Fix vulkan viewport issue

KPageTable: Perform ranged invalidation when unmapping code memory

.ci/scripts/transifex/docker.sh

@@ -0,0 +1,25 @@
+#!/bin/bash -e
+
+# Setup RC file for tx
+cat << EOF > ~/.transifexrc
+[https://www.transifex.com]
+hostname = https://www.transifex.com
+username = api
+password = $TRANSIFEX_API_TOKEN
+EOF
+
+
+set -x
+
+echo -e "\e[1m\e[33mBuild tools information:\e[0m"
+cmake --version
+gcc -v
+tx --version
+
+mkdir build && cd build
+cmake .. -DENABLE_QT_TRANSLATION=ON -DGENERATE_QT_TRANSLATION=ON -DCMAKE_BUILD_TYPE=Release -DENABLE_SDL2=OFF
+make translation
+cd ..
+
+cd dist/languages
+tx push -s

.ci/templates/merge-private.yml

@@ -1,6 +1,8 @@
 jobs:
 - job: merge
   displayName: 'pull requests'
+  pool:
+    vmImage: 'ubuntu-latest'
   steps:
   - checkout: self
     submodules: recursive
@@ -24,6 +26,8 @@ jobs:
 - job: upload_source
   displayName: 'upload'
   dependsOn: merge
+  pool:
+    vmImage: 'ubuntu-latest'
   steps:
   - template: ./sync-source.yml
     parameters:

.ci/templates/merge.yml

@@ -1,6 +1,8 @@
 jobs:
 - job: merge
   displayName: 'pull requests'
+  pool:
+    vmImage: 'ubuntu-latest'
   steps:
   - checkout: self
     submodules: recursive
@@ -23,6 +25,8 @@ jobs:
 - job: upload_source
   displayName: 'upload'
   dependsOn: merge
+  pool:
+    vmImage: 'ubuntu-latest'
   steps:
   - template: ./sync-source.yml
     parameters:

.ci/yuzu-mainline-step2.yml

@@ -66,5 +66,7 @@ stages:
   jobs:
   - job: github
     displayName: 'github'
+    pool:
+      vmImage: ubuntu-latest
     steps:
     - template: ./templates/release-github.yml

.ci/yuzu-patreon-step2.yml

@@ -29,5 +29,7 @@ stages:
   jobs:
     - job: release
       displayName: 'source'
+      pool:
+        vmImage: 'ubuntu-latest'
       steps:
         - template: ./templates/release-private-tag.yml

.github/workflows/ci.yml

@@ -0,0 +1,23 @@
+name: yuzu-ci
+
+on:
+  push:
+    branches: [ master ]
+    tags: [ "*" ]
+  pull_request:
+    branches: [ master ]
+
+jobs:
+  transifex:
+    runs-on: ubuntu-latest
+    container: yuzuemu/build-environments:linux-transifex
+    if: ${{ github.repository == 'yuzu-emu/yuzu' && !github.head_ref }}
+    steps:
+      - uses: actions/checkout@v2
+        with:
+          submodules: recursive
+          fetch-depth: 0   
+      - name: Update Translation
+        run: ./.ci/scripts/transifex/docker.sh
+        env:
+          TRANSIFEX_API_TOKEN: ${{ secrets.TRANSIFEX_API_TOKEN }}

.travis.yml

@@ -1,59 +0,0 @@
-language: cpp
-matrix:
-  include:
-    - os: linux
-      env: NAME="clang-format"
-      sudo: required
-      dist: trusty
-      services: docker
-      install: "./.travis/clang-format/deps.sh"
-      script: "./.travis/clang-format/build.sh"
-    - os: linux
-      env: NAME="linux build"
-      sudo: required
-      dist: trusty
-      services: docker
-      addons:
-        apt:
-          packages:
-            - p7zip-full
-      install: "./.travis/linux/deps.sh"
-      script: "./.travis/linux/build.sh"
-      after_success: "./.travis/linux/upload.sh"
-      cache: ccache
-    - os: osx
-      env: NAME="macos build"
-      sudo: false
-      osx_image: xcode10.2
-      install: "./.travis/macos/deps.sh"
-      script: "./.travis/macos/build.sh"
-      after_success: "./.travis/macos/upload.sh"
-      cache: ccache
-    - os: linux
-      env: NAME="MinGW build"
-      sudo: required
-      dist: trusty
-      services: docker
-      addons:
-        apt:
-          packages:
-            - p7zip-full
-      install: "./.travis/linux-mingw/deps.sh"
-      script: "./.travis/linux-mingw/build.sh"
-      after_success: "./.travis/linux-mingw/upload.sh"
-      cache: ccache
-
-deploy:
-  provider: releases
-  api_key:
-    secure: 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
-  file_glob: true
-  file: "artifacts/*"
-  skip_cleanup: true
-  on:
-    tags: true
-
-notifications:
-  webhooks:
-    urls:
-      - https://api.yuzu-emu.org/code/travis/notify

.travis/clang-format/build.sh

@@ -1,3 +0,0 @@
-#!/bin/bash -ex
-
-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 +0,0 @@
-#!/bin/sh -ex
-
-docker pull citraemu/build-environments:linux-clang-format

.travis/clang-format/docker.sh

@@ -1,5 +0,0 @@
-#!/bin/bash -ex
-
-# Run clang-format
-cd /yuzu
-./.travis/clang-format/script.sh

.travis/clang-format/script.sh

@@ -1,37 +0,0 @@
-#!/bin/bash -ex
-
-if grep -nrI '\s$' src *.yml *.txt *.md Doxyfile .gitignore .gitmodules .travis* dist/*.desktop \
-                 dist/*.svg dist/*.xml; then
-    echo Trailing whitespace found, aborting
-    exit 1
-fi
-
-# Default clang-format points to default 3.5 version one
-CLANG_FORMAT=clang-format-10.0
-$CLANG_FORMAT --version
-
-if [ "$TRAVIS_EVENT_TYPE" = "pull_request" ]; then
-    # Get list of every file modified in this pull request
-    files_to_lint="$(git diff --name-only --diff-filter=ACMRTUXB $TRAVIS_COMMIT_RANGE | grep '^src/[^.]*[.]\(cpp\|h\)$' || true)"
-else
-    # Check everything for branch pushes
-    files_to_lint="$(find src/ -name '*.cpp' -or -name '*.h')"
-fi
-
-# Turn off tracing for this because it's too verbose
-set +x
-
-for f in $files_to_lint; do
-    d=$(diff -u "$f" <($CLANG_FORMAT "$f") || true)
-    if ! [ -z "$d" ]; then
-        echo "!!! $f not compliant to coding style, here is the fix:"
-        echo "$d"
-        fail=1
-    fi
-done
-
-set -x
-
-if [ "$fail" = 1 ]; then
-    exit 1
-fi

.travis/common/post-upload.sh

@@ -1,25 +0,0 @@
-#!/bin/bash -ex
-
-# Copy documentation
-cp license.txt "$REV_NAME"
-cp README.md "$REV_NAME"
-
-tar $COMPRESSION_FLAGS "$ARCHIVE_NAME" "$REV_NAME"
-
-# Find out what release we are building
-if [ -z $TRAVIS_TAG ]; then
-    RELEASE_NAME=head
-else
-    RELEASE_NAME=$(echo $TRAVIS_TAG | cut -d- -f1)
-    if [ "$NAME" = "MinGW build" ]; then
-        RELEASE_NAME="${RELEASE_NAME}-mingw"
-    fi
-fi
-
-mv "$REV_NAME" $RELEASE_NAME
-
-7z a "$REV_NAME.7z" $RELEASE_NAME
-
-# move the compiled archive into the artifacts directory to be uploaded by travis releases
-mv "$ARCHIVE_NAME" artifacts/
-mv "$REV_NAME.7z" artifacts/

.travis/common/pre-upload.sh

@@ -1,6 +0,0 @@
-#!/bin/bash -ex
-
-GITDATE="`git show -s --date=short --format='%ad' | sed 's/-//g'`"
-GITREV="`git show -s --format='%h'`"
-
-mkdir -p artifacts

.travis/common/travis-ci.env

@@ -1,18 +0,0 @@
-# List of environment variables to be shared with Docker containers
-CI
-TRAVIS
-CONTINUOUS_INTEGRATION
-TRAVIS_BRANCH
-TRAVIS_BUILD_ID
-TRAVIS_BUILD_NUMBER
-TRAVIS_COMMIT
-TRAVIS_COMMIT_RANGE
-TRAVIS_EVENT_TYPE
-TRAVIS_JOB_ID
-TRAVIS_JOB_NUMBER
-TRAVIS_REPO_SLUG
-TRAVIS_TAG
-
-# yuzu specific flags
-ENABLE_COMPATIBILITY_REPORTING
-USE_DISCORD_PRESENCE

.travis/linux-mingw/build.sh

@@ -1,3 +0,0 @@
-#!/bin/bash -ex
-mkdir "$HOME/.ccache" || true
-docker run --env-file .travis/common/travis-ci.env -v $(pwd):/yuzu -v "$HOME/.ccache":/root/.ccache yuzuemu/build-environments:linux-mingw /bin/bash -ex /yuzu/.travis/linux-mingw/docker.sh

.travis/linux-mingw/deps.sh

@@ -1,3 +0,0 @@
-#!/bin/sh -ex
-
-docker pull yuzuemu/build-environments:linux-mingw

