name;
+    clock_snapshot.clock_auto_adjustment_enabled = 1;
+    clock_snapshot.ipc_u8 = unknown_u8;
+    IPC::ResponseBuilder rb{ctx, 2};
+    rb.Push(RESULT_SUCCESS);
+    ctx.WriteBuffer(&clock_snapshot, sizeof(ClockSnapshot));
+}
+
+void Module::Interface::CalculateStandardUserSystemClockDifferenceByUser(
+    Kernel::HLERequestContext& ctx) {
+    LOG_DEBUG(Service_Time, "called");
+
+    IPC::RequestParser rp{ctx};
+    const auto snapshot_a = rp.PopRaw<ClockSnapshot>();
+    const auto snapshot_b = rp.PopRaw<ClockSnapshot>();
+    const u64 difference =
+        snapshot_b.user_clock_context.offset - snapshot_a.user_clock_context.offset;
+
+    IPC::ResponseBuilder rb{ctx, 4};
+    rb.Push(RESULT_SUCCESS);
+    rb.PushRaw<u64>(difference);
+}
+
 Module::Interface::Interface(std::shared_ptr<Module> time, const char* name)
     : ServiceFramework(name), time(std::move(time)) {}
 

src/core/hle/service/time/time.h

@@ -5,6 +5,7 @@
 #pragma once
 
 #include <array>
+#include "common/common_funcs.h"
 #include "core/hle/service/service.h"
 
 namespace Service::Time {
@@ -53,6 +54,23 @@ struct SystemClockContext {
 static_assert(sizeof(SystemClockContext) == 0x20,
               "SystemClockContext structure has incorrect size");
 
+struct ClockSnapshot {
+    SystemClockContext user_clock_context;
+    SystemClockContext network_clock_context;
+    s64_le system_posix_time;
+    s64_le network_posix_time;
+    CalendarTime system_calendar_time;
+    CalendarTime network_calendar_time;
+    CalendarAdditionalInfo system_calendar_info;
+    CalendarAdditionalInfo network_calendar_info;
+    SteadyClockTimePoint steady_clock_timepoint;
+    LocationName location_name;
+    u8 clock_auto_adjustment_enabled;
+    u8 ipc_u8;
+    INSERT_PADDING_BYTES(2);
+};
+static_assert(sizeof(ClockSnapshot) == 0xd0, "ClockSnapshot is an invalid size");
+
 class Module final {
 public:
     class Interface : public ServiceFramework<Interface> {
@@ -65,6 +83,8 @@ public:
         void GetStandardSteadyClock(Kernel::HLERequestContext& ctx);
         void GetTimeZoneService(Kernel::HLERequestContext& ctx);
         void GetStandardLocalSystemClock(Kernel::HLERequestContext& ctx);
+        void GetClockSnapshot(Kernel::HLERequestContext& ctx);
+        void CalculateStandardUserSystemClockDifferenceByUser(Kernel::HLERequestContext& ctx);
 
     protected:
         std::shared_ptr<Module> time;

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

@@ -237,6 +237,22 @@ private:
     Data data{};
 };
 
+/// Represents a parcel containing one int '0' as its data
+/// Used by DetachBuffer and Disconnect
+class IGBPEmptyResponseParcel : public Parcel {
+protected:
+    void SerializeData() override {
+        Write(data);
+    }
+
+private:
+    struct Data {
+        u32_le unk_0;
+    };
+
+    Data data{};
+};
+
 class IGBPSetPreallocatedBufferRequestParcel : public Parcel {
 public:
     explicit IGBPSetPreallocatedBufferRequestParcel(std::vector<u8> buffer)
@@ -554,6 +570,12 @@ private:
             ctx.WriteBuffer(response.Serialize());
         } else if (transaction == TransactionId::CancelBuffer) {
             LOG_CRITICAL(Service_VI, "(STUBBED) called, transaction=CancelBuffer");
+        } else if (transaction == TransactionId::Disconnect ||
+                   transaction == TransactionId::DetachBuffer) {
+            const auto buffer = ctx.ReadBuffer();
+
+            IGBPEmptyResponseParcel response{};
+            ctx.WriteBuffer(response.Serialize());
         } else {
             ASSERT_MSG(false, "Unimplemented");
         }

src/core/loader/nro.cpp

@@ -12,10 +12,12 @@
 #include "common/swap.h"
 #include "core/core.h"
 #include "core/file_sys/control_metadata.h"
+#include "core/file_sys/romfs_factory.h"
 #include "core/file_sys/vfs_offset.h"
 #include "core/gdbstub/gdbstub.h"
 #include "core/hle/kernel/process.h"
 #include "core/hle/kernel/vm_manager.h"
+#include "core/hle/service/filesystem/filesystem.h"
 #include "core/loader/nro.h"
 #include "core/loader/nso.h"
 #include "core/memory.h"
@@ -168,17 +170,20 @@ static constexpr u32 PageAlignSize(u32 size) {
                     arg_data.size());
     }
 
-    // Read MOD header
-    ModHeader mod_header{};
     // Default .bss to NRO header bss size if MOD0 section doesn't exist
     u32 bss_size{PageAlignSize(nro_header.bss_size)};
+
+    // Read MOD header
+    ModHeader mod_header{};
     std::memcpy(&mod_header, program_image.data() + nro_header.module_header_offset,
                 sizeof(ModHeader));
+
     const bool has_mod_header{mod_header.magic == Common::MakeMagic('M', 'O', 'D', '0')};
     if (has_mod_header) {
         // Resize program image to include .bss section and page align each section
         bss_size = PageAlignSize(mod_header.bss_end_offset - mod_header.bss_start_offset);
     }
+
     codeset.DataSegment().size += bss_size;
     program_image.resize(static_cast<u32>(program_image.size()) + bss_size);
 
@@ -208,6 +213,9 @@ ResultStatus AppLoader_NRO::Load(Kernel::Process& process) {
         return ResultStatus::ErrorLoadingNRO;
     }
 
+    if (romfs != nullptr)
+        Service::FileSystem::RegisterRomFS(std::make_unique<FileSys::RomFSFactory>(*this));
+
     process.Run(base_address, Kernel::THREADPRIO_DEFAULT, Memory::DEFAULT_STACK_SIZE);
 
     is_loaded = true;

src/core/loader/nso.cpp

@@ -154,7 +154,7 @@ std::optional<VAddr> AppLoader_NSO::LoadModule(const FileSys::VfsFile& file, VAd
     program_image.resize(image_size);
 
     // Apply patches if necessary
-    if (pm && pm->HasNSOPatch(nso_header.build_id)) {
+    if (pm && (pm->HasNSOPatch(nso_header.build_id) || Settings::values.dump_nso)) {
         std::vector<u8> pi_header(program_image.size() + 0x100);
         std::memcpy(pi_header.data(), &nso_header, sizeof(NsoHeader));
         std::memcpy(pi_header.data() + 0x100, program_image.data(), program_image.size());

src/core/settings.cpp

@@ -10,6 +10,56 @@
 
 namespace Settings {
 
+namespace NativeButton {
+const std::array<const char*, NumButtons> mapping = {{
+    "button_a",
+    "button_b",
+    "button_x",
+    "button_y",
+    "button_lstick",
+    "button_rstick",
+    "button_l",
+    "button_r",
+    "button_zl",
+    "button_zr",
+    "button_plus",
+    "button_minus",
+    "button_dleft",
+    "button_dup",
+    "button_dright",
+    "button_ddown",
+    "button_lstick_left",
+    "button_lstick_up",
+    "button_lstick_right",
+    "button_lstick_down",
+    "button_rstick_left",
+    "button_rstick_up",
+    "button_rstick_right",
+    "button_rstick_down",
+    "button_sl",
+    "button_sr",
+    "button_home",
+    "button_screenshot",
+}};
+}
+
+namespace NativeAnalog {
+const std::array<const char*, NumAnalogs> mapping = {{
+    "lstick",
+    "rstick",
+}};
+}
+
+namespace NativeMouseButton {
+const std::array<const char*, NumMouseButtons> mapping = {{
+    "left",
+    "right",
+    "middle",
+    "forward",
+    "back",
+}};
+}
+
 Values values = {};
 
 void Apply() {

src/core/settings.h

@@ -6,6 +6,7 @@
 
 #include <array>
 #include <atomic>
+#include <optional>
 #include <string>
 #include "common/common_types.h"
 
@@ -59,36 +60,7 @@ constexpr int BUTTON_NS_END = NumButtons;
 constexpr int NUM_BUTTONS_HID = BUTTON_HID_END - BUTTON_HID_BEGIN;
 constexpr int NUM_BUTTONS_NS = BUTTON_NS_END - BUTTON_NS_BEGIN;
 
-static const std::array<const char*, NumButtons> mapping = {{
-    "button_a",
-    "button_b",
-    "button_x",
-    "button_y",
-    "button_lstick",
-    "button_rstick",
-    "button_l",
-    "button_r",
-    "button_zl",
-    "button_zr",
-    "button_plus",
-    "button_minus",
-    "button_dleft",
-    "button_dup",
-    "button_dright",
-    "button_ddown",
-    "button_lstick_left",
-    "button_lstick_up",
-    "button_lstick_right",
-    "button_lstick_down",
-    "button_rstick_left",
-    "button_rstick_up",
-    "button_rstick_right",
-    "button_rstick_down",
-    "button_sl",
-    "button_sr",
-    "button_home",
-    "button_screenshot",
-}};
+extern const std::array<const char*, NumButtons> mapping;
 
 } // namespace NativeButton
 
@@ -104,24 +76,298 @@ constexpr int STICK_HID_BEGIN = LStick;
 constexpr int STICK_HID_END = NumAnalogs;
 constexpr int NUM_STICKS_HID = NumAnalogs;
 
-static const std::array<const char*, NumAnalogs> mapping = {{
-    "lstick",
-    "rstick",
-}};
+extern const std::array<const char*, NumAnalogs> mapping;
 } // namespace NativeAnalog
 
+namespace NativeMouseButton {
+enum Values {
+    Left,
+    Right,
+    Middle,
+    Forward,
+    Back,
+
+    NumMouseButtons,
+};
+
+constexpr int MOUSE_HID_BEGIN = Left;
+constexpr int MOUSE_HID_END = NumMouseButtons;
+constexpr int NUM_MOUSE_HID = NumMouseButtons;
+
+extern const std::array<const char*, NumMouseButtons> mapping;
+} // namespace NativeMouseButton
+
+namespace NativeKeyboard {
+enum Keys {
+    None,
+    Error,
+
+    A = 4,
+    B,
+    C,
+    D,
+    E,
+    F,
+    G,
+    H,
+    I,
+    J,
+    K,
+    L,
+    M,
+    N,
+    O,
+    P,
+    Q,
+    R,
+    S,
+    T,
+    U,
+    V,
+    W,
+    X,
+    Y,
+    Z,
+    N1,
+    N2,
+    N3,
+    N4,
+    N5,
+    N6,
+    N7,
+    N8,
+    N9,
+    N0,
+    Enter,
+    Escape,
+    Backspace,
+    Tab,
+    Space,
+    Minus,
+    Equal,
+    LeftBrace,
+    RightBrace,
+    Backslash,
+    Tilde,
+    Semicolon,
+    Apostrophe,
+    Grave,
+    Comma,
+    Dot,
+    Slash,
+    CapsLockKey,
+
+    F1,
+    F2,
+    F3,
+    F4,
+    F5,
+    F6,
+    F7,
+    F8,
+    F9,
+    F10,
+    F11,
+    F12,
+
+    SystemRequest,
+    ScrollLockKey,
+    Pause,
+    Insert,
+    Home,
+    PageUp,
+    Delete,
+    End,
+    PageDown,
+    Right,
+    Left,
+    Down,
+    Up,
+
+    NumLockKey,
+    KPSlash,
+    KPAsterisk,
+    KPMinus,
+    KPPlus,
+    KPEnter,
+    KP1,
+    KP2,
+    KP3,
+    KP4,
+    KP5,
+    KP6,
+    KP7,
+    KP8,
+    KP9,
+    KP0,
+    KPDot,
+
+    Key102,
+    Compose,
+    Power,
+    KPEqual,
+
+    F13,
+    F14,
+    F15,
+    F16,
+    F17,
+    F18,
+    F19,
+    F20,
+    F21,
+    F22,
+    F23,
+    F24,
+
+    Open,
+    Help,
+    Properties,
+    Front,
+    Stop,
+    Repeat,
+    Undo,
+    Cut,
+    Copy,
+    Paste,
+    Find,
+    Mute,
+    VolumeUp,
+    VolumeDown,
+    CapsLockActive,
+    NumLockActive,
+    ScrollLockActive,
+    KPComma,
+
+    KPLeftParenthesis,
+    KPRightParenthesis,
+
+    LeftControlKey = 0xE0,
+    LeftShiftKey,
+    LeftAltKey,
+    LeftMetaKey,
+    RightControlKey,
+    RightShiftKey,
+    RightAltKey,
+    RightMetaKey,
+
+    MediaPlayPause,
+    MediaStopCD,
+    MediaPrevious,
+    MediaNext,
+    MediaEject,
+    MediaVolumeUp,
+    MediaVolumeDown,
+    MediaMute,
+    MediaWebsite,
+    MediaBack,
+    MediaForward,
+    MediaStop,
+    MediaFind,
+    MediaScrollUp,
+    MediaScrollDown,
+    MediaEdit,
+    MediaSleep,
+    MediaCoffee,
+    MediaRefresh,
+    MediaCalculator,
+
+    NumKeyboardKeys,
+};
+
+static_assert(NumKeyboardKeys == 0xFC, "Incorrect number of keyboard keys.");
+
+enum Modifiers {
+    LeftControl,
+    LeftShift,
+    LeftAlt,
+    LeftMeta,
+    RightControl,
+    RightShift,
+    RightAlt,
+    RightMeta,
+    CapsLock,
+    ScrollLock,
+    NumLock,
+
+    NumKeyboardMods,
+};
+
+constexpr int KEYBOARD_KEYS_HID_BEGIN = None;
+constexpr int KEYBOARD_KEYS_HID_END = NumKeyboardKeys;
+constexpr int NUM_KEYBOARD_KEYS_HID = NumKeyboardKeys;
+
+constexpr int KEYBOARD_MODS_HID_BEGIN = LeftControl;
+constexpr int KEYBOARD_MODS_HID_END = NumKeyboardMods;
+constexpr int NUM_KEYBOARD_MODS_HID = NumKeyboardMods;
+
+} // namespace NativeKeyboard
+
+using ButtonsRaw = std::array<std::string, NativeButton::NumButtons>;
+using AnalogsRaw = std::array<std::string, NativeAnalog::NumAnalogs>;
+using MouseButtonsRaw = std::array<std::string, NativeMouseButton::NumMouseButtons>;
+using KeyboardKeysRaw = std::array<std::string, NativeKeyboard::NumKeyboardKeys>;
+using KeyboardModsRaw = std::array<std::string, NativeKeyboard::NumKeyboardMods>;
+
+constexpr u32 JOYCON_BODY_NEON_RED = 0xFF3C28;
+constexpr u32 JOYCON_BUTTONS_NEON_RED = 0x1E0A0A;
+constexpr u32 JOYCON_BODY_NEON_BLUE = 0x0AB9E6;
+constexpr u32 JOYCON_BUTTONS_NEON_BLUE = 0x001E1E;
+
+enum class ControllerType {
+    ProController,
+    DualJoycon,
+    RightJoycon,
+    LeftJoycon,
+};
+
+struct PlayerInput {
+    bool connected;
+    ControllerType type;
+    ButtonsRaw buttons;
+    AnalogsRaw analogs;
+
+    u32 body_color_right;
+    u32 button_color_right;
+    u32 body_color_left;
+    u32 button_color_left;
+};
+
+struct TouchscreenInput {
+    bool enabled;
+    std::string device;
+
+    u32 finger;
+    u32 diameter_x;
+    u32 diameter_y;
+    u32 rotation_angle;
+};
+
 struct Values {
     // System
     bool use_docked_mode;
     bool enable_nfc;
-    int current_user;
-    int language_index;
+    std::optional<u32> rng_seed;
+    s32 current_user;
+    s32 language_index;
 
