rpcs3_version: Bump to 0.0.27

Update FW 4.90 Latest Version
Updates the latest FW version according to: https://www.playstation.com/en-us/support/hardware/ps3/system-software/ List of changes: This system software update improves system performance.
2023-03-01 21:56:02 +00:00 · 2023-03-01 20:25:24 +01:00 · 2023-03-01 18:03:55 +01:00 · 2023-03-01 10:15:21 +01:00 · 2023-03-01 01:10:06 +03:00 · 2023-03-01 01:10:06 +03:00
1374 changed files with 205661 additions and 79528 deletions
@@ -7,13 +7,21 @@ git submodule -q update --init --depth 1 $(awk '/path/ && !/llvm/ { print $3 }'

 # Prefer newer Clang than in base system (see also .ci/install-freebsd.sh)
 # libc++ isn't in llvm* packages, so download manually
-fetch https://github.com/llvm/llvm-project/releases/download/llvmorg-11.0.0/libcxx-11.0.0.src.tar.xz
-tar xf libcxx-11.0.0.src.tar.xz
-export CC=clang11 CXX=clang++11
-export CXXFLAGS="$CXXFLAGS -nostdinc++ -isystem $PWD/libcxx-11.0.0.src/include"
+fetch https://github.com/llvm/llvm-project/releases/download/llvmorg-15.0.6/llvm-project-15.0.6.src.tar.xz
+tar xf llvm*.tar.xz
+export CC=clang15 CXX=clang++15
+cmake -B libcxx_build -G Ninja -S llvm*/libcxx \
+      -DLLVM_CCACHE_BUILD=ON \
+      -DLIBCXX_INCLUDE_BENCHMARKS=OFF \
+      -DCMAKE_INSTALL_PREFIX:PATH=libcxx_prefix
+cmake --build libcxx_build
+cmake --install libcxx_build
+export CXXFLAGS="$CXXFLAGS -nostdinc++ -isystem$PWD/libcxx_prefix/include/c++/v1"
+export LDFLAGS="$LDFLAGS -nostdlib++ -L$PWD/libcxx_prefix/lib -l:libc++.a -lcxxrt"

 CONFIGURE_ARGS="
 	-DWITH_LLVM=OFF
+	-DUSE_SDL=OFF
 	-DUSE_PRECOMPILED_HEADERS=OFF
 	-DUSE_NATIVE_INSTRUCTIONS=OFF
 	-DUSE_SYSTEM_FFMPEG=ON
@@ -1,11 +1,13 @@
 #!/bin/sh -ex

 # Setup Qt variables
-export QT_BASE_DIR=/opt/qt${QTVERMIN}
-export PATH=$QT_BASE_DIR/bin:$PATH
-export LD_LIBRARY_PATH=$QT_BASE_DIR/lib/x86_64-linux-gnu:$QT_BASE_DIR/lib
+export QT_BASE_DIR=/opt/qt"${QTVERMIN}"
+export PATH="$QT_BASE_DIR"/bin:"$PATH"
+export LD_LIBRARY_PATH="$QT_BASE_DIR"/lib/x86_64-linux-gnu:"$QT_BASE_DIR"/lib

-cd rpcs3 || exit 1
+if [ -z "$CIRRUS_CI" ]; then
+   cd rpcs3 || exit 1
+fi

 # Pull all the submodules except llvm, since it is built separately and we just download that build
 # Note: Tried to use git submodule status, but it takes over 20 seconds
@@ -21,19 +23,19 @@ mkdir build && cd build || exit 1

 if [ "$COMPILER" = "gcc" ]; then
    # These are set in the dockerfile
-    export CC=${GCC_BINARY}
-    export CXX=${GXX_BINARY}
+    export CC="${GCC_BINARY}"
+    export CXX="${GXX_BINARY}"
    export LINKER=gold
    # We need to set the following variables for LTO to link properly
-    export AR=/usr/bin/gcc-ar-$GCCVER
-    export RANLIB=/usr/bin/gcc-ranlib-$GCCVER
+    export AR=/usr/bin/gcc-ar-"$GCCVER"
+    export RANLIB=/usr/bin/gcc-ranlib-"$GCCVER"
    export CFLAGS="-fuse-linker-plugin"
 else
-    export CC=${CLANG_BINARY}
-    export CXX=${CLANGXX_BINARY}
+    export CC="${CLANG_BINARY}"
+    export CXX="${CLANGXX_BINARY}"
    export LINKER=lld
-    export AR=/usr/bin/llvm-ar-$LLVMVER
-    export RANLIB=/usr/bin/llvm-ranlib-$LLVMVER
+    export AR=/usr/bin/llvm-ar-"$LLVMVER"
+    export RANLIB=/usr/bin/llvm-ranlib-"$LLVMVER"
 fi

 export CFLAGS="$CFLAGS -fuse-ld=${LINKER}"
@@ -49,15 +51,19 @@ cmake ..                                               \
    -DCMAKE_AR="$AR"                                   \
    -DCMAKE_RANLIB="$RANLIB"                           \
    -DUSE_SYSTEM_CURL=ON                               \
+    -DUSE_SDL=OFF                                      \
+    -DOpenGL_GL_PREFERENCE=LEGACY                      \
    -G Ninja

 ninja; build_status=$?;

 cd ..

-# If it compiled succesfully let's deploy depending on the build pipeline (Azure Pipelines).
-# Azure publishes PRs as artifacts only.
-{   [ "$IS_AZURE" = "true" ];
+shellcheck .ci/*.sh
+
+# If it compiled succesfully let's deploy.
+# Azure and Cirrus publish PRs as artifacts only.
+{   [ "$CI_HAS_ARTIFACTS" = "true" ];
 } && SHOULD_DEPLOY="true" || SHOULD_DEPLOY="false"

 if [ "$build_status" -eq 0 ] && [ "$SHOULD_DEPLOY" = "true" ]; then
@@ -1,27 +1,69 @@
 #!/bin/sh -ex

-export CCACHE_SLOPPINESS=pch_defines,time_macros
-export CMAKE_PREFIX_PATH=/usr/local/opt/qt5/
-export PATH="/usr/local/opt/ccache/libexec:$PATH"
+brew install -f --overwrite llvm@14 nasm ninja git p7zip create-dmg ccache

-# Setup vulkan and gfx-rs/portability
-curl -sLO https://github.com/gfx-rs/portability/releases/download/latest/gfx-portability-macos-latest.zip
-unzip -: gfx-portability-macos-latest.zip
-curl -sLO https://github.com/KhronosGroup/Vulkan-Headers/archive/sdk-1.1.106.0.zip
-unzip -: sdk-*.zip
-mkdir vulkan-sdk
-ln -s "${PWD}"/Vulkan-Headers*/include vulkan-sdk/include
-mkdir vulkan-sdk/lib
-cp target/release/libportability.dylib vulkan-sdk/lib/libVulkan.dylib
-# Let macdeployqt locate and install Vulkan library
-install_name_tool -id "${PWD}"/vulkan-sdk/lib/libVulkan.dylib vulkan-sdk/lib/libVulkan.dylib
-export VULKAN_SDK="${PWD}/vulkan-sdk"
+#/usr/sbin/softwareupdate --install-rosetta --agree-to-license
+arch -x86_64 /bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"
+arch -x86_64 /usr/local/homebrew/bin/brew install -f --overwrite llvm@14 sdl2 glew cmake

-# Pull all the submodules except llvm
-# Note: Tried to use git submodule status, but it takes over 20 seconds
-# shellcheck disable=SC2046
-git submodule -q update --init $(awk '/path/ && !/llvm/ { print $3 }' .gitmodules)
+#export MACOSX_DEPLOYMENT_TARGET=12.0
+export CXX=clang++
+export CC=clang
+
+export BREW_PATH;
+BREW_PATH="$(brew --prefix)"
+export BREW_X64_PATH;
+BREW_X64_PATH="$("/usr/local/homebrew/bin/brew" --prefix)"
+export BREW_BIN="/usr/local/bin"
+export BREW_SBIN="/usr/local/sbin"
+export CMAKE_EXTRA_OPTS='-DLLVM_TARGETS_TO_BUILD=X86'
+
+export WORKDIR;
+WORKDIR="$(pwd)"
+
+#get Qt
+git clone https://github.com/engnr/qt-downloader.git
+cd qt-downloader
+git checkout f52efee0f18668c6d6de2dec0234b8c4bc54c597
+pip3 install pipenv
+pipenv run pip3 install py7zr requests semantic_version lxml
+pipenv run ./qt-downloader macos desktop 5.15.2 clang_64 --opensource
+cd ..
+
+export Qt5_DIR="$WORKDIR/qt-downloader/5.15.2/clang_64/lib/cmake/Qt5"
+export SDL2_DIR="$BREW_X64_PATH/opt/sdl2/lib/cmake/SDL2"
+
+export PATH="$BREW_PATH/opt/llvm@14/bin:$WORKDIR/qt-downloader/5.15.2/clang_64/bin:$BREW_BIN:$BREW_SBIN:/usr/bin:/bin:/usr/sbin:/sbin:/opt/X11/bin:/Library/Apple/usr/bin:$PATH"
+export LDFLAGS="-L$BREW_X64_PATH/lib -Wl,-rpath,$BREW_X64_PATH/lib"
+export CPPFLAGS="-I$BREW_X64_PATH/include -msse -msse2 -mcx16 -no-pie"
+export LIBRARY_PATH="$BREW_X64_PATH/lib"
+export LD_LIBRARY_PATH="$BREW_X64_PATH/lib"
+
+git submodule update --init --recursive --depth 1
+
+# 3rdparty fixes
+sed -i '' "s/extern const double NSAppKitVersionNumber;/const double NSAppKitVersionNumber = 1343;/g" 3rdparty/hidapi/hidapi/mac/hid.c

 mkdir build && cd build || exit 1
-cmake .. -DWITH_LLVM=OFF -DUSE_NATIVE_INSTRUCTIONS=OFF -G Ninja
-ninja
+
+"$BREW_X64_PATH/bin/cmake" .. \
+    -DUSE_SDL=ON -DUSE_DISCORD_RPC=OFF -DUSE_VULKAN=ON -DUSE_ALSA=OFF -DUSE_PULSE=OFF -DUSE_AUDIOUNIT=ON \
+    -DLLVM_CCACHE_BUILD=OFF -DLLVM_BUILD_RUNTIME=OFF -DLLVM_BUILD_TOOLS=OFF \
+    -DLLVM_INCLUDE_DOCS=OFF -DLLVM_INCLUDE_EXAMPLES=OFF -DLLVM_INCLUDE_TESTS=OFF -DLLVM_INCLUDE_TOOLS=OFF \
+    -DLLVM_INCLUDE_UTILS=OFF -DLLVM_USE_PERF=OFF -DLLVM_ENABLE_Z3_SOLVER=OFF \
+    -DUSE_NATIVE_INSTRUCTIONS=OFF \
+    -DUSE_SYSTEM_MVK=OFF \
+    $CMAKE_EXTRA_OPTS \
+    -DLLVM_TARGET_ARCH=X86_64 -DCMAKE_OSX_ARCHITECTURES=x86_64 -DCMAKE_IGNORE_PATH="$BREW_PATH/lib" \
+    -G Ninja
+
+"$BREW_PATH/bin/ninja"; build_status=$?;
+
+cd ..
+
+{   [ "$CI_HAS_ARTIFACTS" = "true" ];
+} && SHOULD_DEPLOY="true" || SHOULD_DEPLOY="false"
+
+if [ "$build_status" -eq 0 ] && [ "$SHOULD_DEPLOY" = "true" ]; then
+    .ci/deploy-mac.sh
+fi
@@ -12,13 +12,21 @@ if [ "$DEPLOY_APPIMAGE" = "true" ]; then
    ./squashfs-root/AppRun ./appdir/usr/share/applications/*.desktop -bundle-non-qt-libs
    ls ./appdir/usr/lib/
    rm -r ./appdir/usr/share/doc
-    rm ./appdir/usr/lib/libxcb*
    cp "$(readlink -f /lib/x86_64-linux-gnu/libnsl.so.1)" ./appdir/usr/lib/libnsl.so.1
-    export PATH=/rpcs3/build/squashfs-root/usr/bin/:${PATH}
+    export PATH=/rpcs3/build/squashfs-root/usr/bin/:"${PATH}"

    # Embed newer libstdc++ for distros that don't come with it (ubuntu 16.04)
    mkdir -p appdir/usr/optional/ ; mkdir -p appdir/usr/optional/libstdc++/
    cp /usr/lib/x86_64-linux-gnu/libstdc++.so.6 ./appdir/usr/optional/libstdc++/
+
+    # Remove libwayland-client because it has platform-dependent exports and breaks other OSes
+    rm -f ./appdir/usr/lib/libwayland-client.so*
+
+    # Remove libgmodule and libglib due to g_source_set_static_name symbol conflict with host libraries
+    rm -f ./appdir/usr/lib/libgmodule-2.0.so*
+    rm -f ./appdir/usr/lib/libglib-2.0.so*
+
+    # Install latest appimage runner
    rm ./appdir/AppRun
    curl -sL https://github.com/RPCS3/AppImageKit-checkrt/releases/download/continuous2/AppRun-patched-x86_64 -o ./appdir/AppRun
    chmod a+x ./appdir/AppRun
@@ -29,29 +37,31 @@ if [ "$DEPLOY_APPIMAGE" = "true" ]; then
    # This may need updating if you update the compiler or rpcs3 uses newer c++ features
    # See https://github.com/gcc-mirror/gcc/blob/master/libstdc%2B%2B-v3/config/abi/pre/gnu.ver
    # for which definitions correlate to which CXXABI version.
-    # Currently we target a minimum of GLIBCXX_3.4.26 and CXXABI_1.3.11
-    printf "#include <bits/stdc++.h>\nint main(){std::make_exception_ptr(0);std::pmr::get_default_resource();}" | $CXX -x c++ -std=c++2a -o ./appdir/usr/optional/checker -
+    # Currently we target a minimum of GLIBCXX_3.4.29 and CXXABI_1.3.11
+    printf "#include <sstream>\n#include <exception>\n#include <memory_resource>\nint main(){auto x = std::stringbuf();x.get_allocator();std::make_exception_ptr(0);std::pmr::get_default_resource();}" \
+     | $CXX -x c++ -std=c++2a -o ./appdir/usr/optional/checker -

    # Package it up and send it off
-    ./squashfs-root/usr/bin/appimagetool /rpcs3/build/appdir
+    ./squashfs-root/usr/bin/appimagetool "$APPDIR"
+
    ls

-    COMM_TAG="$(grep 'version{.*}' ../rpcs3/rpcs3_version.cpp | awk -F[\{,] '{printf "%d.%d.%d", $2, $3, $4}')"
+    COMM_TAG=$(awk '/version{.*}/ { printf("%d.%d.%d", $5, $6, $7) }' ../rpcs3/rpcs3_version.cpp)
    COMM_COUNT="$(git rev-list --count HEAD)"
    COMM_HASH="$(git rev-parse --short=8 HEAD)"
    RPCS3_APPIMAGE="rpcs3-v${COMM_TAG}-${COMM_COUNT}-${COMM_HASH}_linux64.AppImage"

    mv ./RPCS3*.AppImage "$RPCS3_APPIMAGE"

-    # If we're building using Azure Pipelines, let's copy over the AppImage artifact
+    # If we're building using a CI, let's copy over the AppImage artifact
    if [ -n "$BUILD_ARTIFACTSTAGINGDIRECTORY" ]; then
-        cp "$RPCS3_APPIMAGE" ~/artifacts
+        cp "$RPCS3_APPIMAGE" "$ARTDIR"
    fi

    FILESIZE=$(stat -c %s ./rpcs3*.AppImage)
    SHA256SUM=$(sha256sum ./rpcs3*.AppImage | awk '{ print $1 }')
-    echo "${SHA256SUM};${FILESIZE}B" > /rpcs3/GitHubReleaseMessage.txt
-    
+    echo "${SHA256SUM};${FILESIZE}B" > "$RELEASE_MESSAGE"
+
 fi

 if [ "$DEPLOY_PPA" = "true" ]; then
@@ -0,0 +1,58 @@
+#!/bin/sh -ex
+
+cd build || exit 1
+
+# Gather explicit version number and number of commits
+COMM_TAG=$(awk '/version{.*}/ { printf("%d.%d.%d", $5, $6, $7) }' ../rpcs3/rpcs3_version.cpp)
+COMM_COUNT=$(git rev-list --count HEAD)
+COMM_HASH=$(git rev-parse --short=8 HEAD)
+
+AVVER="${COMM_TAG}-${COMM_COUNT}"
+
+# AVVER is used for GitHub releases, it is the version number.
+echo "AVVER=$AVVER" >> ../.ci/ci-vars.env
+
+cd bin
+mkdir "rpcs3.app/Contents/lib/"
+
+cp "/usr/local/Homebrew/opt/llvm@14/lib/c++/libc++abi.1.0.dylib" "rpcs3.app/Contents/lib/libc++abi.1.dylib"
+
+rm -rf "rpcs3.app/Contents/Frameworks/QtPdf.framework" \
+"rpcs3.app/Contents/Frameworks/QtQml.framework" \
+"rpcs3.app/Contents/Frameworks/QtQmlModels.framework" \
+"rpcs3.app/Contents/Frameworks/QtQuick.framework" \
+"rpcs3.app/Contents/Frameworks/QtVirtualKeyboard.framework" \
+"rpcs3.app/Contents/Plugins/platforminputcontexts" \
+"rpcs3.app/Contents/Plugins/virtualkeyboard"
+
+# Need to do this rename hack due to case insensitive filesystem
+mv rpcs3.app RPCS3_.app
+mv RPCS3_.app RPCS3.app
+
+echo "[InternetShortcut]" > Quickstart.url
+echo "URL=https://rpcs3.net/quickstart" >> Quickstart.url
+echo "IconIndex=0" >> Quickstart.url
+
+DMG_FILEPATH="$BUILD_ARTIFACTSTAGINGDIRECTORY/rpcs3-v${COMM_TAG}-${COMM_COUNT}-${COMM_HASH}_macos.dmg"
+
+/opt/homebrew/bin/create-dmg --volname RPCS3 \
+--window-size 800 400 \
+--icon-size 100 \
+--icon rpcs3.app 200 190 \
+--add-file Quickstart.url Quickstart.url 400 20 \
+--hide-extension rpcs3.app \
+--hide-extension Quickstart.url \
+--app-drop-link 600 185 \
+--skip-jenkins \
+"$DMG_FILEPATH" \
+RPCS3.app
+
+/opt/homebrew/bin/7z a -mx9 rpcs3-v"${COMM_TAG}"-"${COMM_COUNT}"-"${COMM_HASH}"_macos.7z RPCS3.app
+
+FILESIZE=$(stat -f %z "$DMG_FILEPATH")
+SHA256SUM=$(shasum -a 256 "$DMG_FILEPATH" | awk '{ print $1 }')
+cd ..
+echo "${SHA256SUM};${FILESIZE}B" > "$RELEASE_MESSAGE"
+cd bin
+
+mv ./rpcs3*_macos.7z "$ARTDIR"
@@ -4,7 +4,7 @@
 ARTIFACT_DIR="$BUILD_ARTIFACTSTAGINGDIRECTORY"

 # Remove unecessary files
-rm -f ./bin/rpcs3.exp ./bin/rpcs3.lib ./bin/rpcs3.pdb
+rm -f ./bin/rpcs3.exp ./bin/rpcs3.lib ./bin/rpcs3.pdb ./bin/vc_redist.x64.exe

 # Prepare compatibility database for packaging, as well as
 # certificate for ssl (auto-updater)
@@ -20,5 +20,5 @@ sha256sum "$BUILD" | awk '{ print $1 }' | tee "$BUILD.sha256"
 echo "$(cat "$BUILD.sha256");$(stat -c %s "$BUILD")B" > GitHubReleaseMessage.txt

 # Move files to publishing directory
-mv -- "$BUILD" "$ARTIFACT_DIR"
-mv -- "$BUILD.sha256" "$ARTIFACT_DIR"
+cp -- "$BUILD" "$ARTIFACT_DIR"
+cp -- "$BUILD.sha256" "$ARTIFACT_DIR"
@@ -1,10 +1,13 @@
 # Variables set by Azure Pipelines
-IS_AZURE
+CI_HAS_ARTIFACTS
 BUILD_REASON
 BUILD_SOURCEVERSION
 BUILD_ARTIFACTSTAGINGDIRECTORY
 BUILD_REPOSITORY_NAME
 BUILD_SOURCEBRANCHNAME
+APPDIR
+ARTDIR
+RELEASE_MESSAGE
 # Variables for build matrix
 COMPILER
 DEPLOY_APPIMAGE
@@ -10,4 +10,4 @@ while IFS='=' read -r key val; do
    # Skip over lines containing comments.
    [ "${key##\#*}" ] || continue
    echo "##vso[task.setvariable variable=$key]$val"
-done < ".ci/azure-vars.env"
+done < ".ci/ci-vars.env"
@@ -0,0 +1,13 @@
+#!/bin/sh -e
+
+# Export variables for later stages of the Cirrus pipeline
+# Values done in this manner will appear as environment variables
+# in later stages.
+
+# From pure-sh-bible
+# Setting 'IFS' tells 'read' where to split the string.
+while IFS='=' read -r key val; do
+    # Skip over lines containing comments.
+    [ "${key##\#*}" ] || continue
+    export "$key"="$val"
+done < ".ci/ci-vars.env"
@@ -24,7 +24,7 @@ curl -s \
 cat release.json
 id=$(grep '"id"' release.json | cut -d ':' -f2 | head -n1 | awk '{$1=$1;print}')
 id=${id%?}
-echo ${id:?}
+echo "${id:?}"

 upload_file()
 {
@@ -9,10 +9,10 @@ export ASSUME_ALWAYS_YES=true
 pkg info # debug

 # Prefer newer Clang than in base system (see also .ci/build-freebsd.sh)
-pkg install llvm11
+pkg install llvm15

 # Mandatory dependencies (qt5-dbus and qt5-gui are pulled via qt5-widgets)
-pkg install git ccache cmake ninja qt5-qmake qt5-buildtools qt5-widgets qt5-concurrent glew openal-soft ffmpeg
+pkg install git ccache cmake ninja qt5-qmake qt5-buildtools qt5-widgets qt5-concurrent qt5-multimedia qt5-svg glew openal-soft ffmpeg

 # Optional dependencies (libevdev is pulled by qt5-gui)
 pkg install pkgconf alsa-lib pulseaudio sdl2 evdev-proto vulkan-headers vulkan-loader
@@ -7,25 +7,29 @@ PR_NUMBER="$SYSTEM_PULLREQUEST_PULLREQUESTID"

 # Resource/dependency URLs
 # Qt mirrors can be volatile and slow, so we list 2
-QT_HOST="http://mirrors.ocf.berkeley.edu/qt/"
-#QT_HOST="http://qt.mirror.constant.com/"
+#QT_HOST="http://mirrors.ocf.berkeley.edu/qt/"
+QT_HOST="http://qt.mirror.constant.com/"
 QT_URL_VER=$(echo "$QT_VER" | sed "s/\.//g")
-QT_VER_MSVC_UP=$(echo ${QT_VER_MSVC} | tr '[:lower:]' '[:upper:]')
+QT_VER_MSVC_UP=$(echo "${QT_VER_MSVC}" | tr '[:lower:]' '[:upper:]')
 QT_PREFIX="online/qtsdkrepository/windows_x86/desktop/qt${QT_VER_MAIN}_${QT_URL_VER}/qt.qt${QT_VER_MAIN}.${QT_URL_VER}.win64_${QT_VER_MSVC}_64/${QT_VER}-0-${QT_DATE}"
 QT_SUFFIX="-Windows-Windows_10-${QT_VER_MSVC_UP}-Windows-Windows_10-X86_64.7z"
 QT_BASE_URL="${QT_HOST}${QT_PREFIX}qtbase${QT_SUFFIX}"
 QT_WINE_URL="${QT_HOST}${QT_PREFIX}qtwinextras${QT_SUFFIX}"
 QT_DECL_URL="${QT_HOST}${QT_PREFIX}qtdeclarative${QT_SUFFIX}"
 QT_TOOL_URL="${QT_HOST}${QT_PREFIX}qttools${QT_SUFFIX}"
+QT_MM_URL="${QT_HOST}${QT_PREFIX}qtmultimedia${QT_SUFFIX}"
+QT_SVG_URL="${QT_HOST}${QT_PREFIX}qtsvg${QT_SUFFIX}"
 LLVMLIBS_URL='https://github.com/RPCS3/llvm-mirror/releases/download/custom-build-win/llvmlibs_mt.7z'
 GLSLANG_URL='https://github.com/RPCS3/glslang/releases/download/custom-build-win/glslanglibs_mt.7z'
-VULKAN_SDK_URL="https://www.dropbox.com/s/kn005de7gsv5whe/VulkanSDK-1.2.154.1-Installer.exe"
+VULKAN_SDK_URL="https://www.dropbox.com/s/cs77c3iv5mbo0bt/VulkanSDK-${VULKAN_VER}-Installer.exe"

 DEP_URLS="         \
    $QT_BASE_URL   \
    $QT_WINE_URL   \
    $QT_DECL_URL   \
    $QT_TOOL_URL   \
+    $QT_MM_URL     \
+    $QT_SVG_URL    \
    $LLVMLIBS_URL  \
    $GLSLANG_URL   \
    $VULKAN_SDK_URL"
@@ -36,7 +40,7 @@ DEP_URLS="         \
 # Pull all the submodules except llvm, since it is built separately and we just download that build
 # Note: Tried to use git submodule status, but it takes over 20 seconds
 # shellcheck disable=SC2046
-git submodule -q update --init --depth 1 $(awk '/path/ && !/llvm/ { print $3 }' .gitmodules)
+git submodule -q update --init --depth=1 --jobs=8 $(awk '/path/ && !/llvm/ { print $3 }' .gitmodules)

 # Git bash doesn't have rev, so here it is
 rev()
@@ -76,13 +80,13 @@ for url in $DEP_URLS; do
    case "$url" in
    *qt*) checksum=$(curl -L "${url}.sha1"); algo="sha1"; outDir='C:\Qt\' ;;
    *llvm*) checksum=$(curl -L "${url}.sha256"); algo="sha256"; outDir="." ;;
-    *glslang*) checksum=$(curl -L "${url}.sha256"); algo="sha256"; outDir="./lib/Release - LLVM-x64" ;;
+    *glslang*) checksum=$(curl -L "${url}.sha256"); algo="sha256"; outDir="./lib/Release-x64" ;;
    *Vulkan*)
        # Vulkan setup needs to be run in batch environment
        # Need to subshell this or else it doesn't wait
        download_and_verify "$url" "$VULKAN_SDK_SHA" "sha256" "$fileName"
        cp "$CACHE_DIR/$fileName" .
-        _=$(echo "$fileName /S" | cmd)
+        _=$(echo "$fileName --accept-licenses --default-answer --confirm-command install" | cmd)
        continue
    ;;
    *) echo "Unknown url resource: $url"; exit 1 ;;
@@ -110,6 +114,6 @@ fi
 # BUILD is the name of the release artifact
 # AVVER is used for GitHub releases, it is the version number.
 BRANCH="${REPO_NAME}/${REPO_BRANCH}"
-echo "BRANCH=$BRANCH" > .ci/azure-vars.env
-echo "BUILD=$BUILD" >> .ci/azure-vars.env
-echo "AVVER=$AVVER" >> .ci/azure-vars.env
+echo "BRANCH=$BRANCH" > .ci/ci-vars.env
+echo "BUILD=$BUILD" >> .ci/ci-vars.env
+echo "AVVER=$AVVER" >> .ci/ci-vars.env
@@ -1,11 +1,114 @@
 env:
  CIRRUS_CLONE_DEPTH: 0 # Unshallow clone to obtain proper GIT_VERSION
+  BUILD_REPOSITORY_NAME: $CIRRUS_REPO_FULL_NAME
+  SYSTEM_PULLREQUEST_SOURCEBRANCH: $CIRRUS_BRANCH
+  SYSTEM_PULLREQUEST_PULLREQUESTID: $CIRRUS_PR
+  BUILD_SOURCEVERSION: $CIRRUS_CHANGE_IN_REPO
+  BUILD_SOURCEBRANCHNAME: $CIRRUS_BRANCH
+  RPCS3_TOKEN: ENCRYPTED[!a4c3850e29ab150692286a74bec29819d25971a7ec431b86de2a35f7ed90c5b2ab3c93469f9298e30924d843599110e9!]
+
+windows_task:
+  matrix:
+    - name: Cirrus Windows
+      windows_container:
+        image: cirrusci/windowsservercore:visualstudio2019
+        cpu: 8
+        memory: 16G
+  env:
+    CIRRUS_SHELL: "bash"
+    COMPILER: msvc
+    QT_VER_MAIN: '5'
+    BUILD_ARTIFACTSTAGINGDIRECTORY: ${CIRRUS_WORKING_DIR}\artifacts\
+    QT_VER: '5.15.2'
+    QT_VER_MSVC: 'msvc2019'
+    QT_DATE: '202011130602'
+    QTDIR: C:\Qt\${QT_VER}\${QT_VER_MSVC}_64
+    VULKAN_VER: '1.3.224.1'
+    VULKAN_SDK_SHA: '2029e652e39ee6a6036cff3765da31e1e6c595fd2413d3cd111dfab7855621ea'
+    VULKAN_SDK: C:\VulkanSDK\${VULKAN_VER}
+    CACHE_DIR: "./cache"
+    UPLOAD_COMMIT_HASH: 7d09e3be30805911226241afbb14f8cdc2eb054e
+    UPLOAD_REPO_FULL_NAME: "rpcs3/rpcs3-binaries-win"
+  deps_cache:
+    folder: "./cache"
+  #obj_cache:
+  #  folder: "./tmp"
+  #obj2_cache:
+  #  folder: "./rpcs3/x64"
+  setup_script:
+    - './.ci/get_keys-windows.sh'
+    - './.ci/setup-windows.sh'
+#   - choco install -y python # Needed for SPIRV, use either this or make a new Docker image
+#  spirv_script:
+#   - export PATH=${PATH}:"C:\Program Files (x86)\Microsoft Visual Studio\2019\BuildTools\MSBuild\Current\Bin"
+#   - cd "${CIRRUS_WORKING_DIR}/3rdparty/SPIRV"
+#   - msbuild.exe spirv.vcxproj //p:Configuration=Release //m
+  rpcs3_script:
+    - export PATH=${PATH}:"C:\Program Files (x86)\Microsoft Visual Studio\2019\BuildTools\MSBuild\Current\Bin"
+    - msbuild.exe rpcs3.sln //p:Configuration=Release //m
+  deploy_script:
+    - mkdir artifacts
+    - source './.ci/export-cirrus-vars.sh'
+    - './.ci/deploy-windows.sh'
+  artifacts:
+    name: Artifact
+    path: "*.7z*"
+  push_script: |
+    if [ "$CIRRUS_REPO_OWNER" = "RPCS3" ] && [ -z "$CIRRUS_PR" ] && [ "$CIRRUS_BRANCH" = "master" ]; then
+      source './.ci/export-cirrus-vars.sh'
+      './.ci/github-upload.sh'
+    fi;
+
+linux_task:
+  container:
+    image: rpcs3/rpcs3-ci-bionic:1.6
+    cpu: 4
+    memory: 16G
+  env:
+    BUILD_ARTIFACTSTAGINGDIRECTORY: ${CIRRUS_WORKING_DIR}/artifacts
+    ARTDIR: ${CIRRUS_WORKING_DIR}/artifacts/
+    CCACHE_DIR: "/tmp/ccache_dir"
+    CCACHE_MAXSIZE: 300M
+    CI_HAS_ARTIFACTS: true
+    UPLOAD_COMMIT_HASH: d812f1254a1157c80fd402f94446310560f54e5f
+    UPLOAD_REPO_FULL_NAME: "rpcs3/rpcs3-binaries-linux"
+    DEPLOY_APPIMAGE: true
+    APPDIR: "./appdir"
+    RELEASE_MESSAGE: "../GitHubReleaseMessage.txt"
+  ccache_cache:
+    folder: "/tmp/ccache_dir"
+  matrix:
+    - name: Cirrus Linux GCC
+      env:
+        COMPILER: gcc
+      gcc_script:
+        - mkdir artifacts
+        - ".ci/build-linux.sh"
+    - name: Cirrus Linux Clang
+      env:
+        COMPILER: clang
+      clang_script:
+        - mkdir artifacts
+        - ".ci/build-linux.sh"
+  artifacts:
+    name: Artifact
+    path: "artifacts/*"
+  push_script: |
+    if [ "$CIRRUS_REPO_OWNER" = "RPCS3" ] && [ -z "$CIRRUS_PR" ] && [ "$CIRRUS_BRANCH" = "master" ] && [ "$COMPILER" = "gcc" ]; then
+      COMM_TAG=$(awk '/version{.*}/ { printf("%d.%d.%d", $5, $6, $7) }' ./rpcs3/rpcs3_version.cpp)
+      COMM_COUNT=$(git rev-list --count HEAD)
+      COMM_HASH=$(git rev-parse --short=8 HEAD)
+
+      export AVVER="${COMM_TAG}-${COMM_COUNT}"
+
+      .ci/github-upload.sh
+    fi;

 freebsd_task:
  matrix:
-    - name: FreeBSD 12.2
+    - name: Cirrus FreeBSD
      freebsd_instance:
-        image_family: freebsd-12-2
+        image_family: freebsd-13-1
        cpu: 8
        memory: 8G
  env:
@@ -13,7 +116,36 @@ freebsd_task:
    CCACHE_DIR: /tmp/ccache_dir
  ccache_cache:
    folder: /tmp/ccache_dir
-  packages_cache:
-    folder: /var/cache/pkg
  install_script: "sh -ex ./.ci/install-freebsd.sh"
  script: "./.ci/build-freebsd.sh"
+
+macos_task:
+  timeout_in: 12000m
+  ccache_cache:
+    folder: /tmp/ccache_dir
+  matrix:
+    - name: Cirrus macOS
+      macos_instance:
+        image: ghcr.io/cirruslabs/macos-monterey-xcode:14
+      mac_script:
+        - mkdir artifacts
+        - chmod +x ".ci/build-mac.sh"
+        - chmod +x ".ci/deploy-mac.sh"
+        - ".ci/build-mac.sh"
+  env:
+    BUILD_ARTIFACTSTAGINGDIRECTORY: ${CIRRUS_WORKING_DIR}/artifacts
+    ARTDIR: ${CIRRUS_WORKING_DIR}/artifacts/
+    CCACHE_DIR: "/tmp/ccache_dir"
+    CCACHE_MAXSIZE: 300M
+    CI_HAS_ARTIFACTS: true
+    UPLOAD_COMMIT_HASH: 51ae32f468089a8169aaf1567de355ff4a3e0842
+    UPLOAD_REPO_FULL_NAME: "rpcs3/rpcs3-binaries-mac"
+    RELEASE_MESSAGE: "../GitHubReleaseMessage.txt"
+  artifacts:
+    name: Artifact
+    path: "artifacts/*"
+  push_script: |
+    if [ "$CIRRUS_REPO_OWNER" = "RPCS3" ] && [ -z "$CIRRUS_PR" ] && [ "$CIRRUS_BRANCH" = "master" ]; then
+      source './.ci/export-cirrus-vars.sh'
+      .ci/github-upload.sh
+    fi;
@@ -1,5 +1,5 @@
 Standard: c++20
-UseTab: ForIndentation
+UseTab: AlignWithSpaces
 TabWidth: 4
 IndentWidth: 4
 AccessModifierOffset: -4
@@ -16,11 +16,12 @@ AllowShortBlocksOnASingleLine: Never
 AllowShortCaseLabelsOnASingleLine: true
 AllowShortFunctionsOnASingleLine: Empty
 AllowShortLoopsOnASingleLine: false
+AllowShortLambdasOnASingleLine: Empty
 Cpp11BracedListStyle: true
 IndentCaseLabels: false
 SortIncludes: false
 ReflowComments: true
-AlignConsecutiveAssignments: true
+AlignConsecutiveAssignments: false
 AlignTrailingComments: true
 AlignAfterOpenBracket: DontAlign
 ConstructorInitializerAllOnOneLineOrOnePerLine: false
@@ -1,6 +1,6 @@
 # Getting Started

-Before getting started using the emulator, read the [Quickstart Guide](https://rpcs3.net/quickstart) and the [FAQ](https://rpcs3.net/faq). After reading those, if you need support, visit our [Forums](https://forums.rpcs3.net). 
+Before getting started using the emulator, read the [Quickstart Guide](https://rpcs3.net/quickstart). After reading it, if you need support, check out [How to ask for Support](https://github.com/RPCS3/rpcs3/wiki/How-to-ask-for-Support).

 # Issue Reporting

@@ -8,13 +8,12 @@ Before getting started using the emulator, read the [Quickstart Guide](https://r

 **Before reporting an issue:**
 - Check if your system matches all the minimum requirements listed in the [Quickstart Guide](https://rpcs3.net/quickstart);
- Check if the issue is meaningful for the team (e.g. The Last of Us doesn't work is obvious and therefore useless);
- Search older issues/forum threads to see if your issue was already submitted.
- Use understandable English. It doesn't need to be perfect, but clear enough to understand your message.
- While reporting issues, don't forget to include details about your system (OS, CPU, GPU, etc.), as well as the RPCS3.log file.
+- Search older issues/forum threads to see if your issue was already submitted;
+- Use understandable English. It doesn't need to be perfect, but clear enough to understand your message;
+- While reporting issues, please follow the template for the type of issue you've selected (Regression Report, Bug Report or Feature Request), which is prefilled on the issue's textbox.

 Submitting your test results for Commercial Games must be done on our forums. Please read the [Game Compatibility](https://github.com/RPCS3/rpcs3/wiki/Game-Compatibility) wiki page before doing so.

 # Contributing

-Check the [Coding Style Guidelines](https://github.com/RPCS3/rpcs3/wiki/Coding-Style), [Roadmap](https://github.com/RPCS3/rpcs3/wiki/Roadmap) and [Developer Information](https://github.com/RPCS3/rpcs3/wiki/Developer-Information). If you have any questions, hit us up on our [Discord Server](https://discord.me/RPCS3) in the **#development** channel.
+Check the [Coding Style Guidelines](https://github.com/RPCS3/rpcs3/wiki/Coding-Style) and [Developer Information](https://github.com/RPCS3/rpcs3/wiki/Developer-Information). If you have any questions, hit us up on our [Discord Server](https://discord.me/RPCS3) in the **#development** channel.
@@ -1,62 +0,0 @@
---
-name: Regression report
-about: If something used to work before, but now it doesn't
-title: ''
-labels: ''
-assignees: ''
-
---
-
-## Please do not ask for help or report compatibility regressions here, use [RPCS3 Discord server](https://discord.me/RPCS3) or [forums](https://forums.rpcs3.net/) instead.
-
-## Quick summary
-Please briefly describe what has stopped working correctly.
-
-## Details
-Please describe the regression as accurately as possible.
-
-#### 1. Please provide _exact_ build (or commit) information that introduced the regression you're reporting.
-* Please see [How to find the build that caused a regression](https://wiki.rpcs3.net/index.php?title=Help:Using_different_versions_of_RPCS3#How_to_find_the_build_that_caused_a_regression.3F) in our wiki.
-* You can find [History of RPCS3 builds](https://rpcs3.net/compatibility?b) here.
-
-#### 2. Please attach TWO logs:
-* One for the build with regression.
-* One for the build that works as expected.
-
-##### How to obtain RPCS3's log:
-* Run the game until you find the regression.
-* Completely close RPCS3, or move to the next step in case it has crashed.
-* Locate RPCS3's log file:
-	+ ```RPCS3.log``` (It can show up as just RPCS3 and have a notepad icon)
-	or
-	+ ```RPCS3.log.gz``` (It can show up as RPCS3.log and have a zip or rar icon)
-![image](https://user-images.githubusercontent.com/44116740/84433247-aa15fa80-ac36-11ea-913e-6fe25a1d040e.png)
-	On Linux it will be in ```~/.cache/rpcs3/```.
-	On Windows it will be near the executable.
-* Attach the log:
-	+ Drag and drop your log file into the issue.
-	+ Or upload it to the cloud, such as Dropbox, Mega etc. 
-
-#### 3. If you describe graphical regression, please provide an RSX capture and a RenderDoc capture that demonstrate it.
-* To create an RSX capture, use _Create_ _RSX_ _Capture_ under _Utilities_.
-Captures will be stored in RPCS3 folder → captures.
-	+ Compress your capture with 7z, Rar etc. 
-	And drag and drop it into the issue.
-	+ Or upload it to the cloud, such as Dropbox, Mega etc.
-* To create a RenderDoc capture, please refer to [RenderDoc's documentation](https://renderdoc.org/docs/how/how_capture_frame.html).
-
-#### 4. Please attach screenshots of your problem.
-* Enable performance overlay with at least medium level of detail.
-You can find it in _Emulator_ tab in _Settings_.
-
-#### 5. Please provide comparison with real PS3.
-
-#### 6. Please provide your system configuration:
-* OS
-* CPU
-* GPU
-* Driver version
-* etc.
-
-#### Please include.
-* Anything else you deem to be important
@@ -0,0 +1,93 @@
+name: Regression report
+description: If something used to work before, but now it doesn't
+title: "[Regression] Enter a title here"
+labels: []
+body:
+  - type: markdown
+    attributes:
+      value: |
+        # Summary
+        Please do not ask for help or report compatibility regressions here, use [RPCS3 Discord server](https://discord.me/RPCS3) or [forums](https://forums.rpcs3.net/) instead.
+  - type: textarea
+    id: quick-summary
+    attributes:
+      label: Quick summary
+      description: Please briefly describe what has stopped working correctly.
+    validations:
+      required: true
+  - type: textarea
+    id: details
+    attributes:
+      label: Details
+      description: Please describe the regression as accurately as possible. Include screenshots if neccessary.
+    validations:
+      required: false
+  - type: markdown
+    attributes:
+      value: |
+        # Identify the regressed build
+        Please provide the _exact_ build (or commit) information that introduced the regression you're reporting.
+        * Please see [How to find the build that caused a regression](https://wiki.rpcs3.net/index.php?title=Help:Using_different_versions_of_RPCS3#How_to_find_the_build_that_caused_a_regression.3F) in our wiki.
+        * You can find [History of RPCS3 builds](https://rpcs3.net/compatibility?b) here.
+
+        Make sure you're running with settings as close to default as possible
+        * **Do NOT enable any emulator game patches when reporting issues**
+        * Only change settings that are required for the game to work
+  - type: input
+    id: regressed-build
+    attributes:
+      label: Build with regression
+      description: Provide _exact_ build (or commit) information that introduced the regression you're reporting.
+      placeholder: v0.0.23-13948-31df99f7
+    validations:
+      required: true
+  - type: markdown
+    attributes:
+      value: |
+        # Log files
+        Obtaining the log file:
+        * Run the game until you find the regression.
+        * Completely close RPCS3 and locate the log file.
+
+        RPCS3's Log file will be ```RPCS3.log.gz``` (sometimes shows as RPCS3.log with zip icon) or ```RPCS3.log``` (sometimes shows as RPCS3 wtih notepad icon).
+        * On Windows it will be in the RPCS3 directory near the executable
+        * On Linux it will be in ```~/.cache/rpcs3/```
+        * On MacOS it will be in ```~/Library/Caches/rpcs3```. If you're unable to locate it copy paste the path in Spotlight and hit enter.
+  - type: textarea
+    id: logs
+    attributes:
+      label: Attach two log files
+      description: |
+        Attach one file for the build with regression and another for a build that works.
+        Drag & drop the files into this input field, or upload them to another service (f.ex. Dropbox, Mega) and provide a link.
+    validations:
+      required: true
+  - type: markdown
+    attributes:
+      value: |
+        # Other details
+        If you describe a graphical regression, please provide an RSX capture and/or a RenderDoc capture that demonstrate it.
+        * To create an RSX capture, use _Create_ _RSX_ _Capture_ under _Utilities_.
+        * Captures will be stored in RPCS3 folder → captures.
+        * To create a RenderDoc capture, please refer to [RenderDoc's documentation](https://renderdoc.org/docs/how/how_capture_frame.html).
+  - type: textarea
+    id: captures
+    attributes:
+      label: Attach capture files for visual issues
+      description: Compress your capture with 7z, Rar etc. and attach it here, or upload it to the cloud (Dropbox, Mega etc) and add a link to it.
+    validations:
+      required: false
+  - type: textarea
+    id: system-info
+    attributes:
+      label: System configuration
+      description: Provide information about your system, such as operating system, CPU and GPU model, GPU driver version and other information that describes your system configuration.
+    validations:
+      required: false
+  - type: textarea
+    id: other-details
+    attributes:
+      label: Other details
+      description: Include anything else you deem to be important.
+    validations:
+      required: false
@@ -1,55 +0,0 @@
---
-name: Bug report
-about: If something doesn't work correctly in RPCS3
-title: ''
-labels: ''
-assignees: ''
-
---
-
-## Please do not ask for help or report compatibility regressions here, use [RPCS3 Discord server](https://discord.me/RPCS3) or [forums](https://forums.rpcs3.net/) instead.
-
-## Quick summary
-Please briefly describe what is not working correctly.
-
-## Details
-
-Please describe the problem as accurately as possible.
-
-#### 1. Please attach RPCS3's log.
-* Run the game until you find the issue.
-* Completely close RPCS3, or move to the next step in case it has crashed.
-* Locate RPCS3's log file:
-	+ ```RPCS3.log``` (It can show up as just RPCS3 and have a notepad icon)
-	or
-	+ ```RPCS3.log.gz``` (It can show up as RPCS3.log and have a zip or rar icon)
-![image](https://user-images.githubusercontent.com/44116740/84433247-aa15fa80-ac36-11ea-913e-6fe25a1d040e.png)
-	On Linux it will be in ```~/.cache/rpcs3/```.
-	On Windows it will be near the executable.
-* Attach the log:
-	+ Drag and drop your log file into the issue.
-	+ Or upload it to the cloud, such as Dropbox, Mega etc. 
-
-#### 2. If you describe graphical issue, please provide an RSX capture and a RenderDoc capture that demonstrate it.
-* To create an RSX capture, use _Create_ _RSX_ _Capture_ under _Utilities_.
-Captures will be stored in RPCS3 folder → captures.
-	+ Compress your capture with 7z, Rar etc. 
-	And	drag and drop it into the issue.
-	+ Or upload it to the cloud, such as Dropbox, Mega etc.
-* To create a RenderDoc capture, please refer to [RenderDoc's documentation](https://renderdoc.org/docs/how/how_capture_frame.html).
-
-#### 3. Please attach screenshots of your problem.
-* Enable performance overlay with at least medium level of detail.
-You can find it in _Emulator_ tab in _Settings_.
-
-#### 4. Please provide comparison with real PS3.
-
-#### 5. Please provide your system configuration:
-* OS
-* CPU
-* GPU
-* Driver version
-* etc.
-
-#### Please include.
-* Anything else you deem to be important
@@ -0,0 +1,78 @@
+name: Bug report
+description: If something doesn't work correctly in RPCS3
+title: "Enter a title here"
+labels: []
+body:
+  - type: markdown
+    attributes:
+      value: |
+        # Summary
+        Please do not ask for help or report compatibility regressions here, use [RPCS3 Discord server](https://discord.me/RPCS3) or [forums](https://forums.rpcs3.net/) instead.
+  - type: textarea
+    id: quick-summary
+    attributes:
+      label: Quick summary
+      description: Please briefly describe what is not working correctly.
+    validations:
+      required: true
+  - type: textarea
+    id: details
+    attributes:
+      label: Details
+      description: |
+        Please describe the problem as accurately as possible.
+        Provide a comparison with a real PS3, if possible.
+    validations:
+      required: false
+  - type: markdown
+    attributes:
+      value: |
+        # Log files
+
+        Provide a log file that includes the bug you're reporting.
+
+        Obtaining the log file:
+        * Run the game until you find the regression.
+        * Completely close RPCS3 and locate the log file.
+
+        RPCS3's Log file will be ```RPCS3.log.gz``` (sometimes shows as RPCS3.log with zip icon) or ```RPCS3.log``` (sometimes shows as RPCS3 wtih notepad icon).
+        * On Windows it will be in the RPCS3 directory near the executable
+        * On Linux it will be in ```~/.cache/rpcs3/```
+        * On MacOS it will be in ```~/Library/Caches/rpcs3```. If you're unable to locate it copy paste the path in Spotlight and hit enter.
+  - type: textarea
+    id: logs
+    attributes:
+      label: Attach a log file
+      description: |
+        Drag & drop the files into this input field, or upload them to another service (f.ex. Dropbox, Mega) and provide a link.
+    validations:
+      required: true
+  - type: markdown
+    attributes:
+      value: |
+        # Other details
+        If you describe a graphical issue, please provide an RSX capture and/or a RenderDoc capture that demonstrate it.
+        * To create an RSX capture, use _Create_ _RSX_ _Capture_ under _Utilities_.
+        * Captures will be stored in RPCS3 folder → captures.
+        * To create a RenderDoc capture, please refer to [RenderDoc's documentation](https://renderdoc.org/docs/how/how_capture_frame.html).
+  - type: textarea
+    id: captures
+    attributes:
+      label: Attach capture files for visual issues
+      description: Compress your capture with 7z, Rar etc. and attach it here, or upload it to the cloud (Dropbox, Mega etc) and add a link to it.
+    validations:
+      required: false
+  - type: textarea
+    id: system-info
+    attributes:
+      label: System configuration
+      description: Provide information about your system, such as operating system, CPU and GPU model, GPU driver version and other information that describes your system configuration.
+    validations:
+      required: false
+  - type: textarea
+    id: other-details
+    attributes:
+      label: Other details
+      description: Include anything else you deem to be important.
+    validations:
+      required: false
@@ -1,41 +0,0 @@
---
-name: Feature Request
-about: If RPCS3 lacks a feature you would like to see
-title: '[Feature Request] Enter Feature Title Here'
-labels: ''
-assignees: ''
-
---
-
-## Please do not ask for help or report compatibility regressions here, use [RPCS3 Discord server](https://discord.me/RPCS3) or [forums](https://forums.rpcs3.net/) instead.
-
-## Quick summary
-Please briefly describe what feature you would like RPCS3 to have.
-
-## Details
-
-Please describe the feature as accurately as possible.
-
-#### 1. Please describe, what part of RPCS3 would be affected by your feature:
-* Gameplay
-* Debugging
-* UI
-* Patches
-* Installation
-* Github
-* CI
-* etc.
-
-#### 2. Please tell us, why your feature is important to RPCS3.
-
-#### 3. Please attach screenshots of the feature implemented in other projects.
-
-#### 4. Please provide your system configuration:
-* OS
-* CPU
-* GPU
-* Driver version
-* etc.
-
-#### Please include.
-* Anything else you deem to be important
@@ -0,0 +1,36 @@
+name: Feature request
+description: If RPCS3 lacks a feature you would like to see
+title: "[Feature request] Enter a title here"
+labels: []
+body:
+  - type: markdown
+    attributes:
+      value: |
+        Please do not ask for help or report compatibility regressions here, use [RPCS3 Discord server](https://discord.me/RPCS3) or [forums](https://forums.rpcs3.net/) instead.
+  - type: textarea
+    id: quick-summary
+    attributes:
+      label: Quick summary
+      description: Please briefly describe what feature you would like RPCS3 to have.
+    validations:
+      required: true
+  - type: textarea
+    id: details
+    attributes:
+      label: Details
+      description: Please describe the feature as accurately as possible.
+    validations:
+      required: false
+  - type: markdown
+    attributes:
+      value: |
+        Please include the following information:
+        * What part of RPCS3 would be affected by your feature? (Gameplay, Debugging, UI, Patches, Installation, CI etc.)
+        * Why your feature is important to RPCS3.
+        * If the feature is implemented in other projects, attach screenshots.
+        * If this feature is something that a game is trying to use, upload a log file for it.
+
+        RPCS3's Log file will be ```RPCS3.log.gz``` (sometimes shows as RPCS3.log with zip icon) or ```RPCS3.log``` (sometimes shows as RPCS3 wtih notepad icon).
+        * On Windows it will be in the RPCS3 directory near the executable
+        * On Linux it will be in ```~/.cache/rpcs3/```
+        * On MacOS it will be in ```~/Library/Caches/rpcs3```. If you're unable to locate it copy paste the path in Spotlight and hit enter.
@@ -0,0 +1,14 @@
+---
+name: Advanced
+about: For developers and experienced users only
+title: ''
+labels: ''
+assignees: ''
+
+---
+
+## Please do not ask for help or report compatibility regressions here, use [RPCS3 Discord server](https://discord.me/RPCS3) or [forums](https://forums.rpcs3.net/) instead.
+
+You're using the advanced template. You're expected to know what to write in order to fill in all the required information for proper report.
+
+If you're unsure on what to do, please return back to the issue type selection and choose different category.
@@ -1,20 +1,18 @@
 ## How to test a PR build

-Please take into account, that RPCS3 build usually takes some time (about 20 min), so you can't access a build, if a PR was just submitted.
+Please take into account, that RPCS3 build usually takes some time (about 15 mins), so you can't access a build if a PR was just submitted.