.travis/linux-mingw/docker.sh

@@ -1,42 +0,0 @@
-#!/bin/bash -ex
-
-cd /yuzu
-# override Travis CI unreasonable ccache size
-echo 'max_size = 3.0G' > "$HOME/.ccache/ccache.conf"
-
-mkdir build && cd build
-cmake .. -G Ninja -DCMAKE_TOOLCHAIN_FILE="$(pwd)/../CMakeModules/MinGWCross.cmake" -DUSE_CCACHE=ON -DENABLE_COMPATIBILITY_LIST_DOWNLOAD=ON -DCMAKE_BUILD_TYPE=Release
-ninja
-
-# Clean up the dirty hacks
-rm /bin/uname && mv /bin/uname1 /bin/uname
-rm /bin/cmd
-
-ccache -s
-
-echo "Tests skipped"
-#ctest -VV -C Release
-
-echo 'Prepare binaries...'
-cd ..
-mkdir package
-
-QT_PLATFORM_DLL_PATH='/usr/x86_64-w64-mingw32/lib/qt5/plugins/platforms/'
-find build/ -name "yuzu*.exe" -exec cp {} 'package' \;
-
-# copy Qt plugins
-mkdir package/platforms
-cp "${QT_PLATFORM_DLL_PATH}/qwindows.dll" package/platforms/
-cp -rv "${QT_PLATFORM_DLL_PATH}/../mediaservice/" package/
-cp -rv "${QT_PLATFORM_DLL_PATH}/../imageformats/" package/
-rm -f package/mediaservice/*d.dll
-
-for i in package/*.exe; do
-  # we need to process pdb here, however, cv2pdb
-  # does not work here, so we just simply strip all the debug symbols
-  x86_64-w64-mingw32-strip "${i}"
-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/"

.travis/linux-mingw/scan_dll.py

@@ -1,106 +0,0 @@
-import pefile
-import sys
-import re
-import os
-import queue
-import shutil
-
-# constant definitions
-KNOWN_SYS_DLLS = ['WINMM.DLL', 'MSVCRT.DLL', 'VERSION.DLL', 'MPR.DLL',
-                  'DWMAPI.DLL', 'UXTHEME.DLL', 'DNSAPI.DLL', 'IPHLPAPI.DLL']
-# below is for Ubuntu 18.04 with specified PPA enabled, if you are using
-# other distro or different repositories, change the following accordingly
-DLL_PATH = [
-    '/usr/x86_64-w64-mingw32/bin/',
-    '/usr/x86_64-w64-mingw32/lib/',
-    '/usr/lib/gcc/x86_64-w64-mingw32/7.3-posix/'
-]
-
-missing = []
-
-
-def parse_imports(file_name):
-    results = []
-    pe = pefile.PE(file_name, fast_load=True)
-    pe.parse_data_directories()
-
-    for entry in pe.DIRECTORY_ENTRY_IMPORT:
-        current = entry.dll.decode()
-        current_u = current.upper()  # b/c Windows is often case insensitive
-        # here we filter out system dlls
-        # dll w/ names like *32.dll are likely to be system dlls
-        if current_u.upper() not in KNOWN_SYS_DLLS and not re.match(string=current_u, pattern=r'.*32\.DLL'):
-            results.append(current)
-
-    return results
-
-
-def parse_imports_recursive(file_name, path_list=[]):
-    q = queue.Queue()  # create a FIFO queue
-    # file_name can be a string or a list for the convience
-    if isinstance(file_name, str):
-        q.put(file_name)
-    elif isinstance(file_name, list):
-        for i in file_name:
-            q.put(i)
-    full_list = []
-    while q.qsize():
-        current = q.get_nowait()
-        print('> %s' % current)
-        deps = parse_imports(current)
-        # if this dll does not have any import, ignore it
-        if not deps:
-            continue
-        for dep in deps:
-            # the dependency already included in the list, skip
-            if dep in full_list:
-                continue
-            # find the requested dll in the provided paths
-            full_path = find_dll(dep)
-            if not full_path:
-                missing.append(dep)
-                continue
-            full_list.append(dep)
-            q.put(full_path)
-            path_list.append(full_path)
-    return full_list
-
-
-def find_dll(name):
-    for path in DLL_PATH:
-        for root, _, files in os.walk(path):
-            for f in files:
-                if name.lower() == f.lower():
-                    return os.path.join(root, f)
-
-
-def deploy(name, dst, dry_run=False):
-    dlls_path = []
-    parse_imports_recursive(name, dlls_path)
-    for dll_entry in dlls_path:
-        if not dry_run:
-            shutil.copy(dll_entry, dst)
-        else:
-            print('[Dry-Run] Copy %s to %s' % (dll_entry, dst))
-    print('Deploy completed.')
-    return dlls_path
-
-
-def main():
-    if len(sys.argv) < 3:
-        print('Usage: %s [files to examine ...] [target deploy directory]')
-        return 1
-    to_deploy = sys.argv[1:-1]
-    tgt_dir = sys.argv[-1]
-    if not os.path.isdir(tgt_dir):
-        print('%s is not a directory.' % tgt_dir)
-        return 1
-    print('Scanning dependencies...')
-    deploy(to_deploy, tgt_dir)
-    if missing:
-        print('Following DLLs are not found: %s' % ('\n'.join(missing)))
-    return 0
-
-
-if __name__ == '__main__':
-    main()

.travis/linux-mingw/upload.sh

@@ -1,13 +0,0 @@
-#!/bin/bash -ex
-
-. .travis/common/pre-upload.sh
-
-REV_NAME="yuzu-windows-mingw-${GITDATE}-${GITREV}"
-ARCHIVE_NAME="${REV_NAME}.tar.gz"
-COMPRESSION_FLAGS="-czvf"
-
-mkdir "$REV_NAME"
-# get around the permission issues
-cp -r package/* "$REV_NAME"
-
-. .travis/common/post-upload.sh

.travis/linux/build.sh

@@ -1,4 +0,0 @@
-#!/bin/bash -ex
-
-mkdir -p "$HOME/.ccache"
-docker run -e ENABLE_COMPATIBILITY_REPORTING --env-file .travis/common/travis-ci.env -v $(pwd):/yuzu -v "$HOME/.ccache":/home/yuzu/.ccache yuzuemu/build-environments:linux-fresh /bin/bash /yuzu/.travis/linux/docker.sh

.travis/linux/deps.sh

@@ -1,3 +0,0 @@
-#!/bin/sh -ex
-
-docker pull yuzuemu/build-environments:linux-fresh

.travis/linux/docker.sh

@@ -1,11 +0,0 @@
-#!/bin/bash -ex
-
-cd /yuzu
-
-mkdir build && cd build
-cmake .. -G Ninja -DYUZU_USE_QT_WEB_ENGINE=ON -DCMAKE_BUILD_TYPE=Release -DCMAKE_C_COMPILER=/usr/lib/ccache/gcc -DCMAKE_CXX_COMPILER=/usr/lib/ccache/g++ -DYUZU_ENABLE_COMPATIBILITY_REPORTING=${ENABLE_COMPATIBILITY_REPORTING:-"OFF"} -DENABLE_COMPATIBILITY_LIST_DOWNLOAD=ON -DUSE_DISCORD_PRESENCE=ON
-ninja
-
-ccache -s
-
-ctest -VV -C Release

.travis/linux/upload.sh

@@ -1,14 +0,0 @@
-#!/bin/bash -ex
-
-. .travis/common/pre-upload.sh
-
-REV_NAME="yuzu-linux-${GITDATE}-${GITREV}"
-ARCHIVE_NAME="${REV_NAME}.tar.xz"
-COMPRESSION_FLAGS="-cJvf"
-
-mkdir "$REV_NAME"
-
-cp build/bin/yuzu-cmd "$REV_NAME"
-cp build/bin/yuzu "$REV_NAME"
-
-. .travis/common/post-upload.sh

.travis/macos/build.sh

@@ -1,17 +0,0 @@
-#!/bin/bash -ex
-
-set -o pipefail
-
-export MACOSX_DEPLOYMENT_TARGET=10.14
-export Qt5_DIR=$(brew --prefix)/opt/qt5
-export PATH="/usr/local/opt/ccache/libexec:$PATH"
-
-# TODO: Build using ninja instead of make
-mkdir build && cd build
-cmake --version
-cmake .. -DYUZU_USE_QT_WEB_ENGINE=ON -DCMAKE_BUILD_TYPE=Release -DENABLE_COMPATIBILITY_LIST_DOWNLOAD=ON -DYUZU_ENABLE_COMPATIBILITY_REPORTING=${ENABLE_COMPATIBILITY_REPORTING:-"OFF"} -DUSE_DISCORD_PRESENCE=ON
-make -j4
-
-ccache -s
-
-ctest -VV -C Release

.travis/macos/deps.sh

@@ -1,6 +0,0 @@
-#!/bin/sh -ex
-
-brew update
-brew install p7zip qt5 sdl2 ccache
-brew outdated cmake || brew upgrade cmake
-pip3 install macpack

.travis/macos/upload.sh

@@ -1,28 +0,0 @@
-#!/bin/bash -ex
-
-. .travis/common/pre-upload.sh
-
-REV_NAME="yuzu-osx-${GITDATE}-${GITREV}"
-ARCHIVE_NAME="${REV_NAME}.tar.gz"
-COMPRESSION_FLAGS="-czvf"
-
-mkdir "$REV_NAME"
-
-cp build/bin/yuzu-cmd "$REV_NAME"
-cp -r build/bin/yuzu.app "$REV_NAME"
-
-# move libs into folder for deployment
-macpack "${REV_NAME}/yuzu.app/Contents/MacOS/yuzu" -d "../Frameworks"
-# move qt frameworks into app bundle for deployment
-$(brew --prefix)/opt/qt5/bin/macdeployqt "${REV_NAME}/yuzu.app" -executable="${REV_NAME}/yuzu.app/Contents/MacOS/yuzu"
-
-# move libs into folder for deployment
-macpack "${REV_NAME}/yuzu-cmd" -d "libs"
-
-# Make the launching script executable
-chmod +x ${REV_NAME}/yuzu.app/Contents/MacOS/yuzu
-
-# Verify loader instructions
-find "$REV_NAME" -exec otool -L {} \;
-
-. .travis/common/post-upload.sh

src/common/fs/path_util.cpp

@@ -82,32 +82,35 @@ public:
 
 private:
     PathManagerImpl() {
+        fs::path yuzu_path;
+        fs::path yuzu_path_cache;
+        fs::path yuzu_path_config;
+
 #ifdef _WIN32
-        auto yuzu_path = GetExeDirectory() / PORTABLE_DIR;
+        yuzu_path = GetExeDirectory() / PORTABLE_DIR;
 
         if (!IsDir(yuzu_path)) {
             yuzu_path = GetAppDataRoamingDirectory() / YUZU_DIR;
         }
 