     // Controls
-    std::array<std::string, NativeButton::NumButtons> buttons;
-    std::array<std::string, NativeAnalog::NumAnalogs> analogs;
+    std::array<PlayerInput, 10> players;
+
+    bool mouse_enabled;
+    std::string mouse_device;
+    MouseButtonsRaw mouse_buttons;
+
+    bool keyboard_enabled;
+    KeyboardKeysRaw keyboard_keys;
+    KeyboardModsRaw keyboard_mods;
+
+    bool debug_pad_enabled;
+    ButtonsRaw debug_pad_buttons;
+    AnalogsRaw debug_pad_analogs;
+
     std::string motion_device;
-    std::string touch_device;
+    TouchscreenInput touchscreen;
     std::atomic_bool is_device_reload_pending{true};
 
     // Core
@@ -157,6 +403,7 @@ struct Values {
     bool use_gdbstub;
     u16 gdbstub_port;
     std::string program_args;
+    bool dump_nso;
 
     // WebService
     bool enable_telemetry;

src/video_core/CMakeLists.txt

@@ -21,6 +21,7 @@ add_library(video_core STATIC
     macro_interpreter.h
     memory_manager.cpp
     memory_manager.h
+    rasterizer_cache.cpp
     rasterizer_cache.h
     rasterizer_interface.h
     renderer_base.cpp

src/video_core/engines/maxwell_3d.cpp

@@ -34,8 +34,8 @@ void Maxwell3D::InitializeRegisterDefaults() {
     // Depth range near/far is not always set, but is expected to be the default 0.0f, 1.0f. This is
     // needed for ARMS.
     for (std::size_t viewport{}; viewport < Regs::NumViewports; ++viewport) {
-        regs.viewport[viewport].depth_range_near = 0.0f;
-        regs.viewport[viewport].depth_range_far = 1.0f;
+        regs.viewports[viewport].depth_range_near = 0.0f;
+        regs.viewports[viewport].depth_range_far = 1.0f;
     }
     // Doom and Bomberman seems to use the uninitialized registers and just enable blend
     // so initialize blend registers with sane values
@@ -53,6 +53,22 @@ void Maxwell3D::InitializeRegisterDefaults() {
         regs.independent_blend[blend_index].factor_source_a = Regs::Blend::Factor::One;
         regs.independent_blend[blend_index].factor_dest_a = Regs::Blend::Factor::Zero;
     }
+    regs.stencil_front_op_fail = Regs::StencilOp::Keep;
+    regs.stencil_front_op_zfail = Regs::StencilOp::Keep;
+    regs.stencil_front_op_zpass = Regs::StencilOp::Keep;
+    regs.stencil_front_func_func = Regs::ComparisonOp::Always;
+    regs.stencil_front_func_mask = 0xFFFFFFFF;
+    regs.stencil_front_mask = 0xFFFFFFFF;
+    regs.stencil_two_side_enable = 1;
+    regs.stencil_back_op_fail = Regs::StencilOp::Keep;
+    regs.stencil_back_op_zfail = Regs::StencilOp::Keep;
+    regs.stencil_back_op_zpass = Regs::StencilOp::Keep;
+    regs.stencil_back_func_func = Regs::ComparisonOp::Always;
+    regs.stencil_back_func_mask = 0xFFFFFFFF;
+    regs.stencil_back_mask = 0xFFFFFFFF;
+    // TODO(Rodrigo): Most games do not set a point size. I think this is a case of a
+    // register carrying a default value. Assume it's OpenGL's default (1).
+    regs.point_size = 1.0f;
 }
 
 void Maxwell3D::CallMacroMethod(u32 method, std::vector<u32> parameters) {
@@ -108,7 +124,13 @@ void Maxwell3D::WriteReg(u32 method, u32 value, u32 remaining_params) {
         debug_context->OnEvent(Tegra::DebugContext::Event::MaxwellCommandLoaded, nullptr);
     }
 
-    regs.reg_array[method] = value;
+    if (regs.reg_array[method] != value) {
+        regs.reg_array[method] = value;
+        if (method >= MAXWELL3D_REG_INDEX(vertex_attrib_format) &&
+            method < MAXWELL3D_REG_INDEX(vertex_attrib_format) + regs.vertex_attrib_format.size()) {
+            dirty_flags.vertex_attrib_format = true;
+        }
+    }
 
     switch (method) {
     case MAXWELL3D_REG_INDEX(macros.data): {

src/video_core/engines/maxwell_3d.h

@@ -345,6 +345,14 @@ public:
             Invert = 6,
             IncrWrap = 7,
             DecrWrap = 8,
+            KeepOGL = 0x1E00,
+            ZeroOGL = 0,
+            ReplaceOGL = 0x1E01,
+            IncrOGL = 0x1E02,
+            DecrOGL = 0x1E03,
+            InvertOGL = 0x150A,
+            IncrWrapOGL = 0x8507,
+            DecrWrapOGL = 0x8508,
         };
 
         enum class MemoryLayout : u32 {
@@ -472,6 +480,67 @@ public:
             };
         };
 
+        struct ViewportTransform {
+            f32 scale_x;
+            f32 scale_y;
+            f32 scale_z;
+            f32 translate_x;
+            f32 translate_y;
+            f32 translate_z;
+            INSERT_PADDING_WORDS(2);
+
+            MathUtil::Rectangle<s32> GetRect() const {
+                return {
+                    GetX(),               // left
+                    GetY() + GetHeight(), // top
+                    GetX() + GetWidth(),  // right
+                    GetY()                // bottom
+                };
+            };
+
+            s32 GetX() const {
+                return static_cast<s32>(std::max(0.0f, translate_x - std::fabs(scale_x)));
+            }
+
+            s32 GetY() const {
+                return static_cast<s32>(std::max(0.0f, translate_y - std::fabs(scale_y)));
+            }
+
+            s32 GetWidth() const {
+                return static_cast<s32>(translate_x + std::fabs(scale_x)) - GetX();
+            }
+
+            s32 GetHeight() const {
+                return static_cast<s32>(translate_y + std::fabs(scale_y)) - GetY();
+            }
+        };
+
+        struct ScissorTest {
+            u32 enable;
+            union {
+                BitField<0, 16, u32> min_x;
+                BitField<16, 16, u32> max_x;
+            };
+            union {
+                BitField<0, 16, u32> min_y;
+                BitField<16, 16, u32> max_y;
+            };
+            u32 fill;
+        };
+
+        struct ViewPort {
+            union {
+                BitField<0, 16, u32> x;
+                BitField<16, 16, u32> width;
+            };
+            union {
+                BitField<0, 16, u32> y;
+                BitField<16, 16, u32> height;
+            };
+            float depth_range_near;
+            float depth_range_far;
+        };
+
         bool IsShaderConfigEnabled(std::size_t index) const {
             // The VertexB is always enabled.
             if (index == static_cast<std::size_t>(Regs::ShaderProgram::VertexB)) {
@@ -497,55 +566,11 @@ public:
 
                 INSERT_PADDING_WORDS(0x2E);
 
-                RenderTargetConfig rt[NumRenderTargets];
+                std::array<RenderTargetConfig, NumRenderTargets> rt;
 
-                struct {
-                    f32 scale_x;
-                    f32 scale_y;
-                    f32 scale_z;
-                    f32 translate_x;
-                    f32 translate_y;
-                    f32 translate_z;
-                    INSERT_PADDING_WORDS(2);
+                std::array<ViewportTransform, NumViewports> viewport_transform;
 
-                    MathUtil::Rectangle<s32> GetRect() const {
-                        return {
-                            GetX(),               // left
-                            GetY() + GetHeight(), // top
-                            GetX() + GetWidth(),  // right
-                            GetY()                // bottom
-                        };
-                    };
-
-                    s32 GetX() const {
-                        return static_cast<s32>(std::max(0.0f, translate_x - std::fabs(scale_x)));
-                    }
-
-                    s32 GetY() const {
-                        return static_cast<s32>(std::max(0.0f, translate_y - std::fabs(scale_y)));
-                    }
-
-                    s32 GetWidth() const {
-                        return static_cast<s32>(translate_x + std::fabs(scale_x)) - GetX();
-                    }
-
-                    s32 GetHeight() const {
-                        return static_cast<s32>(translate_y + std::fabs(scale_y)) - GetY();
-                    }
-                } viewport_transform[NumViewports];
-
-                struct {
-                    union {
-                        BitField<0, 16, u32> x;
-                        BitField<16, 16, u32> width;
-                    };
-                    union {
-                        BitField<0, 16, u32> y;
-                        BitField<16, 16, u32> height;
-                    };
-                    float depth_range_near;
-                    float depth_range_far;
-                } viewport[NumViewports];
+                std::array<ViewPort, NumViewports> viewports;
 
                 INSERT_PADDING_WORDS(0x1D);
 
@@ -563,19 +588,9 @@ public:
 
                 INSERT_PADDING_WORDS(0x17);
 
-                struct {
-                    u32 enable;
-                    union {
-                        BitField<0, 16, u32> min_x;
-                        BitField<16, 16, u32> max_x;
-                    };
-                    union {
-                        BitField<0, 16, u32> min_y;
-                        BitField<16, 16, u32> max_y;
-                    };
-                } scissor_test;
+                std::array<ScissorTest, NumViewports> scissor_test;
 
-                INSERT_PADDING_WORDS(0x52);
+                INSERT_PADDING_WORDS(0x15);
 
                 s32 stencil_back_func_ref;
                 u32 stencil_back_mask;
@@ -692,7 +707,9 @@ public:
                 u32 stencil_front_func_mask;
                 u32 stencil_front_mask;
 
-                INSERT_PADDING_WORDS(0x3);
+                INSERT_PADDING_WORDS(0x2);
+
+                u32 frag_color_clamp;
 
                 union {
                     BitField<4, 1, u32> triangle_rast_flip;
@@ -710,7 +727,12 @@ public:
 
                 u32 zeta_enable;
 
-                INSERT_PADDING_WORDS(0x8);
+                union {
+                    BitField<0, 1, u32> alpha_to_coverage;
+                    BitField<4, 1, u32> alpha_to_one;
+                } multisample_control;
+
+                INSERT_PADDING_WORDS(0x7);
 
                 struct {
                     u32 tsc_address_high;
@@ -1004,6 +1026,12 @@ public:
     State state{};
     MemoryManager& memory_manager;
 
+    struct DirtyFlags {
+        bool vertex_attrib_format = true;
+    };
+
+    DirtyFlags dirty_flags;
+
     /// Reads a register value located at the input method address
     u32 GetRegisterValue(u32 method) const;
 
@@ -1086,8 +1114,8 @@ private:
 ASSERT_REG_POSITION(macros, 0x45);
 ASSERT_REG_POSITION(tfb_enabled, 0x1D1);
 ASSERT_REG_POSITION(rt, 0x200);
-ASSERT_REG_POSITION(viewport_transform[0], 0x280);
-ASSERT_REG_POSITION(viewport, 0x300);
+ASSERT_REG_POSITION(viewport_transform, 0x280);
+ASSERT_REG_POSITION(viewports, 0x300);
 ASSERT_REG_POSITION(vertex_buffer, 0x35D);
 ASSERT_REG_POSITION(clear_color[0], 0x360);
 ASSERT_REG_POSITION(clear_depth, 0x364);
@@ -1122,10 +1150,12 @@ ASSERT_REG_POSITION(stencil_front_func_func, 0x4E4);
 ASSERT_REG_POSITION(stencil_front_func_ref, 0x4E5);
 ASSERT_REG_POSITION(stencil_front_func_mask, 0x4E6);
 ASSERT_REG_POSITION(stencil_front_mask, 0x4E7);
+ASSERT_REG_POSITION(frag_color_clamp, 0x4EA);
 ASSERT_REG_POSITION(screen_y_control, 0x4EB);
 ASSERT_REG_POSITION(vb_element_base, 0x50D);
 ASSERT_REG_POSITION(point_size, 0x546);
 ASSERT_REG_POSITION(zeta_enable, 0x54E);
+ASSERT_REG_POSITION(multisample_control, 0x54F);
 ASSERT_REG_POSITION(tsc, 0x557);
 ASSERT_REG_POSITION(tic, 0x55D);
 ASSERT_REG_POSITION(stencil_two_side_enable, 0x565);

src/video_core/rasterizer_cache.cpp

@@ -0,0 +1,7 @@
+// Copyright 2018 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include "video_core/rasterizer_cache.h"
+
+RasterizerCacheObject::~RasterizerCacheObject() = default;

src/video_core/rasterizer_cache.h

@@ -5,6 +5,7 @@
 #pragma once
 
 #include <set>
+#include <unordered_map>
 
 #include <boost/icl/interval_map.hpp>
 #include <boost/range/iterator_range_core.hpp>
@@ -15,6 +16,8 @@
 
 class RasterizerCacheObject {
 public:
+    virtual ~RasterizerCacheObject();
+
     /// Gets the address of the shader in guest memory, required for cache management
     virtual VAddr GetAddr() const = 0;
 
@@ -86,29 +89,25 @@ public:
 
     /// Invalidates everything in the cache
     void InvalidateAll() {
-        while (object_cache.begin() != object_cache.end()) {
-            Unregister(*object_cache.begin()->second.begin());
+        while (interval_cache.begin() != interval_cache.end()) {
+            Unregister(*interval_cache.begin()->second.begin());
         }
     }
 
 protected:
     /// Tries to get an object from the cache with the specified address
     T TryGet(VAddr addr) const {
-        const ObjectInterval interval{addr};
-        for (auto& pair : boost::make_iterator_range(object_cache.equal_range(interval))) {
-            for (auto& cached_object : pair.second) {
-                if (cached_object->GetAddr() == addr) {
-                    return cached_object;
-                }
-            }
-        }
+        const auto iter = map_cache.find(addr);
+        if (iter != map_cache.end())
+            return iter->second;
         return nullptr;
     }
 
     /// Register an object into the cache
     void Register(const T& object) {
         object->SetIsRegistered(true);
-        object_cache.add({GetInterval(object), ObjectSet{object}});
+        interval_cache.add({GetInterval(object), ObjectSet{object}});
+        map_cache.insert({object->GetAddr(), object});
         rasterizer.UpdatePagesCachedCount(object->GetAddr(), object->GetSizeInBytes(), 1);
     }
 
@@ -116,13 +115,13 @@ protected:
     void Unregister(const T& object) {
         object->SetIsRegistered(false);
         rasterizer.UpdatePagesCachedCount(object->GetAddr(), object->GetSizeInBytes(), -1);
-
         // Only flush if use_accurate_gpu_emulation is enabled, as it incurs a performance hit
         if (Settings::values.use_accurate_gpu_emulation) {
             FlushObject(object);
         }
 
-        object_cache.subtract({GetInterval(object), ObjectSet{object}});
+        interval_cache.subtract({GetInterval(object), ObjectSet{object}});
+        map_cache.erase(object->GetAddr());
     }
 
