diff --git a/.github/workflows/build-ton-linux-android-tonlib.yml b/.github/workflows/build-ton-linux-android-tonlib.yml index ae1be22f..bbd95661 100644 --- a/.github/workflows/build-ton-linux-android-tonlib.yml +++ b/.github/workflows/build-ton-linux-android-tonlib.yml @@ -17,7 +17,7 @@ jobs: sudo apt-get update sudo apt-get install -y build-essential git cmake ninja-build automake libtool texinfo autoconf libgflags-dev \ zlib1g-dev libssl-dev libreadline-dev libmicrohttpd-dev pkg-config libgsl-dev python3 python3-dev \ - libtool autoconf libsodium-dev libsecp256k1-dev + libtool autoconf libsodium-dev libsecp256k1-dev liblz4-dev - name: Build TON run: | @@ -29,4 +29,4 @@ jobs: uses: actions/upload-artifact@master with: name: tonlib-android - path: artifacts \ No newline at end of file + path: artifacts diff --git a/.github/workflows/build-ton-linux-x86-64-shared.yml b/.github/workflows/build-ton-linux-x86-64-shared.yml index 166de660..ce0ade64 100644 --- a/.github/workflows/build-ton-linux-x86-64-shared.yml +++ b/.github/workflows/build-ton-linux-x86-64-shared.yml @@ -19,7 +19,7 @@ jobs: - name: Install system libraries run: | sudo apt-get update - sudo apt-get install -y build-essential git cmake ninja-build zlib1g-dev libsecp256k1-dev libmicrohttpd-dev libsodium-dev + sudo apt-get install -y build-essential git cmake ninja-build zlib1g-dev libsecp256k1-dev libmicrohttpd-dev libsodium-dev liblz4-dev - name: Install clang-16 run: | diff --git a/.github/workflows/build-ton-wasm-emscripten.yml b/.github/workflows/build-ton-wasm-emscripten.yml index 16156b07..92107ffd 100644 --- a/.github/workflows/build-ton-wasm-emscripten.yml +++ b/.github/workflows/build-ton-wasm-emscripten.yml @@ -15,7 +15,7 @@ jobs: - name: Install system libraries run: | sudo apt-get update - sudo apt-get install -y build-essential git openssl cmake ninja-build zlib1g-dev libssl-dev libsecp256k1-dev libmicrohttpd-dev libsodium-dev + sudo apt-get install -y build-essential git openssl cmake ninja-build zlib1g-dev libssl-dev libsecp256k1-dev libmicrohttpd-dev libsodium-dev liblz4-dev - name: Build TON WASM artifacts run: | @@ -27,4 +27,4 @@ jobs: uses: actions/upload-artifact@master with: name: ton-wasm-binaries - path: artifacts \ No newline at end of file + path: artifacts diff --git a/.github/workflows/create-release.yml b/.github/workflows/create-release.yml index 3d248cfe..50d2661b 100644 --- a/.github/workflows/create-release.yml +++ b/.github/workflows/create-release.yml @@ -191,6 +191,14 @@ jobs: asset_name: tonlibjson.dll tag: ${{ steps.tag.outputs.TAG }} + - name: Upload Windows 2019 single artifact - libemulator + uses: svenstaro/upload-release-action@v2 + with: + repo_token: ${{ secrets.GITHUB_TOKEN }} + file: artifacts/ton-win-binaries/emulator.dll + asset_name: libemulator.dll + tag: ${{ steps.tag.outputs.TAG }} + - name: Upload Windows 2019 single artifact - tonlib-cli uses: svenstaro/upload-release-action@v2 with: @@ -273,6 +281,14 @@ jobs: asset_name: tonlibjson-mac-x86-64.dylib tag: ${{ steps.tag.outputs.TAG }} + - name: Upload Mac x86-64 single artifact - libemulator + uses: svenstaro/upload-release-action@v2 + with: + repo_token: ${{ secrets.GITHUB_TOKEN }} + file: artifacts/ton-x86_64-macos-binaries/libemulator.dylib + asset_name: libemulator-mac-x86-64.dylib + tag: ${{ steps.tag.outputs.TAG }} + - name: Upload Mac x86-64 single artifact - tonlib-cli uses: svenstaro/upload-release-action@v2 with: @@ -355,6 +371,14 @@ jobs: asset_name: tonlibjson-linux-x86_64.so tag: ${{ steps.tag.outputs.TAG }} + - name: Upload Linux x86-64 single artifact - libemulator + uses: svenstaro/upload-release-action@v2 + with: + repo_token: ${{ secrets.GITHUB_TOKEN }} + file: artifacts/ton-x86_64-linux-binaries/libemulator.so + asset_name: libemulator-linux-x86_64.so + tag: ${{ steps.tag.outputs.TAG }} + - name: Upload Linux x86-64 single artifact - tonlib-cli uses: svenstaro/upload-release-action@v2 with: diff --git a/assembly/cicd/jenkins/test-builds.groovy b/assembly/cicd/jenkins/test-builds.groovy index 960ac8db..a959d75a 100644 --- a/assembly/cicd/jenkins/test-builds.groovy +++ b/assembly/cicd/jenkins/test-builds.groovy @@ -8,7 +8,7 @@ pipeline { label 'Ubuntu_x86-64' } steps { - timeout(time: 90, unit: 'MINUTES') { + timeout(time: 180, unit: 'MINUTES') { sh ''' cp assembly/native/build-ubuntu-shared.sh . chmod +x build-ubuntu-shared.sh @@ -27,7 +27,7 @@ pipeline { label 'Ubuntu_x86-64' } steps { - timeout(time: 90, unit: 'MINUTES') { + timeout(time: 180, unit: 'MINUTES') { sh ''' cp assembly/nix/build-linux-x86-64-nix.sh . chmod +x build-linux-x86-64-nix.sh @@ -46,7 +46,7 @@ pipeline { label 'Ubuntu_arm64' } steps { - timeout(time: 90, unit: 'MINUTES') { + timeout(time: 180, unit: 'MINUTES') { sh ''' cp assembly/native/build-ubuntu-shared.sh . chmod +x build-ubuntu-shared.sh @@ -65,7 +65,7 @@ pipeline { label 'Ubuntu_arm64' } steps { - timeout(time: 90, unit: 'MINUTES') { + timeout(time: 180, unit: 'MINUTES') { sh ''' cp assembly/nix/build-linux-arm64-nix.sh . chmod +x build-linux-arm64-nix.sh @@ -84,7 +84,7 @@ pipeline { label 'macOS_12.7_x86-64' } steps { - timeout(time: 90, unit: 'MINUTES') { + timeout(time: 180, unit: 'MINUTES') { sh ''' cp assembly/native/build-macos-shared.sh . chmod +x build-macos-shared.sh @@ -103,7 +103,7 @@ pipeline { label 'macOS_12.7_x86-64' } steps { - timeout(time: 90, unit: 'MINUTES') { + timeout(time: 180, unit: 'MINUTES') { sh ''' cp assembly/nix/build-macos-nix.sh . chmod +x build-macos-nix.sh @@ -122,7 +122,7 @@ pipeline { label 'macOS_12.6-arm64-m1' } steps { - timeout(time: 90, unit: 'MINUTES') { + timeout(time: 180, unit: 'MINUTES') { sh ''' cp assembly/native/build-macos-shared.sh . chmod +x build-macos-shared.sh @@ -141,7 +141,7 @@ pipeline { label 'macOS_12.6-arm64-m1' } steps { - timeout(time: 90, unit: 'MINUTES') { + timeout(time: 180, unit: 'MINUTES') { sh ''' cp assembly/nix/build-macos-nix.sh . chmod +x build-macos-nix.sh @@ -160,7 +160,7 @@ pipeline { label 'macOS_13.2-arm64-m2' } steps { - timeout(time: 90, unit: 'MINUTES') { + timeout(time: 180, unit: 'MINUTES') { sh ''' cp assembly/native/build-macos-shared.sh . chmod +x build-macos-shared.sh @@ -179,10 +179,10 @@ pipeline { label 'Windows_x86-64' } steps { - timeout(time: 90, unit: 'MINUTES') { + timeout(time: 180, unit: 'MINUTES') { bat ''' copy assembly\\native\\build-windows.bat . - build-windows.bat + build-windows.bat -t ''' bat ''' cd artifacts @@ -197,7 +197,7 @@ pipeline { label 'Ubuntu_x86-64' } steps { - timeout(time: 90, unit: 'MINUTES') { + timeout(time: 180, unit: 'MINUTES') { sh ''' cp assembly/android/build-android-tonlib.sh . chmod +x build-android-tonlib.sh @@ -216,7 +216,7 @@ pipeline { label 'Ubuntu_x86-64' } steps { - timeout(time: 90, unit: 'MINUTES') { + timeout(time: 180, unit: 'MINUTES') { sh ''' cd assembly/wasm chmod +x fift-func-wasm-build-ubuntu.sh diff --git a/assembly/native/build-macos-portable.sh b/assembly/native/build-macos-portable.sh index a4187d4c..32a09e45 100644 --- a/assembly/native/build-macos-portable.sh +++ b/assembly/native/build-macos-portable.sh @@ -37,6 +37,21 @@ else fi export CCACHE_DISABLE=1 +if [ ! -d "lz4" ]; then +git clone https://github.com/lz4/lz4.git +cd lz4 +lz4Path=`pwd` +git checkout v1.9.4 +make -j12 +test $? -eq 0 || { echo "Can't compile lz4"; exit 1; } +cd .. +# ./lib/liblz4.a +# ./lib +else + lz4Path=$(pwd)/lz4 + echo "Using compiled lz4" +fi + if [ ! -d "secp256k1" ]; then git clone https://github.com/bitcoin-core/secp256k1.git cd secp256k1 @@ -128,7 +143,10 @@ cmake -GNinja .. \ -DSODIUM_LIBRARY_RELEASE=$sodiumPath/src/libsodium/.libs/libsodium.a \ -DMHD_FOUND=1 \ -DMHD_INCLUDE_DIR=$libmicrohttpdPath/src/include \ --DMHD_LIBRARY=$libmicrohttpdPath/src/microhttpd/.libs/libmicrohttpd.a +-DMHD_LIBRARY=$libmicrohttpdPath/src/microhttpd/.libs/libmicrohttpd.a \ +-DLZ4_FOUND=1 \ +-DLZ4_INCLUDE_DIRS=$lz4Path/lib \ +-DLZ4_LIBRARIES=$lz4Path/lib/liblz4.a test $? -eq 0 || { echo "Can't configure ton"; exit 1; } diff --git a/assembly/native/build-macos-shared.sh b/assembly/native/build-macos-shared.sh index 7b4f90ee..0f16eeda 100644 --- a/assembly/native/build-macos-shared.sh +++ b/assembly/native/build-macos-shared.sh @@ -51,6 +51,19 @@ else echo "Using compiled secp256k1" fi +if [ ! -d "lz4" ]; then + git clone https://github.com/lz4/lz4 + cd lz4 + lz4Path=`pwd` + git checkout v1.9.4 + make -j12 + test $? -eq 0 || { echo "Can't compile lz4"; exit 1; } + cd .. +else + lz4Path=$(pwd)/lz4 + echo "Using compiled lz4" +fi + brew unlink openssl@1.1 brew install openssl@3 brew unlink openssl@3 && brew link --overwrite openssl@3 @@ -59,7 +72,10 @@ cmake -GNinja -DCMAKE_BUILD_TYPE=Release .. \ -DCMAKE_CXX_FLAGS="-stdlib=libc++" \ -DSECP256K1_FOUND=1 \ -DSECP256K1_INCLUDE_DIR=$secp256k1Path/include \ --DSECP256K1_LIBRARY=$secp256k1Path/.libs/libsecp256k1.a +-DSECP256K1_LIBRARY=$secp256k1Path/.libs/libsecp256k1.a \ +-DLZ4_FOUND=1 \ +-DLZ4_LIBRARIES=$lz4Path/lib/liblz4.a \ +-DLZ4_INCLUDE_DIRS=$lz4Path/lib test $? -eq 0 || { echo "Can't configure ton"; exit 1; } diff --git a/assembly/native/build-ubuntu-portable.sh b/assembly/native/build-ubuntu-portable.sh index 81dbe710..a3a11f1b 100644 --- a/assembly/native/build-ubuntu-portable.sh +++ b/assembly/native/build-ubuntu-portable.sh @@ -28,6 +28,20 @@ export CC=$(which clang-16) export CXX=$(which clang++-16) export CCACHE_DISABLE=1 +if [ ! -d "lz4" ]; then +git clone https://github.com/lz4/lz4.git +cd lz4 +lz4Path=`pwd` +git checkout v1.9.4 +make -j12 +test $? -eq 0 || { echo "Can't compile lz4"; exit 1; } +cd .. +# ./lib/liblz4.a +# ./lib +else + lz4Path=$(pwd)/lz4 + echo "Using compiled lz4" +fi if [ ! -d "secp256k1" ]; then git clone https://github.com/bitcoin-core/secp256k1.git @@ -120,7 +134,11 @@ cmake -GNinja .. \ -DSODIUM_LIBRARY_RELEASE=$sodiumPath/src/libsodium/.libs/libsodium.a \ -DMHD_FOUND=1 \ -DMHD_INCLUDE_DIR=$libmicrohttpdPath/src/include \ --DMHD_LIBRARY=$libmicrohttpdPath/src/microhttpd/.libs/libmicrohttpd.a +-DMHD_LIBRARY=$libmicrohttpdPath/src/microhttpd/.libs/libmicrohttpd.a \ +-DLZ4_FOUND=1 \ +-DLZ4_INCLUDE_DIRS=$lz4Path/lib \ +-DLZ4_LIBRARIES=$lz4Path/lib/liblz4.a + test $? -eq 0 || { echo "Can't configure ton"; exit 1; } diff --git a/assembly/native/build-ubuntu-shared.sh b/assembly/native/build-ubuntu-shared.sh index 12c819cd..7d5326a4 100644 --- a/assembly/native/build-ubuntu-shared.sh +++ b/assembly/native/build-ubuntu-shared.sh @@ -1,7 +1,7 @@ #/bin/bash #sudo apt-get update -#sudo apt-get install -y build-essential git cmake ninja-build zlib1g-dev libsecp256k1-dev libmicrohttpd-dev libsodium-dev +#sudo apt-get install -y build-essential git cmake ninja-build zlib1g-dev libsecp256k1-dev libmicrohttpd-dev libsodium-dev liblz4-dev with_tests=false with_artifacts=false @@ -119,4 +119,4 @@ if [ "$with_tests" = true ]; then cd build # ctest --output-on-failure -E "test-catchain|test-actors|test-smartcont|test-adnl|test-validator-session-state|test-dht|test-rldp" ctest --output-on-failure --timeout 1800 -fi \ No newline at end of file +fi diff --git a/assembly/native/build-windows.bat b/assembly/native/build-windows.bat index 2e3f2082..9b7322e1 100644 --- a/assembly/native/build-windows.bat +++ b/assembly/native/build-windows.bat @@ -14,24 +14,26 @@ choco feature enable -n allowEmptyChecksums echo Installing pkgconfiglite... choco install -y pkgconfiglite -IF errorlevel 1 ( +IF %errorlevel% NEQ 0 ( echo Can't install pkgconfiglite exit /b %errorlevel% ) echo Installing ninja... choco install -y ninja -IF errorlevel 1 ( +IF %errorlevel% NEQ 0 ( echo Can't install ninja exit /b %errorlevel% ) if not exist "zlib" ( git clone https://github.com/madler/zlib.git -cd zlib\contrib\vstudio\vc14 +cd zlib +git checkout v1.3.1 +cd contrib\vstudio\vc14 msbuild zlibstat.vcxproj /p:Configuration=ReleaseWithoutAsm /p:platform=x64 -p:PlatformToolset=v143 -IF errorlevel 1 ( +IF %errorlevel% NEQ 0 ( echo Can't install zlib exit /b %errorlevel% ) @@ -40,23 +42,45 @@ cd ..\..\..\.. echo Using zlib... ) -if not exist "secp256k1" ( -git clone https://github.com/libbitcoin/secp256k1.git -cd secp256k1\builds\msvc\vs2017 -msbuild /p:Configuration=StaticRelease -p:PlatformToolset=v143 -p:Platform=x64 -IF errorlevel 1 ( - echo Can't install secp256k1 +if not exist "lz4" ( +git clone https://github.com/lz4/lz4.git +cd lz4 +git checkout v1.9.4 +cd build\VS2017\liblz4 +msbuild liblz4.vcxproj /p:Configuration=Release /p:platform=x64 -p:PlatformToolset=v143 +dir /s +IF %errorlevel% NEQ 0 ( + echo Can't install lz4 exit /b %errorlevel% ) cd ..\..\..\.. ) else ( +echo Using lz4... +) + +if not exist "secp256k1" ( +git clone https://github.com/bitcoin-core/secp256k1.git +cd secp256k1 +git checkout v0.3.2 +cmake -G "Visual Studio 17 2022" -A x64 -S . -B build -DSECP256K1_ENABLE_MODULE_RECOVERY=ON -DBUILD_SHARED_LIBS=OFF +IF %errorlevel% NEQ 0 ( + echo Can't configure secp256k1 + exit /b %errorlevel% +) +cmake --build build --config Release +IF %errorlevel% NEQ 0 ( + echo Can't install secp256k1 + exit /b %errorlevel% +) +cd .. +) else ( echo Using secp256k1... ) if not exist "libsodium" ( curl -Lo libsodium-1.0.18-stable-msvc.zip https://download.libsodium.org/libsodium/releases/libsodium-1.0.18-stable-msvc.zip -IF errorlevel 1 ( +IF %errorlevel% NEQ 0 ( echo Can't download libsodium exit /b %errorlevel% ) @@ -67,7 +91,7 @@ echo Using libsodium... if not exist "openssl-3.1.4" ( curl -Lo openssl-3.1.4.zip https://github.com/neodiX42/precompiled-openssl-win64/raw/main/openssl-3.1.4.zip -IF errorlevel 1 ( +IF %errorlevel% NEQ 0 ( echo Can't download OpenSSL exit /b %errorlevel% ) @@ -78,7 +102,7 @@ echo Using openssl... if not exist "libmicrohttpd-0.9.77-w32-bin" ( curl -Lo libmicrohttpd-0.9.77-w32-bin.zip https://github.com/neodiX42/precompiled-openssl-win64/raw/main/libmicrohttpd-0.9.77-w32-bin.zip -IF errorlevel 1 ( +IF %errorlevel% NEQ 0 ( echo Can't download libmicrohttpd exit /b %errorlevel% ) @@ -89,7 +113,7 @@ echo Using libmicrohttpd... if not exist "readline-5.0-1-lib" ( curl -Lo readline-5.0-1-lib.zip https://github.com/neodiX42/precompiled-openssl-win64/raw/main/readline-5.0-1-lib.zip -IF errorlevel 1 ( +IF %errorlevel% NEQ 0 ( echo Can't download readline exit /b %errorlevel% ) @@ -110,7 +134,10 @@ cmake -GNinja -DCMAKE_BUILD_TYPE=Release ^ -DSODIUM_USE_STATIC_LIBS=1 ^ -DSECP256K1_FOUND=1 ^ -DSECP256K1_INCLUDE_DIR=%root%\secp256k1\include ^ --DSECP256K1_LIBRARY=%root%\secp256k1\bin\x64\Release\v143\static\secp256k1.lib ^ +-DSECP256K1_LIBRARY=%root%\secp256k1\build\src\Release\libsecp256k1.lib ^ +-DLZ4_FOUND=1 ^ +-DLZ4_INCLUDE_DIRS=%root%\lz4\lib ^ +-DLZ4_LIBRARIES=%root%\lz4\build\VS2017\liblz4\bin\x64_Release\liblz4_static.lib ^ -DMHD_FOUND=1 ^ -DMHD_LIBRARY=%root%\libmicrohttpd-0.9.77-w32-bin\x86_64\VS2019\Release-static\libmicrohttpd.lib ^ -DMHD_INCLUDE_DIR=%root%\libmicrohttpd-0.9.77-w32-bin\x86_64\VS2019\Release-static ^ @@ -118,10 +145,12 @@ cmake -GNinja -DCMAKE_BUILD_TYPE=Release ^ -DZLIB_INCLUDE_DIR=%root%\zlib ^ -DZLIB_LIBRARIES=%root%\zlib\contrib\vstudio\vc14\x64\ZlibStatReleaseWithoutAsm\zlibstat.lib ^ -DOPENSSL_FOUND=1 ^ --DOPENSSL_INCLUDE_DIR=%root%/openssl-3.1.4/x64/include ^ --DOPENSSL_CRYPTO_LIBRARY=%root%/openssl-3.1.4/x64/lib/libcrypto_static.lib ^ +-DOPENSSL_INCLUDE_DIR=%root%\openssl-3.1.4\x64\include ^ +-DOPENSSL_CRYPTO_LIBRARY=%root%\openssl-3.1.4\x64\lib\libcrypto_static.lib ^ +-DREADLINE_INCLUDE_DIR=%root%\readline-5.0-1-lib\include ^ +-DREADLINE_LIBRARY=%root%\readline-5.0-1-lib\lib\readline.lib ^ -DCMAKE_CXX_FLAGS="/DTD_WINDOWS=1 /EHsc /bigobj" .. -IF errorlevel 1 ( +IF %errorlevel% NEQ 0 ( echo Can't configure TON exit /b %errorlevel% ) @@ -133,7 +162,7 @@ json2tlo dht-server http-proxy rldp-http-proxy adnl-proxy create-state create-ha test-ed25519 test-ed25519-crypto test-bigint test-vm test-fift test-cells test-smartcont test-net ^ test-tdactor test-tdutils test-tonlib-offline test-adnl test-dht test-rldp test-rldp2 test-catchain ^ test-fec test-tddb test-db test-validator-session-state -IF errorlevel 1 ( +IF %errorlevel% NEQ 0 ( echo Can't compile TON exit /b %errorlevel% ) @@ -141,14 +170,14 @@ IF errorlevel 1 ( ninja storage-daemon storage-daemon-cli blockchain-explorer fift func tonlib tonlibjson ^ tonlib-cli validator-engine lite-client pow-miner validator-engine-console generate-random-id ^ json2tlo dht-server http-proxy rldp-http-proxy adnl-proxy create-state create-hardfork emulator -IF errorlevel 1 ( +IF %errorlevel% NEQ 0 ( echo Can't compile TON exit /b %errorlevel% ) ) copy validator-engine\validator-engine.exe test -IF errorlevel 1 ( +IF %errorlevel% NEQ 0 ( echo validator-engine.exe does not exist exit /b %errorlevel% ) @@ -156,8 +185,8 @@ IF errorlevel 1 ( IF "%1"=="-t" ( echo Running tests... REM ctest -C Release --output-on-failure -E "test-catchain|test-actors|test-validator-session-state" - ctest -C Release --output-on-failure --timeout 1800 - IF errorlevel 1 ( + ctest -C Release --output-on-failure -E "test-bigint" --timeout 1800 + IF %errorlevel% NEQ 0 ( echo Some tests failed exit /b %errorlevel% ) diff --git a/assembly/nix/linux-arm64-static.nix b/assembly/nix/linux-arm64-static.nix index 5e834269..8c2749b0 100644 --- a/assembly/nix/linux-arm64-static.nix +++ b/assembly/nix/linux-arm64-static.nix @@ -22,7 +22,7 @@ stdenv.mkDerivation { buildInputs = with pkgs; [ - pkgsStatic.openssl microhttpdmy pkgsStatic.zlib pkgsStatic.libsodium.dev pkgsStatic.secp256k1 glibc.static + pkgsStatic.openssl microhttpdmy pkgsStatic.zlib pkgsStatic.libsodium.dev pkgsStatic.secp256k1 glibc.static pkgsStatic.lz4 ]; makeStatic = true; diff --git a/assembly/nix/linux-arm64-tonlib.nix b/assembly/nix/linux-arm64-tonlib.nix index a753423b..ae62ca26 100644 --- a/assembly/nix/linux-arm64-tonlib.nix +++ b/assembly/nix/linux-arm64-tonlib.nix @@ -21,7 +21,7 @@ pkgs.llvmPackages_16.stdenv.mkDerivation { buildInputs = with pkgs; [ - pkgsStatic.openssl microhttpdmy pkgsStatic.zlib pkgsStatic.libsodium.dev pkgsStatic.secp256k1 + pkgsStatic.openssl microhttpdmy pkgsStatic.zlib pkgsStatic.libsodium.dev pkgsStatic.secp256k1 pkgsStatic.lz4 ]; dontAddStaticConfigureFlags = false; diff --git a/assembly/nix/linux-x86-64-static.nix b/assembly/nix/linux-x86-64-static.nix index 5e834269..8c2749b0 100644 --- a/assembly/nix/linux-x86-64-static.nix +++ b/assembly/nix/linux-x86-64-static.nix @@ -22,7 +22,7 @@ stdenv.mkDerivation { buildInputs = with pkgs; [ - pkgsStatic.openssl microhttpdmy pkgsStatic.zlib pkgsStatic.libsodium.dev pkgsStatic.secp256k1 glibc.static + pkgsStatic.openssl microhttpdmy pkgsStatic.zlib pkgsStatic.libsodium.dev pkgsStatic.secp256k1 glibc.static pkgsStatic.lz4 ]; makeStatic = true; diff --git a/assembly/nix/linux-x86-64-tonlib.nix b/assembly/nix/linux-x86-64-tonlib.nix index ac183d2b..5a6e43e8 100644 --- a/assembly/nix/linux-x86-64-tonlib.nix +++ b/assembly/nix/linux-x86-64-tonlib.nix @@ -34,7 +34,7 @@ stdenv227.mkDerivation { buildInputs = with pkgs; [ - pkgsStatic.openssl pkgsStatic.zlib pkgsStatic.libmicrohttpd.dev pkgsStatic.libsodium.dev pkgsStatic.secp256k1 + pkgsStatic.openssl pkgsStatic.zlib pkgsStatic.libmicrohttpd.dev pkgsStatic.libsodium.dev pkgsStatic.secp256k1 pkgsStatic.lz4 ]; dontAddStaticConfigureFlags = false; diff --git a/assembly/nix/macos-static.nix b/assembly/nix/macos-static.nix index a3d4667f..be15579c 100644 --- a/assembly/nix/macos-static.nix +++ b/assembly/nix/macos-static.nix @@ -25,6 +25,7 @@ pkgs.llvmPackages_14.stdenv.mkDerivation { (openssl.override { static = true; }).dev (zlib.override { shared = false; }).dev (libiconv.override { enableStatic = true; enableShared = false; }) + (lz4.override { enableStatic = true; enableShared = false; }).dev ]; diff --git a/assembly/nix/macos-tonlib.nix b/assembly/nix/macos-tonlib.nix index c362de4e..4f7b7620 100644 --- a/assembly/nix/macos-tonlib.nix +++ b/assembly/nix/macos-tonlib.nix @@ -23,6 +23,7 @@ pkgs.llvmPackages_14.stdenv.mkDerivation { (openssl.override { static = true; }).dev (zlib.override { shared = false; }).dev (libiconv.override { enableStatic = true; enableShared = false; }) + (lz4.override { enableStatic = true; enableShared = false; }).dev ]; dontAddStaticConfigureFlags = true; @@ -52,4 +53,4 @@ pkgs.llvmPackages_14.stdenv.mkDerivation { install_name_tool -change "$(otool -L "$fn" | grep libc++abi.1 | cut -d' ' -f1 | xargs)" libc++abi.dylib "$fn" done ''; -} \ No newline at end of file +} diff --git a/assembly/wasm/fift-func-wasm-build-ubuntu.sh b/assembly/wasm/fift-func-wasm-build-ubuntu.sh index 9ca23cc0..6daf2d4c 100644 --- a/assembly/wasm/fift-func-wasm-build-ubuntu.sh +++ b/assembly/wasm/fift-func-wasm-build-ubuntu.sh @@ -41,6 +41,11 @@ cd zlib ZLIB_DIR=`pwd` cd .. +git clone https://github.com/lz4/lz4.git +cd lz4 +LZ4_DIR=`pwd` +cd .. + git clone https://github.com/bitcoin-core/secp256k1.git cd secp256k1 ./autogen.sh @@ -101,7 +106,10 @@ cd ../zlib emconfigure ./configure --static emmake make -j16 test $? -eq 0 || { echo "Can't compile zlib with emmake "; exit 1; } -ZLIB_DIR=`pwd` + +cd ../lz4 +emmake make -j16 +test $? -eq 0 || { echo "Can't compile lz4 with emmake "; exit 1; } cd ../secp256k1 @@ -121,6 +129,9 @@ emcmake cmake -DUSE_EMSCRIPTEN=ON -DCMAKE_BUILD_TYPE=Release \ -DZLIB_FOUND=1 \ -DZLIB_LIBRARIES=$ZLIB_DIR/libz.a \ -DZLIB_INCLUDE_DIR=$ZLIB_DIR \ +-DLZ4_FOUND=1 \ +-DLZ4_LIBRARIES=$LZ4_DIR/lib/liblz4.a \ +-DLZ4_INCLUDE_DIRS=$LZ4_DIR/lib \ -DOPENSSL_FOUND=1 \ -DOPENSSL_ROOT_DIR=$OPENSSL_DIR \ -DOPENSSL_INCLUDE_DIR=$OPENSSL_DIR/include \ diff --git a/catchain/catchain-receiver.cpp b/catchain/catchain-receiver.cpp index c8206de9..82779e3b 100644 --- a/catchain/catchain-receiver.cpp +++ b/catchain/catchain-receiver.cpp @@ -520,7 +520,8 @@ void CatChainReceiverImpl::start_up() { } td::actor::send_closure(overlay_manager_, &overlay::Overlays::create_private_overlay, get_source(local_idx_)->get_adnl_id(), overlay_full_id_.clone(), std::move(ids), - make_callback(), overlay::OverlayPrivacyRules{0, 0, std::move(root_keys)}); + make_callback(), overlay::OverlayPrivacyRules{0, 0, std::move(root_keys)}, + R"({ "type": "catchain" })"); CHECK(root_block_); diff --git a/common/global-version.h b/common/global-version.h index 0a90ab85..a6775ffa 100644 --- a/common/global-version.h +++ b/common/global-version.h @@ -19,6 +19,6 @@ namespace ton { // See doc/GlobalVersions.md -const int SUPPORTED_VERSION = 6; +const int SUPPORTED_VERSION = 7; } diff --git a/create-hardfork/create-hardfork.cpp b/create-hardfork/create-hardfork.cpp index 5f0b93be..c5f1add8 100644 --- a/create-hardfork/create-hardfork.cpp +++ b/create-hardfork/create-hardfork.cpp @@ -213,7 +213,7 @@ class HardforkCreator : public td::actor::Actor { ton::validator::ValidatorManagerHardforkFactory::create(opts, shard_, shard_top_block_id_, db_root_); for (auto &msg : ext_msgs_) { td::actor::send_closure(validator_manager_, &ton::validator::ValidatorManager::new_external_message, - std::move(msg)); + std::move(msg), 0); } for (auto &topmsg : top_shard_descrs_) { td::actor::send_closure(validator_manager_, &ton::validator::ValidatorManager::new_shard_block, ton::BlockIdExt{}, diff --git a/crypto/block/transaction.cpp b/crypto/block/transaction.cpp index 88966179..24013a40 100644 --- a/crypto/block/transaction.cpp +++ b/crypto/block/transaction.cpp @@ -898,6 +898,9 @@ bool Transaction::prepare_storage_phase(const StoragePhaseConfig& cfg, bool forc res->fees_collected = to_pay; res->fees_due = td::zero_refint(); balance -= std::move(to_pay); + if (cfg.global_version >= 7) { + due_payment = td::zero_refint(); + } } else if (acc_status == Account::acc_frozen && !force_collect && to_pay < cfg.delete_due_limit) { // do not collect fee res->last_paid_updated = (res->is_special ? 