 - Open a PR you want to test
- Scroll to the very bottom and locate "Checks" section
- Expand all checks
+- Scroll to the very bottom and locate the **Checks** section
+- Click on **Show all checks**
 	You are supposed to see something like this
-	![image](https://user-images.githubusercontent.com/44116740/102790039-f98f6d00-43b5-11eb-9134-2359b732508d.png)
- Click on **Linux_Build** or **Windows_Build** __Details__
- Click **View more details on Azure Pipelines** at the very bottom
-	![image](https://user-images.githubusercontent.com/44116740/102790281-45daad00-43b6-11eb-90e7-0fa8c37c31c6.png)
- Click on *1 artifact produced*
-	![image](https://user-images.githubusercontent.com/44116740/102790526-9a7e2800-43b6-11eb-9925-be1ea1c8bdad.png)
- Select a build for your OS and click three dots on the right, Click **Download artifacts**
-	![image](https://user-images.githubusercontent.com/44116740/102790636-c9949980-43b6-11eb-9692-1e3ba567b9be.png)
-	
- Congratulations! You are now downloading RPCS3 build for specific PR.
+	![image](https://user-images.githubusercontent.com/10283761/116630952-2cd99e00-a94c-11eb-933e-986d6020ca92.png)
+- Click on __Details__ on either **Cirrus Linux GCC** or **Cirrus Windows**
+- Click **View more details on Cirrus CI** at the very bottom
+	![image](https://user-images.githubusercontent.com/10283761/116631111-5e526980-a94c-11eb-95f7-751e6f15e1ea.png)
+- Click on the download button for **Artifact** on the **Artifacts** block
+	![image](https://user-images.githubusercontent.com/10283761/116631322-bee1a680-a94c-11eb-89a3-be365783582e.png)

-Please note, that PR builds are not supposed to be stable, because they contain new changesets. 
+- Congratulations! You are now downloading an RPCS3 build for that specific PR.
+
+__Please note that PR builds are not supposed to be stable because they contain new changesets.__
@@ -9,7 +9,7 @@
 # Compiled Static libraries
 *.lai
 *.la
-# *.a # Commented out since OpenAL Soft's binaries use this extension.
+*.a
 *.opensdf
 *.sdf
 *.suo
@@ -37,14 +37,17 @@
 /rpcs3/Debug
 /rpcs3/Release
 /llvm_build
-/3rdparty/llvm_build
-/Vulkan/Vulkan-build
-/Vulkan/glslang-build
-/Vulkan/spirv-tools-build

 !/bin
 /bin/*

+# Audio DLLs
+!/bin/soft_oal.dll
+!/bin/xaudio2_9redist.dll
+
+# Test Programs
+!/bin/test/
+
 # Themes
 !/bin/GuiConfigs/
 /bin/GuiConfigs/*.ini
@@ -65,6 +68,8 @@
 *.vcxproj.user
 enc_temp_folder/*
 CMakeSettings.json
+*PVS-Studio*
+PVS/*

 # Ignore other system generated files
 x64/*
@@ -117,6 +122,9 @@ yaml-cpp.pc
 /3rdparty/libusb_cmake/config.h
 /3rdparty/libusb_cmake/libusb-1.0.pc