     /// Returns a ticks counter used for tracking when cached objects were last modified
@@ -139,7 +138,7 @@ private:
 
         std::vector<T> objects;
         const ObjectInterval interval{addr, addr + size};
-        for (auto& pair : boost::make_iterator_range(object_cache.equal_range(interval))) {
+        for (auto& pair : boost::make_iterator_range(interval_cache.equal_range(interval))) {
             for (auto& cached_object : pair.second) {
                 if (!cached_object) {
                     continue;
@@ -165,15 +164,17 @@ private:
     }
 
     using ObjectSet = std::set<T>;
-    using ObjectCache = boost::icl::interval_map<VAddr, ObjectSet>;
-    using ObjectInterval = typename ObjectCache::interval_type;
+    using ObjectCache = std::unordered_map<VAddr, T>;
+    using IntervalCache = boost::icl::interval_map<VAddr, ObjectSet>;
+    using ObjectInterval = typename IntervalCache::interval_type;
 
     static auto GetInterval(const T& object) {
         return ObjectInterval::right_open(object->GetAddr(),
                                           object->GetAddr() + object->GetSizeInBytes());
     }
 
-    ObjectCache object_cache; ///< Cache of objects
-    u64 modified_ticks{};     ///< Counter of cache state ticks, used for in-order flushing
+    ObjectCache map_cache;
+    IntervalCache interval_cache; ///< Cache of objects
+    u64 modified_ticks{};         ///< Counter of cache state ticks, used for in-order flushing
     VideoCore::RasterizerInterface& rasterizer;
 };

src/video_core/renderer_base.cpp

@@ -5,7 +5,6 @@
 #include "core/frontend/emu_window.h"
 #include "core/settings.h"
 #include "video_core/renderer_base.h"
-#include "video_core/renderer_opengl/gl_rasterizer.h"
 
 namespace VideoCore {
 

src/video_core/renderer_opengl/gl_rasterizer.cpp

@@ -33,7 +33,8 @@ using Maxwell = Tegra::Engines::Maxwell3D::Regs;
 using PixelFormat = VideoCore::Surface::PixelFormat;
 using SurfaceType = VideoCore::Surface::SurfaceType;
 
-MICROPROFILE_DEFINE(OpenGL_VAO, "OpenGL", "Vertex Array Setup", MP_RGB(128, 128, 192));
+MICROPROFILE_DEFINE(OpenGL_VAO, "OpenGL", "Vertex Format Setup", MP_RGB(128, 128, 192));
+MICROPROFILE_DEFINE(OpenGL_VB, "OpenGL", "Vertex Buffer Setup", MP_RGB(128, 128, 192));
 MICROPROFILE_DEFINE(OpenGL_Shader, "OpenGL", "Shader Setup", MP_RGB(128, 128, 192));
 MICROPROFILE_DEFINE(OpenGL_UBO, "OpenGL", "Const Buffer Setup", MP_RGB(128, 128, 192));
 MICROPROFILE_DEFINE(OpenGL_Index, "OpenGL", "Index Buffer Setup", MP_RGB(128, 128, 192));
@@ -106,8 +107,6 @@ RasterizerOpenGL::RasterizerOpenGL(Core::Frontend::EmuWindow& window, ScreenInfo
 
     ASSERT_MSG(has_ARB_separate_shader_objects, "has_ARB_separate_shader_objects is unsupported");
     OpenGLState::ApplyDefaultState();
-    // Clipping plane 0 is always enabled for PICA fixed clip plane z <= 0
-    state.clip_distance[0] = true;
 
     // Create render framebuffer
     framebuffer.Create();
@@ -123,18 +122,23 @@ RasterizerOpenGL::RasterizerOpenGL(Core::Frontend::EmuWindow& window, ScreenInfo
 
 RasterizerOpenGL::~RasterizerOpenGL() {}
 
-void RasterizerOpenGL::SetupVertexArrays() {
-    MICROPROFILE_SCOPE(OpenGL_VAO);
-    const auto& gpu = Core::System::GetInstance().GPU().Maxwell3D();
+void RasterizerOpenGL::SetupVertexFormat() {
+    auto& gpu = Core::System::GetInstance().GPU().Maxwell3D();
     const auto& regs = gpu.regs;
 
+    if (!gpu.dirty_flags.vertex_attrib_format)
+        return;
+    gpu.dirty_flags.vertex_attrib_format = false;
+
+    MICROPROFILE_SCOPE(OpenGL_VAO);
+
     auto [iter, is_cache_miss] = vertex_array_cache.try_emplace(regs.vertex_attrib_format);
     auto& VAO = iter->second;
 
     if (is_cache_miss) {
         VAO.Create();
         state.draw.vertex_array = VAO.handle;
-        state.Apply();
+        state.ApplyVertexBufferState();
 
         // The index buffer binding is stored within the VAO. Stupid OpenGL, but easy to work
         // around.
@@ -176,8 +180,13 @@ void RasterizerOpenGL::SetupVertexArrays() {
         }
     }
     state.draw.vertex_array = VAO.handle;
-    state.draw.vertex_buffer = buffer_cache.GetHandle();
-    state.Apply();
+    state.ApplyVertexBufferState();
+}
+
+void RasterizerOpenGL::SetupVertexBuffer() {
+    MICROPROFILE_SCOPE(OpenGL_VB);
+    const auto& gpu = Core::System::GetInstance().GPU().Maxwell3D();
+    const auto& regs = gpu.regs;
 
     // Upload all guest vertex arrays sequentially to our buffer
     for (u32 index = 0; index < Maxwell::NumVertexArrays; ++index) {
@@ -204,6 +213,9 @@ void RasterizerOpenGL::SetupVertexArrays() {
             glVertexBindingDivisor(index, 0);
         }
     }
+
+    // Implicit set by glBindVertexBuffer. Stupid glstate handling...
+    state.draw.vertex_buffer = buffer_cache.GetHandle();
 }
 
 DrawParameters RasterizerOpenGL::SetupDraw() {
@@ -328,8 +340,6 @@ void RasterizerOpenGL::SetupShaders(GLenum primitive_mode) {
             index++;
         }
     }
-
-    state.Apply();
 }
 
 std::size_t RasterizerOpenGL::CalculateVertexArraysSize() const {
@@ -398,8 +408,8 @@ void RasterizerOpenGL::UpdatePagesCachedCount(VAddr addr, u64 size, int delta) {
         cached_pages.add({pages_interval, delta});
 }
 
-void RasterizerOpenGL::ConfigureFramebuffers(bool using_color_fb, bool using_depth_fb,
-                                             bool preserve_contents,
+void RasterizerOpenGL::ConfigureFramebuffers(OpenGLState& current_state, bool using_color_fb,
+                                             bool using_depth_fb, bool preserve_contents,
                                              std::optional<std::size_t> single_color_target) {
     MICROPROFILE_SCOPE(OpenGL_Framebuffer);
     const auto& regs = Core::System::GetInstance().GPU().Maxwell3D().regs;
@@ -415,9 +425,9 @@ void RasterizerOpenGL::ConfigureFramebuffers(bool using_color_fb, bool using_dep
     ASSERT_MSG(regs.rt_separate_frag_data == 0, "Unimplemented");
 
     // Bind the framebuffer surfaces
-    state.draw.draw_framebuffer = framebuffer.handle;
-    state.Apply();
-    state.framebuffer_srgb.enabled = regs.framebuffer_srgb != 0;
+    current_state.draw.draw_framebuffer = framebuffer.handle;
+    current_state.ApplyFramebufferState();
+    current_state.framebuffer_srgb.enabled = regs.framebuffer_srgb != 0;
 
     if (using_color_fb) {
         if (single_color_target) {
@@ -495,10 +505,7 @@ void RasterizerOpenGL::ConfigureFramebuffers(bool using_color_fb, bool using_dep
         glFramebufferTexture2D(GL_DRAW_FRAMEBUFFER, GL_DEPTH_STENCIL_ATTACHMENT, GL_TEXTURE_2D, 0,
                                0);
     }
-
-    SyncViewport();
-
-    state.Apply();
+    SyncViewport(current_state);
 }
 
 void RasterizerOpenGL::Clear() {
@@ -511,22 +518,23 @@ void RasterizerOpenGL::Clear() {
     bool use_stencil{};
 
     OpenGLState clear_state;
-    clear_state.draw.draw_framebuffer = framebuffer.handle;
-    clear_state.color_mask[0].red_enabled = regs.clear_buffers.R ? GL_TRUE : GL_FALSE;
-    clear_state.color_mask[0].green_enabled = regs.clear_buffers.G ? GL_TRUE : GL_FALSE;
-    clear_state.color_mask[0].blue_enabled = regs.clear_buffers.B ? GL_TRUE : GL_FALSE;
-    clear_state.color_mask[0].alpha_enabled = regs.clear_buffers.A ? GL_TRUE : GL_FALSE;
-
     if (regs.clear_buffers.R || regs.clear_buffers.G || regs.clear_buffers.B ||
         regs.clear_buffers.A) {
         use_color = true;
     }
+    if (use_color) {
+        clear_state.color_mask[0].red_enabled = regs.clear_buffers.R ? GL_TRUE : GL_FALSE;
+        clear_state.color_mask[0].green_enabled = regs.clear_buffers.G ? GL_TRUE : GL_FALSE;
+        clear_state.color_mask[0].blue_enabled = regs.clear_buffers.B ? GL_TRUE : GL_FALSE;
+        clear_state.color_mask[0].alpha_enabled = regs.clear_buffers.A ? GL_TRUE : GL_FALSE;
+    }
     if (regs.clear_buffers.Z) {
         ASSERT_MSG(regs.zeta_enable != 0, "Tried to clear Z but buffer is not enabled!");
         use_depth = true;
 
         // Always enable the depth write when clearing the depth buffer. The depth write mask is
-        // ignored when clearing the buffer in the Switch, but OpenGL obeys it so we set it to true.
+        // ignored when clearing the buffer in the Switch, but OpenGL obeys it so we set it to
+        // true.
         clear_state.depth.test_enabled = true;
         clear_state.depth.test_func = GL_ALWAYS;
     }
@@ -543,11 +551,8 @@ void RasterizerOpenGL::Clear() {
 
     ScopeAcquireGLContext acquire_context{emu_window};
 
-    ConfigureFramebuffers(use_color, use_depth || use_stencil, false,
+    ConfigureFramebuffers(clear_state, use_color, use_depth || use_stencil, false,
                           regs.clear_buffers.RT.Value());
-    // Copy the sRGB setting to the clear state to avoid problem with
-    // specific driver implementations
-    clear_state.framebuffer_srgb.enabled = state.framebuffer_srgb.enabled;
     clear_state.Apply();
 
     if (use_color) {
@@ -573,8 +578,10 @@ void RasterizerOpenGL::DrawArrays() {
 
     ScopeAcquireGLContext acquire_context{emu_window};
 
-    ConfigureFramebuffers();
+    ConfigureFramebuffers(state);
     SyncColorMask();
+    SyncFragmentColorClampState();
+    SyncMultiSampleState();
     SyncDepthTestState();
     SyncStencilTestState();
     SyncBlendState();
@@ -594,7 +601,7 @@ void RasterizerOpenGL::DrawArrays() {
     const bool is_indexed = accelerate_draw == AccelDraw::Indexed;
 
     state.draw.vertex_buffer = buffer_cache.GetHandle();
-    state.Apply();
+    state.ApplyVertexBufferState();
 
     std::size_t buffer_size = CalculateVertexArraysSize();
 
@@ -621,7 +628,8 @@ void RasterizerOpenGL::DrawArrays() {
 
     buffer_cache.Map(buffer_size);
 
-    SetupVertexArrays();
+    SetupVertexFormat();
+    SetupVertexBuffer();
     DrawParameters params = SetupDraw();
     SetupShaders(params.primitive_mode);
 
@@ -634,7 +642,7 @@ void RasterizerOpenGL::DrawArrays() {
     params.DispatchDraw();
 
     // Disable scissor test
-    state.scissor.enabled = false;
+    state.viewports[0].scissor.enabled = false;
 
     accelerate_draw = AccelDraw::Disabled;
 
@@ -727,7 +735,6 @@ void RasterizerOpenGL::SamplerInfo::Create() {
 
 void RasterizerOpenGL::SamplerInfo::SyncWithConfig(const Tegra::Texture::TSCEntry& config) {
     const GLuint s = sampler.handle;
-
     if (mag_filter != config.mag_filter) {
         mag_filter = config.mag_filter;
         glSamplerParameteri(
@@ -769,15 +776,51 @@ void RasterizerOpenGL::SamplerInfo::SyncWithConfig(const Tegra::Texture::TSCEntr
                             MaxwellToGL::DepthCompareFunc(depth_compare_func));
     }
 
-    if (wrap_u == Tegra::Texture::WrapMode::Border || wrap_v == Tegra::Texture::WrapMode::Border ||
-        wrap_p == Tegra::Texture::WrapMode::Border) {
-        const GLvec4 new_border_color = {{config.border_color_r, config.border_color_g,
-                                          config.border_color_b, config.border_color_a}};
-        if (border_color != new_border_color) {
-            border_color = new_border_color;
-            glSamplerParameterfv(s, GL_TEXTURE_BORDER_COLOR, border_color.data());
+    GLvec4 new_border_color;
+    if (config.srgb_conversion) {
+        new_border_color[0] = config.srgb_border_color_r / 255.0f;
+        new_border_color[1] = config.srgb_border_color_g / 255.0f;
+        new_border_color[2] = config.srgb_border_color_g / 255.0f;
+    } else {
+        new_border_color[0] = config.border_color_r;
+        new_border_color[1] = config.border_color_g;
+        new_border_color[2] = config.border_color_b;
+    }
+    new_border_color[3] = config.border_color_a;
+
+    if (border_color != new_border_color) {
+        border_color = new_border_color;
+        glSamplerParameterfv(s, GL_TEXTURE_BORDER_COLOR, border_color.data());
+    }
+
+    const float anisotropic_max = static_cast<float>(1 << config.max_anisotropy.Value());
+    if (anisotropic_max != max_anisotropic) {
+        max_anisotropic = anisotropic_max;
+        if (GLAD_GL_ARB_texture_filter_anisotropic) {
+            glSamplerParameterf(s, GL_TEXTURE_MAX_ANISOTROPY, max_anisotropic);
+        } else if (GLAD_GL_EXT_texture_filter_anisotropic) {
+            glSamplerParameterf(s, GL_TEXTURE_MAX_ANISOTROPY_EXT, max_anisotropic);
         }
     }
+    const float lod_min = static_cast<float>(config.min_lod_clamp.Value()) / 256.0f;
+    if (lod_min != min_lod) {
+        min_lod = lod_min;
+        glSamplerParameterf(s, GL_TEXTURE_MIN_LOD, min_lod);
+    }
+
+    const float lod_max = static_cast<float>(config.max_lod_clamp.Value()) / 256.0f;
+    if (lod_max != max_lod) {
+        max_lod = lod_max;
+        glSamplerParameterf(s, GL_TEXTURE_MAX_LOD, max_lod);
+    }
+    const u32 bias = config.mip_lod_bias.Value();
+    // Sign extend the 13-bit value.
+    const u32 mask = 1U << (13 - 1);
+    const float bias_lod = static_cast<s32>((bias ^ mask) - mask) / 256.f;
+    if (lod_bias != bias_lod) {
+        lod_bias = bias_lod;
+        glSamplerParameterf(s, GL_TEXTURE_LOD_BIAS, lod_bias);
+    }
 }
 
 u32 RasterizerOpenGL::SetupConstBuffers(Maxwell::ShaderStage stage, Shader& shader,
@@ -897,17 +940,17 @@ u32 RasterizerOpenGL::SetupTextures(Maxwell::ShaderStage stage, Shader& shader,
     return current_unit + static_cast<u32>(entries.size());
 }
 