-        GenerateYuzuPath(YuzuPath::YuzuDir, yuzu_path);
-        GenerateYuzuPath(YuzuPath::CacheDir, yuzu_path / CACHE_DIR);
-        GenerateYuzuPath(YuzuPath::ConfigDir, yuzu_path / CONFIG_DIR);
+        yuzu_path_cache = yuzu_path / CACHE_DIR;
+        yuzu_path_config = yuzu_path / CONFIG_DIR;
 #else
-        auto yuzu_path = GetCurrentDir() / PORTABLE_DIR;
+        yuzu_path = GetCurrentDir() / PORTABLE_DIR;
 
         if (Exists(yuzu_path) && IsDir(yuzu_path)) {
-            GenerateYuzuPath(YuzuPath::YuzuDir, yuzu_path);
-            GenerateYuzuPath(YuzuPath::CacheDir, yuzu_path / CACHE_DIR);
-            GenerateYuzuPath(YuzuPath::ConfigDir, yuzu_path / CONFIG_DIR);
+            yuzu_path_cache = yuzu_path / CACHE_DIR;
+            yuzu_path_config = yuzu_path / CONFIG_DIR;
         } else {
             yuzu_path = GetDataDirectory("XDG_DATA_HOME") / YUZU_DIR;
-
-            GenerateYuzuPath(YuzuPath::YuzuDir, yuzu_path);
-            GenerateYuzuPath(YuzuPath::CacheDir, GetDataDirectory("XDG_CACHE_HOME") / YUZU_DIR);
-            GenerateYuzuPath(YuzuPath::ConfigDir, GetDataDirectory("XDG_CONFIG_HOME") / YUZU_DIR);
+            yuzu_path_cache = GetDataDirectory("XDG_CACHE_HOME") / YUZU_DIR;
+            yuzu_path_config = GetDataDirectory("XDG_CONFIG_HOME") / YUZU_DIR;
         }
 #endif
 
+        GenerateYuzuPath(YuzuPath::YuzuDir, yuzu_path);
+        GenerateYuzuPath(YuzuPath::CacheDir, yuzu_path_cache);
+        GenerateYuzuPath(YuzuPath::ConfigDir, yuzu_path_config);
         GenerateYuzuPath(YuzuPath::DumpDir, yuzu_path / DUMP_DIR);
         GenerateYuzuPath(YuzuPath::KeysDir, yuzu_path / KEYS_DIR);
         GenerateYuzuPath(YuzuPath::LoadDir, yuzu_path / LOAD_DIR);

src/common/string_util.cpp

@@ -180,20 +180,20 @@ std::wstring UTF8ToUTF16W(const std::string& input) {
 
 #endif
 
-std::string StringFromFixedZeroTerminatedBuffer(const char* buffer, std::size_t max_len) {
+std::string StringFromFixedZeroTerminatedBuffer(std::string_view buffer, std::size_t max_len) {
     std::size_t len = 0;
-    while (len < max_len && buffer[len] != '\0')
+    while (len < buffer.length() && len < max_len && buffer[len] != '\0') {
         ++len;
-
-    return std::string(buffer, len);
+    }
+    return std::string(buffer.begin(), buffer.begin() + len);
 }
 
 std::u16string UTF16StringFromFixedZeroTerminatedBuffer(std::u16string_view buffer,
                                                         std::size_t max_len) {
     std::size_t len = 0;
-    while (len < max_len && buffer[len] != '\0')
+    while (len < buffer.length() && len < max_len && buffer[len] != '\0') {
         ++len;
-
+    }
     return std::u16string(buffer.begin(), buffer.begin() + len);
 }
 

src/common/string_util.h

@@ -63,7 +63,7 @@ template <typename InIt>
  * Creates a std::string from a fixed-size NUL-terminated char buffer. If the buffer isn't
  * NUL-terminated then the string ends at max_len characters.
  */
-[[nodiscard]] std::string StringFromFixedZeroTerminatedBuffer(const char* buffer,
+[[nodiscard]] std::string StringFromFixedZeroTerminatedBuffer(std::string_view buffer,
                                                               std::size_t max_len);
 
 /**

src/core/core.cpp

@@ -139,27 +139,47 @@ struct System::Impl {
         : kernel{system}, fs_controller{system}, memory{system},
           cpu_manager{system}, reporter{system}, applet_manager{system}, time_manager{system} {}
 
-    ResultStatus Run() {
-        status = ResultStatus::Success;
+    SystemResultStatus Run() {
+        std::unique_lock<std::mutex> lk(suspend_guard);
+        status = SystemResultStatus::Success;
 
         kernel.Suspend(false);
         core_timing.SyncPause(false);
         cpu_manager.Pause(false);
+        is_paused = false;
 
         return status;
     }
 
-    ResultStatus Pause() {
-        status = ResultStatus::Success;
+    SystemResultStatus Pause() {
+        std::unique_lock<std::mutex> lk(suspend_guard);
+        status = SystemResultStatus::Success;
 
         core_timing.SyncPause(true);
         kernel.Suspend(true);
         cpu_manager.Pause(true);
+        is_paused = true;
 
         return status;
     }
 
-    ResultStatus Init(System& system, Frontend::EmuWindow& emu_window) {
+    std::unique_lock<std::mutex> StallCPU() {
+        std::unique_lock<std::mutex> lk(suspend_guard);
+        kernel.Suspend(true);
+        core_timing.SyncPause(true);
+        cpu_manager.Pause(true);
+        return lk;
+    }
+
+    void UnstallCPU() {
+        if (!is_paused) {
+            core_timing.SyncPause(false);
+            kernel.Suspend(false);
+            cpu_manager.Pause(false);
+        }
+    }
+
+    SystemResultStatus Init(System& system, Frontend::EmuWindow& emu_window) {
         LOG_DEBUG(Core, "initialized OK");
 
         device_memory = std::make_unique<Core::DeviceMemory>();
@@ -197,7 +217,7 @@ struct System::Impl {
 
         gpu_core = VideoCore::CreateGPU(emu_window, system);
         if (!gpu_core) {
-            return ResultStatus::ErrorVideoCore;
+            return SystemResultStatus::ErrorVideoCore;
         }
 
         service_manager = std::make_shared<Service::SM::ServiceManager>(kernel);
@@ -217,21 +237,22 @@ struct System::Impl {
 
         LOG_DEBUG(Core, "Initialized OK");
 
-        return ResultStatus::Success;
+        return SystemResultStatus::Success;
     }
 
-    ResultStatus Load(System& system, Frontend::EmuWindow& emu_window, const std::string& filepath,
-                      u64 program_id, std::size_t program_index) {
+    SystemResultStatus Load(System& system, Frontend::EmuWindow& emu_window,
+                            const std::string& filepath, u64 program_id,
+                            std::size_t program_index) {
         app_loader = Loader::GetLoader(system, GetGameFileFromPath(virtual_filesystem, filepath),
                                        program_id, program_index);
 
         if (!app_loader) {
             LOG_CRITICAL(Core, "Failed to obtain loader for {}!", filepath);
-            return ResultStatus::ErrorGetLoader;
+            return SystemResultStatus::ErrorGetLoader;
         }
 
-        ResultStatus init_result{Init(system, emu_window)};
-        if (init_result != ResultStatus::Success) {
+        SystemResultStatus init_result{Init(system, emu_window)};
+        if (init_result != SystemResultStatus::Success) {
             LOG_CRITICAL(Core, "Failed to initialize system (Error {})!",
                          static_cast<int>(init_result));
             Shutdown();
@@ -249,8 +270,8 @@ struct System::Impl {
             LOG_CRITICAL(Core, "Failed to load ROM (Error {})!", load_result);
             Shutdown();
 
-            return static_cast<ResultStatus>(static_cast<u32>(ResultStatus::ErrorLoader) +
-                                             static_cast<u32>(load_result));
+            return static_cast<SystemResultStatus>(
+                static_cast<u32>(SystemResultStatus::ErrorLoader) + static_cast<u32>(load_result));
         }
         AddGlueRegistrationForProcess(*app_loader, *main_process);
         kernel.MakeCurrentProcess(main_process.get());
@@ -282,7 +303,7 @@ struct System::Impl {
         GetAndResetPerfStats();
         perf_stats->BeginSystemFrame();
 
-        status = ResultStatus::Success;
+        status = SystemResultStatus::Success;
         return status;
     }
 
@@ -355,7 +376,7 @@ struct System::Impl {
         arp_manager.Register(launch.title_id, launch, std::move(nacp_data));
     }
 
-    void SetStatus(ResultStatus new_status, const char* details = nullptr) {
+    void SetStatus(SystemResultStatus new_status, const char* details = nullptr) {
         status = new_status;
         if (details) {
             status_details = details;
@@ -366,6 +387,9 @@ struct System::Impl {
         return perf_stats->GetAndResetStats(core_timing.GetGlobalTimeUs());
     }
 
+    std::mutex suspend_guard;
+    bool is_paused{};
+
     Timing::CoreTiming core_timing;
     Kernel::KernelCore kernel;
     /// RealVfsFilesystem instance
@@ -411,7 +435,7 @@ struct System::Impl {
     /// Network instance
     Network::NetworkInstance network_instance;
 
-    ResultStatus status = ResultStatus::Success;
+    SystemResultStatus status = SystemResultStatus::Success;
     std::string status_details = "";
 
     std::unique_ptr<Core::PerfStats> perf_stats;
@@ -428,22 +452,9 @@ struct System::Impl {
 };
 
 System::System() : impl{std::make_unique<Impl>(*this)} {}
+
 System::~System() = default;
 
-System& System::GetInstance() {
-    if (!s_instance) {
-        throw std::runtime_error("Using System instance before its initialization");
-    }
-    return *s_instance;
-}
-
-void System::InitializeGlobalInstance() {
-    if (s_instance) {
-        throw std::runtime_error("Reinitializing Global System instance.");
-    }
-    s_instance = std::unique_ptr<System>(new System);
-}
-
 CpuManager& System::GetCpuManager() {
     return impl->cpu_manager;
 }
@@ -452,16 +463,16 @@ const CpuManager& System::GetCpuManager() const {
     return impl->cpu_manager;
 }
 