+# miniupnp
+/3rdparty/miniupnp/x64/*
+
 # ssl certificate
 cacert.pem

@@ -1,78 +1,85 @@
 [submodule "rpcs3-ffmpeg"]
 	path = 3rdparty/ffmpeg
-	url = https://github.com/RPCS3/ffmpeg-core
+	url = ../../RPCS3/ffmpeg-core.git
 	ignore = dirty
 [submodule "asmjit"]
-	path = asmjit
-	url = https://github.com/asmjit/asmjit
-	branch = oldstable
+	path = 3rdparty/asmjit/asmjit
+	url = ../../RPCS3/asmjit.git
+	branch = aarch64
 	ignore = dirty
 [submodule "llvm"]
 	path = llvm
-	url = https://github.com/RPCS3/llvm-mirror
+	url = ../../RPCS3/llvm-mirror.git
 	ignore = dirty
-[submodule "Vulkan/glslang"]
-	path = Vulkan/glslang
-	url = https://github.com/KhronosGroup/glslang.git
+[submodule "3rdparty/glslang"]
+	path = 3rdparty/glslang/glslang
+	url = ../../KhronosGroup/glslang.git
 	ignore = dirty
-[submodule "Vulkan/spirv-tools"]
-	path = Vulkan/spirv-tools
-	url = https://github.com/KhronosGroup/SPIRV-Tools.git
+[submodule "3rdparty/SPIRV-Tools"]
+	path = 3rdparty/SPIRV/SPIRV-Tools
+	url = ../../KhronosGroup/SPIRV-Tools.git
 	ignore = dirty
-[submodule "Vulkan/spirv-headers"]
-	path = Vulkan/spirv-headers
-	url = https://github.com/KhronosGroup/SPIRV-Headers.git
-	ignore = dirty
-[submodule "3rdparty/cereal"]
-	path = 3rdparty/cereal
-	url = https://github.com/RPCS3/cereal.git
+[submodule "3rdparty/SPIRV-Headers"]
+	path = 3rdparty/SPIRV/SPIRV-Headers
+	url = ../../KhronosGroup/SPIRV-Headers.git
 	ignore = dirty
 [submodule "3rdparty/zlib"]
-	path = 3rdparty/zlib
-	url = https://github.com/madler/zlib
+	path = 3rdparty/zlib/zlib
+	url = ../../madler/zlib
 	ignore = dirty
 [submodule "3rdparty/hidapi"]
-	path = 3rdparty/hidapi
-	url = https://github.com/RPCS3/hidapi
+	path = 3rdparty/hidapi/hidapi
+	url = ../../RPCS3/hidapi.git
 	branch = master
 	ignore = dirty
 [submodule "3rdparty/pugixml"]
 	path = 3rdparty/pugixml
-	url = https://github.com/zeux/pugixml
+	url = ../../zeux/pugixml.git
 	ignore = dirty
 [submodule "3rdparty/xxHash"]
 	path = 3rdparty/xxHash
-	url = https://github.com/Cyan4973/xxHash
+	url = ../../Cyan4973/xxHash.git
 	ignore = dirty
 [submodule "3rdparty/yaml-cpp"]
-	path = 3rdparty/yaml-cpp
-	url = https://github.com/RPCS3/yaml-cpp.git
+	path = 3rdparty/yaml-cpp/yaml-cpp
+	url = ../../RPCS3/yaml-cpp.git
 	ignore = dirty
 [submodule "3rdparty/libpng"]
-	path = 3rdparty/libpng
-	url = https://github.com/glennrp/libpng.git
+	path = 3rdparty/libpng/libpng
+	url = ../../glennrp/libpng.git
 	ignore = dirty
 [submodule "3rdparty/libusb"]
-	path = 3rdparty/libusb
-	url = https://github.com/libusb/libusb.git
+	path = 3rdparty/libusb/libusb
+	url = ../../libusb/libusb.git
 	ignore = dirty
 [submodule "3rdparty/FAudio"]
 	path = 3rdparty/FAudio
-	url = https://github.com/FNA-XNA/FAudio.git
-	ignore = dirty
-[submodule "3rdparty/span"]
-	path = 3rdparty/span
-	url = https://github.com/tcbrindle/span
+	url = ../../FNA-XNA/FAudio.git
 	ignore = dirty
 [submodule "3rdparty/curl"]
-	path = 3rdparty/curl
-	url = https://github.com/curl/curl.git
+	path = 3rdparty/curl/curl
+	url = ../../curl/curl.git
 	ignore = dirty
 [submodule "3rdparty/wolfssl"]
-	path = 3rdparty/wolfssl
-	url = https://github.com/wolfSSL/wolfssl.git
+	path = 3rdparty/wolfssl/wolfssl
+	url = ../../wolfSSL/wolfssl.git
 	ignore = dirty
 [submodule "3rdparty/flatbuffers"]
 	path = 3rdparty/flatbuffers
-	url = https://github.com/google/flatbuffers.git
+	url = ../../google/flatbuffers.git
 	ignore = dirty
+[submodule "3rdparty/cubeb/cubeb"]
+	path = 3rdparty/cubeb/cubeb
+	url = ../../mozilla/cubeb.git
+	ignore = dirty
+[submodule "3rdparty/SoundTouch/soundtouch"]
+	path = 3rdparty/SoundTouch/soundtouch
+	url = ../../RPCS3/soundtouch.git
+	ignore = dirty
+[submodule "3rdparty/libsdl-org/SDL"]
+	path = 3rdparty/libsdl-org/SDL
+	url = https://github.com/libsdl-org/SDL.git
+	ignore = dirty
+[submodule "3rdparty/miniupnp/miniupnp"]
+	path = 3rdparty/miniupnp/miniupnp
+	url = https://github.com/miniupnp/miniupnp.git
@@ -1,128 +0,0 @@
-/* 7zCrc.c -- CRC32 init
-2017-06-06 : Igor Pavlov : Public domain */
-
-#include "Precomp.h"
-
-#include "7zCrc.h"
-#include "CpuArch.h"
-
-#define kCrcPoly 0xEDB88320
-
-#ifdef MY_CPU_LE
-  #define CRC_NUM_TABLES 8
-#else
-  #define CRC_NUM_TABLES 9
-
-  #define CRC_UINT32_SWAP(v) ((v >> 24) | ((v >> 8) & 0xFF00) | ((v << 8) & 0xFF0000) | (v << 24))
-
-  UInt32 MY_FAST_CALL CrcUpdateT1_BeT4(UInt32 v, const void *data, size_t size, const UInt32 *table);
-  UInt32 MY_FAST_CALL CrcUpdateT1_BeT8(UInt32 v, const void *data, size_t size, const UInt32 *table);
-#endif
-
-#ifndef MY_CPU_BE
-  UInt32 MY_FAST_CALL CrcUpdateT4(UInt32 v, const void *data, size_t size, const UInt32 *table);
-  UInt32 MY_FAST_CALL CrcUpdateT8(UInt32 v, const void *data, size_t size, const UInt32 *table);
-#endif
-
-typedef UInt32 (MY_FAST_CALL *CRC_FUNC)(UInt32 v, const void *data, size_t size, const UInt32 *table);
-
-CRC_FUNC g_CrcUpdateT4;
-CRC_FUNC g_CrcUpdateT8;
-CRC_FUNC g_CrcUpdate;
-
-UInt32 g_CrcTable[256 * CRC_NUM_TABLES];
-
-UInt32 MY_FAST_CALL CrcUpdate(UInt32 v, const void *data, size_t size)
-{
-  return g_CrcUpdate(v, data, size, g_CrcTable);
-}
-
-UInt32 MY_FAST_CALL CrcCalc(const void *data, size_t size)
-{
-  return g_CrcUpdate(CRC_INIT_VAL, data, size, g_CrcTable) ^ CRC_INIT_VAL;
-}
-
-#define CRC_UPDATE_BYTE_2(crc, b) (table[((crc) ^ (b)) & 0xFF] ^ ((crc) >> 8))
-
-UInt32 MY_FAST_CALL CrcUpdateT1(UInt32 v, const void *data, size_t size, const UInt32 *table)
-{
-  const Byte *p = (const Byte *)data;
-  const Byte *pEnd = p + size;
-  for (; p != pEnd; p++)
-    v = CRC_UPDATE_BYTE_2(v, *p);
-  return v;
-}
-
-void MY_FAST_CALL CrcGenerateTable()
-{
-  UInt32 i;
-  for (i = 0; i < 256; i++)
-  {
-    UInt32 r = i;
-    unsigned j;
-    for (j = 0; j < 8; j++)
-      r = (r >> 1) ^ (kCrcPoly & ((UInt32)0 - (r & 1)));
-    g_CrcTable[i] = r;
-  }
-  for (i = 256; i < 256 * CRC_NUM_TABLES; i++)
-  {
-    UInt32 r = g_CrcTable[(size_t)i - 256];
-    g_CrcTable[i] = g_CrcTable[r & 0xFF] ^ (r >> 8);
-  }
-
-  #if CRC_NUM_TABLES < 4
-  
-  g_CrcUpdate = CrcUpdateT1;
-  
-  #else
- 
-  #ifdef MY_CPU_LE
-
-    g_CrcUpdateT4 = CrcUpdateT4;
-    g_CrcUpdate = CrcUpdateT4;
-
-    #if CRC_NUM_TABLES >= 8
-      g_CrcUpdateT8 = CrcUpdateT8;
-  
-      #ifdef MY_CPU_X86_OR_AMD64
-      if (!CPU_Is_InOrder())
-      #endif
-        g_CrcUpdate = CrcUpdateT8;
-    #endif
-
-  #else
-  {
-    #ifndef MY_CPU_BE
-    UInt32 k = 0x01020304;
-    const Byte *p = (const Byte *)&k;
-    if (p[0] == 4 && p[1] == 3)
-    {
-      g_CrcUpdateT4 = CrcUpdateT4;
-      g_CrcUpdate = CrcUpdateT4;
-      #if CRC_NUM_TABLES >= 8
-      g_CrcUpdateT8 = CrcUpdateT8;
-      g_CrcUpdate = CrcUpdateT8;
-      #endif
-    }
-    else if (p[0] != 1 || p[1] != 2)
-      g_CrcUpdate = CrcUpdateT1;
-    else
-    #endif
-    {
-      for (i = 256 * CRC_NUM_TABLES - 1; i >= 256; i--)
-      {
-        UInt32 x = g_CrcTable[(size_t)i - 256];
-        g_CrcTable[i] = CRC_UINT32_SWAP(x);
-      }
-      g_CrcUpdateT4 = CrcUpdateT1_BeT4;
-      g_CrcUpdate = CrcUpdateT1_BeT4;
-      #if CRC_NUM_TABLES >= 8
-      g_CrcUpdateT8 = CrcUpdateT1_BeT8;
-      g_CrcUpdate = CrcUpdateT1_BeT8;
-      #endif
-    }
-  }
-  #endif
-
-  #endif
-}
@@ -19,89 +19,89 @@
    </ProjectConfiguration>
  </ItemGroup>
  <ItemGroup>
-    <ClInclude Include="7z\7z.h" />
-    <ClInclude Include="7z\7zAlloc.h" />
-    <ClInclude Include="7z\7zBuf.h" />
-    <ClInclude Include="7z\7zCrc.h" />
-    <ClInclude Include="7z\7zFile.h" />
-    <ClInclude Include="7z\7zTypes.h" />
-    <ClInclude Include="7z\7zVersion.h" />
-    <ClInclude Include="7z\Aes.h" />
-    <ClInclude Include="7z\Alloc.h" />
-    <ClInclude Include="7z\Bcj2.h" />
-    <ClInclude Include="7z\Bra.h" />
-    <ClInclude Include="7z\Compiler.h" />
-    <ClInclude Include="7z\CpuArch.h" />
-    <ClInclude Include="7z\Delta.h" />
-    <ClInclude Include="7z\DllSecur.h" />
-    <ClInclude Include="7z\LzFind.h" />
-    <ClInclude Include="7z\LzFindMt.h" />
-    <ClInclude Include="7z\LzHash.h" />
-    <ClInclude Include="7z\Lzma2Dec.h" />
-    <ClInclude Include="7z\Lzma2DecMt.h" />
-    <ClInclude Include="7z\Lzma2Enc.h" />
-    <ClInclude Include="7z\Lzma86.h" />
-    <ClInclude Include="7z\LzmaDec.h" />
-    <ClInclude Include="7z\LzmaEnc.h" />
-    <ClInclude Include="7z\LzmaLib.h" />
-    <ClInclude Include="7z\MtCoder.h" />
-    <ClInclude Include="7z\MtDec.h" />
-    <ClInclude Include="7z\Ppmd.h" />
-    <ClInclude Include="7z\Ppmd7.h" />
-    <ClInclude Include="7z\Precomp.h" />
-    <ClInclude Include="7z\RotateDefs.h" />
-    <ClInclude Include="7z\Sha256.h" />
-    <ClInclude Include="7z\Sort.h" />
-    <ClInclude Include="7z\Threads.h" />
-    <ClInclude Include="7z\Xz.h" />
-    <ClInclude Include="7z\XzCrc64.h" />
-    <ClInclude Include="7z\XzEnc.h" />
+    <ClInclude Include="src\7z.h" />
+    <ClInclude Include="src\7zAlloc.h" />
+    <ClInclude Include="src\7zBuf.h" />
+    <ClInclude Include="src\7zCrc.h" />
+    <ClInclude Include="src\7zFile.h" />
+    <ClInclude Include="src\7zTypes.h" />
+    <ClInclude Include="src\7zVersion.h" />
+    <ClInclude Include="src\Aes.h" />
+    <ClInclude Include="src\Alloc.h" />
+    <ClInclude Include="src\Bcj2.h" />
+    <ClInclude Include="src\Bra.h" />
+    <ClInclude Include="src\Compiler.h" />
+    <ClInclude Include="src\CpuArch.h" />
+    <ClInclude Include="src\Delta.h" />
+    <ClInclude Include="src\DllSecur.h" />
+    <ClInclude Include="src\LzFind.h" />
+    <ClInclude Include="src\LzFindMt.h" />
+    <ClInclude Include="src\LzHash.h" />
+    <ClInclude Include="src\Lzma2Dec.h" />
+    <ClInclude Include="src\Lzma2DecMt.h" />
+    <ClInclude Include="src\Lzma2Enc.h" />
+    <ClInclude Include="src\Lzma86.h" />
+    <ClInclude Include="src\LzmaDec.h" />
+    <ClInclude Include="src\LzmaEnc.h" />
+    <ClInclude Include="src\LzmaLib.h" />
+    <ClInclude Include="src\MtCoder.h" />
+    <ClInclude Include="src\MtDec.h" />
+    <ClInclude Include="src\Ppmd.h" />
+    <ClInclude Include="src\Ppmd7.h" />
+    <ClInclude Include="src\Precomp.h" />
+    <ClInclude Include="src\RotateDefs.h" />
+    <ClInclude Include="src\Sha256.h" />
+    <ClInclude Include="src\Sort.h" />
+    <ClInclude Include="src\Threads.h" />
+    <ClInclude Include="src\Xz.h" />
+    <ClInclude Include="src\XzCrc64.h" />
+    <ClInclude Include="src\XzEnc.h" />
  </ItemGroup>
  <ItemGroup>
-    <ClCompile Include="7z\7zAlloc.c" />
-    <ClCompile Include="7z\7zArcIn.c" />
-    <ClCompile Include="7z\7zBuf.c" />
-    <ClCompile Include="7z\7zBuf2.c" />
-    <ClCompile Include="7z\7zCrc.c" />
-    <ClCompile Include="7z\7zCrcOpt.c" />
-    <ClCompile Include="7z\7zDec.c" />
-    <ClCompile Include="7z\7zFile.c" />
-    <ClCompile Include="7z\7zStream.c" />
-    <ClCompile Include="7z\Aes.c" />
-    <ClCompile Include="7z\AesOpt.c" />
-    <ClCompile Include="7z\Alloc.c" />
-    <ClCompile Include="7z\Bcj2.c" />
-    <ClCompile Include="7z\Bcj2Enc.c" />
-    <ClCompile Include="7z\Bra.c" />
-    <ClCompile Include="7z\Bra86.c" />
-    <ClCompile Include="7z\BraIA64.c" />
-    <ClCompile Include="7z\CpuArch.c" />
-    <ClCompile Include="7z\Delta.c" />
-    <ClCompile Include="7z\DllSecur.c" />
-    <ClCompile Include="7z\LzFind.c" />
-    <ClCompile Include="7z\LzFindMt.c" />
-    <ClCompile Include="7z\Lzma2Dec.c" />
-    <ClCompile Include="7z\Lzma2DecMt.c" />
-    <ClCompile Include="7z\Lzma2Enc.c" />
-    <ClCompile Include="7z\Lzma86Dec.c" />
-    <ClCompile Include="7z\Lzma86Enc.c" />
-    <ClCompile Include="7z\LzmaDec.c" />
-    <ClCompile Include="7z\LzmaEnc.c" />
-    <ClCompile Include="7z\LzmaLib.c" />
-    <ClCompile Include="7z\MtCoder.c" />
-    <ClCompile Include="7z\MtDec.c" />
-    <ClCompile Include="7z\Ppmd7.c" />
-    <ClCompile Include="7z\Ppmd7Dec.c" />
-    <ClCompile Include="7z\Ppmd7Enc.c" />
-    <ClCompile Include="7z\Sha256.c" />
-    <ClCompile Include="7z\Sort.c" />
-    <ClCompile Include="7z\Threads.c" />
-    <ClCompile Include="7z\Xz.c" />
-    <ClCompile Include="7z\XzCrc64.c" />
-    <ClCompile Include="7z\XzCrc64Opt.c" />
-    <ClCompile Include="7z\XzDec.c" />
-    <ClCompile Include="7z\XzEnc.c" />
-    <ClCompile Include="7z\XzIn.c" />
+    <ClCompile Include="src\7zAlloc.c" />
+    <ClCompile Include="src\7zArcIn.c" />
+    <ClCompile Include="src\7zBuf.c" />
+    <ClCompile Include="src\7zBuf2.c" />
+    <ClCompile Include="src\7zCrc.c" />
+    <ClCompile Include="src\7zCrcOpt.c" />
+    <ClCompile Include="src\7zDec.c" />
+    <ClCompile Include="src\7zFile.c" />
+    <ClCompile Include="src\7zStream.c" />
+    <ClCompile Include="src\Aes.c" />
+    <ClCompile Include="src\AesOpt.c" />
+    <ClCompile Include="src\Alloc.c" />
+    <ClCompile Include="src\Bcj2.c" />
+    <ClCompile Include="src\Bcj2Enc.c" />
+    <ClCompile Include="src\Bra.c" />
+    <ClCompile Include="src\Bra86.c" />
+    <ClCompile Include="src\BraIA64.c" />
+    <ClCompile Include="src\CpuArch.c" />
+    <ClCompile Include="src\Delta.c" />
+    <ClCompile Include="src\DllSecur.c" />
+    <ClCompile Include="src\LzFind.c" />
+    <ClCompile Include="src\LzFindMt.c" />
+    <ClCompile Include="src\Lzma2Dec.c" />
+    <ClCompile Include="src\Lzma2DecMt.c" />
+    <ClCompile Include="src\Lzma2Enc.c" />
+    <ClCompile Include="src\Lzma86Dec.c" />
+    <ClCompile Include="src\Lzma86Enc.c" />
+    <ClCompile Include="src\LzmaDec.c" />
+    <ClCompile Include="src\LzmaEnc.c" />
+    <ClCompile Include="src\LzmaLib.c" />
+    <ClCompile Include="src\MtCoder.c" />
+    <ClCompile Include="src\MtDec.c" />
+    <ClCompile Include="src\Ppmd7.c" />
+    <ClCompile Include="src\Ppmd7Dec.c" />
+    <ClCompile Include="src\Ppmd7Enc.c" />
+    <ClCompile Include="src\Sha256.c" />
+    <ClCompile Include="src\Sort.c" />
+    <ClCompile Include="src\Threads.c" />
+    <ClCompile Include="src\Xz.c" />
+    <ClCompile Include="src\XzCrc64.c" />
+    <ClCompile Include="src\XzCrc64Opt.c" />
+    <ClCompile Include="src\XzDec.c" />
+    <ClCompile Include="src\XzEnc.c" />
+    <ClCompile Include="src\XzIn.c" />
  </ItemGroup>
  <PropertyGroup Label="Globals">
    <VCProjectVersion>16.0</VCProjectVersion>
@@ -109,12 +109,12 @@
    <Keyword>Win32Proj</Keyword>
    <RootNamespace>My7zlib</RootNamespace>
  </PropertyGroup>
-  <Import Project="..\common_default.props" />
+  <Import Project="$(SolutionDir)\buildfiles\msvc\common_default.props" />
  <Import Project="$(VCTargetsPath)\Microsoft.Cpp.Default.props" />
  <PropertyGroup>
    <PreferredToolArchitecture>x64</PreferredToolArchitecture>
  </PropertyGroup>
-  <Import Project="..\common_default_macros.props" />
+  <Import Project="$(SolutionDir)\buildfiles\msvc\common_default_macros.props" />
  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'" Label="Configuration">
    <ConfigurationType>StaticLibrary</ConfigurationType>
    <UseDebugLibraries>true</UseDebugLibraries>
@@ -15,249 +15,249 @@
    </Filter>
  </ItemGroup>
  <ItemGroup>
-    <ClInclude Include="7z\7z.h">
+    <ClInclude Include"src\7z.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\7zAlloc.h">
+    <ClInclude Include"src\7zAlloc.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\7zBuf.h">
+    <ClInclude Include"src\7zBuf.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\7zCrc.h">
+    <ClInclude Include"src\7zCrc.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\7zFile.h">
+    <ClInclude Include"src\7zFile.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\7zTypes.h">
+    <ClInclude Include"src\7zTypes.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\7zVersion.h">
+    <ClInclude Include"src\7zVersion.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\Aes.h">
+    <ClInclude Include"src\Aes.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\Alloc.h">
+    <ClInclude Include"src\Alloc.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\Bcj2.h">
+    <ClInclude Include"src\Bcj2.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\Bra.h">
+    <ClInclude Include"src\Bra.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\Compiler.h">
+    <ClInclude Include"src\Compiler.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\CpuArch.h">
+    <ClInclude Include"src\CpuArch.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\Delta.h">
+    <ClInclude Include"src\Delta.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\DllSecur.h">
+    <ClInclude Include"src\DllSecur.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\LzFind.h">
+    <ClInclude Include"src\LzFind.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\LzFindMt.h">
+    <ClInclude Include"src\LzFindMt.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\LzHash.h">
+    <ClInclude Include"src\LzHash.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\Lzma2Dec.h">
+    <ClInclude Include"src\Lzma2Dec.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\Lzma2DecMt.h">
+    <ClInclude Include"src\Lzma2DecMt.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\Lzma2Enc.h">
+    <ClInclude Include"src\Lzma2Enc.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\Lzma86.h">
+    <ClInclude Include"src\Lzma86.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\LzmaDec.h">
+    <ClInclude Include"src\LzmaDec.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\LzmaEnc.h">
+    <ClInclude Include"src\LzmaEnc.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\LzmaLib.h">
+    <ClInclude Include"src\LzmaLib.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\MtCoder.h">
+    <ClInclude Include"src\MtCoder.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\MtDec.h">
+    <ClInclude Include"src\MtDec.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\Ppmd.h">
+    <ClInclude Include"src\Ppmd.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\Ppmd7.h">
+    <ClInclude Include"src\Ppmd7.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\Precomp.h">
+    <ClInclude Include"src\Precomp.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\RotateDefs.h">
+    <ClInclude Include"src\RotateDefs.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\Sha256.h">
+    <ClInclude Include"src\Sha256.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\Sort.h">
+    <ClInclude Include"src\Sort.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\Threads.h">
+    <ClInclude Include"src\Threads.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\Xz.h">
+    <ClInclude Include"src\Xz.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\XzCrc64.h">
+    <ClInclude Include"src\XzCrc64.h">
      <Filter>Header Files</Filter>
    </ClInclude>
-    <ClInclude Include="7z\XzEnc.h">
+    <ClInclude Include"src\XzEnc.h">
      <Filter>Header Files</Filter>
    </ClInclude>
  </ItemGroup>
  <ItemGroup>
-    <ClCompile Include="7z\7zAlloc.c">
+    <ClCompile Include"src\7zAlloc.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\7zArcIn.c">
+    <ClCompile Include"src\7zArcIn.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\7zBuf.c">
+    <ClCompile Include"src\7zBuf.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\7zBuf2.c">
+    <ClCompile Include"src\7zBuf2.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\7zCrc.c">
+    <ClCompile Include"src\7zCrc.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\7zCrcOpt.c">
+    <ClCompile Include"src\7zCrcOpt.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\7zDec.c">
+    <ClCompile Include"src\7zDec.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\7zFile.c">
+    <ClCompile Include"src\7zFile.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\7zStream.c">
+    <ClCompile Include"src\7zStream.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\Aes.c">
+    <ClCompile Include"src\Aes.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\AesOpt.c">
+    <ClCompile Include"src\AesOpt.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\Alloc.c">
+    <ClCompile Include"src\Alloc.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\Bcj2.c">
+    <ClCompile Include"src\Bcj2.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\Bcj2Enc.c">
+    <ClCompile Include"src\Bcj2Enc.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\Bra.c">
+    <ClCompile Include"src\Bra.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\Bra86.c">
+    <ClCompile Include"src\Bra86.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\BraIA64.c">
+    <ClCompile Include"src\BraIA64.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\CpuArch.c">
+    <ClCompile Include"src\CpuArch.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\Delta.c">
+    <ClCompile Include"src\Delta.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\DllSecur.c">
+    <ClCompile Include"src\DllSecur.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\LzFind.c">
+    <ClCompile Include"src\LzFind.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\LzFindMt.c">
+    <ClCompile Include"src\LzFindMt.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\Lzma2Dec.c">
+    <ClCompile Include"src\Lzma2Dec.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\Lzma2DecMt.c">
+    <ClCompile Include"src\Lzma2DecMt.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\Lzma2Enc.c">
+    <ClCompile Include"src\Lzma2Enc.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\Lzma86Dec.c">
+    <ClCompile Include"src\Lzma86Dec.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\Lzma86Enc.c">
+    <ClCompile Include"src\Lzma86Enc.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\LzmaDec.c">
+    <ClCompile Include"src\LzmaDec.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\LzmaEnc.c">
+    <ClCompile Include"src\LzmaEnc.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\LzmaLib.c">
+    <ClCompile Include"src\LzmaLib.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\MtCoder.c">
+    <ClCompile Include"src\MtCoder.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\MtDec.c">
+    <ClCompile Include"src\MtDec.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\Ppmd7.c">
+    <ClCompile Include"src\Ppmd7.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\Ppmd7Dec.c">
+    <ClCompile Include"src\Ppmd7Dec.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\Ppmd7Enc.c">
+    <ClCompile Include"src\Ppmd7Enc.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\Sha256.c">
+    <ClCompile Include"src\Sha256.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\Sort.c">
+    <ClCompile Include"src\Sort.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\Threads.c">
+    <ClCompile Include"src\Threads.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\Xz.c">
+    <ClCompile Include"src\Xz.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\XzCrc64.c">
+    <ClCompile Include"src\XzCrc64.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\XzCrc64Opt.c">
+    <ClCompile Include"src\XzCrc64Opt.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\XzDec.c">
+    <ClCompile Include"src\XzDec.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\XzEnc.c">
+    <ClCompile Include"src\XzEnc.c">
      <Filter>Source Files</Filter>
    </ClCompile>
-    <ClCompile Include="7z\XzIn.c">
+    <ClCompile Include"src\XzIn.c">
      <Filter>Source Files</Filter>
    </ClCompile>
  </ItemGroup>
@@ -1,184 +0,0 @@
-/* AesOpt.c -- Intel's AES
-2017-06-08 : Igor Pavlov : Public domain */
-
-#include "Precomp.h"
-
-#include "CpuArch.h"
-
-#ifdef MY_CPU_X86_OR_AMD64
-#if (_MSC_VER > 1500) || (_MSC_FULL_VER >= 150030729)
-#define USE_INTEL_AES
-#endif
-#endif
-
-#ifdef USE_INTEL_AES
-
-#include <wmmintrin.h>
-
-void MY_FAST_CALL AesCbc_Encode_Intel(__m128i *p, __m128i *data, size_t numBlocks)
-{
-  __m128i m = *p;
-  for (; numBlocks != 0; numBlocks--, data++)
-  {
-    UInt32 numRounds2 = *(const UInt32 *)(p + 1) - 1;
-    const __m128i *w = p + 3;
-    m = _mm_xor_si128(m, *data);
-    m = _mm_xor_si128(m, p[2]);
-    do
-    {
-      m = _mm_aesenc_si128(m, w[0]);
-      m = _mm_aesenc_si128(m, w[1]);
-      w += 2;
-    }
-    while (--numRounds2 != 0);
-    m = _mm_aesenc_si128(m, w[0]);
-    m = _mm_aesenclast_si128(m, w[1]);
-    *data = m;
-  }
-  *p = m;
-}
-
-#define NUM_WAYS 3
-
-#define AES_OP_W(op, n) { \
-    const __m128i t = w[n]; \
-    m0 = op(m0, t); \
-    m1 = op(m1, t); \
-    m2 = op(m2, t); \
-    }
-
-#define AES_DEC(n) AES_OP_W(_mm_aesdec_si128, n)
-#define AES_DEC_LAST(n) AES_OP_W(_mm_aesdeclast_si128, n)
-#define AES_ENC(n) AES_OP_W(_mm_aesenc_si128, n)
-#define AES_ENC_LAST(n) AES_OP_W(_mm_aesenclast_si128, n)
-
-void MY_FAST_CALL AesCbc_Decode_Intel(__m128i *p, __m128i *data, size_t numBlocks)
-{
-  __m128i iv = *p;
-  for (; numBlocks >= NUM_WAYS; numBlocks -= NUM_WAYS, data += NUM_WAYS)
-  {
-    UInt32 numRounds2 = *(const UInt32 *)(p + 1);
-    const __m128i *w = p + numRounds2 * 2;
-    __m128i m0, m1, m2;
-    {
-      const __m128i t = w[2];
-      m0 = _mm_xor_si128(t, data[0]);
-      m1 = _mm_xor_si128(t, data[1]);
-      m2 = _mm_xor_si128(t, data[2]);
-    }
-    numRounds2--;
-    do
-    {
-      AES_DEC(1)
-      AES_DEC(0)
-      w -= 2;
-    }
-    while (--numRounds2 != 0);
-    AES_DEC(1)
-    AES_DEC_LAST(0)
-
-    {
-      __m128i t;
-      t = _mm_xor_si128(m0, iv); iv = data[0]; data[0] = t;
-      t = _mm_xor_si128(m1, iv); iv = data[1]; data[1] = t;
-      t = _mm_xor_si128(m2, iv); iv = data[2]; data[2] = t;
-    }
-  }
-  for (; numBlocks != 0; numBlocks--, data++)
-  {
-    UInt32 numRounds2 = *(const UInt32 *)(p + 1);
-    const __m128i *w = p + numRounds2 * 2;
-    __m128i m = _mm_xor_si128(w[2], *data);
-    numRounds2--;
-    do
-    {
-      m = _mm_aesdec_si128(m, w[1]);
-      m = _mm_aesdec_si128(m, w[0]);
-      w -= 2;
-    }
-    while (--numRounds2 != 0);
-    m = _mm_aesdec_si128(m, w[1]);
-    m = _mm_aesdeclast_si128(m, w[0]);
-
-    m = _mm_xor_si128(m, iv);
-    iv = *data;
-    *data = m;
-  }
-  *p = iv;
-}
-
-void MY_FAST_CALL AesCtr_Code_Intel(__m128i *p, __m128i *data, size_t numBlocks)
-{
-  __m128i ctr = *p;
-  __m128i one;
-  one.m128i_u64[0] = 1;
-  one.m128i_u64[1] = 0;
-  for (; numBlocks >= NUM_WAYS; numBlocks -= NUM_WAYS, data += NUM_WAYS)
-  {
-    UInt32 numRounds2 = *(const UInt32 *)(p + 1) - 1;
-    const __m128i *w = p;
-    __m128i m0, m1, m2;
-    {
-      const __m128i t = w[2];
-      ctr = _mm_add_epi64(ctr, one); m0 = _mm_xor_si128(ctr, t);
-      ctr = _mm_add_epi64(ctr, one); m1 = _mm_xor_si128(ctr, t);
-      ctr = _mm_add_epi64(ctr, one); m2 = _mm_xor_si128(ctr, t);
-    }
-    w += 3;
-    do
-    {
-      AES_ENC(0)
-      AES_ENC(1)
-      w += 2;
-    }
-    while (--numRounds2 != 0);
-    AES_ENC(0)
-    AES_ENC_LAST(1)
-    data[0] = _mm_xor_si128(data[0], m0);
-    data[1] = _mm_xor_si128(data[1], m1);
-    data[2] = _mm_xor_si128(data[2], m2);
-  }
-  for (; numBlocks != 0; numBlocks--, data++)
-  {
-    UInt32 numRounds2 = *(const UInt32 *)(p + 1) - 1;
-    const __m128i *w = p;
-    __m128i m;
-    ctr = _mm_add_epi64(ctr, one);
-    m = _mm_xor_si128(ctr, p[2]);
-    w += 3;
-    do
-    {
-      m = _mm_aesenc_si128(m, w[0]);
-      m = _mm_aesenc_si128(m, w[1]);
-      w += 2;
-    }
-    while (--numRounds2 != 0);
-    m = _mm_aesenc_si128(m, w[0]);
-    m = _mm_aesenclast_si128(m, w[1]);
-    *data = _mm_xor_si128(*data, m);
-  }
-  *p = ctr;
-}
-
-#else
-
-void MY_FAST_CALL AesCbc_Encode(UInt32 *ivAes, Byte *data, size_t numBlocks);
-void MY_FAST_CALL AesCbc_Decode(UInt32 *ivAes, Byte *data, size_t numBlocks);
-void MY_FAST_CALL AesCtr_Code(UInt32 *ivAes, Byte *data, size_t numBlocks);
-
-void MY_FAST_CALL AesCbc_Encode_Intel(UInt32 *p, Byte *data, size_t numBlocks)
-{
-  AesCbc_Encode(p, data, numBlocks);
-}
-
-void MY_FAST_CALL AesCbc_Decode_Intel(UInt32 *p, Byte *data, size_t numBlocks)
-{
-  AesCbc_Decode(p, data, numBlocks);
-}
-
-void MY_FAST_CALL AesCtr_Code_Intel(UInt32 *p, Byte *data, size_t numBlocks)
-{
-  AesCtr_Code(p, data, numBlocks);
-}
-
-#endif
@@ -0,0 +1,58 @@
+# 7z sdk
+if(WIN32)
+	add_library(3rdparty_7z STATIC EXCLUDE_FROM_ALL
+		src/Sha256.c
+		src/Ppmd7Dec.c
+		src/XzDec.c
+		src/XzEnc.c
+		src/Lzma2Dec.c
+		src/XzCrc64.c
+		src/DllSecur.c
+		src/Lzma2DecMt.c
+		src/BraIA64.c
+		src/7zFile.c
+		src/7zStream.c
+		src/Lzma86Enc.c
+		src/Threads.c
+		src/7zAlloc.c
+		src/LzmaEnc.c
+		src/MtCoder.c
+		src/Lzma86Dec.c
+		src/Sort.c
+		src/LzFindMt.c
+		src/7zDec.c
+		src/Bcj2.c
+		src/Ppmd7.c
+		src/Bra86.c
+		src/Bcj2Enc.c
+		src/7zBuf.c
+		src/Aes.c
+		src/AesOpt.c
+		src/XzCrc64Opt.c
+		src/7zArcIn.c
+		src/Lzma2Enc.c
+		src/Bra.c
+		src/7zCrcOpt.c
+		src/7zBuf2.c
+		src/LzFind.c
+		src/Ppmd7Enc.c
+		src/CpuArch.c
+		src/Delta.c
+		src/XzIn.c
+		src/Alloc.c
+		src/Xz.c
+		src/LzmaDec.c
+		src/LzmaLib.c
+		src/7zCrc.c
+		src/MtDec.c)
+	target_include_directories(3rdparty_7z INTERFACE
+		$<BUILD_INTERFACE:${CMAKE_CURRENT_SOURCE_DIR}/src>
+		$<INSTALL_INTERFACE:/src>)
+
+	target_include_directories(3rdparty_7z INTERFACE 7z)
+
+	set_property(TARGET 3rdparty_7z PROPERTY FOLDER "3rdparty/")
+
+else()
+	add_library(3rdparty_7z INTERFACE)
+endif()
@@ -1,218 +0,0 @@
-/* CpuArch.c -- CPU specific code
-2018-02-18: Igor Pavlov : Public domain */
-
-#include "Precomp.h"
-
-#include "CpuArch.h"
-
-#ifdef MY_CPU_X86_OR_AMD64
-
-#if (defined(_MSC_VER) && !defined(MY_CPU_AMD64)) || defined(__GNUC__)
-#define USE_ASM
-#endif
-
-#if !defined(USE_ASM) && _MSC_VER >= 1500
-#include <intrin.h>
-#endif
-
-#if defined(USE_ASM) && !defined(MY_CPU_AMD64)
-static UInt32 CheckFlag(UInt32 flag)
-{
-  #ifdef _MSC_VER
-  __asm pushfd;
-  __asm pop EAX;
-  __asm mov EDX, EAX;
-  __asm xor EAX, flag;
-  __asm push EAX;
-  __asm popfd;
-  __asm pushfd;
-  __asm pop EAX;
-  __asm xor EAX, EDX;
-  __asm push EDX;
-  __asm popfd;
-  __asm and flag, EAX;
-  #else
-  __asm__ __volatile__ (
-    "pushf\n\t"
-    "pop  %%EAX\n\t"
-    "movl %%EAX,%%EDX\n\t"
-    "xorl %0,%%EAX\n\t"
-    "push %%EAX\n\t"
-    "popf\n\t"
-    "pushf\n\t"
-    "pop  %%EAX\n\t"
-    "xorl %%EDX,%%EAX\n\t"
-    "push %%EDX\n\t"
-    "popf\n\t"
-    "andl %%EAX, %0\n\t":
-    "=c" (flag) : "c" (flag) :
-    "%eax", "%edx");
-  #endif
-  return flag;
-}
-#define CHECK_CPUID_IS_SUPPORTED if (CheckFlag(1 << 18) == 0 || CheckFlag(1 << 21) == 0) return False;
-#else
-#define CHECK_CPUID_IS_SUPPORTED
-#endif
-
-void MyCPUID(UInt32 function, UInt32 *a, UInt32 *b, UInt32 *c, UInt32 *d)
-{
-  #ifdef USE_ASM
-
-  #ifdef _MSC_VER
-
-  UInt32 a2, b2, c2, d2;
-  __asm xor EBX, EBX;
-  __asm xor ECX, ECX;
-  __asm xor EDX, EDX;
-  __asm mov EAX, function;
-  __asm cpuid;
-  __asm mov a2, EAX;
-  __asm mov b2, EBX;
-  __asm mov c2, ECX;
-  __asm mov d2, EDX;
-
-  *a = a2;
-  *b = b2;
-  *c = c2;
-  *d = d2;
-
-  #else
-
-  __asm__ __volatile__ (
-  #if defined(MY_CPU_AMD64) && defined(__PIC__)
-    "mov %%rbx, %%rdi;"
-    "cpuid;"
-    "xchg %%rbx, %%rdi;"
-    : "=a" (*a) ,
-      "=D" (*b) ,
-  #elif defined(MY_CPU_X86) && defined(__PIC__)
-    "mov %%ebx, %%edi;"
-    "cpuid;"
-    "xchgl %%ebx, %%edi;"
-    : "=a" (*a) ,
-      "=D" (*b) ,
-  #else
-    "cpuid"
-    : "=a" (*a) ,
-      "=b" (*b) ,
-  #endif
-      "=c" (*c) ,
-      "=d" (*d)
-    : "0" (function)) ;
-
-  #endif
-  
-  #else
-
-  int CPUInfo[4];
-  __cpuid(CPUInfo, function);
-  *a = CPUInfo[0];
-  *b = CPUInfo[1];
-  *c = CPUInfo[2];
-  *d = CPUInfo[3];
-
-  #endif
-}
-
-BoolInt x86cpuid_CheckAndRead(Cx86cpuid *p)
-{
-  CHECK_CPUID_IS_SUPPORTED
-  MyCPUID(0, &p->maxFunc, &p->vendor[0], &p->vendor[2], &p->vendor[1]);
-  MyCPUID(1, &p->ver, &p->b, &p->c, &p->d);
-  return True;
-}
-
-static const UInt32 kVendors[][3] =
-{
-  { 0x756E6547, 0x49656E69, 0x6C65746E},
-  { 0x68747541, 0x69746E65, 0x444D4163},
-  { 0x746E6543, 0x48727561, 0x736C7561}
-};
-
-int x86cpuid_GetFirm(const Cx86cpuid *p)
-{
-  unsigned i;
-  for (i = 0; i < sizeof(kVendors) / sizeof(kVendors[i]); i++)
-  {
-    const UInt32 *v = kVendors[i];
-    if (v[0] == p->vendor[0] &&
-        v[1] == p->vendor[1] &&
-        v[2] == p->vendor[2])
-      return (int)i;
-  }
-  return -1;
-}
-
-BoolInt CPU_Is_InOrder()
-{
-  Cx86cpuid p;
-  int firm;
-  UInt32 family, model;
-  if (!x86cpuid_CheckAndRead(&p))
-    return True;
-
-  family = x86cpuid_GetFamily(p.ver);
-  model = x86cpuid_GetModel(p.ver);
-  
-  firm = x86cpuid_GetFirm(&p);
-
-  switch (firm)
-  {
-    case CPU_FIRM_INTEL: return (family < 6 || (family == 6 && (
-        /* In-Order Atom CPU */
-           model == 0x1C  /* 45 nm, N4xx, D4xx, N5xx, D5xx, 230, 330 */
-        || model == 0x26  /* 45 nm, Z6xx */
-        || model == 0x27  /* 32 nm, Z2460 */
-        || model == 0x35  /* 32 nm, Z2760 */
-        || model == 0x36  /* 32 nm, N2xxx, D2xxx */
-        )));
-    case CPU_FIRM_AMD: return (family < 5 || (family == 5 && (model < 6 || model == 0xA)));
-    case CPU_FIRM_VIA: return (family < 6 || (family == 6 && model < 0xF));
-  }
-  return True;
-}
-
-#if !defined(MY_CPU_AMD64) && defined(_WIN32)
-#include <windows.h>
-static BoolInt CPU_Sys_Is_SSE_Supported()
-{
-  OSVERSIONINFO vi;
-  vi.dwOSVersionInfoSize = sizeof(vi);
-  if (!GetVersionEx(&vi))
-    return False;
-  return (vi.dwMajorVersion >= 5);
-}
-#define CHECK_SYS_SSE_SUPPORT if (!CPU_Sys_Is_SSE_Supported()) return False;
-#else
-#define CHECK_SYS_SSE_SUPPORT
-#endif
-
-BoolInt CPU_Is_Aes_Supported()
-{
-  Cx86cpuid p;
-  CHECK_SYS_SSE_SUPPORT
-  if (!x86cpuid_CheckAndRead(&p))
-    return False;
-  return (p.c >> 25) & 1;
-}
-
-BoolInt CPU_IsSupported_PageGB()
-{
-  Cx86cpuid cpuid;
-  if (!x86cpuid_CheckAndRead(&cpuid))
-    return False;
-  {
-    UInt32 d[4] = { 0 };
-    MyCPUID(0x80000000, &d[0], &d[1], &d[2], &d[3]);
-    if (d[0] < 0x80000001)
-      return False;
-  }
-  {
-    UInt32 d[4] = { 0 };
-    MyCPUID(0x80000001, &d[0], &d[1], &d[2], &d[3]);
-    return (d[3] >> 26) & 1;
-  }
-}
-
-#endif
@@ -1,64 +0,0 @@
-/* Delta.c -- Delta converter
-2009-05-26 : Igor Pavlov : Public domain */
-
-#include "Precomp.h"
-
-#include "Delta.h"
-
-void Delta_Init(Byte *state)
-{
-  unsigned i;
-  for (i = 0; i < DELTA_STATE_SIZE; i++)
-    state[i] = 0;
-}
-
-static void MyMemCpy(Byte *dest, const Byte *src, unsigned size)
-{
-  unsigned i;
-  for (i = 0; i < size; i++)
-    dest[i] = src[i];
-}
-
-void Delta_Encode(Byte *state, unsigned delta, Byte *data, SizeT size)
-{
-  Byte buf[DELTA_STATE_SIZE];
-  unsigned j = 0;
-  MyMemCpy(buf, state, delta);
-  {
-    SizeT i;
-    for (i = 0; i < size;)
-    {
-      for (j = 0; j < delta && i < size; i++, j++)
-      {
-        Byte b = data[i];
-        data[i] = (Byte)(b - buf[j]);
-        buf[j] = b;
-      }
-    }
-  }
-  if (j == delta)
-    j = 0;
-  MyMemCpy(state, buf + j, delta - j);
-  MyMemCpy(state + delta - j, buf, j);
-}
-
-void Delta_Decode(Byte *state, unsigned delta, Byte *data, SizeT size)
-{
-  Byte buf[DELTA_STATE_SIZE];
-  unsigned j = 0;
-  MyMemCpy(buf, state, delta);
-  {
-    SizeT i;
-    for (i = 0; i < size;)
-    {
-      for (j = 0; j < delta && i < size; i++, j++)
-      {
-        buf[j] = data[i] = (Byte)(buf[j] + data[i]);
-      }
-    }
-  }
-  if (j == delta)
-    j = 0;
-  MyMemCpy(state, buf + j, delta - j);
-  MyMemCpy(state + delta - j, buf, j);
-}
@@ -1,853 +0,0 @@
-/* LzFindMt.c -- multithreaded Match finder for LZ algorithms
-2018-12-29 : Igor Pavlov : Public domain */
-
-#include "Precomp.h"
-
-#include "LzHash.h"
-
-#include "LzFindMt.h"
-
-static void MtSync_Construct(CMtSync *p)
-{
-  p->wasCreated = False;
-  p->csWasInitialized = False;
-  p->csWasEntered = False;
-  Thread_Construct(&p->thread);
-  Event_Construct(&p->canStart);
-  Event_Construct(&p->wasStarted);
-  Event_Construct(&p->wasStopped);
-  Semaphore_Construct(&p->freeSemaphore);
-  Semaphore_Construct(&p->filledSemaphore);
-}
-
-static void MtSync_GetNextBlock(CMtSync *p)
-{
-  if (p->needStart)
-  {
-    p->numProcessedBlocks = 1;
-    p->needStart = False;
-    p->stopWriting = False;
-    p->exit = False;
-    Event_Reset(&p->wasStarted);
-    Event_Reset(&p->wasStopped);
-
-    Event_Set(&p->canStart);
-    Event_Wait(&p->wasStarted);
-
-    // if (mt) MatchFinder_Init_LowHash(mt->MatchFinder);
-  }
-  else
-  {
-    CriticalSection_Leave(&p->cs);
-    p->csWasEntered = False;
-    p->numProcessedBlocks++;
-    Semaphore_Release1(&p->freeSemaphore);
-  }
-  Semaphore_Wait(&p->filledSemaphore);
-  CriticalSection_Enter(&p->cs);
-  p->csWasEntered = True;
-}
-
-/* MtSync_StopWriting must be called if Writing was started */
-
-static void MtSync_StopWriting(CMtSync *p)
-{
-  UInt32 myNumBlocks = p->numProcessedBlocks;
-  if (!Thread_WasCreated(&p->thread) || p->needStart)
-    return;
-  p->stopWriting = True;
-  if (p->csWasEntered)
-  {
-    CriticalSection_Leave(&p->cs);
-    p->csWasEntered = False;
-  }
-  Semaphore_Release1(&p->freeSemaphore);
- 
-  Event_Wait(&p->wasStopped);
-
-  while (myNumBlocks++ != p->numProcessedBlocks)
-  {
-    Semaphore_Wait(&p->filledSemaphore);
-    Semaphore_Release1(&p->freeSemaphore);
-  }
-  p->needStart = True;
-}
-
-static void MtSync_Destruct(CMtSync *p)
-{
-  if (Thread_WasCreated(&p->thread))
-  {
-    MtSync_StopWriting(p);
-    p->exit = True;
-    if (p->needStart)
-      Event_Set(&p->canStart);
-    Thread_Wait(&p->thread);
-    Thread_Close(&p->thread);
-  }
-  if (p->csWasInitialized)
-  {
-    CriticalSection_Delete(&p->cs);
-    p->csWasInitialized = False;
-  }
-
-  Event_Close(&p->canStart);
-  Event_Close(&p->wasStarted);
-  Event_Close(&p->wasStopped);
-  Semaphore_Close(&p->freeSemaphore);
-  Semaphore_Close(&p->filledSemaphore);
-
-  p->wasCreated = False;
-}
-
-#define RINOK_THREAD(x) { if ((x) != 0) return SZ_ERROR_THREAD; }
-
-static SRes MtSync_Create2(CMtSync *p, THREAD_FUNC_TYPE startAddress, void *obj, UInt32 numBlocks)
-{
-  if (p->wasCreated)
-    return SZ_OK;
-
-  RINOK_THREAD(CriticalSection_Init(&p->cs));
-  p->csWasInitialized = True;
-
-  RINOK_THREAD(AutoResetEvent_CreateNotSignaled(&p->canStart));
-  RINOK_THREAD(AutoResetEvent_CreateNotSignaled(&p->wasStarted));
-  RINOK_THREAD(AutoResetEvent_CreateNotSignaled(&p->wasStopped));
-  
-  RINOK_THREAD(Semaphore_Create(&p->freeSemaphore, numBlocks, numBlocks));
-  RINOK_THREAD(Semaphore_Create(&p->filledSemaphore, 0, numBlocks));
-
-  p->needStart = True;
-  
-  RINOK_THREAD(Thread_Create(&p->thread, startAddress, obj));
-  p->wasCreated = True;
-  return SZ_OK;
-}
-
-static SRes MtSync_Create(CMtSync *p, THREAD_FUNC_TYPE startAddress, void *obj, UInt32 numBlocks)
-{
-  SRes res = MtSync_Create2(p, startAddress, obj, numBlocks);
-  if (res != SZ_OK)
-    MtSync_Destruct(p);
-  return res;
-}
-
-void MtSync_Init(CMtSync *p) { p->needStart = True; }
-
-#define kMtMaxValForNormalize 0xFFFFFFFF
-
-#define DEF_GetHeads2(name, v, action) \
-  static void GetHeads ## name(const Byte *p, UInt32 pos, \
-      UInt32 *hash, UInt32 hashMask, UInt32 *heads, UInt32 numHeads, const UInt32 *crc) \
-    { action; for (; numHeads != 0; numHeads--) { \
-      const UInt32 value = (v); p++; *heads++ = pos - hash[value]; hash[value] = pos++;  } }
-
-#define DEF_GetHeads(name, v) DEF_GetHeads2(name, v, ;)
-
-DEF_GetHeads2(2,  (p[0] | ((UInt32)p[1] << 8)), UNUSED_VAR(hashMask); UNUSED_VAR(crc); )
-DEF_GetHeads(3,  (crc[p[0]] ^ p[1] ^ ((UInt32)p[2] << 8)) & hashMask)
-DEF_GetHeads(4,  (crc[p[0]] ^ p[1] ^ ((UInt32)p[2] << 8) ^ (crc[p[3]] << 5)) & hashMask)
-DEF_GetHeads(4b, (crc[p[0]] ^ p[1] ^ ((UInt32)p[2] << 8) ^ ((UInt32)p[3] << 16)) & hashMask)
-/* DEF_GetHeads(5,  (crc[p[0]] ^ p[1] ^ ((UInt32)p[2] << 8) ^ (crc[p[3]] << 5) ^ (crc[p[4]] << 3)) & hashMask) */
-
-static void HashThreadFunc(CMatchFinderMt *mt)
-{
-  CMtSync *p = &mt->hashSync;
-  for (;;)
-  {
-    UInt32 numProcessedBlocks = 0;
-    Event_Wait(&p->canStart);
-    Event_Set(&p->wasStarted);
-
-    MatchFinder_Init_HighHash(mt->MatchFinder);
-
-    for (;;)
-    {
-      if (p->exit)
-        return;
-      if (p->stopWriting)
-      {
-        p->numProcessedBlocks = numProcessedBlocks;
-        Event_Set(&p->wasStopped);
-        break;
-      }
-
-      {
-        CMatchFinder *mf = mt->MatchFinder;
-        if (MatchFinder_NeedMove(mf))
-        {
-          CriticalSection_Enter(&mt->btSync.cs);
-          CriticalSection_Enter(&mt->hashSync.cs);
-          {
-            const Byte *beforePtr = Inline_MatchFinder_GetPointerToCurrentPos(mf);
-            ptrdiff_t offset;
-            MatchFinder_MoveBlock(mf);
-            offset = beforePtr - Inline_MatchFinder_GetPointerToCurrentPos(mf);
-            mt->pointerToCurPos -= offset;
-            mt->buffer -= offset;
-          }
-          CriticalSection_Leave(&mt->btSync.cs);
-          CriticalSection_Leave(&mt->hashSync.cs);
-          continue;
-        }
-
-        Semaphore_Wait(&p->freeSemaphore);
-
-        MatchFinder_ReadIfRequired(mf);
-        if (mf->pos > (kMtMaxValForNormalize - kMtHashBlockSize))
-        {
-          UInt32 subValue = (mf->pos - mf->historySize - 1);
-          MatchFinder_ReduceOffsets(mf, subValue);
-          MatchFinder_Normalize3(subValue, mf->hash + mf->fixedHashSize, (size_t)mf->hashMask + 1);
-        }
-        {
-          UInt32 *heads = mt->hashBuf + ((numProcessedBlocks++) & kMtHashNumBlocksMask) * kMtHashBlockSize;
-          UInt32 num = mf->streamPos - mf->pos;
-          heads[0] = 2;
-          heads[1] = num;
-          if (num >= mf->numHashBytes)
-          {
-            num = num - mf->numHashBytes + 1;
-            if (num > kMtHashBlockSize - 2)
-              num = kMtHashBlockSize - 2;
-            mt->GetHeadsFunc(mf->buffer, mf->pos, mf->hash + mf->fixedHashSize, mf->hashMask, heads + 2, num, mf->crc);
-            heads[0] = 2 + num;
-          }
-          mf->pos += num;
-          mf->buffer += num;
-        }
-      }
-
-      Semaphore_Release1(&p->filledSemaphore);
-    }
-  }
-}
-
-static void MatchFinderMt_GetNextBlock_Hash(CMatchFinderMt *p)
-{
-  MtSync_GetNextBlock(&p->hashSync);
-  p->hashBufPosLimit = p->hashBufPos = ((p->hashSync.numProcessedBlocks - 1) & kMtHashNumBlocksMask) * kMtHashBlockSize;
-  p->hashBufPosLimit += p->hashBuf[p->hashBufPos++];
-  p->hashNumAvail = p->hashBuf[p->hashBufPos++];
-}
-
-#define kEmptyHashValue 0
-
-#define MFMT_GM_INLINE
-
-#ifdef MFMT_GM_INLINE
-
-/*
-  we use size_t for _cyclicBufferPos instead of UInt32
-  to eliminate "movsx" BUG in old MSVC x64 compiler.
-*/
-
-MY_NO_INLINE
-static UInt32 *GetMatchesSpecN(UInt32 lenLimit, UInt32 pos, const Byte *cur, CLzRef *son,
-    size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 _cutValue,
-    UInt32 *distances, UInt32 _maxLen, const UInt32 *hash, const UInt32 *limit, UInt32 size, UInt32 *posRes)
-{
-  do
-  {
-  UInt32 *_distances = ++distances;
-  UInt32 delta = *hash++;
-
-  CLzRef *ptr0 = son + ((size_t)_cyclicBufferPos << 1) + 1;
-  CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1);
-  unsigned len0 = 0, len1 = 0;
-  UInt32 cutValue = _cutValue;
-  unsigned maxLen = (unsigned)_maxLen;
-
-  /*
-  if (size > 1)
-  {
-    UInt32 delta = *hash;
-    if (delta < _cyclicBufferSize)
-    {
-      UInt32 cyc1 = _cyclicBufferPos + 1;
-      CLzRef *pair = son + ((size_t)(cyc1 - delta + ((delta > cyc1) ? _cyclicBufferSize : 0)) << 1);
-      Byte b = *(cur + 1 - delta);
-      _distances[0] = pair[0];
-      _distances[1] = b;
-    }
-  }
-  */
-  if (cutValue == 0 || delta >= _cyclicBufferSize)
-  {
-    *ptr0 = *ptr1 = kEmptyHashValue;
-  }
-  else
-  for(;;)
-  {
-    {
-      CLzRef *pair = son + ((size_t)(_cyclicBufferPos - delta + ((_cyclicBufferPos < delta) ? _cyclicBufferSize : 0)) << 1);
-      const Byte *pb = cur - delta;
-      unsigned len = (len0 < len1 ? len0 : len1);
-      UInt32 pair0 = *pair;
-      if (pb[len] == cur[len])
-      {
-        if (++len != lenLimit && pb[len] == cur[len])
-          while (++len != lenLimit)
-            if (pb[len] != cur[len])
-              break;
-        if (maxLen < len)
-        {
-          maxLen = len;
-          *distances++ = (UInt32)len;
-          *distances++ = delta - 1;
-          if (len == lenLimit)
-          {
-            UInt32 pair1 = pair[1];
-            *ptr1 = pair0;
-            *ptr0 = pair1;
-            break;
-          }
-        }
-      }
-      {
-        UInt32 curMatch = pos - delta;
-        // delta = pos - *pair;
-        // delta = pos - pair[((UInt32)pb[len] - (UInt32)cur[len]) >> 31];
-        if (pb[len] < cur[len])
-        {
-          delta = pos - pair[1];
-          *ptr1 = curMatch;
-          ptr1 = pair + 1;
-          len1 = len;
-        }
-        else
-        {
-          delta = pos - *pair;
-          *ptr0 = curMatch;
-          ptr0 = pair;
-          len0 = len;
-        }
-      }
-    }
-    if (--cutValue == 0 || delta >= _cyclicBufferSize)
-    {
-      *ptr0 = *ptr1 = kEmptyHashValue;
-      break;
-    }
-  }
-  pos++;
-  _cyclicBufferPos++;
-  cur++;
-  {
-    UInt32 num = (UInt32)(distances - _distances);
-    _distances[-1] = num;
-  }
-  }
-  while (distances < limit && --size != 0);
-  *posRes = pos;
-  return distances;
-}
-
-#endif
-
-
-
-static void BtGetMatches(CMatchFinderMt *p, UInt32 *distances)
-{
-  UInt32 numProcessed = 0;
-  UInt32 curPos = 2;
-  UInt32 limit = kMtBtBlockSize - (p->matchMaxLen * 2); //  * 2
-  
-  distances[1] = p->hashNumAvail;
-  
-  while (curPos < limit)
-  {
-    if (p->hashBufPos == p->hashBufPosLimit)
-    {
-      MatchFinderMt_GetNextBlock_Hash(p);
-      distances[1] = numProcessed + p->hashNumAvail;
-      if (p->hashNumAvail >= p->numHashBytes)
-        continue;
-      distances[0] = curPos + p->hashNumAvail;
-      distances += curPos;
-      for (; p->hashNumAvail != 0; p->hashNumAvail--)
-        *distances++ = 0;
-      return;
-    }
-    {
-      UInt32 size = p->hashBufPosLimit - p->hashBufPos;
-      UInt32 lenLimit = p->matchMaxLen;
-      UInt32 pos = p->pos;
-      UInt32 cyclicBufferPos = p->cyclicBufferPos;
-      if (lenLimit >= p->hashNumAvail)
-        lenLimit = p->hashNumAvail;
-      {
-        UInt32 size2 = p->hashNumAvail - lenLimit + 1;
-        if (size2 < size)
-          size = size2;
-        size2 = p->cyclicBufferSize - cyclicBufferPos;
-        if (size2 < size)
-          size = size2;
-      }
-      
-      #ifndef MFMT_GM_INLINE
-      while (curPos < limit && size-- != 0)
-      {
-        UInt32 *startDistances = distances + curPos;
-        UInt32 num = (UInt32)(GetMatchesSpec1(lenLimit, pos - p->hashBuf[p->hashBufPos++],
-            pos, p->buffer, p->son, cyclicBufferPos, p->cyclicBufferSize, p->cutValue,
-            startDistances + 1, p->numHashBytes - 1) - startDistances);
-        *startDistances = num - 1;
-        curPos += num;
-        cyclicBufferPos++;
-        pos++;
-        p->buffer++;
-      }
-      #else
-      {
-        UInt32 posRes;
-        curPos = (UInt32)(GetMatchesSpecN(lenLimit, pos, p->buffer, p->son, cyclicBufferPos, p->cyclicBufferSize, p->cutValue,
-            distances + curPos, p->numHashBytes - 1, p->hashBuf + p->hashBufPos,
-            distances + limit,
-            size, &posRes) - distances);
-        p->hashBufPos += posRes - pos;
-        cyclicBufferPos += posRes - pos;
-        p->buffer += posRes - pos;
-        pos = posRes;
-      }
-      #endif
-
-      numProcessed += pos - p->pos;
-      p->hashNumAvail -= pos - p->pos;
-      p->pos = pos;
-      if (cyclicBufferPos == p->cyclicBufferSize)
-        cyclicBufferPos = 0;
-      p->cyclicBufferPos = cyclicBufferPos;
-    }
-  }
-  
-  distances[0] = curPos;
-}
-
-static void BtFillBlock(CMatchFinderMt *p, UInt32 globalBlockIndex)
-{
-  CMtSync *sync = &p->hashSync;
-  if (!sync->needStart)
-  {
-    CriticalSection_Enter(&sync->cs);
-    sync->csWasEntered = True;
-  }
-  
-  BtGetMatches(p, p->btBuf + (globalBlockIndex & kMtBtNumBlocksMask) * kMtBtBlockSize);
-
-  if (p->pos > kMtMaxValForNormalize - kMtBtBlockSize)
-  {
-    UInt32 subValue = p->pos - p->cyclicBufferSize;
-    MatchFinder_Normalize3(subValue, p->son, (size_t)p->cyclicBufferSize * 2);
-    p->pos -= subValue;
-  }
-
-  if (!sync->needStart)
-  {
-    CriticalSection_Leave(&sync->cs);
-    sync->csWasEntered = False;
-  }
-}
-
-void BtThreadFunc(CMatchFinderMt *mt)
-{
-  CMtSync *p = &mt->btSync;
-  for (;;)
-  {
-    UInt32 blockIndex = 0;
-    Event_Wait(&p->canStart);
-    Event_Set(&p->wasStarted);
-    for (;;)
-    {
-      if (p->exit)
-        return;
-      if (p->stopWriting)
-      {
-        p->numProcessedBlocks = blockIndex;
-        MtSync_StopWriting(&mt->hashSync);
-        Event_Set(&p->wasStopped);
-        break;
-      }
-      Semaphore_Wait(&p->freeSemaphore);
-      BtFillBlock(mt, blockIndex++);
-      Semaphore_Release1(&p->filledSemaphore);
-    }
-  }
-}
-
-void MatchFinderMt_Construct(CMatchFinderMt *p)
-{
-  p->hashBuf = NULL;
-  MtSync_Construct(&p->hashSync);
-  MtSync_Construct(&p->btSync);
-}
-
-static void MatchFinderMt_FreeMem(CMatchFinderMt *p, ISzAllocPtr alloc)
-{
-  ISzAlloc_Free(alloc, p->hashBuf);
-  p->hashBuf = NULL;
-}
-
-void MatchFinderMt_Destruct(CMatchFinderMt *p, ISzAllocPtr alloc)
-{
-  MtSync_Destruct(&p->hashSync);
-  MtSync_Destruct(&p->btSync);
-  MatchFinderMt_FreeMem(p, alloc);
-}
-
-#define kHashBufferSize (kMtHashBlockSize * kMtHashNumBlocks)
-#define kBtBufferSize (kMtBtBlockSize * kMtBtNumBlocks)
-
-static THREAD_FUNC_RET_TYPE THREAD_FUNC_CALL_TYPE HashThreadFunc2(void *p) { HashThreadFunc((CMatchFinderMt *)p);  return 0; }
-static THREAD_FUNC_RET_TYPE THREAD_FUNC_CALL_TYPE BtThreadFunc2(void *p)
-{
-  Byte allocaDummy[0x180];
-  unsigned i = 0;
-  for (i = 0; i < 16; i++)
-    allocaDummy[i] = (Byte)0;
-  if (allocaDummy[0] == 0)
-    BtThreadFunc((CMatchFinderMt *)p);
-  return 0;
-}
-
-SRes MatchFinderMt_Create(CMatchFinderMt *p, UInt32 historySize, UInt32 keepAddBufferBefore,
-    UInt32 matchMaxLen, UInt32 keepAddBufferAfter, ISzAllocPtr alloc)
-{
-  CMatchFinder *mf = p->MatchFinder;
-  p->historySize = historySize;
-  if (kMtBtBlockSize <= matchMaxLen * 4)
-    return SZ_ERROR_PARAM;
-  if (!p->hashBuf)
-  {
-    p->hashBuf = (UInt32 *)ISzAlloc_Alloc(alloc, (kHashBufferSize + kBtBufferSize) * sizeof(UInt32));
-    if (!p->hashBuf)
-      return SZ_ERROR_MEM;
-    p->btBuf = p->hashBuf + kHashBufferSize;
-  }
-  keepAddBufferBefore += (kHashBufferSize + kBtBufferSize);
-  keepAddBufferAfter += kMtHashBlockSize;
-  if (!MatchFinder_Create(mf, historySize, keepAddBufferBefore, matchMaxLen, keepAddBufferAfter, alloc))
-    return SZ_ERROR_MEM;
-
-  RINOK(MtSync_Create(&p->hashSync, HashThreadFunc2, p, kMtHashNumBlocks));
-  RINOK(MtSync_Create(&p->btSync, BtThreadFunc2, p, kMtBtNumBlocks));
-  return SZ_OK;
-}
-
-/* Call it after ReleaseStream / SetStream */
-static void MatchFinderMt_Init(CMatchFinderMt *p)
-{
-  CMatchFinder *mf = p->MatchFinder;
-  
-  p->btBufPos =
-  p->btBufPosLimit = 0;
-  p->hashBufPos =
-  p->hashBufPosLimit = 0;
-
-  /* Init without data reading. We don't want to read data in this thread */
-  MatchFinder_Init_3(mf, False);
-  MatchFinder_Init_LowHash(mf);
-  
-  p->pointerToCurPos = Inline_MatchFinder_GetPointerToCurrentPos(mf);
-  p->btNumAvailBytes = 0;
-  p->lzPos = p->historySize + 1;
-
-  p->hash = mf->hash;
-  p->fixedHashSize = mf->fixedHashSize;
-  p->crc = mf->crc;
-
-  p->son = mf->son;
-  p->matchMaxLen = mf->matchMaxLen;
-  p->numHashBytes = mf->numHashBytes;
-  p->pos = mf->pos;
-  p->buffer = mf->buffer;
-  p->cyclicBufferPos = mf->cyclicBufferPos;
-  p->cyclicBufferSize = mf->cyclicBufferSize;
-  p->cutValue = mf->cutValue;
-}
-
-/* ReleaseStream is required to finish multithreading */
-void MatchFinderMt_ReleaseStream(CMatchFinderMt *p)
-{
-  MtSync_StopWriting(&p->btSync);
-  /* p->MatchFinder->ReleaseStream(); */
-}
-
-static void MatchFinderMt_Normalize(CMatchFinderMt *p)
-{
-  MatchFinder_Normalize3(p->lzPos - p->historySize - 1, p->hash, p->fixedHashSize);
-  p->lzPos = p->historySize + 1;
-}
-
-static void MatchFinderMt_GetNextBlock_Bt(CMatchFinderMt *p)
-{
-  UInt32 blockIndex;
-  MtSync_GetNextBlock(&p->btSync);
-  blockIndex = ((p->btSync.numProcessedBlocks - 1) & kMtBtNumBlocksMask);
-  p->btBufPosLimit = p->btBufPos = blockIndex * kMtBtBlockSize;
-  p->btBufPosLimit += p->btBuf[p->btBufPos++];
-  p->btNumAvailBytes = p->btBuf[p->btBufPos++];
-  if (p->lzPos >= kMtMaxValForNormalize - kMtBtBlockSize)
-    MatchFinderMt_Normalize(p);
-}
-
-static const Byte * MatchFinderMt_GetPointerToCurrentPos(CMatchFinderMt *p)
-{
-  return p->pointerToCurPos;
-}
-
-#define GET_NEXT_BLOCK_IF_REQUIRED if (p->btBufPos == p->btBufPosLimit) MatchFinderMt_GetNextBlock_Bt(p);
-
-static UInt32 MatchFinderMt_GetNumAvailableBytes(CMatchFinderMt *p)
-{
-  GET_NEXT_BLOCK_IF_REQUIRED;
-  return p->btNumAvailBytes;
-}
-
-static UInt32 * MixMatches2(CMatchFinderMt *p, UInt32 matchMinPos, UInt32 *distances)
-{
-  UInt32 h2, curMatch2;
-  UInt32 *hash = p->hash;
-  const Byte *cur = p->pointerToCurPos;
-  UInt32 lzPos = p->lzPos;
-  MT_HASH2_CALC
-      
-  curMatch2 = hash[h2];
-  hash[h2] = lzPos;
-
-  if (curMatch2 >= matchMinPos)
-    if (cur[(ptrdiff_t)curMatch2 - lzPos] == cur[0])
-    {
-      *distances++ = 2;
-      *distances++ = lzPos - curMatch2 - 1;
-    }
-  
-  return distances;
-}
-
-static UInt32 * MixMatches3(CMatchFinderMt *p, UInt32 matchMinPos, UInt32 *distances)
-{
-  UInt32 h2, h3, curMatch2, curMatch3;
-  UInt32 *hash = p->hash;
-  const Byte *cur = p->pointerToCurPos;
-  UInt32 lzPos = p->lzPos;
-  MT_HASH3_CALC
-
-  curMatch2 = hash[                h2];
-  curMatch3 = (hash + kFix3HashSize)[h3];
-  
-  hash[                h2] = lzPos;
-  (hash + kFix3HashSize)[h3] = lzPos;
-
-  if (curMatch2 >= matchMinPos && cur[(ptrdiff_t)curMatch2 - lzPos] == cur[0])
-  {
-    distances[1] = lzPos - curMatch2 - 1;
-    if (cur[(ptrdiff_t)curMatch2 - lzPos + 2] == cur[2])
-    {
-      distances[0] = 3;
-      return distances + 2;
-    }
-    distances[0] = 2;
-    distances += 2;
-  }
-  
-  if (curMatch3 >= matchMinPos && cur[(ptrdiff_t)curMatch3 - lzPos] == cur[0])
-  {
-    *distances++ = 3;
-    *distances++ = lzPos - curMatch3 - 1;
-  }
-  
-  return distances;
-}
-
-/*
-static UInt32 *MixMatches4(CMatchFinderMt *p, UInt32 matchMinPos, UInt32 *distances)
-{
-  UInt32 h2, h3, h4, curMatch2, curMatch3, curMatch4;
-  UInt32 *hash = p->hash;
-  const Byte *cur = p->pointerToCurPos;
-  UInt32 lzPos = p->lzPos;
-  MT_HASH4_CALC
-      
-  curMatch2 = hash[                h2];
-  curMatch3 = (hash + kFix3HashSize)[h3];
-  curMatch4 = (hash + kFix4HashSize)[h4];
-  
-  hash[                h2] = lzPos;
-  (hash + kFix3HashSize)[h3] = lzPos;
-  (hash + kFix4HashSize)[h4] = lzPos;
-
-  if (curMatch2 >= matchMinPos && cur[(ptrdiff_t)curMatch2 - lzPos] == cur[0])
-  {
-    distances[1] = lzPos - curMatch2 - 1;
-    if (cur[(ptrdiff_t)curMatch2 - lzPos + 2] == cur[2])
-    {
-      distances[0] = (cur[(ptrdiff_t)curMatch2 - lzPos + 3] == cur[3]) ? 4 : 3;
-      return distances + 2;
-    }
-    distances[0] = 2;
-    distances += 2;
-  }
-  
-  if (curMatch3 >= matchMinPos && cur[(ptrdiff_t)curMatch3 - lzPos] == cur[0])
-  {
-    distances[1] = lzPos - curMatch3 - 1;
-    if (cur[(ptrdiff_t)curMatch3 - lzPos + 3] == cur[3])
-    {
-      distances[0] = 4;
-      return distances + 2;
-    }
-    distances[0] = 3;
-    distances += 2;
-  }
-
-  if (curMatch4 >= matchMinPos)
-    if (
-      cur[(ptrdiff_t)curMatch4 - lzPos] == cur[0] &&
-      cur[(ptrdiff_t)curMatch4 - lzPos + 3] == cur[3]
-      )
-    {
-      *distances++ = 4;
-      *distances++ = lzPos - curMatch4 - 1;
-    }
-  
-  return distances;
-}
-*/
-
-#define INCREASE_LZ_POS p->lzPos++; p->pointerToCurPos++;
-
-static UInt32 MatchFinderMt2_GetMatches(CMatchFinderMt *p, UInt32 *distances)
-{
-  const UInt32 *btBuf = p->btBuf + p->btBufPos;
-  UInt32 len = *btBuf++;
-  p->btBufPos += 1 + len;
-  p->btNumAvailBytes--;
-  {
-    UInt32 i;
-    for (i = 0; i < len; i += 2)
-    {
-      UInt32 v0 = btBuf[0];
-      UInt32 v1 = btBuf[1];
-      btBuf += 2;
-      distances[0] = v0;
-      distances[1] = v1;
-      distances += 2;
-    }
-  }
-  INCREASE_LZ_POS
-  return len;
-}
-
-static UInt32 MatchFinderMt_GetMatches(CMatchFinderMt *p, UInt32 *distances)
-{
-  const UInt32 *btBuf = p->btBuf + p->btBufPos;
-  UInt32 len = *btBuf++;
-  p->btBufPos += 1 + len;
-
-  if (len == 0)
-  {
-    /* change for bt5 ! */
-    if (p->btNumAvailBytes-- >= 4)
-      len = (UInt32)(p->MixMatchesFunc(p, p->lzPos - p->historySize, distances) - (distances));
-  }
-  else
-  {
-    /* Condition: there are matches in btBuf with length < p->numHashBytes */
-    UInt32 *distances2;
-    p->btNumAvailBytes--;
-    distances2 = p->MixMatchesFunc(p, p->lzPos - btBuf[1], distances);
-    do
-    {
-      UInt32 v0 = btBuf[0];
-      UInt32 v1 = btBuf[1];
-      btBuf += 2;
-      distances2[0] = v0;
-      distances2[1] = v1;
-      distances2 += 2;
-    }
-    while ((len -= 2) != 0);
-    len = (UInt32)(distances2 - (distances));
-  }
-  INCREASE_LZ_POS
-  return len;
-}
-
-#define SKIP_HEADER2_MT  do { GET_NEXT_BLOCK_IF_REQUIRED
-#define SKIP_HEADER_MT(n) SKIP_HEADER2_MT if (p->btNumAvailBytes-- >= (n)) { const Byte *cur = p->pointerToCurPos; UInt32 *hash = p->hash;
-#define SKIP_FOOTER_MT } INCREASE_LZ_POS p->btBufPos += p->btBuf[p->btBufPos] + 1; } while (--num != 0);
-
-static void MatchFinderMt0_Skip(CMatchFinderMt *p, UInt32 num)
-{
-  SKIP_HEADER2_MT { p->btNumAvailBytes--;
-  SKIP_FOOTER_MT
-}
-
-static void MatchFinderMt2_Skip(CMatchFinderMt *p, UInt32 num)
-{
-  SKIP_HEADER_MT(2)
-      UInt32 h2;
-      MT_HASH2_CALC
-      hash[h2] = p->lzPos;
-  SKIP_FOOTER_MT
-}
-
-static void MatchFinderMt3_Skip(CMatchFinderMt *p, UInt32 num)
-{
-  SKIP_HEADER_MT(3)
-      UInt32 h2, h3;
-      MT_HASH3_CALC
-      (hash + kFix3HashSize)[h3] =
-      hash[                h2] =
-        p->lzPos;
-  SKIP_FOOTER_MT
-}
-
-/*
-static void MatchFinderMt4_Skip(CMatchFinderMt *p, UInt32 num)
-{
-  SKIP_HEADER_MT(4)
-      UInt32 h2, h3, h4;
-      MT_HASH4_CALC
-      (hash + kFix4HashSize)[h4] =
-      (hash + kFix3HashSize)[h3] =
-      hash[                h2] =
-        p->lzPos;
-  SKIP_FOOTER_MT
-}
-*/
-
-void MatchFinderMt_CreateVTable(CMatchFinderMt *p, IMatchFinder *vTable)
-{
-  vTable->Init = (Mf_Init_Func)MatchFinderMt_Init;
-  vTable->GetNumAvailableBytes = (Mf_GetNumAvailableBytes_Func)MatchFinderMt_GetNumAvailableBytes;
-  vTable->GetPointerToCurrentPos = (Mf_GetPointerToCurrentPos_Func)MatchFinderMt_GetPointerToCurrentPos;
-  vTable->GetMatches = (Mf_GetMatches_Func)MatchFinderMt_GetMatches;
-  
-  switch (p->MatchFinder->numHashBytes)
-  {
-    case 2:
-      p->GetHeadsFunc = GetHeads2;
-      p->MixMatchesFunc = (Mf_Mix_Matches)NULL;
-      vTable->Skip = (Mf_Skip_Func)MatchFinderMt0_Skip;
-      vTable->GetMatches = (Mf_GetMatches_Func)MatchFinderMt2_GetMatches;
-      break;
-    case 3:
-      p->GetHeadsFunc = GetHeads3;
-      p->MixMatchesFunc = (Mf_Mix_Matches)MixMatches2;
-      vTable->Skip = (Mf_Skip_Func)MatchFinderMt2_Skip;
-      break;
-    default:
-    /* case 4: */
-      p->GetHeadsFunc = p->MatchFinder->bigHash ? GetHeads4b : GetHeads4;
-      p->MixMatchesFunc = (Mf_Mix_Matches)MixMatches3;
-      vTable->Skip = (Mf_Skip_Func)MatchFinderMt3_Skip;
-      break;
-    /*
-    default:
-      p->GetHeadsFunc = GetHeads5;
-      p->MixMatchesFunc = (Mf_Mix_Matches)MixMatches4;
-      vTable->Skip = (Mf_Skip_Func)MatchFinderMt4_Skip;
-      break;
-    */
-  }
-}
@@ -1,57 +0,0 @@
-/* LzHash.h -- HASH functions for LZ algorithms
-2015-04-12 : Igor Pavlov : Public domain */
-
-#ifndef __LZ_HASH_H
-#define __LZ_HASH_H
-
-#define kHash2Size (1 << 10)
-#define kHash3Size (1 << 16)
-#define kHash4Size (1 << 20)
-
-#define kFix3HashSize (kHash2Size)
-#define kFix4HashSize (kHash2Size + kHash3Size)
-#define kFix5HashSize (kHash2Size + kHash3Size + kHash4Size)
-
-#define HASH2_CALC hv = cur[0] | ((UInt32)cur[1] << 8);
-
-#define HASH3_CALC { \
-  UInt32 temp = p->crc[cur[0]] ^ cur[1]; \
-  h2 = temp & (kHash2Size - 1); \
-  hv = (temp ^ ((UInt32)cur[2] << 8)) & p->hashMask; }
-
-#define HASH4_CALC { \
-  UInt32 temp = p->crc[cur[0]] ^ cur[1]; \
-  h2 = temp & (kHash2Size - 1); \
-  temp ^= ((UInt32)cur[2] << 8); \
-  h3 = temp & (kHash3Size - 1); \
-  hv = (temp ^ (p->crc[cur[3]] << 5)) & p->hashMask; }
-
-#define HASH5_CALC { \
-  UInt32 temp = p->crc[cur[0]] ^ cur[1]; \
-  h2 = temp & (kHash2Size - 1); \
-  temp ^= ((UInt32)cur[2] << 8); \
-  h3 = temp & (kHash3Size - 1); \
-  temp ^= (p->crc[cur[3]] << 5); \
-  h4 = temp & (kHash4Size - 1); \
-  hv = (temp ^ (p->crc[cur[4]] << 3)) & p->hashMask; }
-
-/* #define HASH_ZIP_CALC hv = ((cur[0] | ((UInt32)cur[1] << 8)) ^ p->crc[cur[2]]) & 0xFFFF; */
-#define HASH_ZIP_CALC hv = ((cur[2] | ((UInt32)cur[0] << 8)) ^ p->crc[cur[1]]) & 0xFFFF;
-
-
-#define MT_HASH2_CALC \
-  h2 = (p->crc[cur[0]] ^ cur[1]) & (kHash2Size - 1);
-
-#define MT_HASH3_CALC { \
-  UInt32 temp = p->crc[cur[0]] ^ cur[1]; \
-  h2 = temp & (kHash2Size - 1); \
-  h3 = (temp ^ ((UInt32)cur[2] << 8)) & (kHash3Size - 1); }
-
-#define MT_HASH4_CALC { \
-  UInt32 temp = p->crc[cur[0]] ^ cur[1]; \
-  h2 = temp & (kHash2Size - 1); \
-  temp ^= ((UInt32)cur[2] << 8); \
-  h3 = temp & (kHash3Size - 1); \
-  h4 = (temp ^ (p->crc[cur[3]] << 5)) & (kHash4Size - 1); }
-
-#endif
@@ -1,85 +0,0 @@
-/* Ppmd.h -- PPMD codec common code
-2017-04-03 : Igor Pavlov : Public domain
-This code is based on PPMd var.H (2001): Dmitry Shkarin : Public domain */
-
-#ifndef __PPMD_H
-#define __PPMD_H
-
-#include "CpuArch.h"
-
-EXTERN_C_BEGIN
-
-#ifdef MY_CPU_32BIT
-  #define PPMD_32BIT
-#endif
-
-#define PPMD_INT_BITS 7
-#define PPMD_PERIOD_BITS 7
-#define PPMD_BIN_SCALE (1 << (PPMD_INT_BITS + PPMD_PERIOD_BITS))
-
-#define PPMD_GET_MEAN_SPEC(summ, shift, round) (((summ) + (1 << ((shift) - (round)))) >> (shift))
-#define PPMD_GET_MEAN(summ) PPMD_GET_MEAN_SPEC((summ), PPMD_PERIOD_BITS, 2)
-#define PPMD_UPDATE_PROB_0(prob) ((prob) + (1 << PPMD_INT_BITS) - PPMD_GET_MEAN(prob))
-#define PPMD_UPDATE_PROB_1(prob) ((prob) - PPMD_GET_MEAN(prob))
-
-#define PPMD_N1 4
-#define PPMD_N2 4
-#define PPMD_N3 4
-#define PPMD_N4 ((128 + 3 - 1 * PPMD_N1 - 2 * PPMD_N2 - 3 * PPMD_N3) / 4)
-#define PPMD_NUM_INDEXES (PPMD_N1 + PPMD_N2 + PPMD_N3 + PPMD_N4)
-
-#pragma pack(push, 1)
-/* Most compilers works OK here even without #pragma pack(push, 1), but some GCC compilers need it. */
-
-/* SEE-contexts for PPM-contexts with masked symbols */
-typedef struct
-{
-  UInt16 Summ; /* Freq */
-  Byte Shift;  /* Speed of Freq change; low Shift is for fast change */
-  Byte Count;  /* Count to next change of Shift */
-} CPpmd_See;
-
-#define Ppmd_See_Update(p)  if ((p)->Shift < PPMD_PERIOD_BITS && --(p)->Count == 0) \
-    { (p)->Summ <<= 1; (p)->Count = (Byte)(3 << (p)->Shift++); }
-
-typedef struct
-{
-  Byte Symbol;
-  Byte Freq;
-  UInt16 SuccessorLow;
-  UInt16 SuccessorHigh;
-} CPpmd_State;
-
-#pragma pack(pop)
-
-typedef
-  #ifdef PPMD_32BIT
-    CPpmd_State *
-  #else
-    UInt32
-  #endif
-  CPpmd_State_Ref;
-
-typedef
-  #ifdef PPMD_32BIT
-    void *
-  #else
-    UInt32
-  #endif
-  CPpmd_Void_Ref;
-
-typedef
-  #ifdef PPMD_32BIT
-    Byte *
-  #else
-    UInt32
-  #endif
-  CPpmd_Byte_Ref;
-
-#define PPMD_SetAllBitsIn256Bytes(p) \
-  { size_t z; for (z = 0; z < 256 / sizeof(p[0]); z += 8) { \
-  p[z+7] = p[z+6] = p[z+5] = p[z+4] = p[z+3] = p[z+2] = p[z+1] = p[z+0] = ~(size_t)0; }}
-
-EXTERN_C_END
- 
-#endif
@@ -1,712 +0,0 @@
-/* Ppmd7.c -- PPMdH codec
-2018-07-04 : Igor Pavlov : Public domain
-This code is based on PPMd var.H (2001): Dmitry Shkarin : Public domain */
-
-#include "Precomp.h"
-
-#include <string.h>
-
-#include "Ppmd7.h"
-
-const Byte PPMD7_kExpEscape[16] = { 25, 14, 9, 7, 5, 5, 4, 4, 4, 3, 3, 3, 2, 2, 2, 2 };
-static const UInt16 kInitBinEsc[] = { 0x3CDD, 0x1F3F, 0x59BF, 0x48F3, 0x64A1, 0x5ABC, 0x6632, 0x6051};
-
-#define MAX_FREQ 124
-#define UNIT_SIZE 12
-
-#define U2B(nu) ((UInt32)(nu) * UNIT_SIZE)
-#define U2I(nu) (p->Units2Indx[(size_t)(nu) - 1])
-#define I2U(indx) (p->Indx2Units[indx])
-
-#ifdef PPMD_32BIT
-  #define REF(ptr) (ptr)
-#else
-  #define REF(ptr) ((UInt32)((Byte *)(ptr) - (p)->Base))
-#endif
-
-#define STATS_REF(ptr) ((CPpmd_State_Ref)REF(ptr))
-
-#define CTX(ref) ((CPpmd7_Context *)Ppmd7_GetContext(p, ref))
-#define STATS(ctx) Ppmd7_GetStats(p, ctx)
-#define ONE_STATE(ctx) Ppmd7Context_OneState(ctx)
-#define SUFFIX(ctx) CTX((ctx)->Suffix)
-
-typedef CPpmd7_Context * CTX_PTR;
-
-struct CPpmd7_Node_;
-
-typedef
-  #ifdef PPMD_32BIT
-    struct CPpmd7_Node_ *
-  #else
-    UInt32
-  #endif
-  CPpmd7_Node_Ref;
-
-typedef struct CPpmd7_Node_
-{
-  UInt16 Stamp; /* must be at offset 0 as CPpmd7_Context::NumStats. Stamp=0 means free */
-  UInt16 NU;
-  CPpmd7_Node_Ref Next; /* must be at offset >= 4 */
-  CPpmd7_Node_Ref Prev;
-} CPpmd7_Node;
-
-#ifdef PPMD_32BIT
-  #define NODE(ptr) (ptr)
-#else
-  #define NODE(offs) ((CPpmd7_Node *)(p->Base + (offs)))
-#endif
-
-void Ppmd7_Construct(CPpmd7 *p)
-{
-  unsigned i, k, m;
-
-  p->Base = 0;
-
-  for (i = 0, k = 0; i < PPMD_NUM_INDEXES; i++)
-  {
-    unsigned step = (i >= 12 ? 4 : (i >> 2) + 1);
-    do { p->Units2Indx[k++] = (Byte)i; } while (--step);
-    p->Indx2Units[i] = (Byte)k;
-  }
-
-  p->NS2BSIndx[0] = (0 << 1);
-  p->NS2BSIndx[1] = (1 << 1);
-  memset(p->NS2BSIndx + 2, (2 << 1), 9);
-  memset(p->NS2BSIndx + 11, (3 << 1), 256 - 11);
-
-  for (i = 0; i < 3; i++)
-    p->NS2Indx[i] = (Byte)i;
-  for (m = i, k = 1; i < 256; i++)
-  {
-    p->NS2Indx[i] = (Byte)m;
-    if (--k == 0)
-      k = (++m) - 2;
-  }
-
-  memset(p->HB2Flag, 0, 0x40);
-  memset(p->HB2Flag + 0x40, 8, 0x100 - 0x40);
-}
-
-void Ppmd7_Free(CPpmd7 *p, ISzAllocPtr alloc)
-{
-  ISzAlloc_Free(alloc, p->Base);
-  p->Size = 0;
-  p->Base = 0;
-}
-
-BoolInt Ppmd7_Alloc(CPpmd7 *p, UInt32 size, ISzAllocPtr alloc)
-{
-  if (!p->Base || p->Size != size)
-  {
-    size_t size2;
-    Ppmd7_Free(p, alloc);
-    size2 = 0
-      #ifndef PPMD_32BIT
-      + UNIT_SIZE
-      #endif
-      ;
-    p->AlignOffset =
-      #ifdef PPMD_32BIT
-        (4 - size) & 3;
-      #else
-        4 - (size & 3);
-      #endif
-    if ((p->Base = (Byte *)ISzAlloc_Alloc(alloc, p->AlignOffset + size + size2)) == 0)
-      return False;
-    p->Size = size;
-  }
-  return True;
-}
-
-static void InsertNode(CPpmd7 *p, void *node, unsigned indx)
-{
-  *((CPpmd_Void_Ref *)node) = p->FreeList[indx];
-  p->FreeList[indx] = REF(node);
-}
-
-static void *RemoveNode(CPpmd7 *p, unsigned indx)
-{
-  CPpmd_Void_Ref *node = (CPpmd_Void_Ref *)Ppmd7_GetPtr(p, p->FreeList[indx]);
-  p->FreeList[indx] = *node;
-  return node;
-}
-
-static void SplitBlock(CPpmd7 *p, void *ptr, unsigned oldIndx, unsigned newIndx)
-{
-  unsigned i, nu = I2U(oldIndx) - I2U(newIndx);
-  ptr = (Byte *)ptr + U2B(I2U(newIndx));
-  if (I2U(i = U2I(nu)) != nu)
-  {
-    unsigned k = I2U(--i);
-    InsertNode(p, ((Byte *)ptr) + U2B(k), nu - k - 1);
-  }
-  InsertNode(p, ptr, i);
-}
-
-static void GlueFreeBlocks(CPpmd7 *p)
-{
-  #ifdef PPMD_32BIT
-  CPpmd7_Node headItem;
-  CPpmd7_Node_Ref head = &headItem;
-  #else
-  CPpmd7_Node_Ref head = p->AlignOffset + p->Size;
-  #endif
-  
-  CPpmd7_Node_Ref n = head;
-  unsigned i;
-
-  p->GlueCount = 255;
-
-  /* create doubly-linked list of free blocks */
-  for (i = 0; i < PPMD_NUM_INDEXES; i++)
-  {
-    UInt16 nu = I2U(i);
-    CPpmd7_Node_Ref next = (CPpmd7_Node_Ref)p->FreeList[i];
-    p->FreeList[i] = 0;
-    while (next != 0)
-    {
-      CPpmd7_Node *node = NODE(next);
-      node->Next = n;
-      n = NODE(n)->Prev = next;
-      next = *(const CPpmd7_Node_Ref *)node;
-      node->Stamp = 0;
-      node->NU = (UInt16)nu;
-    }
-  }
-  NODE(head)->Stamp = 1;
-  NODE(head)->Next = n;
-  NODE(n)->Prev = head;
-  if (p->LoUnit != p->HiUnit)
-    ((CPpmd7_Node *)p->LoUnit)->Stamp = 1;
-  
-  /* Glue free blocks */
-  while (n != head)
-  {
-    CPpmd7_Node *node = NODE(n);
-    UInt32 nu = (UInt32)node->NU;
-    for (;;)
-    {
-      CPpmd7_Node *node2 = NODE(n) + nu;
-      nu += node2->NU;
-      if (node2->Stamp != 0 || nu >= 0x10000)
-        break;
-      NODE(node2->Prev)->Next = node2->Next;
-      NODE(node2->Next)->Prev = node2->Prev;
-      node->NU = (UInt16)nu;
-    }
-    n = node->Next;
-  }
-  
-  /* Fill lists of free blocks */
-  for (n = NODE(head)->Next; n != head;)
-  {
-    CPpmd7_Node *node = NODE(n);
-    unsigned nu;
-    CPpmd7_Node_Ref next = node->Next;
-    for (nu = node->NU; nu > 128; nu -= 128, node += 128)
-      InsertNode(p, node, PPMD_NUM_INDEXES - 1);
-    if (I2U(i = U2I(nu)) != nu)
-    {
-      unsigned k = I2U(--i);
-      InsertNode(p, node + k, nu - k - 1);
-    }
-    InsertNode(p, node, i);
-    n = next;
-  }
-}
-
-static void *AllocUnitsRare(CPpmd7 *p, unsigned indx)
-{
-  unsigned i;
-  void *retVal;
-  if (p->GlueCount == 0)
-  {
-    GlueFreeBlocks(p);
-    if (p->FreeList[indx] != 0)
-      return RemoveNode(p, indx);
-  }
-  i = indx;
-  do
-  {
-    if (++i == PPMD_NUM_INDEXES)
-    {
-      UInt32 numBytes = U2B(I2U(indx));
-      p->GlueCount--;
-      return ((UInt32)(p->UnitsStart - p->Text) > numBytes) ? (p->UnitsStart -= numBytes) : (NULL);
-    }
-  }
-  while (p->FreeList[i] == 0);
-  retVal = RemoveNode(p, i);
-  SplitBlock(p, retVal, i, indx);
-  return retVal;
-}
-
-static void *AllocUnits(CPpmd7 *p, unsigned indx)
-{
-  UInt32 numBytes;
-  if (p->FreeList[indx] != 0)
-    return RemoveNode(p, indx);
-  numBytes = U2B(I2U(indx));
-  if (numBytes <= (UInt32)(p->HiUnit - p->LoUnit))
-  {
-    void *retVal = p->LoUnit;
-    p->LoUnit += numBytes;
-    return retVal;
-  }
-  return AllocUnitsRare(p, indx);
-}
-
-#define MyMem12Cpy(dest, src, num) \
-  { UInt32 *d = (UInt32 *)dest; const UInt32 *s = (const UInt32 *)src; UInt32 n = num; \
-    do { d[0] = s[0]; d[1] = s[1]; d[2] = s[2]; s += 3; d += 3; } while (--n); }
-
-static void *ShrinkUnits(CPpmd7 *p, void *oldPtr, unsigned oldNU, unsigned newNU)
-{
-  unsigned i0 = U2I(oldNU);
-  unsigned i1 = U2I(newNU);
-  if (i0 == i1)
-    return oldPtr;
-  if (p->FreeList[i1] != 0)
-  {
-    void *ptr = RemoveNode(p, i1);
-    MyMem12Cpy(ptr, oldPtr, newNU);
-    InsertNode(p, oldPtr, i0);
-    return ptr;
-  }
-  SplitBlock(p, oldPtr, i0, i1);
-  return oldPtr;
-}
-
-#define SUCCESSOR(p) ((CPpmd_Void_Ref)((p)->SuccessorLow | ((UInt32)(p)->SuccessorHigh << 16)))
-
-static void SetSuccessor(CPpmd_State *p, CPpmd_Void_Ref v)
-{
-  (p)->SuccessorLow = (UInt16)((UInt32)(v) & 0xFFFF);
-  (p)->SuccessorHigh = (UInt16)(((UInt32)(v) >> 16) & 0xFFFF);
-}
-
-static void RestartModel(CPpmd7 *p)
-{
-  unsigned i, k, m;
-
-  memset(p->FreeList, 0, sizeof(p->FreeList));
-  p->Text = p->Base + p->AlignOffset;
-  p->HiUnit = p->Text + p->Size;
-  p->LoUnit = p->UnitsStart = p->HiUnit - p->Size / 8 / UNIT_SIZE * 7 * UNIT_SIZE;
-  p->GlueCount = 0;
-
-  p->OrderFall = p->MaxOrder;
-  p->RunLength = p->InitRL = -(Int32)((p->MaxOrder < 12) ? p->MaxOrder : 12) - 1;
-  p->PrevSuccess = 0;
-
-  p->MinContext = p->MaxContext = (CTX_PTR)(p->HiUnit -= UNIT_SIZE); /* AllocContext(p); */
-  p->MinContext->Suffix = 0;
-  p->MinContext->NumStats = 256;
-  p->MinContext->SummFreq = 256 + 1;
-  p->FoundState = (CPpmd_State *)p->LoUnit; /* AllocUnits(p, PPMD_NUM_INDEXES - 1); */
-  p->LoUnit += U2B(256 / 2);
-  p->MinContext->Stats = REF(p->FoundState);
-  for (i = 0; i < 256; i++)
-  {
-    CPpmd_State *s = &p->FoundState[i];
-    s->Symbol = (Byte)i;
-    s->Freq = 1;
-    SetSuccessor(s, 0);
-  }
-
-  for (i = 0; i < 128; i++)
-    for (k = 0; k < 8; k++)
-    {
-      UInt16 *dest = p->BinSumm[i] + k;
-      UInt16 val = (UInt16)(PPMD_BIN_SCALE - kInitBinEsc[k] / (i + 2));
-      for (m = 0; m < 64; m += 8)
-        dest[m] = val;
-    }
-  
-  for (i = 0; i < 25; i++)
-    for (k = 0; k < 16; k++)
-    {
-      CPpmd_See *s = &p->See[i][k];
-      s->Summ = (UInt16)((5 * i + 10) << (s->Shift = PPMD_PERIOD_BITS - 4));
-      s->Count = 4;
-    }
-}
-
-void Ppmd7_Init(CPpmd7 *p, unsigned maxOrder)
-{
-  p->MaxOrder = maxOrder;
-  RestartModel(p);
-  p->DummySee.Shift = PPMD_PERIOD_BITS;
-  p->DummySee.Summ = 0; /* unused */
-  p->DummySee.Count = 64; /* unused */
-}
-
-static CTX_PTR CreateSuccessors(CPpmd7 *p, BoolInt skip)
-{
-  CPpmd_State upState;
-  CTX_PTR c = p->MinContext;
-  CPpmd_Byte_Ref upBranch = (CPpmd_Byte_Ref)SUCCESSOR(p->FoundState);
-  CPpmd_State *ps[PPMD7_MAX_ORDER];
-  unsigned numPs = 0;
-  
-  if (!skip)
-    ps[numPs++] = p->FoundState;
-  
-  while (c->Suffix)
-  {
-    CPpmd_Void_Ref successor;
-    CPpmd_State *s;
-    c = SUFFIX(c);
-    if (c->NumStats != 1)
-    {
-      for (s = STATS(c); s->Symbol != p->FoundState->Symbol; s++);
-    }
-    else
-      s = ONE_STATE(c);
-    successor = SUCCESSOR(s);
-    if (successor != upBranch)
-    {
-      c = CTX(successor);
-      if (numPs == 0)
-        return c;
-      break;
-    }
-    ps[numPs++] = s;
-  }
-  
-  upState.Symbol = *(const Byte *)Ppmd7_GetPtr(p, upBranch);
-  SetSuccessor(&upState, upBranch + 1);
-  
-  if (c->NumStats == 1)
-    upState.Freq = ONE_STATE(c)->Freq;
-  else
-  {
-    UInt32 cf, s0;
-    CPpmd_State *s;
-    for (s = STATS(c); s->Symbol != upState.Symbol; s++);
-    cf = s->Freq - 1;
-    s0 = c->SummFreq - c->NumStats - cf;
-    upState.Freq = (Byte)(1 + ((2 * cf <= s0) ? (5 * cf > s0) : ((2 * cf + 3 * s0 - 1) / (2 * s0))));
-  }
-
-  do
-  {
-    /* Create Child */
-    CTX_PTR c1; /* = AllocContext(p); */
-    if (p->HiUnit != p->LoUnit)
-      c1 = (CTX_PTR)(p->HiUnit -= UNIT_SIZE);
-    else if (p->FreeList[0] != 0)
-      c1 = (CTX_PTR)RemoveNode(p, 0);
-    else
-    {
-      c1 = (CTX_PTR)AllocUnitsRare(p, 0);
-      if (!c1)
-        return NULL;
-    }
-    c1->NumStats = 1;
-    *ONE_STATE(c1) = upState;
-    c1->Suffix = REF(c);
-    SetSuccessor(ps[--numPs], REF(c1));
-    c = c1;
-  }
-  while (numPs != 0);
-  
-  return c;
-}
-
-static void SwapStates(CPpmd_State *t1, CPpmd_State *t2)
-{
-  CPpmd_State tmp = *t1;
-  *t1 = *t2;
-  *t2 = tmp;
-}
-
-static void UpdateModel(CPpmd7 *p)
-{
-  CPpmd_Void_Ref successor, fSuccessor = SUCCESSOR(p->FoundState);
-  CTX_PTR c;
-  unsigned s0, ns;
-  
-  if (p->FoundState->Freq < MAX_FREQ / 4 && p->MinContext->Suffix != 0)
-  {
-    c = SUFFIX(p->MinContext);
-    
-    if (c->NumStats == 1)
-    {
-      CPpmd_State *s = ONE_STATE(c);
-      if (s->Freq < 32)
-        s->Freq++;
-    }
-    else
-    {
-      CPpmd_State *s = STATS(c);
-      if (s->Symbol != p->FoundState->Symbol)
-      {
-        do { s++; } while (s->Symbol != p->FoundState->Symbol);
-        if (s[0].Freq >= s[-1].Freq)
-        {
-          SwapStates(&s[0], &s[-1]);
-          s--;
-        }
-      }
-      if (s->Freq < MAX_FREQ - 9)
-      {
-        s->Freq += 2;
-        c->SummFreq += 2;
-      }
-    }
-  }
-
-  if (p->OrderFall == 0)
-  {
-    p->MinContext = p->MaxContext = CreateSuccessors(p, True);
-    if (p->MinContext == 0)
-    {
-      RestartModel(p);
-      return;
-    }
-    SetSuccessor(p->FoundState, REF(p->MinContext));
-    return;
-  }
-  
-  *p->Text++ = p->FoundState->Symbol;
-  successor = REF(p->Text);
-  if (p->Text >= p->UnitsStart)
-  {
-    RestartModel(p);
-    return;
-  }
-  
-  if (fSuccessor)
-  {
-    if (fSuccessor <= successor)
-    {
-      CTX_PTR cs = CreateSuccessors(p, False);
-      if (cs == NULL)
-      {
-        RestartModel(p);
-        return;
-      }
-      fSuccessor = REF(cs);
-    }
-    if (--p->OrderFall == 0)
-    {
-      successor = fSuccessor;
-      p->Text -= (p->MaxContext != p->MinContext);
-    }
-  }
-  else
-  {
-    SetSuccessor(p->FoundState, successor);
-    fSuccessor = REF(p->MinContext);
-  }
-  
-  s0 = p->MinContext->SummFreq - (ns = p->MinContext->NumStats) - (p->FoundState->Freq - 1);
-  
-  for (c = p->MaxContext; c != p->MinContext; c = SUFFIX(c))
-  {
-    unsigned ns1;
-    UInt32 cf, sf;
-    if ((ns1 = c->NumStats) != 1)
-    {
-      if ((ns1 & 1) == 0)
-      {
-        /* Expand for one UNIT */
-        unsigned oldNU = ns1 >> 1;
-        unsigned i = U2I(oldNU);
-        if (i != U2I((size_t)oldNU + 1))
-        {
-          void *ptr = AllocUnits(p, i + 1);
-          void *oldPtr;
-          if (!ptr)
-          {
-            RestartModel(p);
-            return;
-          }
-          oldPtr = STATS(c);
-          MyMem12Cpy(ptr, oldPtr, oldNU);
-          InsertNode(p, oldPtr, i);
-          c->Stats = STATS_REF(ptr);
-        }
-      }
-      c->SummFreq = (UInt16)(c->SummFreq + (2 * ns1 < ns) + 2 * ((4 * ns1 <= ns) & (c->SummFreq <= 8 * ns1)));
-    }
-    else
-    {
-      CPpmd_State *s = (CPpmd_State*)AllocUnits(p, 0);
-      if (!s)
-      {
-        RestartModel(p);
-        return;
-      }
-      *s = *ONE_STATE(c);
-      c->Stats = REF(s);
-      if (s->Freq < MAX_FREQ / 4 - 1)
-        s->Freq <<= 1;
-      else
-        s->Freq = MAX_FREQ - 4;
-      c->SummFreq = (UInt16)(s->Freq + p->InitEsc + (ns > 3));
-    }
-    cf = 2 * (UInt32)p->FoundState->Freq * (c->SummFreq + 6);
-    sf = (UInt32)s0 + c->SummFreq;
-    if (cf < 6 * sf)
-    {
-      cf = 1 + (cf > sf) + (cf >= 4 * sf);
-      c->SummFreq += 3;
-    }
-    else
-    {
-      cf = 4 + (cf >= 9 * sf) + (cf >= 12 * sf) + (cf >= 15 * sf);
-      c->SummFreq = (UInt16)(c->SummFreq + cf);
-    }
-    {
-      CPpmd_State *s = STATS(c) + ns1;
-      SetSuccessor(s, successor);
-      s->Symbol = p->FoundState->Symbol;
-      s->Freq = (Byte)cf;
-      c->NumStats = (UInt16)(ns1 + 1);
-    }
-  }
-  p->MaxContext = p->MinContext = CTX(fSuccessor);
-}
-  
-static void Rescale(CPpmd7 *p)
-{
-  unsigned i, adder, sumFreq, escFreq;
-  CPpmd_State *stats = STATS(p->MinContext);
-  CPpmd_State *s = p->FoundState;
-  {
-    CPpmd_State tmp = *s;
-    for (; s != stats; s--)
-      s[0] = s[-1];
-    *s = tmp;
-  }
-  escFreq = p->MinContext->SummFreq - s->Freq;
-  s->Freq += 4;
-  adder = (p->OrderFall != 0);
-  s->Freq = (Byte)((s->Freq + adder) >> 1);
-  sumFreq = s->Freq;
-  
-  i = p->MinContext->NumStats - 1;
-  do
-  {
-    escFreq -= (++s)->Freq;
-    s->Freq = (Byte)((s->Freq + adder) >> 1);
-    sumFreq += s->Freq;
-    if (s[0].Freq > s[-1].Freq)
-    {
-      CPpmd_State *s1 = s;
-      CPpmd_State tmp = *s1;
-      do
-        s1[0] = s1[-1];
-      while (--s1 != stats && tmp.Freq > s1[-1].Freq);
-      *s1 = tmp;
-    }
-  }
-  while (--i);
-  
-  if (s->Freq == 0)
-  {
-    unsigned numStats = p->MinContext->NumStats;
-    unsigned n0, n1;
-    do { i++; } while ((--s)->Freq == 0);
-    escFreq += i;
-    p->MinContext->NumStats = (UInt16)(p->MinContext->NumStats - i);
-    if (p->MinContext->NumStats == 1)
-    {
-      CPpmd_State tmp = *stats;
-      do
-      {
-        tmp.Freq = (Byte)(tmp.Freq - (tmp.Freq >> 1));
-        escFreq >>= 1;
-      }
-      while (escFreq > 1);
-      InsertNode(p, stats, U2I(((numStats + 1) >> 1)));
-      *(p->FoundState = ONE_STATE(p->MinContext)) = tmp;
-      return;
-    }
-    n0 = (numStats + 1) >> 1;
-    n1 = (p->MinContext->NumStats + 1) >> 1;
-    if (n0 != n1)
-      p->MinContext->Stats = STATS_REF(ShrinkUnits(p, stats, n0, n1));
-  }
-  p->MinContext->SummFreq = (UInt16)(sumFreq + escFreq - (escFreq >> 1));
-  p->FoundState = STATS(p->MinContext);
-}
-
-CPpmd_See *Ppmd7_MakeEscFreq(CPpmd7 *p, unsigned numMasked, UInt32 *escFreq)
-{
-  CPpmd_See *see;
-  unsigned nonMasked = p->MinContext->NumStats - numMasked;
-  if (p->MinContext->NumStats != 256)
-  {
-    see = p->See[(unsigned)p->NS2Indx[(size_t)nonMasked - 1]] +
-        (nonMasked < (unsigned)SUFFIX(p->MinContext)->NumStats - p->MinContext->NumStats) +
-        2 * (unsigned)(p->MinContext->SummFreq < 11 * p->MinContext->NumStats) +
-        4 * (unsigned)(numMasked > nonMasked) +
-        p->HiBitsFlag;
-    {
-      unsigned r = (see->Summ >> see->Shift);
-      see->Summ = (UInt16)(see->Summ - r);
-      *escFreq = r + (r == 0);
-    }
-  }
-  else
-  {
-    see = &p->DummySee;
-    *escFreq = 1;
-  }
-  return see;
-}
-
-static void NextContext(CPpmd7 *p)
-{
-  CTX_PTR c = CTX(SUCCESSOR(p->FoundState));
-  if (p->OrderFall == 0 && (Byte *)c > p->Text)
-    p->MinContext = p->MaxContext = c;
-  else
-    UpdateModel(p);
-}
-
-void Ppmd7_Update1(CPpmd7 *p)
-{
-  CPpmd_State *s = p->FoundState;
-  s->Freq += 4;
-  p->MinContext->SummFreq += 4;
-  if (s[0].Freq > s[-1].Freq)
-  {
-    SwapStates(&s[0], &s[-1]);
-    p->FoundState = --s;
-    if (s->Freq > MAX_FREQ)
-      Rescale(p);
-  }
-  NextContext(p);
-}
-
-void Ppmd7_Update1_0(CPpmd7 *p)
-{
-  p->PrevSuccess = (2 * p->FoundState->Freq > p->MinContext->SummFreq);
-  p->RunLength += p->PrevSuccess;
-  p->MinContext->SummFreq += 4;
-  if ((p->FoundState->Freq += 4) > MAX_FREQ)
-    Rescale(p);
-  NextContext(p);
-}
-
-void Ppmd7_UpdateBin(CPpmd7 *p)
-{
-  p->FoundState->Freq = (Byte)(p->FoundState->Freq + (p->FoundState->Freq < 128 ? 1: 0));
-  p->PrevSuccess = 1;
-  p->RunLength++;
-  NextContext(p);
-}
-
-void Ppmd7_Update2(CPpmd7 *p)
-{
-  p->MinContext->SummFreq += 4;
-  if ((p->FoundState->Freq += 4) > MAX_FREQ)
-    Rescale(p);
-  p->RunLength = p->InitRL;
-  UpdateModel(p);
-}
@@ -1,142 +0,0 @@
-/* Ppmd7.h -- PPMdH compression codec
-2018-07-04 : Igor Pavlov : Public domain
-This code is based on PPMd var.H (2001): Dmitry Shkarin : Public domain */
-
-/* This code supports virtual RangeDecoder and includes the implementation
-of RangeCoder from 7z, instead of RangeCoder from original PPMd var.H.
-If you need the compatibility with original PPMd var.H, you can use external RangeDecoder */
-
-#ifndef __PPMD7_H
-#define __PPMD7_H
-
-#include "Ppmd.h"
-
-EXTERN_C_BEGIN
-
-#define PPMD7_MIN_ORDER 2
-#define PPMD7_MAX_ORDER 64
-
-#define PPMD7_MIN_MEM_SIZE (1 << 11)
-#define PPMD7_MAX_MEM_SIZE (0xFFFFFFFF - 12 * 3)
-
-struct CPpmd7_Context_;
-
-typedef
-  #ifdef PPMD_32BIT
-    struct CPpmd7_Context_ *
-  #else
-    UInt32
-  #endif
-  CPpmd7_Context_Ref;
-
-typedef struct CPpmd7_Context_
-{
-  UInt16 NumStats;
-  UInt16 SummFreq;
-  CPpmd_State_Ref Stats;
-  CPpmd7_Context_Ref Suffix;
-} CPpmd7_Context;
-
-#define Ppmd7Context_OneState(p) ((CPpmd_State *)&(p)->SummFreq)
-
-typedef struct
-{
-  CPpmd7_Context *MinContext, *MaxContext;
-  CPpmd_State *FoundState;
-  unsigned OrderFall, InitEsc, PrevSuccess, MaxOrder, HiBitsFlag;
-  Int32 RunLength, InitRL; /* must be 32-bit at least */
-
-  UInt32 Size;
-  UInt32 GlueCount;
-  Byte *Base, *LoUnit, *HiUnit, *Text, *UnitsStart;
-  UInt32 AlignOffset;
-
-  Byte Indx2Units[PPMD_NUM_INDEXES];
-  Byte Units2Indx[128];
-  CPpmd_Void_Ref FreeList[PPMD_NUM_INDEXES];
-  Byte NS2Indx[256], NS2BSIndx[256], HB2Flag[256];
-  CPpmd_See DummySee, See[25][16];
-  UInt16 BinSumm[128][64];
-} CPpmd7;
-
-void Ppmd7_Construct(CPpmd7 *p);
-BoolInt Ppmd7_Alloc(CPpmd7 *p, UInt32 size, ISzAllocPtr alloc);
-void Ppmd7_Free(CPpmd7 *p, ISzAllocPtr alloc);
-void Ppmd7_Init(CPpmd7 *p, unsigned maxOrder);
-#define Ppmd7_WasAllocated(p) ((p)->Base != NULL)
-
-
-/* ---------- Internal Functions ---------- */
-
-extern const Byte PPMD7_kExpEscape[16];
-
-#ifdef PPMD_32BIT
-  #define Ppmd7_GetPtr(p, ptr) (ptr)
-  #define Ppmd7_GetContext(p, ptr) (ptr)
-  #define Ppmd7_GetStats(p, ctx) ((ctx)->Stats)
-#else
-  #define Ppmd7_GetPtr(p, offs) ((void *)((p)->Base + (offs)))
-  #define Ppmd7_GetContext(p, offs) ((CPpmd7_Context *)Ppmd7_GetPtr((p), (offs)))
-  #define Ppmd7_GetStats(p, ctx) ((CPpmd_State *)Ppmd7_GetPtr((p), ((ctx)->Stats)))
-#endif
-
-void Ppmd7_Update1(CPpmd7 *p);
-void Ppmd7_Update1_0(CPpmd7 *p);
-void Ppmd7_Update2(CPpmd7 *p);
-void Ppmd7_UpdateBin(CPpmd7 *p);
-
-#define Ppmd7_GetBinSumm(p) \
-    &p->BinSumm[(size_t)(unsigned)Ppmd7Context_OneState(p->MinContext)->Freq - 1][p->PrevSuccess + \
-    p->NS2BSIndx[(size_t)Ppmd7_GetContext(p, p->MinContext->Suffix)->NumStats - 1] + \
-    (p->HiBitsFlag = p->HB2Flag[p->FoundState->Symbol]) + \
-    2 * p->HB2Flag[(unsigned)Ppmd7Context_OneState(p->MinContext)->Symbol] + \
-    ((p->RunLength >> 26) & 0x20)]
-
-CPpmd_See *Ppmd7_MakeEscFreq(CPpmd7 *p, unsigned numMasked, UInt32 *scale);
-
-
-/* ---------- Decode ---------- */
-
-typedef struct IPpmd7_RangeDec IPpmd7_RangeDec;
-
-struct IPpmd7_RangeDec
-{
-  UInt32 (*GetThreshold)(const IPpmd7_RangeDec *p, UInt32 total);
-  void (*Decode)(const IPpmd7_RangeDec *p, UInt32 start, UInt32 size);
-  UInt32 (*DecodeBit)(const IPpmd7_RangeDec *p, UInt32 size0);
-};
-
-typedef struct
-{
-  IPpmd7_RangeDec vt;
-  UInt32 Range;
-  UInt32 Code;
-  IByteIn *Stream;
-} CPpmd7z_RangeDec;
-
-void Ppmd7z_RangeDec_CreateVTable(CPpmd7z_RangeDec *p);
-BoolInt Ppmd7z_RangeDec_Init(CPpmd7z_RangeDec *p);
-#define Ppmd7z_RangeDec_IsFinishedOK(p) ((p)->Code == 0)
-
-int Ppmd7_DecodeSymbol(CPpmd7 *p, const IPpmd7_RangeDec *rc);
-
-
-/* ---------- Encode ---------- */
-
-typedef struct
-{
-  UInt64 Low;
-  UInt32 Range;
-  Byte Cache;
-  UInt64 CacheSize;
-  IByteOut *Stream;
-} CPpmd7z_RangeEnc;
-
-void Ppmd7z_RangeEnc_Init(CPpmd7z_RangeEnc *p);
-void Ppmd7z_RangeEnc_FlushData(CPpmd7z_RangeEnc *p);
-
-void Ppmd7_EncodeSymbol(CPpmd7 *p, CPpmd7z_RangeEnc *rc, int symbol);
-
-EXTERN_C_END
- 
-#endif
@@ -1,191 +0,0 @@
-/* Ppmd7Dec.c -- PPMdH Decoder
-2018-07-04 : Igor Pavlov : Public domain
-This code is based on PPMd var.H (2001): Dmitry Shkarin : Public domain */
-
-#include "Precomp.h"
-
-#include "Ppmd7.h"
-
-#define kTopValue (1 << 24)
-
-BoolInt Ppmd7z_RangeDec_Init(CPpmd7z_RangeDec *p)
-{
-  unsigned i;
-  p->Code = 0;
-  p->Range = 0xFFFFFFFF;
-  if (IByteIn_Read(p->Stream) != 0)
-    return False;
-  for (i = 0; i < 4; i++)
-    p->Code = (p->Code << 8) | IByteIn_Read(p->Stream);
-  return (p->Code < 0xFFFFFFFF);
-}
-
-#define GET_Ppmd7z_RangeDec CPpmd7z_RangeDec *p = CONTAINER_FROM_VTBL(pp, CPpmd7z_RangeDec, vt);
- 
-static UInt32 Range_GetThreshold(const IPpmd7_RangeDec *pp, UInt32 total)
-{
-  GET_Ppmd7z_RangeDec
-  return p->Code / (p->Range /= total);
-}
-
-static void Range_Normalize(CPpmd7z_RangeDec *p)
-{
-  if (p->Range < kTopValue)
-  {
-    p->Code = (p->Code << 8) | IByteIn_Read(p->Stream);
-    p->Range <<= 8;
-    if (p->Range < kTopValue)
-    {
-      p->Code = (p->Code << 8) | IByteIn_Read(p->Stream);
-      p->Range <<= 8;
-    }
-  }
-}
-
-static void Range_Decode(const IPpmd7_RangeDec *pp, UInt32 start, UInt32 size)
-{
-  GET_Ppmd7z_RangeDec
-  p->Code -= start * p->Range;
-  p->Range *= size;
-  Range_Normalize(p);
-}
-
-static UInt32 Range_DecodeBit(const IPpmd7_RangeDec *pp, UInt32 size0)
-{
-  GET_Ppmd7z_RangeDec
-  UInt32 newBound = (p->Range >> 14) * size0;
-  UInt32 symbol;
-  if (p->Code < newBound)
-  {
-    symbol = 0;
-    p->Range = newBound;
-  }
-  else
-  {
-    symbol = 1;
-    p->Code -= newBound;
-    p->Range -= newBound;
-  }
-  Range_Normalize(p);
-  return symbol;
-}
-
-void Ppmd7z_RangeDec_CreateVTable(CPpmd7z_RangeDec *p)
-{
-  p->vt.GetThreshold = Range_GetThreshold;
-  p->vt.Decode = Range_Decode;
-  p->vt.DecodeBit = Range_DecodeBit;
-}
-
-
-#define MASK(sym) ((signed char *)charMask)[sym]
-
-int Ppmd7_DecodeSymbol(CPpmd7 *p, const IPpmd7_RangeDec *rc)
-{
-  size_t charMask[256 / sizeof(size_t)];
-  if (p->MinContext->NumStats != 1)
-  {
-    CPpmd_State *s = Ppmd7_GetStats(p, p->MinContext);
-    unsigned i;
-    UInt32 count, hiCnt;
-    if ((count = rc->GetThreshold(rc, p->MinContext->SummFreq)) < (hiCnt = s->Freq))
-    {
-      Byte symbol;
-      rc->Decode(rc, 0, s->Freq);
-      p->FoundState = s;
-      symbol = s->Symbol;
-      Ppmd7_Update1_0(p);
-      return symbol;
-    }
-    p->PrevSuccess = 0;
-    i = p->MinContext->NumStats - 1;
-    do
-    {
-      if ((hiCnt += (++s)->Freq) > count)
-      {
-        Byte symbol;
-        rc->Decode(rc, hiCnt - s->Freq, s->Freq);
-        p->FoundState = s;
-        symbol = s->Symbol;
-        Ppmd7_Update1(p);
-        return symbol;
-      }
-    }
-    while (--i);
-    if (count >= p->MinContext->SummFreq)
-      return -2;
-    p->HiBitsFlag = p->HB2Flag[p->FoundState->Symbol];
-    rc->Decode(rc, hiCnt, p->MinContext->SummFreq - hiCnt);
-    PPMD_SetAllBitsIn256Bytes(charMask);
-    MASK(s->Symbol) = 0;
-    i = p->MinContext->NumStats - 1;
-    do { MASK((--s)->Symbol) = 0; } while (--i);
-  }
-  else
-  {
-    UInt16 *prob = Ppmd7_GetBinSumm(p);
-    if (rc->DecodeBit(rc, *prob) == 0)
-    {
-      Byte symbol;
-      *prob = (UInt16)PPMD_UPDATE_PROB_0(*prob);
-      symbol = (p->FoundState = Ppmd7Context_OneState(p->MinContext))->Symbol;
-      Ppmd7_UpdateBin(p);
-      return symbol;
-    }
-    *prob = (UInt16)PPMD_UPDATE_PROB_1(*prob);
-    p->InitEsc = PPMD7_kExpEscape[*prob >> 10];
-    PPMD_SetAllBitsIn256Bytes(charMask);
-    MASK(Ppmd7Context_OneState(p->MinContext)->Symbol) = 0;
-    p->PrevSuccess = 0;
-  }
-  for (;;)
-  {
-    CPpmd_State *ps[256], *s;
-    UInt32 freqSum, count, hiCnt;
-    CPpmd_See *see;
-    unsigned i, num, numMasked = p->MinContext->NumStats;
-    do
-    {
-      p->OrderFall++;
-      if (!p->MinContext->Suffix)
-        return -1;
-      p->MinContext = Ppmd7_GetContext(p, p->MinContext->Suffix);
-    }
-    while (p->MinContext->NumStats == numMasked);
-    hiCnt = 0;
-    s = Ppmd7_GetStats(p, p->MinContext);
-    i = 0;
-    num = p->MinContext->NumStats - numMasked;
-    do
-    {
-      int k = (int)(MASK(s->Symbol));
-      hiCnt += (s->Freq & k);
-      ps[i] = s++;
-      i -= k;
-    }
-    while (i != num);
-    
-    see = Ppmd7_MakeEscFreq(p, numMasked, &freqSum);
-    freqSum += hiCnt;
-    count = rc->GetThreshold(rc, freqSum);
-    
-    if (count < hiCnt)
-    {
-      Byte symbol;
-      CPpmd_State **pps = ps;
-      for (hiCnt = 0; (hiCnt += (*pps)->Freq) <= count; pps++);
-      s = *pps;
-      rc->Decode(rc, hiCnt - s->Freq, s->Freq);
-      Ppmd_See_Update(see);
-      p->FoundState = s;
-      symbol = s->Symbol;
-      Ppmd7_Update2(p);
-      return symbol;
-    }
-    if (count >= freqSum)
-      return -2;
-    rc->Decode(rc, hiCnt, freqSum - hiCnt);
-    see->Summ = (UInt16)(see->Summ + freqSum);
-    do { MASK(ps[--i]->Symbol) = 0; } while (i != 0);
-  }
-}
@@ -1,187 +0,0 @@
-/* Ppmd7Enc.c -- PPMdH Encoder
-2017-04-03 : Igor Pavlov : Public domain
-This code is based on PPMd var.H (2001): Dmitry Shkarin : Public domain */
-
-#include "Precomp.h"
-
-#include "Ppmd7.h"
-
-#define kTopValue (1 << 24)
-
-void Ppmd7z_RangeEnc_Init(CPpmd7z_RangeEnc *p)
-{
-  p->Low = 0;
-  p->Range = 0xFFFFFFFF;
-  p->Cache = 0;
-  p->CacheSize = 1;
-}
-
-static void RangeEnc_ShiftLow(CPpmd7z_RangeEnc *p)
-{
-  if ((UInt32)p->Low < (UInt32)0xFF000000 || (unsigned)(p->Low >> 32) != 0)
-  {
-    Byte temp = p->Cache;
-    do
-    {
-      IByteOut_Write(p->Stream, (Byte)(temp + (Byte)(p->Low >> 32)));
-      temp = 0xFF;
-    }
-    while (--p->CacheSize != 0);
-    p->Cache = (Byte)((UInt32)p->Low >> 24);
-  }
-  p->CacheSize++;
-  p->Low = (UInt32)p->Low << 8;
-}
-
-static void RangeEnc_Encode(CPpmd7z_RangeEnc *p, UInt32 start, UInt32 size, UInt32 total)
-{
-  p->Low += start * (p->Range /= total);
-  p->Range *= size;
-  while (p->Range < kTopValue)
-  {
-    p->Range <<= 8;
-    RangeEnc_ShiftLow(p);
-  }
-}
-
-static void RangeEnc_EncodeBit_0(CPpmd7z_RangeEnc *p, UInt32 size0)
-{
-  p->Range = (p->Range >> 14) * size0;
-  while (p->Range < kTopValue)
-  {
-    p->Range <<= 8;
-    RangeEnc_ShiftLow(p);
-  }
-}
-
-static void RangeEnc_EncodeBit_1(CPpmd7z_RangeEnc *p, UInt32 size0)
-{
-  UInt32 newBound = (p->Range >> 14) * size0;
-  p->Low += newBound;
-  p->Range -= newBound;
-  while (p->Range < kTopValue)
-  {
-    p->Range <<= 8;
-    RangeEnc_ShiftLow(p);
-  }
-}
-
-void Ppmd7z_RangeEnc_FlushData(CPpmd7z_RangeEnc *p)
-{
-  unsigned i;
-  for (i = 0; i < 5; i++)
-    RangeEnc_ShiftLow(p);
-}
-
-
-#define MASK(sym) ((signed char *)charMask)[sym]
-
-void Ppmd7_EncodeSymbol(CPpmd7 *p, CPpmd7z_RangeEnc *rc, int symbol)
-{
-  size_t charMask[256 / sizeof(size_t)];
-  if (p->MinContext->NumStats != 1)
-  {
-    CPpmd_State *s = Ppmd7_GetStats(p, p->MinContext);
-    UInt32 sum;
-    unsigned i;
-    if (s->Symbol == symbol)
-    {
-      RangeEnc_Encode(rc, 0, s->Freq, p->MinContext->SummFreq);
-      p->FoundState = s;
-      Ppmd7_Update1_0(p);
-      return;
-    }
-    p->PrevSuccess = 0;
-    sum = s->Freq;
-    i = p->MinContext->NumStats - 1;
-    do
-    {
-      if ((++s)->Symbol == symbol)
-      {
-        RangeEnc_Encode(rc, sum, s->Freq, p->MinContext->SummFreq);
-        p->FoundState = s;
-        Ppmd7_Update1(p);
-        return;
-      }
-      sum += s->Freq;
-    }
-    while (--i);
-    
-    p->HiBitsFlag = p->HB2Flag[p->FoundState->Symbol];
-    PPMD_SetAllBitsIn256Bytes(charMask);
-    MASK(s->Symbol) = 0;
-    i = p->MinContext->NumStats - 1;
-    do { MASK((--s)->Symbol) = 0; } while (--i);
-    RangeEnc_Encode(rc, sum, p->MinContext->SummFreq - sum, p->MinContext->SummFreq);
-  }
-  else
-  {
-    UInt16 *prob = Ppmd7_GetBinSumm(p);
-    CPpmd_State *s = Ppmd7Context_OneState(p->MinContext);
-    if (s->Symbol == symbol)
-    {
-      RangeEnc_EncodeBit_0(rc, *prob);
-      *prob = (UInt16)PPMD_UPDATE_PROB_0(*prob);
-      p->FoundState = s;
-      Ppmd7_UpdateBin(p);
-      return;
-    }
-    else
-    {
-      RangeEnc_EncodeBit_1(rc, *prob);
-      *prob = (UInt16)PPMD_UPDATE_PROB_1(*prob);
-      p->InitEsc = PPMD7_kExpEscape[*prob >> 10];
-      PPMD_SetAllBitsIn256Bytes(charMask);
-      MASK(s->Symbol) = 0;
-      p->PrevSuccess = 0;
-    }
-  }
-  for (;;)
-  {
-    UInt32 escFreq;
-    CPpmd_See *see;
-    CPpmd_State *s;
-    UInt32 sum;
-    unsigned i, numMasked = p->MinContext->NumStats;
-    do
-    {
-      p->OrderFall++;
-      if (!p->MinContext->Suffix)
-        return; /* EndMarker (symbol = -1) */
-      p->MinContext = Ppmd7_GetContext(p, p->MinContext->Suffix);
-    }
-    while (p->MinContext->NumStats == numMasked);
-    
-    see = Ppmd7_MakeEscFreq(p, numMasked, &escFreq);
-    s = Ppmd7_GetStats(p, p->MinContext);
-    sum = 0;
-    i = p->MinContext->NumStats;
-    do
-    {
-      int cur = s->Symbol;
-      if (cur == symbol)
-      {
-        UInt32 low = sum;
-        CPpmd_State *s1 = s;
-        do
-        {
-          sum += (s->Freq & (int)(MASK(s->Symbol)));
-          s++;
-        }
-        while (--i);
-        RangeEnc_Encode(rc, low, s1->Freq, sum + escFreq);
-        Ppmd_See_Update(see);
-        p->FoundState = s1;
-        Ppmd7_Update2(p);
-        return;
-      }
-      sum += (s->Freq & (int)(MASK(cur)));
-      MASK(cur) = 0;
-      s++;
-    }
-    while (--i);
-    
-    RangeEnc_Encode(rc, sum, escFreq, sum + escFreq);
-    see->Summ = (UInt16)(see->Summ + sum + escFreq);
-  }
-}
@@ -1,248 +0,0 @@
-/* Crypto/Sha256.c -- SHA-256 Hash
-2017-04-03 : Igor Pavlov : Public domain
-This code is based on public domain code from Wei Dai's Crypto++ library. */
-
-#include "Precomp.h"
-
-#include <string.h>
-
-#include "CpuArch.h"
-#include "RotateDefs.h"
-#include "Sha256.h"
-
-/* define it for speed optimization */
-#ifndef _SFX
-#define _SHA256_UNROLL
-#define _SHA256_UNROLL2
-#endif
-
-/* #define _SHA256_UNROLL2 */
-
-void Sha256_Init(CSha256 *p)
-{
-  p->state[0] = 0x6a09e667;
-  p->state[1] = 0xbb67ae85;
-  p->state[2] = 0x3c6ef372;
-  p->state[3] = 0xa54ff53a;
-  p->state[4] = 0x510e527f;
-  p->state[5] = 0x9b05688c;
-  p->state[6] = 0x1f83d9ab;
-  p->state[7] = 0x5be0cd19;
-  p->count = 0;
-}
-
-#define S0(x) (rotrFixed(x, 2) ^ rotrFixed(x,13) ^ rotrFixed(x, 22))
-#define S1(x) (rotrFixed(x, 6) ^ rotrFixed(x,11) ^ rotrFixed(x, 25))
-#define s0(x) (rotrFixed(x, 7) ^ rotrFixed(x,18) ^ (x >> 3))
-#define s1(x) (rotrFixed(x,17) ^ rotrFixed(x,19) ^ (x >> 10))
-
-#define blk0(i) (W[i])
-#define blk2(i) (W[i] += s1(W[((i)-2)&15]) + W[((i)-7)&15] + s0(W[((i)-15)&15]))
-
-#define Ch(x,y,z) (z^(x&(y^z)))
-#define Maj(x,y,z) ((x&y)|(z&(x|y)))
-
-#ifdef _SHA256_UNROLL2
-
-#define R(a,b,c,d,e,f,g,h, i) \
-    h += S1(e) + Ch(e,f,g) + K[(i)+(size_t)(j)] + (j ? blk2(i) : blk0(i)); \
-    d += h; \
-    h += S0(a) + Maj(a, b, c)
-
-#define RX_8(i) \
-  R(a,b,c,d,e,f,g,h, i); \
-  R(h,a,b,c,d,e,f,g, i+1); \
-  R(g,h,a,b,c,d,e,f, i+2); \
-  R(f,g,h,a,b,c,d,e, i+3); \
-  R(e,f,g,h,a,b,c,d, i+4); \
-  R(d,e,f,g,h,a,b,c, i+5); \
-  R(c,d,e,f,g,h,a,b, i+6); \
-  R(b,c,d,e,f,g,h,a, i+7)
-
-#define RX_16  RX_8(0); RX_8(8);
-
-#else
-
-#define a(i) T[(0-(i))&7]
-#define b(i) T[(1-(i))&7]
-#define c(i) T[(2-(i))&7]
-#define d(i) T[(3-(i))&7]
-#define e(i) T[(4-(i))&7]
-#define f(i) T[(5-(i))&7]
-#define g(i) T[(6-(i))&7]
-#define h(i) T[(7-(i))&7]
-
-#define R(i) \
-    h(i) += S1(e(i)) + Ch(e(i),f(i),g(i)) + K[(i)+(size_t)(j)] + (j ? blk2(i) : blk0(i)); \
-    d(i) += h(i); \
-    h(i) += S0(a(i)) + Maj(a(i), b(i), c(i)) \
-
-#ifdef _SHA256_UNROLL
-
-#define RX_8(i)  R(i+0); R(i+1); R(i+2); R(i+3); R(i+4); R(i+5); R(i+6); R(i+7);
-#define RX_16  RX_8(0); RX_8(8);
-
-#else
-
-#define RX_16  unsigned i; for (i = 0; i < 16; i++) { R(i); }
-
-#endif
-
-#endif
-
-static const UInt32 K[64] = {
-  0x428a2f98, 0x71374491, 0xb5c0fbcf, 0xe9b5dba5,
-  0x3956c25b, 0x59f111f1, 0x923f82a4, 0xab1c5ed5,
-  0xd807aa98, 0x12835b01, 0x243185be, 0x550c7dc3,
-  0x72be5d74, 0x80deb1fe, 0x9bdc06a7, 0xc19bf174,
-  0xe49b69c1, 0xefbe4786, 0x0fc19dc6, 0x240ca1cc,
-  0x2de92c6f, 0x4a7484aa, 0x5cb0a9dc, 0x76f988da,
-  0x983e5152, 0xa831c66d, 0xb00327c8, 0xbf597fc7,
-  0xc6e00bf3, 0xd5a79147, 0x06ca6351, 0x14292967,
-  0x27b70a85, 0x2e1b2138, 0x4d2c6dfc, 0x53380d13,
-  0x650a7354, 0x766a0abb, 0x81c2c92e, 0x92722c85,
-  0xa2bfe8a1, 0xa81a664b, 0xc24b8b70, 0xc76c51a3,
-  0xd192e819, 0xd6990624, 0xf40e3585, 0x106aa070,
-  0x19a4c116, 0x1e376c08, 0x2748774c, 0x34b0bcb5,
-  0x391c0cb3, 0x4ed8aa4a, 0x5b9cca4f, 0x682e6ff3,
-  0x748f82ee, 0x78a5636f, 0x84c87814, 0x8cc70208,
-  0x90befffa, 0xa4506ceb, 0xbef9a3f7, 0xc67178f2
-};
-
-static void Sha256_WriteByteBlock(CSha256 *p)
-{
-  UInt32 W[16];
-  unsigned j;
-  UInt32 *state;
-
-  #ifdef _SHA256_UNROLL2
-  UInt32 a,b,c,d,e,f,g,h;
-  #else
-  UInt32 T[8];
-  #endif
-
-  for (j = 0; j < 16; j += 4)
-  {
-    const Byte *ccc = p->buffer + j * 4;
-    W[j    ] = GetBe32(ccc);
-    W[j + 1] = GetBe32(ccc + 4);
-    W[j + 2] = GetBe32(ccc + 8);
-    W[j + 3] = GetBe32(ccc + 12);
-  }
-
-  state = p->state;
-
-  #ifdef _SHA256_UNROLL2
-  a = state[0];
-  b = state[1];
-  c = state[2];
-  d = state[3];
-  e = state[4];
-  f = state[5];
-  g = state[6];
-  h = state[7];
-  #else
-  for (j = 0; j < 8; j++)
-    T[j] = state[j];
-  #endif
-
-  for (j = 0; j < 64; j += 16)
-  {
-    RX_16
-  }
-
-  #ifdef _SHA256_UNROLL2
-  state[0] += a;
-  state[1] += b;
-  state[2] += c;
-  state[3] += d;
-  state[4] += e;
-  state[5] += f;
-  state[6] += g;
-  state[7] += h;
-  #else
-  for (j = 0; j < 8; j++)
-    state[j] += T[j];
-  #endif
-  
-  /* Wipe variables */
-  /* memset(W, 0, sizeof(W)); */
-  /* memset(T, 0, sizeof(T)); */
-}
-
-#undef S0
-#undef S1
-#undef s0
-#undef s1
-
-void Sha256_Update(CSha256 *p, const Byte *data, size_t size)
-{
-  if (size == 0)
-    return;
-
-  {
-    unsigned pos = (unsigned)p->count & 0x3F;
-    unsigned num;
-    
-    p->count += size;
-    
-    num = 64 - pos;
-    if (num > size)
-    {
-      memcpy(p->buffer + pos, data, size);
-      return;
-    }
-    
-    size -= num;
-    memcpy(p->buffer + pos, data, num);
-    data += num;
-  }
-
-  for (;;)
-  {
-    Sha256_WriteByteBlock(p);
-    if (size < 64)
-      break;
-    size -= 64;
-    memcpy(p->buffer, data, 64);
-    data += 64;
-  }
-
-  if (size != 0)
-    memcpy(p->buffer, data, size);
-}
-
-void Sha256_Final(CSha256 *p, Byte *digest)
-{
-  unsigned pos = (unsigned)p->count & 0x3F;
-  unsigned i;
-  
-  p->buffer[pos++] = 0x80;
-  
-  while (pos != (64 - 8))
-  {
-    pos &= 0x3F;
-    if (pos == 0)
-      Sha256_WriteByteBlock(p);
-    p->buffer[pos++] = 0;
-  }
-
-  {
-    UInt64 numBits = (p->count << 3);
-    SetBe32(p->buffer + 64 - 8, (UInt32)(numBits >> 32));
-    SetBe32(p->buffer + 64 - 4, (UInt32)(numBits));
-  }
-  
-  Sha256_WriteByteBlock(p);
-
-  for (i = 0; i < 8; i += 2)
-  {
-    UInt32 v0 = p->state[i];
-    UInt32 v1 = p->state[i + 1];
-    SetBe32(digest    , v0);
-    SetBe32(digest + 4, v1);
-    digest += 8;
-  }
-  
-  Sha256_Init(p);
-}
@@ -1,26 +0,0 @@
-/* Sha256.h -- SHA-256 Hash
-2013-01-18 : Igor Pavlov : Public domain */
-
-#ifndef __CRYPTO_SHA256_H
-#define __CRYPTO_SHA256_H
-
-#include "7zTypes.h"
-
-EXTERN_C_BEGIN
-
-#define SHA256_DIGEST_SIZE 32
-
-typedef struct
-{
-  UInt32 state[8];
-  UInt64 count;
-  Byte buffer[64];
-} CSha256;
-
-void Sha256_Init(CSha256 *p);
-void Sha256_Update(CSha256 *p, const Byte *data, size_t size);
-void Sha256_Final(CSha256 *p, Byte *digest);
-
-EXTERN_C_END
-
-#endif
@@ -1,95 +0,0 @@
-/* Threads.c -- multithreading library
-2017-06-26 : Igor Pavlov : Public domain */
-
-#include "Precomp.h"
-
-#ifndef UNDER_CE
-#include <process.h>
-#endif
-
-#include "Threads.h"
-
-static WRes GetError()
-{
-  DWORD res = GetLastError();
-  return res ? (WRes)res : 1;
-}
-
-static WRes HandleToWRes(HANDLE h) { return (h != NULL) ? 0 : GetError(); }
-static WRes BOOLToWRes(BOOL v) { return v ? 0 : GetError(); }
-
-WRes HandlePtr_Close(HANDLE *p)
-{
-  if (*p != NULL)
-  {
-    if (!CloseHandle(*p))
-      return GetError();
-    *p = NULL;
-  }
-  return 0;
-}
-
-WRes Handle_WaitObject(HANDLE h) { return (WRes)WaitForSingleObject(h, INFINITE); }
-
-WRes Thread_Create(CThread *p, THREAD_FUNC_TYPE func, LPVOID param)
-{
-  /* Windows Me/98/95: threadId parameter may not be NULL in _beginthreadex/CreateThread functions */
-  
-  #ifdef UNDER_CE
-  
-  DWORD threadId;
-  *p = CreateThread(0, 0, func, param, 0, &threadId);
-
-  #else
-
-  unsigned threadId;
-  *p = (HANDLE)_beginthreadex(NULL, 0, func, param, 0, &threadId);
-   
-  #endif
-
-  /* maybe we must use errno here, but probably GetLastError() is also OK. */
-  return HandleToWRes(*p);
-}
-
-static WRes Event_Create(CEvent *p, BOOL manualReset, int signaled)
-{
-  *p = CreateEvent(NULL, manualReset, (signaled ? TRUE : FALSE), NULL);
-  return HandleToWRes(*p);
-}
-
-WRes Event_Set(CEvent *p) { return BOOLToWRes(SetEvent(*p)); }
-WRes Event_Reset(CEvent *p) { return BOOLToWRes(ResetEvent(*p)); }
-
-WRes ManualResetEvent_Create(CManualResetEvent *p, int signaled) { return Event_Create(p, TRUE, signaled); }
-WRes AutoResetEvent_Create(CAutoResetEvent *p, int signaled) { return Event_Create(p, FALSE, signaled); }
-WRes ManualResetEvent_CreateNotSignaled(CManualResetEvent *p) { return ManualResetEvent_Create(p, 0); }
-WRes AutoResetEvent_CreateNotSignaled(CAutoResetEvent *p) { return AutoResetEvent_Create(p, 0); }
-
-
-WRes Semaphore_Create(CSemaphore *p, UInt32 initCount, UInt32 maxCount)
-{
-  *p = CreateSemaphore(NULL, (LONG)initCount, (LONG)maxCount, NULL);
-  return HandleToWRes(*p);
-}
-
-static WRes Semaphore_Release(CSemaphore *p, LONG releaseCount, LONG *previousCount)
-  { return BOOLToWRes(ReleaseSemaphore(*p, releaseCount, previousCount)); }
-WRes Semaphore_ReleaseN(CSemaphore *p, UInt32 num)
-  { return Semaphore_Release(p, (LONG)num, NULL); }
-WRes Semaphore_Release1(CSemaphore *p) { return Semaphore_ReleaseN(p, 1); }
-
-WRes CriticalSection_Init(CCriticalSection *p)
-{
-  /* InitializeCriticalSection can raise only STATUS_NO_MEMORY exception */
-  #ifdef _MSC_VER
-  __try
-  #endif
-  {
-    InitializeCriticalSection(p);
-    /* InitializeCriticalSectionAndSpinCount(p, 0); */
-  }
-  #ifdef _MSC_VER
-  __except (EXCEPTION_EXECUTE_HANDLER) { return 1; }
-  #endif
-  return 0;
-}
@@ -1,68 +0,0 @@
-/* Threads.h -- multithreading library
-2017-06-18 : Igor Pavlov : Public domain */
-
-#ifndef __7Z_THREADS_H
-#define __7Z_THREADS_H
-
-#ifdef _WIN32
-#include <windows.h>
-#endif
-
-#include "7zTypes.h"
-
-EXTERN_C_BEGIN
-
-WRes HandlePtr_Close(HANDLE *h);
-WRes Handle_WaitObject(HANDLE h);
-
-typedef HANDLE CThread;
-#define Thread_Construct(p) *(p) = NULL
-#define Thread_WasCreated(p) (*(p) != NULL)
-#define Thread_Close(p) HandlePtr_Close(p)
-#define Thread_Wait(p) Handle_WaitObject(*(p))
-
-typedef
-#ifdef UNDER_CE
-  DWORD
-#else
-  unsigned
-#endif
-  THREAD_FUNC_RET_TYPE;
-
-#define THREAD_FUNC_CALL_TYPE MY_STD_CALL
-#define THREAD_FUNC_DECL THREAD_FUNC_RET_TYPE THREAD_FUNC_CALL_TYPE
-typedef THREAD_FUNC_RET_TYPE (THREAD_FUNC_CALL_TYPE * THREAD_FUNC_TYPE)(void *);
-WRes Thread_Create(CThread *p, THREAD_FUNC_TYPE func, LPVOID param);
-
-typedef HANDLE CEvent;
-typedef CEvent CAutoResetEvent;
-typedef CEvent CManualResetEvent;
-#define Event_Construct(p) *(p) = NULL
-#define Event_IsCreated(p) (*(p) != NULL)
-#define Event_Close(p) HandlePtr_Close(p)
-#define Event_Wait(p) Handle_WaitObject(*(p))
-WRes Event_Set(CEvent *p);
-WRes Event_Reset(CEvent *p);
-WRes ManualResetEvent_Create(CManualResetEvent *p, int signaled);
-WRes ManualResetEvent_CreateNotSignaled(CManualResetEvent *p);
-WRes AutoResetEvent_Create(CAutoResetEvent *p, int signaled);
-WRes AutoResetEvent_CreateNotSignaled(CAutoResetEvent *p);
-
-typedef HANDLE CSemaphore;
-#define Semaphore_Construct(p) *(p) = NULL
-#define Semaphore_IsCreated(p) (*(p) != NULL)
-#define Semaphore_Close(p) HandlePtr_Close(p)
-#define Semaphore_Wait(p) Handle_WaitObject(*(p))
-WRes Semaphore_Create(CSemaphore *p, UInt32 initCount, UInt32 maxCount);
-WRes Semaphore_ReleaseN(CSemaphore *p, UInt32 num);
-WRes Semaphore_Release1(CSemaphore *p);
-
-typedef CRITICAL_SECTION CCriticalSection;
-WRes CriticalSection_Init(CCriticalSection *p);
-#define CriticalSection_Delete(p) DeleteCriticalSection(p)
-#define CriticalSection_Enter(p) EnterCriticalSection(p)
-#define CriticalSection_Leave(p) LeaveCriticalSection(p)
-
-EXTERN_C_END
-
-#endif
@@ -1,5 +1,5 @@
 /* 7z.h -- 7z interface
-2017-04-03 : Igor Pavlov : Public domain */
+2018-07-02 : Igor Pavlov : Public domain */