-void RasterizerOpenGL::SyncViewport() {
+void RasterizerOpenGL::SyncViewport(OpenGLState& current_state) {
     const auto& regs = Core::System::GetInstance().GPU().Maxwell3D().regs;
-    for (size_t i = 0; i < Tegra::Engines::Maxwell3D::Regs::NumRenderTargets; i++) {
+    for (std::size_t i = 0; i < Tegra::Engines::Maxwell3D::Regs::NumViewports; i++) {
         const MathUtil::Rectangle<s32> viewport_rect{regs.viewport_transform[i].GetRect()};
-        auto& viewport = state.viewports[i];
+        auto& viewport = current_state.viewports[i];
         viewport.x = viewport_rect.left;
         viewport.y = viewport_rect.bottom;
-        viewport.width = static_cast<GLsizei>(viewport_rect.GetWidth());
-        viewport.height = static_cast<GLsizei>(viewport_rect.GetHeight());
-        viewport.depth_range_far = regs.viewport[i].depth_range_far;
-        viewport.depth_range_near = regs.viewport[i].depth_range_near;
+        viewport.width = static_cast<GLfloat>(viewport_rect.GetWidth());
+        viewport.height = static_cast<GLfloat>(viewport_rect.GetHeight());
+        viewport.depth_range_far = regs.viewports[i].depth_range_far;
+        viewport.depth_range_near = regs.viewports[i].depth_range_near;
     }
 }
 
@@ -970,9 +1013,6 @@ void RasterizerOpenGL::SyncStencilTestState() {
         return;
     }
 
-    // TODO(bunnei): Verify behavior when this is not set
-    ASSERT(regs.stencil_two_side_enable);
-
     state.stencil.front.test_func = MaxwellToGL::ComparisonOp(regs.stencil_front_func_func);
     state.stencil.front.test_ref = regs.stencil_front_func_ref;
     state.stencil.front.test_mask = regs.stencil_front_func_mask;
@@ -980,19 +1020,30 @@ void RasterizerOpenGL::SyncStencilTestState() {
     state.stencil.front.action_depth_fail = MaxwellToGL::StencilOp(regs.stencil_front_op_zfail);
     state.stencil.front.action_depth_pass = MaxwellToGL::StencilOp(regs.stencil_front_op_zpass);
     state.stencil.front.write_mask = regs.stencil_front_mask;
-
-    state.stencil.back.test_func = MaxwellToGL::ComparisonOp(regs.stencil_back_func_func);
-    state.stencil.back.test_ref = regs.stencil_back_func_ref;
-    state.stencil.back.test_mask = regs.stencil_back_func_mask;
-    state.stencil.back.action_stencil_fail = MaxwellToGL::StencilOp(regs.stencil_back_op_fail);
-    state.stencil.back.action_depth_fail = MaxwellToGL::StencilOp(regs.stencil_back_op_zfail);
-    state.stencil.back.action_depth_pass = MaxwellToGL::StencilOp(regs.stencil_back_op_zpass);
-    state.stencil.back.write_mask = regs.stencil_back_mask;
+    if (regs.stencil_two_side_enable) {
+        state.stencil.back.test_func = MaxwellToGL::ComparisonOp(regs.stencil_back_func_func);
+        state.stencil.back.test_ref = regs.stencil_back_func_ref;
+        state.stencil.back.test_mask = regs.stencil_back_func_mask;
+        state.stencil.back.action_stencil_fail = MaxwellToGL::StencilOp(regs.stencil_back_op_fail);
+        state.stencil.back.action_depth_fail = MaxwellToGL::StencilOp(regs.stencil_back_op_zfail);
+        state.stencil.back.action_depth_pass = MaxwellToGL::StencilOp(regs.stencil_back_op_zpass);
+        state.stencil.back.write_mask = regs.stencil_back_mask;
+    } else {
+        state.stencil.back.test_func = GL_ALWAYS;
+        state.stencil.back.test_ref = 0;
+        state.stencil.back.test_mask = 0xFFFFFFFF;
+        state.stencil.back.write_mask = 0xFFFFFFFF;
+        state.stencil.back.action_stencil_fail = GL_KEEP;
+        state.stencil.back.action_depth_fail = GL_KEEP;
+        state.stencil.back.action_depth_pass = GL_KEEP;
+    }
 }
 
 void RasterizerOpenGL::SyncColorMask() {
     const auto& regs = Core::System::GetInstance().GPU().Maxwell3D().regs;
-    for (size_t i = 0; i < Tegra::Engines::Maxwell3D::Regs::NumRenderTargets; i++) {
+    const std::size_t count =
+        regs.independent_blend_enable ? Tegra::Engines::Maxwell3D::Regs::NumRenderTargets : 1;
+    for (std::size_t i = 0; i < count; i++) {
         const auto& source = regs.color_mask[regs.color_mask_common ? 0 : i];
         auto& dest = state.color_mask[i];
         dest.red_enabled = (source.R == 0) ? GL_FALSE : GL_TRUE;
@@ -1002,6 +1053,17 @@ void RasterizerOpenGL::SyncColorMask() {
     }
 }
 
+void RasterizerOpenGL::SyncMultiSampleState() {
+    const auto& regs = Core::System::GetInstance().GPU().Maxwell3D().regs;
+    state.multisample_control.alpha_to_coverage = regs.multisample_control.alpha_to_coverage != 0;
+    state.multisample_control.alpha_to_one = regs.multisample_control.alpha_to_one != 0;
+}
+
+void RasterizerOpenGL::SyncFragmentColorClampState() {
+    const auto& regs = Core::System::GetInstance().GPU().Maxwell3D().regs;
+    state.fragment_color_clamp.enabled = regs.frag_color_clamp != 0;
+}
+
 void RasterizerOpenGL::SyncBlendState() {
     const auto& regs = Core::System::GetInstance().GPU().Maxwell3D().regs;
 
@@ -1013,43 +1075,40 @@ void RasterizerOpenGL::SyncBlendState() {
     state.independant_blend.enabled = regs.independent_blend_enable;
     if (!state.independant_blend.enabled) {
         auto& blend = state.blend[0];
-        blend.enabled = regs.blend.enable[0] != 0;
-        blend.separate_alpha = regs.blend.separate_alpha;
-        blend.rgb_equation = MaxwellToGL::BlendEquation(regs.blend.equation_rgb);
-        blend.src_rgb_func = MaxwellToGL::BlendFunc(regs.blend.factor_source_rgb);
-        blend.dst_rgb_func = MaxwellToGL::BlendFunc(regs.blend.factor_dest_rgb);
-        if (blend.separate_alpha) {
-            blend.a_equation = MaxwellToGL::BlendEquation(regs.blend.equation_a);
-            blend.src_a_func = MaxwellToGL::BlendFunc(regs.blend.factor_source_a);
-            blend.dst_a_func = MaxwellToGL::BlendFunc(regs.blend.factor_dest_a);
+        const auto& src = regs.blend;
+        blend.enabled = src.enable[0] != 0;
+        if (blend.enabled) {
+            blend.rgb_equation = MaxwellToGL::BlendEquation(src.equation_rgb);
+            blend.src_rgb_func = MaxwellToGL::BlendFunc(src.factor_source_rgb);
+            blend.dst_rgb_func = MaxwellToGL::BlendFunc(src.factor_dest_rgb);
+            blend.a_equation = MaxwellToGL::BlendEquation(src.equation_a);
+            blend.src_a_func = MaxwellToGL::BlendFunc(src.factor_source_a);
+            blend.dst_a_func = MaxwellToGL::BlendFunc(src.factor_dest_a);
         }
-        for (size_t i = 1; i < Tegra::Engines::Maxwell3D::Regs::NumRenderTargets; i++) {
+        for (std::size_t i = 1; i < Tegra::Engines::Maxwell3D::Regs::NumRenderTargets; i++) {
             state.blend[i].enabled = false;
         }
         return;
     }
 
-    for (size_t i = 0; i < Tegra::Engines::Maxwell3D::Regs::NumRenderTargets; i++) {
+    for (std::size_t i = 0; i < Tegra::Engines::Maxwell3D::Regs::NumRenderTargets; i++) {
         auto& blend = state.blend[i];
+        const auto& src = regs.independent_blend[i];
         blend.enabled = regs.blend.enable[i] != 0;
         if (!blend.enabled)
             continue;
-        blend.separate_alpha = regs.independent_blend[i].separate_alpha;
-        blend.rgb_equation = MaxwellToGL::BlendEquation(regs.independent_blend[i].equation_rgb);
-        blend.src_rgb_func = MaxwellToGL::BlendFunc(regs.independent_blend[i].factor_source_rgb);
-        blend.dst_rgb_func = MaxwellToGL::BlendFunc(regs.independent_blend[i].factor_dest_rgb);
-        if (blend.separate_alpha) {
-            blend.a_equation = MaxwellToGL::BlendEquation(regs.independent_blend[i].equation_a);
-            blend.src_a_func = MaxwellToGL::BlendFunc(regs.independent_blend[i].factor_source_a);
-            blend.dst_a_func = MaxwellToGL::BlendFunc(regs.independent_blend[i].factor_dest_a);
-        }
+        blend.rgb_equation = MaxwellToGL::BlendEquation(src.equation_rgb);
+        blend.src_rgb_func = MaxwellToGL::BlendFunc(src.factor_source_rgb);
+        blend.dst_rgb_func = MaxwellToGL::BlendFunc(src.factor_dest_rgb);
+        blend.a_equation = MaxwellToGL::BlendEquation(src.equation_a);
+        blend.src_a_func = MaxwellToGL::BlendFunc(src.factor_source_a);
+        blend.dst_a_func = MaxwellToGL::BlendFunc(src.factor_dest_a);
     }
 }
 
 void RasterizerOpenGL::SyncLogicOpState() {
     const auto& regs = Core::System::GetInstance().GPU().Maxwell3D().regs;
 
-    // TODO(Subv): Support more than just render target 0.
     state.logic_op.enabled = regs.logic_op.enable != 0;
 
     if (!state.logic_op.enabled)
@@ -1062,19 +1121,21 @@ void RasterizerOpenGL::SyncLogicOpState() {
 }
 
 void RasterizerOpenGL::SyncScissorTest() {
-    // TODO: what is the correct behavior here, a single scissor for all targets
-    // or scissor disabled for the rest of the targets?
     const auto& regs = Core::System::GetInstance().GPU().Maxwell3D().regs;
-    state.scissor.enabled = (regs.scissor_test.enable != 0);
-    if (regs.scissor_test.enable == 0) {
-        return;
+    for (std::size_t i = 0; i < Tegra::Engines::Maxwell3D::Regs::NumViewports; i++) {
+        const auto& src = regs.scissor_test[i];
+        auto& dst = state.viewports[i].scissor;
+        dst.enabled = (src.enable != 0);
+        if (dst.enabled == 0) {
+            return;
+        }
+        const u32 width = src.max_x - src.min_x;
+        const u32 height = src.max_y - src.min_y;
+        dst.x = src.min_x;
+        dst.y = src.min_y;
+        dst.width = width;
+        dst.height = height;
     }
-    const u32 width = regs.scissor_test.max_x - regs.scissor_test.min_x;
-    const u32 height = regs.scissor_test.max_y - regs.scissor_test.min_y;
-    state.scissor.x = regs.scissor_test.min_x;
-    state.scissor.y = regs.scissor_test.min_y;
-    state.scissor.width = width;
-    state.scissor.height = height;
 }
 
 void RasterizerOpenGL::SyncTransformFeedback() {
@@ -1088,20 +1149,15 @@ void RasterizerOpenGL::SyncTransformFeedback() {
 
 void RasterizerOpenGL::SyncPointState() {
     const auto& regs = Core::System::GetInstance().GPU().Maxwell3D().regs;
-
-    // TODO(Rodrigo): Most games do not set a point size. I think this is a case of a
-    // register carrying a default value. For now, if the point size is zero, assume it's
-    // OpenGL's default (1).
-    state.point.size = regs.point_size == 0 ? 1 : regs.point_size;
+    state.point.size = regs.point_size;
 }
 
 void RasterizerOpenGL::CheckAlphaTests() {
     const auto& regs = Core::System::GetInstance().GPU().Maxwell3D().regs;
 
     if (regs.alpha_test_enabled != 0 && regs.rt_control.count > 1) {
-        LOG_CRITICAL(
-            Render_OpenGL,
-            "Alpha Testing is enabled with Multiple Render Targets, this behavior is undefined.");
+        LOG_CRITICAL(Render_OpenGL, "Alpha Testing is enabled with Multiple Render Targets, "
+                                    "this behavior is undefined.");
         UNREACHABLE();
     }
 }

src/video_core/renderer_opengl/gl_rasterizer.h

@@ -88,7 +88,7 @@ private:
         /// SamplerInfo struct.
         void Create();
         /// Syncs the sampler object with the config, updating any necessary state.
-        void SyncWithConfig(const Tegra::Texture::TSCEntry& config);
+        void SyncWithConfig(const Tegra::Texture::TSCEntry& info);
 
     private:
         Tegra::Texture::TextureFilter mag_filter;
@@ -100,6 +100,10 @@ private:
         bool uses_depth_compare;
         Tegra::Texture::DepthCompareFunc depth_compare_func;
         GLvec4 border_color;
+        float min_lod;
+        float max_lod;
+        float lod_bias;
+        float max_anisotropic;
     };
 
     /**
@@ -109,8 +113,8 @@ private:
      * @param preserve_contents If true, tries to preserve data from a previously used framebuffer.
      * @param single_color_target Specifies if a single color buffer target should be used.
      */
-    void ConfigureFramebuffers(bool use_color_fb = true, bool using_depth_fb = true,
-                               bool preserve_contents = true,
+    void ConfigureFramebuffers(OpenGLState& current_state, bool use_color_fb = true,
+                               bool using_depth_fb = true, bool preserve_contents = true,
                                std::optional<std::size_t> single_color_target = {});
 
     /*
@@ -134,7 +138,7 @@ private:
                       GLenum primitive_mode, u32 current_unit);
 
     /// Syncs the viewport and depth range to match the guest state
-    void SyncViewport();
+    void SyncViewport(OpenGLState& current_state);
 
     /// Syncs the clip enabled status to match the guest state
     void SyncClipEnabled();
@@ -160,6 +164,12 @@ private:
     /// Syncs the LogicOp state to match the guest state
     void SyncLogicOpState();
 
+    /// Syncs the the color clamp state
+    void SyncFragmentColorClampState();
+
+    /// Syncs the alpha coverage and alpha to one
+    void SyncMultiSampleState();
+
     /// Syncs the scissor test state to match the guest state
     void SyncScissorTest();
 
@@ -207,7 +217,8 @@ private:
 
     std::size_t CalculateIndexBufferSize() const;
 
-    void SetupVertexArrays();
+    void SetupVertexFormat();
+    void SetupVertexBuffer();
 
     DrawParameters SetupDraw();
 

src/video_core/renderer_opengl/gl_rasterizer_cache.cpp

@@ -314,6 +314,8 @@ static constexpr std::array<FormatTuple, VideoCore::Surface::MaxPixelFormat> tex
     {GL_SRGB8_ALPHA8, GL_RGBA, GL_UNSIGNED_BYTE, ComponentType::UNorm, false}, // ASTC_2D_5X4_SRGB
     {GL_RGBA8, GL_RGBA, GL_UNSIGNED_BYTE, ComponentType::UNorm, false},        // ASTC_2D_5X5
     {GL_SRGB8_ALPHA8, GL_RGBA, GL_UNSIGNED_BYTE, ComponentType::UNorm, false}, // ASTC_2D_5X5_SRGB
+    {GL_RGBA8, GL_RGBA, GL_UNSIGNED_BYTE, ComponentType::UNorm, false},        // ASTC_2D_10X8
+    {GL_SRGB8_ALPHA8, GL_RGBA, GL_UNSIGNED_BYTE, ComponentType::UNorm, false}, // ASTC_2D_10X8_SRGB
 