-System::ResultStatus System::Run() {
+SystemResultStatus System::Run() {
     return impl->Run();
 }
 
-System::ResultStatus System::Pause() {
+SystemResultStatus System::Pause() {
     return impl->Pause();
 }
 
-System::ResultStatus System::SingleStep() {
-    return ResultStatus::Success;
+SystemResultStatus System::SingleStep() {
+    return SystemResultStatus::Success;
 }
 
 void System::InvalidateCpuInstructionCaches() {
@@ -476,8 +487,16 @@ void System::Shutdown() {
     impl->Shutdown();
 }
 
-System::ResultStatus System::Load(Frontend::EmuWindow& emu_window, const std::string& filepath,
-                                  u64 program_id, std::size_t program_index) {
+std::unique_lock<std::mutex> System::StallCPU() {
+    return impl->StallCPU();
+}
+
+void System::UnstallCPU() {
+    impl->UnstallCPU();
+}
+
+SystemResultStatus System::Load(Frontend::EmuWindow& emu_window, const std::string& filepath,
+                                u64 program_id, std::size_t program_index) {
     return impl->Load(*this, emu_window, filepath, program_id, program_index);
 }
 
@@ -637,7 +656,7 @@ Loader::ResultStatus System::GetGameName(std::string& out) const {
     return impl->GetGameName(out);
 }
 
-void System::SetStatus(ResultStatus new_status, const char* details) {
+void System::SetStatus(SystemResultStatus new_status, const char* details) {
     impl->SetStatus(new_status, details);
 }
 

src/core/core.h

@@ -7,6 +7,7 @@
 #include <cstddef>
 #include <functional>
 #include <memory>
+#include <mutex>
 #include <string>
 #include <vector>
 
@@ -104,55 +105,49 @@ struct PerfStatsResults;
 FileSys::VirtualFile GetGameFileFromPath(const FileSys::VirtualFilesystem& vfs,
                                          const std::string& path);
 
+/// Enumeration representing the return values of the System Initialize and Load process.
+enum class SystemResultStatus : u32 {
+    Success,             ///< Succeeded
+    ErrorNotInitialized, ///< Error trying to use core prior to initialization
+    ErrorGetLoader,      ///< Error finding the correct application loader
+    ErrorSystemFiles,    ///< Error in finding system files
+    ErrorSharedFont,     ///< Error in finding shared font
+    ErrorVideoCore,      ///< Error in the video core
+    ErrorUnknown,        ///< Any other error
+    ErrorLoader,         ///< The base for loader errors (too many to repeat)
+};
+
 class System {
 public:
     using CurrentBuildProcessID = std::array<u8, 0x20>;
 
+    explicit System();
+
+    ~System();
+
     System(const System&) = delete;
     System& operator=(const System&) = delete;
 
     System(System&&) = delete;
     System& operator=(System&&) = delete;
 
-    ~System();
-
-    /**
-     * Gets the instance of the System singleton class.
-     * @returns Reference to the instance of the System singleton class.
-     */
-    [[deprecated("Use of the global system instance is deprecated")]] static System& GetInstance();
-
-    static void InitializeGlobalInstance();
-
-    /// Enumeration representing the return values of the System Initialize and Load process.
-    enum class ResultStatus : u32 {
-        Success,             ///< Succeeded
-        ErrorNotInitialized, ///< Error trying to use core prior to initialization
-        ErrorGetLoader,      ///< Error finding the correct application loader
-        ErrorSystemFiles,    ///< Error in finding system files
-        ErrorSharedFont,     ///< Error in finding shared font
-        ErrorVideoCore,      ///< Error in the video core
-        ErrorUnknown,        ///< Any other error
-        ErrorLoader,         ///< The base for loader errors (too many to repeat)
-    };
-
     /**
      * Run the OS and Application
      * This function will start emulation and run the relevant devices
      */
-    [[nodiscard]] ResultStatus Run();
+    [[nodiscard]] SystemResultStatus Run();
 
     /**
      * Pause the OS and Application
      * This function will pause emulation and stop the relevant devices
      */
-    [[nodiscard]] ResultStatus Pause();
+    [[nodiscard]] SystemResultStatus Pause();
 
     /**
      * Step the CPU one instruction
      * @return Result status, indicating whether or not the operation succeeded.
      */
-    [[nodiscard]] ResultStatus SingleStep();
+    [[nodiscard]] SystemResultStatus SingleStep();
 
     /**
      * Invalidate the CPU instruction caches
@@ -166,16 +161,20 @@ public:
     /// Shutdown the emulated system.
     void Shutdown();
 
+    std::unique_lock<std::mutex> StallCPU();
+    void UnstallCPU();
+
     /**
      * Load an executable application.
      * @param emu_window Reference to the host-system window used for video output and keyboard
      *                   input.
      * @param filepath String path to the executable application to load on the host file system.
      * @param program_index Specifies the index within the container of the program to launch.
-     * @returns ResultStatus code, indicating if the operation succeeded.
+     * @returns SystemResultStatus code, indicating if the operation succeeded.
      */
-    [[nodiscard]] ResultStatus Load(Frontend::EmuWindow& emu_window, const std::string& filepath,
-                                    u64 program_id = 0, std::size_t program_index = 0);
+    [[nodiscard]] SystemResultStatus Load(Frontend::EmuWindow& emu_window,
+                                          const std::string& filepath, u64 program_id = 0,
+                                          std::size_t program_index = 0);
 
     /**
      * Indicates if the emulated system is powered on (all subsystems initialized and able to run an
@@ -301,7 +300,7 @@ public:
     /// Gets the name of the current game
     [[nodiscard]] Loader::ResultStatus GetGameName(std::string& out) const;
 
-    void SetStatus(ResultStatus new_status, const char* details);
+    void SetStatus(SystemResultStatus new_status, const char* details);
 
     [[nodiscard]] const std::string& GetStatusDetails() const;
 
@@ -403,12 +402,8 @@ public:
     void ApplySettings();
 
 private:
-    System();
-
     struct Impl;
     std::unique_ptr<Impl> impl;
-
-    inline static std::unique_ptr<System> s_instance{};
 };
 
 } // namespace Core

src/core/hle/ipc_helpers.h

@@ -4,17 +4,14 @@
 
 #pragma once
 
-#include <array>
 #include <cstring>
 #include <memory>
-#include <tuple>
 #include <type_traits>
 #include <utility>
 #include "common/assert.h"
 #include "common/common_types.h"
 #include "core/hle/ipc.h"
 #include "core/hle/kernel/hle_ipc.h"
-#include "core/hle/kernel/k_client_port.h"
 #include "core/hle/kernel/k_process.h"
 #include "core/hle/kernel/k_resource_limit.h"
 #include "core/hle/kernel/k_session.h"

src/core/hle/kernel/hle_ipc.cpp

@@ -15,6 +15,7 @@
 #include "common/logging/log.h"
 #include "core/hle/ipc_helpers.h"
 #include "core/hle/kernel/hle_ipc.h"
+#include "core/hle/kernel/k_auto_object.h"
 #include "core/hle/kernel/k_handle_table.h"
 #include "core/hle/kernel/k_process.h"
 #include "core/hle/kernel/k_readable_event.h"

src/core/hle/kernel/hle_ipc.h

@@ -17,7 +17,6 @@
 #include "common/concepts.h"
 #include "common/swap.h"
 #include "core/hle/ipc.h"
-#include "core/hle/kernel/k_auto_object.h"
 #include "core/hle/kernel/svc_common.h"
 
 union ResultCode;
@@ -38,6 +37,7 @@ namespace Kernel {
 
 class Domain;
 class HLERequestContext;
+class KAutoObject;
 class KernelCore;
 class KHandleTable;
 class KProcess;

src/core/hle/kernel/k_page_table.cpp

@@ -363,6 +363,8 @@ ResultCode KPageTable::UnmapProcessCodeMemory(VAddr dst_addr, VAddr src_addr, st
     block_manager->Update(src_addr, num_pages, KMemoryState::Normal,
                           KMemoryPermission::ReadAndWrite);
 
+    system.InvalidateCpuInstructionCacheRange(dst_addr, size);
+
     return ResultSuccess;
 }
 

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

@@ -5,7 +5,7 @@
 #pragma once
 
 #include <memory>
-#include "core/hle/kernel/hle_ipc.h"
+
 #include "core/hle/service/service.h"
 
 namespace Service {

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

@@ -5,7 +5,7 @@
 #pragma once
 
 #include <memory>
-#include "core/hle/kernel/hle_ipc.h"
+
 #include "core/hle/service/service.h"
 
 namespace Service {

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

@@ -24,6 +24,7 @@
 #include "core/hle/service/am/applets/applet_web_browser.h"
 #include "core/hle/service/filesystem/filesystem.h"
 #include "core/hle/service/ns/pl_u.h"
+#include "core/loader/loader.h"
 
 namespace Service::AM::Applets {
 
@@ -122,6 +123,15 @@ FileSys::VirtualFile GetOfflineRomFS(Core::System& system, u64 title_id,
         const auto nca = system.GetContentProvider().GetEntry(title_id, nca_type);
 
         if (nca == nullptr) {
+            if (nca_type == FileSys::ContentRecordType::HtmlDocument) {
+                LOG_WARNING(Service_AM, "Falling back to AppLoader to get the RomFS.");
+                FileSys::VirtualFile romfs;
+                system.GetAppLoader().ReadManualRomFS(romfs);
+                if (romfs != nullptr) {
+                    return romfs;
+                }
+            }
+
             LOG_ERROR(Service_AM,
                       "NCA of type={} with title_id={:016X} is not found in the ContentProvider!",
                       nca_type, title_id);

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

@@ -5,7 +5,6 @@
 #include "common/logging/log.h"
 #include "core/core.h"
 #include "core/hle/ipc_helpers.h"
-#include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/kernel/k_event.h"
 #include "core/hle/service/audio/audin_u.h"
 