 #ifndef __7Z_H
 #define __7Z_H
@@ -91,6 +91,8 @@ typedef struct
  UInt64 *CoderUnpackSizes;       // for all coders in all folders

  Byte *CodersData;
+
+  UInt64 RangeLimit;
 } CSzAr;

 UInt64 SzAr_GetFolderUnpackSize(const CSzAr *p, UInt32 folderIndex);
@@ -1,5 +1,5 @@
 /* 7zArcIn.c -- 7z Input functions
-2018-12-31 : Igor Pavlov : Public domain */
+2021-02-09 : Igor Pavlov : Public domain */

 #include "Precomp.h"

@@ -75,7 +75,7 @@ static SRes SzBitUi32s_Alloc(CSzBitUi32s *p, size_t num, ISzAllocPtr alloc)
  return SZ_OK;
 }

-void SzBitUi32s_Free(CSzBitUi32s *p, ISzAllocPtr alloc)
+static void SzBitUi32s_Free(CSzBitUi32s *p, ISzAllocPtr alloc)
 {
  ISzAlloc_Free(alloc, p->Defs); p->Defs = NULL;
  ISzAlloc_Free(alloc, p->Vals); p->Vals = NULL;
@@ -83,7 +83,7 @@ void SzBitUi32s_Free(CSzBitUi32s *p, ISzAllocPtr alloc)

 #define SzBitUi64s_Init(p) { (p)->Defs = NULL; (p)->Vals = NULL; }

-void SzBitUi64s_Free(CSzBitUi64s *p, ISzAllocPtr alloc)
+static void SzBitUi64s_Free(CSzBitUi64s *p, ISzAllocPtr alloc)
 {
  ISzAlloc_Free(alloc, p->Defs); p->Defs = NULL;
  ISzAlloc_Free(alloc, p->Vals); p->Vals = NULL;
@@ -105,6 +105,8 @@ static void SzAr_Init(CSzAr *p)
  p->CoderUnpackSizes = NULL;

  p->CodersData = NULL;
+
+  p->RangeLimit = 0;
 }

 static void SzAr_Free(CSzAr *p, ISzAllocPtr alloc)
@@ -502,7 +504,7 @@ SRes SzGetNextFolderItem(CSzFolder *f, CSzData *sd)
        return SZ_ERROR_ARCHIVE;
      if (propsSize >= 0x80)
        return SZ_ERROR_UNSUPPORTED;
-      coder->PropsOffset = sd->Data - dataStart;
+      coder->PropsOffset = (size_t)(sd->Data - dataStart);
      coder->PropsSize = (Byte)propsSize;
      sd->Data += (size_t)propsSize;
      sd->Size -= (size_t)propsSize;
@@ -677,7 +679,7 @@ static SRes ReadUnpackInfo(CSzAr *p,
  {
    UInt32 numCoders, ci, numInStreams = 0;
    
-    p->FoCodersOffsets[fo] = sd.Data - startBufPtr;
+    p->FoCodersOffsets[fo] = (size_t)(sd.Data - startBufPtr);
    
    RINOK(SzReadNumber32(&sd, &numCoders));
    if (numCoders == 0 || numCoders > k_Scan_NumCoders_MAX)
@@ -797,7 +799,7 @@ static SRes ReadUnpackInfo(CSzAr *p,
  p->FoToCoderUnpackSizes[fo] = numCodersOutStreams;
  
  {
-    size_t dataSize = sd.Data - startBufPtr;
+    const size_t dataSize = (size_t)(sd.Data - startBufPtr);
    p->FoStartPackStreamIndex[fo] = packStreamIndex;
    p->FoCodersOffsets[fo] = dataSize;
    MY_ALLOC_ZE_AND_CPY(p->CodersData, dataSize, startBufPtr, alloc);
@@ -885,7 +887,7 @@ static SRes ReadSubStreamsInfo(CSzAr *p, CSzData *sd, CSubStreamInfo *ssi)
        if (numStreams != 1 || !SzBitWithVals_Check(&p->FolderCRCs, i))
          numSubDigests += numStreams;
      }
-      ssi->sdNumSubStreams.Size = sd->Data - ssi->sdNumSubStreams.Data;
+      ssi->sdNumSubStreams.Size = (size_t)(sd->Data - ssi->sdNumSubStreams.Data);
      continue;
    }
    if (type == k7zIdCRC || type == k7zIdSize || type == k7zIdEnd)
@@ -907,7 +909,7 @@ static SRes ReadSubStreamsInfo(CSzAr *p, CSzData *sd, CSubStreamInfo *ssi)
  {
    ssi->sdSizes.Data = sd->Data;
    RINOK(SkipNumbers(sd, numUnpackSizesInData));
-    ssi->sdSizes.Size = sd->Data - ssi->sdSizes.Data;
+    ssi->sdSizes.Size = (size_t)(sd->Data - ssi->sdSizes.Data);
    RINOK(ReadID(sd, &type));
  }

@@ -919,7 +921,7 @@ static SRes ReadSubStreamsInfo(CSzAr *p, CSzData *sd, CSubStreamInfo *ssi)
    {
      ssi->sdCRCs.Data = sd->Data;
      RINOK(SkipBitUi32s(sd, numSubDigests));
-      ssi->sdCRCs.Size = sd->Data - ssi->sdCRCs.Data;
+      ssi->sdCRCs.Size = (size_t)(sd->Data - ssi->sdCRCs.Data);
    }
    else
    {
@@ -947,7 +949,11 @@ static SRes SzReadStreamsInfo(CSzAr *p,
  if (type == k7zIdPackInfo)
  {
    RINOK(ReadNumber(sd, dataOffset));
+    if (*dataOffset > p->RangeLimit)
+      return SZ_ERROR_ARCHIVE;
    RINOK(ReadPackInfo(p, sd, alloc));
+    if (p->PackPositions[p->NumPackStreams] > p->RangeLimit - *dataOffset)
+      return SZ_ERROR_ARCHIVE;
    RINOK(ReadID(sd, &type));
  }
  if (type == k7zIdUnpackInfo)
@@ -1028,12 +1034,12 @@ static SRes SzReadFileNames(const Byte *data, size_t size, UInt32 numFiles, size
      return SZ_ERROR_ARCHIVE;
    for (p = data + pos;
      #ifdef _WIN32
-      *(const UInt16 *)p != 0
+      *(const UInt16 *)(const void *)p != 0
      #else
      p[0] != 0 || p[1] != 0
      #endif
      ; p += 2);
-    pos = p - data + 2;
+    pos = (size_t)(p - data) + 2;
    *offsets++ = (pos >> 1);
  }
  while (--numFiles);
@@ -1133,6 +1139,8 @@ static SRes SzReadHeader2(
    SRes res;
    
    SzAr_Init(&tempAr);
+    tempAr.RangeLimit = p->db.RangeLimit;
+
    res = SzReadAndDecodePackedStreams(inStream, sd, tempBufs, NUM_ADDITIONAL_STREAMS_MAX,
        p->startPosAfterHeader, &tempAr, allocTemp);
    *numTempBufs = tempAr.NumFolders;
@@ -1526,11 +1534,13 @@ static SRes SzArEx_Open2(
  nextHeaderSize = GetUi64(header + 20);
  nextHeaderCRC = GetUi32(header + 28);

-  p->startPosAfterHeader = startArcPos + k7zStartHeaderSize;
+  p->startPosAfterHeader = (UInt64)startArcPos + k7zStartHeaderSize;
  
  if (CrcCalc(header + 12, 20) != GetUi32(header + 8))
    return SZ_ERROR_CRC;

+  p->db.RangeLimit = nextHeaderOffset;
+
  nextHeaderSizeT = (size_t)nextHeaderSize;
  if (nextHeaderSizeT != nextHeaderSize)
    return SZ_ERROR_MEM;
@@ -1543,13 +1553,13 @@ static SRes SzArEx_Open2(
  {
    Int64 pos = 0;
    RINOK(ILookInStream_Seek(inStream, &pos, SZ_SEEK_END));
-    if ((UInt64)pos < startArcPos + nextHeaderOffset ||
-        (UInt64)pos < startArcPos + k7zStartHeaderSize + nextHeaderOffset ||
-        (UInt64)pos < startArcPos + k7zStartHeaderSize + nextHeaderOffset + nextHeaderSize)
+    if ((UInt64)pos < (UInt64)startArcPos + nextHeaderOffset ||
+        (UInt64)pos < (UInt64)startArcPos + k7zStartHeaderSize + nextHeaderOffset ||
+        (UInt64)pos < (UInt64)startArcPos + k7zStartHeaderSize + nextHeaderOffset + nextHeaderSize)
      return SZ_ERROR_INPUT_EOF;
  }

-  RINOK(LookInStream_SeekTo(inStream, startArcPos + k7zStartHeaderSize + nextHeaderOffset));
+  RINOK(LookInStream_SeekTo(inStream, (UInt64)startArcPos + k7zStartHeaderSize + nextHeaderOffset));

  if (!Buf_Create(&buf, nextHeaderSizeT, allocTemp))
    return SZ_ERROR_MEM;
@@ -1575,6 +1585,8 @@ static SRes SzArEx_Open2(
        Buf_Init(&tempBuf);
        
        SzAr_Init(&tempAr);
+        tempAr.RangeLimit = p->db.RangeLimit;
+
        res = SzReadAndDecodePackedStreams(inStream, &sd, &tempBuf, 1, p->startPosAfterHeader, &tempAr, allocTemp);
        SzAr_Free(&tempAr, allocTemp);
       
@@ -0,0 +1,322 @@
+/* 7zCrc.c -- CRC32 init
+2021-04-01 : Igor Pavlov : Public domain */
+
+#include "Precomp.h"
+
+#include "7zCrc.h"
+#include "CpuArch.h"
+
+#define kCrcPoly 0xEDB88320
+
+#ifdef MY_CPU_LE
+  #define CRC_NUM_TABLES 8
+#else
+  #define CRC_NUM_TABLES 9
+
+  #define CRC_UINT32_SWAP(v) ((v >> 24) | ((v >> 8) & 0xFF00) | ((v << 8) & 0xFF0000) | (v << 24))
+
+  UInt32 MY_FAST_CALL CrcUpdateT1_BeT4(UInt32 v, const void *data, size_t size, const UInt32 *table);
+  UInt32 MY_FAST_CALL CrcUpdateT1_BeT8(UInt32 v, const void *data, size_t size, const UInt32 *table);
+#endif
+
+#ifndef MY_CPU_BE
+  UInt32 MY_FAST_CALL CrcUpdateT4(UInt32 v, const void *data, size_t size, const UInt32 *table);
+  UInt32 MY_FAST_CALL CrcUpdateT8(UInt32 v, const void *data, size_t size, const UInt32 *table);
+#endif
+
+typedef UInt32 (MY_FAST_CALL *CRC_FUNC)(UInt32 v, const void *data, size_t size, const UInt32 *table);
+
+extern
+CRC_FUNC g_CrcUpdateT4;
+CRC_FUNC g_CrcUpdateT4;
+extern
+CRC_FUNC g_CrcUpdateT8;
+CRC_FUNC g_CrcUpdateT8;
+extern
+CRC_FUNC g_CrcUpdateT0_32;
+CRC_FUNC g_CrcUpdateT0_32;
+extern
+CRC_FUNC g_CrcUpdateT0_64;
+CRC_FUNC g_CrcUpdateT0_64;
+extern
+CRC_FUNC g_CrcUpdate;
+CRC_FUNC g_CrcUpdate;
+
+UInt32 g_CrcTable[256 * CRC_NUM_TABLES];
+
+UInt32 MY_FAST_CALL CrcUpdate(UInt32 v, const void *data, size_t size)
+{
+  return g_CrcUpdate(v, data, size, g_CrcTable);
+}
+
+UInt32 MY_FAST_CALL CrcCalc(const void *data, size_t size)
+{
+  return g_CrcUpdate(CRC_INIT_VAL, data, size, g_CrcTable) ^ CRC_INIT_VAL;
+}
+
+#define CRC_UPDATE_BYTE_2(crc, b) (table[((crc) ^ (b)) & 0xFF] ^ ((crc) >> 8))
+
+UInt32 MY_FAST_CALL CrcUpdateT1(UInt32 v, const void *data, size_t size, const UInt32 *table);
+UInt32 MY_FAST_CALL CrcUpdateT1(UInt32 v, const void *data, size_t size, const UInt32 *table)
+{
+  const Byte *p = (const Byte *)data;
+  const Byte *pEnd = p + size;
+  for (; p != pEnd; p++)
+    v = CRC_UPDATE_BYTE_2(v, *p);
+  return v;
+}
+
+
+/* ---------- hardware CRC ---------- */
+
+#ifdef MY_CPU_LE
+
+#if defined(MY_CPU_ARM_OR_ARM64)
+
+// #pragma message("ARM*")
+
+  #if defined(_MSC_VER)
+    #if defined(MY_CPU_ARM64)
+    #if (_MSC_VER >= 1910)
+        #define USE_ARM64_CRC
+    #endif
+    #endif
+  #elif (defined(__clang__) && (__clang_major__ >= 3)) \
+     || (defined(__GNUC__) && (__GNUC__ > 4))
+      #if !defined(__ARM_FEATURE_CRC32)
+        #define __ARM_FEATURE_CRC32 1
+          #if (!defined(__clang__) || (__clang_major__ > 3)) // fix these numbers
+            #define ATTRIB_CRC __attribute__((__target__("arch=armv8-a+crc")))
+          #endif
+      #endif
+      #if defined(__ARM_FEATURE_CRC32)
+        #define USE_ARM64_CRC
+        #include <arm_acle.h>
+      #endif
+  #endif
+
+#else
+
+// no hardware CRC
+
+// #define USE_CRC_EMU
+
+#ifdef USE_CRC_EMU
+
+#pragma message("ARM64 CRC emulation")
+
+MY_FORCE_INLINE
+UInt32 __crc32b(UInt32 v, UInt32 data)
+{
+  const UInt32 *table = g_CrcTable;
+  v = CRC_UPDATE_BYTE_2(v, (Byte)data);
+  return v;
+}
+
+MY_FORCE_INLINE
+UInt32 __crc32w(UInt32 v, UInt32 data)
+{
+  const UInt32 *table = g_CrcTable;
+  v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8;
+  v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8;
+  v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8;
+  v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8;
+  return v;
+}
+
+MY_FORCE_INLINE
+UInt32 __crc32d(UInt32 v, UInt64 data)
+{
+  const UInt32 *table = g_CrcTable;
+  v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8;
+  v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8;
+  v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8;
+  v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8;
+  v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8;
+  v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8;
+  v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8;
+  v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8;
+  return v;
+}
+
+#endif // USE_CRC_EMU
+
+#endif // defined(MY_CPU_ARM64) && defined(MY_CPU_LE)
+
+
+
+#if defined(USE_ARM64_CRC) || defined(USE_CRC_EMU)
+
+#define T0_32_UNROLL_BYTES (4 * 4)
+#define T0_64_UNROLL_BYTES (4 * 8)
+
+#ifndef ATTRIB_CRC
+#define ATTRIB_CRC
+#endif
+// #pragma message("USE ARM HW CRC")
+
+ATTRIB_CRC
+UInt32 MY_FAST_CALL CrcUpdateT0_32(UInt32 v, const void *data, size_t size, const UInt32 *table);
+ATTRIB_CRC
+UInt32 MY_FAST_CALL CrcUpdateT0_32(UInt32 v, const void *data, size_t size, const UInt32 *table)
+{
+  const Byte *p = (const Byte *)data;
+  UNUSED_VAR(table);
+
+  for (; size != 0 && ((unsigned)(ptrdiff_t)p & (T0_32_UNROLL_BYTES - 1)) != 0; size--)
+    v = __crc32b(v, *p++);
+
+  if (size >= T0_32_UNROLL_BYTES)
+  {
+    const Byte *lim = p + size;
+    size &= (T0_32_UNROLL_BYTES - 1);
+    lim -= size;
+    do
+    {
+      v = __crc32w(v, *(const UInt32 *)(const void *)(p));
+      v = __crc32w(v, *(const UInt32 *)(const void *)(p + 4)); p += 2 * 4;
+      v = __crc32w(v, *(const UInt32 *)(const void *)(p));
+      v = __crc32w(v, *(const UInt32 *)(const void *)(p + 4)); p += 2 * 4;
+    }
+    while (p != lim);
+  }
+  
+  for (; size != 0; size--)
+    v = __crc32b(v, *p++);
+
+  return v;
+}
+
+ATTRIB_CRC
+UInt32 MY_FAST_CALL CrcUpdateT0_64(UInt32 v, const void *data, size_t size, const UInt32 *table);
+ATTRIB_CRC
+UInt32 MY_FAST_CALL CrcUpdateT0_64(UInt32 v, const void *data, size_t size, const UInt32 *table)
+{
+  const Byte *p = (const Byte *)data;
+  UNUSED_VAR(table);
+
+  for (; size != 0 && ((unsigned)(ptrdiff_t)p & (T0_64_UNROLL_BYTES - 1)) != 0; size--)
+    v = __crc32b(v, *p++);
+
+  if (size >= T0_64_UNROLL_BYTES)
+  {
+    const Byte *lim = p + size;
+    size &= (T0_64_UNROLL_BYTES - 1);
+    lim -= size;
+    do
+    {
+      v = __crc32d(v, *(const UInt64 *)(const void *)(p));
+      v = __crc32d(v, *(const UInt64 *)(const void *)(p + 8)); p += 2 * 8;
+      v = __crc32d(v, *(const UInt64 *)(const void *)(p));
+      v = __crc32d(v, *(const UInt64 *)(const void *)(p + 8)); p += 2 * 8;
+    }
+    while (p != lim);
+  }
+  
+  for (; size != 0; size--)
+    v = __crc32b(v, *p++);
+
+  return v;
+}
+
+#endif // defined(USE_ARM64_CRC) || defined(USE_CRC_EMU)
+
+#endif // MY_CPU_LE
+
+
+
+
+void MY_FAST_CALL CrcGenerateTable()
+{
+  UInt32 i;
+  for (i = 0; i < 256; i++)
+  {
+    UInt32 r = i;
+    unsigned j;
+    for (j = 0; j < 8; j++)
+      r = (r >> 1) ^ (kCrcPoly & ((UInt32)0 - (r & 1)));
+    g_CrcTable[i] = r;
+  }
+  for (i = 256; i < 256 * CRC_NUM_TABLES; i++)
+  {
+    UInt32 r = g_CrcTable[(size_t)i - 256];
+    g_CrcTable[i] = g_CrcTable[r & 0xFF] ^ (r >> 8);
+  }
+
+  #if CRC_NUM_TABLES < 4
+  
+  g_CrcUpdate = CrcUpdateT1;
+  
+  #else
+ 
+  #ifdef MY_CPU_LE
+
+    g_CrcUpdateT4 = CrcUpdateT4;
+    g_CrcUpdate = CrcUpdateT4;
+
+    #if CRC_NUM_TABLES >= 8
+      g_CrcUpdateT8 = CrcUpdateT8;
+  
+      #ifdef MY_CPU_X86_OR_AMD64
+      if (!CPU_Is_InOrder())
+      #endif
+        g_CrcUpdate = CrcUpdateT8;
+    #endif
+
+  #else
+  {
+    #ifndef MY_CPU_BE
+    UInt32 k = 0x01020304;
+    const Byte *p = (const Byte *)&k;
+    if (p[0] == 4 && p[1] == 3)
+    {
+      g_CrcUpdateT4 = CrcUpdateT4;
+      g_CrcUpdate = CrcUpdateT4;
+      #if CRC_NUM_TABLES >= 8
+      g_CrcUpdateT8 = CrcUpdateT8;
+      g_CrcUpdate = CrcUpdateT8;
+      #endif
+    }
+    else if (p[0] != 1 || p[1] != 2)
+      g_CrcUpdate = CrcUpdateT1;
+    else
+    #endif
+    {
+      for (i = 256 * CRC_NUM_TABLES - 1; i >= 256; i--)
+      {
+        UInt32 x = g_CrcTable[(size_t)i - 256];
+        g_CrcTable[i] = CRC_UINT32_SWAP(x);
+      }
+      g_CrcUpdateT4 = CrcUpdateT1_BeT4;
+      g_CrcUpdate = CrcUpdateT1_BeT4;
+      #if CRC_NUM_TABLES >= 8
+      g_CrcUpdateT8 = CrcUpdateT1_BeT8;
+      g_CrcUpdate = CrcUpdateT1_BeT8;
+      #endif
+    }
+  }
+  #endif
+  #endif
+
+  #ifdef MY_CPU_LE
+    #ifdef USE_ARM64_CRC
+      if (CPU_IsSupported_CRC32())
+      {
+        g_CrcUpdateT0_32 = CrcUpdateT0_32;
+        g_CrcUpdateT0_64 = CrcUpdateT0_64;
+        g_CrcUpdate =
+          #if defined(MY_CPU_ARM)
+            CrcUpdateT0_32;
+          #else
+            CrcUpdateT0_64;
+          #endif
+      }
+    #endif
+    
+    #ifdef USE_CRC_EMU
+      g_CrcUpdateT0_32 = CrcUpdateT0_32;
+      g_CrcUpdateT0_64 = CrcUpdateT0_64;
+      g_CrcUpdate = CrcUpdateT0_64;
+    #endif
+  #endif
+}
@@ -1,5 +1,5 @@
 /* 7zCrcOpt.c -- CRC32 calculation
-2017-04-03 : Igor Pavlov : Public domain */
+2021-02-09 : Igor Pavlov : Public domain */

 #include "Precomp.h"

@@ -9,6 +9,7 @@

 #define CRC_UPDATE_BYTE_2(crc, b) (table[((crc) ^ (b)) & 0xFF] ^ ((crc) >> 8))

+UInt32 MY_FAST_CALL CrcUpdateT4(UInt32 v, const void *data, size_t size, const UInt32 *table);
 UInt32 MY_FAST_CALL CrcUpdateT4(UInt32 v, const void *data, size_t size, const UInt32 *table)
 {
  const Byte *p = (const Byte *)data;
@@ -16,7 +17,7 @@ UInt32 MY_FAST_CALL CrcUpdateT4(UInt32 v, const void *data, size_t size, const U
    v = CRC_UPDATE_BYTE_2(v, *p);
  for (; size >= 4; size -= 4, p += 4)
  {
-    v ^= *(const UInt32 *)p;
+    v ^= *(const UInt32 *)(const void *)p;
    v =
          (table + 0x300)[((v      ) & 0xFF)]
        ^ (table + 0x200)[((v >>  8) & 0xFF)]
@@ -28,6 +29,7 @@ UInt32 MY_FAST_CALL CrcUpdateT4(UInt32 v, const void *data, size_t size, const U
  return v;
 }

+UInt32 MY_FAST_CALL CrcUpdateT8(UInt32 v, const void *data, size_t size, const UInt32 *table);
 UInt32 MY_FAST_CALL CrcUpdateT8(UInt32 v, const void *data, size_t size, const UInt32 *table)
 {
  const Byte *p = (const Byte *)data;
@@ -36,13 +38,13 @@ UInt32 MY_FAST_CALL CrcUpdateT8(UInt32 v, const void *data, size_t size, const U
  for (; size >= 8; size -= 8, p += 8)
  {
    UInt32 d;
-    v ^= *(const UInt32 *)p;
+    v ^= *(const UInt32 *)(const void *)p;
    v =
          (table + 0x700)[((v      ) & 0xFF)]
        ^ (table + 0x600)[((v >>  8) & 0xFF)]
        ^ (table + 0x500)[((v >> 16) & 0xFF)]
        ^ (table + 0x400)[((v >> 24))];
-    d = *((const UInt32 *)p + 1);
+    d = *((const UInt32 *)(const void *)p + 1);
    v ^=
          (table + 0x300)[((d      ) & 0xFF)]
        ^ (table + 0x200)[((d >>  8) & 0xFF)]
@@ -72,7 +74,7 @@ UInt32 MY_FAST_CALL CrcUpdateT1_BeT4(UInt32 v, const void *data, size_t size, co
    v = CRC_UPDATE_BYTE_2_BE(v, *p);
  for (; size >= 4; size -= 4, p += 4)
  {
-    v ^= *(const UInt32 *)p;
+    v ^= *(const UInt32 *)(const void *)p;
    v =
          (table + 0x000)[((v      ) & 0xFF)]
        ^ (table + 0x100)[((v >>  8) & 0xFF)]
@@ -94,13 +96,13 @@ UInt32 MY_FAST_CALL CrcUpdateT1_BeT8(UInt32 v, const void *data, size_t size, co
  for (; size >= 8; size -= 8, p += 8)
  {
    UInt32 d;
-    v ^= *(const UInt32 *)p;
+    v ^= *(const UInt32 *)(const void *)p;
    v =
          (table + 0x400)[((v      ) & 0xFF)]
        ^ (table + 0x500)[((v >>  8) & 0xFF)]
        ^ (table + 0x600)[((v >> 16) & 0xFF)]
        ^ (table + 0x700)[((v >> 24))];
-    d = *((const UInt32 *)p + 1);
+    d = *((const UInt32 *)(const void *)p + 1);
    v ^=
          (table + 0x000)[((d      ) & 0xFF)]
        ^ (table + 0x100)[((d >>  8) & 0xFF)]
@@ -1,5 +1,5 @@
 /* 7zDec.c -- Decoding from 7z folder
-2019-02-02 : Igor Pavlov : Public domain */
+2021-02-09 : Igor Pavlov : Public domain */

 #include "Precomp.h"

@@ -21,17 +21,20 @@
 #endif

 #define k_Copy 0
-#define k_Delta 3
+#ifndef _7Z_NO_METHOD_LZMA2
 #define k_LZMA2 0x21
+#endif
 #define k_LZMA  0x30101
-#define k_BCJ   0x3030103
 #define k_BCJ2  0x303011B
+#ifndef _7Z_NO_METHODS_FILTERS
+#define k_Delta 3
+#define k_BCJ   0x3030103
 #define k_PPC   0x3030205
 #define k_IA64  0x3030401
 #define k_ARM   0x3030501
 #define k_ARMT  0x3030701
 #define k_SPARC 0x3030805
-
+#endif