     // Depth formats
     {GL_DEPTH_COMPONENT32F, GL_DEPTH_COMPONENT, GL_FLOAT, ComponentType::Float, false}, // Z32F
@@ -379,11 +381,8 @@ void MortonCopy(u32 stride, u32 block_height, u32 height, u32 block_depth, u32 d
     const u32 tile_size_y{GetDefaultBlockHeight(format)};
 
     if (morton_to_gl) {
-        const std::vector<u8> data =
-            Tegra::Texture::UnswizzleTexture(addr, tile_size_x, tile_size_y, bytes_per_pixel,
-                                             stride, height, depth, block_height, block_depth);
-        const std::size_t size_to_copy{std::min(gl_buffer_size, data.size())};
-        memcpy(gl_buffer, data.data(), size_to_copy);
+        Tegra::Texture::UnswizzleTexture(gl_buffer, addr, tile_size_x, tile_size_y, bytes_per_pixel,
+                                         stride, height, depth, block_height, block_depth);
     } else {
         Tegra::Texture::CopySwizzledData((stride + tile_size_x - 1) / tile_size_x,
                                          (height + tile_size_y - 1) / tile_size_y, depth,
@@ -456,6 +455,8 @@ static constexpr GLConversionArray morton_to_gl_fns = {
         MortonCopy<true, PixelFormat::ASTC_2D_5X4_SRGB>,
         MortonCopy<true, PixelFormat::ASTC_2D_5X5>,
         MortonCopy<true, PixelFormat::ASTC_2D_5X5_SRGB>,
+        MortonCopy<true, PixelFormat::ASTC_2D_10X8>,
+        MortonCopy<true, PixelFormat::ASTC_2D_10X8_SRGB>,
         MortonCopy<true, PixelFormat::Z32F>,
         MortonCopy<true, PixelFormat::Z16>,
         MortonCopy<true, PixelFormat::Z24S8>,
@@ -526,6 +527,8 @@ static constexpr GLConversionArray gl_to_morton_fns = {
         nullptr,
         nullptr,
         nullptr,
+        nullptr,
+        nullptr,
         MortonCopy<false, PixelFormat::Z32F>,
         MortonCopy<false, PixelFormat::Z16>,
         MortonCopy<false, PixelFormat::Z24S8>,
@@ -544,8 +547,8 @@ void SwizzleFunc(const GLConversionArray& functions, const SurfaceParams& params
     if (params.is_layered) {
         u64 offset = params.GetMipmapLevelOffset(mip_level);
         u64 offset_gl = 0;
-        u64 layer_size = params.LayerMemorySize();
-        u64 gl_size = params.LayerSizeGL(mip_level);
+        const u64 layer_size = params.LayerMemorySize();
+        const u64 gl_size = params.LayerSizeGL(mip_level);
         for (u32 i = 0; i < params.depth; i++) {
             functions[static_cast<std::size_t>(params.pixel_format)](
                 params.MipWidth(mip_level), params.MipBlockHeight(mip_level),
@@ -555,7 +558,7 @@ void SwizzleFunc(const GLConversionArray& functions, const SurfaceParams& params
             offset_gl += gl_size;
         }
     } else {
-        u64 offset = params.GetMipmapLevelOffset(mip_level);
+        const u64 offset = params.GetMipmapLevelOffset(mip_level);
         functions[static_cast<std::size_t>(params.pixel_format)](
             params.MipWidth(mip_level), params.MipBlockHeight(mip_level),
             params.MipHeight(mip_level), params.MipBlockDepth(mip_level), depth, gl_buffer.data(),
@@ -580,7 +583,7 @@ static bool BlitSurface(const Surface& src_surface, const Surface& dst_surface,
     state.draw.draw_framebuffer = draw_fb_handle;
     // Set sRGB enabled if the destination surfaces need it
     state.framebuffer_srgb.enabled = dst_params.srgb_conversion;
-    state.Apply();
+    state.ApplyFramebufferState();
 
     u32 buffers{};
 
@@ -709,18 +712,18 @@ static void FastCopySurface(const Surface& src_surface, const Surface& dst_surfa
 
 MICROPROFILE_DEFINE(OpenGL_CopySurface, "OpenGL", "CopySurface", MP_RGB(128, 192, 64));
 static void CopySurface(const Surface& src_surface, const Surface& dst_surface,
-                        GLuint copy_pbo_handle, GLenum src_attachment = 0,
-                        GLenum dst_attachment = 0, std::size_t cubemap_face = 0) {
+                        const GLuint copy_pbo_handle, const GLenum src_attachment = 0,
+                        const GLenum dst_attachment = 0, const std::size_t cubemap_face = 0) {
     MICROPROFILE_SCOPE(OpenGL_CopySurface);
     ASSERT_MSG(dst_attachment == 0, "Unimplemented");
 
     const auto& src_params{src_surface->GetSurfaceParams()};
     const auto& dst_params{dst_surface->GetSurfaceParams()};
 
-    auto source_format = GetFormatTuple(src_params.pixel_format, src_params.component_type);
-    auto dest_format = GetFormatTuple(dst_params.pixel_format, dst_params.component_type);
+    const auto source_format = GetFormatTuple(src_params.pixel_format, src_params.component_type);
+    const auto dest_format = GetFormatTuple(dst_params.pixel_format, dst_params.component_type);
 
-    std::size_t buffer_size = std::max(src_params.size_in_bytes, dst_params.size_in_bytes);
+    const std::size_t buffer_size = std::max(src_params.size_in_bytes, dst_params.size_in_bytes);
 
     glBindBuffer(GL_PIXEL_PACK_BUFFER, copy_pbo_handle);
     glBufferData(GL_PIXEL_PACK_BUFFER, buffer_size, nullptr, GL_STREAM_DRAW_ARB);
@@ -744,13 +747,10 @@ static void CopySurface(const Surface& src_surface, const Surface& dst_surface,
             LOG_DEBUG(HW_GPU, "Trying to upload extra texture data from the CPU during "
                               "reinterpretation but the texture is tiled.");
         }
-        std::size_t remaining_size = dst_params.size_in_bytes - src_params.size_in_bytes;
-        std::vector<u8> data(remaining_size);
-        std::memcpy(data.data(), Memory::GetPointer(dst_params.addr + src_params.size_in_bytes),
-                    data.size());
+        const std::size_t remaining_size = dst_params.size_in_bytes - src_params.size_in_bytes;
 
         glBufferSubData(GL_PIXEL_PACK_BUFFER, src_params.size_in_bytes, remaining_size,
-                        data.data());
+                        Memory::GetPointer(dst_params.addr + src_params.size_in_bytes));
     }
 
     glBindBuffer(GL_PIXEL_PACK_BUFFER, 0);
@@ -932,7 +932,9 @@ static void ConvertFormatAsNeeded_LoadGLBuffer(std::vector<u8>& data, PixelForma
     case PixelFormat::ASTC_2D_8X8_SRGB:
     case PixelFormat::ASTC_2D_8X5_SRGB:
     case PixelFormat::ASTC_2D_5X4_SRGB:
-    case PixelFormat::ASTC_2D_5X5_SRGB: {
+    case PixelFormat::ASTC_2D_5X5_SRGB:
+    case PixelFormat::ASTC_2D_10X8:
+    case PixelFormat::ASTC_2D_10X8_SRGB: {
         // Convert ASTC pixel formats to RGBA8, as most desktop GPUs do not support ASTC.
         u32 block_width{};
         u32 block_height{};
@@ -967,7 +969,11 @@ static void ConvertFormatAsNeeded_FlushGLBuffer(std::vector<u8>& data, PixelForm
     case PixelFormat::ASTC_2D_4X4:
     case PixelFormat::ASTC_2D_8X8:
     case PixelFormat::ASTC_2D_4X4_SRGB:
-    case PixelFormat::ASTC_2D_8X8_SRGB: {
+    case PixelFormat::ASTC_2D_8X8_SRGB:
+    case PixelFormat::ASTC_2D_5X5:
+    case PixelFormat::ASTC_2D_5X5_SRGB:
+    case PixelFormat::ASTC_2D_10X8:
+    case PixelFormat::ASTC_2D_10X8_SRGB: {
         LOG_CRITICAL(HW_GPU, "Conversion of format {} after texture flushing is not implemented",
                      static_cast<u32>(pixel_format));
         UNREACHABLE();
@@ -1336,6 +1342,7 @@ Surface RasterizerCacheOpenGL::RecreateSurface(const Surface& old_surface,
         break;
     case SurfaceTarget::TextureCubemap:
     case SurfaceTarget::Texture3D:
+    case SurfaceTarget::Texture2DArray:
     case SurfaceTarget::TextureCubeArray:
         AccurateCopySurface(old_surface, new_surface);
         break;

src/video_core/renderer_opengl/gl_shader_cache.cpp

@@ -121,12 +121,16 @@ GLint CachedShader::GetUniformLocation(const GLShader::SamplerEntry& sampler) {
 }
 
 GLuint CachedShader::LazyGeometryProgram(OGLProgram& target_program,
-                                         const std::string& glsl_topology,
+                                         const std::string& glsl_topology, u32 max_vertices,
                                          const std::string& debug_name) {
     if (target_program.handle != 0) {
         return target_program.handle;
     }
-    const std::string source{geometry_programs.code + "layout (" + glsl_topology + ") in;\n"};
+    std::string source = "#version 430 core\n";
+    source += "layout (" + glsl_topology + ") in;\n";
+    source += "#define MAX_VERTEX_INPUT " + std::to_string(max_vertices) + '\n';
+    source += geometry_programs.code;
+
     OGLShader shader;
     shader.Create(source.c_str(), GL_GEOMETRY_SHADER);
     target_program.Create(true, shader.handle);

src/video_core/renderer_opengl/gl_shader_cache.h

@@ -48,22 +48,23 @@ public:
         }
         switch (primitive_mode) {
         case GL_POINTS:
-            return LazyGeometryProgram(geometry_programs.points, "points", "ShaderPoints");
+            return LazyGeometryProgram(geometry_programs.points, "points", 1, "ShaderPoints");
         case GL_LINES:
         case GL_LINE_STRIP:
-            return LazyGeometryProgram(geometry_programs.lines, "lines", "ShaderLines");
+            return LazyGeometryProgram(geometry_programs.lines, "lines", 2, "ShaderLines");
         case GL_LINES_ADJACENCY:
         case GL_LINE_STRIP_ADJACENCY:
-            return LazyGeometryProgram(geometry_programs.lines_adjacency, "lines_adjacency",
+            return LazyGeometryProgram(geometry_programs.lines_adjacency, "lines_adjacency", 4,
                                        "ShaderLinesAdjacency");
         case GL_TRIANGLES:
         case GL_TRIANGLE_STRIP:
         case GL_TRIANGLE_FAN:
-            return LazyGeometryProgram(geometry_programs.triangles, "triangles", "ShaderTriangles");
+            return LazyGeometryProgram(geometry_programs.triangles, "triangles", 3,
+                                       "ShaderTriangles");
         case GL_TRIANGLES_ADJACENCY:
         case GL_TRIANGLE_STRIP_ADJACENCY:
             return LazyGeometryProgram(geometry_programs.triangles_adjacency, "triangles_adjacency",
-                                       "ShaderLines");
+                                       6, "ShaderTrianglesAdjacency");
         default:
             UNREACHABLE_MSG("Unknown primitive mode.");
         }
@@ -78,7 +79,7 @@ public:
 private:
     /// Generates a geometry shader or returns one that already exists.
     GLuint LazyGeometryProgram(OGLProgram& target_program, const std::string& glsl_topology,
-                               const std::string& debug_name);
+                               u32 max_vertices, const std::string& debug_name);
 
     VAddr addr;
     Maxwell::ShaderProgram program_type;

src/video_core/renderer_opengl/gl_shader_decompiler.cpp

@@ -494,10 +494,10 @@ public:
             // instruction for now.
             if (stage == Maxwell3D::Regs::ShaderStage::Geometry) {
                 // TODO(Rodrigo): nouveau sets some attributes after setting emitting a geometry
-                // shader. These instructions use a dirty register as buffer index. To avoid some
-                // drivers from complaining for the out of boundary writes, guard them.
-                const std::string buf_index{"min(" + GetRegisterAsInteger(buf_reg) + ", " +
-                                            std::to_string(MAX_GEOMETRY_BUFFERS - 1) + ')'};
+                // shader. These instructions use a dirty register as buffer index, to avoid some
+                // drivers from complaining about out of boundary writes, guard them.
+                const std::string buf_index{"((" + GetRegisterAsInteger(buf_reg) + ") % " +
+                                            std::to_string(MAX_GEOMETRY_BUFFERS) + ')'};
                 shader.AddLine("amem[" + buf_index + "][" +
                                std::to_string(static_cast<u32>(attribute)) + ']' +
                                GetSwizzle(elem) + " = " + src + ';');
@@ -811,7 +811,11 @@ private:
                                   std::optional<Register> vertex = {}) {
         auto GeometryPass = [&](const std::string& name) {
             if (stage == Maxwell3D::Regs::ShaderStage::Geometry && vertex) {
-                return "gs_" + name + '[' + GetRegisterAsInteger(*vertex, 0, false) + ']';
+                // TODO(Rodrigo): Guard geometry inputs against out of bound reads. Some games set
+                // an 0x80000000 index for those and the shader fails to build. Find out why this
+                // happens and what's its intent.
+                return "gs_" + name + '[' + GetRegisterAsInteger(*vertex, 0, false) +
+                       " % MAX_VERTEX_INPUT]";
             }
             return name;
         };
@@ -2742,12 +2746,12 @@ private:
                 }
                 case 3: {
                     if (is_array) {
-                        UNIMPLEMENTED_MSG("3-coordinate arrays not fully implemented");
-                        const std::string x = regs.GetRegisterAsFloat(instr.gpr8);
-                        const std::string y = regs.GetRegisterAsFloat(instr.gpr20);
-                        coord = "vec2 coords = vec2(" + x + ", " + y + ");";
-                        texture_type = Tegra::Shader::TextureType::Texture2D;
-                        is_array = false;
+                        const std::string index = regs.GetRegisterAsInteger(instr.gpr8);
+                        const std::string x = regs.GetRegisterAsFloat(instr.gpr8.Value() + 1);
+                        const std::string y = regs.GetRegisterAsFloat(instr.gpr8.Value() + 2);
+                        const std::string z = regs.GetRegisterAsFloat(instr.gpr20);
+                        coord =
+                            "vec4 coords = vec4(" + x + ", " + y + ", " + z + ", " + index + ");";
                     } else {
                         const std::string x = regs.GetRegisterAsFloat(instr.gpr8);
                         const std::string y = regs.GetRegisterAsFloat(instr.gpr8.Value() + 1);
@@ -2777,7 +2781,11 @@ private:
                     break;
                 }
                 case Tegra::Shader::TextureProcessMode::LZ: {
-                    texture = "textureLod(" + sampler + ", coords, 0.0)";
+                    if (depth_compare && is_array) {
+                        texture = "texture(" + sampler + ", coords)";
+                    } else {
+                        texture = "textureLod(" + sampler + ", coords, 0.0)";
+                    }
                     break;
                 }
                 case Tegra::Shader::TextureProcessMode::LL: {

src/video_core/renderer_opengl/gl_shader_gen.cpp

@@ -82,8 +82,8 @@ void main() {
 }
 
 ProgramResult GenerateGeometryShader(const ShaderSetup& setup) {
-    std::string out = "#version 430 core\n";
-    out += "#extension GL_ARB_separate_shader_objects : enable\n\n";
+    // Version is intentionally skipped in shader generation, it's added by the lazy compilation.
+    std::string out = "#extension GL_ARB_separate_shader_objects : enable\n\n";
     out += Decompiler::GetCommonDeclarations();
     out += "bool exec_geometry();\n";
 