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

@@ -13,7 +13,6 @@
 #include "common/swap.h"
 #include "core/core.h"
 #include "core/hle/ipc_helpers.h"
-#include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/kernel/k_event.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/service/audio/audout_u.h"

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

@@ -15,7 +15,6 @@
 #include "common/string_util.h"
 #include "core/core.h"
 #include "core/hle/ipc_helpers.h"
-#include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/kernel/k_event.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/service/audio/audren_u.h"

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

@@ -13,7 +13,6 @@
 #include "common/assert.h"
 #include "common/logging/log.h"
 #include "core/hle/ipc_helpers.h"
-#include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/service/audio/hwopus.h"
 
 namespace Service::Audio {

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

@@ -5,7 +5,6 @@
 #include "common/logging/log.h"
 #include "core/core.h"
 #include "core/hle/ipc_helpers.h"
-#include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/kernel/k_event.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/service/btdrv/btdrv.h"

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

@@ -7,7 +7,6 @@
 #include "common/logging/log.h"
 #include "core/core.h"
 #include "core/hle/ipc_helpers.h"
-#include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/kernel/k_event.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/service/btm/btm.h"

src/core/hle/service/caps/caps.h

@@ -4,7 +4,8 @@
 
 #pragma once
 
-#include "core/hle/service/service.h"
+#include "common/common_funcs.h"
+#include "common/common_types.h"
 
 namespace Core {
 class System;

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

@@ -5,7 +5,6 @@
 #include <memory>
 
 #include "core/hle/ipc_helpers.h"
-#include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/service/fgm/fgm.h"
 #include "core/hle/service/service.h"
 #include "core/hle/service/sm/sm.h"

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

@@ -3,7 +3,6 @@
 // Refer to the license.txt file included.
 
 #include "core/hle/service/filesystem/fsp_ldr.h"
-#include "core/hle/service/service.h"
 
 namespace Service::FileSystem {
 

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

@@ -3,7 +3,6 @@
 // Refer to the license.txt file included.
 
 #include "core/hle/service/filesystem/fsp_pr.h"
-#include "core/hle/service/service.h"
 
 namespace Service::FileSystem {
 

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

@@ -8,13 +8,11 @@
 #include "core/core.h"
 #include "core/file_sys/control_metadata.h"
 #include "core/hle/ipc_helpers.h"
-#include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/kernel/k_process.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/service/glue/arp.h"
 #include "core/hle/service/glue/errors.h"
 #include "core/hle/service/glue/glue_manager.h"
-#include "core/hle/service/service.h"
 
 namespace Service::Glue {
 

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

@@ -8,12 +8,9 @@
 #include "common/settings.h"
 #include "core/core.h"
 #include "core/core_timing.h"
-#include "core/core_timing_util.h"
 #include "core/frontend/emu_window.h"
 #include "core/frontend/input.h"
-#include "core/hardware_properties.h"
 #include "core/hle/ipc_helpers.h"
-#include "core/hle/kernel/k_client_port.h"
 #include "core/hle/kernel/k_readable_event.h"
 #include "core/hle/kernel/k_shared_memory.h"
 #include "core/hle/kernel/k_transfer_memory.h"
@@ -23,7 +20,6 @@
 #include "core/hle/service/hid/hid.h"
 #include "core/hle/service/hid/irs.h"
 #include "core/hle/service/hid/xcd.h"
-#include "core/hle/service/service.h"
 #include "core/memory.h"
 
 #include "core/hle/service/hid/controllers/console_sixaxis.h"

src/core/hle/service/lbl/lbl.cpp

@@ -7,7 +7,6 @@
 
 #include "common/logging/log.h"
 #include "core/hle/ipc_helpers.h"
-#include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/service/lbl/lbl.h"
 #include "core/hle/service/service.h"
 #include "core/hle/service/sm/sm.h"

src/core/hle/service/mii/mii.cpp

@@ -6,7 +6,6 @@
 
 #include "common/logging/log.h"
 #include "core/hle/ipc_helpers.h"
-#include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/service/mii/mii.h"
 #include "core/hle/service/mii/mii_manager.h"
 #include "core/hle/service/service.h"

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

@@ -7,7 +7,6 @@
 #include "common/logging/log.h"
 #include "common/settings.h"
 #include "core/hle/ipc_helpers.h"
-#include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/service/nfc/nfc.h"
 #include "core/hle/service/service.h"
 #include "core/hle/service/sm/sm.h"

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

@@ -9,7 +9,6 @@
 #include "core/file_sys/patch_manager.h"
 #include "core/file_sys/vfs.h"
 #include "core/hle/ipc_helpers.h"
-#include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/service/ns/errors.h"
 #include "core/hle/service/ns/language.h"
 #include "core/hle/service/ns/ns.h"

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

@@ -92,6 +92,7 @@ NvResult nvhost_ctrl::IocCtrlEventWait(const std::vector<u8>& input, std::vector
     if (syncpoint_manager.IsSyncpointExpired(params.syncpt_id, params.threshold)) {
         params.value = syncpoint_manager.GetSyncpointMin(params.syncpt_id);
         std::memcpy(output.data(), &params, sizeof(params));
+        events_interface.failed[event_id] = false;
         return NvResult::Success;
     }
 
@@ -99,6 +100,7 @@ NvResult nvhost_ctrl::IocCtrlEventWait(const std::vector<u8>& input, std::vector
         syncpoint_manager.IsSyncpointExpired(params.syncpt_id, params.threshold)) {
         params.value = new_value;
         std::memcpy(output.data(), &params, sizeof(params));
+        events_interface.failed[event_id] = false;
         return NvResult::Success;
     }
 
@@ -117,6 +119,7 @@ NvResult nvhost_ctrl::IocCtrlEventWait(const std::vector<u8>& input, std::vector
         event.event->GetWritableEvent().Signal();
         params.value = current_syncpoint_value;
         std::memcpy(output.data(), &params, sizeof(params));
+        events_interface.failed[event_id] = false;
         return NvResult::Success;
     }
     const u32 target_value = current_syncpoint_value - diff;
@@ -146,6 +149,16 @@ NvResult nvhost_ctrl::IocCtrlEventWait(const std::vector<u8>& input, std::vector
     }
     params.value |= event_id;
     event.event->GetWritableEvent().Clear();
+    if (events_interface.failed[event_id]) {
+        {
+            auto lk = system.StallCPU();
+            gpu.WaitFence(params.syncpt_id, target_value);
+            system.UnstallCPU();
+        }
+        std::memcpy(output.data(), &params, sizeof(params));
+        events_interface.failed[event_id] = false;
+        return NvResult::Success;
+    }
     gpu.RegisterSyncptInterrupt(params.syncpt_id, target_value);
     std::memcpy(output.data(), &params, sizeof(params));
     return NvResult::Timeout;
@@ -201,6 +214,7 @@ NvResult nvhost_ctrl::IocCtrlClearEventWait(const std::vector<u8>& input, std::v
     if (events_interface.status[event_id] == EventState::Waiting) {
         events_interface.LiberateEvent(event_id);
     }
+    events_interface.failed[event_id] = true;
 
     syncpoint_manager.RefreshSyncpoint(events_interface.events[event_id].fence.id);
 

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

@@ -49,6 +49,8 @@ struct EventInterface {
     std::array<EventState, MaxNvEvents> status{};
     // Tells if an NVEvent is registered or not
     std::array<bool, MaxNvEvents> registered{};
+    // Tells the NVEvent that it has failed.
+    std::array<bool, MaxNvEvents> failed{};
     // When an NVEvent is waiting on GPU interrupt, this is the sync_point
     // associated with it.
     std::array<u32, MaxNvEvents> assigned_syncpt{};

src/core/hle/service/olsc/olsc.cpp

@@ -3,7 +3,6 @@
 // Refer to the license.txt file included.
 
 #include "core/hle/ipc_helpers.h"
-#include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/service/olsc/olsc.h"
 #include "core/hle/service/service.h"
 #include "core/hle/service/sm/sm.h"

src/core/hle/service/set/set_sys.cpp

@@ -7,7 +7,6 @@
 #include "core/file_sys/errors.h"
 #include "core/file_sys/system_archive/system_version.h"
 #include "core/hle/ipc_helpers.h"
-#include "core/hle/kernel/k_client_port.h"
 #include "core/hle/service/filesystem/filesystem.h"
 #include "core/hle/service/set/set_sys.h"
 

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

@@ -5,11 +5,9 @@
 #pragma once
 
 #include <memory>
-#include <string_view>
 #include <vector>
 
 #include "common/common_types.h"
-#include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/service/service.h"
 #include "core/hle/service/sockets/sockets.h"
 

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

@@ -4,7 +4,6 @@
 
 #pragma once
 
-#include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/service/service.h"
 
 namespace Core {

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

@@ -4,13 +4,17 @@
 
 #pragma once
 
+#include "common/common_funcs.h"
 #include "common/common_types.h"
-#include "core/hle/service/service.h"
 
 namespace Core {
 class System;
 }
 
+namespace Service::SM {
+class ServiceManager;
+}
+
 namespace Service::Sockets {
 
 enum class Errno : u32 {

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

@@ -3,10 +3,8 @@
 // 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 "common/settings.h"

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

@@ -3,7 +3,6 @@
 // Refer to the license.txt file included.
 