0 : account.last_paid); @@ -1318,7 +1321,7 @@ Ref Transaction::prepare_vm_c7(const ComputePhaseConfig& cfg) const { vm::StackEntry::maybe(cfg.global_config) // global_config:(Maybe Cell) ] = SmartContractInfo; }; if (cfg.global_version >= 4) { - tuple.push_back(new_code); // code:Cell + tuple.push_back(vm::StackEntry::maybe(new_code)); // code:Cell if (msg_balance_remaining.is_valid()) { tuple.push_back(msg_balance_remaining.as_vm_tuple()); // in_msg_value:[Integer (Maybe Cell)] } else { @@ -1335,11 +1338,10 @@ Ref Transaction::prepare_vm_c7(const ComputePhaseConfig& cfg) const { // Inside validator, collator and liteserver checking external message contexts // prev_blocks_info is always not null, since get_prev_blocks_info() // may only return tuple or raise Error (See crypto/block/mc-config.cpp#2223) - tuple.push_back(cfg.prev_blocks_info.not_null() ? vm::StackEntry(cfg.prev_blocks_info) : vm::StackEntry()); + tuple.push_back(vm::StackEntry::maybe(cfg.prev_blocks_info)); } if (cfg.global_version >= 6) { - tuple.push_back(cfg.unpacked_config_tuple.not_null() ? vm::StackEntry(cfg.unpacked_config_tuple) - : vm::StackEntry()); // unpacked_config_tuple:[...] + tuple.push_back(vm::StackEntry::maybe(cfg.unpacked_config_tuple)); // unpacked_config_tuple:[...] tuple.push_back(due_payment.not_null() ? due_payment : td::zero_refint()); // due_payment:Integer tuple.push_back(compute_phase->precompiled_gas_usage ? vm::StackEntry(td::make_refint(compute_phase->precompiled_gas_usage.value())) @@ -3668,6 +3670,7 @@ td::Status FetchConfigParams::fetch_config_params( compute_phase_cfg->mc_gas_prices = std::move(mc_gas_prices); compute_phase_cfg->special_gas_full = config.get_global_version() >= 5; storage_phase_cfg->enable_due_payment = config.get_global_version() >= 4; + storage_phase_cfg->global_version = config.get_global_version(); compute_phase_cfg->block_rand_seed = *rand_seed; compute_phase_cfg->max_vm_data_depth = size_limits.max_vm_data_depth; compute_phase_cfg->global_config = config.get_root_cell(); diff --git a/crypto/block/transaction.h b/crypto/block/transaction.h index 42cd84a8..6d8e8a29 100644 --- a/crypto/block/transaction.h +++ b/crypto/block/transaction.h @@ -82,6 +82,7 @@ struct StoragePhaseConfig { td::RefInt256 freeze_due_limit; td::RefInt256 delete_due_limit; bool enable_due_payment{false}; + int global_version = 0; StoragePhaseConfig() = default; StoragePhaseConfig(const std::vector* _pricing, td::RefInt256 freeze_limit = {}, td::RefInt256 delete_limit = {}) diff --git a/crypto/smc-envelope/SmartContract.cpp b/crypto/smc-envelope/SmartContract.cpp index 13e5c70b..2578a951 100644 --- a/crypto/smc-envelope/SmartContract.cpp +++ b/crypto/smc-envelope/SmartContract.cpp @@ -157,11 +157,11 @@ td::Ref prepare_vm_c7(SmartContract::Args args, td::Ref cod td::make_refint(0), //TODO: // trans_lt:Integer std::move(rand_seed_int), // rand_seed:Integer block::CurrencyCollection(args.balance).as_vm_tuple(), // balance_remaining:[Integer (Maybe Cell)] - vm::load_cell_slice_ref(address), // myself:MsgAddressInt - vm::StackEntry::maybe(config) //vm::StackEntry::maybe(td::Ref()) + vm::load_cell_slice_ref(address), // myself:MsgAddressInt + vm::StackEntry::maybe(config) // vm::StackEntry::maybe(td::Ref()) }; if (args.config && args.config.value()->get_global_version() >= 4) { - tuple.push_back(code.not_null() ? code : vm::StackEntry{}); // code:Cell + tuple.push_back(vm::StackEntry::maybe(code)); // code:Cell tuple.push_back(block::CurrencyCollection::zero().as_vm_tuple()); // in_msg_value:[Integer (Maybe Cell)] tuple.push_back(td::zero_refint()); // storage_fees:Integer @@ -174,7 +174,12 @@ td::Ref prepare_vm_c7(SmartContract::Args args, td::Ref cod if (args.config && args.config.value()->get_global_version() >= 6) { tuple.push_back(args.config.value()->get_unpacked_config_tuple(now)); // unpacked_config_tuple tuple.push_back(td::zero_refint()); // due_payment - tuple.push_back(vm::StackEntry()); // precompiled_gas_usage:Integer + // precomiled_gas_usage:(Maybe Integer) + td::optional precompiled; + if (code.not_null()) { + precompiled = args.config.value()->get_precompiled_contracts_config().get_contract(code->get_hash().bits()); + } + tuple.push_back(precompiled ? td::make_refint(precompiled.value().gas_usage) : vm::StackEntry()); } auto tuple_ref = td::make_cnt_ref>(std::move(tuple)); //LOG(DEBUG) << "SmartContractInfo initialized with " << vm::StackEntry(tuple).to_string(); diff --git a/crypto/smc-envelope/WalletInterface.cpp b/crypto/smc-envelope/WalletInterface.cpp index c76ac4a2..e02759b2 100644 --- a/crypto/smc-envelope/WalletInterface.cpp +++ b/crypto/smc-envelope/WalletInterface.cpp @@ -71,7 +71,7 @@ void WalletInterface::store_gift_message(vm::CellBuilder &cb, const Gift &gift) cb.store_zeroes(1); if (gift.is_encrypted) { - cb.store_long(0x2167da4b, 32); + cb.store_long(EncryptedCommentOp, 32); } else { cb.store_long(0, 32); } diff --git a/crypto/smc-envelope/WalletInterface.h b/crypto/smc-envelope/WalletInterface.h index b7f80630..c4e1f270 100644 --- a/crypto/smc-envelope/WalletInterface.h +++ b/crypto/smc-envelope/WalletInterface.h @@ -33,6 +33,7 @@ namespace ton { class WalletInterface : public SmartContract { public: + static constexpr uint32_t EncryptedCommentOp = 0x2167da4b; struct Gift { block::StdAddress destination; td::int64 gramms; diff --git a/doc/ConfigParam-HOWTO b/doc/ConfigParam-HOWTO index 8432657a..639014bb 100644 --- a/doc/ConfigParam-HOWTO +++ b/doc/ConfigParam-HOWTO @@ -13,7 +13,7 @@ All configuration parameters are combined into a *configuration dictionary* - a We see that, apart from the configuration dictionary, `ConfigParams` contains `config_addr` -- 256-bit address of the configuration smart contract in the masterchain. More details on the configuration smart contract will be provided later. -The configuration dictionary containing the active values of all configuration parameters is available via special TVM register *c7* to all smart contracts when their code is executed in a transaction. More precisely, when a smart contract is executed, *c7* is initialized by a Tuple, the only element of which is a Tuple with several "context" values useful for the execution of the smart contract, such as the current Unix time (as registered in the block header). The tenth entry of this Tuple (i.e., the one with zero-based index 9) contains a Cell representing the configuration dictionary. Therefore, it can be accesses by means of TVM instructions "PUSH c7; FIRST; INDEX 9", or by equivalent instruction "CONFIGROOT". In fact, special TVM instructions "CONFIGPARAM" and "CONFIGOPTPARAM" combine the previous actions with a dictionary lookup, returning any configuration parameter by its index. We refer to the TVM documentation for more details on these instructions. What is relevant here is that all configuration parameters are easily accessible from all smart contracts (masterchain or shardchain), and smart contracts may inspect them and use them to perform specific checks. For instance, a smart contract might extract workchain data storage prices from a configuration parameter to compute the price for storing a chunk of user-provided data. +The configuration dictionary containing the active values of all configuration parameters is available via special TVM register *c7* to all smart contracts when their code is executed in a transaction. More precisely, when a smart contract is executed, *c7* is initialized by a Tuple, the only element of which is a Tuple with several "context" values useful for the execution of the smart contract, such as the current Unix time (as registered in the block header). The tenth entry of this Tuple (i.e., the one with zero-based index 9) contains a Cell representing the configuration dictionary. Therefore, it can be accessed by means of TVM instructions "PUSH c7; FIRST; INDEX 9", or by equivalent instruction "CONFIGROOT". In fact, special TVM instructions "CONFIGPARAM" and "CONFIGOPTPARAM" combine the previous actions with a dictionary lookup, returning any configuration parameter by its index. We refer to the TVM documentation for more details on these instructions. What is relevant here is that all configuration parameters are easily accessible from all smart contracts (masterchain or shardchain), and smart contracts may inspect them and use them to perform specific checks. For instance, a smart contract might extract workchain data storage prices from a configuration parameter to compute the price for storing a chunk of user-provided data. The values of configuration parameters are not arbitrary. In fact, if the configuration parameter index *i* is non-negative, then the value of this parameter must be a valid value of TL-B type (ConfigParam i). This restriction is enforced by the validators, which will not accept changes to configuration parameters with non-negative indices unless they are valid values of the corresponding TL-B type. diff --git a/doc/catchain-dos.md b/doc/catchain-dos.md index 028cb95a..d256b157 100644 --- a/doc/catchain-dos.md +++ b/doc/catchain-dos.md @@ -13,8 +13,8 @@ For catchain protocol version higher or equal to 2 hash covers catchain block de ### Catchain block size All catchain messages, except `REJECT` message, have (and had) a limited size. After update `REJECT` message will be limited to 1024 bytes. At the same time one block contains at most number of block candidates per round messages. That way, 16kb limit per catchain block should be enough to prevent DoS. ### Limiting block height -Another potential DoS is related to a situation when a malbehaviouring node sends too many catchain blocks. Note that limiting of maximal number of blocks per second is not a good solution since it will hinder synchronization after node disconnect. -At the same time, catchain groups exist for quite short period of time (around a few hunderd seconds), while number of blocks production is determined by "natural block production speed" on the one hand and number of blocks generated to decrease dependencies size on the other. In any case, total number of blocks is limited by `catchain_lifetime * natural_block_production_speed * (1 + number_of_catchain_participants / max_dependencies_size)`. +Another potential DoS is related to a situation when a misbehaviouring node sends too many catchain blocks. Note that limiting of maximal number of blocks per second is not a good solution since it will hinder synchronization after node disconnect. +At the same time, catchain groups exist for quite short period of time (around a few hundred seconds), while number of blocks production is determined by "natural block production speed" on the one hand and number of blocks generated to decrease dependencies size on the other. In any case, total number of blocks is limited by `catchain_lifetime * natural_block_production_speed * (1 + number_of_catchain_participants / max_dependencies_size)`. To prevent DoS attack we limit maximal height of the block which will be processed by node by `catchain_lifetime * natural_block_production_speed * (1 + number_of_catchain_participants / max_dependencies_size)`, where `catchain_lifetime` is set by `ConfigParam 28` (`CatchainConfig`), `natural_block_production_speed` and `max_dependencies_size` are set by `ConfigParam 29` (`ConsensusConfig`) (`natural_block_production_speed` is calculated as `catchain_max_blocks_coeff / 1000`) and `number_of_catchain_participants` is set from catchain group configuration. By default, `catchain_max_blocks_coeff` is set to zero: special value which means that there is no limitation on catchain block height. It is recommended to set `catchain_max_blocks_coeff` to `10000`: we expect that natural production rate is about one block per 3 seconds, so we set the coefficient to allow 30 times higher block production than expected. At the same time, this number is low enough to prevent resource-intensive attacks. diff --git a/example/android/build.sh b/example/android/build.sh index 06217255..cd75f9d7 100755 --- a/example/android/build.sh +++ b/example/android/build.sh @@ -4,6 +4,7 @@ pushd . SECP256K1_INCLUDE_DIR=$(pwd)/third_party/secp256k1/include OPENSSL_DIR=$(pwd)/third_party/crypto/ +LZ4_INCLUDE_DIR=$(pwd)/third_party/lz4/include if [ $ARCH == "arm" ] then @@ -12,6 +13,7 @@ then SODIUM_LIBRARY_RELEASE=$(pwd)/third_party/libsodium/libsodium-android-armv7-a/lib/libsodium.a SECP256K1_LIBRARY=$(pwd)/third_party/secp256k1/armv7/libsecp256k1.a BLST_LIBRARY=$(pwd)/third_party/blst/armv7/libblst.a + LZ4_LIBRARY=$(pwd)/third_party/lz4/armv7/liblz4.a elif [ $ARCH == "x86" ] then ABI=$ARCH @@ -19,6 +21,7 @@ then SODIUM_LIBRARY_RELEASE=$(pwd)/third_party/libsodium/libsodium-android-i686/lib/libsodium.a SECP256K1_LIBRARY=$(pwd)/third_party/secp256k1/i686/libsecp256k1.a BLST_LIBRARY=$(pwd)/third_party/blst/i686/libblst.a + LZ4_LIBRARY=$(pwd)/third_party/lz4/i686/liblz4.a TARGET=i686-linux-android21 elif [ $ARCH == "x86_64" ] then @@ -27,6 +30,7 @@ then SODIUM_LIBRARY_RELEASE=$(pwd)/third_party/libsodium/libsodium-android-westmere/lib/libsodium.a SECP256K1_LIBRARY=$(pwd)/third_party/secp256k1/x86-64/libsecp256k1.a BLST_LIBRARY=$(pwd)/third_party/blst/x86-64/libblst.a + LZ4_LIBRARY=$(pwd)/third_party/lz4/x86-64/liblz4.a elif [ $ARCH == "arm64" ] then ABI="arm64-v8a" @@ -34,6 +38,7 @@ then SODIUM_LIBRARY_RELEASE=$(pwd)/third_party/libsodium/libsodium-android-armv8-a/lib/libsodium.a SECP256K1_LIBRARY=$(pwd)/third_party/secp256k1/armv8/libsecp256k1.a BLST_LIBRARY=$(pwd)/third_party/blst/armv8/libblst.a + LZ4_LIBRARY=$(pwd)/third_party/lz4/armv8/liblz4.a fi ORIG_ARCH=$ARCH @@ -56,6 +61,9 @@ cmake .. -GNinja \ -DSECP256K1_FOUND=1 \ -DSECP256K1_INCLUDE_DIR=${SECP256K1_INCLUDE_DIR} \ -DSECP256K1_LIBRARY=${SECP256K1_LIBRARY} \ +-DLZ4_FOUND=1 \ +-DLZ4_INCLUDE_DIRS=${LZ4_INCLUDE_DIR} \ +-DLZ4_LIBRARIES=${LZ4_LIBRARY} \ -DSODIUM_INCLUDE_DIR=${SODIUM_INCLUDE_DIR} \ -DSODIUM_LIBRARY_RELEASE=${SODIUM_LIBRARY_RELEASE} \ -DSODIUM_USE_STATIC_LIBS=1 \ diff --git a/example/android/third_party/lz4/armv7/liblz4.a b/example/android/third_party/lz4/armv7/liblz4.a new file mode 100644 index 00000000..311fbca4 Binary files /dev/null and b/example/android/third_party/lz4/armv7/liblz4.a differ diff --git a/example/android/third_party/lz4/armv7/liblz4.so b/example/android/third_party/lz4/armv7/liblz4.so new file mode 100644 index 00000000..f98cffe8 Binary files /dev/null and b/example/android/third_party/lz4/armv7/liblz4.so differ diff --git a/example/android/third_party/lz4/armv8/liblz4.a b/example/android/third_party/lz4/armv8/liblz4.a new file mode 100644 index 00000000..6db4a5d4 Binary files /dev/null and b/example/android/third_party/lz4/armv8/liblz4.a differ diff --git a/example/android/third_party/lz4/armv8/liblz4.so b/example/android/third_party/lz4/armv8/liblz4.so new file mode 100644 index 00000000..f22c7945 Binary files /dev/null and b/example/android/third_party/lz4/armv8/liblz4.so differ diff --git a/example/android/third_party/lz4/i686/liblz4.a b/example/android/third_party/lz4/i686/liblz4.a new file mode 100644 index 00000000..2d86f7ef Binary files /dev/null and b/example/android/third_party/lz4/i686/liblz4.a differ diff --git a/example/android/third_party/lz4/i686/liblz4.so b/example/android/third_party/lz4/i686/liblz4.so new file mode 100644 index 00000000..bed09629 Binary files /dev/null and b/example/android/third_party/lz4/i686/liblz4.so differ diff --git a/example/android/third_party/lz4/include/lz4.c b/example/android/third_party/lz4/include/lz4.c new file mode 100644 index 00000000..b7dd32a7 --- /dev/null +++ b/example/android/third_party/lz4/include/lz4.c @@ -0,0 +1,2826 @@ +/* + LZ4 - Fast LZ compression algorithm + Copyright (C) 2011-2023, Yann Collet. + + BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) + + Redistribution and use in source and binary forms, with or without + modification, are permitted provided that the following conditions are + met: + + * Redistributions of source code must retain the above copyright + notice, this list of conditions and the following disclaimer. + * Redistributions in binary form must reproduce the above + copyright notice, this list of conditions and the following disclaimer + in the documentation and/or other materials provided with the + distribution. + + THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + + You can contact the author at : + - LZ4 homepage : http://www.lz4.org + - LZ4 source repository : https://github.com/lz4/lz4 +*/ + +/*-************************************ +* Tuning parameters +**************************************/ +/* + * LZ4_HEAPMODE : + * Select how stateless compression functions like `LZ4_compress_default()` + * allocate memory for their hash table, + * in memory stack (0:default, fastest), or in memory heap (1:requires malloc()). + */ +#ifndef LZ4_HEAPMODE +# define LZ4_HEAPMODE 0 +#endif + +/* + * LZ4_ACCELERATION_DEFAULT : + * Select "acceleration" for LZ4_compress_fast() when parameter value <= 0 + */ +#define LZ4_ACCELERATION_DEFAULT 1 +/* + * LZ4_ACCELERATION_MAX : + * Any "acceleration" value higher than this threshold + * get treated as LZ4_ACCELERATION_MAX instead (fix #876) + */ +#define LZ4_ACCELERATION_MAX 65537 + + +/*-************************************ +* CPU Feature Detection +**************************************/ +/* LZ4_FORCE_MEMORY_ACCESS + * By default, access to unaligned memory is controlled by `memcpy()`, which is safe and portable. + * Unfortunately, on some target/compiler combinations, the generated assembly is sub-optimal. + * The below switch allow to select different access method for improved performance. + * Method 0 (default) : use `memcpy()`. Safe and portable. + * Method 1 : `__packed` statement. It depends on compiler extension (ie, not portable). + * This method is safe if your compiler supports it, and *generally* as fast or faster than `memcpy`. + * Method 2 : direct access. This method is portable but violate C standard. + * It can generate buggy code on targets which assembly generation depends on alignment. + * But in some circumstances, it's the only known way to get the most performance (ie GCC + ARMv6) + * See https://fastcompression.blogspot.fr/2015/08/accessing-unaligned-memory.html for details. + * Prefer these methods in priority order (0 > 1 > 2) + */ +#ifndef LZ4_FORCE_MEMORY_ACCESS /* can be defined externally */ +# if defined(__GNUC__) && \ + ( defined(__ARM_ARCH_6__) || defined(__ARM_ARCH_6J__) || defined(__ARM_ARCH_6K__) \ + || defined(__ARM_ARCH_6Z__) || defined(__ARM_ARCH_6ZK__) || defined(__ARM_ARCH_6T2__) ) +# define LZ4_FORCE_MEMORY_ACCESS 2 +# elif (defined(__INTEL_COMPILER) && !defined(_WIN32)) || defined(__GNUC__) || defined(_MSC_VER) +# define LZ4_FORCE_MEMORY_ACCESS 1 +# endif +#endif + +/* + * LZ4_FORCE_SW_BITCOUNT + * Define this parameter if your target system or compiler does not support hardware bit count + */ +#if defined(_MSC_VER) && defined(_WIN32_WCE) /* Visual Studio for WinCE doesn't support Hardware bit count */ +# undef LZ4_FORCE_SW_BITCOUNT /* avoid double def */ +# define LZ4_FORCE_SW_BITCOUNT +#endif + + + +/*-************************************ +* Dependency +**************************************/ +/* + * LZ4_SRC_INCLUDED: + * Amalgamation flag, whether lz4.c is included + */ +#ifndef LZ4_SRC_INCLUDED +# define LZ4_SRC_INCLUDED 1 +#endif + +#ifndef LZ4_DISABLE_DEPRECATE_WARNINGS +# define LZ4_DISABLE_DEPRECATE_WARNINGS /* due to LZ4_decompress_safe_withPrefix64k */ +#endif + +#ifndef LZ4_STATIC_LINKING_ONLY +# define LZ4_STATIC_LINKING_ONLY +#endif +#include "lz4.h" +/* see also "memory routines" below */ + + +/*-************************************ +* Compiler Options +**************************************/ +#if defined(_MSC_VER) && (_MSC_VER >= 1400) /* Visual Studio 2005+ */ +# include /* only present in VS2005+ */ +# pragma warning(disable : 4127) /* disable: C4127: conditional expression is constant */ +# pragma warning(disable : 6237) /* disable: C6237: conditional expression is always 0 */ +#endif /* _MSC_VER */ + +#ifndef LZ4_FORCE_INLINE +# if defined (_MSC_VER) && !defined (__clang__) /* MSVC */ +# define LZ4_FORCE_INLINE static __forceinline +# else +# if defined (__cplusplus) || defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L /* C99 */ +# if defined (__GNUC__) || defined (__clang__) +# define LZ4_FORCE_INLINE static inline __attribute__((always_inline)) +# else +# define LZ4_FORCE_INLINE static inline +# endif +# else +# define LZ4_FORCE_INLINE static +# endif /* __STDC_VERSION__ */ +# endif /* _MSC_VER */ +#endif /* LZ4_FORCE_INLINE */ + +/* LZ4_FORCE_O2 and LZ4_FORCE_INLINE + * gcc on ppc64le generates an unrolled SIMDized loop for LZ4_wildCopy8, + * together with a simple 8-byte copy loop as a fall-back path. + * However, this optimization hurts the decompression speed by >30%, + * because the execution does not go to the optimized loop + * for typical compressible data, and all of the preamble checks + * before going to the fall-back path become useless overhead. + * This optimization happens only with the -O3 flag, and -O2 generates + * a simple 8-byte copy loop. + * With gcc on ppc64le, all of the LZ4_decompress_* and LZ4_wildCopy8 + * functions are annotated with __attribute__((optimize("O2"))), + * and also LZ4_wildCopy8 is forcibly inlined, so that the O2 attribute + * of LZ4_wildCopy8 does not affect the compression speed. + */ +#if defined(__PPC64__) && defined(__LITTLE_ENDIAN__) && defined(__GNUC__) && !defined(__clang__) +# define LZ4_FORCE_O2 __attribute__((optimize("O2"))) +# undef LZ4_FORCE_INLINE +# define LZ4_FORCE_INLINE static __inline __attribute__((optimize("O2"),always_inline)) +#else +# define LZ4_FORCE_O2 +#endif + +#if (defined(__GNUC__) && (__GNUC__ >= 3)) || (defined(__INTEL_COMPILER) && (__INTEL_COMPILER >= 800)) || defined(__clang__) +# define expect(expr,value) (__builtin_expect ((expr),(value)) ) +#else +# define expect(expr,value) (expr) +#endif + +#ifndef likely +#define likely(expr) expect((expr) != 0, 1) +#endif +#ifndef unlikely +#define unlikely(expr) expect((expr) != 0, 0) +#endif + +/* Should the alignment test prove unreliable, for some reason, + * it can be disabled by setting LZ4_ALIGN_TEST to 0 */ +#ifndef LZ4_ALIGN_TEST /* can be externally provided */ +# define LZ4_ALIGN_TEST 1 +#endif + + +/*-************************************ +* Memory routines +**************************************/ + +/*! LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION : + * Disable relatively high-level LZ4/HC functions that use dynamic memory + * allocation functions (malloc(), calloc(), free()). + * + * Note that this is a compile-time switch. And since it disables + * public/stable LZ4 v1 API functions, we don't recommend using this + * symbol to generate a library for distribution. + * + * The following public functions are removed when this symbol is defined. + * - lz4 : LZ4_createStream, LZ4_freeStream, + * LZ4_createStreamDecode, LZ4_freeStreamDecode, LZ4_create (deprecated) + * - lz4hc : LZ4_createStreamHC, LZ4_freeStreamHC, + * LZ4_createHC (deprecated), LZ4_freeHC (deprecated) + * - lz4frame, lz4file : All LZ4F_* functions + */ +#if defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION) +# define ALLOC(s) lz4_error_memory_allocation_is_disabled +# define ALLOC_AND_ZERO(s) lz4_error_memory_allocation_is_disabled +# define FREEMEM(p) lz4_error_memory_allocation_is_disabled +#elif defined(LZ4_USER_MEMORY_FUNCTIONS) +/* memory management functions can be customized by user project. + * Below functions must exist somewhere in the Project + * and be available at link time */ +void* LZ4_malloc(size_t s); +void* LZ4_calloc(size_t n, size_t s); +void LZ4_free(void* p); +# define ALLOC(s) LZ4_malloc(s) +# define ALLOC_AND_ZERO(s) LZ4_calloc(1,s) +# define FREEMEM(p) LZ4_free(p) +#else +# include /* malloc, calloc, free */ +# define ALLOC(s) malloc(s) +# define ALLOC_AND_ZERO(s) calloc(1,s) +# define FREEMEM(p) free(p) +#endif + +#if ! LZ4_FREESTANDING +# include /* memset, memcpy */ +#endif +#if !defined(LZ4_memset) +# define LZ4_memset(p,v,s) memset((p),(v),(s)) +#endif +#define MEM_INIT(p,v,s) LZ4_memset((p),(v),(s)) + + +/*-************************************ +* Common Constants +**************************************/ +#define MINMATCH 4 + +#define WILDCOPYLENGTH 8 +#define LASTLITERALS 5 /* see ../doc/lz4_Block_format.md#parsing-restrictions */ +#define MFLIMIT 12 /* see ../doc/lz4_Block_format.md#parsing-restrictions */ +#define MATCH_SAFEGUARD_DISTANCE ((2*WILDCOPYLENGTH) - MINMATCH) /* ensure it's possible to write 2 x wildcopyLength without overflowing output buffer */ +#define FASTLOOP_SAFE_DISTANCE 64 +static const int LZ4_minLength = (MFLIMIT+1); + +#define KB *(1 <<10) +#define MB *(1 <<20) +#define GB *(1U<<30) + +#define LZ4_DISTANCE_ABSOLUTE_MAX 65535 +#if (LZ4_DISTANCE_MAX > LZ4_DISTANCE_ABSOLUTE_MAX) /* max supported by LZ4 format */ +# error "LZ4_DISTANCE_MAX is too big : must be <= 65535" +#endif + +#define ML_BITS 4 +#define ML_MASK ((1U<=1) +# include +#else +# ifndef assert +# define assert(condition) ((void)0) +# endif +#endif + +#define LZ4_STATIC_ASSERT(c) { enum { LZ4_static_assert = 1/(int)(!!