src/video_core/renderer_opengl/gl_shader_manager.h

@@ -67,6 +67,7 @@ public:
         glUseProgramStages(pipeline.handle, GL_FRAGMENT_SHADER_BIT, fs);
         state.draw.shader_program = 0;
         state.draw.program_pipeline = pipeline.handle;
+        state.geometry_shaders.enabled = (gs != 0);
     }
 
 private:

src/video_core/renderer_opengl/gl_state.cpp

@@ -14,7 +14,10 @@ OpenGLState OpenGLState::cur_state;
 bool OpenGLState::s_rgb_used;
 OpenGLState::OpenGLState() {
     // These all match default OpenGL values
+    geometry_shaders.enabled = false;
     framebuffer_srgb.enabled = false;
+    multisample_control.alpha_to_coverage = false;
+    multisample_control.alpha_to_one = false;
     cull.enabled = false;
     cull.mode = GL_BACK;
     cull.front_face = GL_CCW;
@@ -50,12 +53,12 @@ OpenGLState::OpenGLState() {
         item.height = 0;
         item.depth_range_near = 0.0f;
         item.depth_range_far = 1.0f;
+        item.scissor.enabled = false;
+        item.scissor.x = 0;
+        item.scissor.y = 0;
+        item.scissor.width = 0;
+        item.scissor.height = 0;
     }
-    scissor.enabled = false;
-    scissor.x = 0;
-    scissor.y = 0;
-    scissor.width = 0;
-    scissor.height = 0;
     for (auto& item : blend) {
         item.enabled = true;
         item.rgb_equation = GL_FUNC_ADD;
@@ -88,6 +91,7 @@ OpenGLState::OpenGLState() {
     clip_distance = {};
 
     point.size = 1;
+    fragment_color_clamp.enabled = false;
 }
 
 void OpenGLState::ApplyDefaultState() {
@@ -136,7 +140,7 @@ void OpenGLState::ApplyCulling() const {
 }
 
 void OpenGLState::ApplyColorMask() const {
-    if (GLAD_GL_ARB_viewport_array) {
+    if (GLAD_GL_ARB_viewport_array && independant_blend.enabled) {
         for (size_t i = 0; i < Tegra::Engines::Maxwell3D::Regs::NumRenderTargets; i++) {
             const auto& updated = color_mask[i];
             const auto& current = cur_state.color_mask[i];
@@ -230,26 +234,10 @@ void OpenGLState::ApplyStencilTest() const {
     }
 }
 
-void OpenGLState::ApplyScissor() const {
-    const bool scissor_changed = scissor.enabled != cur_state.scissor.enabled;
-    if (scissor_changed) {
-        if (scissor.enabled) {
-            glEnable(GL_SCISSOR_TEST);
-        } else {
-            glDisable(GL_SCISSOR_TEST);
-        }
-    }
-    if (scissor.enabled &&
-        (scissor_changed || scissor.x != cur_state.scissor.x || scissor.y != cur_state.scissor.y ||
-         scissor.width != cur_state.scissor.width || scissor.height != cur_state.scissor.height)) {
-        glScissor(scissor.x, scissor.y, scissor.width, scissor.height);
-    }
-}
-
 void OpenGLState::ApplyViewport() const {
-    if (GLAD_GL_ARB_viewport_array) {
-        for (GLuint i = 0;
-             i < static_cast<GLuint>(Tegra::Engines::Maxwell3D::Regs::NumRenderTargets); i++) {
+    if (GLAD_GL_ARB_viewport_array && geometry_shaders.enabled) {
+        for (GLuint i = 0; i < static_cast<GLuint>(Tegra::Engines::Maxwell3D::Regs::NumViewports);
+             i++) {
             const auto& current = cur_state.viewports[i];
             const auto& updated = viewports[i];
             if (updated.x != current.x || updated.y != current.y ||
@@ -260,18 +248,50 @@ void OpenGLState::ApplyViewport() const {
                 updated.depth_range_far != current.depth_range_far) {
                 glDepthRangeIndexed(i, updated.depth_range_near, updated.depth_range_far);
             }
+            const bool scissor_changed = updated.scissor.enabled != current.scissor.enabled;
+            if (scissor_changed) {
+                if (updated.scissor.enabled) {
+                    glEnablei(GL_SCISSOR_TEST, i);
+                } else {
+                    glDisablei(GL_SCISSOR_TEST, i);
+                }
+            }
+            if (updated.scissor.enabled &&
+                (scissor_changed || updated.scissor.x != current.scissor.x ||
+                 updated.scissor.y != current.scissor.y ||
+                 updated.scissor.width != current.scissor.width ||
+                 updated.scissor.height != current.scissor.height)) {
+                glScissorIndexed(i, updated.scissor.x, updated.scissor.y, updated.scissor.width,
+                                 updated.scissor.height);
+            }
         }
     } else {
         const auto& current = cur_state.viewports[0];
         const auto& updated = viewports[0];
         if (updated.x != current.x || updated.y != current.y || updated.width != current.width ||
             updated.height != current.height) {
-            glViewport(updated.x, updated.y, updated.width, updated.height);
+            glViewport(static_cast<GLint>(updated.x), static_cast<GLint>(updated.y),
+                       static_cast<GLsizei>(updated.width), static_cast<GLsizei>(updated.height));
         }
         if (updated.depth_range_near != current.depth_range_near ||
             updated.depth_range_far != current.depth_range_far) {
             glDepthRange(updated.depth_range_near, updated.depth_range_far);
         }
+        const bool scissor_changed = updated.scissor.enabled != current.scissor.enabled;
+        if (scissor_changed) {
+            if (updated.scissor.enabled) {
+                glEnable(GL_SCISSOR_TEST);
+            } else {
+                glDisable(GL_SCISSOR_TEST);
+            }
+        }
+        if (updated.scissor.enabled && (scissor_changed || updated.scissor.x != current.scissor.x ||
+                                        updated.scissor.y != current.scissor.y ||
+                                        updated.scissor.width != current.scissor.width ||
+                                        updated.scissor.height != current.scissor.height)) {
+            glScissor(updated.scissor.x, updated.scissor.y, updated.scissor.width,
+                      updated.scissor.height);
+        }
     }
 }
 
@@ -289,31 +309,20 @@ void OpenGLState::ApplyGlobalBlending() const {
     if (!updated.enabled) {
         return;
     }
-    if (updated.separate_alpha) {
-        if (blend_changed || updated.src_rgb_func != current.src_rgb_func ||
-            updated.dst_rgb_func != current.dst_rgb_func ||
-            updated.src_a_func != current.src_a_func || updated.dst_a_func != current.dst_a_func) {
-            glBlendFuncSeparate(updated.src_rgb_func, updated.dst_rgb_func, updated.src_a_func,
-                                updated.dst_a_func);
-        }
+    if (blend_changed || updated.src_rgb_func != current.src_rgb_func ||
+        updated.dst_rgb_func != current.dst_rgb_func || updated.src_a_func != current.src_a_func ||
+        updated.dst_a_func != current.dst_a_func) {
+        glBlendFuncSeparate(updated.src_rgb_func, updated.dst_rgb_func, updated.src_a_func,
+                            updated.dst_a_func);
+    }
 
-        if (blend_changed || updated.rgb_equation != current.rgb_equation ||
-            updated.a_equation != current.a_equation) {
-            glBlendEquationSeparate(updated.rgb_equation, updated.a_equation);
-        }
-    } else {
-        if (blend_changed || updated.src_rgb_func != current.src_rgb_func ||
-            updated.dst_rgb_func != current.dst_rgb_func) {
-            glBlendFunc(updated.src_rgb_func, updated.dst_rgb_func);
-        }
-
-        if (blend_changed || updated.rgb_equation != current.rgb_equation) {
-            glBlendEquation(updated.rgb_equation);
-        }
+    if (blend_changed || updated.rgb_equation != current.rgb_equation ||
+        updated.a_equation != current.a_equation) {
+        glBlendEquationSeparate(updated.rgb_equation, updated.a_equation);
     }
 }
 
-void OpenGLState::ApplyTargetBlending(int target, bool force) const {
+void OpenGLState::ApplyTargetBlending(std::size_t target, bool force) const {
     const Blend& updated = blend[target];
     const Blend& current = cur_state.blend[target];
     const bool blend_changed = updated.enabled != current.enabled || force;
@@ -327,29 +336,17 @@ void OpenGLState::ApplyTargetBlending(int target, bool force) const {
     if (!updated.enabled) {
         return;
     }
-    if (updated.separate_alpha) {
-        if (blend_changed || updated.src_rgb_func != current.src_rgb_func ||
-            updated.dst_rgb_func != current.dst_rgb_func ||
-            updated.src_a_func != current.src_a_func || updated.dst_a_func != current.dst_a_func) {
-            glBlendFuncSeparateiARB(static_cast<GLuint>(target), updated.src_rgb_func,
-                                    updated.dst_rgb_func, updated.src_a_func, updated.dst_a_func);
-        }
+    if (blend_changed || updated.src_rgb_func != current.src_rgb_func ||
+        updated.dst_rgb_func != current.dst_rgb_func || updated.src_a_func != current.src_a_func ||
+        updated.dst_a_func != current.dst_a_func) {
+        glBlendFuncSeparateiARB(static_cast<GLuint>(target), updated.src_rgb_func,
+                                updated.dst_rgb_func, updated.src_a_func, updated.dst_a_func);
+    }
 
-        if (blend_changed || updated.rgb_equation != current.rgb_equation ||
-            updated.a_equation != current.a_equation) {
-            glBlendEquationSeparateiARB(static_cast<GLuint>(target), updated.rgb_equation,
-                                        updated.a_equation);
-        }
-    } else {
-        if (blend_changed || updated.src_rgb_func != current.src_rgb_func ||
-            updated.dst_rgb_func != current.dst_rgb_func) {
-            glBlendFunciARB(static_cast<GLuint>(target), updated.src_rgb_func,
-                            updated.dst_rgb_func);
-        }
-
-        if (blend_changed || updated.rgb_equation != current.rgb_equation) {
-            glBlendEquationiARB(static_cast<GLuint>(target), updated.rgb_equation);
-        }
+    if (blend_changed || updated.rgb_equation != current.rgb_equation ||
+        updated.a_equation != current.a_equation) {
+        glBlendEquationSeparateiARB(static_cast<GLuint>(target), updated.rgb_equation,
+                                    updated.a_equation);
     }
 }
 
@@ -427,7 +424,7 @@ void OpenGLState::ApplySamplers() const {
     }
 }
 
-void OpenGLState::Apply() const {
+void OpenGLState::ApplyFramebufferState() const {
     // Framebuffer
     if (draw.read_framebuffer != cur_state.draw.read_framebuffer) {
         glBindFramebuffer(GL_READ_FRAMEBUFFER, draw.read_framebuffer);
@@ -435,7 +432,9 @@ void OpenGLState::Apply() const {
     if (draw.draw_framebuffer != cur_state.draw.draw_framebuffer) {
         glBindFramebuffer(GL_DRAW_FRAMEBUFFER, draw.draw_framebuffer);
     }
+}
 
+void OpenGLState::ApplyVertexBufferState() const {
     // Vertex array
     if (draw.vertex_array != cur_state.draw.vertex_array) {
         glBindVertexArray(draw.vertex_array);
@@ -445,7 +444,11 @@ void OpenGLState::Apply() const {
     if (draw.vertex_buffer != cur_state.draw.vertex_buffer) {
         glBindBuffer(GL_ARRAY_BUFFER, draw.vertex_buffer);
     }
+}
 
+void OpenGLState::Apply() const {
+    ApplyFramebufferState();
+    ApplyVertexBufferState();
     // Uniform buffer
     if (draw.uniform_buffer != cur_state.draw.uniform_buffer) {
         glBindBuffer(GL_UNIFORM_BUFFER, draw.uniform_buffer);
@@ -474,9 +477,29 @@ void OpenGLState::Apply() const {
     if (point.size != cur_state.point.size) {
         glPointSize(point.size);
     }
+    if (GLAD_GL_ARB_color_buffer_float) {
+        if (fragment_color_clamp.enabled != cur_state.fragment_color_clamp.enabled) {
+            glClampColor(GL_CLAMP_FRAGMENT_COLOR_ARB,
+                         fragment_color_clamp.enabled ? GL_TRUE : GL_FALSE);
+        }
+    }
+    if (multisample_control.alpha_to_coverage != cur_state.multisample_control.alpha_to_coverage) {
+        if (multisample_control.alpha_to_coverage) {
+            glEnable(GL_SAMPLE_ALPHA_TO_COVERAGE);
+        } else {
+            glDisable(GL_SAMPLE_ALPHA_TO_COVERAGE);
+        }
+    }
+    if (multisample_control.alpha_to_one != cur_state.multisample_control.alpha_to_one) {
+        if (multisample_control.alpha_to_one) {
+            glEnable(GL_SAMPLE_ALPHA_TO_ONE);
+        } else {
+            glDisable(GL_SAMPLE_ALPHA_TO_ONE);
+        }
+    }
+
     ApplyColorMask();
     ApplyViewport();
-    ApplyScissor();
     ApplyStencilTest();
     ApplySRgb();
     ApplyCulling();

src/video_core/renderer_opengl/gl_state.h

@@ -39,6 +39,19 @@ public:
         bool enabled; // GL_FRAMEBUFFER_SRGB
     } framebuffer_srgb;
 
+    struct {
+        bool alpha_to_coverage; // GL_ALPHA_TO_COVERAGE
+        bool alpha_to_one;      // GL_ALPHA_TO_ONE
+    } multisample_control;
+
+    struct {
+        bool enabled; // GL_CLAMP_FRAGMENT_COLOR_ARB
+    } fragment_color_clamp;
+
+    struct {
+        bool enabled; // viewports arrays are only supported when geometry shaders are enabled.
+    } geometry_shaders;
+
     struct {
         bool enabled;      // GL_CULL_FACE
         GLenum mode;       // GL_CULL_FACE_MODE
@@ -79,7 +92,6 @@ public:
 
     struct Blend {
         bool enabled;        // GL_BLEND
-        bool separate_alpha; // Independent blend enabled
         GLenum rgb_equation; // GL_BLEND_EQUATION_RGB
         GLenum a_equation;   // GL_BLEND_EQUATION_ALPHA
         GLenum src_rgb_func; // GL_BLEND_SRC_RGB
@@ -150,16 +162,15 @@ public:
         GLfloat height;
         GLfloat depth_range_near; // GL_DEPTH_RANGE
         GLfloat depth_range_far;  // GL_DEPTH_RANGE
+        struct {
+            bool enabled; // GL_SCISSOR_TEST
+            GLint x;
+            GLint y;
+            GLsizei width;
+            GLsizei height;
+        } scissor;
     };
-    std::array<viewport, Tegra::Engines::Maxwell3D::Regs::NumRenderTargets> viewports;
-
-    struct {
-        bool enabled; // GL_SCISSOR_TEST
-        GLint x;
-        GLint y;
-        GLsizei width;
-        GLsizei height;
-    } scissor;
+    std::array<viewport, Tegra::Engines::Maxwell3D::Regs::NumViewports> viewports;
 
     struct {
         float size; // GL_POINT_SIZE
@@ -181,6 +192,10 @@ public:
     }
     /// Apply this state as the current OpenGL state
     void Apply() const;
+    /// Apply only the state afecting the framebuffer
+    void ApplyFramebufferState() const;
+    /// Apply only the state afecting the vertex buffer
+    void ApplyVertexBufferState() const;
     /// Set the initial OpenGL state
     static void ApplyDefaultState();
     /// Resets any references to the given resource
@@ -204,13 +219,12 @@ private:
     void ApplyPrimitiveRestart() const;
     void ApplyStencilTest() const;
     void ApplyViewport() const;
-    void ApplyTargetBlending(int target, bool force) const;
+    void ApplyTargetBlending(std::size_t target, bool force) const;
     void ApplyGlobalBlending() const;
     void ApplyBlending() const;
     void ApplyLogicOp() const;
     void ApplyTextures() const;
     void ApplySamplers() const;
-    void ApplyScissor() const;
 };
 