 #include "core/hle/ipc_helpers.h"
-#include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/service/service.h"
 #include "core/hle/service/sm/sm.h"
 #include "core/hle/service/ssl/ssl.h"

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

@@ -8,11 +8,11 @@
 #include "core/core_timing_util.h"
 #include "core/hardware_properties.h"
 #include "core/hle/ipc_helpers.h"
-#include "core/hle/kernel/k_client_port.h"
 #include "core/hle/kernel/k_scheduler.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/service/time/time.h"
 #include "core/hle/service/time/time_interface.h"
+#include "core/hle/service/time/time_manager.h"
 #include "core/hle/service/time/time_sharedmemory.h"
 #include "core/hle/service/time/time_zone_service.h"
 

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

@@ -6,7 +6,6 @@
 
 #include "core/hle/service/service.h"
 #include "core/hle/service/time/clock_types.h"
-#include "core/hle/service/time/time_manager.h"
 
 namespace Core {
 class System;

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

@@ -6,7 +6,6 @@
 
 #include "common/logging/log.h"
 #include "core/hle/ipc_helpers.h"
-#include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/service/service.h"
 #include "core/hle/service/sm/sm.h"
 #include "core/hle/service/usb/usb.h"

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

@@ -524,7 +524,9 @@ private:
         Disconnect = 11,
 
         AllocateBuffers = 13,
-        SetPreallocatedBuffer = 14
+        SetPreallocatedBuffer = 14,
+
+        GetBufferHistory = 17
     };
 
     void TransactParcel(Kernel::HLERequestContext& ctx) {
@@ -641,6 +643,14 @@ private:
             ctx.WriteBuffer(response.Serialize());
             break;
         }
+        case TransactionId::GetBufferHistory: {
+            LOG_WARNING(Service_VI, "(STUBBED) called, transaction=GetBufferHistory");
+            [[maybe_unused]] const auto buffer = ctx.ReadBuffer();
+
+            IGBPEmptyResponseParcel response{};
+            ctx.WriteBuffer(response.Serialize());
+            break;
+        }
         default:
             ASSERT_MSG(false, "Unimplemented");
         }

src/core/hle/service/vi/vi.h

@@ -4,7 +4,6 @@
 
 #pragma once
 
-#include <memory>
 #include "common/common_types.h"
 
 namespace Core {

src/shader_recompiler/ir_opt/texture_pass.cpp

@@ -492,7 +492,8 @@ void TexturePass(Environment& env, IR::Program& program) {
             const auto insert_point{IR::Block::InstructionList::s_iterator_to(*inst)};
             IR::IREmitter ir{*texture_inst.block, insert_point};
             const IR::U32 shift{ir.Imm32(std::countr_zero(DESCRIPTOR_SIZE))};
-            inst->SetArg(0, ir.ShiftRightArithmetic(cbuf.dynamic_offset, shift));
+            inst->SetArg(0, ir.SMin(ir.ShiftRightArithmetic(cbuf.dynamic_offset, shift),
+                                    ir.Imm32(DESCRIPTOR_SIZE - 1)));
         } else {
             inst->SetArg(0, IR::Value{});
         }

src/video_core/command_classes/codecs/h264.cpp

@@ -20,6 +20,8 @@
 
 #include <array>
 #include <bit>
+
+#include "common/settings.h"
 #include "video_core/command_classes/codecs/h264.h"
 #include "video_core/gpu.h"
 #include "video_core/memory_manager.h"
@@ -96,7 +98,10 @@ const std::vector<u8>& H264::ComposeFrameHeader(const NvdecCommon::NvdecRegister
                            (context.h264_parameter_set.frame_mbs_only_flag ? 1 : 2);
 
     // TODO (ameerj): Where do we get this number, it seems to be particular for each stream
-    writer.WriteUe(6); // Max number of reference frames
+    const auto nvdec_decoding = Settings::values.nvdec_emulation.GetValue();
+    const bool uses_gpu_decoding = nvdec_decoding == Settings::NvdecEmulation::GPU;
+    const u32 max_num_ref_frames = uses_gpu_decoding ? 6u : 16u;
+    writer.WriteUe(max_num_ref_frames);
     writer.WriteBit(false);
     writer.WriteUe(context.h264_parameter_set.pic_width_in_mbs - 1);
     writer.WriteUe(pic_height - 1);

src/video_core/command_classes/vic.cpp

@@ -16,6 +16,7 @@ extern "C" {
 }
 
 #include "common/assert.h"
+#include "common/bit_field.h"
 #include "common/logging/log.h"
 
 #include "video_core/command_classes/nvdec.h"
@@ -26,6 +27,25 @@ extern "C" {
 #include "video_core/textures/decoders.h"
 
 namespace Tegra {
+namespace {
+enum class VideoPixelFormat : u64_le {
+    RGBA8 = 0x1f,
+    BGRA8 = 0x20,
+    RGBX8 = 0x23,
+    YUV420 = 0x44,
+};
+} // Anonymous namespace
+
+union VicConfig {
+    u64_le raw{};
+    BitField<0, 7, VideoPixelFormat> pixel_format;
+    BitField<7, 2, u64_le> chroma_loc_horiz;
+    BitField<9, 2, u64_le> chroma_loc_vert;
+    BitField<11, 4, u64_le> block_linear_kind;
+    BitField<15, 4, u64_le> block_linear_height_log2;
+    BitField<32, 14, u64_le> surface_width_minus1;
+    BitField<46, 14, u64_le> surface_height_minus1;
+};
 
 Vic::Vic(GPU& gpu_, std::shared_ptr<Nvdec> nvdec_processor_)
     : gpu(gpu_),
@@ -65,134 +85,155 @@ void Vic::Execute() {
     if (!frame) {
         return;
     }
-    const auto pixel_format = static_cast<VideoPixelFormat>(config.pixel_format.Value());
-    switch (pixel_format) {
+    const u64 surface_width = config.surface_width_minus1 + 1;
+    const u64 surface_height = config.surface_height_minus1 + 1;
+    if (static_cast<u64>(frame->width) != surface_width ||
+        static_cast<u64>(frame->height) != surface_height) {
+        // TODO: Properly support multiple video streams with differing frame dimensions
+        LOG_WARNING(Service_NVDRV, "Frame dimensions {}x{} don't match surface dimensions {}x{}",
+                    frame->width, frame->height, surface_width, surface_height);
+    }
+    switch (config.pixel_format) {
+    case VideoPixelFormat::RGBA8:
     case VideoPixelFormat::BGRA8:
-    case VideoPixelFormat::RGBA8: {
-        LOG_TRACE(Service_NVDRV, "Writing RGB Frame");
-
-        if (scaler_ctx == nullptr || frame->width != scaler_width ||
-            frame->height != scaler_height) {
-            const AVPixelFormat target_format =
-                (pixel_format == VideoPixelFormat::RGBA8) ? AV_PIX_FMT_RGBA : AV_PIX_FMT_BGRA;
-
-            sws_freeContext(scaler_ctx);
-            scaler_ctx = nullptr;
-
-            // Frames are decoded into either YUV420 or NV12 formats. Convert to desired format
-            scaler_ctx = sws_getContext(frame->width, frame->height,
-                                        static_cast<AVPixelFormat>(frame->format), frame->width,
-                                        frame->height, target_format, 0, nullptr, nullptr, nullptr);
-
-            scaler_width = frame->width;
-            scaler_height = frame->height;
-        }
-        // Get Converted frame
-        const u32 width = static_cast<u32>(frame->width);
-        const u32 height = static_cast<u32>(frame->height);
-        const std::size_t linear_size = width * height * 4;
-
-        // Only allocate frame_buffer once per stream, as the size is not expected to change
-        if (!converted_frame_buffer) {
-            converted_frame_buffer = AVMallocPtr{static_cast<u8*>(av_malloc(linear_size)), av_free};
-        }
-        const std::array<int, 4> converted_stride{frame->width * 4, frame->height * 4, 0, 0};
-        u8* const converted_frame_buf_addr{converted_frame_buffer.get()};
-
-        sws_scale(scaler_ctx, frame->data, frame->linesize, 0, frame->height,
-                  &converted_frame_buf_addr, converted_stride.data());
-
-        const u32 blk_kind = static_cast<u32>(config.block_linear_kind);
-        if (blk_kind != 0) {
-            // swizzle pitch linear to block linear
-            const u32 block_height = static_cast<u32>(config.block_linear_height_log2);
-            const auto size =
-                Tegra::Texture::CalculateSize(true, 4, width, height, 1, block_height, 0);
-            luma_buffer.resize(size);
-            Tegra::Texture::SwizzleSubrect(width, height, width * 4, width, 4, luma_buffer.data(),
-                                           converted_frame_buffer.get(), block_height, 0, 0);
-
-            gpu.MemoryManager().WriteBlock(output_surface_luma_address, luma_buffer.data(), size);
-        } else {
-            // send pitch linear frame
-            gpu.MemoryManager().WriteBlock(output_surface_luma_address, converted_frame_buf_addr,
-                                           linear_size);
-        }
+    case VideoPixelFormat::RGBX8:
+        WriteRGBFrame(frame, config);
         break;
-    }
-    case VideoPixelFormat::Yuv420: {
-        LOG_TRACE(Service_NVDRV, "Writing YUV420 Frame");
-
-        const std::size_t surface_width = config.surface_width_minus1 + 1;
-        const std::size_t surface_height = config.surface_height_minus1 + 1;
-        const auto frame_width = std::min(surface_width, static_cast<size_t>(frame->width));
-        const auto frame_height = std::min(surface_height, static_cast<size_t>(frame->height));
-        const std::size_t aligned_width = (surface_width + 0xff) & ~0xffUL;
-
-        const auto stride = static_cast<size_t>(frame->linesize[0]);
-
-        luma_buffer.resize(aligned_width * surface_height);
-        chroma_buffer.resize(aligned_width * surface_height / 2);
-
-        // Populate luma buffer
-        const u8* luma_src = frame->data[0];
-        for (std::size_t y = 0; y < frame_height; ++y) {
-            const std::size_t src = y * stride;
-            const std::size_t dst = y * aligned_width;
-            for (std::size_t x = 0; x < frame_width; ++x) {
-                luma_buffer[dst + x] = luma_src[src + x];
-            }
-        }
-        gpu.MemoryManager().WriteBlock(output_surface_luma_address, luma_buffer.data(),
-                                       luma_buffer.size());
-
-        // Chroma
-        const std::size_t half_height = frame_height / 2;
-        const auto half_stride = static_cast<size_t>(frame->linesize[1]);
-
-        switch (frame->format) {
-        case AV_PIX_FMT_YUV420P: {
-            // Frame from FFmpeg software
-            // Populate chroma buffer from both channels with interleaving.
-            const std::size_t half_width = frame_width / 2;
-            const u8* chroma_b_src = frame->data[1];
-            const u8* chroma_r_src = frame->data[2];
-            for (std::size_t y = 0; y < half_height; ++y) {
-                const std::size_t src = y * half_stride;
-                const std::size_t dst = y * aligned_width;
-
-                for (std::size_t x = 0; x < half_width; ++x) {
-                    chroma_buffer[dst + x * 2] = chroma_b_src[src + x];
-                    chroma_buffer[dst + x * 2 + 1] = chroma_r_src[src + x];
-                }
-            }
-            break;
-        }
-        case AV_PIX_FMT_NV12: {
-            // Frame from VA-API hardware
-            // This is already interleaved so just copy
-            const u8* chroma_src = frame->data[1];
-            for (std::size_t y = 0; y < half_height; ++y) {
-                const std::size_t src = y * stride;
-                const std::size_t dst = y * aligned_width;
-                for (std::size_t x = 0; x < frame_width; ++x) {
-                    chroma_buffer[dst + x] = chroma_src[src + x];
-                }
-            }
-            break;
-        }
-        default:
-            UNREACHABLE();
-            break;
-        }
-        gpu.MemoryManager().WriteBlock(output_surface_chroma_address, chroma_buffer.data(),
-                                       chroma_buffer.size());
+    case VideoPixelFormat::YUV420:
+        WriteYUVFrame(frame, config);
         break;
-    }
     default:
-        UNIMPLEMENTED_MSG("Unknown video pixel format {}", config.pixel_format.Value());
+        UNIMPLEMENTED_MSG("Unknown video pixel format {:X}", config.pixel_format.Value());
         break;
     }
 }
 