(c)) }; } /* use after variable declarations */ + +#if defined(LZ4_DEBUG) && (LZ4_DEBUG>=2) +# include + static int g_debuglog_enable = 1; +# define DEBUGLOG(l, ...) { \ + if ((g_debuglog_enable) && (l<=LZ4_DEBUG)) { \ + fprintf(stderr, __FILE__ " %i: ", __LINE__); \ + fprintf(stderr, __VA_ARGS__); \ + fprintf(stderr, " \n"); \ + } } +#else +# define DEBUGLOG(l, ...) {} /* disabled */ +#endif + +static int LZ4_isAligned(const void* ptr, size_t alignment) +{ + return ((size_t)ptr & (alignment -1)) == 0; +} + + +/*-************************************ +* Types +**************************************/ +#include +#if defined(__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) +# include + typedef uint8_t BYTE; + typedef uint16_t U16; + typedef uint32_t U32; + typedef int32_t S32; + typedef uint64_t U64; + typedef uintptr_t uptrval; +#else +# if UINT_MAX != 4294967295UL +# error "LZ4 code (when not C++ or C99) assumes that sizeof(int) == 4" +# endif + typedef unsigned char BYTE; + typedef unsigned short U16; + typedef unsigned int U32; + typedef signed int S32; + typedef unsigned long long U64; + typedef size_t uptrval; /* generally true, except OpenVMS-64 */ +#endif + +#if defined(__x86_64__) + typedef U64 reg_t; /* 64-bits in x32 mode */ +#else + typedef size_t reg_t; /* 32-bits in x32 mode */ +#endif + +typedef enum { + notLimited = 0, + limitedOutput = 1, + fillOutput = 2 +} limitedOutput_directive; + + +/*-************************************ +* Reading and writing into memory +**************************************/ + +/** + * LZ4 relies on memcpy with a constant size being inlined. In freestanding + * environments, the compiler can't assume the implementation of memcpy() is + * standard compliant, so it can't apply its specialized memcpy() inlining + * logic. When possible, use __builtin_memcpy() to tell the compiler to analyze + * memcpy() as if it were standard compliant, so it can inline it in freestanding + * environments. This is needed when decompressing the Linux Kernel, for example. + */ +#if !defined(LZ4_memcpy) +# if defined(__GNUC__) && (__GNUC__ >= 4) +# define LZ4_memcpy(dst, src, size) __builtin_memcpy(dst, src, size) +# else +# define LZ4_memcpy(dst, src, size) memcpy(dst, src, size) +# endif +#endif + +#if !defined(LZ4_memmove) +# if defined(__GNUC__) && (__GNUC__ >= 4) +# define LZ4_memmove __builtin_memmove +# else +# define LZ4_memmove memmove +# endif +#endif + +static unsigned LZ4_isLittleEndian(void) +{ + const union { U32 u; BYTE c[4]; } one = { 1 }; /* don't use static : performance detrimental */ + return one.c[0]; +} + +#if defined(__GNUC__) || defined(__INTEL_COMPILER) +#define LZ4_PACK( __Declaration__ ) __Declaration__ __attribute__((__packed__)) +#elif defined(_MSC_VER) +#define LZ4_PACK( __Declaration__ ) __pragma( pack(push, 1) ) __Declaration__ __pragma( pack(pop)) +#endif + +#if defined(LZ4_FORCE_MEMORY_ACCESS) && (LZ4_FORCE_MEMORY_ACCESS==2) +/* lie to the compiler about data alignment; use with caution */ + +static U16 LZ4_read16(const void* memPtr) { return *(const U16*) memPtr; } +static U32 LZ4_read32(const void* memPtr) { return *(const U32*) memPtr; } +static reg_t LZ4_read_ARCH(const void* memPtr) { return *(const reg_t*) memPtr; } + +static void LZ4_write16(void* memPtr, U16 value) { *(U16*)memPtr = value; } +static void LZ4_write32(void* memPtr, U32 value) { *(U32*)memPtr = value; } + +#elif defined(LZ4_FORCE_MEMORY_ACCESS) && (LZ4_FORCE_MEMORY_ACCESS==1) + +/* __pack instructions are safer, but compiler specific, hence potentially problematic for some compilers */ +/* currently only defined for gcc and icc */ +LZ4_PACK(typedef struct { U16 u16; }) LZ4_unalign16; +LZ4_PACK(typedef struct { U32 u32; }) LZ4_unalign32; +LZ4_PACK(typedef struct { reg_t uArch; }) LZ4_unalignST; + +static U16 LZ4_read16(const void* ptr) { return ((const LZ4_unalign16*)ptr)->u16; } +static U32 LZ4_read32(const void* ptr) { return ((const LZ4_unalign32*)ptr)->u32; } +static reg_t LZ4_read_ARCH(const void* ptr) { return ((const LZ4_unalignST*)ptr)->uArch; } + +static void LZ4_write16(void* memPtr, U16 value) { ((LZ4_unalign16*)memPtr)->u16 = value; } +static void LZ4_write32(void* memPtr, U32 value) { ((LZ4_unalign32*)memPtr)->u32 = value; } + +#else /* safe and portable access using memcpy() */ + +static U16 LZ4_read16(const void* memPtr) +{ + U16 val; LZ4_memcpy(&val, memPtr, sizeof(val)); return val; +} + +static U32 LZ4_read32(const void* memPtr) +{ + U32 val; LZ4_memcpy(&val, memPtr, sizeof(val)); return val; +} + +static reg_t LZ4_read_ARCH(const void* memPtr) +{ + reg_t val; LZ4_memcpy(&val, memPtr, sizeof(val)); return val; +} + +static void LZ4_write16(void* memPtr, U16 value) +{ + LZ4_memcpy(memPtr, &value, sizeof(value)); +} + +static void LZ4_write32(void* memPtr, U32 value) +{ + LZ4_memcpy(memPtr, &value, sizeof(value)); +} + +#endif /* LZ4_FORCE_MEMORY_ACCESS */ + + +static U16 LZ4_readLE16(const void* memPtr) +{ + if (LZ4_isLittleEndian()) { + return LZ4_read16(memPtr); + } else { + const BYTE* p = (const BYTE*)memPtr; + return (U16)((U16)p[0] + (p[1]<<8)); + } +} + +#ifdef LZ4_STATIC_LINKING_ONLY_ENDIANNESS_INDEPENDENT_OUTPUT +static U32 LZ4_readLE32(const void* memPtr) +{ + if (LZ4_isLittleEndian()) { + return LZ4_read32(memPtr); + } else { + const BYTE* p = (const BYTE*)memPtr; + return (U32)p[0] + (p[1]<<8) + (p[2]<<16) + (p[3]<<24); + } +} +#endif + +static void LZ4_writeLE16(void* memPtr, U16 value) +{ + if (LZ4_isLittleEndian()) { + LZ4_write16(memPtr, value); + } else { + BYTE* p = (BYTE*)memPtr; + p[0] = (BYTE) value; + p[1] = (BYTE)(value>>8); + } +} + +/* customized variant of memcpy, which can overwrite up to 8 bytes beyond dstEnd */ +LZ4_FORCE_INLINE +void LZ4_wildCopy8(void* dstPtr, const void* srcPtr, void* dstEnd) +{ + BYTE* d = (BYTE*)dstPtr; + const BYTE* s = (const BYTE*)srcPtr; + BYTE* const e = (BYTE*)dstEnd; + + do { LZ4_memcpy(d,s,8); d+=8; s+=8; } while (d= 16. */ +LZ4_FORCE_INLINE void +LZ4_wildCopy32(void* dstPtr, const void* srcPtr, void* dstEnd) +{ + BYTE* d = (BYTE*)dstPtr; + const BYTE* s = (const BYTE*)srcPtr; + BYTE* const e = (BYTE*)dstEnd; + + do { LZ4_memcpy(d,s,16); LZ4_memcpy(d+16,s+16,16); d+=32; s+=32; } while (d= dstPtr + MINMATCH + * - there is at least 8 bytes available to write after dstEnd */ +LZ4_FORCE_INLINE void +LZ4_memcpy_using_offset(BYTE* dstPtr, const BYTE* srcPtr, BYTE* dstEnd, const size_t offset) +{ + BYTE v[8]; + + assert(dstEnd >= dstPtr + MINMATCH); + + switch(offset) { + case 1: + MEM_INIT(v, *srcPtr, 8); + break; + case 2: + LZ4_memcpy(v, srcPtr, 2); + LZ4_memcpy(&v[2], srcPtr, 2); +#if defined(_MSC_VER) && (_MSC_VER <= 1937) /* MSVC 2022 ver 17.7 or earlier */ +# pragma warning(push) +# pragma warning(disable : 6385) /* warning C6385: Reading invalid data from 'v'. */ +#endif + LZ4_memcpy(&v[4], v, 4); +#if defined(_MSC_VER) && (_MSC_VER <= 1937) /* MSVC 2022 ver 17.7 or earlier */ +# pragma warning(pop) +#endif + break; + case 4: + LZ4_memcpy(v, srcPtr, 4); + LZ4_memcpy(&v[4], srcPtr, 4); + break; + default: + LZ4_memcpy_using_offset_base(dstPtr, srcPtr, dstEnd, offset); + return; + } + + LZ4_memcpy(dstPtr, v, 8); + dstPtr += 8; + while (dstPtr < dstEnd) { + LZ4_memcpy(dstPtr, v, 8); + dstPtr += 8; + } +} +#endif + + +/*-************************************ +* Common functions +**************************************/ +static unsigned LZ4_NbCommonBytes (reg_t val) +{ + assert(val != 0); + if (LZ4_isLittleEndian()) { + if (sizeof(val) == 8) { +# if defined(_MSC_VER) && (_MSC_VER >= 1800) && (defined(_M_AMD64) && !defined(_M_ARM64EC)) && !defined(LZ4_FORCE_SW_BITCOUNT) +/*-************************************************************************************************* +* ARM64EC is a Microsoft-designed ARM64 ABI compatible with AMD64 applications on ARM64 Windows 11. +* The ARM64EC ABI does not support AVX/AVX2/AVX512 instructions, nor their relevant intrinsics +* including _tzcnt_u64. Therefore, we need to neuter the _tzcnt_u64 code path for ARM64EC. +****************************************************************************************************/ +# if defined(__clang__) && (__clang_major__ < 10) + /* Avoid undefined clang-cl intrinsics issue. + * See https://github.com/lz4/lz4/pull/1017 for details. */ + return (unsigned)__builtin_ia32_tzcnt_u64(val) >> 3; +# else + /* x64 CPUS without BMI support interpret `TZCNT` as `REP BSF` */ + return (unsigned)_tzcnt_u64(val) >> 3; +# endif +# elif defined(_MSC_VER) && defined(_WIN64) && !defined(LZ4_FORCE_SW_BITCOUNT) + unsigned long r = 0; + _BitScanForward64(&r, (U64)val); + return (unsigned)r >> 3; +# elif (defined(__clang__) || (defined(__GNUC__) && ((__GNUC__ > 3) || \ + ((__GNUC__ == 3) && (__GNUC_MINOR__ >= 4))))) && \ + !defined(LZ4_FORCE_SW_BITCOUNT) + return (unsigned)__builtin_ctzll((U64)val) >> 3; +# else + const U64 m = 0x0101010101010101ULL; + val ^= val - 1; + return (unsigned)(((U64)((val & (m - 1)) * m)) >> 56); +# endif + } else /* 32 bits */ { +# if defined(_MSC_VER) && (_MSC_VER >= 1400) && !defined(LZ4_FORCE_SW_BITCOUNT) + unsigned long r; + _BitScanForward(&r, (U32)val); + return (unsigned)r >> 3; +# elif (defined(__clang__) || (defined(__GNUC__) && ((__GNUC__ > 3) || \ + ((__GNUC__ == 3) && (__GNUC_MINOR__ >= 4))))) && \ + !defined(__TINYC__) && !defined(LZ4_FORCE_SW_BITCOUNT) + return (unsigned)__builtin_ctz((U32)val) >> 3; +# else + const U32 m = 0x01010101; + return (unsigned)((((val - 1) ^ val) & (m - 1)) * m) >> 24; +# endif + } + } else /* Big Endian CPU */ { + if (sizeof(val)==8) { +# if (defined(__clang__) || (defined(__GNUC__) && ((__GNUC__ > 3) || \ + ((__GNUC__ == 3) && (__GNUC_MINOR__ >= 4))))) && \ + !defined(__TINYC__) && !defined(LZ4_FORCE_SW_BITCOUNT) + return (unsigned)__builtin_clzll((U64)val) >> 3; +# else +#if 1 + /* this method is probably faster, + * but adds a 128 bytes lookup table */ + static const unsigned char ctz7_tab[128] = { + 7, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0, + 4, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0, + 5, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0, + 4, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0, + 6, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0, + 4, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0, + 5, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0, + 4, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0, + }; + U64 const mask = 0x0101010101010101ULL; + U64 const t = (((val >> 8) - mask) | val) & mask; + return ctz7_tab[(t * 0x0080402010080402ULL) >> 57]; +#else + /* this method doesn't consume memory space like the previous one, + * but it contains several branches, + * that may end up slowing execution */ + static const U32 by32 = sizeof(val)*4; /* 32 on 64 bits (goal), 16 on 32 bits. + Just to avoid some static analyzer complaining about shift by 32 on 32-bits target. + Note that this code path is never triggered in 32-bits mode. */ + unsigned r; + if (!(val>>by32)) { r=4; } else { r=0; val>>=by32; } + if (!(val>>16)) { r+=2; val>>=8; } else { val>>=24; } + r += (!val); + return r; +#endif +# endif + } else /* 32 bits */ { +# if (defined(__clang__) || (defined(__GNUC__) && ((__GNUC__ > 3) || \ + ((__GNUC__ == 3) && (__GNUC_MINOR__ >= 4))))) && \ + !defined(LZ4_FORCE_SW_BITCOUNT) + return (unsigned)__builtin_clz((U32)val) >> 3; +# else + val >>= 8; + val = ((((val + 0x00FFFF00) | 0x00FFFFFF) + val) | + (val + 0x00FF0000)) >> 24; + return (unsigned)val ^ 3; +# endif + } + } +} + + +#define STEPSIZE sizeof(reg_t) +LZ4_FORCE_INLINE +unsigned LZ4_count(const BYTE* pIn, const BYTE* pMatch, const BYTE* pInLimit) +{ + const BYTE* const pStart = pIn; + + if (likely(pIn < pInLimit-(STEPSIZE-1))) { + reg_t const diff = LZ4_read_ARCH(pMatch) ^ LZ4_read_ARCH(pIn); + if (!diff) { + pIn+=STEPSIZE; pMatch+=STEPSIZE; + } else { + return LZ4_NbCommonBytes(diff); + } } + + while (likely(pIn < pInLimit-(STEPSIZE-1))) { + reg_t const diff = LZ4_read_ARCH(pMatch) ^ LZ4_read_ARCH(pIn); + if (!diff) { pIn+=STEPSIZE; pMatch+=STEPSIZE; continue; } + pIn += LZ4_NbCommonBytes(diff); + return (unsigned)(pIn - pStart); + } + + if ((STEPSIZE==8) && (pIn<(pInLimit-3)) && (LZ4_read32(pMatch) == LZ4_read32(pIn))) { pIn+=4; pMatch+=4; } + if ((pIn<(pInLimit-1)) && (LZ4_read16(pMatch) == LZ4_read16(pIn))) { pIn+=2; pMatch+=2; } + if ((pIn compression run slower on incompressible data */ + + +/*-************************************ +* Local Structures and types +**************************************/ +typedef enum { clearedTable = 0, byPtr, byU32, byU16 } tableType_t; + +/** + * This enum distinguishes several different modes of accessing previous + * content in the stream. + * + * - noDict : There is no preceding content. + * - withPrefix64k : Table entries up to ctx->dictSize before the current blob + * blob being compressed are valid and refer to the preceding + * content (of length ctx->dictSize), which is available + * contiguously preceding in memory the content currently + * being compressed. + * - usingExtDict : Like withPrefix64k, but the preceding content is somewhere + * else in memory, starting at ctx->dictionary with length + * ctx->dictSize. + * - usingDictCtx : Everything concerning the preceding content is + * in a separate context, pointed to by ctx->dictCtx. + * ctx->dictionary, ctx->dictSize, and table entries + * in the current context that refer to positions + * preceding the beginning of the current compression are + * ignored. Instead, ctx->dictCtx->dictionary and ctx->dictCtx + * ->dictSize describe the location and size of the preceding + * content, and matches are found by looking in the ctx + * ->dictCtx->hashTable. + */ +typedef enum { noDict = 0, withPrefix64k, usingExtDict, usingDictCtx } dict_directive; +typedef enum { noDictIssue = 0, dictSmall } dictIssue_directive; + + +/*-************************************ +* Local Utils +**************************************/ +int LZ4_versionNumber (void) { return LZ4_VERSION_NUMBER; } +const char* LZ4_versionString(void) { return LZ4_VERSION_STRING; } +int LZ4_compressBound(int isize) { return LZ4_COMPRESSBOUND(isize); } +int LZ4_sizeofState(void) { return sizeof(LZ4_stream_t); } + + +/*-**************************************** +* Internal Definitions, used only in Tests +*******************************************/ +#if defined (__cplusplus) +extern "C" { +#endif + +int LZ4_compress_forceExtDict (LZ4_stream_t* LZ4_dict, const char* source, char* dest, int srcSize); + +int LZ4_decompress_safe_forceExtDict(const char* source, char* dest, + int compressedSize, int maxOutputSize, + const void* dictStart, size_t dictSize); +int LZ4_decompress_safe_partial_forceExtDict(const char* source, char* dest, + int compressedSize, int targetOutputSize, int dstCapacity, + const void* dictStart, size_t dictSize); +#if defined (__cplusplus) +} +#endif + +/*-****************************** +* Compression functions +********************************/ +LZ4_FORCE_INLINE U32 LZ4_hash4(U32 sequence, tableType_t const tableType) +{ + if (tableType == byU16) + return ((sequence * 2654435761U) >> ((MINMATCH*8)-(LZ4_HASHLOG+1))); + else + return ((sequence * 2654435761U) >> ((MINMATCH*8)-LZ4_HASHLOG)); +} + +LZ4_FORCE_INLINE U32 LZ4_hash5(U64 sequence, tableType_t const tableType) +{ + const U32 hashLog = (tableType == byU16) ? LZ4_HASHLOG+1 : LZ4_HASHLOG; + if (LZ4_isLittleEndian()) { + const U64 prime5bytes = 889523592379ULL; + return (U32)(((sequence << 24) * prime5bytes) >> (64 - hashLog)); + } else { + const U64 prime8bytes = 11400714785074694791ULL; + return (U32)(((sequence >> 24) * prime8bytes) >> (64 - hashLog)); + } +} + +LZ4_FORCE_INLINE U32 LZ4_hashPosition(const void* const p, tableType_t const tableType) +{ + if ((sizeof(reg_t)==8) && (tableType != byU16)) return LZ4_hash5(LZ4_read_ARCH(p), tableType); + +#ifdef LZ4_STATIC_LINKING_ONLY_ENDIANNESS_INDEPENDENT_OUTPUT + return LZ4_hash4(LZ4_readLE32(p), tableType); +#else + return LZ4_hash4(LZ4_read32(p), tableType); +#endif +} + +LZ4_FORCE_INLINE void LZ4_clearHash(U32 h, void* tableBase, tableType_t const tableType) +{ + switch (tableType) + { + default: /* fallthrough */ + case clearedTable: { /* illegal! */ assert(0); return; } + case byPtr: { const BYTE** hashTable = (const BYTE**)tableBase; hashTable[h] = NULL; return; } + case byU32: { U32* hashTable = (U32*) tableBase; hashTable[h] = 0; return; } + case byU16: { U16* hashTable = (U16*) tableBase; hashTable[h] = 0; return; } + } +} + +LZ4_FORCE_INLINE void LZ4_putIndexOnHash(U32 idx, U32 h, void* tableBase, tableType_t const tableType) +{ + switch (tableType) + { + default: /* fallthrough */ + case clearedTable: /* fallthrough */ + case byPtr: { /* illegal! */ assert(0); return; } + case byU32: { U32* hashTable = (U32*) tableBase; hashTable[h] = idx; return; } + case byU16: { U16* hashTable = (U16*) tableBase; assert(idx < 65536); hashTable[h] = (U16)idx; return; } + } +} + +/* LZ4_putPosition*() : only used in byPtr mode */ +LZ4_FORCE_INLINE void LZ4_putPositionOnHash(const BYTE* p, U32 h, + void* tableBase, tableType_t const tableType) +{ + const BYTE** const hashTable = (const BYTE**)tableBase; + assert(tableType == byPtr); (void)tableType; + hashTable[h] = p; +} + +LZ4_FORCE_INLINE void LZ4_putPosition(const BYTE* p, void* tableBase, tableType_t tableType) +{ + U32 const h = LZ4_hashPosition(p, tableType); + LZ4_putPositionOnHash(p, h, tableBase, tableType); +} + +/* LZ4_getIndexOnHash() : + * Index of match position registered in hash table. + * hash position must be calculated by using base+index, or dictBase+index. + * Assumption 1 : only valid if tableType == byU32 or byU16. + * Assumption 2 : h is presumed valid (within limits of hash table) + */ +LZ4_FORCE_INLINE U32 LZ4_getIndexOnHash(U32 h, const void* tableBase, tableType_t tableType) +{ + LZ4_STATIC_ASSERT(LZ4_MEMORY_USAGE > 2); + if (tableType == byU32) { + const U32* const hashTable = (const U32*) tableBase; + assert(h < (1U << (LZ4_MEMORY_USAGE-2))); + return hashTable[h]; + } + if (tableType == byU16) { + const U16* const hashTable = (const U16*) tableBase; + assert(h < (1U << (LZ4_MEMORY_USAGE-1))); + return hashTable[h]; + } + assert(0); return 0; /* forbidden case */ +} + +static const BYTE* LZ4_getPositionOnHash(U32 h, const void* tableBase, tableType_t tableType) +{ + assert(tableType == byPtr); (void)tableType; + { const BYTE* const* hashTable = (const BYTE* const*) tableBase; return hashTable[h]; } +} + +LZ4_FORCE_INLINE const BYTE* +LZ4_getPosition(const BYTE* p, + const void* tableBase, tableType_t tableType) +{ + U32 const h = LZ4_hashPosition(p, tableType); + return LZ4_getPositionOnHash(h, tableBase, tableType); +} + +LZ4_FORCE_INLINE void +LZ4_prepareTable(LZ4_stream_t_internal* const cctx, + const int inputSize, + const tableType_t tableType) { + /* If the table hasn't been used, it's guaranteed to be zeroed out, and is + * therefore safe to use no matter what mode we're in. Otherwise, we figure + * out if it's safe to leave as is or whether it needs to be reset. + */ + if ((tableType_t)cctx->tableType != clearedTable) { + assert(inputSize >= 0); + if ((tableType_t)cctx->tableType != tableType + || ((tableType == byU16) && cctx->currentOffset + (unsigned)inputSize >= 0xFFFFU) + || ((tableType == byU32) && cctx->currentOffset > 1 GB) + || tableType == byPtr + || inputSize >= 4 KB) + { + DEBUGLOG(4, "LZ4_prepareTable: Resetting table in %p", cctx); + MEM_INIT(cctx->hashTable, 0, LZ4_HASHTABLESIZE); + cctx->currentOffset = 0; + cctx->tableType = (U32)clearedTable; + } else { + DEBUGLOG(4, "LZ4_prepareTable: Re-use hash table (no reset)"); + } + } + + /* Adding a gap, so all previous entries are > LZ4_DISTANCE_MAX back, + * is faster than compressing without a gap. + * However, compressing with currentOffset == 0 is faster still, + * so we preserve that case. + */ + if (cctx->currentOffset != 0 && tableType == byU32) { + DEBUGLOG(5, "LZ4_prepareTable: adding 64KB to currentOffset"); + cctx->currentOffset += 64 KB; + } + + /* Finally, clear history */ + cctx->dictCtx = NULL; + cctx->dictionary = NULL; + cctx->dictSize = 0; +} + +/** LZ4_compress_generic_validated() : + * inlined, to ensure branches are decided at compilation time. + * The following conditions are presumed already validated: + * - source != NULL + * - inputSize > 0 + */ +LZ4_FORCE_INLINE int LZ4_compress_generic_validated( + LZ4_stream_t_internal* const cctx, + const char* const source, + char* const dest, + const int inputSize, + int* inputConsumed, /* only written when outputDirective == fillOutput */ + const int maxOutputSize, + const limitedOutput_directive outputDirective, + const tableType_t tableType, + const dict_directive dictDirective, + const dictIssue_directive dictIssue, + const int acceleration) +{ + int result; + const BYTE* ip = (const BYTE*)source; + + U32 const startIndex = cctx->currentOffset; + const BYTE* base = (const BYTE*)source - startIndex; + const BYTE* lowLimit; + + const LZ4_stream_t_internal* dictCtx = (const LZ4_stream_t_internal*) cctx->dictCtx; + const BYTE* const dictionary = + dictDirective == usingDictCtx ? dictCtx->dictionary : cctx->dictionary; + const U32 dictSize = + dictDirective == usingDictCtx ? dictCtx->dictSize : cctx->dictSize; + const U32 dictDelta = + (dictDirective == usingDictCtx) ? startIndex - dictCtx->currentOffset : 0; /* make indexes in dictCtx comparable with indexes in current context */ + + int const maybe_extMem = (dictDirective == usingExtDict) || (dictDirective == usingDictCtx); + U32 const prefixIdxLimit = startIndex - dictSize; /* used when dictDirective == dictSmall */ + const BYTE* const dictEnd = dictionary ? dictionary + dictSize : dictionary; + const BYTE* anchor = (const BYTE*) source; + const BYTE* const iend = ip + inputSize; + const BYTE* const mflimitPlusOne = iend - MFLIMIT + 1; + const BYTE* const matchlimit = iend - LASTLITERALS; + + /* the dictCtx currentOffset is indexed on the start of the dictionary, + * while a dictionary in the current context precedes the currentOffset */ + const BYTE* dictBase = (dictionary == NULL) ? NULL : + (dictDirective == usingDictCtx) ? + dictionary + dictSize - dictCtx->currentOffset : + dictionary + dictSize - startIndex; + + BYTE* op = (BYTE*) dest; + BYTE* const olimit = op + maxOutputSize; + + U32 offset = 0; + U32 forwardH; + + DEBUGLOG(5, "LZ4_compress_generic_validated: srcSize=%i, tableType=%u", inputSize, tableType); + assert(ip != NULL); + if (tableType == byU16) assert(inputSize= 1); + + lowLimit = (const BYTE*)source - (dictDirective == withPrefix64k ? dictSize : 0); + + /* Update context state */ + if (dictDirective == usingDictCtx) { + /* Subsequent linked blocks can't use the dictionary. */ + /* Instead, they use the block we just compressed. */ + cctx->dictCtx = NULL; + cctx->dictSize = (U32)inputSize; + } else { + cctx->dictSize += (U32)inputSize; + } + cctx->currentOffset += (U32)inputSize; + cctx->tableType = (U32)tableType; + + if (inputSizehashTable, byPtr); + } else { + LZ4_putIndexOnHash(startIndex, h, cctx->hashTable, tableType); + } } + ip++; forwardH = LZ4_hashPosition(ip, tableType); + + /* Main Loop */ + for ( ; ; ) { + const BYTE* match; + BYTE* token; + const BYTE* filledIp; + + /* Find a match */ + if (tableType == byPtr) { + const BYTE* forwardIp = ip; + int step = 1; + int searchMatchNb = acceleration << LZ4_skipTrigger; + do { + U32 const h = forwardH; + ip = forwardIp; + forwardIp += step; + step = (searchMatchNb++ >> LZ4_skipTrigger); + + if (unlikely(forwardIp > mflimitPlusOne)) goto _last_literals; + assert(ip < mflimitPlusOne); + + match = LZ4_getPositionOnHash(h, cctx->hashTable, tableType); + forwardH = LZ4_hashPosition(forwardIp, tableType); + LZ4_putPositionOnHash(ip, h, cctx->hashTable, tableType); + + } while ( (match+LZ4_DISTANCE_MAX < ip) + || (LZ4_read32(match) != LZ4_read32(ip)) ); + + } else { /* byU32, byU16 */ + + const BYTE* forwardIp = ip; + int step = 1; + int searchMatchNb = acceleration << LZ4_skipTrigger; + do { + U32 const h = forwardH; + U32 const current = (U32)(forwardIp - base); + U32 matchIndex = LZ4_getIndexOnHash(h, cctx->hashTable, tableType); + assert(matchIndex <= current); + assert(forwardIp - base < (ptrdiff_t)(2 GB - 1)); + ip = forwardIp; + forwardIp += step; + step = (searchMatchNb++ >> LZ4_skipTrigger); + + if (unlikely(forwardIp > mflimitPlusOne)) goto _last_literals; + assert(ip < mflimitPlusOne); + + if (dictDirective == usingDictCtx) { + if (matchIndex < startIndex) { + /* there was no match, try the dictionary */ + assert(tableType == byU32); + matchIndex = LZ4_getIndexOnHash(h, dictCtx->hashTable, byU32); + match = dictBase + matchIndex; + matchIndex += dictDelta; /* make dictCtx index comparable with current context */ + lowLimit = dictionary; + } else { + match = base + matchIndex; + lowLimit = (const BYTE*)source; + } + } else if (dictDirective == usingExtDict) { + if (matchIndex < startIndex) { + DEBUGLOG(7, "extDict candidate: matchIndex=%5u < startIndex=%5u", matchIndex, startIndex); + assert(startIndex - matchIndex >= MINMATCH); + assert(dictBase); + match = dictBase + matchIndex; + lowLimit = dictionary; + } else { + match = base + matchIndex; + lowLimit = (const BYTE*)source; + } + } else { /* single continuous memory segment */ + match = base + matchIndex; + } + forwardH = LZ4_hashPosition(forwardIp, tableType); + LZ4_putIndexOnHash(current, h, cctx->hashTable, tableType); + + DEBUGLOG(7, "candidate at pos=%u (offset=%u \n", matchIndex, current - matchIndex); + if ((dictIssue == dictSmall) && (matchIndex < prefixIdxLimit)) { continue; } /* match outside of valid area */ + assert(matchIndex < current); + if ( ((tableType != byU16) || (LZ4_DISTANCE_MAX < LZ4_DISTANCE_ABSOLUTE_MAX)) + && (matchIndex+LZ4_DISTANCE_MAX < current)) { + continue; + } /* too far */ + assert((current - matchIndex) <= LZ4_DISTANCE_MAX); /* match now expected within distance */ + + if (LZ4_read32(match) == LZ4_read32(ip)) { + if (maybe_extMem) offset = current - matchIndex; + break; /* match found */ + } + + } while(1); + } + + /* Catch up */ + filledIp = ip; + assert(ip > anchor); /* this is always true as ip has been advanced before entering the main loop */ + if ((match > lowLimit) && unlikely(ip[-1] == match[-1])) { + do { ip--; match--; } while (((ip > anchor) & (match > lowLimit)) && (unlikely(ip[-1] == match[-1]))); + } + + /* Encode Literals */ + { unsigned const litLength = (unsigned)(ip - anchor); + token = op++; + if ((outputDirective == limitedOutput) && /* Check output buffer overflow */ + (unlikely(op + litLength + (2 + 1 + LASTLITERALS) + (litLength/255) > olimit)) ) { + return 0; /* cannot compress within `dst` budget. Stored indexes in hash table are nonetheless fine */ + } + if ((outputDirective == fillOutput) && + (unlikely(op + (litLength+240)/255 /* litlen */ + litLength /* literals */ + 2 /* offset */ + 1 /* token */ + MFLIMIT - MINMATCH /* min last literals so last match is <= end - MFLIMIT */ > olimit))) { + op--; + goto _last_literals; + } + if (litLength >= RUN_MASK) { + int len = (int)(litLength - RUN_MASK); + *token = (RUN_MASK<= 255 ; len-=255) *op++ = 255; + *op++ = (BYTE)len; + } + else *token = (BYTE)(litLength< olimit)) { + /* the match was too close to the end, rewind and go to last literals */ + op = token; + goto _last_literals; + } + + /* Encode Offset */ + if (maybe_extMem) { /* static test */ + DEBUGLOG(6, " with offset=%u (ext if > %i)", offset, (int)(ip - (const BYTE*)source)); + assert(offset <= LZ4_DISTANCE_MAX && offset > 0); + LZ4_writeLE16(op, (U16)offset); op+=2; + } else { + DEBUGLOG(6, " with offset=%u (same segment)", (U32)(ip - match)); + assert(ip-match <= LZ4_DISTANCE_MAX); + LZ4_writeLE16(op, (U16)(ip - match)); op+=2; + } + + /* Encode MatchLength */ + { unsigned matchCode; + + if ( (dictDirective==usingExtDict || dictDirective==usingDictCtx) + && (lowLimit==dictionary) /* match within extDict */ ) { + const BYTE* limit = ip + (dictEnd-match); + assert(dictEnd > match); + if (limit > matchlimit) limit = matchlimit; + matchCode = LZ4_count(ip+MINMATCH, match+MINMATCH, limit); + ip += (size_t)matchCode + MINMATCH; + if (ip==limit) { + unsigned const more = LZ4_count(limit, (const BYTE*)source, matchlimit); + matchCode += more; + ip += more; + } + DEBUGLOG(6, " with matchLength=%u starting in extDict", matchCode+MINMATCH); + } else { + matchCode = LZ4_count(ip+MINMATCH, match+MINMATCH, matchlimit); + ip += (size_t)matchCode + MINMATCH; + DEBUGLOG(6, " with matchLength=%u", matchCode+MINMATCH); + } + + if ((outputDirective) && /* Check output buffer overflow */ + (unlikely(op + (1 + LASTLITERALS) + (matchCode+240)/255 > olimit)) ) { + if (outputDirective == fillOutput) { + /* Match description too long : reduce it */ + U32 newMatchCode = 15 /* in token */ - 1 /* to avoid needing a zero byte */ + ((U32)(olimit - op) - 1 - LASTLITERALS) * 255; + ip -= matchCode - newMatchCode; + assert(newMatchCode < matchCode); + matchCode = newMatchCode; + if (unlikely(ip <= filledIp)) { + /* We have already filled up to filledIp so if ip ends up less than filledIp + * we have positions in the hash table beyond the current position. This is + * a problem if we reuse the hash table. So we have to remove these positions + * from the hash table. + */ + const BYTE* ptr; + DEBUGLOG(5, "Clearing %u positions", (U32)(filledIp - ip)); + for (ptr = ip; ptr <= filledIp; ++ptr) { + U32 const h = LZ4_hashPosition(ptr, tableType); + LZ4_clearHash(h, cctx->hashTable, tableType); + } + } + } else { + assert(outputDirective == limitedOutput); + return 0; /* cannot compress within `dst` budget. Stored indexes in hash table are nonetheless fine */ + } + } + if (matchCode >= ML_MASK) { + *token += ML_MASK; + matchCode -= ML_MASK; + LZ4_write32(op, 0xFFFFFFFF); + while (matchCode >= 4*255) { + op+=4; + LZ4_write32(op, 0xFFFFFFFF); + matchCode -= 4*255; + } + op += matchCode / 255; + *op++ = (BYTE)(matchCode % 255); + } else + *token += (BYTE)(matchCode); + } + /* Ensure we have enough space for the last literals. */ + assert(!(outputDirective == fillOutput && op + 1 + LASTLITERALS > olimit)); + + anchor = ip; + + /* Test end of chunk */ + if (ip >= mflimitPlusOne) break; + + /* Fill table */ + { U32 const h = LZ4_hashPosition(ip-2, tableType); + if (tableType == byPtr) { + LZ4_putPositionOnHash(ip-2, h, cctx->hashTable, byPtr); + } else { + U32 const idx = (U32)((ip-2) - base); + LZ4_putIndexOnHash(idx, h, cctx->hashTable, tableType); + } } + + /* Test next position */ + if (tableType == byPtr) { + + match = LZ4_getPosition(ip, cctx->hashTable, tableType); + LZ4_putPosition(ip, cctx->hashTable, tableType); + if ( (match+LZ4_DISTANCE_MAX >= ip) + && (LZ4_read32(match) == LZ4_read32(ip)) ) + { token=op++; *token=0; goto _next_match; } + + } else { /* byU32, byU16 */ + + U32 const h = LZ4_hashPosition(ip, tableType); + U32 const current = (U32)(ip-base); + U32 matchIndex = LZ4_getIndexOnHash(h, cctx->hashTable, tableType); + assert(matchIndex < current); + if (dictDirective == usingDictCtx) { + if (matchIndex < startIndex) { + /* there was no match, try the dictionary */ + assert(tableType == byU32); + matchIndex = LZ4_getIndexOnHash(h, dictCtx->hashTable, byU32); + match = dictBase + matchIndex; + lowLimit = dictionary; /* required for match length counter */ + matchIndex += dictDelta; + } else { + match = base + matchIndex; + lowLimit = (const BYTE*)source; /* required for match length counter */ + } + } else if (dictDirective==usingExtDict) { + if (matchIndex < startIndex) { + assert(dictBase); + match = dictBase + matchIndex; + lowLimit = dictionary; /* required for match length counter */ + } else { + match = base + matchIndex; + lowLimit = (const BYTE*)source; /* required for match length counter */ + } + } else { /* single memory segment */ + match = base + matchIndex; + } + LZ4_putIndexOnHash(current, h, cctx->hashTable, tableType); + assert(matchIndex < current); + if ( ((dictIssue==dictSmall) ? (matchIndex >= prefixIdxLimit) : 1) + && (((tableType==byU16) && (LZ4_DISTANCE_MAX == LZ4_DISTANCE_ABSOLUTE_MAX)) ? 1 : (matchIndex+LZ4_DISTANCE_MAX >= current)) + && (LZ4_read32(match) == LZ4_read32(ip)) ) { + token=op++; + *token=0; + if (maybe_extMem) offset = current - matchIndex; + DEBUGLOG(6, "seq.start:%i, literals=%u, match.start:%i", + (int)(anchor-(const BYTE*)source), 0, (int)(ip-(const BYTE*)source)); + goto _next_match; + } + } + + /* Prepare next loop */ + forwardH = LZ4_hashPosition(++ip, tableType); + + } + +_last_literals: + /* Encode Last Literals */ + { size_t lastRun = (size_t)(iend - anchor); + if ( (outputDirective) && /* Check output buffer overflow */ + (op + lastRun + 1 + ((lastRun+255-RUN_MASK)/255) > olimit)) { + if (outputDirective == fillOutput) { + /* adapt lastRun to fill 'dst' */ + assert(olimit >= op); + lastRun = (size_t)(olimit-op) - 1/*token*/; + lastRun -= (lastRun + 256 - RUN_MASK) / 256; /*additional length tokens*/ + } else { + assert(outputDirective == limitedOutput); + return 0; /* cannot compress within `dst` budget. Stored indexes in hash table are nonetheless fine */ + } + } + DEBUGLOG(6, "Final literal run : %i literals", (int)lastRun); + if (lastRun >= RUN_MASK) { + size_t accumulator = lastRun - RUN_MASK; + *op++ = RUN_MASK << ML_BITS; + for(; accumulator >= 255 ; accumulator-=255) *op++ = 255; + *op++ = (BYTE) accumulator; + } else { + *op++ = (BYTE)(lastRun< 0); + DEBUGLOG(5, "LZ4_compress_generic: compressed %i bytes into %i bytes", inputSize, result); + return result; +} + +/** LZ4_compress_generic() : + * inlined, to ensure branches are decided at compilation time; + * takes care of src == (NULL, 0) + * and forward the rest to LZ4_compress_generic_validated */ +LZ4_FORCE_INLINE int LZ4_compress_generic( + LZ4_stream_t_internal* const cctx, + const char* const src, + char* const dst, + const int srcSize, + int *inputConsumed, /* only written when outputDirective == fillOutput */ + const int dstCapacity, + const limitedOutput_directive outputDirective, + const tableType_t tableType, + const dict_directive dictDirective, + const dictIssue_directive dictIssue, + const int acceleration) +{ + DEBUGLOG(5, "LZ4_compress_generic: srcSize=%i, dstCapacity=%i", + srcSize, dstCapacity); + + if ((U32)srcSize > (U32)LZ4_MAX_INPUT_SIZE) { return 0; } /* Unsupported srcSize, too large (or negative) */ + if (srcSize == 0) { /* src == NULL supported if srcSize == 0 */ + if (outputDirective != notLimited && dstCapacity <= 0) return 0; /* no output, can't write anything */ + DEBUGLOG(5, "Generating an empty block"); + assert(outputDirective == notLimited || dstCapacity >= 1); + assert(dst != NULL); + dst[0] = 0; + if (outputDirective == fillOutput) { + assert (inputConsumed != NULL); + *inputConsumed = 0; + } + return 1; + } + assert(src != NULL); + + return LZ4_compress_generic_validated(cctx, src, dst, srcSize, + inputConsumed, /* only written into if outputDirective == fillOutput */ + dstCapacity, outputDirective, + tableType, dictDirective, dictIssue, acceleration); +} + + +int LZ4_compress_fast_extState(void* state, const char* source, char* dest, int inputSize, int maxOutputSize, int acceleration) +{ + LZ4_stream_t_internal* const ctx = & LZ4_initStream(state, sizeof(LZ4_stream_t)) -> internal_donotuse; + assert(ctx != NULL); + if (acceleration < 1) acceleration = LZ4_ACCELERATION_DEFAULT; + if (acceleration > LZ4_ACCELERATION_MAX) acceleration = LZ4_ACCELERATION_MAX; + if (maxOutputSize >= LZ4_compressBound(inputSize)) { + if (inputSize < LZ4_64Klimit) { + return LZ4_compress_generic(ctx, source, dest, inputSize, NULL, 0, notLimited, byU16, noDict, noDictIssue, acceleration); + } else { + const tableType_t tableType = ((sizeof(void*)==4) && ((uptrval)source > LZ4_DISTANCE_MAX)) ? byPtr : byU32; + return LZ4_compress_generic(ctx, source, dest, inputSize, NULL, 0, notLimited, tableType, noDict, noDictIssue, acceleration); + } + } else { + if (inputSize < LZ4_64Klimit) { + return LZ4_compress_generic(ctx, source, dest, inputSize, NULL, maxOutputSize, limitedOutput, byU16, noDict, noDictIssue, acceleration); + } else { + const tableType_t tableType = ((sizeof(void*)==4) && ((uptrval)source > LZ4_DISTANCE_MAX)) ? byPtr : byU32; + return LZ4_compress_generic(ctx, source, dest, inputSize, NULL, maxOutputSize, limitedOutput, tableType, noDict, noDictIssue, acceleration); + } + } +} + +/** + * LZ4_compress_fast_extState_fastReset() : + * A variant of LZ4_compress_fast_extState(). + * + * Using this variant avoids an expensive initialization step. It is only safe + * to call if the state buffer is known to be correctly initialized already + * (see comment in lz4.h on LZ4_resetStream_fast() for a definition of + * "correctly initialized"). + */ +int LZ4_compress_fast_extState_fastReset(void* state, const char* src, char* dst, int srcSize, int dstCapacity, int acceleration) +{ + LZ4_stream_t_internal* const ctx = &((LZ4_stream_t*)state)->internal_donotuse; + if (acceleration < 1) acceleration = LZ4_ACCELERATION_DEFAULT; + if (acceleration > LZ4_ACCELERATION_MAX) acceleration = LZ4_ACCELERATION_MAX; + assert(ctx != NULL); + + if (dstCapacity >= LZ4_compressBound(srcSize)) { + if (srcSize < LZ4_64Klimit) { + const tableType_t tableType = byU16; + LZ4_prepareTable(ctx, srcSize, tableType); + if (ctx->currentOffset) { + return LZ4_compress_generic(ctx, src, dst, srcSize, NULL, 0, notLimited, tableType, noDict, dictSmall, acceleration); + } else { + return LZ4_compress_generic(ctx, src, dst, srcSize, NULL, 0, notLimited, tableType, noDict, noDictIssue, acceleration); + } + } else { + const tableType_t tableType = ((sizeof(void*)==4) && ((uptrval)src > LZ4_DISTANCE_MAX)) ? byPtr : byU32; + LZ4_prepareTable(ctx, srcSize, tableType); + return LZ4_compress_generic(ctx, src, dst, srcSize, NULL, 0, notLimited, tableType, noDict, noDictIssue, acceleration); + } + } else { + if (srcSize < LZ4_64Klimit) { + const tableType_t tableType = byU16; + LZ4_prepareTable(ctx, srcSize, tableType); + if (ctx->currentOffset) { + return LZ4_compress_generic(ctx, src, dst, srcSize, NULL, dstCapacity, limitedOutput, tableType, noDict, dictSmall, acceleration); + } else { + return LZ4_compress_generic(ctx, src, dst, srcSize, NULL, dstCapacity, limitedOutput, tableType, noDict, noDictIssue, acceleration); + } + } else { + const tableType_t tableType = ((sizeof(void*)==4) && ((uptrval)src > LZ4_DISTANCE_MAX)) ? byPtr : byU32; + LZ4_prepareTable(ctx, srcSize, tableType); + return LZ4_compress_generic(ctx, src, dst, srcSize, NULL, dstCapacity, limitedOutput, tableType, noDict, noDictIssue, acceleration); + } + } +} + + +int LZ4_compress_fast(const char* src, char* dest, int srcSize, int dstCapacity, int acceleration) +{ + int result; +#if (LZ4_HEAPMODE) + LZ4_stream_t* const ctxPtr = (LZ4_stream_t*)ALLOC(sizeof(LZ4_stream_t)); /* malloc-calloc always properly aligned */ + if (ctxPtr == NULL) return 0; +#else + LZ4_stream_t ctx; + LZ4_stream_t* const ctxPtr = &ctx; +#endif + result = LZ4_compress_fast_extState(ctxPtr, src, dest, srcSize, dstCapacity, acceleration); + +#if (LZ4_HEAPMODE) + FREEMEM(ctxPtr); +#endif + return result; +} + + +int LZ4_compress_default(const char* src, char* dst, int srcSize, int dstCapacity) +{ + return LZ4_compress_fast(src, dst, srcSize, dstCapacity, 1); +} + + +/* Note!: This function leaves the stream in an unclean/broken state! + * It is not safe to subsequently use the same state with a _fastReset() or + * _continue() call without resetting it. */ +static int LZ4_compress_destSize_extState_internal(LZ4_stream_t* state, const char* src, char* dst, int* srcSizePtr, int targetDstSize, int acceleration) +{ + void* const s = LZ4_initStream(state, sizeof (*state)); + assert(s != NULL); (void)s; + + if (targetDstSize >= LZ4_compressBound(*srcSizePtr)) { /* compression success is guaranteed */ + return LZ4_compress_fast_extState(state, src, dst, *srcSizePtr, targetDstSize, acceleration); + } else { + if (*srcSizePtr < LZ4_64Klimit) { + return LZ4_compress_generic(&state->internal_donotuse, src, dst, *srcSizePtr, srcSizePtr, targetDstSize, fillOutput, byU16, noDict, noDictIssue, acceleration); + } else { + tableType_t const addrMode = ((sizeof(void*)==4) && ((uptrval)src > LZ4_DISTANCE_MAX)) ? byPtr : byU32; + return LZ4_compress_generic(&state->internal_donotuse, src, dst, *srcSizePtr, srcSizePtr, targetDstSize, fillOutput, addrMode, noDict, noDictIssue, acceleration); + } } +} + +int LZ4_compress_destSize_extState(void* state, const char* src, char* dst, int* srcSizePtr, int targetDstSize, int acceleration) +{ + int const r = LZ4_compress_destSize_extState_internal((LZ4_stream_t*)state, src, dst, srcSizePtr, targetDstSize, acceleration); + /* clean the state on exit */ + LZ4_initStream(state, sizeof (LZ4_stream_t)); + return r; +} + + +int LZ4_compress_destSize(const char* src, char* dst, int* srcSizePtr, int targetDstSize) +{ +#if (LZ4_HEAPMODE) + LZ4_stream_t* const ctx = (LZ4_stream_t*)ALLOC(sizeof(LZ4_stream_t)); /* malloc-calloc always properly aligned */ + if (ctx == NULL) return 0; +#else + LZ4_stream_t ctxBody; + LZ4_stream_t* const ctx = &ctxBody; +#endif + + int result = LZ4_compress_destSize_extState_internal(ctx, src, dst, srcSizePtr, targetDstSize, 1); + +#if (LZ4_HEAPMODE) + FREEMEM(ctx); +#endif + return result; +} + + + +/*-****************************** +* Streaming functions +********************************/ + +#if !defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION) +LZ4_stream_t* LZ4_createStream(void) +{ + LZ4_stream_t* const lz4s = (LZ4_stream_t*)ALLOC(sizeof(LZ4_stream_t)); + LZ4_STATIC_ASSERT(sizeof(LZ4_stream_t) >= sizeof(LZ4_stream_t_internal)); + DEBUGLOG(4, "LZ4_createStream %p", lz4s); + if (lz4s == NULL) return NULL; + LZ4_initStream(lz4s, sizeof(*lz4s)); + return lz4s; +} +#endif + +static size_t LZ4_stream_t_alignment(void) +{ +#if LZ4_ALIGN_TEST + typedef struct { char c; LZ4_stream_t t; } t_a; + return sizeof(t_a) - sizeof(LZ4_stream_t); +#else + return 1; /* effectively disabled */ +#endif +} + +LZ4_stream_t* LZ4_initStream (void* buffer, size_t size) +{ + DEBUGLOG(5, "LZ4_initStream"); + if (buffer == NULL) { return NULL; } + if (size < sizeof(LZ4_stream_t)) { return NULL; } + if (!LZ4_isAligned(buffer, LZ4_stream_t_alignment())) return NULL; + MEM_INIT(buffer, 0, sizeof(LZ4_stream_t_internal)); + return (LZ4_stream_t*)buffer; +} + +/* resetStream is now deprecated, + * prefer initStream() which is more general */ +void LZ4_resetStream (LZ4_stream_t* LZ4_stream) +{ + DEBUGLOG(5, "LZ4_resetStream (ctx:%p)", LZ4_stream); + MEM_INIT(LZ4_stream, 0, sizeof(LZ4_stream_t_internal)); +} + +void LZ4_resetStream_fast(LZ4_stream_t* ctx) { + LZ4_prepareTable(&(ctx->internal_donotuse), 0, byU32); +} + +#if !defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION) +int LZ4_freeStream (LZ4_stream_t* LZ4_stream) +{ + if (!LZ4_stream) return 0; /* support free on NULL */ + DEBUGLOG(5, "LZ4_freeStream %p", LZ4_stream); + FREEMEM(LZ4_stream); + return (0); +} +#endif + + +typedef enum { _ld_fast, _ld_slow } LoadDict_mode_e; +#define HASH_UNIT sizeof(reg_t) +int LZ4_loadDict_internal(LZ4_stream_t* LZ4_dict, + const char* dictionary, int dictSize, + LoadDict_mode_e _ld) +{ + LZ4_stream_t_internal* const dict = &LZ4_dict->internal_donotuse; + const tableType_t tableType = byU32; + const BYTE* p = (const BYTE*)dictionary; + const BYTE* const dictEnd = p + dictSize; + U32 idx32; + + DEBUGLOG(4, "LZ4_loadDict (%i bytes from %p into %p)", dictSize, dictionary, LZ4_dict); + + /* It's necessary to reset the context, + * and not just continue it with prepareTable() + * to avoid any risk of generating overflowing matchIndex + * when compressing using this dictionary */ + LZ4_resetStream(LZ4_dict); + + /* We always increment the offset by 64 KB, since, if the dict is longer, + * we truncate it to the last 64k, and if it's shorter, we still want to + * advance by a whole window length so we can provide the guarantee that + * there are only valid offsets in the window, which allows an optimization + * in LZ4_compress_fast_continue() where it uses noDictIssue even when the + * dictionary isn't a full 64k. */ + dict->currentOffset += 64 KB; + + if (dictSize < (int)HASH_UNIT) { + return 0; + } + + if ((dictEnd - p) > 64 KB) p = dictEnd - 64 KB; + dict->dictionary = p; + dict->dictSize = (U32)(dictEnd - p); + dict->tableType = (U32)tableType; + idx32 = dict->currentOffset - dict->dictSize; + + while (p <= dictEnd-HASH_UNIT) { + U32 const h = LZ4_hashPosition(p, tableType); + /* Note: overwriting => favors positions end of dictionary */ + LZ4_putIndexOnHash(idx32, h, dict->hashTable, tableType); + p+=3; idx32+=3; + } + + if (_ld == _ld_slow) { + /* Fill hash table with additional references, to improve compression capability */ + p = dict->dictionary; + idx32 = dict->currentOffset - dict->dictSize; + while (p <= dictEnd-HASH_UNIT) { + U32 const h = LZ4_hashPosition(p, tableType); + U32 const limit = dict->currentOffset - 64 KB; + if (LZ4_getIndexOnHash(h, dict->hashTable, tableType) <= limit) { + /* Note: not overwriting => favors positions beginning of dictionary */ + LZ4_putIndexOnHash(idx32, h, dict->hashTable, tableType); + } + p++; idx32++; + } + } + + return (int)dict->dictSize; +} + +int LZ4_loadDict(LZ4_stream_t* LZ4_dict, const char* dictionary, int dictSize) +{ + return LZ4_loadDict_internal(LZ4_dict, dictionary, dictSize, _ld_fast); +} + +int LZ4_loadDictSlow(LZ4_stream_t* LZ4_dict, const char* dictionary, int dictSize) +{ + return LZ4_loadDict_internal(LZ4_dict, dictionary, dictSize, _ld_slow); +} + +void LZ4_attach_dictionary(LZ4_stream_t* workingStream, const LZ4_stream_t* dictionaryStream) +{ + const LZ4_stream_t_internal* dictCtx = (dictionaryStream == NULL) ? NULL : + &(dictionaryStream->internal_donotuse); + + DEBUGLOG(4, "LZ4_attach_dictionary (%p, %p, size %u)", + workingStream, dictionaryStream, + dictCtx != NULL ? dictCtx->dictSize : 0); + + if (dictCtx != NULL) { + /* If the current offset is zero, we will never look in the + * external dictionary context, since there is no value a table + * entry can take that indicate a miss. In that case, we need + * to bump the offset to something non-zero. + */ + if (workingStream->internal_donotuse.currentOffset == 0) { + workingStream->internal_donotuse.currentOffset = 64 KB; + } + + /* Don't actually attach an empty dictionary. + */ + if (dictCtx->dictSize == 0) { + dictCtx = NULL; + } + } + workingStream->internal_donotuse.dictCtx = dictCtx; +} + + +static void LZ4_renormDictT(LZ4_stream_t_internal* LZ4_dict, int nextSize) +{ + assert(nextSize >= 0); + if (LZ4_dict->currentOffset + (unsigned)nextSize > 0x80000000) { /* potential ptrdiff_t overflow (32-bits mode) */ + /* rescale hash table */ + U32 const delta = LZ4_dict->currentOffset - 64 KB; + const BYTE* dictEnd = LZ4_dict->dictionary + LZ4_dict->dictSize; + int i; + DEBUGLOG(4, "LZ4_renormDictT"); + for (i=0; ihashTable[i] < delta) LZ4_dict->hashTable[i]=0; + else LZ4_dict->hashTable[i] -= delta; + } + LZ4_dict->currentOffset = 64 KB; + if (LZ4_dict->dictSize > 64 KB) LZ4_dict->dictSize = 64 KB; + LZ4_dict->dictionary = dictEnd - LZ4_dict->dictSize; + } +} + + +int LZ4_compress_fast_continue (LZ4_stream_t* LZ4_stream, + const char* source, char* dest, + int inputSize, int maxOutputSize, + int acceleration) +{ + const tableType_t tableType = byU32; + LZ4_stream_t_internal* const streamPtr = &LZ4_stream->internal_donotuse; + const char* dictEnd = streamPtr->dictSize ? (const char*)streamPtr->dictionary + streamPtr->dictSize : NULL; + + DEBUGLOG(5, "LZ4_compress_fast_continue (inputSize=%i, dictSize=%u)", inputSize, streamPtr->dictSize); + + LZ4_renormDictT(streamPtr, inputSize); /* fix index overflow */ + if (acceleration < 1) acceleration = LZ4_ACCELERATION_DEFAULT; + if (acceleration > LZ4_ACCELERATION_MAX) acceleration = LZ4_ACCELERATION_MAX; + + /* invalidate tiny dictionaries */ + if ( (streamPtr->dictSize < 4) /* tiny dictionary : not enough for a hash */ + && (dictEnd != source) /* prefix mode */ + && (inputSize > 0) /* tolerance : don't lose history, in case next invocation would use prefix mode */ + && (streamPtr->dictCtx == NULL) /* usingDictCtx */ + ) { + DEBUGLOG(5, "LZ4_compress_fast_continue: dictSize(%u) at addr:%p is too small", streamPtr->dictSize, streamPtr->dictionary); + /* remove dictionary existence from history, to employ faster prefix mode */ + streamPtr->dictSize = 0; + streamPtr->dictionary = (const BYTE*)source; + dictEnd = source; + } + + /* Check overlapping input/dictionary space */ + { const char* const sourceEnd = source + inputSize; + if ((sourceEnd > (const char*)streamPtr->dictionary) && (sourceEnd < dictEnd)) { + streamPtr->dictSize = (U32)(dictEnd - sourceEnd); + if (streamPtr->dictSize > 64 KB) streamPtr->dictSize = 64 KB; + if (streamPtr->dictSize < 4) streamPtr->dictSize = 0; + streamPtr->dictionary = (const BYTE*)dictEnd - streamPtr->dictSize; + } + } + + /* prefix mode : source data follows dictionary */ + if (dictEnd == source) { + if ((streamPtr->dictSize < 64 KB) && (streamPtr->dictSize < streamPtr->currentOffset)) + return LZ4_compress_generic(streamPtr, source, dest, inputSize, NULL, maxOutputSize, limitedOutput, tableType, withPrefix64k, dictSmall, acceleration); + else + return LZ4_compress_generic(streamPtr, source, dest, inputSize, NULL, maxOutputSize, limitedOutput, tableType, withPrefix64k, noDictIssue, acceleration); + } + + /* external dictionary mode */ + { int result; + if (streamPtr->dictCtx) { + /* We depend here on the fact that dictCtx'es (produced by + * LZ4_loadDict) guarantee that their tables contain no references + * to offsets between dictCtx->currentOffset - 64 KB and + * dictCtx->currentOffset - dictCtx->dictSize. This makes it safe + * to use noDictIssue even when the dict isn't a full 64 KB. + */ + if (inputSize > 4 KB) { + /* For compressing large blobs, it is faster to pay the setup + * cost to copy the dictionary's tables into the active context, + * so that the compression loop is only looking into one table. + */ + LZ4_memcpy(streamPtr, streamPtr->dictCtx, sizeof(*streamPtr)); + result = LZ4_compress_generic(streamPtr, source, dest, inputSize, NULL, maxOutputSize, limitedOutput, tableType, usingExtDict, noDictIssue, acceleration); + } else { + result = LZ4_compress_generic(streamPtr, source, dest, inputSize, NULL, maxOutputSize, limitedOutput, tableType, usingDictCtx, noDictIssue, acceleration); + } + } else { /* small data <= 4 KB */ + if ((streamPtr->dictSize < 64 KB) && (streamPtr->dictSize < streamPtr->currentOffset)) { + result = LZ4_compress_generic(streamPtr, source, dest, inputSize, NULL, maxOutputSize, limitedOutput, tableType, usingExtDict, dictSmall, acceleration); + } else { + result = LZ4_compress_generic(streamPtr, source, dest, inputSize, NULL, maxOutputSize, limitedOutput, tableType, usingExtDict, noDictIssue, acceleration); + } + } + streamPtr->dictionary = (const BYTE*)source; + streamPtr->dictSize = (U32)inputSize; + return result; + } +} + + +/* Hidden debug function, to force-test external dictionary mode */ +int LZ4_compress_forceExtDict (LZ4_stream_t* LZ4_dict, const char* source, char* dest, int srcSize) +{ + LZ4_stream_t_internal* const streamPtr = &LZ4_dict->internal_donotuse; + int result; + + LZ4_renormDictT(streamPtr, srcSize); + + if ((streamPtr->dictSize < 64 KB) && (streamPtr->dictSize < streamPtr->currentOffset)) { + result = LZ4_compress_generic(streamPtr, source, dest, srcSize, NULL, 0, notLimited, byU32, usingExtDict, dictSmall, 1); + } else { + result = LZ4_compress_generic(streamPtr, source, dest, srcSize, NULL, 0, notLimited, byU32, usingExtDict, noDictIssue, 1); + } + + streamPtr->dictionary = (const BYTE*)source; + streamPtr->dictSize = (U32)srcSize; + + return result; +} + + +/*! LZ4_saveDict() : + * If previously compressed data block is not guaranteed to remain available at its memory location, + * save it into a safer place (char* safeBuffer). + * Note : no need to call LZ4_loadDict() afterwards, dictionary is immediately usable, + * one can therefore call LZ4_compress_fast_continue() right after. + * @return : saved dictionary size in bytes (necessarily <= dictSize), or 0 if error. + */ +int LZ4_saveDict (LZ4_stream_t* LZ4_dict, char* safeBuffer, int dictSize) +{ + LZ4_stream_t_internal* const dict = &LZ4_dict->internal_donotuse; + + DEBUGLOG(5, "LZ4_saveDict : dictSize=%i, safeBuffer=%p", dictSize, safeBuffer); + + if ((U32)dictSize > 64 KB) { dictSize = 64 KB; } /* useless to define a dictionary > 64 KB */ + if ((U32)dictSize > dict->dictSize) { dictSize = (int)dict->dictSize; } + + if (safeBuffer == NULL) assert(dictSize == 0); + if (dictSize > 0) { + const BYTE* const previousDictEnd = dict->dictionary + dict->dictSize; + assert(dict->dictionary); + LZ4_memmove(safeBuffer, previousDictEnd - dictSize, (size_t)dictSize); + } + + dict->dictionary = (const BYTE*)safeBuffer; + dict->dictSize = (U32)dictSize; + + return dictSize; +} + + + +/*-******************************* + * Decompression functions + ********************************/ + +typedef enum { decode_full_block = 0, partial_decode = 1 } earlyEnd_directive; + +#undef MIN +#define MIN(a,b) ( (a) < (b) ? (a) : (b) ) + + +/* variant for decompress_unsafe() + * does not know end of input + * presumes input is well formed + * note : will consume at least one byte */ +static size_t read_long_length_no_check(const BYTE** pp) +{ + size_t b, l = 0; + do { b = **pp; (*pp)++; l += b; } while (b==255); + DEBUGLOG(6, "read_long_length_no_check: +length=%zu using %zu input bytes", l, l/255 + 1) + return l; +} + +/* core decoder variant for LZ4_decompress_fast*() + * for legacy support only : these entry points are deprecated. + * - Presumes input is correctly formed (no defense vs malformed inputs) + * - Does not know input size (presume input buffer is "large enough") + * - Decompress a full block (only) + * @return : nb of bytes read from input. + * Note : this variant is not optimized for speed, just for maintenance. + * the goal is to remove support of decompress_fast*() variants by v2.0 +**/ +LZ4_FORCE_INLINE int +LZ4_decompress_unsafe_generic( + const BYTE* const istart, + BYTE* const ostart, + int decompressedSize, + + size_t prefixSize, + const BYTE* const dictStart, /* only if dict==usingExtDict */ + const size_t dictSize /* note: =0 if dictStart==NULL */ + ) +{ + const BYTE* ip = istart; + BYTE* op = (BYTE*)ostart; + BYTE* const oend = ostart + decompressedSize; + const BYTE* const prefixStart = ostart - prefixSize; + + DEBUGLOG(5, "LZ4_decompress_unsafe_generic"); + if (dictStart == NULL) assert(dictSize == 0); + + while (1) { + /* start new sequence */ + unsigned token = *ip++; + + /* literals */ + { size_t ll = token >> ML_BITS; + if (ll==15) { + /* long literal length */ + ll += read_long_length_no_check(&ip); + } + if ((size_t)(oend-op) < ll) return -1; /* output buffer overflow */ + LZ4_memmove(op, ip, ll); /* support in-place decompression */ + op += ll; + ip += ll; + if ((size_t)(oend-op) < MFLIMIT) { + if (op==oend) break; /* end of block */ + DEBUGLOG(5, "invalid: literals end at distance %zi from end of block", oend-op); + /* incorrect end of block : + * last match must start at least MFLIMIT==12 bytes before end of output block */ + return -1; + } } + + /* match */ + { size_t ml = token & 15; + size_t const offset = LZ4_readLE16(ip); + ip+=2; + + if (ml==15) { + /* long literal length */ + ml += read_long_length_no_check(&ip); + } + ml += MINMATCH; + + if ((size_t)(oend-op) < ml) return -1; /* output buffer overflow */ + + { const BYTE* match = op - offset; + + /* out of range */ + if (offset > (size_t)(op - prefixStart) + dictSize) { + DEBUGLOG(6, "offset out of range"); + return -1; + } + + /* check special case : extDict */ + if (offset > (size_t)(op - prefixStart)) { + /* extDict scenario */ + const BYTE* const dictEnd = dictStart + dictSize; + const BYTE* extMatch = dictEnd - (offset - (size_t)(op-prefixStart)); + size_t const extml = (size_t)(dictEnd - extMatch); + if (extml > ml) { + /* match entirely within extDict */ + LZ4_memmove(op, extMatch, ml); + op += ml; + ml = 0; + } else { + /* match split between extDict & prefix */ + LZ4_memmove(op, extMatch, extml); + op += extml; + ml -= extml; + } + match = prefixStart; + } + + /* match copy - slow variant, supporting overlap copy */ + { size_t u; + for (u=0; u= ipmax before start of loop. Returns initial_error if so. + * @error (output) - error code. Must be set to 0 before call. +**/ +typedef size_t Rvl_t; +static const Rvl_t rvl_error = (Rvl_t)(-1); +LZ4_FORCE_INLINE Rvl_t +read_variable_length(const BYTE** ip, const BYTE* ilimit, + int initial_check) +{ + Rvl_t s, length = 0; + assert(ip != NULL); + assert(*ip != NULL); + assert(ilimit != NULL); + if (initial_check && unlikely((*ip) >= ilimit)) { /* read limit reached */ + return rvl_error; + } + s = **ip; + (*ip)++; + length += s; + if (unlikely((*ip) > ilimit)) { /* read limit reached */ + return rvl_error; + } + /* accumulator overflow detection (32-bit mode only) */ + if ((sizeof(length) < 8) && unlikely(length > ((Rvl_t)(-1)/2)) ) { + return rvl_error; + } + if (likely(s != 255)) return length; + do { + s = **ip; + (*ip)++; + length += s; + if (unlikely((*ip) > ilimit)) { /* read limit reached */ + return rvl_error; + } + /* accumulator overflow detection (32-bit mode only) */ + if ((sizeof(length) < 8) && unlikely(length > ((Rvl_t)(-1)/2)) ) { + return rvl_error; + } + } while (s == 255); + + return length; +} + +/*! LZ4_decompress_generic() : + * This generic decompression function covers all use cases. + * It shall be instantiated several times, using different sets of directives. + * Note that it is important for performance that this function really get inlined, + * in order to remove useless branches during compilation optimization. + */ +LZ4_FORCE_INLINE int +LZ4_decompress_generic( + const char* const src, + char* const dst, + int srcSize, + int outputSize, /* If endOnInput==endOnInputSize, this value is `dstCapacity` */ + + earlyEnd_directive partialDecoding, /* full, partial */ + dict_directive dict, /* noDict, withPrefix64k, usingExtDict */ + const BYTE* const lowPrefix, /* always <= dst, == dst when no prefix */ + const BYTE* const dictStart, /* only if dict==usingExtDict */ + const size_t dictSize /* note : = 0 if noDict */ + ) +{ + if ((src == NULL) || (outputSize < 0)) { return -1; } + + { const BYTE* ip = (const BYTE*) src; + const BYTE* const iend = ip + srcSize; + + BYTE* op = (BYTE*) dst; + BYTE* const oend = op + outputSize; + BYTE* cpy; + + const BYTE* const dictEnd = (dictStart == NULL) ? NULL : dictStart + dictSize; + + const int checkOffset = (dictSize < (int)(64 KB)); + + + /* Set up the "end" pointers for the shortcut. */ + const BYTE* const shortiend = iend - 14 /*maxLL*/ - 2 /*offset*/; + const BYTE* const shortoend = oend - 14 /*maxLL*/ - 18 /*maxML*/; + + const BYTE* match; + size_t offset; + unsigned token; + size_t length; + + + DEBUGLOG(5, "LZ4_decompress_generic (srcSize:%i, dstSize:%i)", srcSize, outputSize); + + /* Special cases */ + assert(lowPrefix <= op); + if (unlikely(outputSize==0)) { + /* Empty output buffer */ + if (partialDecoding) return 0; + return ((srcSize==1) && (*ip==0)) ? 0 : -1; + } + if (unlikely(srcSize==0)) { return -1; } + + /* LZ4_FAST_DEC_LOOP: + * designed for modern OoO performance cpus, + * where copying reliably 32-bytes is preferable to an unpredictable branch. + * note : fast loop may show a regression for some client arm chips. */ +#if LZ4_FAST_DEC_LOOP + if ((oend - op) < FASTLOOP_SAFE_DISTANCE) { + DEBUGLOG(6, "move to safe decode loop"); + goto safe_decode; + } + + /* Fast loop : decode sequences as long as output < oend-FASTLOOP_SAFE_DISTANCE */ + DEBUGLOG(6, "using fast decode loop"); + while (1) { + /* Main fastloop assertion: We can always wildcopy FASTLOOP_SAFE_DISTANCE */ + assert(oend - op >= FASTLOOP_SAFE_DISTANCE); + assert(ip < iend); + token = *ip++; + length = token >> ML_BITS; /* literal length */ + DEBUGLOG(7, "blockPos%6u: litLength token = %u", (unsigned)(op-(BYTE*)dst), (unsigned)length); + + /* decode literal length */ + if (length == RUN_MASK) { + size_t const addl = read_variable_length(&ip, iend-RUN_MASK, 1); + if (addl == rvl_error) { + DEBUGLOG(6, "error reading long literal length"); + goto _output_error; + } + length += addl; + if (unlikely((uptrval)(op)+length<(uptrval)(op))) { goto _output_error; } /* overflow detection */ + if (unlikely((uptrval)(ip)+length<(uptrval)(ip))) { goto _output_error; } /* overflow detection */ + + /* copy literals */ + LZ4_STATIC_ASSERT(MFLIMIT >= WILDCOPYLENGTH); + if ((op+length>oend-32) || (ip+length>iend-32)) { goto safe_literal_copy; } + LZ4_wildCopy32(op, ip, op+length); + ip += length; op += length; + } else if (ip <= iend-(16 + 1/*max lit + offset + nextToken*/)) { + /* We don't need to check oend, since we check it once for each loop below */ + DEBUGLOG(7, "copy %u bytes in a 16-bytes stripe", (unsigned)length); + /* Literals can only be <= 14, but hope compilers optimize better when copy by a register size */ + LZ4_memcpy(op, ip, 16); + ip += length; op += length; + } else { + goto safe_literal_copy; + } + + /* get offset */ + offset = LZ4_readLE16(ip); ip+=2; + DEBUGLOG(6, "blockPos%6u: offset = %u", (unsigned)(op-(BYTE*)dst), (unsigned)offset); + match = op - offset; + assert(match <= op); /* overflow check */ + + /* get matchlength */ + length = token & ML_MASK; + DEBUGLOG(7, " match length token = %u (len==%u)", (unsigned)length, (unsigned)length+MINMATCH); + + if (length == ML_MASK) { + size_t const addl = read_variable_length(&ip, iend - LASTLITERALS + 1, 0); + if (addl == rvl_error) { + DEBUGLOG(5, "error reading long match length"); + goto _output_error; + } + length += addl; + length += MINMATCH; + DEBUGLOG(7, " long match length == %u", (unsigned)length); + if (unlikely((uptrval)(op)+length<(uptrval)op)) { goto _output_error; } /* overflow detection */ + if (op + length >= oend - FASTLOOP_SAFE_DISTANCE) { + goto safe_match_copy; + } + } else { + length += MINMATCH; + if (op + length >= oend - FASTLOOP_SAFE_DISTANCE) { + DEBUGLOG(7, "moving to safe_match_copy (ml==%u)", (unsigned)length); + goto safe_match_copy; + } + + /* Fastpath check: skip LZ4_wildCopy32 when true */ + if ((dict == withPrefix64k) || (match >= lowPrefix)) { + if (offset >= 8) { + assert(match >= lowPrefix); + assert(match <= op); + assert(op + 18 <= oend); + + LZ4_memcpy(op, match, 8); + LZ4_memcpy(op+8, match+8, 8); + LZ4_memcpy(op+16, match+16, 2); + op += length; + continue; + } } } + + if ( checkOffset && (unlikely(match + dictSize < lowPrefix)) ) { + DEBUGLOG(5, "Error : pos=%zi, offset=%zi => outside buffers", op-lowPrefix, op-match); + goto _output_error; + } + /* match starting within external dictionary */ + if ((dict==usingExtDict) && (match < lowPrefix)) { + assert(dictEnd != NULL); + if (unlikely(op+length > oend-LASTLITERALS)) { + if (partialDecoding) { + DEBUGLOG(7, "partialDecoding: dictionary match, close to dstEnd"); + length = MIN(length, (size_t)(oend-op)); + } else { + DEBUGLOG(6, "end-of-block condition violated") + goto _output_error; + } } + + if (length <= (size_t)(lowPrefix-match)) { + /* match fits entirely within external dictionary : just copy */ + LZ4_memmove(op, dictEnd - (lowPrefix-match), length); + op += length; + } else { + /* match stretches into both external dictionary and current block */ + size_t const copySize = (size_t)(lowPrefix - match); + size_t const restSize = length - copySize; + LZ4_memcpy(op, dictEnd - copySize, copySize); + op += copySize; + if (restSize > (size_t)(op - lowPrefix)) { /* overlap copy */ + BYTE* const endOfMatch = op + restSize; + const BYTE* copyFrom = lowPrefix; + while (op < endOfMatch) { *op++ = *copyFrom++; } + } else { + LZ4_memcpy(op, lowPrefix, restSize); + op += restSize; + } } + continue; + } + + /* copy match within block */ + cpy = op + length; + + assert((op <= oend) && (oend-op >= 32)); + if (unlikely(offset<16)) { + LZ4_memcpy_using_offset(op, match, cpy, offset); + } else { + LZ4_wildCopy32(op, match, cpy); + } + + op = cpy; /* wildcopy correction */ + } + safe_decode: +#endif + + /* Main Loop : decode remaining sequences where output < FASTLOOP_SAFE_DISTANCE */ + DEBUGLOG(6, "using safe decode loop"); + while (1) { + assert(ip < iend); + token = *ip++; + length = token >> ML_BITS; /* literal length */ + DEBUGLOG(7, "blockPos%6u: litLength token = %u", (unsigned)(op-(BYTE*)dst), (unsigned)length); + + /* A two-stage shortcut for the most common case: + * 1) If the literal length is 0..14, and there is enough space, + * enter the shortcut and copy 16 bytes on behalf of the literals + * (in the fast mode, only 8 bytes can be safely copied this way). + * 2) Further if the match length is 4..18, copy 18 bytes in a similar + * manner; but we ensure that there's enough space in the output for + * those 18 bytes earlier, upon entering the shortcut (in other words, + * there is a combined check for both stages). + */ + if ( (length != RUN_MASK) + /* strictly "less than" on input, to re-enter the loop with at least one byte */ + && likely((ip < shortiend) & (op <= shortoend)) ) { + /* Copy the literals */ + LZ4_memcpy(op, ip, 16); + op += length; ip += length; + + /* The second stage: prepare for match copying, decode full info. + * If it doesn't work out, the info won't be wasted. */ + length = token & ML_MASK; /* match length */ + DEBUGLOG(7, "blockPos%6u: matchLength token = %u (len=%u)", (unsigned)(op-(BYTE*)dst), (unsigned)length, (unsigned)length + 4); + offset = LZ4_readLE16(ip); ip += 2; + match = op - offset; + assert(match <= op); /* check overflow */ + + /* Do not deal with overlapping matches. */ + if ( (length != ML_MASK) + && (offset >= 8) + && (dict==withPrefix64k || match >= lowPrefix) ) { + /* Copy the match. */ + LZ4_memcpy(op + 0, match + 0, 8); + LZ4_memcpy(op + 8, match + 8, 8); + LZ4_memcpy(op +16, match +16, 2); + op += length + MINMATCH; + /* Both stages worked, load the next token. */ + continue; + } + + /* The second stage didn't work out, but the info is ready. + * Propel it right to the point of match copying. */ + goto _copy_match; + } + + /* decode literal length */ + if (length == RUN_MASK) { + size_t const addl = read_variable_length(&ip, iend-RUN_MASK, 1); + if (addl == rvl_error) { goto _output_error; } + length += addl; + if (unlikely((uptrval)(op)+length<(uptrval)(op))) { goto _output_error; } /* overflow detection */ + if (unlikely((uptrval)(ip)+length<(uptrval)(ip))) { goto _output_error; } /* overflow detection */ + } + +#if LZ4_FAST_DEC_LOOP + safe_literal_copy: +#endif + /* copy literals */ + cpy = op+length; + + LZ4_STATIC_ASSERT(MFLIMIT >= WILDCOPYLENGTH); + if ((cpy>oend-MFLIMIT) || (ip+length>iend-(2+1+LASTLITERALS))) { + /* We've either hit the input parsing restriction or the output parsing restriction. + * In the normal scenario, decoding a full block, it must be the last sequence, + * otherwise it's an error (invalid input or dimensions). + * In partialDecoding scenario, it's necessary to ensure there is no buffer overflow. + */ + if (partialDecoding) { + /* Since we are partial decoding we may be in this block because of the output parsing + * restriction, which is not valid since the output buffer is allowed to be undersized. + */ + DEBUGLOG(7, "partialDecoding: copying literals, close to input or output end") + DEBUGLOG(7, "partialDecoding: literal length = %u", (unsigned)length); + DEBUGLOG(7, "partialDecoding: remaining space in dstBuffer : %i", (int)(oend - op)); + DEBUGLOG(7, "partialDecoding: remaining space in srcBuffer : %i", (int)(iend - ip)); + /* Finishing in the middle of a literals segment, + * due to lack of input. + */ + if (ip+length > iend) { + length = (size_t)(iend-ip); + cpy = op + length; + } + /* Finishing in the middle of a literals segment, + * due to lack of output space. + */ + if (cpy > oend) { + cpy = oend; + assert(op<=oend); + length = (size_t)(oend-op); + } + } else { + /* We must be on the last sequence (or invalid) because of the parsing limitations + * so check that we exactly consume the input and don't overrun the output buffer. + */ + if ((ip+length != iend) || (cpy > oend)) { + DEBUGLOG(5, "should have been last run of literals") + DEBUGLOG(5, "ip(%p) + length(%i) = %p != iend (%p)", ip, (int)length, ip+length, iend); + DEBUGLOG(5, "or cpy(%p) > (oend-MFLIMIT)(%p)", cpy, oend-MFLIMIT); + DEBUGLOG(5, "after writing %u bytes / %i bytes available", (unsigned)(op-(BYTE*)dst), outputSize); + goto _output_error; + } + } + LZ4_memmove(op, ip, length); /* supports overlapping memory regions, for in-place decompression scenarios */ + ip += length; + op += length; + /* Necessarily EOF when !partialDecoding. + * When partialDecoding, it is EOF if we've either + * filled the output buffer or + * can't proceed with reading an offset for following match. + */ + if (!partialDecoding || (cpy == oend) || (ip >= (iend-2))) { + break; + } + } else { + LZ4_wildCopy8(op, ip, cpy); /* can overwrite up to 8 bytes beyond cpy */ + ip += length; op = cpy; + } + + /* get offset */ + offset = LZ4_readLE16(ip); ip+=2; + match = op - offset; + + /* get matchlength */ + length = token & ML_MASK; + DEBUGLOG(7, "blockPos%6u: matchLength token = %u", (unsigned)(op-(BYTE*)dst), (unsigned)length); + + _copy_match: + if (length == ML_MASK) { + size_t const addl = read_variable_length(&ip, iend - LASTLITERALS + 1, 0); + if (addl == rvl_error) { goto _output_error; } + length += addl; + if (unlikely((uptrval)(op)+length<(uptrval)op)) goto _output_error; /* overflow detection */ + } + length += MINMATCH; + +#if LZ4_FAST_DEC_LOOP + safe_match_copy: +#endif + if ((checkOffset) && (unlikely(match + dictSize < lowPrefix))) goto _output_error; /* Error : offset outside buffers */ + /* match starting within external dictionary */ + if ((dict==usingExtDict) && (match < lowPrefix)) { + assert(dictEnd != NULL); + if (unlikely(op+length > oend-LASTLITERALS)) { + if (partialDecoding) length = MIN(length, (size_t)(oend-op)); + else goto _output_error; /* doesn't respect parsing restriction */ + } + + if (length <= (size_t)(lowPrefix-match)) { + /* match fits entirely within external dictionary : just copy */ + LZ4_memmove(op, dictEnd - (lowPrefix-match), length); + op += length; + } else { + /* match stretches into both external dictionary and current block */ + size_t const copySize = (size_t)(lowPrefix - match); + size_t const restSize = length - copySize; + LZ4_memcpy(op, dictEnd - copySize, copySize); + op += copySize; + if (restSize > (size_t)(op - lowPrefix)) { /* overlap copy */ + BYTE* const endOfMatch = op + restSize; + const BYTE* copyFrom = lowPrefix; + while (op < endOfMatch) *op++ = *copyFrom++; + } else { + LZ4_memcpy(op, lowPrefix, restSize); + op += restSize; + } } + continue; + } + assert(match >= lowPrefix); + + /* copy match within block */ + cpy = op + length; + + /* partialDecoding : may end anywhere within the block */ + assert(op<=oend); + if (partialDecoding && (cpy > oend-MATCH_SAFEGUARD_DISTANCE)) { + size_t const mlen = MIN(length, (size_t)(oend-op)); + const BYTE* const matchEnd = match + mlen; + BYTE* const copyEnd = op + mlen; + if (matchEnd > op) { /* overlap copy */ + while (op < copyEnd) { *op++ = *match++; } + } else { + LZ4_memcpy(op, match, mlen); + } + op = copyEnd; + if (op == oend) { break; } + continue; + } + + if (unlikely(offset<8)) { + LZ4_write32(op, 0); /* silence msan warning when offset==0 */ + op[0] = match[0]; + op[1] = match[1]; + op[2] = match[2]; + op[3] = match[3]; + match += inc32table[offset]; + LZ4_memcpy(op+4, match, 4); + match -= dec64table[offset]; + } else { + LZ4_memcpy(op, match, 8); + match += 8; + } + op += 8; + + if (unlikely(cpy > oend-MATCH_SAFEGUARD_DISTANCE)) { + BYTE* const oCopyLimit = oend - (WILDCOPYLENGTH-1); + if (cpy > oend-LASTLITERALS) { goto _output_error; } /* Error : last LASTLITERALS bytes must be literals (uncompressed) */ + if (op < oCopyLimit) { + LZ4_wildCopy8(op, match, oCopyLimit); + match += oCopyLimit - op; + op = oCopyLimit; + } + while (op < cpy) { *op++ = *match++; } + } else { + LZ4_memcpy(op, match, 8); + if (length > 16) { LZ4_wildCopy8(op+8, match+8, cpy); } + } + op = cpy; /* wildcopy correction */ + } + + /* end of decoding */ + DEBUGLOG(5, "decoded %i bytes", (int) (((char*)op)-dst)); + return (int) (((char*)op)-dst); /* Nb of output bytes decoded */ + + /* Overflow error detected */ + _output_error: + return (int) (-(((const char*)ip)-src))-1; + } +} + + +/*===== Instantiate the API decoding functions. =====*/ + +LZ4_FORCE_O2 +int LZ4_decompress_safe(const char* source, char* dest, int compressedSize, int maxDecompressedSize) +{ + return LZ4_decompress_generic(source, dest, compressedSize, maxDecompressedSize, + decode_full_block, noDict, + (BYTE*)dest, NULL, 0); +} + +LZ4_FORCE_O2 +int LZ4_decompress_safe_partial(const char* src, char* dst, int compressedSize, int targetOutputSize, int dstCapacity) +{ + dstCapacity = MIN(targetOutputSize, dstCapacity); + return LZ4_decompress_generic(src, dst, compressedSize, dstCapacity, + partial_decode, + noDict, (BYTE*)dst, NULL, 0); +} + +LZ4_FORCE_O2 +int LZ4_decompress_fast(const char* source, char* dest, int originalSize) +{ + DEBUGLOG(5, "LZ4_decompress_fast"); + return LZ4_decompress_unsafe_generic( + (const BYTE*)source, (BYTE*)dest, originalSize, + 0, NULL, 0); +} + +/*===== Instantiate a few more decoding cases, used more than once. =====*/ + +LZ4_FORCE_O2 /* Exported, an obsolete API function. */ +int LZ4_decompress_safe_withPrefix64k(const char* source, char* dest, int compressedSize, int maxOutputSize) +{ + return LZ4_decompress_generic(source, dest, compressedSize, maxOutputSize, + decode_full_block, withPrefix64k, + (BYTE*)dest - 64 KB, NULL, 0); +} + +LZ4_FORCE_O2 +static int LZ4_decompress_safe_partial_withPrefix64k(const char* source, char* dest, int compressedSize, int targetOutputSize, int dstCapacity) +{ + dstCapacity = MIN(targetOutputSize, dstCapacity); + return LZ4_decompress_generic(source, dest, compressedSize, dstCapacity, + partial_decode, withPrefix64k, + (BYTE*)dest - 64 KB, NULL, 0); +} + +/* Another obsolete API function, paired with the previous one. */ +int LZ4_decompress_fast_withPrefix64k(const char* source, char* dest, int originalSize) +{ + return LZ4_decompress_unsafe_generic( + (const BYTE*)source, (BYTE*)dest, originalSize, + 64 KB, NULL, 0); +} + +LZ4_FORCE_O2 +static int LZ4_decompress_safe_withSmallPrefix(const char* source, char* dest, int compressedSize, int maxOutputSize, + size_t prefixSize) +{ + return LZ4_decompress_generic(source, dest, compressedSize, maxOutputSize, + decode_full_block, noDict, + (BYTE*)dest-prefixSize, NULL, 0); +} + +LZ4_FORCE_O2 +static int LZ4_decompress_safe_partial_withSmallPrefix(const char* source, char* dest, int compressedSize, int targetOutputSize, int dstCapacity, + size_t prefixSize) +{ + dstCapacity = MIN(targetOutputSize, dstCapacity); + return LZ4_decompress_generic(source, dest, compressedSize, dstCapacity, + partial_decode, noDict, + (BYTE*)dest-prefixSize, NULL, 0); +} + +LZ4_FORCE_O2 +int LZ4_decompress_safe_forceExtDict(const char* source, char* dest, + int compressedSize, int maxOutputSize, + const void* dictStart, size_t dictSize) +{ + DEBUGLOG(5, "LZ4_decompress_safe_forceExtDict"); + return LZ4_decompress_generic(source, dest, compressedSize, maxOutputSize, + decode_full_block, usingExtDict, + (BYTE*)dest, (const BYTE*)dictStart, dictSize); +} + +LZ4_FORCE_O2 +int LZ4_decompress_safe_partial_forceExtDict(const char* source, char* dest, + int compressedSize, int targetOutputSize, int dstCapacity, + const void* dictStart, size_t dictSize) +{ + dstCapacity = MIN(targetOutputSize, dstCapacity); + return LZ4_decompress_generic(source, dest, compressedSize, dstCapacity, + partial_decode, usingExtDict, + (BYTE*)dest, (const BYTE*)dictStart, dictSize); +} + +LZ4_FORCE_O2 +static int LZ4_decompress_fast_extDict(const char* source, char* dest, int originalSize, + const void* dictStart, size_t dictSize) +{ + return LZ4_decompress_unsafe_generic( + (const BYTE*)source, (BYTE*)dest, originalSize, + 0, (const BYTE*)dictStart, dictSize); +} + +/* The "double dictionary" mode, for use with e.g. ring buffers: the first part + * of the dictionary is passed as prefix, and the second via dictStart + dictSize. + * These routines are used only once, in LZ4_decompress_*_continue(). + */ +LZ4_FORCE_INLINE +int LZ4_decompress_safe_doubleDict(const char* source, char* dest, int compressedSize, int maxOutputSize, + size_t prefixSize, const void* dictStart, size_t dictSize) +{ + return LZ4_decompress_generic(source, dest, compressedSize, maxOutputSize, + decode_full_block, usingExtDict, + (BYTE*)dest-prefixSize, (const BYTE*)dictStart, dictSize); +} + +/*===== streaming decompression functions =====*/ + +#if !defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION) +LZ4_streamDecode_t* LZ4_createStreamDecode(void) +{ + LZ4_STATIC_ASSERT(sizeof(LZ4_streamDecode_t) >= sizeof(LZ4_streamDecode_t_internal)); + return (LZ4_streamDecode_t*) ALLOC_AND_ZERO(sizeof(LZ4_streamDecode_t)); +} + +int LZ4_freeStreamDecode (LZ4_streamDecode_t* LZ4_stream) +{ + if (LZ4_stream == NULL) { return 0; } /* support free on NULL */ + FREEMEM(LZ4_stream); + return 0; +} +#endif + +/*! LZ4_setStreamDecode() : + * Use this function to instruct where to find the dictionary. + * This function is not necessary if previous data is still available where it was decoded. + * Loading a size of 0 is allowed (same effect as no dictionary). + * @return : 1 if OK, 0 if error + */ +int LZ4_setStreamDecode (LZ4_streamDecode_t* LZ4_streamDecode, const char* dictionary, int dictSize) +{ + LZ4_streamDecode_t_internal* lz4sd = &LZ4_streamDecode->internal_donotuse; + lz4sd->prefixSize = (size_t)dictSize; + if (dictSize) { + assert(dictionary != NULL); + lz4sd->prefixEnd = (const BYTE*) dictionary + dictSize; + } else { + lz4sd->prefixEnd = (const BYTE*) dictionary; + } + lz4sd->externalDict = NULL; + lz4sd->extDictSize = 0; + return 1; +} + +/*! LZ4_decoderRingBufferSize() : + * when setting a ring buffer for streaming decompression (optional scenario), + * provides the minimum size of this ring buffer + * to be compatible with any source respecting maxBlockSize condition. + * Note : in a ring buffer scenario, + * blocks are presumed decompressed next to each other. + * When not enough space remains for next block (remainingSize < maxBlockSize), + * decoding resumes from beginning of ring buffer. + * @return : minimum ring buffer size, + * or 0 if there is an error (invalid maxBlockSize). + */ +int LZ4_decoderRingBufferSize(int maxBlockSize) +{ + if (maxBlockSize < 0) return 0; + if (maxBlockSize > LZ4_MAX_INPUT_SIZE) return 0; + if (maxBlockSize < 16) maxBlockSize = 16; + return LZ4_DECODER_RING_BUFFER_SIZE(maxBlockSize); +} + +/* +*_continue() : + These decoding functions allow decompression of multiple blocks in "streaming" mode. + Previously decoded blocks must still be available at the memory position where they were decoded. + If it's not possible, save the relevant part of decoded data into a safe buffer, + and indicate where it stands using LZ4_setStreamDecode() +*/ +LZ4_FORCE_O2 +int LZ4_decompress_safe_continue (LZ4_streamDecode_t* LZ4_streamDecode, const char* source, char* dest, int compressedSize, int maxOutputSize) +{ + LZ4_streamDecode_t_internal* lz4sd = &LZ4_streamDecode->internal_donotuse; + int result; + + if (lz4sd->prefixSize == 0) { + /* The first call, no dictionary yet. */ + assert(lz4sd->extDictSize == 0); + result = LZ4_decompress_safe(source, dest, compressedSize, maxOutputSize); + if (result <= 0) return result; + lz4sd->prefixSize = (size_t)result; + lz4sd->prefixEnd = (BYTE*)dest + result; + } else if (lz4sd->prefixEnd == (BYTE*)dest) { + /* They're rolling the current segment. */ + if (lz4sd->prefixSize >= 64 KB - 1) + result = LZ4_decompress_safe_withPrefix64k(source, dest, compressedSize, maxOutputSize); + else if (lz4sd->extDictSize == 0) + result = LZ4_decompress_safe_withSmallPrefix(source, dest, compressedSize, maxOutputSize, + lz4sd->prefixSize); + else + result = LZ4_decompress_safe_doubleDict(source, dest, compressedSize, maxOutputSize, + lz4sd->prefixSize, lz4sd->externalDict, lz4sd->extDictSize); + if (result <= 0) return result; + lz4sd->prefixSize += (size_t)result; + lz4sd->prefixEnd += result; + } else { + /* The buffer wraps around, or they're switching to another buffer. */ + lz4sd->extDictSize = lz4sd->prefixSize; + lz4sd->externalDict = lz4sd->prefixEnd - lz4sd->extDictSize; + result = LZ4_decompress_safe_forceExtDict(source, dest, compressedSize, maxOutputSize, + lz4sd->externalDict, lz4sd->extDictSize); + if (result <= 0) return result; + lz4sd->prefixSize = (size_t)result; + lz4sd->prefixEnd = (BYTE*)dest + result; + } + + return result; +} + +LZ4_FORCE_O2 int +LZ4_decompress_fast_continue (LZ4_streamDecode_t* LZ4_streamDecode, + const char* source, char* dest, int originalSize) +{ + LZ4_streamDecode_t_internal* const lz4sd = + (assert(LZ4_streamDecode!=NULL), &LZ4_streamDecode->internal_donotuse); + int result; + + DEBUGLOG(5, "LZ4_decompress_fast_continue (toDecodeSize=%i)", originalSize); + assert(originalSize >= 0); + + if (lz4sd->prefixSize == 0) { + DEBUGLOG(5, "first invocation : no prefix nor extDict"); + assert(lz4sd->extDictSize == 0); + result = LZ4_decompress_fast(source, dest, originalSize); + if (result <= 0) return result; + lz4sd->prefixSize = (size_t)originalSize; + lz4sd->prefixEnd = (BYTE*)dest + originalSize; + } else if (lz4sd->prefixEnd == (BYTE*)dest) { + DEBUGLOG(5, "continue using existing prefix"); + result = LZ4_decompress_unsafe_generic( + (const BYTE*)source, (BYTE*)dest, originalSize, + lz4sd->prefixSize, + lz4sd->externalDict, lz4sd->extDictSize); + if (result <= 0) return result; + lz4sd->prefixSize += (size_t)originalSize; + lz4sd->prefixEnd += originalSize; + } else { + DEBUGLOG(5, "prefix becomes extDict"); + lz4sd->extDictSize = lz4sd->prefixSize; + lz4sd->externalDict = lz4sd->prefixEnd - lz4sd->extDictSize; + result = LZ4_decompress_fast_extDict(source, dest, originalSize, + lz4sd->externalDict, lz4sd->extDictSize); + if (result <= 0) return result; + lz4sd->prefixSize = (size_t)originalSize; + lz4sd->prefixEnd = (BYTE*)dest + originalSize; + } + + return result; +} + + +/* +Advanced decoding functions : +*_usingDict() : + These decoding functions work the same as "_continue" ones, + the dictionary must be explicitly provided within parameters +*/ + +int LZ4_decompress_safe_usingDict(const char* source, char* dest, int compressedSize, int maxOutputSize, const char* dictStart, int dictSize) +{ + if (dictSize==0) + return LZ4_decompress_safe(source, dest, compressedSize, maxOutputSize); + if (dictStart+dictSize == dest) { + if (dictSize >= 64 KB - 1) { + return LZ4_decompress_safe_withPrefix64k(source, dest, compressedSize, maxOutputSize); + } + assert(dictSize >= 0); + return LZ4_decompress_safe_withSmallPrefix(source, dest, compressedSize, maxOutputSize, (size_t)dictSize); + } + assert(dictSize >= 0); + return LZ4_decompress_safe_forceExtDict(source, dest, compressedSize, maxOutputSize, dictStart, (size_t)dictSize); +} + +int LZ4_decompress_safe_partial_usingDict(const char* source, char* dest, int compressedSize, int targetOutputSize, int dstCapacity, const char* dictStart, int dictSize) +{ + if (dictSize==0) + return LZ4_decompress_safe_partial(source, dest, compressedSize, targetOutputSize, dstCapacity); + if (dictStart+dictSize == dest) { + if (dictSize >= 64 KB - 1) { + return LZ4_decompress_safe_partial_withPrefix64k(source, dest, compressedSize, targetOutputSize, dstCapacity); + } + assert(dictSize >= 0); + return LZ4_decompress_safe_partial_withSmallPrefix(source, dest, compressedSize, targetOutputSize, dstCapacity, (size_t)dictSize); + } + assert(dictSize >= 0); + return LZ4_decompress_safe_partial_forceExtDict(source, dest, compressedSize, targetOutputSize, dstCapacity, dictStart, (size_t)dictSize); +} + +int LZ4_decompress_fast_usingDict(const char* source, char* dest, int originalSize, const char* dictStart, int dictSize) +{ + if (dictSize==0 || dictStart+dictSize == dest) + return LZ4_decompress_unsafe_generic( + (const BYTE*)source, (BYTE*)dest, originalSize, + (size_t)dictSize, NULL, 0); + assert(dictSize >= 0); + return LZ4_decompress_fast_extDict(source, dest, originalSize, dictStart, (size_t)dictSize); +} + + +/*=************************************************* +* Obsolete Functions +***************************************************/ +/* obsolete compression functions */ +int LZ4_compress_limitedOutput(const char* source, char* dest, int inputSize, int maxOutputSize) +{ + return LZ4_compress_default(source, dest, inputSize, maxOutputSize); +} +int LZ4_compress(const char* src, char* dest, int srcSize) +{ + return LZ4_compress_default(src, dest, srcSize, LZ4_compressBound(srcSize)); +} +int LZ4_compress_limitedOutput_withState (void* state, const char* src, char* dst, int srcSize, int dstSize) +{ + return LZ4_compress_fast_extState(state, src, dst, srcSize, dstSize, 1); +} +int LZ4_compress_withState (void* state, const char* src, char* dst, int srcSize) +{ + return LZ4_compress_fast_extState(state, src, dst, srcSize, LZ4_compressBound(srcSize), 1); +} +int LZ4_compress_limitedOutput_continue (LZ4_stream_t* LZ4_stream, const char* src, char* dst, int srcSize, int dstCapacity) +{ + return LZ4_compress_fast_continue(LZ4_stream, src, dst, srcSize, dstCapacity, 1); +} +int LZ4_compress_continue (LZ4_stream_t* LZ4_stream, const char* source, char* dest, int inputSize) +{ + return LZ4_compress_fast_continue(LZ4_stream, source, dest, inputSize, LZ4_compressBound(inputSize), 1); +} + +/* +These decompression functions are deprecated and should no longer be used. +They are only provided here for compatibility with older user programs. +- LZ4_uncompress is totally equivalent to LZ4_decompress_fast +- LZ4_uncompress_unknownOutputSize is totally equivalent to LZ4_decompress_safe +*/ +int LZ4_uncompress (const char* source, char* dest, int outputSize) +{ + return LZ4_decompress_fast(source, dest, outputSize); +} +int LZ4_uncompress_unknownOutputSize (const char* source, char* dest, int isize, int maxOutputSize) +{ + return LZ4_decompress_safe(source, dest, isize, maxOutputSize); +} + +/* Obsolete Streaming functions */ + +int LZ4_sizeofStreamState(void) { return sizeof(LZ4_stream_t); } + +int LZ4_resetStreamState(void* state, char* inputBuffer) +{ + (void)inputBuffer; + LZ4_resetStream((LZ4_stream_t*)state); + return 0; +} + +#if !defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION) +void* LZ4_create (char* inputBuffer) +{ + (void)inputBuffer; + return LZ4_createStream(); +} +#endif + +char* LZ4_slideInputBuffer (void* state) +{ + /* avoid const char * -> char * conversion warning */ + return (char *)(uptrval)((LZ4_stream_t*)state)->internal_donotuse.dictionary; +} + +#endif /* LZ4_COMMONDEFS_ONLY */ diff --git a/example/android/third_party/lz4/include/lz4.h b/example/android/third_party/lz4/include/lz4.h new file mode 100644 index 00000000..5c799723 --- /dev/null +++ b/example/android/third_party/lz4/include/lz4.h @@ -0,0 +1,879 @@ +/* + * LZ4 - Fast LZ compression algorithm + * Header File + * Copyright (C) 2011-2023, Yann Collet. + + BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) + + Redistribution and use in source and binary forms, with or without + modification, are permitted provided that the following conditions are + met: + + * Redistributions of source code must retain the above copyright + notice, this list of conditions and the following disclaimer. + * Redistributions in binary form must reproduce the above + copyright notice, this list of conditions and the following disclaimer + in the documentation and/or other materials provided with the + distribution. + + THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + + You can contact the author at : + - LZ4 homepage : http://www.lz4.org + - LZ4 source repository : https://github.com/lz4/lz4 +*/ +#if defined (__cplusplus) +extern "C" { +#endif + +#ifndef LZ4_H_2983827168210 +#define LZ4_H_2983827168210 + +/* --- Dependency --- */ +#include /* size_t */ + + +/** + Introduction + + LZ4 is lossless compression algorithm, providing compression speed >500 MB/s per core, + scalable with multi-cores CPU. It features an extremely fast decoder, with speed in + multiple GB/s per core, typically reaching RAM speed limits on multi-core systems. + + The LZ4 compression library provides in-memory compression and decompression functions. + It gives full buffer control to user. + Compression can be done in: + - a single step (described as Simple Functions) + - a single step, reusing a context (described in Advanced Functions) + - unbounded multiple steps (described as Streaming compression) + + lz4.h generates and decodes LZ4-compressed blocks (doc/lz4_Block_format.md). + Decompressing such a compressed block requires additional metadata. + Exact metadata depends on exact decompression function. + For the typical case of LZ4_decompress_safe(), + metadata includes block's compressed size, and maximum bound of decompressed size. + Each application is free to encode and pass such metadata in whichever way it wants. + + lz4.h only handle blocks, it can not generate Frames. + + Blocks are different from Frames (doc/lz4_Frame_format.md). + Frames bundle both blocks and metadata in a specified manner. + Embedding metadata is required for compressed data to be self-contained and portable. + Frame format is delivered through a companion API, declared in lz4frame.h. + The `lz4` CLI can only manage frames. +*/ + +/*^*************************************************************** +* Export parameters +*****************************************************************/ +/* +* LZ4_DLL_EXPORT : +* Enable exporting of functions when building a Windows DLL +* LZ4LIB_VISIBILITY : +* Control library symbols visibility. +*/ +#ifndef LZ4LIB_VISIBILITY +# if defined(__GNUC__) && (__GNUC__ >= 4) +# define LZ4LIB_VISIBILITY __attribute__ ((visibility ("default"))) +# else +# define LZ4LIB_VISIBILITY +# endif +#endif +#if defined(LZ4_DLL_EXPORT) && (LZ4_DLL_EXPORT==1) +# define LZ4LIB_API __declspec(dllexport) LZ4LIB_VISIBILITY +#elif defined(LZ4_DLL_IMPORT) && (LZ4_DLL_IMPORT==1) +# define LZ4LIB_API __declspec(dllimport) LZ4LIB_VISIBILITY /* It isn't required but allows to generate better code, saving a function pointer load from the IAT and an indirect jump.