 } // namespace OpenGL

src/video_core/renderer_opengl/maxwell_to_gl.h

@@ -159,10 +159,8 @@ inline GLenum TextureFilterMode(Tegra::Texture::TextureFilter filter_mode,
         }
     }
     }
-    LOG_CRITICAL(Render_OpenGL, "Unimplemented texture filter mode={}",
-                 static_cast<u32>(filter_mode));
-    UNREACHABLE();
-    return {};
+    LOG_ERROR(Render_OpenGL, "Unimplemented texture filter mode={}", static_cast<u32>(filter_mode));
+    return GL_LINEAR;
 }
 
 inline GLenum WrapMode(Tegra::Texture::WrapMode wrap_mode) {
@@ -182,10 +180,15 @@ inline GLenum WrapMode(Tegra::Texture::WrapMode wrap_mode) {
         return GL_CLAMP_TO_BORDER;
     case Tegra::Texture::WrapMode::MirrorOnceClampToEdge:
         return GL_MIRROR_CLAMP_TO_EDGE;
+    case Tegra::Texture::WrapMode::MirrorOnceBorder:
+        if (GL_EXT_texture_mirror_clamp) {
+            return GL_MIRROR_CLAMP_TO_BORDER_EXT;
+        } else {
+            return GL_MIRROR_CLAMP_TO_EDGE;
+        }
     }
-    LOG_CRITICAL(Render_OpenGL, "Unimplemented texture wrap mode={}", static_cast<u32>(wrap_mode));
-    UNREACHABLE();
-    return {};
+    LOG_ERROR(Render_OpenGL, "Unimplemented texture wrap mode={}", static_cast<u32>(wrap_mode));
+    return GL_REPEAT;
 }
 
 inline GLenum DepthCompareFunc(Tegra::Texture::DepthCompareFunc func) {
@@ -207,10 +210,9 @@ inline GLenum DepthCompareFunc(Tegra::Texture::DepthCompareFunc func) {
     case Tegra::Texture::DepthCompareFunc::Always:
         return GL_ALWAYS;
     }
-    LOG_CRITICAL(Render_OpenGL, "Unimplemented texture depth compare function ={}",
-                 static_cast<u32>(func));
-    UNREACHABLE();
-    return {};
+    LOG_ERROR(Render_OpenGL, "Unimplemented texture depth compare function ={}",
+              static_cast<u32>(func));
+    return GL_GREATER;
 }
 
 inline GLenum BlendEquation(Maxwell::Blend::Equation equation) {
@@ -226,9 +228,8 @@ inline GLenum BlendEquation(Maxwell::Blend::Equation equation) {
     case Maxwell::Blend::Equation::Max:
         return GL_MAX;
     }
-    LOG_CRITICAL(Render_OpenGL, "Unimplemented blend equation={}", static_cast<u32>(equation));
-    UNREACHABLE();
-    return {};
+    LOG_ERROR(Render_OpenGL, "Unimplemented blend equation={}", static_cast<u32>(equation));
+    return GL_FUNC_ADD;
 }
 
 inline GLenum BlendFunc(Maxwell::Blend::Factor factor) {
@@ -291,9 +292,8 @@ inline GLenum BlendFunc(Maxwell::Blend::Factor factor) {
     case Maxwell::Blend::Factor::OneMinusConstantAlphaGL:
         return GL_ONE_MINUS_CONSTANT_ALPHA;
     }
-    LOG_CRITICAL(Render_OpenGL, "Unimplemented blend factor={}", static_cast<u32>(factor));
-    UNREACHABLE();
-    return {};
+    LOG_ERROR(Render_OpenGL, "Unimplemented blend factor={}", static_cast<u32>(factor));
+    return GL_ZERO;
 }
 
 inline GLenum SwizzleSource(Tegra::Texture::SwizzleSource source) {
@@ -312,9 +312,8 @@ inline GLenum SwizzleSource(Tegra::Texture::SwizzleSource source) {
     case Tegra::Texture::SwizzleSource::OneFloat:
         return GL_ONE;
     }
-    LOG_CRITICAL(Render_OpenGL, "Unimplemented swizzle source={}", static_cast<u32>(source));
-    UNREACHABLE();
-    return {};
+    LOG_ERROR(Render_OpenGL, "Unimplemented swizzle source={}", static_cast<u32>(source));
+    return GL_ZERO;
 }
 
 inline GLenum ComparisonOp(Maxwell::ComparisonOp comparison) {
@@ -344,33 +343,39 @@ inline GLenum ComparisonOp(Maxwell::ComparisonOp comparison) {
     case Maxwell::ComparisonOp::AlwaysOld:
         return GL_ALWAYS;
     }
-    LOG_CRITICAL(Render_OpenGL, "Unimplemented comparison op={}", static_cast<u32>(comparison));
-    UNREACHABLE();
-    return {};
+    LOG_ERROR(Render_OpenGL, "Unimplemented comparison op={}", static_cast<u32>(comparison));
+    return GL_ALWAYS;
 }
 
 inline GLenum StencilOp(Maxwell::StencilOp stencil) {
     switch (stencil) {
     case Maxwell::StencilOp::Keep:
+    case Maxwell::StencilOp::KeepOGL:
         return GL_KEEP;
     case Maxwell::StencilOp::Zero:
+    case Maxwell::StencilOp::ZeroOGL:
         return GL_ZERO;
     case Maxwell::StencilOp::Replace:
+    case Maxwell::StencilOp::ReplaceOGL:
         return GL_REPLACE;
     case Maxwell::StencilOp::Incr:
+    case Maxwell::StencilOp::IncrOGL:
         return GL_INCR;
     case Maxwell::StencilOp::Decr:
+    case Maxwell::StencilOp::DecrOGL:
         return GL_DECR;
     case Maxwell::StencilOp::Invert:
+    case Maxwell::StencilOp::InvertOGL:
         return GL_INVERT;
     case Maxwell::StencilOp::IncrWrap:
+    case Maxwell::StencilOp::IncrWrapOGL:
         return GL_INCR_WRAP;
     case Maxwell::StencilOp::DecrWrap:
+    case Maxwell::StencilOp::DecrWrapOGL:
         return GL_DECR_WRAP;
     }
-    LOG_CRITICAL(Render_OpenGL, "Unimplemented stencil op={}", static_cast<u32>(stencil));
-    UNREACHABLE();
-    return {};
+    LOG_ERROR(Render_OpenGL, "Unimplemented stencil op={}", static_cast<u32>(stencil));
+    return GL_KEEP;
 }
 
 inline GLenum FrontFace(Maxwell::Cull::FrontFace front_face) {
@@ -380,9 +385,8 @@ inline GLenum FrontFace(Maxwell::Cull::FrontFace front_face) {
     case Maxwell::Cull::FrontFace::CounterClockWise:
         return GL_CCW;
     }
-    LOG_CRITICAL(Render_OpenGL, "Unimplemented front face cull={}", static_cast<u32>(front_face));
-    UNREACHABLE();
-    return {};
+    LOG_ERROR(Render_OpenGL, "Unimplemented front face cull={}", static_cast<u32>(front_face));
+    return GL_CCW;
 }
 
 inline GLenum CullFace(Maxwell::Cull::CullFace cull_face) {
@@ -394,9 +398,8 @@ inline GLenum CullFace(Maxwell::Cull::CullFace cull_face) {
     case Maxwell::Cull::CullFace::FrontAndBack:
         return GL_FRONT_AND_BACK;
     }
-    LOG_CRITICAL(Render_OpenGL, "Unimplemented cull face={}", static_cast<u32>(cull_face));
-    UNREACHABLE();
-    return {};
+    LOG_ERROR(Render_OpenGL, "Unimplemented cull face={}", static_cast<u32>(cull_face));
+    return GL_BACK;
 }
 
 inline GLenum LogicOp(Maxwell::LogicOperation operation) {
@@ -434,9 +437,8 @@ inline GLenum LogicOp(Maxwell::LogicOperation operation) {
     case Maxwell::LogicOperation::Set:
         return GL_SET;
     }
-    LOG_CRITICAL(Render_OpenGL, "Unimplemented logic operation={}", static_cast<u32>(operation));
-    UNREACHABLE();
-    return {};
+    LOG_ERROR(Render_OpenGL, "Unimplemented logic operation={}", static_cast<u32>(operation));
+    return GL_COPY;
 }
 
 } // namespace MaxwellToGL

src/video_core/surface.cpp

@@ -306,6 +306,8 @@ PixelFormat PixelFormatFromTextureFormat(Tegra::Texture::TextureFormat format,
         return is_srgb ? PixelFormat::ASTC_2D_8X8_SRGB : PixelFormat::ASTC_2D_8X8;
     case Tegra::Texture::TextureFormat::ASTC_2D_8X5:
         return is_srgb ? PixelFormat::ASTC_2D_8X5_SRGB : PixelFormat::ASTC_2D_8X5;
+    case Tegra::Texture::TextureFormat::ASTC_2D_10X8:
+        return is_srgb ? PixelFormat::ASTC_2D_10X8_SRGB : PixelFormat::ASTC_2D_10X8;
     case Tegra::Texture::TextureFormat::R16_G16:
         switch (component_type) {
         case Tegra::Texture::ComponentType::FLOAT:
@@ -453,6 +455,8 @@ bool IsPixelFormatASTC(PixelFormat format) {
     case PixelFormat::ASTC_2D_5X5_SRGB:
     case PixelFormat::ASTC_2D_8X8_SRGB:
     case PixelFormat::ASTC_2D_8X5_SRGB:
+    case PixelFormat::ASTC_2D_10X8:
+    case PixelFormat::ASTC_2D_10X8_SRGB:
         return true;
     default:
         return false;

src/video_core/surface.h

@@ -74,19 +74,21 @@ enum class PixelFormat {
     ASTC_2D_5X4_SRGB = 56,
     ASTC_2D_5X5 = 57,
     ASTC_2D_5X5_SRGB = 58,
+    ASTC_2D_10X8 = 59,
+    ASTC_2D_10X8_SRGB = 60,
 
     MaxColorFormat,
 
     // Depth formats
-    Z32F = 59,
-    Z16 = 60,
+    Z32F = 61,
+    Z16 = 62,
 
     MaxDepthFormat,
 
     // DepthStencil formats
-    Z24S8 = 61,
-    S8Z24 = 62,
-    Z32FS8 = 63,
+    Z24S8 = 63,
+    S8Z24 = 64,
+    Z32FS8 = 65,
 
     MaxDepthStencilFormat,
 
@@ -193,6 +195,8 @@ static constexpr u32 GetCompressionFactor(PixelFormat format) {
         4, // ASTC_2D_5X4_SRGB
         4, // ASTC_2D_5X5
         4, // ASTC_2D_5X5_SRGB
+        4, // ASTC_2D_10X8
+        4, // ASTC_2D_10X8_SRGB
         1, // Z32F
         1, // Z16
         1, // Z24S8
@@ -208,70 +212,72 @@ static constexpr u32 GetDefaultBlockWidth(PixelFormat format) {
     if (format == PixelFormat::Invalid)
         return 0;
     constexpr std::array<u32, MaxPixelFormat> block_width_table = {{
-        1, // ABGR8U
-        1, // ABGR8S
-        1, // ABGR8UI
-        1, // B5G6R5U
-        1, // A2B10G10R10U
-        1, // A1B5G5R5U
-        1, // R8U
-        1, // R8UI
-        1, // RGBA16F
-        1, // RGBA16U
-        1, // RGBA16UI
-        1, // R11FG11FB10F
-        1, // RGBA32UI
-        4, // DXT1
-        4, // DXT23
-        4, // DXT45
-        4, // DXN1
-        4, // DXN2UNORM
-        4, // DXN2SNORM
-        4, // BC7U
-        4, // BC6H_UF16
-        4, // BC6H_SF16
-        4, // ASTC_2D_4X4
-        1, // G8R8U
-        1, // G8R8S
-        1, // BGRA8
-        1, // RGBA32F
-        1, // RG32F
-        1, // R32F
-        1, // R16F
-        1, // R16U
-        1, // R16S
-        1, // R16UI
-        1, // R16I
-        1, // RG16
-        1, // RG16F
-        1, // RG16UI
-        1, // RG16I
-        1, // RG16S
-        1, // RGB32F
-        1, // RGBA8_SRGB
-        1, // RG8U
-        1, // RG8S
-        1, // RG32UI
-        1, // R32UI
-        8, // ASTC_2D_8X8
-        8, // ASTC_2D_8X5
-        5, // ASTC_2D_5X4
-        1, // BGRA8_SRGB
-        4, // DXT1_SRGB
-        4, // DXT23_SRGB
-        4, // DXT45_SRGB
-        4, // BC7U_SRGB
-        4, // ASTC_2D_4X4_SRGB
-        8, // ASTC_2D_8X8_SRGB
-        8, // ASTC_2D_8X5_SRGB
-        5, // ASTC_2D_5X4_SRGB
-        5, // ASTC_2D_5X5
-        5, // ASTC_2D_5X5_SRGB
-        1, // Z32F
-        1, // Z16
-        1, // Z24S8
-        1, // S8Z24
-        1, // Z32FS8
+        1,  // ABGR8U
+        1,  // ABGR8S
+        1,  // ABGR8UI
+        1,  // B5G6R5U
+        1,  // A2B10G10R10U
+        1,  // A1B5G5R5U
+        1,  // R8U
+        1,  // R8UI
+        1,  // RGBA16F
+        1,  // RGBA16U
+        1,  // RGBA16UI
+        1,  // R11FG11FB10F
+        1,  // RGBA32UI
+        4,  // DXT1
+        4,  // DXT23
+        4,  // DXT45
+        4,  // DXN1
+        4,  // DXN2UNORM
+        4,  // DXN2SNORM
+        4,  // BC7U
+        4,  // BC6H_UF16
+        4,  // BC6H_SF16
+        4,  // ASTC_2D_4X4
+        1,  // G8R8U
+        1,  // G8R8S
+        1,  // BGRA8
+        1,  // RGBA32F
+        1,  // RG32F
+        1,  // R32F
+        1,  // R16F
+        1,  // R16U
+        1,  // R16S
+        1,  // R16UI
+        1,  // R16I
+        1,  // RG16
+        1,  // RG16F
+        1,  // RG16UI
+        1,  // RG16I
+        1,  // RG16S
+        1,  // RGB32F
+        1,  // RGBA8_SRGB
+        1,  // RG8U
+        1,  // RG8S
+        1,  // RG32UI
+        1,  // R32UI
+        8,  // ASTC_2D_8X8
+        8,  // ASTC_2D_8X5
+        5,  // ASTC_2D_5X4
+        1,  // BGRA8_SRGB
+        4,  // DXT1_SRGB
+        4,  // DXT23_SRGB
+        4,  // DXT45_SRGB
+        4,  // BC7U_SRGB
+        4,  // ASTC_2D_4X4_SRGB
+        8,  // ASTC_2D_8X8_SRGB
+        8,  // ASTC_2D_8X5_SRGB
+        5,  // ASTC_2D_5X4_SRGB
+        5,  // ASTC_2D_5X5
+        5,  // ASTC_2D_5X5_SRGB
+        10, // ASTC_2D_10X8
+        10, // ASTC_2D_10X8_SRGB
+        1,  // Z32F
+        1,  // Z16
+        1,  // Z24S8
+        1,  // S8Z24
+        1,  // Z32FS8
     }};
     ASSERT(static_cast<std::size_t>(format) < block_width_table.size());
     return block_width_table[static_cast<std::size_t>(format)];
@@ -341,6 +347,8 @@ static constexpr u32 GetDefaultBlockHeight(PixelFormat format) {
         4, // ASTC_2D_5X4_SRGB
         5, // ASTC_2D_5X5
         5, // ASTC_2D_5X5_SRGB
+        8, // ASTC_2D_10X8
+        8, // ASTC_2D_10X8_SRGB
         1, // Z32F
         1, // Z16
         1, // Z24S8
@@ -416,6 +424,8 @@ static constexpr u32 GetFormatBpp(PixelFormat format) {
         128, // ASTC_2D_5X4_SRGB
         128, // ASTC_2D_5X5
         128, // ASTC_2D_5X5_SRGB
+        128, // ASTC_2D_10X8
+        128, // ASTC_2D_10X8_SRGB
         32,  // Z32F
         16,  // Z16
         32,  // Z24S8

src/video_core/textures/decoders.cpp

@@ -37,25 +37,28 @@ struct alignas(64) SwizzleTable {
     std::array<std::array<u16, M>, N> values{};
 };
 