 #ifdef _7ZIP_PPMD_SUPPPORT

@@ -56,7 +59,7 @@ static Byte ReadByte(const IByteIn *pp)
    return *p->cur++;
  if (p->res == SZ_OK)
  {
-    size_t size = p->cur - p->begin;
+    size_t size = (size_t)(p->cur - p->begin);
    p->processed += size;
    p->res = ILookInStream_Skip(p->inStream, size);
    size = (1 << 25);
@@ -101,28 +104,32 @@ static SRes SzDecodePpmd(const Byte *props, unsigned propsSize, UInt64 inSize, c
    Ppmd7_Init(&ppmd, order);
  }
  {
-    CPpmd7z_RangeDec rc;
-    Ppmd7z_RangeDec_CreateVTable(&rc);
-    rc.Stream = &s.vt;
-    if (!Ppmd7z_RangeDec_Init(&rc))
+    ppmd.rc.dec.Stream = &s.vt;
+    if (!Ppmd7z_RangeDec_Init(&ppmd.rc.dec))
      res = SZ_ERROR_DATA;
-    else if (s.extra)
-      res = (s.res != SZ_OK ? s.res : SZ_ERROR_DATA);
-    else
+    else if (!s.extra)
    {
-      SizeT i;
-      for (i = 0; i < outSize; i++)
+      Byte *buf = outBuffer;
+      const Byte *lim = buf + outSize;
+      for (; buf != lim; buf++)
      {
-        int sym = Ppmd7_DecodeSymbol(&ppmd, &rc.vt);
+        int sym = Ppmd7z_DecodeSymbol(&ppmd);
        if (s.extra || sym < 0)
          break;
-        outBuffer[i] = (Byte)sym;
+        *buf = (Byte)sym;
      }
-      if (i != outSize)
-        res = (s.res != SZ_OK ? s.res : SZ_ERROR_DATA);
-      else if (s.processed + (s.cur - s.begin) != inSize || !Ppmd7z_RangeDec_IsFinishedOK(&rc))
+      if (buf != lim)
        res = SZ_ERROR_DATA;
+      else if (!Ppmd7z_RangeDec_IsFinishedOK(&ppmd.rc.dec))
+      {
+        /* if (Ppmd7z_DecodeSymbol(&ppmd) != PPMD7_SYM_END || !Ppmd7z_RangeDec_IsFinishedOK(&ppmd.rc.dec)) */
+        res = SZ_ERROR_DATA;
+      }
    }
+    if (s.extra)
+      res = (s.res != SZ_OK ? s.res : SZ_ERROR_DATA);
+    else if (s.processed + (size_t)(s.cur - s.begin) != inSize)
+      res = SZ_ERROR_DATA;
  }
  Ppmd7_Free(&ppmd, allocMain);
  return res;
@@ -365,7 +372,9 @@ static SRes CheckSupportedFolder(const CSzFolder *f)
  return SZ_ERROR_UNSUPPORTED;
 }

+#ifndef _7Z_NO_METHODS_FILTERS
 #define CASE_BRA_CONV(isa) case k_ ## isa: isa ## _Convert(outBuffer, outSize, 0, 0); break;
+#endif

 static SRes SzFolder_Decode2(const CSzFolder *folder,
    const Byte *propsData,
@@ -1,5 +1,5 @@
 /* 7zFile.c -- File IO
-2017-04-03 : Igor Pavlov : Public domain */
+2021-04-29 : Igor Pavlov : Public domain */

 #include "Precomp.h"

@@ -7,9 +7,19 @@

 #ifndef USE_WINDOWS_FILE

-#ifndef UNDER_CE
-#include <errno.h>
-#endif
+  #include <errno.h>
+
+  #ifndef USE_FOPEN
+    #include <stdio.h>
+    #include <fcntl.h>
+    #ifdef _WIN32
+      #include <io.h>
+      typedef int ssize_t;
+      typedef int off_t;
+    #else
+      #include <unistd.h>
+    #endif
+  #endif

 #else

@@ -23,30 +33,36 @@
   And message can be "Network connection was lost"
 */

-#define kChunkSizeMax (1 << 22)
-
 #endif

+#define kChunkSizeMax (1 << 22)
+
 void File_Construct(CSzFile *p)
 {
  #ifdef USE_WINDOWS_FILE
  p->handle = INVALID_HANDLE_VALUE;
-  #else
+  #elif defined(USE_FOPEN)
  p->file = NULL;
+  #else
+  p->fd = -1;
  #endif
 }

 #if !defined(UNDER_CE) || !defined(USE_WINDOWS_FILE)
+
 static WRes File_Open(CSzFile *p, const char *name, int writeMode)
 {
  #ifdef USE_WINDOWS_FILE
+  
  p->handle = CreateFileA(name,
      writeMode ? GENERIC_WRITE : GENERIC_READ,
      FILE_SHARE_READ, NULL,
      writeMode ? CREATE_ALWAYS : OPEN_EXISTING,
      FILE_ATTRIBUTE_NORMAL, NULL);
  return (p->handle != INVALID_HANDLE_VALUE) ? 0 : GetLastError();
-  #else
+  
+  #elif defined(USE_FOPEN)
+  
  p->file = fopen(name, writeMode ? "wb+" : "rb");
  return (p->file != 0) ? 0 :
    #ifdef UNDER_CE
@@ -54,13 +70,34 @@ static WRes File_Open(CSzFile *p, const char *name, int writeMode)
    #else
    errno;
    #endif
+  
+  #else
+
+  int flags = (writeMode ? (O_CREAT | O_EXCL | O_WRONLY) : O_RDONLY);
+  #ifdef O_BINARY
+  flags |= O_BINARY;
+  #endif
+  p->fd = open(name, flags, 0666);
+  return (p->fd != -1) ? 0 : errno;
+
  #endif
 }

 WRes InFile_Open(CSzFile *p, const char *name) { return File_Open(p, name, 0); }
-WRes OutFile_Open(CSzFile *p, const char *name) { return File_Open(p, name, 1); }
+
+WRes OutFile_Open(CSzFile *p, const char *name)
+{
+  #if defined(USE_WINDOWS_FILE) || defined(USE_FOPEN)
+  return File_Open(p, name, 1);
+  #else
+  p->fd = creat(name, 0666);
+  return (p->fd != -1) ? 0 : errno;
+  #endif
+}
+
 #endif

+
 #ifdef USE_WINDOWS_FILE
 static WRes File_OpenW(CSzFile *p, const WCHAR *name, int writeMode)
 {
@@ -78,74 +115,124 @@ WRes OutFile_OpenW(CSzFile *p, const WCHAR *name) { return File_OpenW(p, name, 1
 WRes File_Close(CSzFile *p)
 {
  #ifdef USE_WINDOWS_FILE
+  
  if (p->handle != INVALID_HANDLE_VALUE)
  {
    if (!CloseHandle(p->handle))
      return GetLastError();
    p->handle = INVALID_HANDLE_VALUE;
  }
-  #else
+  
+  #elif defined(USE_FOPEN)
+
  if (p->file != NULL)
  {
    int res = fclose(p->file);
    if (res != 0)
+    {
+      if (res == EOF)
+        return errno;
      return res;
+    }
    p->file = NULL;
  }
+
+  #else
+
+  if (p->fd != -1)
+  {
+    if (close(p->fd) != 0)
+      return errno;
+    p->fd = -1;
+  }
+
  #endif
+
  return 0;
 }

+
 WRes File_Read(CSzFile *p, void *data, size_t *size)
 {
  size_t originalSize = *size;
+  *size = 0;
  if (originalSize == 0)
    return 0;

  #ifdef USE_WINDOWS_FILE

-  *size = 0;
  do
  {
-    DWORD curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : (DWORD)originalSize;
+    const DWORD curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : (DWORD)originalSize;
    DWORD processed = 0;
-    BOOL res = ReadFile(p->handle, data, curSize, &processed, NULL);
+    const BOOL res = ReadFile(p->handle, data, curSize, &processed, NULL);
    data = (void *)((Byte *)data + processed);
    originalSize -= processed;
    *size += processed;
    if (!res)
      return GetLastError();
+    // debug : we can break here for partial reading mode
+    if (processed == 0)
+      break;
+  }
+  while (originalSize > 0);
+
+  #elif defined(USE_FOPEN)
+
+  do
+  {
+    const size_t curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : originalSize;
+    const size_t processed = fread(data, 1, curSize, p->file);
+    data = (void *)((Byte *)data + (size_t)processed);
+    originalSize -= processed;
+    *size += processed;
+    if (processed != curSize)
+      return ferror(p->file);
+    // debug : we can break here for partial reading mode
    if (processed == 0)
      break;
  }
  while (originalSize > 0);
-  return 0;

  #else
-  
-  *size = fread(data, 1, originalSize, p->file);
-  if (*size == originalSize)
-    return 0;
-  return ferror(p->file);
-  
+
+  do
+  {
+    const size_t curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : originalSize;
+    const ssize_t processed = read(p->fd, data, curSize);
+    if (processed == -1)
+      return errno;
+    if (processed == 0)
+      break;
+    data = (void *)((Byte *)data + (size_t)processed);
+    originalSize -= (size_t)processed;
+    *size += (size_t)processed;
+    // debug : we can break here for partial reading mode
+    // break;
+  }
+  while (originalSize > 0);
+
  #endif
+
+  return 0;
 }

+
 WRes File_Write(CSzFile *p, const void *data, size_t *size)
 {
  size_t originalSize = *size;
+  *size = 0;
  if (originalSize == 0)
    return 0;
  
  #ifdef USE_WINDOWS_FILE

-  *size = 0;
  do
  {
-    DWORD curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : (DWORD)originalSize;
+    const DWORD curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : (DWORD)originalSize;
    DWORD processed = 0;
-    BOOL res = WriteFile(p->handle, data, curSize, &processed, NULL);
-    data = (void *)((Byte *)data + processed);
+    const BOOL res = WriteFile(p->handle, data, curSize, &processed, NULL);
+    data = (const void *)((const Byte *)data + processed);
    originalSize -= processed;
    *size += processed;
    if (!res)
@@ -154,26 +241,52 @@ WRes File_Write(CSzFile *p, const void *data, size_t *size)
      break;
  }
  while (originalSize > 0);
-  return 0;
+
+  #elif defined(USE_FOPEN)
+
+  do
+  {
+    const size_t curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : originalSize;
+    const size_t processed = fwrite(data, 1, curSize, p->file);
+    data = (void *)((Byte *)data + (size_t)processed);
+    originalSize -= processed;
+    *size += processed;
+    if (processed != curSize)
+      return ferror(p->file);
+    if (processed == 0)
+      break;
+  }
+  while (originalSize > 0);

  #else

-  *size = fwrite(data, 1, originalSize, p->file);
-  if (*size == originalSize)
-    return 0;
-  return ferror(p->file);
-  
+  do
+  {
+    const size_t curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : originalSize;
+    const ssize_t processed = write(p->fd, data, curSize);
+    if (processed == -1)
+      return errno;
+    if (processed == 0)
+      break;
+    data = (void *)((Byte *)data + (size_t)processed);
+    originalSize -= (size_t)processed;
+    *size += (size_t)processed;
+  }
+  while (originalSize > 0);
+
  #endif
+
+  return 0;
 }

+
 WRes File_Seek(CSzFile *p, Int64 *pos, ESzSeek origin)
 {
  #ifdef USE_WINDOWS_FILE

-  LARGE_INTEGER value;
  DWORD moveMethod;
-  value.LowPart = (DWORD)*pos;
-  value.HighPart = (LONG)((UInt64)*pos >> 16 >> 16); /* for case when UInt64 is 32-bit only */
+  UInt32 low = (UInt32)*pos;
+  LONG high = (LONG)((UInt64)*pos >> 16 >> 16); /* for case when UInt64 is 32-bit only */
  switch (origin)
  {
    case SZ_SEEK_SET: moveMethod = FILE_BEGIN; break;
@@ -181,34 +294,52 @@ WRes File_Seek(CSzFile *p, Int64 *pos, ESzSeek origin)
    case SZ_SEEK_END: moveMethod = FILE_END; break;
    default: return ERROR_INVALID_PARAMETER;
  }
-  value.LowPart = SetFilePointer(p->handle, value.LowPart, &value.HighPart, moveMethod);
-  if (value.LowPart == 0xFFFFFFFF)
+  low = SetFilePointer(p->handle, (LONG)low, &high, moveMethod);
+  if (low == (UInt32)0xFFFFFFFF)
  {
    WRes res = GetLastError();
    if (res != NO_ERROR)
      return res;
  }
-  *pos = ((Int64)value.HighPart << 32) | value.LowPart;
+  *pos = ((Int64)high << 32) | low;
  return 0;

  #else
  
-  int moveMethod;
-  int res;
+  int moveMethod; // = origin;
+
  switch (origin)
  {
    case SZ_SEEK_SET: moveMethod = SEEK_SET; break;
    case SZ_SEEK_CUR: moveMethod = SEEK_CUR; break;
    case SZ_SEEK_END: moveMethod = SEEK_END; break;
-    default: return 1;
+    default: return EINVAL;
  }
-  res = fseek(p->file, (long)*pos, moveMethod);
-  *pos = ftell(p->file);
-  return res;
  
-  #endif
+  #if defined(USE_FOPEN)
+  {
+    int res = fseek(p->file, (long)*pos, moveMethod);
+    if (res == -1)
+      return errno;
+    *pos = ftell(p->file);
+    if (*pos == -1)
+      return errno;
+    return 0;
+  }
+  #else
+  {
+    off_t res = lseek(p->fd, (off_t)*pos, moveMethod);
+    if (res == -1)
+      return errno;
+    *pos = res;
+    return 0;
+  }
+  
+  #endif // USE_FOPEN
+  #endif // USE_WINDOWS_FILE
 }

+
 WRes File_GetLength(CSzFile *p, UInt64 *length)
 {
  #ifdef USE_WINDOWS_FILE
@@ -224,13 +355,31 @@ WRes File_GetLength(CSzFile *p, UInt64 *length)
  *length = (((UInt64)sizeHigh) << 32) + sizeLow;
  return 0;
  
-  #else
+  #elif defined(USE_FOPEN)
  
  long pos = ftell(p->file);
  int res = fseek(p->file, 0, SEEK_END);
  *length = ftell(p->file);
  fseek(p->file, pos, SEEK_SET);
  return res;
+
+  #else
+
+  off_t pos;
+  *length = 0;
+  pos = lseek(p->fd, 0, SEEK_CUR);
+  if (pos != -1)
+  {
+    const off_t len2 = lseek(p->fd, 0, SEEK_END);
+    const off_t res2 = lseek(p->fd, pos, SEEK_SET);
+    if (len2 != -1)
+    {
+      *length = (UInt64)len2;
+      if (res2 != -1)
+        return 0;
+    }
+  }
+  return errno;
  
  #endif
 }
@@ -241,7 +390,9 @@ WRes File_GetLength(CSzFile *p, UInt64 *length)
 static SRes FileSeqInStream_Read(const ISeqInStream *pp, void *buf, size_t *size)
 {
  CFileSeqInStream *p = CONTAINER_FROM_VTBL(pp, CFileSeqInStream, vt);
-  return File_Read(&p->file, buf, size) == 0 ? SZ_OK : SZ_ERROR_READ;
+  WRes wres = File_Read(&p->file, buf, size);
+  p->wres = wres;
+  return (wres == 0) ? SZ_OK : SZ_ERROR_READ;
 }

 void FileSeqInStream_CreateVTable(CFileSeqInStream *p)
@@ -255,13 +406,17 @@ void FileSeqInStream_CreateVTable(CFileSeqInStream *p)
 static SRes FileInStream_Read(const ISeekInStream *pp, void *buf, size_t *size)
 {
  CFileInStream *p = CONTAINER_FROM_VTBL(pp, CFileInStream, vt);
-  return (File_Read(&p->file, buf, size) == 0) ? SZ_OK : SZ_ERROR_READ;
+  WRes wres = File_Read(&p->file, buf, size);
+  p->wres = wres;
+  return (wres == 0) ? SZ_OK : SZ_ERROR_READ;
 }

 static SRes FileInStream_Seek(const ISeekInStream *pp, Int64 *pos, ESzSeek origin)
 {
  CFileInStream *p = CONTAINER_FROM_VTBL(pp, CFileInStream, vt);
-  return File_Seek(&p->file, pos, origin);
+  WRes wres = File_Seek(&p->file, pos, origin);
+  p->wres = wres;
+  return (wres == 0) ? SZ_OK : SZ_ERROR_READ;
 }

 void FileInStream_CreateVTable(CFileInStream *p)
@@ -276,7 +431,8 @@ void FileInStream_CreateVTable(CFileInStream *p)
 static size_t FileOutStream_Write(const ISeqOutStream *pp, const void *data, size_t size)
 {
  CFileOutStream *p = CONTAINER_FROM_VTBL(pp, CFileOutStream, vt);
-  File_Write(&p->file, data, &size);
+  WRes wres = File_Write(&p->file, data, &size);
+  p->wres = wres;
  return size;
 }

@@ -1,17 +1,20 @@
 /* 7zFile.h -- File IO
-2017-04-03 : Igor Pavlov : Public domain */
+2021-02-15 : Igor Pavlov : Public domain */

 #ifndef __7Z_FILE_H
 #define __7Z_FILE_H

 #ifdef _WIN32
 #define USE_WINDOWS_FILE
+// #include <windows.h>
 #endif

 #ifdef USE_WINDOWS_FILE
 #include <windows.h>
 #else
-#include <stdio.h>
+// note: USE_FOPEN mode is limited to 32-bit file size
+// #define USE_FOPEN
+// #include <stdio.h>
 #endif

 #include "7zTypes.h"
@@ -24,8 +27,10 @@ typedef struct
 {
  #ifdef USE_WINDOWS_FILE
  HANDLE handle;
-  #else
+  #elif defined(USE_FOPEN)
  FILE *file;
+  #else
+  int fd;
  #endif
 } CSzFile;

@@ -56,6 +61,7 @@ typedef struct
 {
  ISeqInStream vt;
  CSzFile file;
+  WRes wres;
 } CFileSeqInStream;

 void FileSeqInStream_CreateVTable(CFileSeqInStream *p);
@@ -65,6 +71,7 @@ typedef struct
 {
  ISeekInStream vt;
  CSzFile file;
+  WRes wres;
 } CFileInStream;

 void FileInStream_CreateVTable(CFileInStream *p);
@@ -74,6 +81,7 @@ typedef struct
 {
  ISeqOutStream vt;
  CSzFile file;
+  WRes wres;
 } CFileOutStream;

 void FileOutStream_CreateVTable(CFileOutStream *p);
@@ -1,5 +1,5 @@
 /* 7zStream.c -- 7z Stream functions
-2017-04-03 : Igor Pavlov : Public domain */
+2021-02-09 : Igor Pavlov : Public domain */

 #include "Precomp.h"

@@ -37,7 +37,7 @@ SRes SeqInStream_ReadByte(const ISeqInStream *stream, Byte *buf)

 SRes LookInStream_SeekTo(const ILookInStream *stream, UInt64 offset)
 {
-  Int64 t = offset;
+  Int64 t = (Int64)offset;
  return ILookInStream_Seek(stream, &t, SZ_SEEK_SET);
 }

@@ -1,11 +1,13 @@
 /* 7zTypes.h -- Basic types
-2018-08-04 : Igor Pavlov : Public domain */
+2022-04-01 : Igor Pavlov : Public domain */

 #ifndef __7Z_TYPES_H
 #define __7Z_TYPES_H

 #ifdef _WIN32
 /* #include <windows.h> */
+#else
+#include <errno.h>
 #endif

 #include <stddef.h>
@@ -43,18 +45,112 @@ EXTERN_C_BEGIN
 typedef int SRes;


+#ifdef _MSC_VER
+  #if _MSC_VER > 1200
+    #define MY_ALIGN(n) __declspec(align(n))
+  #else
+    #define MY_ALIGN(n)
+  #endif
+#else
+  #define MY_ALIGN(n) __attribute__ ((aligned(n)))
+#endif
+
+
 #ifdef _WIN32

 /* typedef DWORD WRes; */
 typedef unsigned WRes;
 #define MY_SRes_HRESULT_FROM_WRes(x) HRESULT_FROM_WIN32(x)

-#else
+// #define MY_HRES_ERROR__INTERNAL_ERROR  MY_SRes_HRESULT_FROM_WRes(ERROR_INTERNAL_ERROR)

+#else // _WIN32
+
+// #define ENV_HAVE_LSTAT
 typedef int WRes;
-#define MY__FACILITY_WIN32 7
-#define MY__FACILITY__WRes MY__FACILITY_WIN32
-#define MY_SRes_HRESULT_FROM_WRes(x) ((HRESULT)(x) <= 0 ? ((HRESULT)(x)) : ((HRESULT) (((x) & 0x0000FFFF) | (MY__FACILITY__WRes << 16) | 0x80000000)))
+
+// (FACILITY_ERRNO = 0x800) is 7zip's FACILITY constant to represent (errno) errors in HRESULT
+#define MY__FACILITY_ERRNO  0x800
+#define MY__FACILITY_WIN32  7
+#define MY__FACILITY__WRes  MY__FACILITY_ERRNO
+
+#define MY_HRESULT_FROM_errno_CONST_ERROR(x) ((HRESULT)( \
+          ( (HRESULT)(x) & 0x0000FFFF) \
+          | (MY__FACILITY__WRes << 16)  \
+          | (HRESULT)0x80000000 ))
+
+#define MY_SRes_HRESULT_FROM_WRes(x) \
+  ((HRESULT)(x) <= 0 ? ((HRESULT)(x)) : MY_HRESULT_FROM_errno_CONST_ERROR(x))
+
+// we call macro HRESULT_FROM_WIN32 for system errors (WRes) that are (errno)
+#define HRESULT_FROM_WIN32(x) MY_SRes_HRESULT_FROM_WRes(x)
+
+/*
+#define ERROR_FILE_NOT_FOUND             2L
+#define ERROR_ACCESS_DENIED              5L
+#define ERROR_NO_MORE_FILES              18L
+#define ERROR_LOCK_VIOLATION             33L
+#define ERROR_FILE_EXISTS                80L
+#define ERROR_DISK_FULL                  112L
+#define ERROR_NEGATIVE_SEEK              131L
+#define ERROR_ALREADY_EXISTS             183L
+#define ERROR_DIRECTORY                  267L
+#define ERROR_TOO_MANY_POSTS             298L
+
+#define ERROR_INTERNAL_ERROR             1359L
+#define ERROR_INVALID_REPARSE_DATA       4392L
+#define ERROR_REPARSE_TAG_INVALID        4393L
+#define ERROR_REPARSE_TAG_MISMATCH       4394L
+*/
+
+// we use errno equivalents for some WIN32 errors:
+
+#define ERROR_INVALID_PARAMETER     EINVAL
+#define ERROR_INVALID_FUNCTION      EINVAL
+#define ERROR_ALREADY_EXISTS        EEXIST
+#define ERROR_FILE_EXISTS           EEXIST
+#define ERROR_PATH_NOT_FOUND        ENOENT
+#define ERROR_FILE_NOT_FOUND        ENOENT
+#define ERROR_DISK_FULL             ENOSPC
+// #define ERROR_INVALID_HANDLE        EBADF
+
+// we use FACILITY_WIN32 for errors that has no errno equivalent
+// Too many posts were made to a semaphore.
+#define ERROR_TOO_MANY_POSTS        ((HRESULT)0x8007012AL)
+#define ERROR_INVALID_REPARSE_DATA  ((HRESULT)0x80071128L)
+#define ERROR_REPARSE_TAG_INVALID   ((HRESULT)0x80071129L)
+
+// if (MY__FACILITY__WRes != FACILITY_WIN32),
+// we use FACILITY_WIN32 for COM errors:
+#define E_OUTOFMEMORY               ((HRESULT)0x8007000EL)
+#define E_INVALIDARG                ((HRESULT)0x80070057L)
+#define MY__E_ERROR_NEGATIVE_SEEK   ((HRESULT)0x80070083L)
+
+/*
+// we can use FACILITY_ERRNO for some COM errors, that have errno equivalents:
+#define E_OUTOFMEMORY             MY_HRESULT_FROM_errno_CONST_ERROR(ENOMEM)
+#define E_INVALIDARG              MY_HRESULT_FROM_errno_CONST_ERROR(EINVAL)
+#define MY__E_ERROR_NEGATIVE_SEEK MY_HRESULT_FROM_errno_CONST_ERROR(EINVAL)
+*/
+
+#define TEXT(quote) quote
+
+#define FILE_ATTRIBUTE_READONLY       0x0001
+#define FILE_ATTRIBUTE_HIDDEN         0x0002
+#define FILE_ATTRIBUTE_SYSTEM         0x0004
+#define FILE_ATTRIBUTE_DIRECTORY      0x0010
+#define FILE_ATTRIBUTE_ARCHIVE        0x0020
+#define FILE_ATTRIBUTE_DEVICE         0x0040
+#define FILE_ATTRIBUTE_NORMAL         0x0080
+#define FILE_ATTRIBUTE_TEMPORARY      0x0100
+#define FILE_ATTRIBUTE_SPARSE_FILE    0x0200
+#define FILE_ATTRIBUTE_REPARSE_POINT  0x0400
+#define FILE_ATTRIBUTE_COMPRESSED     0x0800
+#define FILE_ATTRIBUTE_OFFLINE        0x1000
+#define FILE_ATTRIBUTE_NOT_CONTENT_INDEXED 0x2000
+#define FILE_ATTRIBUTE_ENCRYPTED      0x4000
+
+#define FILE_ATTRIBUTE_UNIX_EXTENSION 0x8000   /* trick for Unix */

 #endif

@@ -63,6 +159,10 @@ typedef int WRes;
 #define RINOK(x) { int __result__ = (x); if (__result__ != 0) return __result__; }
 #endif

+#ifndef RINOK_WRes
+#define RINOK_WRes(x) { WRes __result__ = (x); if (__result__ != 0) return __result__; }
+#endif
+
 typedef unsigned char Byte;
 typedef short Int16;
 typedef unsigned short UInt16;
@@ -75,6 +175,40 @@ typedef int Int32;
 typedef unsigned int UInt32;
 #endif

+
+#ifndef _WIN32
+
+typedef int INT;
+typedef Int32 INT32;
+typedef unsigned int UINT;
+typedef UInt32 UINT32;
+typedef INT32 LONG;   // LONG, ULONG and DWORD must be 32-bit for _WIN32 compatibility
+typedef UINT32 ULONG;
+
+#undef DWORD
+typedef UINT32 DWORD;
+
+#define VOID void
+
+#define HRESULT LONG
+
+typedef void *LPVOID;
+// typedef void VOID;
+// typedef ULONG_PTR DWORD_PTR, *PDWORD_PTR;
+// gcc / clang on Unix  : sizeof(long==sizeof(void*) in 32 or 64 bits)
+typedef          long  INT_PTR;
+typedef unsigned long  UINT_PTR;
+typedef          long  LONG_PTR;
+typedef unsigned long  DWORD_PTR;
+
+typedef size_t SIZE_T;
+
+#endif //  _WIN32
+
+
+#define MY_HRES_ERROR__INTERNAL_ERROR  ((HRESULT)0x8007054FL)
+
+
 #ifdef _SZ_NO_INT_64

 /* define _SZ_NO_INT_64, if your compiler doesn't support 64-bit integers.
@@ -128,25 +262,37 @@ typedef int BoolInt;
 #define MY_CDECL __cdecl
 #define MY_FAST_CALL __fastcall

-#else
+#else //  _MSC_VER

-#define MY_NO_INLINE
-#define MY_FORCE_INLINE
-#define MY_CDECL
-#define MY_FAST_CALL
-
-/* inline keyword : for C++ / C99 */
-
-/* GCC, clang: */
-/*
-#if defined (__GNUC__) && (__GNUC__ >= 4)
-#define MY_FORCE_INLINE __attribute__((always_inline))
+#if (defined(__GNUC__) && (__GNUC__ >= 4)) \
+    || (defined(__clang__) && (__clang_major__ >= 4)) \
+    || defined(__INTEL_COMPILER) \
+    || defined(__xlC__)
 #define MY_NO_INLINE __attribute__((noinline))
+// #define MY_FORCE_INLINE __attribute__((always_inline)) inline
+#else
+#define MY_NO_INLINE
 #endif
-*/

+#define MY_FORCE_INLINE
+
+
+#define MY_CDECL
+
+#if  defined(_M_IX86) \
+  || defined(__i386__)
+// #define MY_FAST_CALL __attribute__((fastcall))
+// #define MY_FAST_CALL __attribute__((cdecl))
+#define MY_FAST_CALL
+#elif defined(MY_CPU_AMD64)
+// #define MY_FAST_CALL __attribute__((ms_abi))
+#define MY_FAST_CALL
+#else
+#define MY_FAST_CALL
 #endif

+#endif //  _MSC_VER
+

 /* The following interfaces use first parameter as pointer to structure */

@@ -335,12 +481,11 @@ struct ISzAlloc
    GCC 4.8.1 : classes with non-public variable members"
 */

-#define MY_container_of(ptr, type, m) ((type *)((char *)(1 ? (ptr) : &((type *)0)->m) - MY_offsetof(type, m)))
-
+#define MY_container_of(ptr, type, m) ((type *)(void *)((char *)(void *)(1 ? (ptr) : &((type *)0)->m) - MY_offsetof(type, m)))

 #endif

-#define CONTAINER_FROM_VTBL_SIMPLE(ptr, type, m) ((type *)(ptr))
+#define CONTAINER_FROM_VTBL_SIMPLE(ptr, type, m) ((type *)(void *)(ptr))

 /*
 #define CONTAINER_FROM_VTBL(ptr, type, m) CONTAINER_FROM_VTBL_SIMPLE(ptr, type, m)
@@ -353,6 +498,7 @@ struct ISzAlloc
 */


+#define MY_memset_0_ARRAY(a) memset((a), 0, sizeof(a))

 #ifdef _WIN32

@@ -370,6 +516,14 @@ struct ISzAlloc

 #endif

+#define k_PropVar_TimePrec_0        0
+#define k_PropVar_TimePrec_Unix     1
+#define k_PropVar_TimePrec_DOS      2
+#define k_PropVar_TimePrec_HighPrec 3
+#define k_PropVar_TimePrec_Base     16
+#define k_PropVar_TimePrec_100ns (k_PropVar_TimePrec_Base + 7)
+#define k_PropVar_TimePrec_1ns   (k_PropVar_TimePrec_Base + 9)
+
 EXTERN_C_END

 #endif
@@ -1,7 +1,7 @@
-#define MY_VER_MAJOR 19
-#define MY_VER_MINOR 00
+#define MY_VER_MAJOR 22
+#define MY_VER_MINOR 01
 #define MY_VER_BUILD 0
-#define MY_VERSION_NUMBERS "19.00"
+#define MY_VERSION_NUMBERS "22.01"
 #define MY_VERSION MY_VERSION_NUMBERS

 #ifdef MY_CPU_NAME
@@ -10,12 +10,12 @@
  #define MY_VERSION_CPU MY_VERSION
 #endif

-#define MY_DATE "2019-02-21"
+#define MY_DATE "2022-07-15"
 #undef MY_COPYRIGHT
 #undef MY_VERSION_COPYRIGHT_DATE
 #define MY_AUTHOR_NAME "Igor Pavlov"
 #define MY_COPYRIGHT_PD "Igor Pavlov : Public domain"
-#define MY_COPYRIGHT_CR "Copyright (c) 1999-2018 Igor Pavlov"
+#define MY_COPYRIGHT_CR "Copyright (c) 1999-2022 Igor Pavlov"

 #ifdef USE_COPYRIGHT_CR
  #define MY_COPYRIGHT MY_COPYRIGHT_CR
@@ -1,10 +1,17 @@
 /* Aes.c -- AES encryption / decryption
-2017-01-24 : Igor Pavlov : Public domain */
+2021-05-13 : Igor Pavlov : Public domain */

 #include "Precomp.h"

-#include "Aes.h"
 #include "CpuArch.h"
+#include "Aes.h"
+
+AES_CODE_FUNC g_AesCbc_Decode;
+#ifndef _SFX
+AES_CODE_FUNC g_AesCbc_Encode;
+AES_CODE_FUNC g_AesCtr_Code;
+UInt32 g_Aes_SupportedFunctions_Flags;
+#endif

 static UInt32 T[256 * 4];
 static const Byte Sbox[256] = {
@@ -25,23 +32,10 @@ static const Byte Sbox[256] = {
  0xe1, 0xf8, 0x98, 0x11, 0x69, 0xd9, 0x8e, 0x94, 0x9b, 0x1e, 0x87, 0xe9, 0xce, 0x55, 0x28, 0xdf,
  0x8c, 0xa1, 0x89, 0x0d, 0xbf, 0xe6, 0x42, 0x68, 0x41, 0x99, 0x2d, 0x0f, 0xb0, 0x54, 0xbb, 0x16};

-void MY_FAST_CALL AesCbc_Encode(UInt32 *ivAes, Byte *data, size_t numBlocks);
-void MY_FAST_CALL AesCbc_Decode(UInt32 *ivAes, Byte *data, size_t numBlocks);
-void MY_FAST_CALL AesCtr_Code(UInt32 *ivAes, Byte *data, size_t numBlocks);
-
-void MY_FAST_CALL AesCbc_Encode_Intel(UInt32 *ivAes, Byte *data, size_t numBlocks);
-void MY_FAST_CALL AesCbc_Decode_Intel(UInt32 *ivAes, Byte *data, size_t numBlocks);
-void MY_FAST_CALL AesCtr_Code_Intel(UInt32 *ivAes, Byte *data, size_t numBlocks);
-
-AES_CODE_FUNC g_AesCbc_Encode;
-AES_CODE_FUNC g_AesCbc_Decode;
-AES_CODE_FUNC g_AesCtr_Code;

 static UInt32 D[256 * 4];
 static Byte InvS[256];

-static const Byte Rcon[11] = { 0x00, 0x01, 0x02, 0x04, 0x08, 0x10, 0x20, 0x40, 0x80, 0x1b, 0x36 };
-
 #define xtime(x) ((((x) << 1) ^ (((x) & 0x80) != 0 ? 0x1B : 0)) & 0xFF)

 #define Ui32(a0, a1, a2, a3) ((UInt32)(a0) | ((UInt32)(a1) << 8) | ((UInt32)(a2) << 16) | ((UInt32)(a3) << 24))
@@ -57,6 +51,36 @@ static const Byte Rcon[11] = { 0x00, 0x01, 0x02, 0x04, 0x08, 0x10, 0x20, 0x40, 0
 #define DD(x) (D + (x << 8))


+// #define _SHOW_AES_STATUS
+
+#ifdef MY_CPU_X86_OR_AMD64
+  #define USE_HW_AES
+#elif defined(MY_CPU_ARM_OR_ARM64) && defined(MY_CPU_LE)
+  #if defined(__clang__)
+    #if (__clang_major__ >= 8) // fix that check
+      #define USE_HW_AES
+    #endif
+  #elif defined(__GNUC__)
+    #if (__GNUC__ >= 6) // fix that check
+      #define USE_HW_AES
+    #endif
+  #elif defined(_MSC_VER)
+    #if _MSC_VER >= 1910
+      #define USE_HW_AES
+    #endif
+  #endif
+#endif
+
+#ifdef USE_HW_AES
+#ifdef _SHOW_AES_STATUS
+#include <stdio.h>
+#define _PRF(x) x
+#else
+#define _PRF(x)
+#endif
+#endif
+
+
 void AesGenTables(void)
 {
  unsigned i;
@@ -90,18 +114,48 @@ void AesGenTables(void)
    }
  }
  
-  g_AesCbc_Encode = AesCbc_Encode;
-  g_AesCbc_Decode = AesCbc_Decode;
-  g_AesCtr_Code = AesCtr_Code;
-  
-  #ifdef MY_CPU_X86_OR_AMD64
-  if (CPU_Is_Aes_Supported())
  {
-    g_AesCbc_Encode = AesCbc_Encode_Intel;
-    g_AesCbc_Decode = AesCbc_Decode_Intel;
-    g_AesCtr_Code = AesCtr_Code_Intel;
+  AES_CODE_FUNC d = AesCbc_Decode;
+  #ifndef _SFX
+  AES_CODE_FUNC e = AesCbc_Encode;
+  AES_CODE_FUNC c = AesCtr_Code;
+  UInt32 flags = 0;
+  #endif
+  
+  #ifdef USE_HW_AES
+  if (CPU_IsSupported_AES())
+  {
+    // #pragma message ("AES HW")
+    _PRF(printf("\n===AES HW\n"));
+    d = AesCbc_Decode_HW;
+
+    #ifndef _SFX
+    e = AesCbc_Encode_HW;
+    c = AesCtr_Code_HW;
+    flags = k_Aes_SupportedFunctions_HW;
+    #endif
+
+    #ifdef MY_CPU_X86_OR_AMD64
+    if (CPU_IsSupported_VAES_AVX2())
+    {
+      _PRF(printf("\n===vaes avx2\n"));
+      d = AesCbc_Decode_HW_256;
+      #ifndef _SFX
+      c = AesCtr_Code_HW_256;
+      flags |= k_Aes_SupportedFunctions_HW_256;
+      #endif
+    }
+    #endif
  }
  #endif
+
+  g_AesCbc_Decode = d;
+  #ifndef _SFX
+  g_AesCbc_Encode = e;
+  g_AesCtr_Code = c;
+  g_Aes_SupportedFunctions_Flags = flags;
+  #endif
+  }
 }


@@ -142,8 +196,11 @@ void AesGenTables(void)

 void MY_FAST_CALL Aes_SetKey_Enc(UInt32 *w, const Byte *key, unsigned keySize)
 {
-  unsigned i, wSize;
-  wSize = keySize + 28;
+  unsigned i, m;
+  const UInt32 *wLim;
+  UInt32 t;
+  UInt32 rcon = 1;
+  
  keySize /= 4;
  w[0] = ((UInt32)keySize / 2) + 3;
  w += 4;
@@ -151,16 +208,26 @@ void MY_FAST_CALL Aes_SetKey_Enc(UInt32 *w, const Byte *key, unsigned keySize)
  for (i = 0; i < keySize; i++, key += 4)
    w[i] = GetUi32(key);

-  for (; i < wSize; i++)
+  t = w[(size_t)keySize - 1];
+  wLim = w + (size_t)keySize * 3 + 28;
+  m = 0;
+  do
  {
-    UInt32 t = w[(size_t)i - 1];
-    unsigned rem = i % keySize;
-    if (rem == 0)
-      t = Ui32(Sbox[gb1(t)] ^ Rcon[i / keySize], Sbox[gb2(t)], Sbox[gb3(t)], Sbox[gb0(t)]);
-    else if (keySize > 6 && rem == 4)
+    if (m == 0)
+    {
+      t = Ui32(Sbox[gb1(t)] ^ rcon, Sbox[gb2(t)], Sbox[gb3(t)], Sbox[gb0(t)]);
+      rcon <<= 1;
+      if (rcon & 0x100)
+        rcon = 0x1b;
+      m = keySize;
+    }
+    else if (m == 4 && keySize > 6)
      t = Ui32(Sbox[gb0(t)], Sbox[gb1(t)], Sbox[gb2(t)], Sbox[gb3(t)]);
-    w[i] = w[i - keySize] ^ t;
+    m--;
+    t ^= w[0];
+    w[keySize] = t;
  }
+  while (++w != wLim);
 }

 void MY_FAST_CALL Aes_SetKey_Dec(UInt32 *w, const Byte *key, unsigned keySize)
@@ -184,6 +251,7 @@ void MY_FAST_CALL Aes_SetKey_Dec(UInt32 *w, const Byte *key, unsigned keySize)
  src and dest are pointers to 4 UInt32 words.
  src and dest can point to same block */

+// MY_FORCE_INLINE
 static void Aes_Encode(const UInt32 *w, UInt32 *dest, const UInt32 *src)
 {
  UInt32 s[4];
@@ -207,6 +275,7 @@ static void Aes_Encode(const UInt32 *w, UInt32 *dest, const UInt32 *src)
  FT4(0); FT4(1); FT4(2); FT4(3);
 }

+MY_FORCE_INLINE
 static void Aes_Decode(const UInt32 *w, UInt32 *dest, const UInt32 *src)
 {
  UInt32 s[4];
@@ -294,12 +363,12 @@ void MY_FAST_CALL AesCtr_Code(UInt32 *p, Byte *data, size_t numBlocks)
      UInt32 t = temp[i];

      #ifdef MY_CPU_LE_UNALIGN
-        *((UInt32 *)data) ^= t;
+        *((UInt32 *)(void *)data) ^= t;
      #else
-        data[0] ^= (t & 0xFF);
-        data[1] ^= ((t >> 8) & 0xFF);
-        data[2] ^= ((t >> 16) & 0xFF);
-        data[3] ^= ((t >> 24));
+        data[0] = (Byte)(data[0] ^ (t & 0xFF));
+        data[1] = (Byte)(data[1] ^ ((t >> 8) & 0xFF));
+        data[2] = (Byte)(data[2] ^ ((t >> 16) & 0xFF));
+        data[3] = (Byte)(data[3] ^ ((t >> 24)));
      #endif
    }
  }
@@ -1,5 +1,5 @@
 /* Aes.h -- AES encryption / decryption
-2013-01-18 : Igor Pavlov : Public domain */
+2018-04-28 : Igor Pavlov : Public domain */

 #ifndef __AES_H
 #define __AES_H
@@ -26,12 +26,34 @@ void MY_FAST_CALL Aes_SetKey_Dec(UInt32 *aes, const Byte *key, unsigned keySize)

 /* ivAes - 16-byte aligned pointer to iv+keyMode+roundKeys sequence: UInt32[AES_NUM_IVMRK_WORDS] */
 void AesCbc_Init(UInt32 *ivAes, const Byte *iv); /* iv size is AES_BLOCK_SIZE */
+
 /* data - 16-byte aligned pointer to data */
 /* numBlocks - the number of 16-byte blocks in data array */
 typedef void (MY_FAST_CALL *AES_CODE_FUNC)(UInt32 *ivAes, Byte *data, size_t numBlocks);
-extern AES_CODE_FUNC g_AesCbc_Encode;
+
 extern AES_CODE_FUNC g_AesCbc_Decode;
+#ifndef _SFX
+extern AES_CODE_FUNC g_AesCbc_Encode;
 extern AES_CODE_FUNC g_AesCtr_Code;
+#define k_Aes_SupportedFunctions_HW     (1 << 2)
+#define k_Aes_SupportedFunctions_HW_256 (1 << 3)
+extern UInt32 g_Aes_SupportedFunctions_Flags;
+#endif
+
+
+#define DECLARE__AES_CODE_FUNC(funcName) \
+    void MY_FAST_CALL funcName(UInt32 *ivAes, Byte *data, size_t numBlocks);
+
+DECLARE__AES_CODE_FUNC (AesCbc_Encode)
+DECLARE__AES_CODE_FUNC (AesCbc_Decode)
+DECLARE__AES_CODE_FUNC (AesCtr_Code)
+
+DECLARE__AES_CODE_FUNC (AesCbc_Encode_HW)
+DECLARE__AES_CODE_FUNC (AesCbc_Decode_HW)
+DECLARE__AES_CODE_FUNC (AesCtr_Code_HW)
+
+DECLARE__AES_CODE_FUNC (AesCbc_Decode_HW_256)
+DECLARE__AES_CODE_FUNC (AesCtr_Code_HW_256)