*/ +#else +# define LZ4LIB_API LZ4LIB_VISIBILITY +#endif + +/*! LZ4_FREESTANDING : + * When this macro is set to 1, it enables "freestanding mode" that is + * suitable for typical freestanding environment which doesn't support + * standard C library. + * + * - LZ4_FREESTANDING is a compile-time switch. + * - It requires the following macros to be defined: + * LZ4_memcpy, LZ4_memmove, LZ4_memset. + * - It only enables LZ4/HC functions which don't use heap. + * All LZ4F_* functions are not supported. + * - See tests/freestanding.c to check its basic setup. + */ +#if defined(LZ4_FREESTANDING) && (LZ4_FREESTANDING == 1) +# define LZ4_HEAPMODE 0 +# define LZ4HC_HEAPMODE 0 +# define LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION 1 +# if !defined(LZ4_memcpy) +# error "LZ4_FREESTANDING requires macro 'LZ4_memcpy'." +# endif +# if !defined(LZ4_memset) +# error "LZ4_FREESTANDING requires macro 'LZ4_memset'." +# endif +# if !defined(LZ4_memmove) +# error "LZ4_FREESTANDING requires macro 'LZ4_memmove'." +# endif +#elif ! defined(LZ4_FREESTANDING) +# define LZ4_FREESTANDING 0 +#endif + + +/*------ Version ------*/ +#define LZ4_VERSION_MAJOR 1 /* for breaking interface changes */ +#define LZ4_VERSION_MINOR 9 /* for new (non-breaking) interface capabilities */ +#define LZ4_VERSION_RELEASE 5 /* for tweaks, bug-fixes, or development */ + +#define LZ4_VERSION_NUMBER (LZ4_VERSION_MAJOR *100*100 + LZ4_VERSION_MINOR *100 + LZ4_VERSION_RELEASE) + +#define LZ4_LIB_VERSION LZ4_VERSION_MAJOR.LZ4_VERSION_MINOR.LZ4_VERSION_RELEASE +#define LZ4_QUOTE(str) #str +#define LZ4_EXPAND_AND_QUOTE(str) LZ4_QUOTE(str) +#define LZ4_VERSION_STRING LZ4_EXPAND_AND_QUOTE(LZ4_LIB_VERSION) /* requires v1.7.3+ */ + +LZ4LIB_API int LZ4_versionNumber (void); /**< library version number; useful to check dll version; requires v1.3.0+ */ +LZ4LIB_API const char* LZ4_versionString (void); /**< library version string; useful to check dll version; requires v1.7.5+ */ + + +/*-************************************ +* Tuning memory usage +**************************************/ +/*! + * LZ4_MEMORY_USAGE : + * Can be selected at compile time, by setting LZ4_MEMORY_USAGE. + * Memory usage formula : N->2^N Bytes (examples : 10 -> 1KB; 12 -> 4KB ; 16 -> 64KB; 20 -> 1MB) + * Increasing memory usage improves compression ratio, generally at the cost of speed. + * Reduced memory usage may improve speed at the cost of ratio, thanks to better cache locality. + * Default value is 14, for 16KB, which nicely fits into most L1 caches. + */ +#ifndef LZ4_MEMORY_USAGE +# define LZ4_MEMORY_USAGE LZ4_MEMORY_USAGE_DEFAULT +#endif + +/* These are absolute limits, they should not be changed by users */ +#define LZ4_MEMORY_USAGE_MIN 10 +#define LZ4_MEMORY_USAGE_DEFAULT 14 +#define LZ4_MEMORY_USAGE_MAX 20 + +#if (LZ4_MEMORY_USAGE < LZ4_MEMORY_USAGE_MIN) +# error "LZ4_MEMORY_USAGE is too small !" +#endif + +#if (LZ4_MEMORY_USAGE > LZ4_MEMORY_USAGE_MAX) +# error "LZ4_MEMORY_USAGE is too large !" +#endif + +/*-************************************ +* Simple Functions +**************************************/ +/*! LZ4_compress_default() : + * Compresses 'srcSize' bytes from buffer 'src' + * into already allocated 'dst' buffer of size 'dstCapacity'. + * Compression is guaranteed to succeed if 'dstCapacity' >= LZ4_compressBound(srcSize). + * It also runs faster, so it's a recommended setting. + * If the function cannot compress 'src' into a more limited 'dst' budget, + * compression stops *immediately*, and the function result is zero. + * In which case, 'dst' content is undefined (invalid). + * srcSize : max supported value is LZ4_MAX_INPUT_SIZE. + * dstCapacity : size of buffer 'dst' (which must be already allocated) + * @return : the number of bytes written into buffer 'dst' (necessarily <= dstCapacity) + * or 0 if compression fails + * Note : This function is protected against buffer overflow scenarios (never writes outside 'dst' buffer, nor read outside 'source' buffer). + */ +LZ4LIB_API int LZ4_compress_default(const char* src, char* dst, int srcSize, int dstCapacity); + +/*! LZ4_decompress_safe() : + * @compressedSize : is the exact complete size of the compressed block. + * @dstCapacity : is the size of destination buffer (which must be already allocated), + * presumed an upper bound of decompressed size. + * @return : the number of bytes decompressed into destination buffer (necessarily <= dstCapacity) + * If destination buffer is not large enough, decoding will stop and output an error code (negative value). + * If the source stream is detected malformed, the function will stop decoding and return a negative result. + * Note 1 : This function is protected against malicious data packets : + * it will never writes outside 'dst' buffer, nor read outside 'source' buffer, + * even if the compressed block is maliciously modified to order the decoder to do these actions. + * In such case, the decoder stops immediately, and considers the compressed block malformed. + * Note 2 : compressedSize and dstCapacity must be provided to the function, the compressed block does not contain them. + * The implementation is free to send / store / derive this information in whichever way is most beneficial. + * If there is a need for a different format which bundles together both compressed data and its metadata, consider looking at lz4frame.h instead. + */ +LZ4LIB_API int LZ4_decompress_safe (const char* src, char* dst, int compressedSize, int dstCapacity); + + +/*-************************************ +* Advanced Functions +**************************************/ +#define LZ4_MAX_INPUT_SIZE 0x7E000000 /* 2 113 929 216 bytes */ +#define LZ4_COMPRESSBOUND(isize) ((unsigned)(isize) > (unsigned)LZ4_MAX_INPUT_SIZE ? 0 : (isize) + ((isize)/255) + 16) + +/*! LZ4_compressBound() : + Provides the maximum size that LZ4 compression may output in a "worst case" scenario (input data not compressible) + This function is primarily useful for memory allocation purposes (destination buffer size). + Macro LZ4_COMPRESSBOUND() is also provided for compilation-time evaluation (stack memory allocation for example). + Note that LZ4_compress_default() compresses faster when dstCapacity is >= LZ4_compressBound(srcSize) + inputSize : max supported value is LZ4_MAX_INPUT_SIZE + return : maximum output size in a "worst case" scenario + or 0, if input size is incorrect (too large or negative) +*/ +LZ4LIB_API int LZ4_compressBound(int inputSize); + +/*! LZ4_compress_fast() : + Same as LZ4_compress_default(), but allows selection of "acceleration" factor. + The larger the acceleration value, the faster the algorithm, but also the lesser the compression. + It's a trade-off. It can be fine tuned, with each successive value providing roughly +~3% to speed. + An acceleration value of "1" is the same as regular LZ4_compress_default() + Values <= 0 will be replaced by LZ4_ACCELERATION_DEFAULT (currently == 1, see lz4.c). + Values > LZ4_ACCELERATION_MAX will be replaced by LZ4_ACCELERATION_MAX (currently == 65537, see lz4.c). +*/ +LZ4LIB_API int LZ4_compress_fast (const char* src, char* dst, int srcSize, int dstCapacity, int acceleration); + + +/*! LZ4_compress_fast_extState() : + * Same as LZ4_compress_fast(), using an externally allocated memory space for its state. + * Use LZ4_sizeofState() to know how much memory must be allocated, + * and allocate it on 8-bytes boundaries (using `malloc()` typically). + * Then, provide this buffer as `void* state` to compression function. + */ +LZ4LIB_API int LZ4_sizeofState(void); +LZ4LIB_API int LZ4_compress_fast_extState (void* state, const char* src, char* dst, int srcSize, int dstCapacity, int acceleration); + +/*! LZ4_compress_destSize() : + * Reverse the logic : compresses as much data as possible from 'src' buffer + * into already allocated buffer 'dst', of size >= 'dstCapacity'. + * This function either compresses the entire 'src' content into 'dst' if it's large enough, + * or fill 'dst' buffer completely with as much data as possible from 'src'. + * note: acceleration parameter is fixed to "default". + * + * *srcSizePtr : in+out parameter. Initially contains size of input. + * Will be modified to indicate how many bytes where read from 'src' to fill 'dst'. + * New value is necessarily <= input value. + * @return : Nb bytes written into 'dst' (necessarily <= dstCapacity) + * or 0 if compression fails. + * + * Note : from v1.8.2 to v1.9.1, this function had a bug (fixed in v1.9.2+): + * the produced compressed content could, in specific circumstances, + * require to be decompressed into a destination buffer larger + * by at least 1 byte than the content to decompress. + * If an application uses `LZ4_compress_destSize()`, + * it's highly recommended to update liblz4 to v1.9.2 or better. + * If this can't be done or ensured, + * the receiving decompression function should provide + * a dstCapacity which is > decompressedSize, by at least 1 byte. + * See https://github.com/lz4/lz4/issues/859 for details + */ +LZ4LIB_API int LZ4_compress_destSize(const char* src, char* dst, int* srcSizePtr, int targetDstSize); + +/*! LZ4_decompress_safe_partial() : + * Decompress an LZ4 compressed block, of size 'srcSize' at position 'src', + * into destination buffer 'dst' of size 'dstCapacity'. + * Up to 'targetOutputSize' bytes will be decoded. + * The function stops decoding on reaching this objective. + * This can be useful to boost performance + * whenever only the beginning of a block is required. + * + * @return : the number of bytes decoded in `dst` (necessarily <= targetOutputSize) + * If source stream is detected malformed, function returns a negative result. + * + * Note 1 : @return can be < targetOutputSize, if compressed block contains less data. + * + * Note 2 : targetOutputSize must be <= dstCapacity + * + * Note 3 : this function effectively stops decoding on reaching targetOutputSize, + * so dstCapacity is kind of redundant. + * This is because in older versions of this function, + * decoding operation would still write complete sequences. + * Therefore, there was no guarantee that it would stop writing at exactly targetOutputSize, + * it could write more bytes, though only up to dstCapacity. + * Some "margin" used to be required for this operation to work properly. + * Thankfully, this is no longer necessary. + * The function nonetheless keeps the same signature, in an effort to preserve API compatibility. + * + * Note 4 : If srcSize is the exact size of the block, + * then targetOutputSize can be any value, + * including larger than the block's decompressed size. + * The function will, at most, generate block's decompressed size. + * + * Note 5 : If srcSize is _larger_ than block's compressed size, + * then targetOutputSize **MUST** be <= block's decompressed size. + * Otherwise, *silent corruption will occur*. + */ +LZ4LIB_API int LZ4_decompress_safe_partial (const char* src, char* dst, int srcSize, int targetOutputSize, int dstCapacity); + + +/*-********************************************* +* Streaming Compression Functions +***********************************************/ +typedef union LZ4_stream_u LZ4_stream_t; /* incomplete type (defined later) */ + +/*! + Note about RC_INVOKED + + - RC_INVOKED is predefined symbol of rc.exe (the resource compiler which is part of MSVC/Visual Studio). + https://docs.microsoft.com/en-us/windows/win32/menurc/predefined-macros + + - Since rc.exe is a legacy compiler, it truncates long symbol (> 30 chars) + and reports warning "RC4011: identifier truncated". + + - To eliminate the warning, we surround long preprocessor symbol with + "#if !defined(RC_INVOKED) ... #endif" block that means + "skip this block when rc.exe is trying to read it". +*/ +#if !defined(RC_INVOKED) /* https://docs.microsoft.com/en-us/windows/win32/menurc/predefined-macros */ +#if !defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION) +LZ4LIB_API LZ4_stream_t* LZ4_createStream(void); +LZ4LIB_API int LZ4_freeStream (LZ4_stream_t* streamPtr); +#endif /* !defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION) */ +#endif + +/*! LZ4_resetStream_fast() : v1.9.0+ + * Use this to prepare an LZ4_stream_t for a new chain of dependent blocks + * (e.g., LZ4_compress_fast_continue()). + * + * An LZ4_stream_t must be initialized once before usage. + * This is automatically done when created by LZ4_createStream(). + * However, should the LZ4_stream_t be simply declared on stack (for example), + * it's necessary to initialize it first, using LZ4_initStream(). + * + * After init, start any new stream with LZ4_resetStream_fast(). + * A same LZ4_stream_t can be re-used multiple times consecutively + * and compress multiple streams, + * provided that it starts each new stream with LZ4_resetStream_fast(). + * + * LZ4_resetStream_fast() is much faster than LZ4_initStream(), + * but is not compatible with memory regions containing garbage data. + * + * Note: it's only useful to call LZ4_resetStream_fast() + * in the context of streaming compression. + * The *extState* functions perform their own resets. + * Invoking LZ4_resetStream_fast() before is redundant, and even counterproductive. + */ +LZ4LIB_API void LZ4_resetStream_fast (LZ4_stream_t* streamPtr); + +/*! LZ4_loadDict() : + * Use this function to reference a static dictionary into LZ4_stream_t. + * The dictionary must remain available during compression. + * LZ4_loadDict() triggers a reset, so any previous data will be forgotten. + * The same dictionary will have to be loaded on decompression side for successful decoding. + * Dictionary are useful for better compression of small data (KB range). + * While LZ4 itself accepts any input as dictionary, dictionary efficiency is also a topic. + * When in doubt, employ the Zstandard's Dictionary Builder. + * Loading a size of 0 is allowed, and is the same as reset. + * @return : loaded dictionary size, in bytes (note: only the last 64 KB are loaded) + */ +LZ4LIB_API int LZ4_loadDict (LZ4_stream_t* streamPtr, const char* dictionary, int dictSize); + +/*! LZ4_loadDictSlow() : v1.9.5+ + * Same as LZ4_loadDict(), + * but uses a bit more cpu to reference the dictionary content more thoroughly. + * This is expected to slightly improve compression ratio. + * The extra-cpu cost is likely worth it if the dictionary is re-used across multiple sessions. + * @return : loaded dictionary size, in bytes (note: only the last 64 KB are loaded) + */ +LZ4LIB_API int LZ4_loadDictSlow(LZ4_stream_t* streamPtr, const char* dictionary, int dictSize); + +/*! LZ4_compress_fast_continue() : + * Compress 'src' content using data from previously compressed blocks, for better compression ratio. + * 'dst' buffer must be already allocated. + * If dstCapacity >= LZ4_compressBound(srcSize), compression is guaranteed to succeed, and runs faster. + * + * @return : size of compressed block + * or 0 if there is an error (typically, cannot fit into 'dst'). + * + * Note 1 : Each invocation to LZ4_compress_fast_continue() generates a new block. + * Each block has precise boundaries. + * Each block must be decompressed separately, calling LZ4_decompress_*() with relevant metadata. + * It's not possible to append blocks together and expect a single invocation of LZ4_decompress_*() to decompress them together. + * + * Note 2 : The previous 64KB of source data is __assumed__ to remain present, unmodified, at same address in memory ! + * + * Note 3 : When input is structured as a double-buffer, each buffer can have any size, including < 64 KB. + * Make sure that buffers are separated, by at least one byte. + * This construction ensures that each block only depends on previous block. + * + * Note 4 : If input buffer is a ring-buffer, it can have any size, including < 64 KB. + * + * Note 5 : After an error, the stream status is undefined (invalid), it can only be reset or freed. + */ +LZ4LIB_API int LZ4_compress_fast_continue (LZ4_stream_t* streamPtr, const char* src, char* dst, int srcSize, int dstCapacity, int acceleration); + +/*! LZ4_saveDict() : + * If last 64KB data cannot be guaranteed to remain available at its current memory location, + * save it into a safer place (char* safeBuffer). + * This is schematically equivalent to a memcpy() followed by LZ4_loadDict(), + * but is much faster, because LZ4_saveDict() doesn't need to rebuild tables. + * @return : saved dictionary size in bytes (necessarily <= maxDictSize), or 0 if error. + */ +LZ4LIB_API int LZ4_saveDict (LZ4_stream_t* streamPtr, char* safeBuffer, int maxDictSize); + + +/*-********************************************** +* Streaming Decompression Functions +* Bufferless synchronous API +************************************************/ +typedef union LZ4_streamDecode_u LZ4_streamDecode_t; /* tracking context */ + +/*! LZ4_createStreamDecode() and LZ4_freeStreamDecode() : + * creation / destruction of streaming decompression tracking context. + * A tracking context can be re-used multiple times. + */ +#if !defined(RC_INVOKED) /* https://docs.microsoft.com/en-us/windows/win32/menurc/predefined-macros */ +#if !defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION) +LZ4LIB_API LZ4_streamDecode_t* LZ4_createStreamDecode(void); +LZ4LIB_API int LZ4_freeStreamDecode (LZ4_streamDecode_t* LZ4_stream); +#endif /* !defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION) */ +#endif + +/*! LZ4_setStreamDecode() : + * An LZ4_streamDecode_t context can be allocated once and re-used multiple times. + * Use this function to start decompression of a new stream of blocks. + * A dictionary can optionally be set. Use NULL or size 0 for a reset order. + * Dictionary is presumed stable : it must remain accessible and unmodified during next decompression. + * @return : 1 if OK, 0 if error + */ +LZ4LIB_API int LZ4_setStreamDecode (LZ4_streamDecode_t* LZ4_streamDecode, const char* dictionary, int dictSize); + +/*! LZ4_decoderRingBufferSize() : v1.8.2+ + * Note : in a ring buffer scenario (optional), + * blocks are presumed decompressed next to each other + * up to the moment there is not enough remaining space for next block (remainingSize < maxBlockSize), + * at which stage it resumes from beginning of ring buffer. + * When setting such a ring buffer for streaming decompression, + * provides the minimum size of this ring buffer + * to be compatible with any source respecting maxBlockSize condition. + * @return : minimum ring buffer size, + * or 0 if there is an error (invalid maxBlockSize). + */ +LZ4LIB_API int LZ4_decoderRingBufferSize(int maxBlockSize); +#define LZ4_DECODER_RING_BUFFER_SIZE(maxBlockSize) (65536 + 14 + (maxBlockSize)) /* for static allocation; maxBlockSize presumed valid */ + +/*! LZ4_decompress_safe_continue() : + * This decoding function allows decompression of consecutive blocks in "streaming" mode. + * The difference with the usual independent blocks is that + * new blocks are allowed to find references into former blocks. + * A block is an unsplittable entity, and must be presented entirely to the decompression function. + * LZ4_decompress_safe_continue() only accepts one block at a time. + * It's modeled after `LZ4_decompress_safe()` and behaves similarly. + * + * @LZ4_streamDecode : decompression state, tracking the position in memory of past data + * @compressedSize : exact complete size of one compressed block. + * @dstCapacity : size of destination buffer (which must be already allocated), + * must be an upper bound of decompressed size. + * @return : number of bytes decompressed into destination buffer (necessarily <= dstCapacity) + * If destination buffer is not large enough, decoding will stop and output an error code (negative value). + * If the source stream is detected malformed, the function will stop decoding and return a negative result. + * + * The last 64KB of previously decoded data *must* remain available and unmodified + * at the memory position where they were previously decoded. + * If less than 64KB of data has been decoded, all the data must be present. + * + * Special : if decompression side sets a ring buffer, it must respect one of the following conditions : + * - Decompression buffer size is _at least_ LZ4_decoderRingBufferSize(maxBlockSize). + * maxBlockSize is the maximum size of any single block. It can have any value > 16 bytes. + * In which case, encoding and decoding buffers do not need to be synchronized. + * Actually, data can be produced by any source compliant with LZ4 format specification, and respecting maxBlockSize. + * - Synchronized mode : + * Decompression buffer size is _exactly_ the same as compression buffer size, + * and follows exactly same update rule (block boundaries at same positions), + * and decoding function is provided with exact decompressed size of each block (exception for last block of the stream), + * _then_ decoding & encoding ring buffer can have any size, including small ones ( < 64 KB). + * - Decompression buffer is larger than encoding buffer, by a minimum of maxBlockSize more bytes. + * In which case, encoding and decoding buffers do not need to be synchronized, + * and encoding ring buffer can have any size, including small ones ( < 64 KB). + * + * Whenever these conditions are not possible, + * save the last 64KB of decoded data into a safe buffer where it can't be modified during decompression, + * then indicate where this data is saved using LZ4_setStreamDecode(), before decompressing next block. +*/ +LZ4LIB_API int +LZ4_decompress_safe_continue (LZ4_streamDecode_t* LZ4_streamDecode, + const char* src, char* dst, + int srcSize, int dstCapacity); + + +/*! LZ4_decompress_safe_usingDict() : + * Works the same as + * a combination of LZ4_setStreamDecode() followed by LZ4_decompress_safe_continue() + * However, it's stateless: it doesn't need any LZ4_streamDecode_t state. + * Dictionary is presumed stable : it must remain accessible and unmodified during decompression. + * Performance tip : Decompression speed can be substantially increased + * when dst == dictStart + dictSize. + */ +LZ4LIB_API int +LZ4_decompress_safe_usingDict(const char* src, char* dst, + int srcSize, int dstCapacity, + const char* dictStart, int dictSize); + +/*! LZ4_decompress_safe_partial_usingDict() : + * Behaves the same as LZ4_decompress_safe_partial() + * with the added ability to specify a memory segment for past data. + * Performance tip : Decompression speed can be substantially increased + * when dst == dictStart + dictSize. + */ +LZ4LIB_API int +LZ4_decompress_safe_partial_usingDict(const char* src, char* dst, + int compressedSize, + int targetOutputSize, int maxOutputSize, + const char* dictStart, int dictSize); + +#endif /* LZ4_H_2983827168210 */ + + +/*^************************************* + * !!!!!! STATIC LINKING ONLY !!!!!! + ***************************************/ + +/*-**************************************************************************** + * Experimental section + * + * Symbols declared in this section must be considered unstable. Their + * signatures or semantics may change, or they may be removed altogether in the + * future. They are therefore only safe to depend on when the caller is + * statically linked against the library. + * + * To protect against unsafe usage, not only are the declarations guarded, + * the definitions are hidden by default + * when building LZ4 as a shared/dynamic library. + * + * In order to access these declarations, + * define LZ4_STATIC_LINKING_ONLY in your application + * before including LZ4's headers. + * + * In order to make their implementations accessible dynamically, you must + * define LZ4_PUBLISH_STATIC_FUNCTIONS when building the LZ4 library. + ******************************************************************************/ + +#ifdef LZ4_STATIC_LINKING_ONLY + +#ifndef LZ4_STATIC_3504398509 +#define LZ4_STATIC_3504398509 + +#ifdef LZ4_PUBLISH_STATIC_FUNCTIONS +# define LZ4LIB_STATIC_API LZ4LIB_API +#else +# define LZ4LIB_STATIC_API +#endif + + +/*! LZ4_compress_fast_extState_fastReset() : + * A variant of LZ4_compress_fast_extState(). + * + * Using this variant avoids an expensive initialization step. + * It is only safe to call if the state