-constexpr auto legacy_swizzle_table = SwizzleTable<8, 64, 1>();
-constexpr auto fast_swizzle_table = SwizzleTable<8, 4, 16>();
+constexpr u32 gob_size_x = 64;
+constexpr u32 gob_size_y = 8;
+constexpr u32 gob_size_z = 1;
+constexpr u32 gob_size = gob_size_x * gob_size_y * gob_size_z;
+constexpr u32 fast_swizzle_align = 16;
+
+constexpr auto legacy_swizzle_table = SwizzleTable<gob_size_y, gob_size_x, gob_size_z>();
+constexpr auto fast_swizzle_table = SwizzleTable<gob_size_y, 4, fast_swizzle_align>();
 
 /**
  * This function manages ALL the GOBs(Group of Bytes) Inside a single block.
  * Instead of going gob by gob, we map the coordinates inside a block and manage from
  * those. Block_Width is assumed to be 1.
  */
-void PreciseProcessBlock(u8* swizzled_data, u8* unswizzled_data, const bool unswizzle,
+void PreciseProcessBlock(u8* const swizzled_data, u8* const unswizzled_data, const bool unswizzle,
                          const u32 x_start, const u32 y_start, const u32 z_start, const u32 x_end,
                          const u32 y_end, const u32 z_end, const u32 tile_offset,
                          const u32 xy_block_size, const u32 layer_z, const u32 stride_x,
                          const u32 bytes_per_pixel, const u32 out_bytes_per_pixel) {
     std::array<u8*, 2> data_ptrs;
     u32 z_address = tile_offset;
-    const u32 gob_size_x = 64;
-    const u32 gob_size_y = 8;
-    const u32 gob_size_z = 1;
-    const u32 gob_size = gob_size_x * gob_size_y * gob_size_z;
+
     for (u32 z = z_start; z < z_end; z++) {
         u32 y_address = z_address;
         u32 pixel_base = layer_z * z + y_start * stride_x;
@@ -81,7 +84,7 @@ void PreciseProcessBlock(u8* swizzled_data, u8* unswizzled_data, const bool unsw
  * Instead of going gob by gob, we map the coordinates inside a block and manage from
  * those. Block_Width is assumed to be 1.
  */
-void FastProcessBlock(u8* swizzled_data, u8* unswizzled_data, const bool unswizzle,
+void FastProcessBlock(u8* const swizzled_data, u8* const unswizzled_data, const bool unswizzle,
                       const u32 x_start, const u32 y_start, const u32 z_start, const u32 x_end,
                       const u32 y_end, const u32 z_end, const u32 tile_offset,
                       const u32 xy_block_size, const u32 layer_z, const u32 stride_x,
@@ -90,23 +93,19 @@ void FastProcessBlock(u8* swizzled_data, u8* unswizzled_data, const bool unswizz
     u32 z_address = tile_offset;
     const u32 x_startb = x_start * bytes_per_pixel;
     const u32 x_endb = x_end * bytes_per_pixel;
-    const u32 copy_size = 16;
-    const u32 gob_size_x = 64;
-    const u32 gob_size_y = 8;
-    const u32 gob_size_z = 1;
-    const u32 gob_size = gob_size_x * gob_size_y * gob_size_z;
+
     for (u32 z = z_start; z < z_end; z++) {
         u32 y_address = z_address;
         u32 pixel_base = layer_z * z + y_start * stride_x;
         for (u32 y = y_start; y < y_end; y++) {
             const auto& table = fast_swizzle_table[y % gob_size_y];
-            for (u32 xb = x_startb; xb < x_endb; xb += copy_size) {
-                const u32 swizzle_offset{y_address + table[(xb / copy_size) % 4]};
+            for (u32 xb = x_startb; xb < x_endb; xb += fast_swizzle_align) {
+                const u32 swizzle_offset{y_address + table[(xb / fast_swizzle_align) % 4]};
                 const u32 out_x = xb * out_bytes_per_pixel / bytes_per_pixel;
                 const u32 pixel_index{out_x + pixel_base};
                 data_ptrs[unswizzle] = swizzled_data + swizzle_offset;
                 data_ptrs[!unswizzle] = unswizzled_data + pixel_index;
-                std::memcpy(data_ptrs[0], data_ptrs[1], copy_size);
+                std::memcpy(data_ptrs[0], data_ptrs[1], fast_swizzle_align);
             }
             pixel_base += stride_x;
             if ((y + 1) % gob_size_y == 0)
@@ -126,23 +125,21 @@ void FastProcessBlock(u8* swizzled_data, u8* unswizzled_data, const bool unswizz
  *  https://envytools.readthedocs.io/en/latest/hw/memory/g80-surface.html#blocklinear-surfaces
  */
 template <bool fast>
-void SwizzledData(u8* swizzled_data, u8* unswizzled_data, const bool unswizzle, const u32 width,
-                  const u32 height, const u32 depth, const u32 bytes_per_pixel,
+void SwizzledData(u8* const swizzled_data, u8* const unswizzled_data, const bool unswizzle,
+                  const u32 width, const u32 height, const u32 depth, const u32 bytes_per_pixel,
                   const u32 out_bytes_per_pixel, const u32 block_height, const u32 block_depth) {
     auto div_ceil = [](const u32 x, const u32 y) { return ((x + y - 1) / y); };
     const u32 stride_x = width * out_bytes_per_pixel;
     const u32 layer_z = height * stride_x;
-    const u32 gob_x_bytes = 64;
-    const u32 gob_elements_x = gob_x_bytes / bytes_per_pixel;
-    const u32 gob_elements_y = 8;
-    const u32 gob_elements_z = 1;
+    const u32 gob_elements_x = gob_size_x / bytes_per_pixel;
+    constexpr u32 gob_elements_y = gob_size_y;
+    constexpr u32 gob_elements_z = gob_size_z;
     const u32 block_x_elements = gob_elements_x;
     const u32 block_y_elements = gob_elements_y * block_height;
     const u32 block_z_elements = gob_elements_z * block_depth;
     const u32 blocks_on_x = div_ceil(width, block_x_elements);
     const u32 blocks_on_y = div_ceil(height, block_y_elements);
     const u32 blocks_on_z = div_ceil(depth, block_z_elements);
-    const u32 gob_size = gob_x_bytes * gob_elements_y * gob_elements_z;
     const u32 xy_block_size = gob_size * block_height;
     const u32 block_size = xy_block_size * block_depth;
     u32 tile_offset = 0;
@@ -171,9 +168,9 @@ void SwizzledData(u8* swizzled_data, u8* unswizzled_data, const bool unswizzle,
 }
 
 void CopySwizzledData(u32 width, u32 height, u32 depth, u32 bytes_per_pixel,
-                      u32 out_bytes_per_pixel, u8* swizzled_data, u8* unswizzled_data,
+                      u32 out_bytes_per_pixel, u8* const swizzled_data, u8* const unswizzled_data,
                       bool unswizzle, u32 block_height, u32 block_depth) {
-    if (bytes_per_pixel % 3 != 0 && (width * bytes_per_pixel) % 16 == 0) {
+    if (bytes_per_pixel % 3 != 0 && (width * bytes_per_pixel) % fast_swizzle_align == 0) {
         SwizzledData<true>(swizzled_data, unswizzled_data, unswizzle, width, height, depth,
                            bytes_per_pixel, out_bytes_per_pixel, block_height, block_depth);
     } else {
@@ -202,6 +199,8 @@ u32 BytesPerPixel(TextureFormat format) {
     case TextureFormat::ASTC_2D_5X4:
     case TextureFormat::ASTC_2D_8X8:
     case TextureFormat::ASTC_2D_8X5:
+    case TextureFormat::ASTC_2D_10X8:
+    case TextureFormat::ASTC_2D_5X5:
     case TextureFormat::A8R8G8B8:
     case TextureFormat::A2B10G10R10:
     case TextureFormat::BF10GF11RF11:
@@ -227,29 +226,38 @@ u32 BytesPerPixel(TextureFormat format) {
     }
 }
 
+void UnswizzleTexture(u8* const unswizzled_data, VAddr address, u32 tile_size_x, u32 tile_size_y,
+                      u32 bytes_per_pixel, u32 width, u32 height, u32 depth, u32 block_height,
+                      u32 block_depth) {
+    CopySwizzledData((width + tile_size_x - 1) / tile_size_x,
+                     (height + tile_size_y - 1) / tile_size_y, depth, bytes_per_pixel,
+                     bytes_per_pixel, Memory::GetPointer(address), unswizzled_data, true,
+                     block_height, block_depth);
+}
+
 std::vector<u8> UnswizzleTexture(VAddr address, u32 tile_size_x, u32 tile_size_y,
                                  u32 bytes_per_pixel, u32 width, u32 height, u32 depth,
                                  u32 block_height, u32 block_depth) {
     std::vector<u8> unswizzled_data(width * height * depth * bytes_per_pixel);
-    CopySwizzledData((width + tile_size_x - 1) / tile_size_x,
-                     (height + tile_size_y - 1) / tile_size_y, depth, bytes_per_pixel,
-                     bytes_per_pixel, Memory::GetPointer(address), unswizzled_data.data(), true,
-                     block_height, block_depth);
+    UnswizzleTexture(unswizzled_data.data(), address, tile_size_x, tile_size_y, bytes_per_pixel,
+                     width, height, depth, block_height, block_depth);
     return unswizzled_data;
 }
 
 void SwizzleSubrect(u32 subrect_width, u32 subrect_height, u32 source_pitch, u32 swizzled_width,
                     u32 bytes_per_pixel, VAddr swizzled_data, VAddr unswizzled_data,
                     u32 block_height) {
-    const u32 image_width_in_gobs{(swizzled_width * bytes_per_pixel + 63) / 64};
+    const u32 image_width_in_gobs{(swizzled_width * bytes_per_pixel + (gob_size_x - 1)) /
+                                  gob_size_x};
     for (u32 line = 0; line < subrect_height; ++line) {
         const u32 gob_address_y =
-            (line / (8 * block_height)) * 512 * block_height * image_width_in_gobs +
-            (line % (8 * block_height) / 8) * 512;
-        const auto& table = legacy_swizzle_table[line % 8];
+            (line / (gob_size_y * block_height)) * gob_size * block_height * image_width_in_gobs +
+            ((line % (gob_size_y * block_height)) / gob_size_y) * gob_size;
+        const auto& table = legacy_swizzle_table[line % gob_size_y];
         for (u32 x = 0; x < subrect_width; ++x) {
-            const u32 gob_address = gob_address_y + (x * bytes_per_pixel / 64) * 512 * block_height;
-            const u32 swizzled_offset = gob_address + table[(x * bytes_per_pixel) % 64];
+            const u32 gob_address =
+                gob_address_y + (x * bytes_per_pixel / gob_size_x) * gob_size * block_height;
+            const u32 swizzled_offset = gob_address + table[(x * bytes_per_pixel) % gob_size_x];
             const VAddr source_line = unswizzled_data + line * source_pitch + x * bytes_per_pixel;
             const VAddr dest_addr = swizzled_data + swizzled_offset;
 
@@ -263,13 +271,13 @@ void UnswizzleSubrect(u32 subrect_width, u32 subrect_height, u32 dest_pitch, u32
                       u32 block_height, u32 offset_x, u32 offset_y) {
     for (u32 line = 0; line < subrect_height; ++line) {
         const u32 y2 = line + offset_y;
-        const u32 gob_address_y =
-            (y2 / (8 * block_height)) * 512 * block_height + (y2 % (8 * block_height) / 8) * 512;
-        const auto& table = legacy_swizzle_table[y2 % 8];
+        const u32 gob_address_y = (y2 / (gob_size_y * block_height)) * gob_size * block_height +
+                                  ((y2 % (gob_size_y * block_height)) / gob_size_y) * gob_size;
+        const auto& table = legacy_swizzle_table[y2 % gob_size_y];
         for (u32 x = 0; x < subrect_width; ++x) {
             const u32 x2 = (x + offset_x) * bytes_per_pixel;
-            const u32 gob_address = gob_address_y + (x2 / 64) * 512 * block_height;
-            const u32 swizzled_offset = gob_address + table[x2 % 64];
+            const u32 gob_address = gob_address_y + (x2 / gob_size_x) * gob_size * block_height;
+            const u32 swizzled_offset = gob_address + table[x2 % gob_size_x];
             const VAddr dest_line = unswizzled_data + line * dest_pitch + x * bytes_per_pixel;
             const VAddr source_addr = swizzled_data + swizzled_offset;
 
@@ -294,6 +302,8 @@ std::vector<u8> DecodeTexture(const std::vector<u8>& texture_data, TextureFormat
     case TextureFormat::BC6H_SF16:
     case TextureFormat::ASTC_2D_4X4:
     case TextureFormat::ASTC_2D_8X8:
+    case TextureFormat::ASTC_2D_5X5:
+    case TextureFormat::ASTC_2D_10X8:
     case TextureFormat::A8R8G8B8:
     case TextureFormat::A2B10G10R10:
     case TextureFormat::A1B5G5R5:
@@ -321,12 +331,9 @@ std::vector<u8> DecodeTexture(const std::vector<u8>& texture_data, TextureFormat
 std::size_t CalculateSize(bool tiled, u32 bytes_per_pixel, u32 width, u32 height, u32 depth,
                           u32 block_height, u32 block_depth) {
     if (tiled) {
-        const u32 gobs_in_x = 64;
-        const u32 gobs_in_y = 8;
-        const u32 gobs_in_z = 1;
-        const u32 aligned_width = Common::AlignUp(width * bytes_per_pixel, gobs_in_x);
-        const u32 aligned_height = Common::AlignUp(height, gobs_in_y * block_height);
-        const u32 aligned_depth = Common::AlignUp(depth, gobs_in_z * block_depth);
+        const u32 aligned_width = Common::AlignUp(width * bytes_per_pixel, gob_size_x);
+        const u32 aligned_height = Common::AlignUp(height, gob_size_y * block_height);
+        const u32 aligned_depth = Common::AlignUp(depth, gob_size_z * block_depth);
         return aligned_width * aligned_height * aligned_depth;
     } else {
         return width * height * depth * bytes_per_pixel;