+void Vic::WriteRGBFrame(const AVFrame* frame, const VicConfig& config) {
+    LOG_TRACE(Service_NVDRV, "Writing RGB Frame");
+
+    if (!scaler_ctx || frame->width != scaler_width || frame->height != scaler_height) {
+        const AVPixelFormat target_format = [pixel_format = config.pixel_format]() {
+            switch (pixel_format) {
+            case VideoPixelFormat::RGBA8:
+                return AV_PIX_FMT_RGBA;
+            case VideoPixelFormat::BGRA8:
+                return AV_PIX_FMT_BGRA;
+            case VideoPixelFormat::RGBX8:
+                return AV_PIX_FMT_RGB0;
+            default:
+                return AV_PIX_FMT_RGBA;
+            }
+        }();
+
+        sws_freeContext(scaler_ctx);
+        // Frames are decoded into either YUV420 or NV12 formats. Convert to desired RGB format
+        scaler_ctx = sws_getContext(frame->width, frame->height,
+                                    static_cast<AVPixelFormat>(frame->format), frame->width,
+                                    frame->height, target_format, 0, nullptr, nullptr, nullptr);
+        scaler_width = frame->width;
+        scaler_height = frame->height;
+        converted_frame_buffer.reset();
+    }
+    if (!converted_frame_buffer) {
+        const size_t frame_size = frame->width * frame->height * 4;
+        converted_frame_buffer = AVMallocPtr{static_cast<u8*>(av_malloc(frame_size)), av_free};
+    }
+    const std::array<int, 4> converted_stride{frame->width * 4, frame->height * 4, 0, 0};
+    u8* const converted_frame_buf_addr{converted_frame_buffer.get()};
+    sws_scale(scaler_ctx, frame->data, frame->linesize, 0, frame->height, &converted_frame_buf_addr,
+              converted_stride.data());
+
+    // Use the minimum of surface/frame dimensions to avoid buffer overflow.
+    const u32 surface_width = static_cast<u32>(config.surface_width_minus1) + 1;
+    const u32 surface_height = static_cast<u32>(config.surface_height_minus1) + 1;
+    const u32 width = std::min(surface_width, static_cast<u32>(frame->width));
+    const u32 height = std::min(surface_height, static_cast<u32>(frame->height));
+    const u32 blk_kind = static_cast<u32>(config.block_linear_kind);
+    if (blk_kind != 0) {
+        // swizzle pitch linear to block linear
+        const u32 block_height = static_cast<u32>(config.block_linear_height_log2);
+        const auto size = Texture::CalculateSize(true, 4, width, height, 1, block_height, 0);
+        luma_buffer.resize(size);
+        Texture::SwizzleSubrect(width, height, width * 4, width, 4, luma_buffer.data(),
+                                converted_frame_buf_addr, block_height, 0, 0);
+
+        gpu.MemoryManager().WriteBlock(output_surface_luma_address, luma_buffer.data(), size);
+    } else {
+        // send pitch linear frame
+        const size_t linear_size = width * height * 4;
+        gpu.MemoryManager().WriteBlock(output_surface_luma_address, converted_frame_buf_addr,
+                                       linear_size);
+    }
+}
+
+void Vic::WriteYUVFrame(const AVFrame* frame, const VicConfig& config) {
+    LOG_TRACE(Service_NVDRV, "Writing YUV420 Frame");
+
+    const std::size_t surface_width = config.surface_width_minus1 + 1;
+    const std::size_t surface_height = config.surface_height_minus1 + 1;
+    const std::size_t aligned_width = (surface_width + 0xff) & ~0xffUL;
+    // Use the minimum of surface/frame dimensions to avoid buffer overflow.
+    const auto frame_width = std::min(surface_width, static_cast<size_t>(frame->width));
+    const auto frame_height = std::min(surface_height, static_cast<size_t>(frame->height));
+
+    const auto stride = static_cast<size_t>(frame->linesize[0]);
+
+    luma_buffer.resize(aligned_width * surface_height);
+    chroma_buffer.resize(aligned_width * surface_height / 2);
+
+    // Populate luma buffer
+    const u8* luma_src = frame->data[0];
+    for (std::size_t y = 0; y < frame_height; ++y) {
+        const std::size_t src = y * stride;
+        const std::size_t dst = y * aligned_width;
+        for (std::size_t x = 0; x < frame_width; ++x) {
+            luma_buffer[dst + x] = luma_src[src + x];
+        }
+    }
+    gpu.MemoryManager().WriteBlock(output_surface_luma_address, luma_buffer.data(),
+                                   luma_buffer.size());
+
+    // Chroma
+    const std::size_t half_height = frame_height / 2;
+    const auto half_stride = static_cast<size_t>(frame->linesize[1]);
+
+    switch (frame->format) {
+    case AV_PIX_FMT_YUV420P: {
+        // Frame from FFmpeg software
+        // Populate chroma buffer from both channels with interleaving.
+        const std::size_t half_width = frame_width / 2;
+        const u8* chroma_b_src = frame->data[1];
+        const u8* chroma_r_src = frame->data[2];
+        for (std::size_t y = 0; y < half_height; ++y) {
+            const std::size_t src = y * half_stride;
+            const std::size_t dst = y * aligned_width;
+
+            for (std::size_t x = 0; x < half_width; ++x) {
+                chroma_buffer[dst + x * 2] = chroma_b_src[src + x];
+                chroma_buffer[dst + x * 2 + 1] = chroma_r_src[src + x];
+            }
+        }
+        break;
+    }
+    case AV_PIX_FMT_NV12: {
+        // Frame from VA-API hardware
+        // This is already interleaved so just copy
+        const u8* chroma_src = frame->data[1];
+        for (std::size_t y = 0; y < half_height; ++y) {
+            const std::size_t src = y * stride;
+            const std::size_t dst = y * aligned_width;
+            for (std::size_t x = 0; x < frame_width; ++x) {
+                chroma_buffer[dst + x] = chroma_src[src + x];
+            }
+        }
+        break;
+    }
+    default:
+        UNREACHABLE();
+        break;
+    }
+    gpu.MemoryManager().WriteBlock(output_surface_chroma_address, chroma_buffer.data(),
+                                   chroma_buffer.size());
+}
+
 } // namespace Tegra

src/video_core/command_classes/vic.h

@@ -6,7 +6,6 @@
 
 #include <memory>
 #include <vector>
-#include "common/bit_field.h"
 #include "common/common_types.h"
 
 struct SwsContext;
@@ -14,6 +13,7 @@ struct SwsContext;
 namespace Tegra {
 class GPU;
 class Nvdec;
+union VicConfig;
 
 class Vic {
 public:
@@ -27,6 +27,7 @@ public:
     };
 
     explicit Vic(GPU& gpu, std::shared_ptr<Nvdec> nvdec_processor);
+
     ~Vic();
 
     /// Write to the device state.
@@ -35,22 +36,9 @@ public:
 private:
     void Execute();
 
-    enum class VideoPixelFormat : u64_le {
-        RGBA8 = 0x1f,
-        BGRA8 = 0x20,
-        Yuv420 = 0x44,
-    };
+    void WriteRGBFrame(const AVFrame* frame, const VicConfig& config);
 
-    union VicConfig {
-        u64_le raw{};
-        BitField<0, 7, u64_le> pixel_format;
-        BitField<7, 2, u64_le> chroma_loc_horiz;
-        BitField<9, 2, u64_le> chroma_loc_vert;
-        BitField<11, 4, u64_le> block_linear_kind;
-        BitField<15, 4, u64_le> block_linear_height_log2;
-        BitField<32, 14, u64_le> surface_width_minus1;
-        BitField<46, 14, u64_le> surface_height_minus1;
-    };
+    void WriteYUVFrame(const AVFrame* frame, const VicConfig& config);
 
     GPU& gpu;
     std::shared_ptr<Tegra::Nvdec> nvdec_processor;

src/video_core/query_cache.h

@@ -258,9 +258,9 @@ private:
 
     void AsyncFlushQuery(VAddr addr) {
         if (!uncommitted_flushes) {
-            uncommitted_flushes = std::make_shared<std::unordered_set<VAddr>>();
+            uncommitted_flushes = std::make_shared<std::vector<VAddr>>();
         }
-        uncommitted_flushes->insert(addr);
+        uncommitted_flushes->push_back(addr);
     }
 
     static constexpr std::uintptr_t PAGE_SIZE = 4096;
@@ -276,8 +276,8 @@ private:
 
     std::array<CounterStream, VideoCore::NumQueryTypes> streams;
 