 EXTERN_C_END

@@ -0,0 +1,776 @@
+/* AesOpt.c -- AES optimized code for x86 AES hardware instructions
+2021-04-01 : Igor Pavlov : Public domain */
+
+#include "Precomp.h"
+
+#include "CpuArch.h"
+
+#ifdef MY_CPU_X86_OR_AMD64
+
+  #if defined(__clang__)
+    #if __clang_major__ > 3 || (__clang_major__ == 3 && __clang_minor__ >= 8)
+      #define USE_INTEL_AES
+        #define ATTRIB_AES __attribute__((__target__("aes")))
+      #if (__clang_major__ >= 8)
+        #define USE_INTEL_VAES
+        #define ATTRIB_VAES __attribute__((__target__("aes,vaes,avx2")))
+      #endif
+    #endif
+  #elif defined(__GNUC__)
+    #if (__GNUC__ > 4) || (__GNUC__ == 4 && __GNUC_MINOR__ >= 4)
+      #define USE_INTEL_AES
+      #ifndef __AES__
+        #define ATTRIB_AES __attribute__((__target__("aes")))
+      #endif
+      #if (__GNUC__ >= 8)
+        #define USE_INTEL_VAES
+        #define ATTRIB_VAES __attribute__((__target__("aes,vaes,avx2")))
+      #endif
+    #endif
+  #elif defined(__INTEL_COMPILER)
+    #if (__INTEL_COMPILER >= 1110)
+      #define USE_INTEL_AES
+      #if (__INTEL_COMPILER >= 1900)
+        #define USE_INTEL_VAES
+      #endif
+    #endif
+  #elif defined(_MSC_VER)
+    #if (_MSC_VER > 1500) || (_MSC_FULL_VER >= 150030729)
+      #define USE_INTEL_AES
+      #if (_MSC_VER >= 1910)
+        #define USE_INTEL_VAES
+      #endif
+    #endif
+  #endif
+
+#ifndef ATTRIB_AES
+  #define ATTRIB_AES
+#endif
+#ifndef ATTRIB_VAES
+  #define ATTRIB_VAES
+#endif
+
+
+#ifdef USE_INTEL_AES
+
+#include <wmmintrin.h>
+
+#ifndef USE_INTEL_VAES
+#define AES_TYPE_keys __m128i
+#define AES_TYPE_data __m128i
+#endif
+
+#define AES_FUNC_START(name) \
+    void MY_FAST_CALL name(__m128i *p, __m128i *data, size_t numBlocks)
+
+#define AES_FUNC_START2(name) \
+AES_FUNC_START (name); \
+ATTRIB_AES \
+AES_FUNC_START (name)
+
+#define MM_OP(op, dest, src)  dest = op(dest, src);
+#define MM_OP_m(op, src)      MM_OP(op, m, src);
+
+#define MM_XOR( dest, src)    MM_OP(_mm_xor_si128,    dest, src);
+#define AVX_XOR(dest, src)    MM_OP(_mm256_xor_si256, dest, src);
+
+
+AES_FUNC_START2 (AesCbc_Encode_HW)
+{
+  __m128i m = *p;
+  const __m128i k0 = p[2];
+  const __m128i k1 = p[3];
+  const UInt32 numRounds2 = *(const UInt32 *)(p + 1) - 1;
+  for (; numBlocks != 0; numBlocks--, data++)
+  {
+    UInt32 r = numRounds2;
+    const __m128i *w = p + 4;
+    __m128i temp = *data;
+    MM_XOR (temp, k0);
+    MM_XOR (m, temp);
+    MM_OP_m (_mm_aesenc_si128, k1);
+    do
+    {
+      MM_OP_m (_mm_aesenc_si128, w[0]);
+      MM_OP_m (_mm_aesenc_si128, w[1]);
+      w += 2;
+    }
+    while (--r);
+    MM_OP_m (_mm_aesenclast_si128, w[0]);
+    *data = m;
+  }
+  *p = m;
+}
+
+
+#define WOP_1(op)
+#define WOP_2(op)   WOP_1 (op)  op (m1, 1);
+#define WOP_3(op)   WOP_2 (op)  op (m2, 2);
+#define WOP_4(op)   WOP_3 (op)  op (m3, 3);
+#ifdef MY_CPU_AMD64
+#define WOP_5(op)   WOP_4 (op)  op (m4, 4);
+#define WOP_6(op)   WOP_5 (op)  op (m5, 5);
+#define WOP_7(op)   WOP_6 (op)  op (m6, 6);
+#define WOP_8(op)   WOP_7 (op)  op (m7, 7);
+#endif
+/*
+#define WOP_9(op)   WOP_8 (op)  op (m8, 8);
+#define WOP_10(op)  WOP_9 (op)  op (m9, 9);
+#define WOP_11(op)  WOP_10(op)  op (m10, 10);
+#define WOP_12(op)  WOP_11(op)  op (m11, 11);
+#define WOP_13(op)  WOP_12(op)  op (m12, 12);
+#define WOP_14(op)  WOP_13(op)  op (m13, 13);
+*/
+
+#ifdef MY_CPU_AMD64
+  #define NUM_WAYS      8
+  #define WOP_M1    WOP_8
+#else
+  #define NUM_WAYS      4
+  #define WOP_M1    WOP_4
+#endif
+
+#define WOP(op)  op (m0, 0);  WOP_M1(op)
+
+
+#define DECLARE_VAR(reg, ii)  __m128i reg
+#define LOAD_data(  reg, ii)  reg = data[ii];
+#define STORE_data( reg, ii)  data[ii] = reg;
+#if (NUM_WAYS > 1)
+#define XOR_data_M1(reg, ii)  MM_XOR (reg, data[ii- 1]);
+#endif
+
+#define AVX__DECLARE_VAR(reg, ii)  __m256i reg
+#define AVX__LOAD_data(  reg, ii)  reg = ((const __m256i *)(const void *)data)[ii];
+#define AVX__STORE_data( reg, ii)  ((__m256i *)(void *)data)[ii] = reg;
+#define AVX__XOR_data_M1(reg, ii)  AVX_XOR (reg, (((const __m256i *)(const void *)(data - 1))[ii]));
+
+#define MM_OP_key(op, reg)  MM_OP(op, reg, key);
+
+#define AES_DEC(      reg, ii)   MM_OP_key (_mm_aesdec_si128,     reg)
+#define AES_DEC_LAST( reg, ii)   MM_OP_key (_mm_aesdeclast_si128, reg)
+#define AES_ENC(      reg, ii)   MM_OP_key (_mm_aesenc_si128,     reg)
+#define AES_ENC_LAST( reg, ii)   MM_OP_key (_mm_aesenclast_si128, reg)
+#define AES_XOR(      reg, ii)   MM_OP_key (_mm_xor_si128,        reg)
+
+
+#define AVX__AES_DEC(      reg, ii)   MM_OP_key (_mm256_aesdec_epi128,     reg)
+#define AVX__AES_DEC_LAST( reg, ii)   MM_OP_key (_mm256_aesdeclast_epi128, reg)
+#define AVX__AES_ENC(      reg, ii)   MM_OP_key (_mm256_aesenc_epi128,     reg)
+#define AVX__AES_ENC_LAST( reg, ii)   MM_OP_key (_mm256_aesenclast_epi128, reg)
+#define AVX__AES_XOR(      reg, ii)   MM_OP_key (_mm256_xor_si256,         reg)
+
+#define CTR_START(reg, ii)  MM_OP (_mm_add_epi64, ctr, one); reg = ctr;
+#define CTR_END(  reg, ii)  MM_XOR (data[ii], reg);
+
+#define AVX__CTR_START(reg, ii)  MM_OP (_mm256_add_epi64, ctr2, two); reg = _mm256_xor_si256(ctr2, key);
+#define AVX__CTR_END(  reg, ii)  AVX_XOR (((__m256i *)(void *)data)[ii], reg);
+
+#define WOP_KEY(op, n) { \
+    const __m128i key = w[n]; \
+    WOP(op); }
+
+#define AVX__WOP_KEY(op, n) { \
+    const __m256i key = w[n]; \
+    WOP(op); }
+
+
+#define WIDE_LOOP_START  \
+    dataEnd = data + numBlocks;  \
+    if (numBlocks >= NUM_WAYS)  \
+    { dataEnd -= NUM_WAYS; do {  \
+
+
+#define WIDE_LOOP_END  \
+    data += NUM_WAYS;  \
+    } while (data <= dataEnd);  \
+    dataEnd += NUM_WAYS; }  \
+
+
+#define SINGLE_LOOP  \
+    for (; data < dataEnd; data++)
+
+
+#define NUM_AES_KEYS_MAX 15
+
+#define WIDE_LOOP_START_AVX(OP)  \
+    dataEnd = data + numBlocks;  \
+    if (numBlocks >= NUM_WAYS * 2)  \
+    { __m256i keys[NUM_AES_KEYS_MAX]; \
+    UInt32 ii; \
+    OP \
+    for (ii = 0; ii < numRounds; ii++) \
+      keys[ii] = _mm256_broadcastsi128_si256(p[ii]); \
+    dataEnd -= NUM_WAYS * 2; do {  \
+
+
+#define WIDE_LOOP_END_AVX(OP)  \
+    data += NUM_WAYS * 2;  \
+    } while (data <= dataEnd);  \
+    dataEnd += NUM_WAYS * 2;  \
+    OP  \
+    _mm256_zeroupper();  \
+    }  \
+
+/* MSVC for x86: If we don't call _mm256_zeroupper(), and -arch:IA32 is not specified,
+   MSVC still can insert vzeroupper instruction. */
+
+
+AES_FUNC_START2 (AesCbc_Decode_HW)
+{
+  __m128i iv = *p;
+  const __m128i *wStart = p + *(const UInt32 *)(p + 1) * 2 + 2 - 1;
+  const __m128i *dataEnd;
+  p += 2;
+  
+  WIDE_LOOP_START
+  {
+    const __m128i *w = wStart;
+    
+    WOP (DECLARE_VAR)
+    WOP (LOAD_data);
+    WOP_KEY (AES_XOR, 1)
+
+    do
+    {
+      WOP_KEY (AES_DEC, 0)
+      w--;
+    }
+    while (w != p);
+    WOP_KEY (AES_DEC_LAST, 0)
+
+    MM_XOR (m0, iv);
+    WOP_M1 (XOR_data_M1)
+    iv = data[NUM_WAYS - 1];
+    WOP (STORE_data);
+  }
+  WIDE_LOOP_END
+
+  SINGLE_LOOP
+  {
+    const __m128i *w = wStart - 1;
+    __m128i m = _mm_xor_si128 (w[2], *data);
+    do
+    {
+      MM_OP_m (_mm_aesdec_si128, w[1]);
+      MM_OP_m (_mm_aesdec_si128, w[0]);
+      w -= 2;
+    }
+    while (w != p);
+    MM_OP_m (_mm_aesdec_si128,     w[1]);
+    MM_OP_m (_mm_aesdeclast_si128, w[0]);
+
+    MM_XOR (m, iv);
+    iv = *data;
+    *data = m;
+  }
+  
+  p[-2] = iv;
+}
+
+
+AES_FUNC_START2 (AesCtr_Code_HW)
+{
+  __m128i ctr = *p;
+  UInt32 numRoundsMinus2 = *(const UInt32 *)(p + 1) * 2 - 1;
+  const __m128i *dataEnd;
+  __m128i one = _mm_cvtsi32_si128(1);
+
+  p += 2;
+  
+  WIDE_LOOP_START
+  {
+    const __m128i *w = p;
+    UInt32 r = numRoundsMinus2;
+    WOP (DECLARE_VAR)
+    WOP (CTR_START);
+    WOP_KEY (AES_XOR, 0)
+    w += 1;
+    do
+    {
+      WOP_KEY (AES_ENC, 0)
+      w += 1;
+    }
+    while (--r);
+    WOP_KEY (AES_ENC_LAST, 0)
+   
+    WOP (CTR_END);
+  }
+  WIDE_LOOP_END
+
+  SINGLE_LOOP
+  {
+    UInt32 numRounds2 = *(const UInt32 *)(p - 2 + 1) - 1;
+    const __m128i *w = p;
+    __m128i m;
+    MM_OP (_mm_add_epi64, ctr, one);
+    m = _mm_xor_si128 (ctr, p[0]);
+    w += 1;
+    do
+    {
+      MM_OP_m (_mm_aesenc_si128, w[0]);
+      MM_OP_m (_mm_aesenc_si128, w[1]);
+      w += 2;
+    }
+    while (--numRounds2);
+    MM_OP_m (_mm_aesenc_si128,     w[0]);
+    MM_OP_m (_mm_aesenclast_si128, w[1]);
+    MM_XOR (*data, m);
+  }
+  
+  p[-2] = ctr;
+}
+
+
+
+#ifdef USE_INTEL_VAES
+
+#if defined(__clang__) && defined(_MSC_VER)
+#define __SSE4_2__
+#define __AES__
+#define __AVX__
+#define __AVX2__
+#define __VAES__
+#define __AVX512F__
+#define __AVX512VL__
+#endif
+
+#include <immintrin.h>
+
+#define VAES_FUNC_START2(name) \
+AES_FUNC_START (name); \
+ATTRIB_VAES \
+AES_FUNC_START (name)
+
+VAES_FUNC_START2 (AesCbc_Decode_HW_256)
+{
+  __m128i iv = *p;
+  const __m128i *dataEnd;
+  UInt32 numRounds = *(const UInt32 *)(p + 1) * 2 + 1;
+  p += 2;
+  
+  WIDE_LOOP_START_AVX(;)
+  {
+    const __m256i *w = keys + numRounds - 2;
+    
+    WOP (AVX__DECLARE_VAR)
+    WOP (AVX__LOAD_data);
+    AVX__WOP_KEY (AVX__AES_XOR, 1)
+
+    do
+    {
+      AVX__WOP_KEY (AVX__AES_DEC, 0)
+      w--;
+    }
+    while (w != keys);
+    AVX__WOP_KEY (AVX__AES_DEC_LAST, 0)
+
+    AVX_XOR (m0, _mm256_setr_m128i(iv, data[0]));
+    WOP_M1 (AVX__XOR_data_M1)
+    iv = data[NUM_WAYS * 2 - 1];
+    WOP (AVX__STORE_data);
+  }
+  WIDE_LOOP_END_AVX(;)
+
+  SINGLE_LOOP
+  {
+    const __m128i *w = p + *(const UInt32 *)(p + 1 - 2) * 2 + 1 - 3;
+    __m128i m = _mm_xor_si128 (w[2], *data);
+    do
+    {
+      MM_OP_m (_mm_aesdec_si128, w[1]);
+      MM_OP_m (_mm_aesdec_si128, w[0]);
+      w -= 2;
+    }
+    while (w != p);
+    MM_OP_m (_mm_aesdec_si128,     w[1]);
+    MM_OP_m (_mm_aesdeclast_si128, w[0]);
+
+    MM_XOR (m, iv);
+    iv = *data;
+    *data = m;
+  }
+  
+  p[-2] = iv;
+}
+
+
+/*
+SSE2: _mm_cvtsi32_si128 : movd
+AVX:  _mm256_setr_m128i            : vinsertf128
+AVX2: _mm256_add_epi64             : vpaddq ymm, ymm, ymm
+      _mm256_extracti128_si256     : vextracti128
+      _mm256_broadcastsi128_si256  : vbroadcasti128
+*/
+
+#define AVX__CTR_LOOP_START  \
+    ctr2 = _mm256_setr_m128i(_mm_sub_epi64(ctr, one), ctr); \
+    two = _mm256_setr_m128i(one, one); \
+    two = _mm256_add_epi64(two, two); \
+
+// two = _mm256_setr_epi64x(2, 0, 2, 0);
+  
+#define AVX__CTR_LOOP_ENC  \
+    ctr = _mm256_extracti128_si256 (ctr2, 1); \
+ 
+VAES_FUNC_START2 (AesCtr_Code_HW_256)
+{
+  __m128i ctr = *p;
+  UInt32 numRounds = *(const UInt32 *)(p + 1) * 2 + 1;
+  const __m128i *dataEnd;
+  __m128i one = _mm_cvtsi32_si128(1);
+  __m256i ctr2, two;
+  p += 2;
+  
+  WIDE_LOOP_START_AVX (AVX__CTR_LOOP_START)
+  {
+    const __m256i *w = keys;
+    UInt32 r = numRounds - 2;
+    WOP (AVX__DECLARE_VAR)
+    AVX__WOP_KEY (AVX__CTR_START, 0);
+
+    w += 1;
+    do
+    {
+      AVX__WOP_KEY (AVX__AES_ENC, 0)
+      w += 1;
+    }
+    while (--r);
+    AVX__WOP_KEY (AVX__AES_ENC_LAST, 0)
+   
+    WOP (AVX__CTR_END);
+  }
+  WIDE_LOOP_END_AVX (AVX__CTR_LOOP_ENC)
+  
+  SINGLE_LOOP
+  {
+    UInt32 numRounds2 = *(const UInt32 *)(p - 2 + 1) - 1;
+    const __m128i *w = p;
+    __m128i m;
+    MM_OP (_mm_add_epi64, ctr, one);
+    m = _mm_xor_si128 (ctr, p[0]);
+    w += 1;
+    do
+    {
+      MM_OP_m (_mm_aesenc_si128, w[0]);
+      MM_OP_m (_mm_aesenc_si128, w[1]);
+      w += 2;
+    }
+    while (--numRounds2);
+    MM_OP_m (_mm_aesenc_si128,     w[0]);
+    MM_OP_m (_mm_aesenclast_si128, w[1]);
+    MM_XOR (*data, m);
+  }
+
+  p[-2] = ctr;
+}
+
+#endif // USE_INTEL_VAES
+
+#else // USE_INTEL_AES
+
+/* no USE_INTEL_AES */
+
+#pragma message("AES  HW_SW stub was used")
+
+#define AES_TYPE_keys UInt32
+#define AES_TYPE_data Byte
+
+#define AES_FUNC_START(name) \
+    void MY_FAST_CALL name(UInt32 *p, Byte *data, size_t numBlocks) \
+
+#define AES_COMPAT_STUB(name) \
+    AES_FUNC_START(name); \
+    AES_FUNC_START(name ## _HW) \
+    { name(p, data, numBlocks); }
+
+AES_COMPAT_STUB (AesCbc_Encode)
+AES_COMPAT_STUB (AesCbc_Decode)
+AES_COMPAT_STUB (AesCtr_Code)
+
+#endif // USE_INTEL_AES
+
+
+#ifndef USE_INTEL_VAES
+
+#pragma message("VAES HW_SW stub was used")
+
+#define VAES_COMPAT_STUB(name) \
+    void MY_FAST_CALL name ## _256(UInt32 *p, Byte *data, size_t numBlocks); \
+    void MY_FAST_CALL name ## _256(UInt32 *p, Byte *data, size_t numBlocks) \
+    { name((AES_TYPE_keys *)(void *)p, (AES_TYPE_data *)(void *)data, numBlocks); }
+
+VAES_COMPAT_STUB (AesCbc_Decode_HW)
+VAES_COMPAT_STUB (AesCtr_Code_HW)
+
+#endif // ! USE_INTEL_VAES
+
+
+#elif defined(MY_CPU_ARM_OR_ARM64) && defined(MY_CPU_LE)
+
+  #if defined(__clang__)
+    #if (__clang_major__ >= 8) // fix that check
+      #define USE_HW_AES
+    #endif
+  #elif defined(__GNUC__)
+    #if (__GNUC__ >= 6) // fix that check
+      #define USE_HW_AES
+    #endif
+  #elif defined(_MSC_VER)
+    #if _MSC_VER >= 1910
+      #define USE_HW_AES
+    #endif
+  #endif
+
+#ifdef USE_HW_AES
+
+// #pragma message("=== AES HW === ")
+
+#if defined(__clang__) || defined(__GNUC__)
+  #ifdef MY_CPU_ARM64
+    #define ATTRIB_AES __attribute__((__target__("+crypto")))
+  #else
+    #define ATTRIB_AES __attribute__((__target__("fpu=crypto-neon-fp-armv8")))
+  #endif
+#else
+  // _MSC_VER
+  // for arm32
+  #define _ARM_USE_NEW_NEON_INTRINSICS
+#endif
+
+#ifndef ATTRIB_AES
+  #define ATTRIB_AES
+#endif
+
+#if defined(_MSC_VER) && defined(MY_CPU_ARM64)
+#include <arm64_neon.h>
+#else
+#include <arm_neon.h>
+#endif
+
+typedef uint8x16_t v128;
+
+#define AES_FUNC_START(name) \
+    void MY_FAST_CALL name(v128 *p, v128 *data, size_t numBlocks)
+
+#define AES_FUNC_START2(name) \
+AES_FUNC_START (name); \
+ATTRIB_AES \
+AES_FUNC_START (name)
+
+#define MM_OP(op, dest, src)  dest = op(dest, src);
+#define MM_OP_m(op, src)      MM_OP(op, m, src);
+#define MM_OP1_m(op)          m = op(m);
+
+#define MM_XOR( dest, src)    MM_OP(veorq_u8, dest, src);
+#define MM_XOR_m( src)        MM_XOR(m, src);
+
+#define AES_E_m(k)     MM_OP_m (vaeseq_u8, k);
+#define AES_E_MC_m(k)  AES_E_m (k);  MM_OP1_m(vaesmcq_u8);
+
+
+AES_FUNC_START2 (AesCbc_Encode_HW)
+{
+  v128 m = *p;
+  const v128 k0 = p[2];
+  const v128 k1 = p[3];
+  const v128 k2 = p[4];
+  const v128 k3 = p[5];
+  const v128 k4 = p[6];
+  const v128 k5 = p[7];
+  const v128 k6 = p[8];
+  const v128 k7 = p[9];
+  const v128 k8 = p[10];
+  const v128 k9 = p[11];
+  const UInt32 numRounds2 = *(const UInt32 *)(p + 1);
+  const v128 *w = p + ((size_t)numRounds2 * 2);
+  const v128 k_z1 = w[1];
+  const v128 k_z0 = w[2];
+  for (; numBlocks != 0; numBlocks--, data++)
+  {
+    MM_XOR_m (*data);
+    AES_E_MC_m (k0)
+    AES_E_MC_m (k1)
+    AES_E_MC_m (k2)
+    AES_E_MC_m (k3)
+    AES_E_MC_m (k4)
+    AES_E_MC_m (k5)
+    AES_E_MC_m (k6)
+    AES_E_MC_m (k7)
+    AES_E_MC_m (k8)
+    if (numRounds2 >= 6)
+    {
+      AES_E_MC_m (k9)
+      AES_E_MC_m (p[12])
+      if (numRounds2 != 6)
+      {
+        AES_E_MC_m (p[13])
+        AES_E_MC_m (p[14])
+      }
+    }
+    AES_E_m  (k_z1);
+    MM_XOR_m (k_z0);
+    *data = m;
+  }
+  *p = m;
+}
+
+
+#define WOP_1(op)
+#define WOP_2(op)   WOP_1 (op)  op (m1, 1);
+#define WOP_3(op)   WOP_2 (op)  op (m2, 2);
+#define WOP_4(op)   WOP_3 (op)  op (m3, 3);
+#define WOP_5(op)   WOP_4 (op)  op (m4, 4);
+#define WOP_6(op)   WOP_5 (op)  op (m5, 5);
+#define WOP_7(op)   WOP_6 (op)  op (m6, 6);
+#define WOP_8(op)   WOP_7 (op)  op (m7, 7);
+
+  #define NUM_WAYS      8
+  #define WOP_M1    WOP_8
+
+#define WOP(op)  op (m0, 0);  WOP_M1(op)
+
+#define DECLARE_VAR(reg, ii)  v128 reg
+#define LOAD_data(  reg, ii)  reg = data[ii];
+#define STORE_data( reg, ii)  data[ii] = reg;
+#if (NUM_WAYS > 1)
+#define XOR_data_M1(reg, ii)  MM_XOR (reg, data[ii- 1]);
+#endif
+
+#define MM_OP_key(op, reg)  MM_OP (op, reg, key);
+
+#define AES_D_m(k)      MM_OP_m (vaesdq_u8, k);
+#define AES_D_IMC_m(k)  AES_D_m (k);  MM_OP1_m (vaesimcq_u8);
+
+#define AES_XOR(   reg, ii)  MM_OP_key (veorq_u8,  reg)
+#define AES_D(     reg, ii)  MM_OP_key (vaesdq_u8, reg)
+#define AES_E(     reg, ii)  MM_OP_key (vaeseq_u8, reg)
+
+#define AES_D_IMC( reg, ii)  AES_D (reg, ii);  reg = vaesimcq_u8(reg)
+#define AES_E_MC(  reg, ii)  AES_E (reg, ii);  reg = vaesmcq_u8(reg)
+
+#define CTR_START(reg, ii)  MM_OP (vaddq_u64, ctr, one);  reg = vreinterpretq_u8_u64(ctr);
+#define CTR_END(  reg, ii)  MM_XOR (data[ii], reg);
+
+#define WOP_KEY(op, n) { \
+    const v128 key = w[n]; \
+    WOP(op); }
+
+#define WIDE_LOOP_START  \
+    dataEnd = data + numBlocks;  \
+    if (numBlocks >= NUM_WAYS)  \
+    { dataEnd -= NUM_WAYS; do {  \
+
+#define WIDE_LOOP_END  \
+    data += NUM_WAYS;  \
+    } while (data <= dataEnd);  \
+    dataEnd += NUM_WAYS; }  \
+
+#define SINGLE_LOOP  \
+    for (; data < dataEnd; data++)
+
+
+AES_FUNC_START2 (AesCbc_Decode_HW)
+{
+  v128 iv = *p;
+  const v128 *wStart = p + ((size_t)*(const UInt32 *)(p + 1)) * 2;
+  const v128 *dataEnd;
+  p += 2;
+  
+  WIDE_LOOP_START
+  {
+    const v128 *w = wStart;
+    WOP (DECLARE_VAR)
+    WOP (LOAD_data);
+    WOP_KEY (AES_D_IMC, 2)
+    do
+    {
+      WOP_KEY (AES_D_IMC, 1)
+      WOP_KEY (AES_D_IMC, 0)
+      w -= 2;
+    }
+    while (w != p);
+    WOP_KEY (AES_D,   1)
+    WOP_KEY (AES_XOR, 0)
+    MM_XOR (m0, iv);
+    WOP_M1 (XOR_data_M1)
+    iv = data[NUM_WAYS - 1];
+    WOP (STORE_data);
+  }
+  WIDE_LOOP_END
+
+  SINGLE_LOOP
+  {
+    const v128 *w = wStart;
+    v128 m = *data;
+    AES_D_IMC_m (w[2])
+    do
+    {
+      AES_D_IMC_m (w[1]);
+      AES_D_IMC_m (w[0]);
+      w -= 2;
+    }
+    while (w != p);
+    AES_D_m  (w[1]);
+    MM_XOR_m (w[0]);
+    MM_XOR_m (iv);
+    iv = *data;
+    *data = m;
+  }
+  
+  p[-2] = iv;
+}
+
+
+AES_FUNC_START2 (AesCtr_Code_HW)
+{
+  uint64x2_t ctr = vreinterpretq_u64_u8(*p);
+  const v128 *wEnd = p + ((size_t)*(const UInt32 *)(p + 1)) * 2;
+  const v128 *dataEnd;
+  uint64x2_t one = vdupq_n_u64(0);
+  one = vsetq_lane_u64(1, one, 0);
+  p += 2;
+  
+  WIDE_LOOP_START
+  {
+    const v128 *w = p;
+    WOP (DECLARE_VAR)
+    WOP (CTR_START);
+    do
+    {
+      WOP_KEY (AES_E_MC, 0)
+      WOP_KEY (AES_E_MC, 1)
+      w += 2;
+    }
+    while (w != wEnd);
+    WOP_KEY (AES_E_MC, 0)
+    WOP_KEY (AES_E,    1)
+    WOP_KEY (AES_XOR,  2)
+    WOP (CTR_END);
+  }
+  WIDE_LOOP_END
+
+  SINGLE_LOOP
+  {
+    const v128 *w = p;
+    v128 m;
+    CTR_START (m, 0);
+    do
+    {
+      AES_E_MC_m (w[0]);
+      AES_E_MC_m (w[1]);
+      w += 2;
+    }
+    while (w != wEnd);
+    AES_E_MC_m (w[0]);
+    AES_E_m    (w[1]);
+    MM_XOR_m   (w[2]);
+    CTR_END (m, 0);
+  }
+  
+  p[-2] = vreinterpretq_u8_u64(ctr);
+}
+
+#endif // USE_HW_AES
+
+#endif // MY_CPU_ARM_OR_ARM64
@@ -1,12 +1,12 @@
 /* Alloc.c -- Memory allocation functions
-2018-04-27 : Igor Pavlov : Public domain */
+2021-07-13 : Igor Pavlov : Public domain */

 #include "Precomp.h"

 #include <stdio.h>

 #ifdef _WIN32
-#include <windows.h>
+#include <Windows.h>
 #endif
 #include <stdlib.h>

@@ -122,7 +122,6 @@ static void PrintAddr(void *p)
 #define Print(s)
 #define PrintLn()
 #define PrintHex(v, align)
-#define PrintDec(v, align)
 #define PrintAddr(p)

 #endif
@@ -133,10 +132,11 @@ void *MyAlloc(size_t size)
 {
  if (size == 0)
    return NULL;
+  PRINT_ALLOC("Alloc    ", g_allocCount, size, NULL);
  #ifdef _SZ_ALLOC_DEBUG
  {
    void *p = malloc(size);
-    PRINT_ALLOC("Alloc    ", g_allocCount, size, p);
+    // PRINT_ALLOC("Alloc    ", g_allocCount, size, p);
    return p;
  }
  #else
@@ -172,14 +172,20 @@ void MidFree(void *address)
  VirtualFree(address, 0, MEM_RELEASE);
 }

-#ifndef MEM_LARGE_PAGES
-#undef _7ZIP_LARGE_PAGES
+#ifdef _7ZIP_LARGE_PAGES
+
+#ifdef MEM_LARGE_PAGES
+  #define MY__MEM_LARGE_PAGES  MEM_LARGE_PAGES
+#else
+  #define MY__MEM_LARGE_PAGES  0x20000000
 #endif

-#ifdef _7ZIP_LARGE_PAGES
+extern
+SIZE_T g_LargePageSize;
 SIZE_T g_LargePageSize = 0;
-typedef SIZE_T (WINAPI *GetLargePageMinimumP)();
-#endif
+typedef SIZE_T (WINAPI *GetLargePageMinimumP)(VOID);
+
+#endif // _7ZIP_LARGE_PAGES

 void SetLargePageSize()
 {
@@ -214,7 +220,7 @@ void *BigAlloc(size_t size)
      size2 = (size + ps) & ~ps;
      if (size2 >= size)
      {
-        void *res = VirtualAlloc(NULL, size2, MEM_COMMIT | MEM_LARGE_PAGES, PAGE_READWRITE);
+        void *res = VirtualAlloc(NULL, size2, MEM_COMMIT | MY__MEM_LARGE_PAGES, PAGE_READWRITE);
        if (res)
          return res;
      }
@@ -241,14 +247,14 @@ static void *SzAlloc(ISzAllocPtr p, size_t size) { UNUSED_VAR(p); return MyAlloc
 static void SzFree(ISzAllocPtr p, void *address) { UNUSED_VAR(p); MyFree(address); }
 const ISzAlloc g_Alloc = { SzAlloc, SzFree };

+#ifdef _WIN32
 static void *SzMidAlloc(ISzAllocPtr p, size_t size) { UNUSED_VAR(p); return MidAlloc(size); }
 static void SzMidFree(ISzAllocPtr p, void *address) { UNUSED_VAR(p); MidFree(address); }
-const ISzAlloc g_MidAlloc = { SzMidAlloc, SzMidFree };
-
 static void *SzBigAlloc(ISzAllocPtr p, size_t size) { UNUSED_VAR(p); return BigAlloc(size); }
 static void SzBigFree(ISzAllocPtr p, void *address) { UNUSED_VAR(p); BigFree(address); }
+const ISzAlloc g_MidAlloc = { SzMidAlloc, SzMidFree };
 const ISzAlloc g_BigAlloc = { SzBigAlloc, SzBigFree };
-
+#endif

 /*
  uintptr_t : <stdint.h> C99 (optional)
@@ -280,13 +286,15 @@ const ISzAlloc g_BigAlloc = { SzBigAlloc, SzBigFree };
 */
 #define MY_ALIGN_PTR_DOWN(p, align) ((void *)((((UIntPtr)(p)) & ~((UIntPtr)(align) - 1))))

-#define MY_ALIGN_PTR_UP_PLUS(p, align) MY_ALIGN_PTR_DOWN(((char *)(p) + (align) + ADJUST_ALLOC_SIZE), align)

-
-#if (_POSIX_C_SOURCE >= 200112L) && !defined(_WIN32)
+#if !defined(_WIN32) && defined(_POSIX_C_SOURCE) && (_POSIX_C_SOURCE >= 200112L)
  #define USE_posix_memalign
 #endif

+#ifndef USE_posix_memalign
+#define MY_ALIGN_PTR_UP_PLUS(p, align) MY_ALIGN_PTR_DOWN(((char *)(p) + (align) + ADJUST_ALLOC_SIZE), align)
+#endif
+
 /*
  This posix_memalign() is for test purposes only.
  We also need special Free() function instead of free(),
@@ -1,5 +1,5 @@
 /* Alloc.h -- Memory allocation functions
-2018-02-19 : Igor Pavlov : Public domain */
+2021-07-13 : Igor Pavlov : Public domain */

 #ifndef __COMMON_ALLOC_H
 #define __COMMON_ALLOC_H
@@ -13,7 +13,7 @@ void MyFree(void *address);

 #ifdef _WIN32

-void SetLargePageSize();
+void SetLargePageSize(void);

 void *MidAlloc(size_t size);
 void MidFree(void *address);
@@ -30,8 +30,15 @@ void BigFree(void *address);
 #endif

 extern const ISzAlloc g_Alloc;
+
+#ifdef _WIN32
 extern const ISzAlloc g_BigAlloc;
 extern const ISzAlloc g_MidAlloc;
+#else
+#define g_BigAlloc g_AlignedAlloc
+#define g_MidAlloc g_AlignedAlloc
+#endif
+
 extern const ISzAlloc g_AlignedAlloc;


@@ -1,5 +1,5 @@
 /* Bcj2.c -- BCJ2 Decoder (Converter for x86 code)
-2018-04-28 : Igor Pavlov : Public domain */
+2021-02-09 : Igor Pavlov : Public domain */

 #include "Precomp.h"

@@ -123,7 +123,7 @@ SRes Bcj2Dec_Decode(CBcj2Dec *p)
        const Byte *src = p->bufs[BCJ2_STREAM_MAIN];
        const Byte *srcLim;
        Byte *dest;
-        SizeT num = p->lims[BCJ2_STREAM_MAIN] - src;
+        SizeT num = (SizeT)(p->lims[BCJ2_STREAM_MAIN] - src);
        
        if (num == 0)
        {
@@ -134,7 +134,7 @@ SRes Bcj2Dec_Decode(CBcj2Dec *p)
        dest = p->dest;
        if (num > (SizeT)(p->destLim - dest))
        {
-          num = p->destLim - dest;
+          num = (SizeT)(p->destLim - dest);
          if (num == 0)
          {
            p->state = BCJ2_DEC_STATE_ORIG;
@@ -168,7 +168,7 @@ SRes Bcj2Dec_Decode(CBcj2Dec *p)
          break;
        }
        
-        num = src - p->bufs[BCJ2_STREAM_MAIN];
+        num = (SizeT)(src - p->bufs[BCJ2_STREAM_MAIN]);
        
        if (src == srcLim)
        {
@@ -228,7 +228,7 @@ SRes Bcj2Dec_Decode(CBcj2Dec *p)
      p->ip += 4;
      val -= p->ip;
      dest = p->dest;
-      rem = p->destLim - dest;
+      rem = (SizeT)(p->destLim - dest);
      
      if (rem < 4)
      {
@@ -1,5 +1,5 @@
 /* Bcj2Enc.c -- BCJ2 Encoder (Converter for x86 code)
-2019-02-02 : Igor Pavlov : Public domain */
+2021-02-09 : Igor Pavlov : Public domain */

 #include "Precomp.h"

@@ -104,7 +104,7 @@ static void Bcj2Enc_Encode_2(CBcj2Enc *p)
        const Byte *src = p->src;
        const Byte *srcLim;
        Byte *dest;
-        SizeT num = p->srcLim - src;
+        SizeT num = (SizeT)(p->srcLim - src);

        if (p->finishMode == BCJ2_ENC_FINISH_MODE_CONTINUE)
        {
@@ -118,7 +118,7 @@ static void Bcj2Enc_Encode_2(CBcj2Enc *p)
        dest = p->bufs[BCJ2_STREAM_MAIN];
        if (num > (SizeT)(p->lims[BCJ2_STREAM_MAIN] - dest))
        {
-          num = p->lims[BCJ2_STREAM_MAIN] - dest;
+          num = (SizeT)(p->lims[BCJ2_STREAM_MAIN] - dest);
          if (num == 0)
          {
            p->state = BCJ2_STREAM_MAIN;
@@ -152,7 +152,7 @@ static void Bcj2Enc_Encode_2(CBcj2Enc *p)
          break;
        }
        
-        num = src - p->src;
+        num = (SizeT)(src - p->src);
        
        if (src == srcLim)
        {
@@ -1,5 +1,5 @@
 /* Bra.c -- Converters for RISC code
-2017-04-04 : Igor Pavlov : Public domain */
+2021-02-09 : Igor Pavlov : Public domain */

 #include "Precomp.h"

@@ -22,7 +22,7 @@ SizeT ARM_Convert(Byte *data, SizeT size, UInt32 ip, int encoding)
    for (;;)
    {
      if (p >= lim)
-        return p - data;
+        return (SizeT)(p - data);
      p += 4;
      if (p[-1] == 0xEB)
        break;
@@ -43,7 +43,7 @@ SizeT ARM_Convert(Byte *data, SizeT size, UInt32 ip, int encoding)
    for (;;)
    {
      if (p >= lim)
-        return p - data;
+        return (SizeT)(p - data);
      p += 4;
      if (p[-1] == 0xEB)
        break;
@@ -78,7 +78,7 @@ SizeT ARMT_Convert(Byte *data, SizeT size, UInt32 ip, int encoding)
    {
      UInt32 b3;
      if (p > lim)
-        return p - data;
+        return (SizeT)(p - data);
      b1 = p[1];
      b3 = p[3];
      p += 2;
@@ -113,7 +113,7 @@ SizeT ARMT_Convert(Byte *data, SizeT size, UInt32 ip, int encoding)
    {
      UInt32 b3;
      if (p > lim)
-        return p - data;
+        return (SizeT)(p - data);
      b1 = p[1];
      b3 = p[3];
      p += 2;
@@ -162,7 +162,7 @@ SizeT PPC_Convert(Byte *data, SizeT size, UInt32 ip, int encoding)
    for (;;)
    {
      if (p >= lim)
-        return p - data;
+        return (SizeT)(p - data);
      p += 4;
      /* if ((v & 0xFC000003) == 0x48000001) */
      if ((p[-4] & 0xFC) == 0x48 && (p[-1] & 3) == 1)
@@ -196,7 +196,7 @@ SizeT SPARC_Convert(Byte *data, SizeT size, UInt32 ip, int encoding)
    for (;;)
    {
      if (p >= lim)
-        return p - data;
+        return (SizeT)(p - data);
      /*
      v = GetBe32(p);
      p += 4;
@@ -1,5 +1,5 @@
 /* Bra86.c -- Converter for x86 code (BCJ)
-2017-04-03 : Igor Pavlov : Public domain */
+2021-02-09 : Igor Pavlov : Public domain */

 #include "Precomp.h"

@@ -25,7 +25,7 @@ SizeT x86_Convert(Byte *data, SizeT size, UInt32 ip, UInt32 *state, int encoding
        break;

    {
-      SizeT d = (SizeT)(p - data - pos);
+      SizeT d = (SizeT)(p - data) - pos;
      pos = (SizeT)(p - data);
      if (p >= limit)
      {
@@ -1,9 +1,13 @@
 /* Compiler.h
-2017-04-03 : Igor Pavlov : Public domain */
+2021-01-05 : Igor Pavlov : Public domain */

 #ifndef __7Z_COMPILER_H
 #define __7Z_COMPILER_H

+  #ifdef __clang__
+    #pragma clang diagnostic ignored "-Wunused-private-field"
+  #endif
+
 #ifdef _MSC_VER

  #ifdef UNDER_CE
@@ -25,6 +29,12 @@
    #pragma warning(disable : 4786) // identifier was truncated to '255' characters in the debug information
  #endif

+  #ifdef __clang__
+    #pragma clang diagnostic ignored "-Wdeprecated-declarations"
+    #pragma clang diagnostic ignored "-Wmicrosoft-exception-spec"
+    // #pragma clang diagnostic ignored "-Wreserved-id-macro"
+  #endif
+
 #endif

 #define UNUSED_VAR(x) (void)x;
@@ -0,0 +1,478 @@
+/* CpuArch.c -- CPU specific code
+2021-07-13 : Igor Pavlov : Public domain */
+
+#include "Precomp.h"
+
+#include "CpuArch.h"
+
+#ifdef MY_CPU_X86_OR_AMD64
+
+#if (defined(_MSC_VER) && !defined(MY_CPU_AMD64)) || defined(__GNUC__)
+#define USE_ASM
+#endif
+
+#if !defined(USE_ASM) && _MSC_VER >= 1500
+#include <intrin.h>
+#endif
+
+#if defined(USE_ASM) && !defined(MY_CPU_AMD64)
+static UInt32 CheckFlag(UInt32 flag)
+{
+  #ifdef _MSC_VER
+  __asm pushfd;
+  __asm pop EAX;
+  __asm mov EDX, EAX;
+  __asm xor EAX, flag;
+  __asm push EAX;
+  __asm popfd;
+  __asm pushfd;
+  __asm pop EAX;
+  __asm xor EAX, EDX;
+  __asm push EDX;
+  __asm popfd;
+  __asm and flag, EAX;
+  #else
+  __asm__ __volatile__ (
+    "pushf\n\t"
+    "pop  %%EAX\n\t"
+    "movl %%EAX,%%EDX\n\t"
+    "xorl %0,%%EAX\n\t"
+    "push %%EAX\n\t"
+    "popf\n\t"
+    "pushf\n\t"
+    "pop  %%EAX\n\t"
+    "xorl %%EDX,%%EAX\n\t"
+    "push %%EDX\n\t"
+    "popf\n\t"
+    "andl %%EAX, %0\n\t":
+    "=c" (flag) : "c" (flag) :
+    "%eax", "%edx");
+  #endif
+  return flag;
+}
+#define CHECK_CPUID_IS_SUPPORTED if (CheckFlag(1 << 18) == 0 || CheckFlag(1 << 21) == 0) return False;
+#else
+#define CHECK_CPUID_IS_SUPPORTED
+#endif
+
+#ifndef USE_ASM
+  #ifdef _MSC_VER
+    #if _MSC_VER >= 1600
+      #define MY__cpuidex  __cpuidex
+    #else
+
+/*
+ __cpuid (function == 4) requires subfunction number in ECX.
+  MSDN: The __cpuid intrinsic clears the ECX register before calling the cpuid instruction.
+   __cpuid() in new MSVC clears ECX.
+   __cpuid() in old MSVC (14.00) doesn't clear ECX
+ We still can use __cpuid for low (function) values that don't require ECX,
+ but __cpuid() in old MSVC will be incorrect for some function values: (function == 4).
+ So here we use the hack for old MSVC to send (subFunction) in ECX register to cpuid instruction,
+ where ECX value is first parameter for FAST_CALL / NO_INLINE function,
+ So the caller of MY__cpuidex_HACK() sets ECX as subFunction, and
+ old MSVC for __cpuid() doesn't change ECX and cpuid instruction gets (subFunction) value.
+ 
+ DON'T remove MY_NO_INLINE and MY_FAST_CALL for MY__cpuidex_HACK() !!!
+*/
+
+static
+MY_NO_INLINE
+void MY_FAST_CALL MY__cpuidex_HACK(UInt32 subFunction, int *CPUInfo, UInt32 function)
+{
+  UNUSED_VAR(subFunction);
+  __cpuid(CPUInfo, function);
+}
+
+      #define MY__cpuidex(info, func, func2)  MY__cpuidex_HACK(func2, info, func)
+      #pragma message("======== MY__cpuidex_HACK WAS USED ========")
+    #endif
+  #else
+     #define MY__cpuidex(info, func, func2)  __cpuid(info, func)
+     #pragma message("======== (INCORRECT ?) cpuid WAS USED ========")
+  #endif
+#endif
+
+
+
+
+void MyCPUID(UInt32 function, UInt32 *a, UInt32 *b, UInt32 *c, UInt32 *d)
+{
+  #ifdef USE_ASM
+
+  #ifdef _MSC_VER
+
+  UInt32 a2, b2, c2, d2;
+  __asm xor EBX, EBX;
+  __asm xor ECX, ECX;
+  __asm xor EDX, EDX;
+  __asm mov EAX, function;
+  __asm cpuid;
+  __asm mov a2, EAX;
+  __asm mov b2, EBX;
+  __asm mov c2, ECX;
+  __asm mov d2, EDX;
+
+  *a = a2;
+  *b = b2;
+  *c = c2;
+  *d = d2;
+
+  #else
+
+  __asm__ __volatile__ (
+  #if defined(MY_CPU_AMD64) && defined(__PIC__)
+    "mov %%rbx, %%rdi;"
+    "cpuid;"
+    "xchg %%rbx, %%rdi;"
+    : "=a" (*a) ,
+      "=D" (*b) ,
+  #elif defined(MY_CPU_X86) && defined(__PIC__)
+    "mov %%ebx, %%edi;"
+    "cpuid;"
+    "xchgl %%ebx, %%edi;"
+    : "=a" (*a) ,
+      "=D" (*b) ,
+  #else
+    "cpuid"
+    : "=a" (*a) ,
+      "=b" (*b) ,
+  #endif
+      "=c" (*c) ,
+      "=d" (*d)
+    : "0" (function), "c"(0) ) ;
+
+  #endif
+  
+  #else
+
+  int CPUInfo[4];
+
+  MY__cpuidex(CPUInfo, (int)function, 0);
+
+  *a = (UInt32)CPUInfo[0];
+  *b = (UInt32)CPUInfo[1];
+  *c = (UInt32)CPUInfo[2];
+  *d = (UInt32)CPUInfo[3];
+
+  #endif
+}
+
+BoolInt x86cpuid_CheckAndRead(Cx86cpuid *p)
+{
+  CHECK_CPUID_IS_SUPPORTED
+  MyCPUID(0, &p->maxFunc, &p->vendor[0], &p->vendor[2], &p->vendor[1]);
+  MyCPUID(1, &p->ver, &p->b, &p->c, &p->d);
+  return True;
+}
+
+static const UInt32 kVendors[][3] =
+{
+  { 0x756E6547, 0x49656E69, 0x6C65746E},
+  { 0x68747541, 0x69746E65, 0x444D4163},
+  { 0x746E6543, 0x48727561, 0x736C7561}
+};
+
+int x86cpuid_GetFirm(const Cx86cpuid *p)
+{
+  unsigned i;
+  for (i = 0; i < sizeof(kVendors) / sizeof(kVendors[i]); i++)
+  {
+    const UInt32 *v = kVendors[i];
+    if (v[0] == p->vendor[0] &&
+        v[1] == p->vendor[1] &&
+        v[2] == p->vendor[2])
+      return (int)i;
+  }
+  return -1;
+}
+
+BoolInt CPU_Is_InOrder()
+{
+  Cx86cpuid p;
+  int firm;
+  UInt32 family, model;
+  if (!x86cpuid_CheckAndRead(&p))
+    return True;
+
+  family = x86cpuid_GetFamily(p.ver);
+  model = x86cpuid_GetModel(p.ver);
+  
+  firm = x86cpuid_GetFirm(&p);
+
+  switch (firm)
+  {
+    case CPU_FIRM_INTEL: return (family < 6 || (family == 6 && (
+        /* In-Order Atom CPU */
+           model == 0x1C  /* 45 nm, N4xx, D4xx, N5xx, D5xx, 230, 330 */
+        || model == 0x26  /* 45 nm, Z6xx */
+        || model == 0x27  /* 32 nm, Z2460 */
+        || model == 0x35  /* 32 nm, Z2760 */
+        || model == 0x36  /* 32 nm, N2xxx, D2xxx */
+        )));
+    case CPU_FIRM_AMD: return (family < 5 || (family == 5 && (model < 6 || model == 0xA)));
+    case CPU_FIRM_VIA: return (family < 6 || (family == 6 && model < 0xF));
+  }
+  return True;
+}
+
+#if !defined(MY_CPU_AMD64) && defined(_WIN32)
+#include <Windows.h>
+static BoolInt CPU_Sys_Is_SSE_Supported()
+{
+  OSVERSIONINFO vi;
+  vi.dwOSVersionInfoSize = sizeof(vi);
+  if (!GetVersionEx(&vi))
+    return False;
+  return (vi.dwMajorVersion >= 5);
+}
+#define CHECK_SYS_SSE_SUPPORT if (!CPU_Sys_Is_SSE_Supported()) return False;
+#else
+#define CHECK_SYS_SSE_SUPPORT
+#endif
+
+
+static UInt32 X86_CPUID_ECX_Get_Flags()
+{
+  Cx86cpuid p;
+  CHECK_SYS_SSE_SUPPORT
+  if (!x86cpuid_CheckAndRead(&p))
+    return 0;
+  return p.c;
+}
+
+BoolInt CPU_IsSupported_AES()
+{
+  return (X86_CPUID_ECX_Get_Flags() >> 25) & 1;
+}
+
+BoolInt CPU_IsSupported_SSSE3()
+{
+  return (X86_CPUID_ECX_Get_Flags() >> 9) & 1;
+}
+
+BoolInt CPU_IsSupported_SSE41()
+{
+  return (X86_CPUID_ECX_Get_Flags() >> 19) & 1;
+}
+
+BoolInt CPU_IsSupported_SHA()
+{
+  Cx86cpuid p;
+  CHECK_SYS_SSE_SUPPORT
+  if (!x86cpuid_CheckAndRead(&p))
+    return False;
+
+  if (p.maxFunc < 7)
+    return False;
+  {
+    UInt32 d[4] = { 0 };
+    MyCPUID(7, &d[0], &d[1], &d[2], &d[3]);
+    return (d[1] >> 29) & 1;
+  }
+}
+
+// #include <stdio.h>
+
+#ifdef _WIN32
+#include <Windows.h>
+#endif
+
+BoolInt CPU_IsSupported_AVX2()
+{
+  Cx86cpuid p;
+  CHECK_SYS_SSE_SUPPORT
+
+  #ifdef _WIN32
+  #define MY__PF_XSAVE_ENABLED  17
+  if (!IsProcessorFeaturePresent(MY__PF_XSAVE_ENABLED))
+    return False;
+  #endif
+
+  if (!x86cpuid_CheckAndRead(&p))
+    return False;
+  if (p.maxFunc < 7)
+    return False;
+  {
+    UInt32 d[4] = { 0 };
+    MyCPUID(7, &d[0], &d[1], &d[2], &d[3]);
+    // printf("\ncpuid(7): ebx=%8x ecx=%8x\n", d[1], d[2]);
+    return 1
+      & (d[1] >> 5); // avx2
+  }
+}
+
+BoolInt CPU_IsSupported_VAES_AVX2()
+{
+  Cx86cpuid p;
+  CHECK_SYS_SSE_SUPPORT
+
+  #ifdef _WIN32
+  #define MY__PF_XSAVE_ENABLED  17
+  if (!IsProcessorFeaturePresent(MY__PF_XSAVE_ENABLED))
+    return False;
+  #endif
+
+  if (!x86cpuid_CheckAndRead(&p))
+    return False;
+  if (p.maxFunc < 7)
+    return False;
+  {
+    UInt32 d[4] = { 0 };
+    MyCPUID(7, &d[0], &d[1], &d[2], &d[3]);
+    // printf("\ncpuid(7): ebx=%8x ecx=%8x\n", d[1], d[2]);
+    return 1
+      & (d[1] >> 5) // avx2
+      // & (d[1] >> 31) // avx512vl
+      & (d[2] >> 9); // vaes // VEX-256/EVEX
+  }
+}
+
+BoolInt CPU_IsSupported_PageGB()
+{
+  Cx86cpuid cpuid;
+  if (!x86cpuid_CheckAndRead(&cpuid))
+    return False;
+  {
+    UInt32 d[4] = { 0 };
+    MyCPUID(0x80000000, &d[0], &d[1], &d[2], &d[3]);
+    if (d[0] < 0x80000001)
+      return False;
+  }
+  {
+    UInt32 d[4] = { 0 };
+    MyCPUID(0x80000001, &d[0], &d[1], &d[2], &d[3]);
+    return (d[3] >> 26) & 1;
+  }
+}
+
+
+#elif defined(MY_CPU_ARM_OR_ARM64)
+
+#ifdef _WIN32
+
+#include <Windows.h>
+
+BoolInt CPU_IsSupported_CRC32()  { return IsProcessorFeaturePresent(PF_ARM_V8_CRC32_INSTRUCTIONS_AVAILABLE) ? 1 : 0; }
+BoolInt CPU_IsSupported_CRYPTO() { return IsProcessorFeaturePresent(PF_ARM_V8_CRYPTO_INSTRUCTIONS_AVAILABLE) ? 1 : 0; }
+BoolInt CPU_IsSupported_NEON()   { return IsProcessorFeaturePresent(PF_ARM_NEON_INSTRUCTIONS_AVAILABLE) ? 1 : 0; }
+
+#else
+
+#if defined(__APPLE__)
+
+/*
+#include <stdio.h>
+#include <string.h>
+static void Print_sysctlbyname(const char *name)
+{
+  size_t bufSize = 256;
+  char buf[256];
+  int res = sysctlbyname(name, &buf, &bufSize, NULL, 0);
+  {
+    int i;
+    printf("\nres = %d : %s : '%s' : bufSize = %d, numeric", res, name, buf, (unsigned)bufSize);
+    for (i = 0; i < 20; i++)
+      printf(" %2x", (unsigned)(Byte)buf[i]);
+
+  }
+}
+*/
+
+static BoolInt My_sysctlbyname_Get_BoolInt(const char *name)
+{
+  UInt32 val = 0;
+  if (My_sysctlbyname_Get_UInt32(name, &val) == 0 && val == 1)
+    return 1;
+  return 0;
+}
+
+  /*
+  Print_sysctlbyname("hw.pagesize");
+  Print_sysctlbyname("machdep.cpu.brand_string");
+  */
+
+BoolInt CPU_IsSupported_CRC32(void)
+{
+  return My_sysctlbyname_Get_BoolInt("hw.optional.armv8_crc32");
+}
+
+BoolInt CPU_IsSupported_NEON(void)
+{
+  return My_sysctlbyname_Get_BoolInt("hw.optional.neon");
+}
+
+#ifdef MY_CPU_ARM64
+#define APPLE_CRYPTO_SUPPORT_VAL 1
+#else
+#define APPLE_CRYPTO_SUPPORT_VAL 0
+#endif
+
+BoolInt CPU_IsSupported_SHA1(void) { return APPLE_CRYPTO_SUPPORT_VAL; }
+BoolInt CPU_IsSupported_SHA2(void) { return APPLE_CRYPTO_SUPPORT_VAL; }
+BoolInt CPU_IsSupported_AES (void) { return APPLE_CRYPTO_SUPPORT_VAL; }
+
+
+#else // __APPLE__
+
+#include <sys/auxv.h>
+
+#define USE_HWCAP
+
+#ifdef USE_HWCAP
+
+#include <asm/hwcap.h>
+
+  #define MY_HWCAP_CHECK_FUNC_2(name1, name2) \
+  BoolInt CPU_IsSupported_ ## name1() { return (getauxval(AT_HWCAP)  & (HWCAP_  ## name2)) ? 1 : 0; }
+
+#ifdef MY_CPU_ARM64
+  #define MY_HWCAP_CHECK_FUNC(name) \
+  MY_HWCAP_CHECK_FUNC_2(name, name)
+  MY_HWCAP_CHECK_FUNC_2(NEON, ASIMD)
+// MY_HWCAP_CHECK_FUNC (ASIMD)
+#elif defined(MY_CPU_ARM)
+  #define MY_HWCAP_CHECK_FUNC(name) \
+  BoolInt CPU_IsSupported_ ## name() { return (getauxval(AT_HWCAP2) & (HWCAP2_ ## name)) ? 1 : 0; }
+  MY_HWCAP_CHECK_FUNC_2(NEON, NEON)
+#endif
+
+#else // USE_HWCAP
+
+  #define MY_HWCAP_CHECK_FUNC(name) \
+  BoolInt CPU_IsSupported_ ## name() { return 0; }
+  MY_HWCAP_CHECK_FUNC(NEON)
+
+#endif // USE_HWCAP
+
+MY_HWCAP_CHECK_FUNC (CRC32)
+MY_HWCAP_CHECK_FUNC (SHA1)
+MY_HWCAP_CHECK_FUNC (SHA2)
+MY_HWCAP_CHECK_FUNC (AES)
+
+#endif // __APPLE__
+#endif // _WIN32
+
+#endif // MY_CPU_ARM_OR_ARM64
+
+
+
+#ifdef __APPLE__
+
+#include <sys/sysctl.h>
+
+int My_sysctlbyname_Get(const char *name, void *buf, size_t *bufSize)
+{
+  return sysctlbyname(name, buf, bufSize, NULL, 0);
+}
+
+int My_sysctlbyname_Get_UInt32(const char *name, UInt32 *val)
+{
+  size_t bufSize = sizeof(*val);
+  int res = My_sysctlbyname_Get(name, val, &bufSize);
+  if (res == 0 && bufSize != sizeof(*val))
+    return EFAULT;
+  return res;
+}
+
+#endif
@@ -1,5 +1,5 @@
 /* CpuArch.h -- CPU specific code
-2018-02-18 : Igor Pavlov : Public domain */
+2022-07-15 : Igor Pavlov : Public domain */

 #ifndef __CPU_ARCH_H
 #define __CPU_ARCH_H
@@ -14,6 +14,10 @@ MY_CPU_BE means that CPU is BIG ENDIAN.
 If MY_CPU_LE and MY_CPU_BE are not defined, we don't know about ENDIANNESS of platform.

 MY_CPU_LE_UNALIGN means that CPU is LITTLE ENDIAN and CPU supports unaligned memory accesses.
+
+MY_CPU_64BIT means that processor can work with 64-bit registers.
+  MY_CPU_64BIT can be used to select fast code branch
+  MY_CPU_64BIT doesn't mean that (sizeof(void *) == 8)
 */

 #if  defined(_M_X64) \
@@ -24,8 +28,10 @@ MY_CPU_LE_UNALIGN means that CPU is LITTLE ENDIAN and CPU supports unaligned mem
  #define MY_CPU_AMD64
  #ifdef __ILP32__
    #define MY_CPU_NAME "x32"
+    #define MY_CPU_SIZEOF_POINTER 4
  #else
    #define MY_CPU_NAME "x64"
+    #define MY_CPU_SIZEOF_POINTER 8
  #endif
  #define MY_CPU_64BIT
 #endif
@@ -35,7 +41,8 @@ MY_CPU_LE_UNALIGN means that CPU is LITTLE ENDIAN and CPU supports unaligned mem
  || defined(__i386__)
  #define MY_CPU_X86
  #define MY_CPU_NAME "x86"
-  #define MY_CPU_32BIT
+  /* #define MY_CPU_32BIT */
+  #define MY_CPU_SIZEOF_POINTER 4
 #endif


@@ -59,8 +66,14 @@ MY_CPU_LE_UNALIGN means that CPU is LITTLE ENDIAN and CPU supports unaligned mem
  || defined(__THUMBEL__) \
  || defined(__THUMBEB__)
  #define MY_CPU_ARM
-  #define MY_CPU_NAME "arm"
-  #define MY_CPU_32BIT
+
+  #if defined(__thumb__) || defined(__THUMBEL__) || defined(_M_ARMT)
+    #define MY_CPU_NAME "armt"
+  #else
+    #define MY_CPU_NAME "arm"
+  #endif
+  /* #define MY_CPU_32BIT */
+  #define MY_CPU_SIZEOF_POINTER 4
 #endif


@@ -84,26 +97,41 @@ MY_CPU_LE_UNALIGN means that CPU is LITTLE ENDIAN and CPU supports unaligned mem


 #if  defined(__ppc64__) \
-  || defined(__powerpc64__)
+  || defined(__powerpc64__) \
+  || defined(__ppc__) \
+  || defined(__powerpc__) \
+  || defined(__PPC__) \
+  || defined(_POWER)
+
+#if  defined(__ppc64__) \
+  || defined(__powerpc64__) \
+  || defined(_LP64) \
+  || defined(__64BIT__)
  #ifdef __ILP32__
    #define MY_CPU_NAME "ppc64-32"
+    #define MY_CPU_SIZEOF_POINTER 4
  #else
    #define MY_CPU_NAME "ppc64"
+    #define MY_CPU_SIZEOF_POINTER 8
  #endif
  #define MY_CPU_64BIT
-#elif defined(__ppc__) \
-  || defined(__powerpc__)
+#else
  #define MY_CPU_NAME "ppc"
-  #define MY_CPU_32BIT
+  #define MY_CPU_SIZEOF_POINTER 4
+  /* #define MY_CPU_32BIT */
+#endif
 #endif


-#if  defined(__sparc64__)
-  #define MY_CPU_NAME "sparc64"
-  #define MY_CPU_64BIT
-#elif defined(__sparc__)
-  #define MY_CPU_NAME "sparc"
-  /* #define MY_CPU_32BIT */
+#if  defined(__riscv) \
+  || defined(__riscv__)
+  #if __riscv_xlen == 32
+    #define MY_CPU_NAME "riscv32"
+  #elif __riscv_xlen == 64
+    #define MY_CPU_NAME "riscv64"
+  #else
+    #define MY_CPU_NAME "riscv"
+  #endif
 #endif


@@ -111,6 +139,10 @@ MY_CPU_LE_UNALIGN means that CPU is LITTLE ENDIAN and CPU supports unaligned mem
 #define MY_CPU_X86_OR_AMD64
 #endif

+#if defined(MY_CPU_ARM) || defined(MY_CPU_ARM64)
+#define MY_CPU_ARM_OR_ARM64
+#endif
+

 #ifdef _WIN32

@@ -170,6 +202,40 @@ MY_CPU_LE_UNALIGN means that CPU is LITTLE ENDIAN and CPU supports unaligned mem
  #error Stop_Compiling_Bad_32_64_BIT
 #endif

+#ifdef __SIZEOF_POINTER__
+  #ifdef MY_CPU_SIZEOF_POINTER
+    #if MY_CPU_SIZEOF_POINTER != __SIZEOF_POINTER__
+      #error Stop_Compiling_Bad_MY_CPU_PTR_SIZE
+    #endif
+  #else
+    #define MY_CPU_SIZEOF_POINTER  __SIZEOF_POINTER__
+  #endif
+#endif
+
+#if defined(MY_CPU_SIZEOF_POINTER) && (MY_CPU_SIZEOF_POINTER == 4)
+#if defined (_LP64)
+      #error Stop_Compiling_Bad_MY_CPU_PTR_SIZE
+#endif
+#endif
+
+#ifdef _MSC_VER
+  #if _MSC_VER >= 1300
+    #define MY_CPU_pragma_pack_push_1   __pragma(pack(push, 1))
+    #define MY_CPU_pragma_pop           __pragma(pack(pop))
+  #else
+    #define MY_CPU_pragma_pack_push_1
+    #define MY_CPU_pragma_pop
+  #endif
+#else
+  #ifdef __xlC__
+    #define MY_CPU_pragma_pack_push_1   _Pragma("pack(1)")
+    #define MY_CPU_pragma_pop           _Pragma("pack()")
+  #else
+    #define MY_CPU_pragma_pack_push_1   _Pragma("pack(push, 1)")
+    #define MY_CPU_pragma_pop           _Pragma("pack(pop)")
+  #endif
+#endif
+

 #ifndef MY_CPU_NAME
  #ifdef MY_CPU_LE
@@ -189,8 +255,12 @@ MY_CPU_LE_UNALIGN means that CPU is LITTLE ENDIAN and CPU supports unaligned mem

 #ifdef MY_CPU_LE
  #if defined(MY_CPU_X86_OR_AMD64) \
-      || defined(MY_CPU_ARM64) \
-      || defined(__ARM_FEATURE_UNALIGNED)
+      || defined(MY_CPU_ARM64)
+    #define MY_CPU_LE_UNALIGN
+    #define MY_CPU_LE_UNALIGN_64
+  #elif defined(__ARM_FEATURE_UNALIGNED)
+    /* gcc9 for 32-bit arm can use LDRD instruction that requires 32-bit alignment.
+       So we can't use unaligned 64-bit operations. */
    #define MY_CPU_LE_UNALIGN
  #endif
 #endif
@@ -200,11 +270,15 @@ MY_CPU_LE_UNALIGN means that CPU is LITTLE ENDIAN and CPU supports unaligned mem

 #define GetUi16(p) (*(const UInt16 *)(const void *)(p))
 #define GetUi32(p) (*(const UInt32 *)(const void *)(p))
+#ifdef MY_CPU_LE_UNALIGN_64
 #define GetUi64(p) (*(const UInt64 *)(const void *)(p))
+#endif

-#define SetUi16(p, v) { *(UInt16 *)(p) = (v); }
-#define SetUi32(p, v) { *(UInt32 *)(p) = (v); }
-#define SetUi64(p, v) { *(UInt64 *)(p) = (v); }
+#define SetUi16(p, v) { *(UInt16 *)(void *)(p) = (v); }
+#define SetUi32(p, v) { *(UInt32 *)(void *)(p) = (v); }
+#ifdef MY_CPU_LE_UNALIGN_64
+#define SetUi64(p, v) { *(UInt64 *)(void *)(p) = (v); }
+#endif

 #else

@@ -218,8 +292,6 @@ MY_CPU_LE_UNALIGN means that CPU is LITTLE ENDIAN and CPU supports unaligned mem
    ((UInt32)((const Byte *)(p))[2] << 16) | \
    ((UInt32)((const Byte *)(p))[3] << 24))

-#define GetUi64(p) (GetUi32(p) | ((UInt64)GetUi32(((const Byte *)(p)) + 4) << 32))
-
 #define SetUi16(p, v) { Byte *_ppp_ = (Byte *)(p); UInt32 _vvv_ = (v); \
    _ppp_[0] = (Byte)_vvv_; \
    _ppp_[1] = (Byte)(_vvv_ >> 8); }
@@ -230,19 +302,29 @@ MY_CPU_LE_UNALIGN means that CPU is LITTLE ENDIAN and CPU supports unaligned mem
    _ppp_[2] = (Byte)(_vvv_ >> 16); \
    _ppp_[3] = (Byte)(_vvv_ >> 24); }

+#endif
+
+
+#ifndef MY_CPU_LE_UNALIGN_64
+
+#define GetUi64(p) (GetUi32(p) | ((UInt64)GetUi32(((const Byte *)(p)) + 4) << 32))
+
 #define SetUi64(p, v) { Byte *_ppp2_ = (Byte *)(p); UInt64 _vvv2_ = (v); \
    SetUi32(_ppp2_    , (UInt32)_vvv2_); \
    SetUi32(_ppp2_ + 4, (UInt32)(_vvv2_ >> 32)); }

 #endif

+
+
+
 #ifdef __has_builtin
  #define MY__has_builtin(x) __has_builtin(x)
 #else
  #define MY__has_builtin(x) 0
 #endif

-#if defined(MY_CPU_LE_UNALIGN) && /* defined(_WIN64) && */ (_MSC_VER >= 1300)
+#if defined(MY_CPU_LE_UNALIGN) && /* defined(_WIN64) && */ defined(_MSC_VER) && (_MSC_VER >= 1300)

 /* Note: we use bswap instruction, that is unsupported in 386 cpu */

@@ -253,8 +335,8 @@ MY_CPU_LE_UNALIGN means that CPU is LITTLE ENDIAN and CPU supports unaligned mem
 #pragma intrinsic(_byteswap_uint64)

 /* #define GetBe16(p) _byteswap_ushort(*(const UInt16 *)(const Byte *)(p)) */
-#define GetBe32(p) _byteswap_ulong(*(const UInt32 *)(const Byte *)(p))
-#define GetBe64(p) _byteswap_uint64(*(const UInt64 *)(const Byte *)(p))
+#define GetBe32(p) _byteswap_ulong (*(const UInt32 *)(const void *)(p))
+#define GetBe64(p) _byteswap_uint64(*(const UInt64 *)(const void *)(p))

 #define SetBe32(p, v) (*(UInt32 *)(void *)(p)) = _byteswap_ulong(v)

@@ -262,9 +344,9 @@ MY_CPU_LE_UNALIGN means that CPU is LITTLE ENDIAN and CPU supports unaligned mem
       (defined(__GNUC__) && (__GNUC__ > 4 || (__GNUC__ == 4 && __GNUC_MINOR__ >= 3))) \
    || (defined(__clang__) && MY__has_builtin(__builtin_bswap16)) )

-/* #define GetBe16(p) __builtin_bswap16(*(const UInt16 *)(const Byte *)(p)) */
-#define GetBe32(p) __builtin_bswap32(*(const UInt32 *)(const Byte *)(p))
-#define GetBe64(p) __builtin_bswap64(*(const UInt64 *)(const Byte *)(p))
+/* #define GetBe16(p) __builtin_bswap16(*(const UInt16 *)(const void *)(p)) */
+#define GetBe32(p) __builtin_bswap32(*(const UInt32 *)(const void *)(p))
+#define GetBe64(p) __builtin_bswap64(*(const UInt64 *)(const void *)(p))

 #define SetBe32(p, v) (*(UInt32 *)(void *)(p)) = __builtin_bswap32(v)

@@ -325,10 +407,37 @@ int x86cpuid_GetFirm(const Cx86cpuid *p);
 #define x86cpuid_GetModel(ver)  (((ver >> 12) &  0xF0) | ((ver >> 4) & 0xF))
 #define x86cpuid_GetStepping(ver) (ver & 0xF)

-BoolInt CPU_Is_InOrder();
-BoolInt CPU_Is_Aes_Supported();
-BoolInt CPU_IsSupported_PageGB();
+BoolInt CPU_Is_InOrder(void);

+BoolInt CPU_IsSupported_AES(void);
+BoolInt CPU_IsSupported_AVX2(void);
+BoolInt CPU_IsSupported_VAES_AVX2(void);
+BoolInt CPU_IsSupported_SSSE3(void);
+BoolInt CPU_IsSupported_SSE41(void);
+BoolInt CPU_IsSupported_SHA(void);
+BoolInt CPU_IsSupported_PageGB(void);
+
+#elif defined(MY_CPU_ARM_OR_ARM64)
+
+BoolInt CPU_IsSupported_CRC32(void);
+BoolInt CPU_IsSupported_NEON(void);
+
+#if defined(_WIN32)
+BoolInt CPU_IsSupported_CRYPTO(void);
+#define CPU_IsSupported_SHA1  CPU_IsSupported_CRYPTO
+#define CPU_IsSupported_SHA2  CPU_IsSupported_CRYPTO
+#define CPU_IsSupported_AES   CPU_IsSupported_CRYPTO
+#else
+BoolInt CPU_IsSupported_SHA1(void);
+BoolInt CPU_IsSupported_SHA2(void);
+BoolInt CPU_IsSupported_AES(void);
+#endif
+
+#endif
+
+#if defined(__APPLE__)
+int My_sysctlbyname_Get(const char *name, void *buf, size_t *bufSize);
+int My_sysctlbyname_Get_UInt32(const char *name, UInt32 *val);
 #endif