-    std::shared_ptr<std::unordered_set<VAddr>> uncommitted_flushes{};
-    std::list<std::shared_ptr<std::unordered_set<VAddr>>> committed_flushes;
+    std::shared_ptr<std::vector<VAddr>> uncommitted_flushes{};
+    std::list<std::shared_ptr<std::vector<VAddr>>> committed_flushes;
 };
 
 template <class QueryCache, class HostCounter>

src/video_core/rasterizer_accelerated.h

@@ -42,7 +42,7 @@ private:
     };
     static_assert(sizeof(CacheEntry) == 8, "CacheEntry should be 8 bytes!");
 
-    std::array<CacheEntry, 0x1000000> cached_pages;
+    std::array<CacheEntry, 0x2000000> cached_pages;
     Core::Memory::Memory& cpu_memory;
 };
 

src/video_core/renderer_vulkan/vk_master_semaphore.h

@@ -21,12 +21,12 @@ public:
 
     /// Returns the current logical tick.
     [[nodiscard]] u64 CurrentTick() const noexcept {
-        return current_tick.load(std::memory_order_relaxed);
+        return current_tick.load(std::memory_order_acquire);
     }
 
     /// Returns the last known GPU tick.
     [[nodiscard]] u64 KnownGpuTick() const noexcept {
-        return gpu_tick.load(std::memory_order_relaxed);
+        return gpu_tick.load(std::memory_order_acquire);
     }
 
     /// Returns the timeline semaphore handle.
@@ -41,12 +41,21 @@ public:
 
     /// Advance to the logical tick and return the old one
     [[nodiscard]] u64 NextTick() noexcept {
-        return current_tick.fetch_add(1, std::memory_order::relaxed);
+        return current_tick.fetch_add(1, std::memory_order_release);
     }
 
     /// Refresh the known GPU tick
     void Refresh() {
-        gpu_tick.store(semaphore.GetCounter(), std::memory_order_relaxed);
+        u64 this_tick{};
+        u64 counter{};
+        do {
+            this_tick = gpu_tick.load(std::memory_order_acquire);
+            counter = semaphore.GetCounter();
+            if (counter < this_tick) {
+                return;
+            }
+        } while (!gpu_tick.compare_exchange_weak(this_tick, counter, std::memory_order_release,
+                                                 std::memory_order_relaxed));
     }
 
     /// Waits for a tick to be hit on the GPU

src/video_core/renderer_vulkan/vk_query_cache.cpp

@@ -117,7 +117,8 @@ u64 HostCounter::BlockingQuery() const {
     cache.GetScheduler().Wait(tick);
     u64 data;
     const VkResult query_result = cache.GetDevice().GetLogical().GetQueryResults(
-        query.first, query.second, 1, sizeof(data), &data, sizeof(data), VK_QUERY_RESULT_64_BIT);
+        query.first, query.second, 1, sizeof(data), &data, sizeof(data),
+        VK_QUERY_RESULT_64_BIT | VK_QUERY_RESULT_WAIT_BIT);
 
     switch (query_result) {
     case VK_SUCCESS:

src/video_core/renderer_vulkan/vk_state_tracker.cpp

@@ -54,6 +54,7 @@ void SetupDirtyViewports(Tables& tables) {
     FillBlock(tables[0], OFF(viewport_transform), NUM(viewport_transform), Viewports);
     FillBlock(tables[0], OFF(viewports), NUM(viewports), Viewports);
     tables[0][OFF(viewport_transform_enabled)] = Viewports;
+    tables[1][OFF(screen_y_control)] = Viewports;
 }
 
 void SetupDirtyScissors(Tables& tables) {

src/yuzu/about_dialog.cpp

@@ -8,7 +8,8 @@
 #include "ui_aboutdialog.h"
 #include "yuzu/about_dialog.h"
 
-AboutDialog::AboutDialog(QWidget* parent) : QDialog(parent), ui(new Ui::AboutDialog) {
+AboutDialog::AboutDialog(QWidget* parent)
+    : QDialog(parent), ui{std::make_unique<Ui::AboutDialog>()} {
     const auto branch_name = std::string(Common::g_scm_branch);
     const auto description = std::string(Common::g_scm_desc);
     const auto build_id = std::string(Common::g_build_id);

src/yuzu/applets/qt_controller.cpp

@@ -37,17 +37,14 @@ constexpr std::array<std::array<bool, 4>, 8> led_patterns{{
 }};
 
 void UpdateController(Settings::ControllerType controller_type, std::size_t npad_index,
-                      bool connected) {
-    Core::System& system{Core::System::GetInstance()};
-
+                      bool connected, Core::System& system) {
     if (!system.IsPoweredOn()) {
         return;
     }
 
-    Service::SM::ServiceManager& sm = system.ServiceManager();
-
     auto& npad =
-        sm.GetService<Service::HID::Hid>("hid")
+        system.ServiceManager()
+            .GetService<Service::HID::Hid>("hid")
             ->GetAppletResource()
             ->GetController<Service::HID::Controller_NPad>(Service::HID::HidController::NPad);
 
@@ -79,10 +76,10 @@ bool IsControllerCompatible(Settings::ControllerType controller_type,
 
 QtControllerSelectorDialog::QtControllerSelectorDialog(
     QWidget* parent, Core::Frontend::ControllerParameters parameters_,
-    InputCommon::InputSubsystem* input_subsystem_)
+    InputCommon::InputSubsystem* input_subsystem_, Core::System& system_)
     : QDialog(parent), ui(std::make_unique<Ui::QtControllerSelectorDialog>()),
       parameters(std::move(parameters_)), input_subsystem{input_subsystem_},
-      input_profiles(std::make_unique<InputProfiles>()) {
+      input_profiles(std::make_unique<InputProfiles>(system_)), system{system_} {
     ui->setupUi(this);
 
     player_widgets = {
@@ -245,7 +242,7 @@ int QtControllerSelectorDialog::exec() {
 void QtControllerSelectorDialog::ApplyConfiguration() {
     const bool pre_docked_mode = Settings::values.use_docked_mode.GetValue();
     Settings::values.use_docked_mode.SetValue(ui->radioDocked->isChecked());
-    OnDockedModeChanged(pre_docked_mode, Settings::values.use_docked_mode.GetValue());
+    OnDockedModeChanged(pre_docked_mode, Settings::values.use_docked_mode.GetValue(), system);
 
     Settings::values.vibration_enabled.SetValue(ui->vibrationGroup->isChecked());
     Settings::values.motion_enabled.SetValue(ui->motionGroup->isChecked());
@@ -293,7 +290,7 @@ void QtControllerSelectorDialog::CallConfigureMotionTouchDialog() {
 }
 
 void QtControllerSelectorDialog::CallConfigureInputProfileDialog() {
-    ConfigureInputProfileDialog dialog(this, input_subsystem, input_profiles.get());
+    ConfigureInputProfileDialog dialog(this, input_subsystem, input_profiles.get(), system);
 
     dialog.setWindowFlags(Qt::Dialog | Qt::CustomizeWindowHint | Qt::WindowTitleHint |
                           Qt::WindowSystemMenuHint);
@@ -533,7 +530,7 @@ void QtControllerSelectorDialog::UpdateControllerState(std::size_t player_index)
     }
 
     // Disconnect the controller first.
-    UpdateController(controller_type, player_index, false);
+    UpdateController(controller_type, player_index, false, system);
 
     player.controller_type = controller_type;
     player.connected = player_connected;
@@ -548,7 +545,7 @@ void QtControllerSelectorDialog::UpdateControllerState(std::size_t player_index)
         }
         handheld.connected = player_groupboxes[player_index]->isChecked() &&
                              controller_type == Settings::ControllerType::Handheld;
-        UpdateController(Settings::ControllerType::Handheld, 8, handheld.connected);
+        UpdateController(Settings::ControllerType::Handheld, 8, handheld.connected, system);
     }
 
     if (!player.connected) {
@@ -560,7 +557,7 @@ void QtControllerSelectorDialog::UpdateControllerState(std::size_t player_index)
     using namespace std::chrono_literals;
     std::this_thread::sleep_for(60ms);
 
-    UpdateController(controller_type, player_index, player_connected);
+    UpdateController(controller_type, player_index, player_connected, system);
 }
 
 void QtControllerSelectorDialog::UpdateLEDPattern(std::size_t player_index) {
@@ -659,7 +656,8 @@ void QtControllerSelectorDialog::DisableUnsupportedPlayers() {
     for (std::size_t index = max_supported_players; index < NUM_PLAYERS; ++index) {
         // Disconnect any unsupported players here and disable or hide them if applicable.
         Settings::values.players.GetValue()[index].connected = false;
-        UpdateController(Settings::values.players.GetValue()[index].controller_type, index, false);
+        UpdateController(Settings::values.players.GetValue()[index].controller_type, index, false,
+                         system);
         // Hide the player widgets when max_supported_controllers is less than or equal to 4.
         if (max_supported_players <= 4) {
             player_widgets[index]->hide();

src/yuzu/applets/qt_controller.h

@@ -7,6 +7,7 @@
 #include <array>
 #include <memory>
 #include <QDialog>
+#include "core/core.h"
 #include "core/frontend/applets/controller.h"
 
 class GMainWindow;
@@ -36,7 +37,8 @@ class QtControllerSelectorDialog final : public QDialog {
 public:
     explicit QtControllerSelectorDialog(QWidget* parent,
                                         Core::Frontend::ControllerParameters parameters_,
-                                        InputCommon::InputSubsystem* input_subsystem_);
+                                        InputCommon::InputSubsystem* input_subsystem_,
+                                        Core::System& system_);
     ~QtControllerSelectorDialog() override;
 
     int exec() override;
@@ -103,6 +105,8 @@ private:
 
     std::unique_ptr<InputProfiles> input_profiles;
 
+    Core::System& system;
+
     // This is true if and only if all parameters are met. Otherwise, this is false.
     // This determines whether the "OK" button can be clicked to exit the applet.
     bool parameters_met{false};

src/yuzu/applets/qt_web_browser.cpp

@@ -3,6 +3,7 @@
 // Refer to the license.txt file included.
 
 #ifdef YUZU_USE_QT_WEB_ENGINE
+#include <QApplication>
 #include <QKeyEvent>
 
 #include <QWebEngineProfile>