 EXTERN_C_END
@@ -0,0 +1,169 @@
+/* Delta.c -- Delta converter
+2021-02-09 : Igor Pavlov : Public domain */
+
+#include "Precomp.h"
+
+#include "Delta.h"
+
+void Delta_Init(Byte *state)
+{
+  unsigned i;
+  for (i = 0; i < DELTA_STATE_SIZE; i++)
+    state[i] = 0;
+}
+
+
+void Delta_Encode(Byte *state, unsigned delta, Byte *data, SizeT size)
+{
+  Byte temp[DELTA_STATE_SIZE];
+
+  if (size == 0)
+    return;
+
+  {
+    unsigned i = 0;
+    do
+      temp[i] = state[i];
+    while (++i != delta);
+  }
+
+  if (size <= delta)
+  {
+    unsigned i = 0, k;
+    do
+    {
+      Byte b = *data;
+      *data++ = (Byte)(b - temp[i]);
+      temp[i] = b;
+    }
+    while (++i != size);
+    
+    k = 0;
+    
+    do
+    {
+      if (i == delta)
+        i = 0;
+      state[k] = temp[i++];
+    }
+    while (++k != delta);
+    
+    return;
+  }
+    
+  {
+    Byte *p = data + size - delta;
+    {
+      unsigned i = 0;
+      do
+        state[i] = *p++;
+      while (++i != delta);
+    }
+    {
+      const Byte *lim = data + delta;
+      ptrdiff_t dif = -(ptrdiff_t)delta;
+      
+      if (((ptrdiff_t)size + dif) & 1)
+      {
+        --p;  *p = (Byte)(*p - p[dif]);
+      }
+
+      while (p != lim)
+      {
+        --p;  *p = (Byte)(*p - p[dif]);
+        --p;  *p = (Byte)(*p - p[dif]);
+      }
+      
+      dif = -dif;
+      
+      do
+      {
+        --p;  *p = (Byte)(*p - temp[--dif]);
+      }
+      while (dif != 0);
+    }
+  }
+}
+
+
+void Delta_Decode(Byte *state, unsigned delta, Byte *data, SizeT size)
+{
+  unsigned i;
+  const Byte *lim;
+
+  if (size == 0)
+    return;
+  
+  i = 0;
+  lim = data + size;
+  
+  if (size <= delta)
+  {
+    do
+      *data = (Byte)(*data + state[i++]);
+    while (++data != lim);
+
+    for (; delta != i; state++, delta--)
+      *state = state[i];
+    data -= i;
+  }
+  else
+  {
+    /*
+    #define B(n) b ## n
+    #define I(n) Byte B(n) = state[n];
+    #define U(n) { B(n) = (Byte)((B(n)) + *data++); data[-1] = (B(n)); }
+    #define F(n) if (data != lim) { U(n) }
+
+    if (delta == 1)
+    {
+      I(0)
+      if ((lim - data) & 1) { U(0) }
+      while (data != lim) { U(0) U(0) }
+      data -= 1;
+    }
+    else if (delta == 2)
+    {
+      I(0) I(1)
+      lim -= 1; while (data < lim) { U(0) U(1) }
+      lim += 1; F(0)
+      data -= 2;
+    }
+    else if (delta == 3)
+    {
+      I(0) I(1) I(2)
+      lim -= 2; while (data < lim) { U(0) U(1) U(2) }
+      lim += 2; F(0) F(1)
+      data -= 3;
+    }
+    else if (delta == 4)
+    {
+      I(0) I(1) I(2) I(3)
+      lim -= 3; while (data < lim) { U(0) U(1) U(2) U(3) }
+      lim += 3; F(0) F(1) F(2)
+      data -= 4;
+    }
+    else
+    */
+    {
+      do
+      {
+        *data = (Byte)(*data + state[i++]);
+        data++;
+      }
+      while (i != delta);
+  
+      {
+        ptrdiff_t dif = -(ptrdiff_t)delta;
+        do
+          *data = (Byte)(*data + data[dif]);
+        while (++data != lim);
+        data += dif;
+      }
+    }
+  }
+
+  do
+    *state++ = *data;
+  while (++data != lim);
+}
@@ -1,16 +1,20 @@
 /* DllSecur.c -- DLL loading security
-2018-02-21 : Igor Pavlov : Public domain */
+2022-07-15 : Igor Pavlov : Public domain */

 #include "Precomp.h"

 #ifdef _WIN32

-#include <windows.h>
+#include <Windows.h>

 #include "DllSecur.h"

 #ifndef UNDER_CE

+#if defined(__GNUC__) && (__GNUC__ >= 8)
+  #pragma GCC diagnostic ignored "-Wcast-function-type"
+#endif
+
 typedef BOOL (WINAPI *Func_SetDefaultDllDirectories)(DWORD DirectoryFlags);

 #define MY_LOAD_LIBRARY_SEARCH_USER_DIRS 0x400
@@ -33,17 +37,19 @@ static const char * const g_Dlls =

 #endif

+// #define MY_CAST_FUNC  (void(*)())
+#define MY_CAST_FUNC
+
 void My_SetDefaultDllDirectories()
 {
  #ifndef UNDER_CE
  
    OSVERSIONINFO vi;
    vi.dwOSVersionInfoSize = sizeof(vi);
-    GetVersionEx(&vi);
    if (!GetVersionEx(&vi) || vi.dwMajorVersion != 6 || vi.dwMinorVersion != 0)
    {
      Func_SetDefaultDllDirectories setDllDirs = (Func_SetDefaultDllDirectories)
-          GetProcAddress(GetModuleHandle(TEXT("kernel32.dll")), "SetDefaultDllDirectories");
+          MY_CAST_FUNC GetProcAddress(GetModuleHandle(TEXT("kernel32.dll")), "SetDefaultDllDirectories");
      if (setDllDirs)
        if (setDllDirs(MY_LOAD_LIBRARY_SEARCH_SYSTEM32 | MY_LOAD_LIBRARY_SEARCH_USER_DIRS))
          return;
@@ -66,7 +72,7 @@ void LoadSecurityDlls()
    if (!GetVersionEx(&vi) || vi.dwMajorVersion != 6 || vi.dwMinorVersion != 0)
    {
      Func_SetDefaultDllDirectories setDllDirs = (Func_SetDefaultDllDirectories)
-          GetProcAddress(GetModuleHandle(TEXT("kernel32.dll")), "SetDefaultDllDirectories");
+          MY_CAST_FUNC GetProcAddress(GetModuleHandle(TEXT("kernel32.dll")), "SetDefaultDllDirectories");
      if (setDllDirs)
        if (setDllDirs(MY_LOAD_LIBRARY_SEARCH_SYSTEM32 | MY_LOAD_LIBRARY_SEARCH_USER_DIRS))
          return;
@@ -10,8 +10,8 @@ EXTERN_C_BEGIN

 #ifdef _WIN32

-void My_SetDefaultDllDirectories();
-void LoadSecurityDlls();
+void My_SetDefaultDllDirectories(void);
+void LoadSecurityDlls(void);

 #endif

@@ -1,5 +1,5 @@
 /* LzFind.h -- Match finder for LZ algorithms
-2017-06-10 : Igor Pavlov : Public domain */
+2021-07-13 : Igor Pavlov : Public domain */

 #ifndef __LZ_FIND_H
 #define __LZ_FIND_H
@@ -15,7 +15,7 @@ typedef struct _CMatchFinder
  Byte *buffer;
  UInt32 pos;
  UInt32 posLimit;
-  UInt32 streamPos;
+  UInt32 streamPos;  /* wrap over Zero is allowed (streamPos < pos). Use (UInt32)(streamPos - pos) */
  UInt32 lenLimit;

  UInt32 cyclicBufferPos;
@@ -51,17 +51,19 @@ typedef struct _CMatchFinder
  UInt64 expectedDataSize;
 } CMatchFinder;

-#define Inline_MatchFinder_GetPointerToCurrentPos(p) ((p)->buffer)
+#define Inline_MatchFinder_GetPointerToCurrentPos(p) ((const Byte *)(p)->buffer)

-#define Inline_MatchFinder_GetNumAvailableBytes(p) ((p)->streamPos - (p)->pos)
+#define Inline_MatchFinder_GetNumAvailableBytes(p) ((UInt32)((p)->streamPos - (p)->pos))

+/*
 #define Inline_MatchFinder_IsFinishedOK(p) \
    ((p)->streamEndWasReached \
        && (p)->streamPos == (p)->pos \
        && (!(p)->directInput || (p)->directInputRem == 0))
+*/
      
 int MatchFinder_NeedMove(CMatchFinder *p);
-Byte *MatchFinder_GetPointerToCurrentPos(CMatchFinder *p);
+/* Byte *MatchFinder_GetPointerToCurrentPos(CMatchFinder *p); */
 void MatchFinder_MoveBlock(CMatchFinder *p);
 void MatchFinder_ReadIfRequired(CMatchFinder *p);

@@ -76,10 +78,21 @@ int MatchFinder_Create(CMatchFinder *p, UInt32 historySize,
    ISzAllocPtr alloc);
 void MatchFinder_Free(CMatchFinder *p, ISzAllocPtr alloc);
 void MatchFinder_Normalize3(UInt32 subValue, CLzRef *items, size_t numItems);
-void MatchFinder_ReduceOffsets(CMatchFinder *p, UInt32 subValue);
+// void MatchFinder_ReduceOffsets(CMatchFinder *p, UInt32 subValue);
+
+/*
+#define Inline_MatchFinder_InitPos(p, val) \
+    (p)->pos = (val); \
+    (p)->streamPos = (val);
+*/
+
+#define Inline_MatchFinder_ReduceOffsets(p, subValue) \
+    (p)->pos -= (subValue); \
+    (p)->streamPos -= (subValue);
+

 UInt32 * GetMatchesSpec1(UInt32 lenLimit, UInt32 curMatch, UInt32 pos, const Byte *buffer, CLzRef *son,
-    UInt32 _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 _cutValue,
+    size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 _cutValue,
    UInt32 *distances, UInt32 maxLen);

 /*
@@ -91,7 +104,7 @@ Conditions:
 typedef void (*Mf_Init_Func)(void *object);
 typedef UInt32 (*Mf_GetNumAvailableBytes_Func)(void *object);
 typedef const Byte * (*Mf_GetPointerToCurrentPos_Func)(void *object);
-typedef UInt32 (*Mf_GetMatches_Func)(void *object, UInt32 *distances);
+typedef UInt32 * (*Mf_GetMatches_Func)(void *object, UInt32 *distances);
 typedef void (*Mf_Skip_Func)(void *object, UInt32);

 typedef struct _IMatchFinder
@@ -101,21 +114,23 @@ typedef struct _IMatchFinder
  Mf_GetPointerToCurrentPos_Func GetPointerToCurrentPos;
  Mf_GetMatches_Func GetMatches;
  Mf_Skip_Func Skip;
-} IMatchFinder;
+} IMatchFinder2;

-void MatchFinder_CreateVTable(CMatchFinder *p, IMatchFinder *vTable);
+void MatchFinder_CreateVTable(CMatchFinder *p, IMatchFinder2 *vTable);

 void MatchFinder_Init_LowHash(CMatchFinder *p);
 void MatchFinder_Init_HighHash(CMatchFinder *p);
-void MatchFinder_Init_3(CMatchFinder *p, int readData);
+void MatchFinder_Init_4(CMatchFinder *p);
 void MatchFinder_Init(CMatchFinder *p);

-UInt32 Bt3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances);
-UInt32 Hc3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances);
+UInt32* Bt3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances);
+UInt32* Hc3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances);

 void Bt3Zip_MatchFinder_Skip(CMatchFinder *p, UInt32 num);
 void Hc3Zip_MatchFinder_Skip(CMatchFinder *p, UInt32 num);

+void LzFindPrepare(void);
+
 EXTERN_C_END

 #endif
@@ -1,5 +1,5 @@
 /* LzFindMt.h -- multithreaded Match finder for LZ algorithms
-2018-07-04 : Igor Pavlov : Public domain */
+2021-07-12 : Igor Pavlov : Public domain */

 #ifndef __LZ_FIND_MT_H
 #define __LZ_FIND_MT_H
@@ -9,31 +9,26 @@

 EXTERN_C_BEGIN

-#define kMtHashBlockSize (1 << 13)
-#define kMtHashNumBlocks (1 << 3)
-#define kMtHashNumBlocksMask (kMtHashNumBlocks - 1)
-
-#define kMtBtBlockSize (1 << 14)
-#define kMtBtNumBlocks (1 << 6)
-#define kMtBtNumBlocksMask (kMtBtNumBlocks - 1)
-
 typedef struct _CMtSync
 {
+  UInt32 numProcessedBlocks;
+  CThread thread;
+  UInt64 affinity;
+
  BoolInt wasCreated;
  BoolInt needStart;
+  BoolInt csWasInitialized;
+  BoolInt csWasEntered;
+
  BoolInt exit;
  BoolInt stopWriting;

-  CThread thread;
  CAutoResetEvent canStart;
-  CAutoResetEvent wasStarted;
  CAutoResetEvent wasStopped;
  CSemaphore freeSemaphore;
  CSemaphore filledSemaphore;
-  BoolInt csWasInitialized;
-  BoolInt csWasEntered;
  CCriticalSection cs;
-  UInt32 numProcessedBlocks;
+  // UInt32 numBlocks_Sent;
 } CMtSync;

 typedef UInt32 * (*Mf_Mix_Matches)(void *p, UInt32 matchMinPos, UInt32 *distances);
@@ -49,18 +44,23 @@ typedef struct _CMatchFinderMt
  /* LZ */
  const Byte *pointerToCurPos;
  UInt32 *btBuf;
-  UInt32 btBufPos;
-  UInt32 btBufPosLimit;
+  const UInt32 *btBufPos;
+  const UInt32 *btBufPosLimit;
  UInt32 lzPos;
  UInt32 btNumAvailBytes;

  UInt32 *hash;
  UInt32 fixedHashSize;
+  // UInt32 hash4Mask;
  UInt32 historySize;
  const UInt32 *crc;

  Mf_Mix_Matches MixMatchesFunc;
-  
+  UInt32 failure_LZ_BT; // failure in BT transfered to LZ
+  // UInt32 failure_LZ_LZ; // failure in LZ tables
+  UInt32 failureBuf[1];
+  // UInt32 crc[256];
+
  /* LZ + BT */
  CMtSync btSync;
  Byte btDummy[kMtCacheLineDummy];
@@ -70,6 +70,8 @@ typedef struct _CMatchFinderMt
  UInt32 hashBufPos;
  UInt32 hashBufPosLimit;
  UInt32 hashNumAvail;
+  UInt32 failure_BT;
+

  CLzRef *son;
  UInt32 matchMaxLen;
@@ -77,7 +79,7 @@ typedef struct _CMatchFinderMt
  UInt32 pos;
  const Byte *buffer;
  UInt32 cyclicBufferPos;
-  UInt32 cyclicBufferSize; /* it must be historySize + 1 */
+  UInt32 cyclicBufferSize; /* it must be = (historySize + 1) */
  UInt32 cutValue;

  /* BT + Hash */
@@ -87,13 +89,19 @@ typedef struct _CMatchFinderMt
  /* Hash */
  Mf_GetHeads GetHeadsFunc;
  CMatchFinder *MatchFinder;
+  // CMatchFinder MatchFinder;
 } CMatchFinderMt;

+// only for Mt part
 void MatchFinderMt_Construct(CMatchFinderMt *p);
 void MatchFinderMt_Destruct(CMatchFinderMt *p, ISzAllocPtr alloc);
+
 SRes MatchFinderMt_Create(CMatchFinderMt *p, UInt32 historySize, UInt32 keepAddBufferBefore,
    UInt32 matchMaxLen, UInt32 keepAddBufferAfter, ISzAllocPtr alloc);
-void MatchFinderMt_CreateVTable(CMatchFinderMt *p, IMatchFinder *vTable);
+void MatchFinderMt_CreateVTable(CMatchFinderMt *p, IMatchFinder2 *vTable);
+
+/* call MatchFinderMt_InitMt() before IMatchFinder::Init() */
+SRes MatchFinderMt_InitMt(CMatchFinderMt *p);
 void MatchFinderMt_ReleaseStream(CMatchFinderMt *p);

 EXTERN_C_END
@@ -0,0 +1,578 @@
+/* LzFindOpt.c -- multithreaded Match finder for LZ algorithms
+2021-07-13 : Igor Pavlov : Public domain */
+
+#include "Precomp.h"
+
+#include "CpuArch.h"
+#include "LzFind.h"
+
+// #include "LzFindMt.h"
+
+// #define LOG_ITERS
+
+// #define LOG_THREAD
+
+#ifdef LOG_THREAD
+#include <stdio.h>
+#define PRF(x) x
+#else
+// #define PRF(x)
+#endif
+
+#ifdef LOG_ITERS
+#include <stdio.h>
+UInt64 g_NumIters_Tree;
+UInt64 g_NumIters_Loop;
+UInt64 g_NumIters_Bytes;
+#define LOG_ITER(x) x
+#else
+#define LOG_ITER(x)
+#endif
+
+// ---------- BT THREAD ----------
+
+#define USE_SON_PREFETCH
+#define USE_LONG_MATCH_OPT
+
+#define kEmptyHashValue 0
+
+// #define CYC_TO_POS_OFFSET 0
+
+// #define CYC_TO_POS_OFFSET 1 // for debug
+
+/*
+MY_NO_INLINE
+UInt32 * MY_FAST_CALL GetMatchesSpecN_1(const Byte *lenLimit, size_t pos, const Byte *cur, CLzRef *son,
+    UInt32 _cutValue, UInt32 *d, size_t _maxLen, const UInt32 *hash, const UInt32 *limit, const UInt32 *size, UInt32 *posRes)
+{
+  do
+  {
+    UInt32 delta;
+    if (hash == size)
+      break;
+    delta = *hash++;
+
+    if (delta == 0 || delta > (UInt32)pos)
+      return NULL;
+
+    lenLimit++;
+
+    if (delta == (UInt32)pos)
+    {
+      CLzRef *ptr1 = son + ((size_t)pos << 1) - CYC_TO_POS_OFFSET * 2;
+      *d++ = 0;
+      ptr1[0] = kEmptyHashValue;
+      ptr1[1] = kEmptyHashValue;
+    }
+else
+{
+  UInt32 *_distances = ++d;
+
+  CLzRef *ptr0 = son + ((size_t)(pos) << 1) - CYC_TO_POS_OFFSET * 2 + 1;
+  CLzRef *ptr1 = son + ((size_t)(pos) << 1) - CYC_TO_POS_OFFSET * 2;
+
+  const Byte *len0 = cur, *len1 = cur;
+  UInt32 cutValue = _cutValue;
+  const Byte *maxLen = cur + _maxLen;
+
+  for (LOG_ITER(g_NumIters_Tree++);;)
+  {
+    LOG_ITER(g_NumIters_Loop++);
+    {
+      const ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)delta;
+      CLzRef *pair = son + ((size_t)(((ptrdiff_t)pos - CYC_TO_POS_OFFSET) + diff) << 1);
+      const Byte *len = (len0 < len1 ? len0 : len1);
+
+    #ifdef USE_SON_PREFETCH
+      const UInt32 pair0 = *pair;
+    #endif
+
+      if (len[diff] == len[0])
+      {
+        if (++len != lenLimit && len[diff] == len[0])
+          while (++len != lenLimit)
+          {
+            LOG_ITER(g_NumIters_Bytes++);
+            if (len[diff] != len[0])
+              break;
+          }
+        if (maxLen < len)
+        {
+          maxLen = len;
+          *d++ = (UInt32)(len - cur);
+          *d++ = delta - 1;
+          
+          if (len == lenLimit)
+          {
+            const UInt32 pair1 = pair[1];
+            *ptr1 =
+              #ifdef USE_SON_PREFETCH
+                pair0;
+              #else
+                pair[0];
+              #endif
+            *ptr0 = pair1;
+
+            _distances[-1] = (UInt32)(d - _distances);
+
+            #ifdef USE_LONG_MATCH_OPT
+
+                if (hash == size || *hash != delta || lenLimit[diff] != lenLimit[0] || d >= limit)
+                  break;
+
+            {
+              for (;;)
+              {
+                hash++;
+                pos++;
+                cur++;
+                lenLimit++;
+                {
+                  CLzRef *ptr = son + ((size_t)(pos) << 1) - CYC_TO_POS_OFFSET * 2;
+                  #if 0
+                  *(UInt64 *)(void *)ptr = ((const UInt64 *)(const void *)ptr)[diff];
+                  #else
+                  const UInt32 p0 = ptr[0 + (diff * 2)];
+                  const UInt32 p1 = ptr[1 + (diff * 2)];
+                  ptr[0] = p0;
+                  ptr[1] = p1;
+                  // ptr[0] = ptr[0 + (diff * 2)];
+                  // ptr[1] = ptr[1 + (diff * 2)];
+                  #endif
+                }
+                // PrintSon(son + 2, pos - 1);
+                // printf("\npos = %x delta = %x\n", pos, delta);
+                len++;
+                *d++ = 2;
+                *d++ = (UInt32)(len - cur);
+                *d++ = delta - 1;
+                if (hash == size || *hash != delta || lenLimit[diff] != lenLimit[0] || d >= limit)
+                  break;
+              }
+            }
+            #endif
+
+            break;
+          }
+        }
+      }
+
+      {
+        const UInt32 curMatch = (UInt32)pos - delta; // (UInt32)(pos + diff);
+        if (len[diff] < len[0])
+        {
+          delta = pair[1];
+          if (delta >= curMatch)
+            return NULL;
+          *ptr1 = curMatch;
+          ptr1 = pair + 1;
+          len1 = len;
+        }
+        else
+        {
+          delta = *pair;
+          if (delta >= curMatch)
+            return NULL;
+          *ptr0 = curMatch;
+          ptr0 = pair;
+          len0 = len;
+        }
+
+        delta = (UInt32)pos - delta;
+ 
+        if (--cutValue == 0 || delta >= pos)
+        {
+          *ptr0 = *ptr1 = kEmptyHashValue;
+          _distances[-1] = (UInt32)(d - _distances);
+          break;
+        }
+      }
+    }
+  } // for (tree iterations)
+}
+    pos++;
+    cur++;
+  }
+  while (d < limit);
+  *posRes = (UInt32)pos;
+  return d;
+}
+*/
+
+/* define cbs if you use 2 functions.
+       GetMatchesSpecN_1() :  (pos <  _cyclicBufferSize)
+       GetMatchesSpecN_2() :  (pos >= _cyclicBufferSize)
+
+  do not define cbs if you use 1 function:
+       GetMatchesSpecN_2()
+*/
+
+// #define cbs _cyclicBufferSize
+
+/*
+  we use size_t for (pos) and (_cyclicBufferPos_ instead of UInt32
+  to eliminate "movsx" BUG in old MSVC x64 compiler.
+*/
+
+UInt32 * MY_FAST_CALL GetMatchesSpecN_2(const Byte *lenLimit, size_t pos, const Byte *cur, CLzRef *son,
+    UInt32 _cutValue, UInt32 *d, size_t _maxLen, const UInt32 *hash, const UInt32 *limit, const UInt32 *size,
+    size_t _cyclicBufferPos, UInt32 _cyclicBufferSize,
+    UInt32 *posRes);
+
+MY_NO_INLINE
+UInt32 * MY_FAST_CALL GetMatchesSpecN_2(const Byte *lenLimit, size_t pos, const Byte *cur, CLzRef *son,
+    UInt32 _cutValue, UInt32 *d, size_t _maxLen, const UInt32 *hash, const UInt32 *limit, const UInt32 *size,
+    size_t _cyclicBufferPos, UInt32 _cyclicBufferSize,
+    UInt32 *posRes)
+{
+  do // while (hash != size)
+  {
+    UInt32 delta;
+    
+  #ifndef cbs
+    UInt32 cbs;
+  #endif
+
+    if (hash == size)
+      break;
+
+    delta = *hash++;
+
+    if (delta == 0)
+      return NULL;
+
+    lenLimit++;
+
+  #ifndef cbs
+    cbs = _cyclicBufferSize;
+    if ((UInt32)pos < cbs)
+    {
+      if (delta > (UInt32)pos)
+        return NULL;
+      cbs = (UInt32)pos;
+    }
+  #endif
+
+    if (delta >= cbs)
+    {
+      CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1);
+      *d++ = 0;
+      ptr1[0] = kEmptyHashValue;
+      ptr1[1] = kEmptyHashValue;
+    }
+else
+{
+  UInt32 *_distances = ++d;
+
+  CLzRef *ptr0 = son + ((size_t)_cyclicBufferPos << 1) + 1;
+  CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1);
+
+  UInt32 cutValue = _cutValue;
+  const Byte *len0 = cur, *len1 = cur;
+  const Byte *maxLen = cur + _maxLen;
+
+  // if (cutValue == 0) { *ptr0 = *ptr1 = kEmptyHashValue; } else
+  for (LOG_ITER(g_NumIters_Tree++);;)
+  {
+    LOG_ITER(g_NumIters_Loop++);
+    {
+      // SPEC code
+      CLzRef *pair = son + ((size_t)((ptrdiff_t)_cyclicBufferPos - (ptrdiff_t)delta
+          + (ptrdiff_t)(UInt32)(_cyclicBufferPos < delta ? cbs : 0)
+          ) << 1);
+
+      const ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)delta;
+      const Byte *len = (len0 < len1 ? len0 : len1);
+
+    #ifdef USE_SON_PREFETCH
+      const UInt32 pair0 = *pair;
+    #endif
+
+      if (len[diff] == len[0])
+      {
+        if (++len != lenLimit && len[diff] == len[0])
+          while (++len != lenLimit)
+          {
+            LOG_ITER(g_NumIters_Bytes++);
+            if (len[diff] != len[0])
+              break;
+          }
+        if (maxLen < len)
+        {
+          maxLen = len;
+          *d++ = (UInt32)(len - cur);
+          *d++ = delta - 1;
+          
+          if (len == lenLimit)
+          {
+            const UInt32 pair1 = pair[1];
+            *ptr1 =
+              #ifdef USE_SON_PREFETCH
+                pair0;
+              #else
+                pair[0];
+              #endif
+            *ptr0 = pair1;
+
+            _distances[-1] = (UInt32)(d - _distances);
+
+            #ifdef USE_LONG_MATCH_OPT
+
+                if (hash == size || *hash != delta || lenLimit[diff] != lenLimit[0] || d >= limit)
+                  break;
+
+            {
+              for (;;)
+              {
+                *d++ = 2;
+                *d++ = (UInt32)(lenLimit - cur);
+                *d++ = delta - 1;
+                cur++;
+                lenLimit++;
+                // SPEC
+                _cyclicBufferPos++;
+                {
+                  // SPEC code
+                  CLzRef *dest = son + ((size_t)(_cyclicBufferPos) << 1);
+                  const CLzRef *src = dest + ((diff
+                      + (ptrdiff_t)(UInt32)((_cyclicBufferPos < delta) ? cbs : 0)) << 1);
+                  // CLzRef *ptr = son + ((size_t)(pos) << 1) - CYC_TO_POS_OFFSET * 2;
+                  #if 0
+                  *(UInt64 *)(void *)dest = *((const UInt64 *)(const void *)src);
+                  #else
+                  const UInt32 p0 = src[0];
+                  const UInt32 p1 = src[1];
+                  dest[0] = p0;
+                  dest[1] = p1;
+                  #endif
+                }
+                pos++;
+                hash++;
+                if (hash == size || *hash != delta || lenLimit[diff] != lenLimit[0] || d >= limit)
+                  break;
+              } // for() end for long matches
+            }
+            #endif
+
+            break; // break from TREE iterations
+          }
+        }
+      }
+      {
+        const UInt32 curMatch = (UInt32)pos - delta; // (UInt32)(pos + diff);
+        if (len[diff] < len[0])
+        {
+          delta = pair[1];
+          *ptr1 = curMatch;
+          ptr1 = pair + 1;
+          len1 = len;
+          if (delta >= curMatch)
+            return NULL;
+        }
+        else
+        {
+          delta = *pair;
+          *ptr0 = curMatch;
+          ptr0 = pair;
+          len0 = len;
+          if (delta >= curMatch)
+            return NULL;
+        }
+        delta = (UInt32)pos - delta;
+ 
+        if (--cutValue == 0 || delta >= cbs)
+        {
+          *ptr0 = *ptr1 = kEmptyHashValue;
+          _distances[-1] = (UInt32)(d - _distances);
+          break;
+        }
+      }
+    }
+  } // for (tree iterations)
+}
+    pos++;
+    _cyclicBufferPos++;
+    cur++;
+  }
+  while (d < limit);
+  *posRes = (UInt32)pos;
+  return d;
+}
+
+
+
+/*
+typedef UInt32 uint32plus; // size_t
+
+UInt32 * MY_FAST_CALL GetMatchesSpecN_3(uint32plus lenLimit, size_t pos, const Byte *cur, CLzRef *son,
+    UInt32 _cutValue, UInt32 *d, uint32plus _maxLen, const UInt32 *hash, const UInt32 *limit, const UInt32 *size,
+    size_t _cyclicBufferPos, UInt32 _cyclicBufferSize,
+    UInt32 *posRes)
+{
+  do // while (hash != size)
+  {
+    UInt32 delta;
+
+  #ifndef cbs
+    UInt32 cbs;
+  #endif
+
+    if (hash == size)
+      break;
+
+    delta = *hash++;
+
+    if (delta == 0)
+      return NULL;
+
+  #ifndef cbs
+    cbs = _cyclicBufferSize;
+    if ((UInt32)pos < cbs)
+    {
+      if (delta > (UInt32)pos)
+        return NULL;
+      cbs = (UInt32)pos;
+    }
+  #endif
+    
+    if (delta >= cbs)
+    {
+      CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1);
+      *d++ = 0;
+      ptr1[0] = kEmptyHashValue;
+      ptr1[1] = kEmptyHashValue;
+    }
+else
+{
+  CLzRef *ptr0 = son + ((size_t)_cyclicBufferPos << 1) + 1;
+  CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1);
+  UInt32 *_distances = ++d;
+  uint32plus len0 = 0, len1 = 0;
+  UInt32 cutValue = _cutValue;
+  uint32plus maxLen = _maxLen;
+  // lenLimit++; // const Byte *lenLimit = cur + _lenLimit;
+
+  for (LOG_ITER(g_NumIters_Tree++);;)
+  {
+    LOG_ITER(g_NumIters_Loop++);
+    {
+      // const ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)delta;
+      CLzRef *pair = son + ((size_t)((ptrdiff_t)_cyclicBufferPos - delta
+          + (ptrdiff_t)(UInt32)(_cyclicBufferPos < delta ? cbs : 0)
+          ) << 1);
+      const Byte *pb = cur - delta;
+      uint32plus len = (len0 < len1 ? len0 : len1);
+
+    #ifdef USE_SON_PREFETCH
+      const UInt32 pair0 = *pair;
+    #endif
+
+      if (pb[len] == cur[len])
+      {
+        if (++len != lenLimit && pb[len] == cur[len])
+          while (++len != lenLimit)
+            if (pb[len] != cur[len])
+              break;
+        if (maxLen < len)
+        {
+          maxLen = len;
+          *d++ = (UInt32)len;
+          *d++ = delta - 1;
+          if (len == lenLimit)
+          {
+            {
+              const UInt32 pair1 = pair[1];
+              *ptr0 = pair1;
+              *ptr1 =
+              #ifdef USE_SON_PREFETCH
+                pair0;
+              #else
+                pair[0];
+              #endif
+            }
+
+            _distances[-1] = (UInt32)(d - _distances);
+
+            #ifdef USE_LONG_MATCH_OPT
+
+                if (hash == size || *hash != delta || pb[lenLimit] != cur[lenLimit] || d >= limit)
+                  break;
+
+            {
+              const ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)delta;
+              for (;;)
+              {
+                *d++ = 2;
+                *d++ = (UInt32)lenLimit;
+                *d++ = delta - 1;
+                _cyclicBufferPos++;
+                {
+                  CLzRef *dest = son + ((size_t)_cyclicBufferPos << 1);
+                  const CLzRef *src = dest + ((diff +
+                      (ptrdiff_t)(UInt32)(_cyclicBufferPos < delta ? cbs : 0)) << 1);
+                #if 0
+                  *(UInt64 *)(void *)dest = *((const UInt64 *)(const void *)src);
+                #else
+                  const UInt32 p0 = src[0];
+                  const UInt32 p1 = src[1];
+                  dest[0] = p0;
+                  dest[1] = p1;
+                #endif
+                }
+                hash++;
+                pos++;
+                cur++;
+                pb++;
+                if (hash == size || *hash != delta || pb[lenLimit] != cur[lenLimit] || d >= limit)
+                  break;
+              }
+            }
+            #endif
+
+            break;
+          }
+        }
+      }
+      {
+        const UInt32 curMatch = (UInt32)pos - delta;
+        if (pb[len] < cur[len])
+        {
+          delta = pair[1];
+          *ptr1 = curMatch;
+          ptr1 = pair + 1;
+          len1 = len;
+        }
+        else
+        {
+          delta = *pair;
+          *ptr0 = curMatch;
+          ptr0 = pair;
+          len0 = len;
+        }
+
+        {
+          if (delta >= curMatch)
+            return NULL;
+          delta = (UInt32)pos - delta;
+          if (delta >= cbs
+              // delta >= _cyclicBufferSize || delta >= pos
+              || --cutValue == 0)
+          {
+            *ptr0 = *ptr1 = kEmptyHashValue;
+            _distances[-1] = (UInt32)(d - _distances);
+            break;
+          }
+        }
+      }
+    }
+  } // for (tree iterations)
+}
+    pos++;
+    _cyclicBufferPos++;
+    cur++;
+  }
+  while (d < limit);
+  *posRes = (UInt32)pos;
+  return d;
+}
+*/
@@ -0,0 +1,34 @@
+/* LzHash.h -- HASH functions for LZ algorithms
+2019-10-30 : Igor Pavlov : Public domain */
+
+#ifndef __LZ_HASH_H
+#define __LZ_HASH_H
+
+/*
+  (kHash2Size >= (1 <<  8)) : Required
+  (kHash3Size >= (1 << 16)) : Required
+*/
+
+#define kHash2Size (1 << 10)
+#define kHash3Size (1 << 16)
+// #define kHash4Size (1 << 20)
+
+#define kFix3HashSize (kHash2Size)
+#define kFix4HashSize (kHash2Size + kHash3Size)
+// #define kFix5HashSize (kHash2Size + kHash3Size + kHash4Size)
+
+/*
+  We use up to 3 crc values for hash:
+    crc0
+    crc1 << Shift_1
+    crc2 << Shift_2
+  (Shift_1 = 5) and (Shift_2 = 10) is good tradeoff.
+  Small values for Shift are not good for collision rate.
+  Big value for Shift_2 increases the minimum size
+  of hash table, that will be slow for small files.
+*/
+
+#define kLzHash_CrcShift_1 5
+#define kLzHash_CrcShift_2 10
+
+#endif
@@ -1,5 +1,5 @@
 /* Lzma2Dec.c -- LZMA2 Decoder
-2019-02-02 : Igor Pavlov : Public domain */
+2021-02-09 : Igor Pavlov : Public domain */

 /* #define SHOW_DEBUG_INFO */

@@ -93,7 +93,8 @@ void Lzma2Dec_Init(CLzma2Dec *p)
  LzmaDec_Init(&p->decoder);
 }

-static ELzma2State Lzma2Dec_UpdateState(CLzma2Dec *p, Byte b)
+// ELzma2State
+static unsigned Lzma2Dec_UpdateState(CLzma2Dec *p, Byte b)
 {
  switch (p->state)
  {
@@ -1,25 +1,25 @@
 /* Lzma2DecMt.c -- LZMA2 Decoder Multi-thread
-2019-02-02 : Igor Pavlov : Public domain */
+2021-04-01 : Igor Pavlov : Public domain */

 #include "Precomp.h"

 // #define SHOW_DEBUG_INFO

+// #define _7ZIP_ST
+
 #ifdef SHOW_DEBUG_INFO
 #include <stdio.h>
 #endif

+#ifndef _7ZIP_ST
 #ifdef SHOW_DEBUG_INFO
 #define PRF(x) x
 #else
 #define PRF(x)
 #endif
-
 #define PRF_STR(s) PRF(printf("\n" s "\n"))
-#define PRF_STR_INT(s, d) PRF(printf("\n" s " %d\n", (unsigned)d))
 #define PRF_STR_INT_2(s, d1, d2) PRF(printf("\n" s " %d %d\n", (unsigned)d1, (unsigned)d2))
-
-// #define _7ZIP_ST
+#endif

 #include "Alloc.h"

@@ -28,10 +28,10 @@

 #ifndef _7ZIP_ST
 #include "MtDec.h"
-#endif
-

 #define LZMA2DECMT_OUT_BLOCK_MAX_DEFAULT (1 << 28)
+#endif
+

 void Lzma2DecMtProps_Init(CLzma2DecMtProps *p)
 {
@@ -255,7 +255,7 @@ static void Lzma2DecMt_MtCallback_Parse(void *obj, unsigned coderIndex, CMtDecCa
        const unsigned kNumAlignBits = 12;
        const unsigned kNumCacheLineBits = 7; /* <= kNumAlignBits */
        t->alloc.numAlignBits = kNumAlignBits;
-        t->alloc.offset = ((UInt32)coderIndex * ((1 << 11) + (1 << 8) + (1 << 6))) & ((1 << kNumAlignBits) - (1 << kNumCacheLineBits));
+        t->alloc.offset = ((UInt32)coderIndex * (((unsigned)1 << 11) + (1 << 8) + (1 << 6))) & (((unsigned)1 << kNumAlignBits) - ((unsigned)1 << kNumCacheLineBits));
        t->alloc.baseAlloc = me->alignOffsetAlloc.baseAlloc;
      }
    }
@@ -527,7 +527,7 @@ static SRes Lzma2DecMt_MtCallback_Code(void *pp, unsigned coderIndex,

 static SRes Lzma2DecMt_MtCallback_Write(void *pp, unsigned coderIndex,
    BoolInt needWriteToStream,
-    const Byte *src, size_t srcSize,
+    const Byte *src, size_t srcSize, BoolInt isCross,
    BoolInt *needContinue, BoolInt *canRecode)
 {
  CLzma2DecMt *me = (CLzma2DecMt *)pp;
@@ -536,12 +536,14 @@ static SRes Lzma2DecMt_MtCallback_Write(void *pp, unsigned coderIndex,
  const Byte *data = t->outBuf;
  BoolInt needContinue2 = True;

+  UNUSED_VAR(src)
+  UNUSED_VAR(srcSize)
+  UNUSED_VAR(isCross)
+
  PRF_STR_INT_2("Write", coderIndex, srcSize);

  *needContinue = False;
  *canRecode = True;
-  UNUSED_VAR(src)
-  UNUSED_VAR(srcSize)

  if (
      // t->parseStatus == LZMA_STATUS_FINISHED_WITH_MARK
@@ -696,7 +698,7 @@ static SRes Lzma2Dec_Decode_ST(CLzma2DecMt *p
        inPos = 0;
        inLim = p->inBufSize;
        inData = p->inBuf;
-        p->readRes = ISeqInStream_Read(p->inStream, (void *)inData, &inLim);
+        p->readRes = ISeqInStream_Read(p->inStream, (void *)(p->inBuf), &inLim);
        // p->readProcessed += inLim;
        // inLim -= 5; p->readWasFinished = True; // for test
        if (inLim == 0 || p->readRes != SZ_OK)
@@ -838,6 +840,7 @@ SRes Lzma2DecMt_Decode(CLzma2DecMtHandle pp,
  p->inProcessed = 0;

  p->readWasFinished = False;
+  p->readRes = SZ_OK;

  *isMT = False;

@@ -856,7 +859,7 @@ SRes Lzma2DecMt_Decode(CLzma2DecMtHandle pp,

  if (p->props.numThreads > 1)
  {
-    IMtDecCallback vt;
+    IMtDecCallback2 vt;

    Lzma2DecMt_FreeSt(p);

@@ -955,7 +958,12 @@ SRes Lzma2DecMt_Decode(CLzma2DecMtHandle pp,
    *inProcessed = p->inProcessed;

    // res = SZ_OK; // for test
-    if (res == SZ_OK && p->readRes != SZ_OK)
+    if (res == SZ_ERROR_INPUT_EOF)
+    {
+      if (p->readRes != SZ_OK)
+        res = p->readRes;
+    }
+    else if (res == SZ_OK && p->readRes != SZ_OK)
      res = p->readRes;
    
    /*
@@ -1,5 +1,5 @@
 /* Lzma2Enc.c -- LZMA2 Encoder
-2018-07-04 : Igor Pavlov : Public domain */
+2021-02-09 : Igor Pavlov : Public domain */

 #include "Precomp.h"

@@ -330,7 +330,7 @@ void Lzma2EncProps_Normalize(CLzma2EncProps *p)
        numBlocks++;
      if (numBlocks < (unsigned)t2)
      {
-        t2r = (unsigned)numBlocks;
+        t2r = (int)numBlocks;
        if (t2r == 0)
          t2r = 1;
        t3 = t1 * t2r;
@@ -632,15 +632,15 @@ static SRes Lzma2Enc_EncodeMt1(
      {
        if (outBuf)
        {
-          size_t destPos = *outBufSize;
+          const size_t destPos = *outBufSize;
          if (destPos >= outLim)
            return SZ_ERROR_OUTPUT_EOF;
-          outBuf[destPos] = 0;
+          outBuf[destPos] = LZMA2_CONTROL_EOF; // 0
          *outBufSize = destPos + 1;
        }
        else
        {
-          Byte b = 0;
+          const Byte b = LZMA2_CONTROL_EOF; // 0;
          if (ISeqOutStream_Write(outStream, &b, 1) != 1)
            return SZ_ERROR_WRITE;
        }
@@ -780,13 +780,13 @@ SRes Lzma2Enc_Encode2(CLzma2EncHandle pp,
      p->outBufSize = destBlockSize;
    }

-    p->mtCoder.numThreadsMax = p->props.numBlockThreads_Max;
+    p->mtCoder.numThreadsMax = (unsigned)p->props.numBlockThreads_Max;
    p->mtCoder.expectedDataSize = p->expectedDataSize;
    
    {
      SRes res = MtCoder_Code(&p->mtCoder);
      if (!outStream)
-        *outBufSize = p->outBuf - outBuf;
+        *outBufSize = (size_t)(p->outBuf - outBuf);
      return res;
    }
  }
@@ -11,8 +11,6 @@
 #include "Bra.h"
 #include "LzmaEnc.h"

-#define SZE_OUT_OVERFLOW SZE_DATA_ERROR
-
 int Lzma86_Encode(Byte *dest, size_t *destLen, const Byte *src, size_t srcLen,
    int level, UInt32 dictSize, int filterMode)
 {
@@ -1,5 +1,5 @@
 /* LzmaDec.c -- LZMA Decoder
-2018-07-04 : Igor Pavlov : Public domain */
+2021-04-01 : Igor Pavlov : Public domain */

 #include "Precomp.h"

@@ -13,10 +13,12 @@

 #define kNumBitModelTotalBits 11
 #define kBitModelTotal (1 << kNumBitModelTotalBits)
-#define kNumMoveBits 5

 #define RC_INIT_SIZE 5

+#ifndef _LZMA_DEC_OPT
+
+#define kNumMoveBits 5
 #define NORMALIZE if (range < kTopValue) { range <<= 8; code = (code << 8) | (*buf++); }

 #define IF_BIT_0(p) ttt = *(p); NORMALIZE; bound = (range >> kNumBitModelTotalBits) * (UInt32)ttt; if (code < bound)
@@ -62,9 +64,10 @@
  probLit = prob + (offs + bit + symbol); \
  GET_BIT2(probLit, symbol, offs ^= bit; , ;)

+#endif // _LZMA_DEC_OPT


-#define NORMALIZE_CHECK if (range < kTopValue) { if (buf >= bufLimit) return DUMMY_ERROR; range <<= 8; code = (code << 8) | (*buf++); }
+#define NORMALIZE_CHECK if (range < kTopValue) { if (buf >= bufLimit) return DUMMY_INPUT_EOF; range <<= 8; code = (code << 8) | (*buf++); }

 #define IF_BIT_0_CHECK(p) ttt = *(p); NORMALIZE_CHECK; bound = (range >> kNumBitModelTotalBits) * (UInt32)ttt; if (code < bound)
 #define UPDATE_0_CHECK range = bound;
@@ -114,6 +117,9 @@
 #define kMatchMinLen 2
 #define kMatchSpecLenStart (kMatchMinLen + kLenNumLowSymbols * 2 + kLenNumHighSymbols)

+#define kMatchSpecLen_Error_Data (1 << 9)
+#define kMatchSpecLen_Error_Fail (kMatchSpecLen_Error_Data - 1)
+
 /* External ASM code needs same CLzmaProb array layout. So don't change it. */

 /* (probs_1664) is faster and better for code size at some platforms */
@@ -166,10 +172,12 @@

 /*
 p->remainLen : shows status of LZMA decoder:
-    < kMatchSpecLenStart : normal remain
-    = kMatchSpecLenStart : finished
-    = kMatchSpecLenStart + 1 : need init range coder
-    = kMatchSpecLenStart + 2 : need init range coder and state
+    < kMatchSpecLenStart  : the number of bytes to be copied with (p->rep0) offset
+    = kMatchSpecLenStart  : the LZMA stream was finished with end mark
+    = kMatchSpecLenStart + 1  : need init range coder
+    = kMatchSpecLenStart + 2  : need init range coder and state
+    = kMatchSpecLen_Error_Fail                : Internal Code Failure
+    = kMatchSpecLen_Error_Data + [0 ... 273]  : LZMA Data Error
 */

 /* ---------- LZMA_DECODE_REAL ---------- */
@@ -188,23 +196,31 @@ In:
  {
    LzmaDec_TryDummy() was called before to exclude LITERAL and MATCH-REP cases.
    So first symbol can be only MATCH-NON-REP. And if that MATCH-NON-REP symbol
-    is not END_OF_PAYALOAD_MARKER, then function returns error code.
+    is not END_OF_PAYALOAD_MARKER, then the function doesn't write any byte to dictionary,
+    the function returns SZ_OK, and the caller can use (p->remainLen) and (p->reps[0]) later.
  }

 Processing:
-  first LZMA symbol will be decoded in any case
-  All checks for limits are at the end of main loop,
-  It will decode new LZMA-symbols while (p->buf < bufLimit && dicPos < limit),
+  The first LZMA symbol will be decoded in any case.
+  All main checks for limits are at the end of main loop,
+  It decodes additional LZMA-symbols while (p->buf < bufLimit && dicPos < limit),
  RangeCoder is still without last normalization when (p->buf < bufLimit) is being checked.
+  But if (p->buf < bufLimit), the caller provided at least (LZMA_REQUIRED_INPUT_MAX + 1) bytes for
+  next iteration  before limit (bufLimit + LZMA_REQUIRED_INPUT_MAX),
+  that is enough for worst case LZMA symbol with one additional RangeCoder normalization for one bit.
+  So that function never reads bufLimit [LZMA_REQUIRED_INPUT_MAX] byte.

 Out:
  RangeCoder is normalized
  Result:
    SZ_OK - OK
-    SZ_ERROR_DATA - Error
-  p->remainLen:
-    < kMatchSpecLenStart : normal remain
-    = kMatchSpecLenStart : finished
+      p->remainLen:
+        < kMatchSpecLenStart : the number of bytes to be copied with (p->reps[0]) offset
+        = kMatchSpecLenStart : the LZMA stream was finished with end mark
+
+    SZ_ERROR_DATA - error, when the MATCH-Symbol refers out of dictionary
+      p->remainLen : undefined
+      p->reps[*]    : undefined
 */


@@ -316,11 +332,6 @@ int MY_FAST_CALL LZMA_DECODE_REAL(CLzmaDec *p, SizeT limit, const Byte *bufLimit
      else
      {
        UPDATE_1(prob);
-        /*
-        // that case was checked before with kBadRepCode
-        if (checkDicSize == 0 && processedPos == 0)
-          return SZ_ERROR_DATA;
-        */
        prob = probs + IsRepG0 + state;
        IF_BIT_0(prob)
        {
@@ -329,6 +340,13 @@ int MY_FAST_CALL LZMA_DECODE_REAL(CLzmaDec *p, SizeT limit, const Byte *bufLimit
          IF_BIT_0(prob)
          {
            UPDATE_0(prob);
+  
+            // that case was checked before with kBadRepCode
+            // if (checkDicSize == 0 && processedPos == 0) { len = kMatchSpecLen_Error_Data + 1; break; }
+            // The caller doesn't allow (dicPos == limit) case here
+            // so we don't need the following check:
+            // if (dicPos == limit) { state = state < kNumLitStates ? 9 : 11; len = 1; break; }
+            
            dic[dicPos] = dic[dicPos - rep0 + (dicPos < rep0 ? dicBufSize : 0)];
            dicPos++;
            processedPos++;
@@ -518,8 +536,10 @@ int MY_FAST_CALL LZMA_DECODE_REAL(CLzmaDec *p, SizeT limit, const Byte *bufLimit
        state = (state < kNumStates + kNumLitStates) ? kNumLitStates : kNumLitStates + 3;
        if (distance >= (checkDicSize == 0 ? processedPos: checkDicSize))
        {
-          p->dicPos = dicPos;
-          return SZ_ERROR_DATA;
+          len += kMatchSpecLen_Error_Data + kMatchMinLen;
+          // len = kMatchSpecLen_Error_Data;
+          // len += kMatchMinLen;
+          break;
        }
      }

@@ -532,8 +552,13 @@ int MY_FAST_CALL LZMA_DECODE_REAL(CLzmaDec *p, SizeT limit, const Byte *bufLimit
        
        if ((rem = limit - dicPos) == 0)
        {
-          p->dicPos = dicPos;
-          return SZ_ERROR_DATA;
+          /*
+          We stop decoding and return SZ_OK, and we can resume decoding later.
+          Any error conditions can be tested later in caller code.
+          For more strict mode we can stop decoding with error
+          // len += kMatchSpecLen_Error_Data;
+          */
+          break;
        }
        
        curLen = ((rem < len) ? (unsigned)rem : len);
@@ -572,7 +597,7 @@ int MY_FAST_CALL LZMA_DECODE_REAL(CLzmaDec *p, SizeT limit, const Byte *bufLimit
  p->buf = buf;
  p->range = range;
  p->code = code;
-  p->remainLen = (UInt32)len;
+  p->remainLen = (UInt32)len; // & (kMatchSpecLen_Error_Data - 1); // we can write real length for error matches too.
  p->dicPos = dicPos;
  p->processedPos = processedPos;
  p->reps[0] = rep0;
@@ -580,40 +605,61 @@ int MY_FAST_CALL LZMA_DECODE_REAL(CLzmaDec *p, SizeT limit, const Byte *bufLimit
  p->reps[2] = rep2;
  p->reps[3] = rep3;
  p->state = (UInt32)state;
-
+  if (len >= kMatchSpecLen_Error_Data)
+    return SZ_ERROR_DATA;
  return SZ_OK;
 }
 #endif

+
+
 static void MY_FAST_CALL LzmaDec_WriteRem(CLzmaDec *p, SizeT limit)
 {
-  if (p->remainLen != 0 && p->remainLen < kMatchSpecLenStart)
+  unsigned len = (unsigned)p->remainLen;
+  if (len == 0 /* || len >= kMatchSpecLenStart */)
+    return;
  {
-    Byte *dic = p->dic;
    SizeT dicPos = p->dicPos;
-    SizeT dicBufSize = p->dicBufSize;
-    unsigned len = (unsigned)p->remainLen;
-    SizeT rep0 = p->reps[0]; /* we use SizeT to avoid the BUG of VC14 for AMD64 */
-    SizeT rem = limit - dicPos;
-    if (rem < len)
-      len = (unsigned)(rem);
+    Byte *dic;
+    SizeT dicBufSize;
+    SizeT rep0;   /* we use SizeT to avoid the BUG of VC14 for AMD64 */
+    {
+      SizeT rem = limit - dicPos;
+      if (rem < len)
+      {
+        len = (unsigned)(rem);
+        if (len == 0)
+          return;
+      }
+    }

    if (p->checkDicSize == 0 && p->prop.dicSize - p->processedPos <= len)
      p->checkDicSize = p->prop.dicSize;

    p->processedPos += (UInt32)len;
    p->remainLen -= (UInt32)len;
-    while (len != 0)
+    dic = p->dic;
+    rep0 = p->reps[0];
+    dicBufSize = p->dicBufSize;
+    do
    {
-      len--;
      dic[dicPos] = dic[dicPos - rep0 + (dicPos < rep0 ? dicBufSize : 0)];
      dicPos++;
    }
+    while (--len);
    p->dicPos = dicPos;
  }
 }


+/*
+At staring of new stream we have one of the following symbols:
+  - Literal        - is allowed
+  - Non-Rep-Match  - is allowed only if it's end marker symbol
+  - Rep-Match      - is not allowed
+We use early check of (RangeCoder:Code) over kBadRepCode to simplify main decoding code
+*/
+
 #define kRange0 0xFFFFFFFF
 #define kBound0 ((kRange0 >> kNumBitModelTotalBits) << (kNumBitModelTotalBits - 1))
 #define kBadRepCode (kBound0 + (((kRange0 - kBound0) >> kNumBitModelTotalBits) << (kNumBitModelTotalBits - 1)))
@@ -621,69 +667,77 @@ static void MY_FAST_CALL LzmaDec_WriteRem(CLzmaDec *p, SizeT limit)
  #error Stop_Compiling_Bad_LZMA_Check
 #endif

+
+/*
+LzmaDec_DecodeReal2():
+  It calls LZMA_DECODE_REAL() and it adjusts limit according (p->checkDicSize).
+
+We correct (p->checkDicSize) after LZMA_DECODE_REAL() and in LzmaDec_WriteRem(),
+and we support the following state of (p->checkDicSize):
+  if (total_processed < p->prop.dicSize) then
+  {
+    (total_processed == p->processedPos)
+    (p->checkDicSize == 0)
+  }
+  else
+    (p->checkDicSize == p->prop.dicSize)
+*/
+
 static int MY_FAST_CALL LzmaDec_DecodeReal2(CLzmaDec *p, SizeT limit, const Byte *bufLimit)
 {
-  do
+  if (p->checkDicSize == 0)
  {
-    SizeT limit2 = limit;
-    if (p->checkDicSize == 0)
-    {
-      UInt32 rem = p->prop.dicSize - p->processedPos;
-      if (limit - p->dicPos > rem)
-        limit2 = p->dicPos + rem;
-
-      if (p->processedPos == 0)
-        if (p->code >= kBadRepCode)
-          return SZ_ERROR_DATA;
-    }
-
-    RINOK(LZMA_DECODE_REAL(p, limit2, bufLimit));
-    
+    UInt32 rem = p->prop.dicSize - p->processedPos;
+    if (limit - p->dicPos > rem)
+      limit = p->dicPos + rem;
+  }
+  {
+    int res = LZMA_DECODE_REAL(p, limit, bufLimit);
    if (p->checkDicSize == 0 && p->processedPos >= p->prop.dicSize)
      p->checkDicSize = p->prop.dicSize;
-    
-    LzmaDec_WriteRem(p, limit);
+    return res;
  }
-  while (p->dicPos < limit && p->buf < bufLimit && p->remainLen < kMatchSpecLenStart);
-
-  return 0;
 }

+
+
 typedef enum
 {
-  DUMMY_ERROR, /* unexpected end of input stream */
+  DUMMY_INPUT_EOF, /* need more input data */
  DUMMY_LIT,
  DUMMY_MATCH,
  DUMMY_REP
 } ELzmaDummy;

-static ELzmaDummy LzmaDec_TryDummy(const CLzmaDec *p, const Byte *buf, SizeT inSize)
+
+#define IS_DUMMY_END_MARKER_POSSIBLE(dummyRes) ((dummyRes) == DUMMY_MATCH)
+
+static ELzmaDummy LzmaDec_TryDummy(const CLzmaDec *p, const Byte *buf, const Byte **bufOut)
 {
  UInt32 range = p->range;
  UInt32 code = p->code;
-  const Byte *bufLimit = buf + inSize;
+  const Byte *bufLimit = *bufOut;
  const CLzmaProb *probs = GET_PROBS;
  unsigned state = (unsigned)p->state;
  ELzmaDummy res;

+  for (;;)
  {
    const CLzmaProb *prob;
    UInt32 bound;
    unsigned ttt;
-    unsigned posState = CALC_POS_STATE(p->processedPos, (1 << p->prop.pb) - 1);
+    unsigned posState = CALC_POS_STATE(p->processedPos, ((unsigned)1 << p->prop.pb) - 1);

    prob = probs + IsMatch + COMBINED_PS_STATE;
    IF_BIT_0_CHECK(prob)
    {
      UPDATE_0_CHECK

-      /* if (bufLimit - buf >= 7) return DUMMY_LIT; */
-
      prob = probs + Literal;
      if (p->checkDicSize != 0 || p->processedPos != 0)
        prob += ((UInt32)LZMA_LIT_SIZE *
-            ((((p->processedPos) & ((1 << (p->prop.lp)) - 1)) << p->prop.lc) +
-            (p->dic[(p->dicPos == 0 ? p->dicBufSize : p->dicPos) - 1] >> (8 - p->prop.lc))));
+            ((((p->processedPos) & (((unsigned)1 << (p->prop.lp)) - 1)) << p->prop.lc) +
+            ((unsigned)p->dic[(p->dicPos == 0 ? p->dicBufSize : p->dicPos) - 1] >> (8 - p->prop.lc))));

      if (state < kNumLitStates)
      {
@@ -735,8 +789,7 @@ static ELzmaDummy LzmaDec_TryDummy(const CLzmaDec *p, const Byte *buf, SizeT inS
          IF_BIT_0_CHECK(prob)
          {
            UPDATE_0_CHECK;
-            NORMALIZE_CHECK;
-            return DUMMY_REP;
+            break;
          }
          else
          {
@@ -812,8 +865,6 @@ static ELzmaDummy LzmaDec_TryDummy(const CLzmaDec *p, const Byte *buf, SizeT inS
        {
          unsigned numDirectBits = ((posSlot >> 1) - 1);

-          /* if (bufLimit - buf >= 8) return DUMMY_MATCH; */
-
          if (posSlot < kEndPosModelIndex)
          {
            prob = probs + SpecPos + ((2 | (posSlot & 1)) << numDirectBits);
@@ -844,12 +895,15 @@ static ELzmaDummy LzmaDec_TryDummy(const CLzmaDec *p, const Byte *buf, SizeT inS
        }
      }
    }
+    break;
  }
  NORMALIZE_CHECK;
+
+  *bufOut = buf;
  return res;
 }

-
+void LzmaDec_InitDicAndState(CLzmaDec *p, BoolInt initDic, BoolInt initState);
 void LzmaDec_InitDicAndState(CLzmaDec *p, BoolInt initDic, BoolInt initState)
 {
  p->remainLen = kMatchSpecLenStart + 1;
@@ -872,16 +926,41 @@ void LzmaDec_Init(CLzmaDec *p)
 }


+/*
+LZMA supports optional end_marker.
+So the decoder can lookahead for one additional LZMA-Symbol to check end_marker.
+That additional LZMA-Symbol can require up to LZMA_REQUIRED_INPUT_MAX bytes in input stream.
+When the decoder reaches dicLimit, it looks (finishMode) parameter:
+  if (finishMode == LZMA_FINISH_ANY), the decoder doesn't lookahead
+  if (finishMode != LZMA_FINISH_ANY), the decoder lookahead, if end_marker is possible for current position
+
+When the decoder lookahead, and the lookahead symbol is not end_marker, we have two ways:
+  1) Strict mode (default) : the decoder returns SZ_ERROR_DATA.
+  2) The relaxed mode (alternative mode) : we could return SZ_OK, and the caller
+     must check (status) value. The caller can show the error,
+     if the end of stream is expected, and the (status) is noit
+     LZMA_STATUS_FINISHED_WITH_MARK or LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK.
+*/
+
+
+#define RETURN__NOT_FINISHED__FOR_FINISH \
+  *status = LZMA_STATUS_NOT_FINISHED; \
+  return SZ_ERROR_DATA; // for strict mode
+  // return SZ_OK; // for relaxed mode
+
+
 SRes LzmaDec_DecodeToDic(CLzmaDec *p, SizeT dicLimit, const Byte *src, SizeT *srcLen,
    ELzmaFinishMode finishMode, ELzmaStatus *status)
 {
  SizeT inSize = *srcLen;
  (*srcLen) = 0;
-  
  *status = LZMA_STATUS_NOT_SPECIFIED;

  if (p->remainLen > kMatchSpecLenStart)
  {
+    if (p->remainLen > kMatchSpecLenStart + 2)
+      return p->remainLen == kMatchSpecLen_Error_Fail ? SZ_ERROR_FAIL : SZ_ERROR_DATA;
+
    for (; inSize > 0 && p->tempBufSize < RC_INIT_SIZE; (*srcLen)++, inSize--)
      p->tempBuf[p->tempBufSize++] = *src++;
    if (p->tempBufSize != 0 && p->tempBuf[0] != 0)
@@ -896,6 +975,12 @@ SRes LzmaDec_DecodeToDic(CLzmaDec *p, SizeT dicLimit, const Byte *src, SizeT *sr
      | ((UInt32)p->tempBuf[2] << 16)
      | ((UInt32)p->tempBuf[3] << 8)
      | ((UInt32)p->tempBuf[4]);
+
+    if (p->checkDicSize == 0
+        && p->processedPos == 0
+        && p->code >= kBadRepCode)
+      return SZ_ERROR_DATA;
+
    p->range = 0xFFFFFFFF;
    p->tempBufSize = 0;

@@ -913,10 +998,21 @@ SRes LzmaDec_DecodeToDic(CLzmaDec *p, SizeT dicLimit, const Byte *src, SizeT *sr
    p->remainLen = 0;
  }

-  LzmaDec_WriteRem(p, dicLimit);
-
-  while (p->remainLen != kMatchSpecLenStart)
+  for (;;)
  {
+    if (p->remainLen == kMatchSpecLenStart)
+    {
+      if (p->code != 0)
+        return SZ_ERROR_DATA;
+      *status = LZMA_STATUS_FINISHED_WITH_MARK;
+      return SZ_OK;
+    }
+
+    LzmaDec_WriteRem(p, dicLimit);
+
+    {
+      // (p->remainLen == 0 || p->dicPos == dicLimit)
+
      int checkEndMarkNow = 0;

      if (p->dicPos >= dicLimit)
@@ -933,92 +1029,174 @@ SRes LzmaDec_DecodeToDic(CLzmaDec *p, SizeT dicLimit, const Byte *src, SizeT *sr
        }
        if (p->remainLen != 0)
        {
-          *status = LZMA_STATUS_NOT_FINISHED;
-          return SZ_ERROR_DATA;
+          RETURN__NOT_FINISHED__FOR_FINISH;
        }
        checkEndMarkNow = 1;
      }

+      // (p->remainLen == 0)
+
      if (p->tempBufSize == 0)
      {
-        SizeT processed;
        const Byte *bufLimit;
+        int dummyProcessed = -1;
+        
        if (inSize < LZMA_REQUIRED_INPUT_MAX || checkEndMarkNow)
        {
-          int dummyRes = LzmaDec_TryDummy(p, src, inSize);
-          if (dummyRes == DUMMY_ERROR)
+          const Byte *bufOut = src + inSize;
+          
+          ELzmaDummy dummyRes = LzmaDec_TryDummy(p, src, &bufOut);
+          
+          if (dummyRes == DUMMY_INPUT_EOF)
          {
-            memcpy(p->tempBuf, src, inSize);
-            p->tempBufSize = (unsigned)inSize;
+            size_t i;
+            if (inSize >= LZMA_REQUIRED_INPUT_MAX)
+              break;
            (*srcLen) += inSize;
+            p->tempBufSize = (unsigned)inSize;
+            for (i = 0; i < inSize; i++)
+              p->tempBuf[i] = src[i];
            *status = LZMA_STATUS_NEEDS_MORE_INPUT;
            return SZ_OK;
          }
-          if (checkEndMarkNow && dummyRes != DUMMY_MATCH)
+ 
+          dummyProcessed = (int)(bufOut - src);
+          if ((unsigned)dummyProcessed > LZMA_REQUIRED_INPUT_MAX)
+            break;
+          
+          if (checkEndMarkNow && !IS_DUMMY_END_MARKER_POSSIBLE(dummyRes))
          {
-            *status = LZMA_STATUS_NOT_FINISHED;
-            return SZ_ERROR_DATA;
+            unsigned i;
+            (*srcLen) += (unsigned)dummyProcessed;
+            p->tempBufSize = (unsigned)dummyProcessed;
+            for (i = 0; i < (unsigned)dummyProcessed; i++)
+              p->tempBuf[i] = src[i];
+            // p->remainLen = kMatchSpecLen_Error_Data;
+            RETURN__NOT_FINISHED__FOR_FINISH;
          }
+          
          bufLimit = src;
+          // we will decode only one iteration
        }
        else
          bufLimit = src + inSize - LZMA_REQUIRED_INPUT_MAX;
+
        p->buf = src;
-        if (LzmaDec_DecodeReal2(p, dicLimit, bufLimit) != 0)
-          return SZ_ERROR_DATA;
-        processed = (SizeT)(p->buf - src);
-        (*srcLen) += processed;
-        src += processed;
-        inSize -= processed;
-      }
-      else
-      {
-        unsigned rem = p->tempBufSize, lookAhead = 0;
-        while (rem < LZMA_REQUIRED_INPUT_MAX && lookAhead < inSize)
-          p->tempBuf[rem++] = src[lookAhead++];
-        p->tempBufSize = rem;
-        if (rem < LZMA_REQUIRED_INPUT_MAX || checkEndMarkNow)
+        
        {
-          int dummyRes = LzmaDec_TryDummy(p, p->tempBuf, (SizeT)rem);
-          if (dummyRes == DUMMY_ERROR)
+          int res = LzmaDec_DecodeReal2(p, dicLimit, bufLimit);
+          
+          SizeT processed = (SizeT)(p->buf - src);
+
+          if (dummyProcessed < 0)
          {
-            (*srcLen) += (SizeT)lookAhead;
-            *status = LZMA_STATUS_NEEDS_MORE_INPUT;
-            return SZ_OK;
+            if (processed > inSize)
+              break;
          }
-          if (checkEndMarkNow && dummyRes != DUMMY_MATCH)
+          else if ((unsigned)dummyProcessed != processed)
+            break;
+
+          src += processed;
+          inSize -= processed;
+          (*srcLen) += processed;
+
+          if (res != SZ_OK)
          {
-            *status = LZMA_STATUS_NOT_FINISHED;
+            p->remainLen = kMatchSpecLen_Error_Data;
            return SZ_ERROR_DATA;
          }
        }
+        continue;
+      }
+
+      {
+        // we have some data in (p->tempBuf)
+        // in strict mode: tempBufSize is not enough for one Symbol decoding.
+        // in relaxed mode: tempBufSize not larger than required for one Symbol decoding.
+
+        unsigned rem = p->tempBufSize;
+        unsigned ahead = 0;
+        int dummyProcessed = -1;
+        
+        while (rem < LZMA_REQUIRED_INPUT_MAX && ahead < inSize)
+          p->tempBuf[rem++] = src[ahead++];
+        
+        // ahead - the size of new data copied from (src) to (p->tempBuf)
+        // rem   - the size of temp buffer including new data from (src)
+        
+        if (rem < LZMA_REQUIRED_INPUT_MAX || checkEndMarkNow)
+        {
+          const Byte *bufOut = p->tempBuf + rem;
+        
+          ELzmaDummy dummyRes = LzmaDec_TryDummy(p, p->tempBuf, &bufOut);
+          
+          if (dummyRes == DUMMY_INPUT_EOF)
+          {
+            if (rem >= LZMA_REQUIRED_INPUT_MAX)
+              break;
+            p->tempBufSize = rem;
+            (*srcLen) += (SizeT)ahead;
+            *status = LZMA_STATUS_NEEDS_MORE_INPUT;
+            return SZ_OK;
+          }
+          
+          dummyProcessed = (int)(bufOut - p->tempBuf);
+
+          if ((unsigned)dummyProcessed < p->tempBufSize)
+            break;
+
+          if (checkEndMarkNow && !IS_DUMMY_END_MARKER_POSSIBLE(dummyRes))
+          {
+            (*srcLen) += (unsigned)dummyProcessed - p->tempBufSize;
+            p->tempBufSize = (unsigned)dummyProcessed;
+            // p->remainLen = kMatchSpecLen_Error_Data;
+            RETURN__NOT_FINISHED__FOR_FINISH;
+          }
+        }
+
        p->buf = p->tempBuf;
-        if (LzmaDec_DecodeReal2(p, dicLimit, p->buf) != 0)
-          return SZ_ERROR_DATA;
        
        {
-          unsigned kkk = (unsigned)(p->buf - p->tempBuf);
-          if (rem < kkk)
-            return SZ_ERROR_FAIL; /* some internal error */
-          rem -= kkk;
-          if (lookAhead < rem)
-            return SZ_ERROR_FAIL; /* some internal error */
-          lookAhead -= rem;
+          // we decode one symbol from (p->tempBuf) here, so the (bufLimit) is equal to (p->buf)
+          int res = LzmaDec_DecodeReal2(p, dicLimit, p->buf);
+
+          SizeT processed = (SizeT)(p->buf - p->tempBuf);
+          rem = p->tempBufSize;
+          
+          if (dummyProcessed < 0)
+          {
+            if (processed > LZMA_REQUIRED_INPUT_MAX)
+              break;
+            if (processed < rem)
+              break;
+          }
+          else if ((unsigned)dummyProcessed != processed)
+            break;
+          
+          processed -= rem;
+
+          src += processed;
+          inSize -= processed;
+          (*srcLen) += processed;
+          p->tempBufSize = 0;
+          
+          if (res != SZ_OK)
+          {
+            p->remainLen = kMatchSpecLen_Error_Data;
+            return SZ_ERROR_DATA;
+          }
        }
-        (*srcLen) += (SizeT)lookAhead;
-        src += lookAhead;
-        inSize -= (SizeT)lookAhead;
-        p->tempBufSize = 0;
      }
+    }
  }
-  
-  if (p->code != 0)
-    return SZ_ERROR_DATA;
-  *status = LZMA_STATUS_FINISHED_WITH_MARK;
-  return SZ_OK;
+
+  /*  Some unexpected error: internal error of code, memory corruption or hardware failure */
+  p->remainLen = kMatchSpecLen_Error_Fail;
+  return SZ_ERROR_FAIL;
 }


+
 SRes LzmaDec_DecodeToBuf(CLzmaDec *p, Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status)
 {
  SizeT outSize = *destLen;
@@ -1,5 +1,5 @@
 /* LzmaDec.h -- LZMA Decoder
-2018-04-21 : Igor Pavlov : Public domain */
+2020-03-19 : Igor Pavlov : Public domain */

 #ifndef __LZMA_DEC_H
 #define __LZMA_DEC_H
@@ -181,6 +181,7 @@ Returns:
      LZMA_STATUS_NEEDS_MORE_INPUT
      LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK
  SZ_ERROR_DATA - Data error
+  SZ_ERROR_FAIL - Some unexpected error: internal error of code, memory corruption or hardware failure
 */

 SRes LzmaDec_DecodeToDic(CLzmaDec *p, SizeT dicLimit,
@@ -223,6 +224,7 @@ Returns:
  SZ_ERROR_MEM  - Memory allocation error
  SZ_ERROR_UNSUPPORTED - Unsupported properties
  SZ_ERROR_INPUT_EOF - It needs more bytes in input buffer (src).
+  SZ_ERROR_FAIL - Some unexpected error: internal error of code, memory corruption or hardware failure
 */

 SRes LzmaDecode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen,
@@ -1,5 +1,5 @@
 /* LzmaEnc.c -- LZMA Encoder
-2019-01-10: Igor Pavlov : Public domain */
+2022-07-15: Igor Pavlov : Public domain */

 #include "Precomp.h"

@@ -12,6 +12,7 @@
 #include <stdio.h>
 #endif

+#include "CpuArch.h"
 #include "LzmaEnc.h"

 #include "LzFind.h"
@@ -19,12 +20,25 @@
 #include "LzFindMt.h"
 #endif

+/* the following LzmaEnc_* declarations is internal LZMA interface for LZMA2 encoder */
+
+SRes LzmaEnc_PrepareForLzma2(CLzmaEncHandle pp, ISeqInStream *inStream, UInt32 keepWindowSize,
+    ISzAllocPtr alloc, ISzAllocPtr allocBig);
+SRes LzmaEnc_MemPrepare(CLzmaEncHandle pp, const Byte *src, SizeT srcLen,
+    UInt32 keepWindowSize, ISzAllocPtr alloc, ISzAllocPtr allocBig);
+SRes LzmaEnc_CodeOneMemBlock(CLzmaEncHandle pp, BoolInt reInit,
+    Byte *dest, size_t *destLen, UInt32 desiredPackSize, UInt32 *unpackSize);
+const Byte *LzmaEnc_GetCurBuf(CLzmaEncHandle pp);
+void LzmaEnc_Finish(CLzmaEncHandle pp);
+void LzmaEnc_SaveState(CLzmaEncHandle pp);
+void LzmaEnc_RestoreState(CLzmaEncHandle pp);
+
 #ifdef SHOW_STAT
 static unsigned g_STAT_OFFSET = 0;
 #endif

-#define kLzmaMaxHistorySize ((UInt32)3 << 29)
-/* #define kLzmaMaxHistorySize ((UInt32)7 << 29) */
+/* for good normalization speed we still reserve 256 MB before 4 GB range */
+#define kLzmaMaxHistorySize ((UInt32)15 << 28)

 #define kNumTopBits 24
 #define kTopValue ((UInt32)1 << kNumTopBits)
@@ -36,7 +50,7 @@ static unsigned g_STAT_OFFSET = 0;

 #define kNumMoveReducingBits 4
 #define kNumBitPriceShiftBits 4
-#define kBitPrice (1 << kNumBitPriceShiftBits)
+// #define kBitPrice (1 << kNumBitPriceShiftBits)

 #define REP_LEN_COUNT 64

@@ -47,6 +61,7 @@ void LzmaEncProps_Init(CLzmaEncProps *p)
  p->reduceSize = (UInt64)(Int64)-1;
  p->lc = p->lp = p->pb = p->algo = p->fb = p->btMode = p->numHashBytes = p->numThreads = -1;
  p->writeEndMark = 0;
+  p->affinity = 0;
 }

 void LzmaEncProps_Normalize(CLzmaEncProps *p)
@@ -55,16 +70,21 @@ void LzmaEncProps_Normalize(CLzmaEncProps *p)
  if (level < 0) level = 5;
  p->level = level;
  
-  if (p->dictSize == 0) p->dictSize = (level <= 5 ? (1 << (level * 2 + 14)) : (level <= 7 ? (1 << 25) : (1 << 26)));
+  if (p->dictSize == 0)
+    p->dictSize =
+      ( level <= 3 ? ((UInt32)1 << (level * 2 + 16)) :
+      ( level <= 6 ? ((UInt32)1 << (level + 19)) :
+      ( level <= 7 ? ((UInt32)1 << 25) : ((UInt32)1 << 26)
+      )));
+
  if (p->dictSize > p->reduceSize)
  {
-    unsigned i;
-    UInt32 reduceSize = (UInt32)p->reduceSize;
-    for (i = 11; i <= 30; i++)
-    {
-      if (reduceSize <= ((UInt32)2 << i)) { p->dictSize = ((UInt32)2 << i); break; }
-      if (reduceSize <= ((UInt32)3 << i)) { p->dictSize = ((UInt32)3 << i); break; }
-    }
+    UInt32 v = (UInt32)p->reduceSize;
+    const UInt32 kReduceMin = ((UInt32)1 << 12);
+    if (v < kReduceMin)
+      v = kReduceMin;
+    if (p->dictSize > v)
+      p->dictSize = v;
  }

  if (p->lc < 0) p->lc = 3;
@@ -74,8 +94,8 @@ void LzmaEncProps_Normalize(CLzmaEncProps *p)
  if (p->algo < 0) p->algo = (level < 5 ? 0 : 1);
  if (p->fb < 0) p->fb = (level < 7 ? 32 : 64);
  if (p->btMode < 0) p->btMode = (p->algo == 0 ? 0 : 1);
-  if (p->numHashBytes < 0) p->numHashBytes = 4;
-  if (p->mc == 0) p->mc = (16 + (p->fb >> 1)) >> (p->btMode ? 0 : 1);
+  if (p->numHashBytes < 0) p->numHashBytes = (p->btMode ? 4 : 5);
+  if (p->mc == 0) p->mc = (16 + ((unsigned)p->fb >> 1)) >> (p->btMode ? 0 : 1);
  
  if (p->numThreads < 0)
    p->numThreads =
@@ -93,18 +113,85 @@ UInt32 LzmaEncProps_GetDictSize(const CLzmaEncProps *props2)
  return props.dictSize;
 }

-#if (_MSC_VER >= 1400)
-/* BSR code is fast for some new CPUs */
-/* #define LZMA_LOG_BSR */
+
+/*
+x86/x64:
+
+BSR:
+  IF (SRC == 0) ZF = 1, DEST is undefined;
+                  AMD : DEST is unchanged;
+  IF (SRC != 0) ZF = 0; DEST is index of top non-zero bit
+  BSR is slow in some processors
+
+LZCNT:
+  IF (SRC  == 0) CF = 1, DEST is size_in_bits_of_register(src) (32 or 64)
+  IF (SRC  != 0) CF = 0, DEST = num_lead_zero_bits
+  IF (DEST == 0) ZF = 1;
+
+LZCNT works only in new processors starting from Haswell.
+if LZCNT is not supported by processor, then it's executed as BSR.
+LZCNT can be faster than BSR, if supported.
+*/
+
+// #define LZMA_LOG_BSR
+
+#if defined(MY_CPU_ARM_OR_ARM64) /* || defined(MY_CPU_X86_OR_AMD64) */
+
+  #if (defined(__clang__) && (__clang_major__ >= 6)) \
+      || (defined(__GNUC__) && (__GNUC__ >= 6))
+      #define LZMA_LOG_BSR
+  #elif defined(_MSC_VER) && (_MSC_VER >= 1300)
+    // #if defined(MY_CPU_ARM_OR_ARM64)
+      #define LZMA_LOG_BSR
+    // #endif
+  #endif
 #endif

+// #include <intrin.h>
+
 #ifdef LZMA_LOG_BSR

-#define kDicLogSizeMaxCompress 32
+#if defined(__clang__) \
+    || defined(__GNUC__)

-#define BSR2_RET(pos, res) { unsigned long zz; _BitScanReverse(&zz, (pos)); res = (zz + zz) + ((pos >> (zz - 1)) & 1); }
+/*
+  C code:                  : (30 - __builtin_clz(x))
+    gcc9/gcc10 for x64 /x86  : 30 - (bsr(x) xor 31)
+    clang10 for x64          : 31 + (bsr(x) xor -32)
+*/

-static unsigned GetPosSlot1(UInt32 pos)
+  #define MY_clz(x)  ((unsigned)__builtin_clz(x))
+  // __lzcnt32
+  // __builtin_ia32_lzcnt_u32
+
+#else  // #if defined(_MSC_VER)
+
+  #ifdef MY_CPU_ARM_OR_ARM64
+
+    #define MY_clz  _CountLeadingZeros
+
+  #else // if defined(MY_CPU_X86_OR_AMD64)
+
+    // #define MY_clz  __lzcnt  // we can use lzcnt (unsupported by old CPU)
+    // _BitScanReverse code is not optimal for some MSVC compilers
+    #define BSR2_RET(pos, res) { unsigned long zz; _BitScanReverse(&zz, (pos)); zz--; \
+      res = (zz + zz) + (pos >> zz); }
+
+  #endif // MY_CPU_X86_OR_AMD64
+
+#endif // _MSC_VER
+
+
+#ifndef BSR2_RET
+
+    #define BSR2_RET(pos, res) { unsigned zz = 30 - MY_clz(pos); \
+      res = (zz + zz) + (pos >> zz); }
+
+#endif
+
+
+unsigned GetPosSlot1(UInt32 pos);
+unsigned GetPosSlot1(UInt32 pos)
 {
  unsigned res;
  BSR2_RET(pos, res);
@@ -113,10 +200,10 @@ static unsigned GetPosSlot1(UInt32 pos)
 #define GetPosSlot2(pos, res) { BSR2_RET(pos, res); }
 #define GetPosSlot(pos, res) { if (pos < 2) res = pos; else BSR2_RET(pos, res); }

-#else

-#define kNumLogBits (9 + sizeof(size_t) / 2)
-/* #define kNumLogBits (11 + sizeof(size_t) / 8 * 3) */
+#else // ! LZMA_LOG_BSR
+
+#define kNumLogBits (11 + sizeof(size_t) / 8 * 3)

 #define kDicLogSizeMaxCompress ((kNumLogBits - 1) * 2 + 7)

@@ -163,7 +250,7 @@ static void LzmaEnc_FastPosInit(Byte *g_FastPos)
 #define GetPosSlot2(pos, res) { BSR2_RET(pos, res); }
 #define GetPosSlot(pos, res) { if (pos < kNumFullDistances) res = p->g_FastPos[pos & (kNumFullDistances - 1)]; else BSR2_RET(pos, res); }

-#endif
+#endif // LZMA_LOG_BSR


 #define LZMA_NUM_REPS 4
@@ -193,7 +280,7 @@ typedef struct

 #define kNumLenToPosStates 4
 #define kNumPosSlotBits 6
-#define kDicLogSizeMin 0
+// #define kDicLogSizeMin 0
 #define kDicLogSizeMax 32
 #define kDistTableSizeMax (kDicLogSizeMax * 2)

@@ -299,7 +386,7 @@ typedef UInt32 CProbPrice;
 typedef struct
 {
  void *matchFinderObj;
-  IMatchFinder matchFinder;
+  IMatchFinder2 matchFinder;

  unsigned optCur;
  unsigned optEnd;
@@ -344,10 +431,14 @@ typedef struct
  // begin of CMatchFinderMt is used in LZ thread
  CMatchFinderMt matchFinderMt;
  // end of CMatchFinderMt is used in BT and HASH threads
+  // #else
+  // CMatchFinder matchFinderBase;
  #endif
-
  CMatchFinder matchFinderBase;

+  
+  // we suppose that we have 8-bytes alignment after CMatchFinder
+ 
  #ifndef _7ZIP_ST
  Byte pad[128];
  #endif
@@ -355,8 +446,10 @@ typedef struct
  // LZ thread
  CProbPrice ProbPrices[kBitModelTotal >> kNumMoveReducingBits];

-  UInt32 matches[LZMA_MATCH_LEN_MAX * 2 + 2 + 1];
+  // we want {len , dist} pairs to be 8-bytes aligned in matches array
+  UInt32 matches[LZMA_MATCH_LEN_MAX * 2 + 2];

+  // we want 8-bytes alignment here
  UInt32 alignPrices[kAlignTableSize];
  UInt32 posSlotPrices[kNumLenToPosStates][kDistTableSizeMax];
  UInt32 distancesPrices[kNumLenToPosStates][kNumFullDistances];
@@ -385,12 +478,19 @@ typedef struct

  CSaveState saveState;

+  // BoolInt mf_Failure;
  #ifndef _7ZIP_ST
  Byte pad2[128];
  #endif
 } CLzmaEnc;


+#define MFB (p->matchFinderBase)
+/*
+#ifndef _7ZIP_ST
+#define MFB (p->matchFinderMt.MatchFinder)
+#endif
+*/

 #define COPY_ARR(dest, src, arr) memcpy(dest->arr, src->arr, sizeof(src->arr));

@@ -455,41 +555,51 @@ SRes LzmaEnc_SetProps(CLzmaEncHandle pp, const CLzmaEncProps *props2)

  if (props.lc > LZMA_LC_MAX
      || props.lp > LZMA_LP_MAX
-      || props.pb > LZMA_PB_MAX
-      || props.dictSize > ((UInt64)1 << kDicLogSizeMaxCompress)
-      || props.dictSize > kLzmaMaxHistorySize)
+      || props.pb > LZMA_PB_MAX)
    return SZ_ERROR_PARAM;

+
+  if (props.dictSize > kLzmaMaxHistorySize)
+    props.dictSize = kLzmaMaxHistorySize;
+
+  #ifndef LZMA_LOG_BSR
+  {
+    const UInt64 dict64 = props.dictSize;
+    if (dict64 > ((UInt64)1 << kDicLogSizeMaxCompress))
+      return SZ_ERROR_PARAM;
+  }
+  #endif
+
  p->dictSize = props.dictSize;
  {
-    unsigned fb = props.fb;
+    unsigned fb = (unsigned)props.fb;
    if (fb < 5)
      fb = 5;
    if (fb > LZMA_MATCH_LEN_MAX)
      fb = LZMA_MATCH_LEN_MAX;
    p->numFastBytes = fb;
  }
-  p->lc = props.lc;
-  p->lp = props.lp;
-  p->pb = props.pb;
+  p->lc = (unsigned)props.lc;
+  p->lp = (unsigned)props.lp;
+  p->pb = (unsigned)props.pb;
  p->fastMode = (props.algo == 0);
  // p->_maxMode = True;
-  p->matchFinderBase.btMode = (Byte)(props.btMode ? 1 : 0);
+  MFB.btMode = (Byte)(props.btMode ? 1 : 0);
  {
    unsigned numHashBytes = 4;
    if (props.btMode)
    {
-      if (props.numHashBytes < 2)
-        numHashBytes = 2;
-      else if (props.numHashBytes < 4)
-        numHashBytes = props.numHashBytes;
+           if (props.numHashBytes <  2) numHashBytes = 2;
+      else if (props.numHashBytes <  4) numHashBytes = (unsigned)props.numHashBytes;
    }
-    p->matchFinderBase.numHashBytes = numHashBytes;
+    if (props.numHashBytes >= 5) numHashBytes = 5;
+
+    MFB.numHashBytes = numHashBytes;
  }

-  p->matchFinderBase.cutValue = props.mc;
+  MFB.cutValue = props.mc;

-  p->writeEndMark = props.writeEndMark;
+  p->writeEndMark = (BoolInt)props.writeEndMark;

  #ifndef _7ZIP_ST
  /*
@@ -500,6 +610,8 @@ SRes LzmaEnc_SetProps(CLzmaEncHandle pp, const CLzmaEncProps *props2)
  }
  */
  p->multiThread = (props.numThreads > 1);
+  p->matchFinderMt.btSync.affinity =
+  p->matchFinderMt.hashSync.affinity = props.affinity;
  #endif

  return SZ_OK;
@@ -509,7 +621,7 @@ SRes LzmaEnc_SetProps(CLzmaEncHandle pp, const CLzmaEncProps *props2)
 void LzmaEnc_SetDataSize(CLzmaEncHandle pp, UInt64 expectedDataSiize)
 {
  CLzmaEnc *p = (CLzmaEnc *)pp;
-  p->matchFinderBase.expectedDataSize = expectedDataSiize;
+  MFB.expectedDataSize = expectedDataSiize;
 }


@@ -536,8 +648,8 @@ static void RangeEnc_Construct(CRangeEnc *p)
  p->bufBase = NULL;
 }

-#define RangeEnc_GetProcessed(p)       ((p)->processed + ((p)->buf - (p)->bufBase) + (p)->cacheSize)
-#define RangeEnc_GetProcessed_sizet(p) ((size_t)(p)->processed + ((p)->buf - (p)->bufBase) + (size_t)(p)->cacheSize)
+#define RangeEnc_GetProcessed(p)       (        (p)->processed + (size_t)((p)->buf - (p)->bufBase) +         (p)->cacheSize)
+#define RangeEnc_GetProcessed_sizet(p) ((size_t)(p)->processed + (size_t)((p)->buf - (p)->bufBase) + (size_t)(p)->cacheSize)

 #define RC_BUF_SIZE (1 << 16)

@@ -556,12 +668,11 @@ static int RangeEnc_Alloc(CRangeEnc *p, ISzAllocPtr alloc)
 static void RangeEnc_Free(CRangeEnc *p, ISzAllocPtr alloc)
 {
  ISzAlloc_Free(alloc, p->bufBase);
-  p->bufBase = 0;
+  p->bufBase = NULL;
 }

 static void RangeEnc_Init(CRangeEnc *p)
 {
-  /* Stream.Init(); */
  p->range = 0xFFFFFFFF;
  p->cache = 0;
  p->low = 0;
@@ -575,12 +686,12 @@ static void RangeEnc_Init(CRangeEnc *p)

 MY_NO_INLINE static void RangeEnc_FlushStream(CRangeEnc *p)
 {
-  size_t num;
-  if (p->res != SZ_OK)
-    return;
-  num = p->buf - p->bufBase;
-  if (num != ISeqOutStream_Write(p->outStream, p->bufBase, num))
-    p->res = SZ_ERROR_WRITE;
+  const size_t num = (size_t)(p->buf - p->bufBase);
+  if (p->res == SZ_OK)
+  {
+    if (num != ISeqOutStream_Write(p->outStream, p->bufBase, num))
+      p->res = SZ_ERROR_WRITE;
+  }
  p->processed += num;
  p->buf = p->bufBase;
 }
@@ -656,7 +767,7 @@ static void RangeEnc_FlushData(CRangeEnc *p)
  range += newBound & mask; \
  mask &= (kBitModelTotal - ((1 << kNumMoveBits) - 1)); \
  mask += ((1 << kNumMoveBits) - 1); \
-  ttt += (Int32)(mask - ttt) >> kNumMoveBits; \
+  ttt += (UInt32)((Int32)(mask - ttt) >> kNumMoveBits); \
  *(prob) = (CLzmaProb)ttt; \
  RC_NORM(p) \
  }
@@ -749,7 +860,7 @@ static void LzmaEnc_InitPriceTables(CProbPrice *ProbPrices)
        bitCount++;
      }
    }
-    ProbPrices[i] = (CProbPrice)((kNumBitModelTotalBits << kCyclesBits) - 15 - bitCount);
+    ProbPrices[i] = (CProbPrice)(((unsigned)kNumBitModelTotalBits << kCyclesBits) - 15 - bitCount);
    // printf("\n%3d: %5d", i, ProbPrices[i]);
  }
 }
@@ -985,7 +1096,11 @@ static unsigned ReadMatchDistances(CLzmaEnc *p, unsigned *numPairsRes)
  
  p->additionalOffset++;
  p->numAvail = p->matchFinder.GetNumAvailableBytes(p->matchFinderObj);
-  numPairs = p->matchFinder.GetMatches(p->matchFinderObj, p->matches);
+  {
+    const UInt32 *d = p->matchFinder.GetMatches(p->matchFinderObj, p->matches);
+    // if (!d) { p->mf_Failure = True; *numPairsRes = 0;  return 0; }
+    numPairs = (unsigned)(d - p->matches);
+  }
  *numPairsRes = numPairs;
  
  #ifdef SHOW_STAT
@@ -1001,7 +1116,7 @@ static unsigned ReadMatchDistances(CLzmaEnc *p, unsigned *numPairsRes)
  if (numPairs == 0)
    return 0;
  {
-    unsigned len = p->matches[(size_t)numPairs - 2];
+    const unsigned len = p->matches[(size_t)numPairs - 2];
    if (len != p->numFastBytes)
      return len;
    {
@@ -1011,7 +1126,7 @@ static unsigned ReadMatchDistances(CLzmaEnc *p, unsigned *numPairsRes)
      {
        const Byte *p1 = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1;
        const Byte *p2 = p1 + len;
-        ptrdiff_t dif = (ptrdiff_t)-1 - p->matches[(size_t)numPairs - 1];
+        const ptrdiff_t dif = (ptrdiff_t)-1 - (ptrdiff_t)p->matches[(size_t)numPairs - 1];
        const Byte *lim = p1 + numAvail;
        for (; p2 != lim && *p2 == p2[dif]; p2++)
        {}
@@ -1167,6 +1282,8 @@ static unsigned GetOptimum(CLzmaEnc *p, UInt32 position)
      repLens[i] = len;
      if (len > repLens[repMaxIndex])
        repMaxIndex = i;
+      if (len == LZMA_MATCH_LEN_MAX) // 21.03 : optimization
+        break;
    }
    
    if (repLens[repMaxIndex] >= p->numFastBytes)
@@ -1179,10 +1296,12 @@ static unsigned GetOptimum(CLzmaEnc *p, UInt32 position)
    }
    
    matches = p->matches;
+    #define MATCHES  matches
+    // #define MATCHES  p->matches
    
    if (mainLen >= p->numFastBytes)
    {
-      p->backRes = matches[(size_t)numPairs - 1] + LZMA_NUM_REPS;
+      p->backRes = MATCHES[(size_t)numPairs - 1] + LZMA_NUM_REPS;
      MOVE_POS(p, mainLen - 1)
      return mainLen;
    }
@@ -1276,13 +1395,13 @@ static unsigned GetOptimum(CLzmaEnc *p, UInt32 position)
        if (len < 2)
          len = 2;
        else
-          while (len > matches[offs])
+          while (len > MATCHES[offs])
            offs += 2;
    
        for (; ; len++)
        {
          COptimal *opt;
-          UInt32 dist = matches[(size_t)offs + 1];
+          UInt32 dist = MATCHES[(size_t)offs + 1];
          UInt32 price = normalMatchPrice + GET_PRICE_LEN(&p->lenEnc, posState, len);
          unsigned lenToPosState = GetLenToPosState(len);
       
@@ -1306,7 +1425,7 @@ static unsigned GetOptimum(CLzmaEnc *p, UInt32 position)
            opt->extra = 0;
          }
          
-          if (len == matches[offs])
+          if (len == MATCHES[offs])
          {
            offs += 2;
            if (offs == numPairs)
@@ -1727,8 +1846,8 @@ static unsigned GetOptimum(CLzmaEnc *p, UInt32 position)
    if (newLen > numAvail)
    {
      newLen = numAvail;
-      for (numPairs = 0; newLen > matches[numPairs]; numPairs += 2);
-      matches[numPairs] = (UInt32)newLen;
+      for (numPairs = 0; newLen > MATCHES[numPairs]; numPairs += 2);
+      MATCHES[numPairs] = (UInt32)newLen;
      numPairs += 2;
    }
    
@@ -1747,9 +1866,9 @@ static unsigned GetOptimum(CLzmaEnc *p, UInt32 position)
      }

      offs = 0;
-      while (startLen > matches[offs])
+      while (startLen > MATCHES[offs])
        offs += 2;
-      dist = matches[(size_t)offs + 1];
+      dist = MATCHES[(size_t)offs + 1];
      
      // if (dist >= kNumFullDistances)
      GetPosSlot2(dist, posSlot);
@@ -1776,7 +1895,7 @@ static unsigned GetOptimum(CLzmaEnc *p, UInt32 position)
          }
        }

-        if (len == matches[offs])
+        if (len == MATCHES[offs])
        {
          // if (p->_maxMode) {
          // MATCH : LIT : REP_0
@@ -1841,7 +1960,7 @@ static unsigned GetOptimum(CLzmaEnc *p, UInt32 position)
          offs += 2;
          if (offs == numPairs)
            break;
-          dist = matches[(size_t)offs + 1];
+          dist = MATCHES[(size_t)offs + 1];
          // if (dist >= kNumFullDistances)
            GetPosSlot2(dist, posSlot);
        }
@@ -2059,8 +2178,23 @@ static SRes CheckErrors(CLzmaEnc *p)
    return p->result;
  if (p->rc.res != SZ_OK)
    p->result = SZ_ERROR_WRITE;
-  if (p->matchFinderBase.result != SZ_OK)
+
+  #ifndef _7ZIP_ST
+  if (
+      // p->mf_Failure ||
+        (p->mtMode &&
+          ( // p->matchFinderMt.failure_LZ_LZ ||
+            p->matchFinderMt.failure_LZ_BT))
+     )
+  {
+    p->result = MY_HRES_ERROR__INTERNAL_ERROR;
+    // printf("\nCheckErrors p->matchFinderMt.failureLZ\n");
+  }
+  #endif
+
+  if (MFB.result != SZ_OK)
    p->result = SZ_ERROR_READ;
+  
  if (p->result != SZ_OK)
    p->finished = True;
  return p->result;
@@ -2198,14 +2332,14 @@ MY_NO_INLINE static void FillDistancesPrices(CLzmaEnc *p)



-void LzmaEnc_Construct(CLzmaEnc *p)
+static void LzmaEnc_Construct(CLzmaEnc *p)
 {
  RangeEnc_Construct(&p->rc);
-  MatchFinder_Construct(&p->matchFinderBase);
+  MatchFinder_Construct(&MFB);
  
  #ifndef _7ZIP_ST
+  p->matchFinderMt.MatchFinder = &MFB;
  MatchFinderMt_Construct(&p->matchFinderMt);
-  p->matchFinderMt.MatchFinder = &p->matchFinderBase;
  #endif

  {
@@ -2221,7 +2355,6 @@ void LzmaEnc_Construct(CLzmaEnc *p)
  LzmaEnc_InitPriceTables(p->ProbPrices);
  p->litProbs = NULL;
  p->saveState.litProbs = NULL;
-
 }

 CLzmaEncHandle LzmaEnc_Create(ISzAllocPtr alloc)
@@ -2233,7 +2366,7 @@ CLzmaEncHandle LzmaEnc_Create(ISzAllocPtr alloc)
  return p;
 }

-void LzmaEnc_FreeLits(CLzmaEnc *p, ISzAllocPtr alloc)
+static void LzmaEnc_FreeLits(CLzmaEnc *p, ISzAllocPtr alloc)
 {
  ISzAlloc_Free(alloc, p->litProbs);
  ISzAlloc_Free(alloc, p->saveState.litProbs);
@@ -2241,13 +2374,13 @@ void LzmaEnc_FreeLits(CLzmaEnc *p, ISzAllocPtr alloc)
  p->saveState.litProbs = NULL;
 }

-void LzmaEnc_Destruct(CLzmaEnc *p, ISzAllocPtr alloc, ISzAllocPtr allocBig)
+static void LzmaEnc_Destruct(CLzmaEnc *p, ISzAllocPtr alloc, ISzAllocPtr allocBig)
 {
  #ifndef _7ZIP_ST
  MatchFinderMt_Destruct(&p->matchFinderMt, allocBig);
  #endif
  
-  MatchFinder_Free(&p->matchFinderBase, allocBig);
+  MatchFinder_Free(&MFB, allocBig);
  LzmaEnc_FreeLits(p, alloc);
  RangeEnc_Free(&p->rc, alloc);
 }
@@ -2259,11 +2392,18 @@ void LzmaEnc_Destroy(CLzmaEncHandle p, ISzAllocPtr alloc, ISzAllocPtr allocBig)
 }


+MY_NO_INLINE
 static SRes LzmaEnc_CodeOneBlock(CLzmaEnc *p, UInt32 maxPackSize, UInt32 maxUnpackSize)
 {
  UInt32 nowPos32, startPos32;
  if (p->needInit)
  {
+    #ifndef _7ZIP_ST
+    if (p->mtMode)
+    {
+      RINOK(MatchFinderMt_InitMt(&p->matchFinderMt));
+    }
+    #endif
    p->matchFinder.Init(p->matchFinderObj);
    p->needInit = 0;
  }
@@ -2521,12 +2661,12 @@ static SRes LzmaEnc_CodeOneBlock(CLzmaEnc *p, UInt32 maxPackSize, UInt32 maxUnpa
          // { int y; for (y = 0; y < 100; y++) {
          FillDistancesPrices(p);
          // }}
-          LenPriceEnc_UpdateTables(&p->lenEnc, 1 << p->pb, &p->lenProbs, p->ProbPrices);
+          LenPriceEnc_UpdateTables(&p->lenEnc, (unsigned)1 << p->pb, &p->lenProbs, p->ProbPrices);
        }
        if (p->repLenEncCounter <= 0)
        {
          p->repLenEncCounter = REP_LEN_COUNT;
-          LenPriceEnc_UpdateTables(&p->repLenEnc, 1 << p->pb, &p->repLenProbs, p->ProbPrices);
+          LenPriceEnc_UpdateTables(&p->repLenEnc, (unsigned)1 << p->pb, &p->repLenProbs, p->ProbPrices);
        }
      }
    
@@ -2559,11 +2699,13 @@ static SRes LzmaEnc_CodeOneBlock(CLzmaEnc *p, UInt32 maxPackSize, UInt32 maxUnpa
 static SRes LzmaEnc_Alloc(CLzmaEnc *p, UInt32 keepWindowSize, ISzAllocPtr alloc, ISzAllocPtr allocBig)
 {
  UInt32 beforeSize = kNumOpts;
+  UInt32 dictSize;
+
  if (!RangeEnc_Alloc(&p->rc, alloc))
    return SZ_ERROR_MEM;

  #ifndef _7ZIP_ST
-  p->mtMode = (p->multiThread && !p->fastMode && (p->matchFinderBase.btMode != 0));
+  p->mtMode = (p->multiThread && !p->fastMode && (MFB.btMode != 0));
  #endif

  {
@@ -2582,36 +2724,56 @@ static SRes LzmaEnc_Alloc(CLzmaEnc *p, UInt32 keepWindowSize, ISzAllocPtr alloc,
    }
  }

-  p->matchFinderBase.bigHash = (Byte)(p->dictSize > kBigHashDicLimit ? 1 : 0);
+  MFB.bigHash = (Byte)(p->dictSize > kBigHashDicLimit ? 1 : 0);

-  if (beforeSize + p->dictSize < keepWindowSize)
-    beforeSize = keepWindowSize - p->dictSize;
+
+  dictSize = p->dictSize;
+  if (dictSize == ((UInt32)2 << 30) ||
+      dictSize == ((UInt32)3 << 30))
+  {
+    /* 21.03 : here we reduce the dictionary for 2 reasons:
+       1) we don't want 32-bit back_distance matches in decoder for 2 GB dictionary.
+       2) we want to elimate useless last MatchFinder_Normalize3() for corner cases,
+          where data size is aligned for 1 GB: 5/6/8 GB.
+          That reducing must be >= 1 for such corner cases. */
+    dictSize -= 1;
+  }
+
+  if (beforeSize + dictSize < keepWindowSize)
+    beforeSize = keepWindowSize - dictSize;
+
+  /* in worst case we can look ahead for
+        max(LZMA_MATCH_LEN_MAX, numFastBytes + 1 + numFastBytes) bytes.
+     we send larger value for (keepAfter) to MantchFinder_Create():
+        (numFastBytes + LZMA_MATCH_LEN_MAX + 1)
+  */

  #ifndef _7ZIP_ST
  if (p->mtMode)
  {
-    RINOK(MatchFinderMt_Create(&p->matchFinderMt, p->dictSize, beforeSize, p->numFastBytes,
-        LZMA_MATCH_LEN_MAX
-        + 1  /* 18.04 */
+    RINOK(MatchFinderMt_Create(&p->matchFinderMt, dictSize, beforeSize,
+        p->numFastBytes, LZMA_MATCH_LEN_MAX + 1 /* 18.04 */
        , allocBig));
    p->matchFinderObj = &p->matchFinderMt;
-    p->matchFinderBase.bigHash = (Byte)(
-        (p->dictSize > kBigHashDicLimit && p->matchFinderBase.hashMask >= 0xFFFFFF) ? 1 : 0);
+    MFB.bigHash = (Byte)(
+        (p->dictSize > kBigHashDicLimit && MFB.hashMask >= 0xFFFFFF) ? 1 : 0);
    MatchFinderMt_CreateVTable(&p->matchFinderMt, &p->matchFinder);
  }
  else
  #endif
  {
-    if (!MatchFinder_Create(&p->matchFinderBase, p->dictSize, beforeSize, p->numFastBytes, LZMA_MATCH_LEN_MAX, allocBig))
+    if (!MatchFinder_Create(&MFB, dictSize, beforeSize,
+        p->numFastBytes, LZMA_MATCH_LEN_MAX + 1 /* 21.03 */
+        , allocBig))
      return SZ_ERROR_MEM;
-    p->matchFinderObj = &p->matchFinderBase;
-    MatchFinder_CreateVTable(&p->matchFinderBase, &p->matchFinder);
+    p->matchFinderObj = &MFB;
+    MatchFinder_CreateVTable(&MFB, &p->matchFinder);
  }
  
  return SZ_OK;
 }

-void LzmaEnc_Init(CLzmaEnc *p)
+static void LzmaEnc_Init(CLzmaEnc *p)
 {
  unsigned i;
  p->state = 0;
@@ -2675,12 +2837,14 @@ void LzmaEnc_Init(CLzmaEnc *p)

  p->additionalOffset = 0;

-  p->pbMask = (1 << p->pb) - 1;
+  p->pbMask = ((unsigned)1 << p->pb) - 1;
  p->lpMask = ((UInt32)0x100 << p->lp) - ((unsigned)0x100 >> p->lc);
+
+  // p->mf_Failure = False;
 }


-void LzmaEnc_InitPrices(CLzmaEnc *p)
+static void LzmaEnc_InitPrices(CLzmaEnc *p)
 {
  if (!p->fastMode)
  {
@@ -2694,8 +2858,8 @@ void LzmaEnc_InitPrices(CLzmaEnc *p)

  p->repLenEncCounter = REP_LEN_COUNT;

-  LenPriceEnc_UpdateTables(&p->lenEnc, 1 << p->pb, &p->lenProbs, p->ProbPrices);
-  LenPriceEnc_UpdateTables(&p->repLenEnc, 1 << p->pb, &p->repLenProbs, p->ProbPrices);
+  LenPriceEnc_UpdateTables(&p->lenEnc, (unsigned)1 << p->pb, &p->lenProbs, p->ProbPrices);
+  LenPriceEnc_UpdateTables(&p->repLenEnc, (unsigned)1 << p->pb, &p->repLenProbs, p->ProbPrices);
 }

 static SRes LzmaEnc_AllocAndInit(CLzmaEnc *p, UInt32 keepWindowSize, ISzAllocPtr alloc, ISzAllocPtr allocBig)
@@ -2719,7 +2883,7 @@ static SRes LzmaEnc_Prepare(CLzmaEncHandle pp, ISeqOutStream *outStream, ISeqInS
    ISzAllocPtr alloc, ISzAllocPtr allocBig)
 {
  CLzmaEnc *p = (CLzmaEnc *)pp;
-  p->matchFinderBase.stream = inStream;
+  MFB.stream = inStream;
  p->needInit = 1;
  p->rc.outStream = outStream;
  return LzmaEnc_AllocAndInit(p, 0, alloc, allocBig);
@@ -2730,16 +2894,16 @@ SRes LzmaEnc_PrepareForLzma2(CLzmaEncHandle pp,
    ISzAllocPtr alloc, ISzAllocPtr allocBig)
 {
  CLzmaEnc *p = (CLzmaEnc *)pp;
-  p->matchFinderBase.stream = inStream;
+  MFB.stream = inStream;
  p->needInit = 1;
  return LzmaEnc_AllocAndInit(p, keepWindowSize, alloc, allocBig);
 }

 static void LzmaEnc_SetInputBuf(CLzmaEnc *p, const Byte *src, SizeT srcLen)
 {
-  p->matchFinderBase.directInput = 1;
-  p->matchFinderBase.bufferBase = (Byte *)src;
-  p->matchFinderBase.directInputRem = srcLen;
+  MFB.directInput = 1;
+  MFB.bufferBase = (Byte *)src;
+  MFB.directInputRem = srcLen;
 }

 SRes LzmaEnc_MemPrepare(CLzmaEncHandle pp, const Byte *src, SizeT srcLen,
@@ -2781,19 +2945,23 @@ static size_t SeqOutStreamBuf_Write(const ISeqOutStream *pp, const void *data, s
    size = p->rem;
    p->overflow = True;
  }
-  memcpy(p->data, data, size);
-  p->rem -= size;
-  p->data += size;
+  if (size != 0)
+  {
+    memcpy(p->data, data, size);
+    p->rem -= size;
+    p->data += size;
+  }
  return size;
 }


+/*
 UInt32 LzmaEnc_GetNumAvailableBytes(CLzmaEncHandle pp)
 {
  const CLzmaEnc *p = (CLzmaEnc *)pp;
  return p->matchFinder.GetNumAvailableBytes(p->matchFinderObj);
 }
-
+*/

 const Byte *LzmaEnc_GetCurBuf(CLzmaEncHandle pp)
 {
@@ -2802,6 +2970,7 @@ const Byte *LzmaEnc_GetCurBuf(CLzmaEncHandle pp)
 }


+// (desiredPackSize == 0) is not allowed
 SRes LzmaEnc_CodeOneMemBlock(CLzmaEncHandle pp, BoolInt reInit,
    Byte *dest, size_t *destLen, UInt32 desiredPackSize, UInt32 *unpackSize)
 {
@@ -2822,14 +2991,10 @@ SRes LzmaEnc_CodeOneMemBlock(CLzmaEncHandle pp, BoolInt reInit,
  if (reInit)
    LzmaEnc_Init(p);
  LzmaEnc_InitPrices(p);
-
-  nowPos64 = p->nowPos64;
  RangeEnc_Init(&p->rc);
  p->rc.outStream = &outStream.vt;
-
-  if (desiredPackSize == 0)
-    return SZ_ERROR_OUTPUT_EOF;
-
+  nowPos64 = p->nowPos64;
+  
  res = LzmaEnc_CodeOneBlock(p, desiredPackSize, *unpackSize);
  
  *unpackSize = (UInt32)(p->nowPos64 - nowPos64);
@@ -2841,6 +3006,7 @@ SRes LzmaEnc_CodeOneMemBlock(CLzmaEncHandle pp, BoolInt reInit,
 }


+MY_NO_INLINE
 static SRes LzmaEnc_Encode2(CLzmaEnc *p, ICompressProgress *progress)
 {
  SRes res = SZ_OK;
@@ -2870,7 +3036,7 @@ static SRes LzmaEnc_Encode2(CLzmaEnc *p, ICompressProgress *progress)
  LzmaEnc_Finish(p);

  /*
-  if (res == SZ_OK && !Inline_MatchFinder_IsFinishedOK(&p->matchFinderBase))
+  if (res == SZ_OK && !Inline_MatchFinder_IsFinishedOK(&MFB))
    res = SZ_ERROR_FAIL;
  }
  */
@@ -2889,35 +3055,43 @@ SRes LzmaEnc_Encode(CLzmaEncHandle pp, ISeqOutStream *outStream, ISeqInStream *i

 SRes LzmaEnc_WriteProperties(CLzmaEncHandle pp, Byte *props, SizeT *size)
 {
-  CLzmaEnc *p = (CLzmaEnc *)pp;
-  unsigned i;
-  UInt32 dictSize = p->dictSize;
  if (*size < LZMA_PROPS_SIZE)
    return SZ_ERROR_PARAM;
  *size = LZMA_PROPS_SIZE;
-  props[0] = (Byte)((p->pb * 5 + p->lp) * 9 + p->lc);
-
-  if (dictSize >= ((UInt32)1 << 22))
  {
-    UInt32 kDictMask = ((UInt32)1 << 20) - 1;
-    if (dictSize < (UInt32)0xFFFFFFFF - kDictMask)
-      dictSize = (dictSize + kDictMask) & ~kDictMask;
-  }
-  else for (i = 11; i <= 30; i++)
-  {
-    if (dictSize <= ((UInt32)2 << i)) { dictSize = (2 << i); break; }
-    if (dictSize <= ((UInt32)3 << i)) { dictSize = (3 << i); break; }
-  }
+    const CLzmaEnc *p = (const CLzmaEnc *)pp;
+    const UInt32 dictSize = p->dictSize;
+    UInt32 v;
+    props[0] = (Byte)((p->pb * 5 + p->lp) * 9 + p->lc);
+    
+    // we write aligned dictionary value to properties for lzma decoder
+    if (dictSize >= ((UInt32)1 << 21))
+    {
+      const UInt32 kDictMask = ((UInt32)1 << 20) - 1;
+      v = (dictSize + kDictMask) & ~kDictMask;
+      if (v < dictSize)
+        v = dictSize;
+    }
+    else
+    {
+      unsigned i = 11 * 2;
+      do
+      {
+        v = (UInt32)(2 + (i & 1)) << (i >> 1);
+        i++;
+      }
+      while (v < dictSize);
+    }

-  for (i = 0; i < 4; i++)
-    props[1 + i] = (Byte)(dictSize >> (8 * i));
-  return SZ_OK;
+    SetUi32(props + 1, v);
+    return SZ_OK;
+  }
 }


 unsigned LzmaEnc_IsWriteEndMark(CLzmaEncHandle pp)
 {
-  return ((CLzmaEnc *)pp)->writeEndMark;
+  return (unsigned)((CLzmaEnc *)pp)->writeEndMark;
 }


@@ -2974,3 +3148,15 @@ SRes LzmaEncode(Byte *dest, SizeT *destLen, const Byte *src, SizeT srcLen,
  LzmaEnc_Destroy(p, alloc, allocBig);
  return res;
 }
+
+
+/*
+#ifndef _7ZIP_ST
+void LzmaEnc_GetLzThreads(CLzmaEncHandle pp, HANDLE lz_threads[2])
+{
+  const CLzmaEnc *p = (CLzmaEnc *)pp;
+  lz_threads[0] = p->matchFinderMt.hashSync.thread;
+  lz_threads[1] = p->matchFinderMt.btSync.thread;
+}
+#endif
+*/
@@ -1,5 +1,5 @@
 /*  LzmaEnc.h -- LZMA Encoder
-2017-07-27 : Igor Pavlov : Public domain */
+2019-10-30 : Igor Pavlov : Public domain */

 #ifndef __LZMA_ENC_H
 #define __LZMA_ENC_H
@@ -29,6 +29,8 @@ typedef struct _CLzmaEncProps

  UInt64 reduceSize; /* estimated size of data that will be compressed. default = (UInt64)(Int64)-1.
                        Encoder uses this value to reduce dictionary size */
+
+  UInt64 affinity;
 } CLzmaEncProps;

 void LzmaEncProps_Init(CLzmaEncProps *p);
--- a/Show More
+++ b/Show More