buffer is known to be correctly initialized already + * (see above comment on LZ4_resetStream_fast() for a definition of "correctly initialized"). + * From a high level, the difference is that + * this function initializes the provided state with a call to something like LZ4_resetStream_fast() + * while LZ4_compress_fast_extState() starts with a call to LZ4_resetStream(). + */ +LZ4LIB_STATIC_API int LZ4_compress_fast_extState_fastReset (void* state, const char* src, char* dst, int srcSize, int dstCapacity, int acceleration); + +/*! LZ4_compress_destSize_extState() : + * Same as LZ4_compress_destSize(), but using an externally allocated state. + * Also: exposes @acceleration + */ +int LZ4_compress_destSize_extState(void* state, const char* src, char* dst, int* srcSizePtr, int targetDstSize, int acceleration); + +/*! LZ4_attach_dictionary() : + * This is an experimental API that allows + * efficient use of a static dictionary many times. + * + * Rather than re-loading the dictionary buffer into a working context before + * each compression, or copying a pre-loaded dictionary's LZ4_stream_t into a + * working LZ4_stream_t, this function introduces a no-copy setup mechanism, + * in which the working stream references the dictionary stream in-place. + * + * Several assumptions are made about the state of the dictionary stream. + * Currently, only streams which have been prepared by LZ4_loadDict() should + * be expected to work. + * + * Alternatively, the provided dictionaryStream may be NULL, + * in which case any existing dictionary stream is unset. + * + * If a dictionary is provided, it replaces any pre-existing stream history. + * The dictionary contents are the only history that can be referenced and + * logically immediately precede the data compressed in the first subsequent + * compression call. + * + * The dictionary will only remain attached to the working stream through the + * first compression call, at the end of which it is cleared. The dictionary + * stream (and source buffer) must remain in-place / accessible / unchanged + * through the completion of the first compression call on the stream. + */ +LZ4LIB_STATIC_API void +LZ4_attach_dictionary(LZ4_stream_t* workingStream, + const LZ4_stream_t* dictionaryStream); + + +/*! In-place compression and decompression + * + * It's possible to have input and output sharing the same buffer, + * for highly constrained memory environments. + * In both cases, it requires input to lay at the end of the buffer, + * and decompression to start at beginning of the buffer. + * Buffer size must feature some margin, hence be larger than final size. + * + * |<------------------------buffer--------------------------------->| + * |<-----------compressed data--------->| + * |<-----------decompressed size------------------>| + * |<----margin---->| + * + * This technique is more useful for decompression, + * since decompressed size is typically larger, + * and margin is short. + * + * In-place decompression will work inside any buffer + * which size is >= LZ4_DECOMPRESS_INPLACE_BUFFER_SIZE(decompressedSize). + * This presumes that decompressedSize > compressedSize. + * Otherwise, it means compression actually expanded data, + * and it would be more efficient to store such data with a flag indicating it's not compressed. + * This can happen when data is not compressible (already compressed, or encrypted). + * + * For in-place compression, margin is larger, as it must be able to cope with both + * history preservation, requiring input data to remain unmodified up to LZ4_DISTANCE_MAX, + * and data expansion, which can happen when input is not compressible. + * As a consequence, buffer size requirements are much higher, + * and memory savings offered by in-place compression are more limited. + * + * There are ways to limit this cost for compression : + * - Reduce history size, by modifying LZ4_DISTANCE_MAX. + * Note that it is a compile-time constant, so all compressions will apply this limit. + * Lower values will reduce compression ratio, except when input_size < LZ4_DISTANCE_MAX, + * so it's a reasonable trick when inputs are known to be small. + * - Require the compressor to deliver a "maximum compressed size". + * This is the `dstCapacity` parameter in `LZ4_compress*()`. + * When this size is < LZ4_COMPRESSBOUND(inputSize), then compression can fail, + * in which case, the return code will be 0 (zero). + * The caller must be ready for these cases to happen, + * and typically design a backup scheme to send data uncompressed. + * The combination of both techniques can significantly reduce + * the amount of margin required for in-place compression. + * + * In-place compression can work in any buffer + * which size is >= (maxCompressedSize) + * with maxCompressedSize == LZ4_COMPRESSBOUND(srcSize) for guaranteed compression success. + * LZ4_COMPRESS_INPLACE_BUFFER_SIZE() depends on both maxCompressedSize and LZ4_DISTANCE_MAX, + * so it's possible to reduce memory requirements by playing with them. + */ + +#define LZ4_DECOMPRESS_INPLACE_MARGIN(compressedSize) (((compressedSize) >> 8) + 32) +#define LZ4_DECOMPRESS_INPLACE_BUFFER_SIZE(decompressedSize) ((decompressedSize) + LZ4_DECOMPRESS_INPLACE_MARGIN(decompressedSize)) /**< note: presumes that compressedSize < decompressedSize. note2: margin is overestimated a bit, since it could use compressedSize instead */ + +#ifndef LZ4_DISTANCE_MAX /* history window size; can be user-defined at compile time */ +# define LZ4_DISTANCE_MAX 65535 /* set to maximum value by default */ +#endif + +#define LZ4_COMPRESS_INPLACE_MARGIN (LZ4_DISTANCE_MAX + 32) /* LZ4_DISTANCE_MAX can be safely replaced by srcSize when it's smaller */ +#define LZ4_COMPRESS_INPLACE_BUFFER_SIZE(maxCompressedSize) ((maxCompressedSize) + LZ4_COMPRESS_INPLACE_MARGIN) /**< maxCompressedSize is generally LZ4_COMPRESSBOUND(inputSize), but can be set to any lower value, with the risk that compression can fail (return code 0(zero)) */ + +#endif /* LZ4_STATIC_3504398509 */ +#endif /* LZ4_STATIC_LINKING_ONLY */ + + + +#ifndef LZ4_H_98237428734687 +#define LZ4_H_98237428734687 + +/*-************************************************************ + * Private Definitions + ************************************************************** + * Do not use these definitions directly. + * They are only exposed to allow static allocation of `LZ4_stream_t` and `LZ4_streamDecode_t`. + * Accessing members will expose user code to API and/or ABI break in future versions of the library. + **************************************************************/ +#define LZ4_HASHLOG (LZ4_MEMORY_USAGE-2) +#define LZ4_HASHTABLESIZE (1 << LZ4_MEMORY_USAGE) +#define LZ4_HASH_SIZE_U32 (1 << LZ4_HASHLOG) /* required as macro for static allocation */ + +#if defined(__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) +# include + typedef int8_t LZ4_i8; + typedef uint8_t LZ4_byte; + typedef uint16_t LZ4_u16; + typedef uint32_t LZ4_u32; +#else + typedef signed char LZ4_i8; + typedef unsigned char LZ4_byte; + typedef unsigned short LZ4_u16; + typedef unsigned int LZ4_u32; +#endif + +/*! LZ4_stream_t : + * Never ever use below internal definitions directly ! + * These definitions are not API/ABI safe, and may change in future versions. + * If you need static allocation, declare or allocate an LZ4_stream_t object. +**/ + +typedef struct LZ4_stream_t_internal LZ4_stream_t_internal; +struct LZ4_stream_t_internal { + LZ4_u32 hashTable[LZ4_HASH_SIZE_U32]; + const LZ4_byte* dictionary; + const LZ4_stream_t_internal* dictCtx; + LZ4_u32 currentOffset; + LZ4_u32 tableType; + LZ4_u32 dictSize; + /* Implicit padding to ensure structure is aligned */ +}; + +#define LZ4_STREAM_MINSIZE ((1UL << (LZ4_MEMORY_USAGE)) + 32) /* static size, for inter-version compatibility */ +union LZ4_stream_u { + char minStateSize[LZ4_STREAM_MINSIZE]; + LZ4_stream_t_internal internal_donotuse; +}; /* previously typedef'd to LZ4_stream_t */ + + +/*! LZ4_initStream() : v1.9.0+ + * An LZ4_stream_t structure must be initialized at least once. + * This is automatically done when invoking LZ4_createStream(), + * but it's not when the structure is simply declared on stack (for example). + * + * Use LZ4_initStream() to properly initialize a newly declared LZ4_stream_t. + * It can also initialize any arbitrary buffer of sufficient size, + * and will @return a pointer of proper type upon initialization. + * + * Note : initialization fails if size and alignment conditions are not respected. + * In which case, the function will @return NULL. + * Note2: An LZ4_stream_t structure guarantees correct alignment and size. + * Note3: Before v1.9.0, use LZ4_resetStream() instead +**/ +LZ4LIB_API LZ4_stream_t* LZ4_initStream (void* stateBuffer, size_t size); + + +/*! LZ4_streamDecode_t : + * Never ever use below internal definitions directly ! + * These definitions are not API/ABI safe, and may change in future versions. + * If you need static allocation, declare or allocate an LZ4_streamDecode_t object. +**/ +typedef struct { + const LZ4_byte* externalDict; + const LZ4_byte* prefixEnd; + size_t extDictSize; + size_t prefixSize; +} LZ4_streamDecode_t_internal; + +#define LZ4_STREAMDECODE_MINSIZE 32 +union LZ4_streamDecode_u { + char minStateSize[LZ4_STREAMDECODE_MINSIZE]; + LZ4_streamDecode_t_internal internal_donotuse; +} ; /* previously typedef'd to LZ4_streamDecode_t */ + + + +/*-************************************ +* Obsolete Functions +**************************************/ + +/*! Deprecation warnings + * + * Deprecated functions make the compiler generate a warning when invoked. + * This is meant to invite users to update their source code. + * Should deprecation warnings be a problem, it is generally possible to disable them, + * typically with -Wno-deprecated-declarations for gcc + * or _CRT_SECURE_NO_WARNINGS in Visual. + * + * Another method is to define LZ4_DISABLE_DEPRECATE_WARNINGS + * before including the header file. + */ +#ifdef LZ4_DISABLE_DEPRECATE_WARNINGS +# define LZ4_DEPRECATED(message) /* disable deprecation warnings */ +#else +# if defined (__cplusplus) && (__cplusplus >= 201402) /* C++14 or greater */ +# define LZ4_DEPRECATED(message) [[deprecated(message)]] +# elif defined(_MSC_VER) +# define LZ4_DEPRECATED(message) __declspec(deprecated(message)) +# elif defined(__clang__) || (defined(__GNUC__) && (__GNUC__ * 10 + __GNUC_MINOR__ >= 45)) +# define LZ4_DEPRECATED(message) __attribute__((deprecated(message))) +# elif defined(__GNUC__) && (__GNUC__ * 10 + __GNUC_MINOR__ >= 31) +# define LZ4_DEPRECATED(message) __attribute__((deprecated)) +# else +# pragma message("WARNING: LZ4_DEPRECATED needs custom implementation for this compiler") +# define LZ4_DEPRECATED(message) /* disabled */ +# endif +#endif /* LZ4_DISABLE_DEPRECATE_WARNINGS */ + +/*! Obsolete compression functions (since v1.7.3) */ +LZ4_DEPRECATED("use LZ4_compress_default() instead") LZ4LIB_API int LZ4_compress (const char* src, char* dest, int srcSize); +LZ4_DEPRECATED("use LZ4_compress_default() instead") LZ4LIB_API int LZ4_compress_limitedOutput (const char* src, char* dest, int srcSize, int maxOutputSize); +LZ4_DEPRECATED("use LZ4_compress_fast_extState() instead") LZ4LIB_API int LZ4_compress_withState (void* state, const char* source, char* dest, int inputSize); +LZ4_DEPRECATED("use LZ4_compress_fast_extState() instead") LZ4LIB_API int LZ4_compress_limitedOutput_withState (void* state, const char* source, char* dest, int inputSize, int maxOutputSize); +LZ4_DEPRECATED("use LZ4_compress_fast_continue() instead") LZ4LIB_API int LZ4_compress_continue (LZ4_stream_t* LZ4_streamPtr, const char* source, char* dest, int inputSize); +LZ4_DEPRECATED("use LZ4_compress_fast_continue() instead") LZ4LIB_API int LZ4_compress_limitedOutput_continue (LZ4_stream_t* LZ4_streamPtr, const char* source, char* dest, int inputSize, int maxOutputSize); + +/*! Obsolete decompression functions (since v1.8.0) */ +LZ4_DEPRECATED("use LZ4_decompress_fast() instead") LZ4LIB_API int LZ4_uncompress (const char* source, char* dest, int outputSize); +LZ4_DEPRECATED("use LZ4_decompress_safe() instead") LZ4LIB_API int LZ4_uncompress_unknownOutputSize (const char* source, char* dest, int isize, int maxOutputSize); + +/* Obsolete streaming functions (since v1.7.0) + * degraded functionality; do not use! + * + * In order to perform streaming compression, these functions depended on data + * that is no longer tracked in the state. They have been preserved as well as + * possible: using them will still produce a correct output. However, they don't + * actually retain any history between compression calls. The compression ratio + * achieved will therefore be no better than compressing each chunk + * independently. + */ +LZ4_DEPRECATED("Use LZ4_createStream() instead") LZ4LIB_API void* LZ4_create (char* inputBuffer); +LZ4_DEPRECATED("Use LZ4_createStream() instead") LZ4LIB_API int LZ4_sizeofStreamState(void); +LZ4_DEPRECATED("Use LZ4_resetStream() instead") LZ4LIB_API int LZ4_resetStreamState(void* state, char* inputBuffer); +LZ4_DEPRECATED("Use LZ4_saveDict() instead") LZ4LIB_API char* LZ4_slideInputBuffer (void* state); + +/*! Obsolete streaming decoding functions (since v1.7.0) */ +LZ4_DEPRECATED("use LZ4_decompress_safe_usingDict() instead") LZ4LIB_API int LZ4_decompress_safe_withPrefix64k (const char* src, char* dst, int compressedSize, int maxDstSize); +LZ4_DEPRECATED("use LZ4_decompress_fast_usingDict() instead") LZ4LIB_API int LZ4_decompress_fast_withPrefix64k (const char* src, char* dst, int originalSize); + +/*! Obsolete LZ4_decompress_fast variants (since v1.9.0) : + * These functions used to be faster than LZ4_decompress_safe(), + * but this is no longer the case. They are now slower. + * This is because LZ4_decompress_fast() doesn't know the input size, + * and therefore must progress more cautiously into the input buffer to not read beyond the end of block. + * On top of that `LZ4_decompress_fast()` is not protected vs malformed or malicious inputs, making it a security liability. + * As a consequence, LZ4_decompress_fast() is strongly discouraged, and deprecated. + * + * The last remaining LZ4_decompress_fast() specificity is that + * it can decompress a block without knowing its compressed size. + * Such functionality can be achieved in a more secure manner + * by employing LZ4_decompress_safe_partial(). + * + * Parameters: + * originalSize : is the uncompressed size to regenerate. + * `dst` must be already allocated, its size must be >= 'originalSize' bytes. + * @return : number of bytes read from source buffer (== compressed size). + * The function expects to finish at block's end exactly. + * If the source stream is detected malformed, the function stops decoding and returns a negative result. + * note : LZ4_decompress_fast*() requires originalSize. Thanks to this information, it never writes past the output buffer. + * However, since it doesn't know its 'src' size, it may read an unknown amount of input, past input buffer bounds. + * Also, since match offsets are not validated, match reads from 'src' may underflow too. + * These issues never happen if input (compressed) data is correct. + * But they may happen if input data is invalid (error or intentional tampering). + * As a consequence, use these functions in trusted environments with trusted data **only**. + */ +LZ4_DEPRECATED("This function is deprecated and unsafe. Consider using LZ4_decompress_safe_partial() instead") +LZ4LIB_API int LZ4_decompress_fast (const char* src, char* dst, int originalSize); +LZ4_DEPRECATED("This function is deprecated and unsafe. Consider migrating towards LZ4_decompress_safe_continue() instead. " + "Note that the contract will change (requires block's compressed size, instead of decompressed size)") +LZ4LIB_API int LZ4_decompress_fast_continue (LZ4_streamDecode_t* LZ4_streamDecode, const char* src, char* dst, int originalSize); +LZ4_DEPRECATED("This function is deprecated and unsafe. Consider using LZ4_decompress_safe_partial_usingDict() instead") +LZ4LIB_API int LZ4_decompress_fast_usingDict (const char* src, char* dst, int originalSize, const char* dictStart, int dictSize); + +/*! LZ4_resetStream() : + * An LZ4_stream_t structure must be initialized at least once. + * This is done with LZ4_initStream(), or LZ4_resetStream(). + * Consider switching to LZ4_initStream(), + * invoking LZ4_resetStream() will trigger deprecation warnings in the future. + */ +LZ4LIB_API void LZ4_resetStream (LZ4_stream_t* streamPtr); + + +#endif /* LZ4_H_98237428734687 */ + + +#if defined (__cplusplus) +} +#endif diff --git a/example/android/third_party/lz4/include/lz4file.c b/example/android/third_party/lz4/include/lz4file.c new file mode 100644 index 00000000..a4197ea8 --- /dev/null +++ b/example/android/third_party/lz4/include/lz4file.c @@ -0,0 +1,341 @@ +/* + * LZ4 file library + * Copyright (C) 2022, Xiaomi Inc. + * + * BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions are + * met: + * + * - Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * - Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following disclaimer + * in the documentation and/or other materials provided with the + * distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * You can contact the author at : + * - LZ4 homepage : http://www.lz4.org + * - LZ4 source repository : https://github.com/lz4/lz4 + */ +#include /* malloc, free */ +#include +#include +#include "lz4.h" +#include "lz4file.h" + +static LZ4F_errorCode_t returnErrorCode(LZ4F_errorCodes code) +{ + return (LZ4F_errorCode_t)-(ptrdiff_t)code; +} +#undef RETURN_ERROR +#define RETURN_ERROR(e) return returnErrorCode(LZ4F_ERROR_ ## e) + +/* ===== read API ===== */ + +struct LZ4_readFile_s { + LZ4F_dctx* dctxPtr; + FILE* fp; + LZ4_byte* srcBuf; + size_t srcBufNext; + size_t srcBufSize; + size_t srcBufMaxSize; +}; + +static void LZ4F_freeReadFile(LZ4_readFile_t* lz4fRead) +{ + if (lz4fRead==NULL) return; + LZ4F_freeDecompressionContext(lz4fRead->dctxPtr); + free(lz4fRead->srcBuf); + free(lz4fRead); +} + +static void LZ4F_freeAndNullReadFile(LZ4_readFile_t** statePtr) +{ + assert(statePtr != NULL); + LZ4F_freeReadFile(*statePtr); + *statePtr = NULL; +} + +LZ4F_errorCode_t LZ4F_readOpen(LZ4_readFile_t** lz4fRead, FILE* fp) +{ + char buf[LZ4F_HEADER_SIZE_MAX]; + size_t consumedSize; + LZ4F_errorCode_t ret; + + if (fp == NULL || lz4fRead == NULL) { + RETURN_ERROR(parameter_null); + } + + *lz4fRead = (LZ4_readFile_t*)calloc(1, sizeof(LZ4_readFile_t)); + if (*lz4fRead == NULL) { + RETURN_ERROR(allocation_failed); + } + + ret = LZ4F_createDecompressionContext(&(*lz4fRead)->dctxPtr, LZ4F_VERSION); + if (LZ4F_isError(ret)) { + LZ4F_freeAndNullReadFile(lz4fRead); + return ret; + } + + (*lz4fRead)->fp = fp; + consumedSize = fread(buf, 1, sizeof(buf), (*lz4fRead)->fp); + if (consumedSize != sizeof(buf)) { + LZ4F_freeAndNullReadFile(lz4fRead); + RETURN_ERROR(io_read); + } + + { LZ4F_frameInfo_t info; + LZ4F_errorCode_t const r = LZ4F_getFrameInfo((*lz4fRead)->dctxPtr, &info, buf, &consumedSize); + if (LZ4F_isError(r)) { + LZ4F_freeAndNullReadFile(lz4fRead); + return r; + } + + switch (info.blockSizeID) { + case LZ4F_default : + case LZ4F_max64KB : + (*lz4fRead)->srcBufMaxSize = 64 * 1024; + break; + case LZ4F_max256KB: + (*lz4fRead)->srcBufMaxSize = 256 * 1024; + break; + case LZ4F_max1MB: + (*lz4fRead)->srcBufMaxSize = 1 * 1024 * 1024; + break; + case LZ4F_max4MB: + (*lz4fRead)->srcBufMaxSize = 4 * 1024 * 1024; + break; + default: + LZ4F_freeAndNullReadFile(lz4fRead); + RETURN_ERROR(maxBlockSize_invalid); + } + } + + (*lz4fRead)->srcBuf = (LZ4_byte*)malloc((*lz4fRead)->srcBufMaxSize); + if ((*lz4fRead)->srcBuf == NULL) { + LZ4F_freeAndNullReadFile(lz4fRead); + RETURN_ERROR(allocation_failed); + } + + (*lz4fRead)->srcBufSize = sizeof(buf) - consumedSize; + memcpy((*lz4fRead)->srcBuf, buf + consumedSize, (*lz4fRead)->srcBufSize); + + return ret; +} + +size_t LZ4F_read(LZ4_readFile_t* lz4fRead, void* buf, size_t size) +{ + LZ4_byte* p = (LZ4_byte*)buf; + size_t next = 0; + + if (lz4fRead == NULL || buf == NULL) + RETURN_ERROR(parameter_null); + + while (next < size) { + size_t srcsize = lz4fRead->srcBufSize - lz4fRead->srcBufNext; + size_t dstsize = size - next; + size_t ret; + + if (srcsize == 0) { + ret = fread(lz4fRead->srcBuf, 1, lz4fRead->srcBufMaxSize, lz4fRead->fp); + if (ret > 0) { + lz4fRead->srcBufSize = ret; + srcsize = lz4fRead->srcBufSize; + lz4fRead->srcBufNext = 0; + } else if (ret == 0) { + break; + } else { + RETURN_ERROR(io_read); + } + } + + ret = LZ4F_decompress(lz4fRead->dctxPtr, + p, &dstsize, + lz4fRead->srcBuf + lz4fRead->srcBufNext, + &srcsize, + NULL); + if (LZ4F_isError(ret)) { + return ret; + } + + lz4fRead->srcBufNext += srcsize; + next += dstsize; + p += dstsize; + } + + return next; +} + +LZ4F_errorCode_t LZ4F_readClose(LZ4_readFile_t* lz4fRead) +{ + if (lz4fRead == NULL) + RETURN_ERROR(parameter_null); + LZ4F_freeReadFile(lz4fRead); + return LZ4F_OK_NoError; +} + +/* ===== write API ===== */ + +struct LZ4_writeFile_s { + LZ4F_cctx* cctxPtr; + FILE* fp; + LZ4_byte* dstBuf; + size_t maxWriteSize; + size_t dstBufMaxSize; + LZ4F_errorCode_t errCode; +}; + +static void LZ4F_freeWriteFile(LZ4_writeFile_t* state) +{ + if (state == NULL) return; + LZ4F_freeCompressionContext(state->cctxPtr); + free(state->dstBuf); + free(state); +} + +static void LZ4F_freeAndNullWriteFile(LZ4_writeFile_t** statePtr) +{ + assert(statePtr != NULL); + LZ4F_freeWriteFile(*statePtr); + *statePtr = NULL; +} + +LZ4F_errorCode_t LZ4F_writeOpen(LZ4_writeFile_t** lz4fWrite, FILE* fp, const LZ4F_preferences_t* prefsPtr) +{ + LZ4_byte buf[LZ4F_HEADER_SIZE_MAX]; + size_t ret; + + if (fp == NULL || lz4fWrite == NULL) + RETURN_ERROR(parameter_null); + + *lz4fWrite = (LZ4_writeFile_t*)calloc(1, sizeof(LZ4_writeFile_t)); + if (*lz4fWrite == NULL) { + RETURN_ERROR(allocation_failed); + } + if (prefsPtr != NULL) { + switch (prefsPtr->frameInfo.blockSizeID) { + case LZ4F_default : + case LZ4F_max64KB : + (*lz4fWrite)->maxWriteSize = 64 * 1024; + break; + case LZ4F_max256KB: + (*lz4fWrite)->maxWriteSize = 256 * 1024; + break; + case LZ4F_max1MB: + (*lz4fWrite)->maxWriteSize = 1 * 1024 * 1024; + break; + case LZ4F_max4MB: + (*lz4fWrite)->maxWriteSize = 4 * 1024 * 1024; + break; + default: + LZ4F_freeAndNullWriteFile(lz4fWrite); + RETURN_ERROR(maxBlockSize_invalid); + } + } else { + (*lz4fWrite)->maxWriteSize = 64 * 1024; + } + + (*lz4fWrite)->dstBufMaxSize = LZ4F_compressBound((*lz4fWrite)->maxWriteSize, prefsPtr); + (*lz4fWrite)->dstBuf = (LZ4_byte*)malloc((*lz4fWrite)->dstBufMaxSize); + if ((*lz4fWrite)->dstBuf == NULL) { + LZ4F_freeAndNullWriteFile(lz4fWrite); + RETURN_ERROR(allocation_failed); + } + + ret = LZ4F_createCompressionContext(&(*lz4fWrite)->cctxPtr, LZ4F_VERSION); + if (LZ4F_isError(ret)) { + LZ4F_freeAndNullWriteFile(lz4fWrite); + return ret; + } + + ret = LZ4F_compressBegin((*lz4fWrite)->cctxPtr, buf, LZ4F_HEADER_SIZE_MAX, prefsPtr); + if (LZ4F_isError(ret)) { + LZ4F_freeAndNullWriteFile(lz4fWrite); + return ret; + } + + if (ret != fwrite(buf, 1, ret, fp)) { + LZ4F_freeAndNullWriteFile(lz4fWrite); + RETURN_ERROR(io_write); + } + + (*lz4fWrite)->fp = fp; + (*lz4fWrite)->errCode = LZ4F_OK_NoError; + return LZ4F_OK_NoError; +} + +size_t LZ4F_write(LZ4_writeFile_t* lz4fWrite, const void* buf, size_t size) +{ + const LZ4_byte* p = (const LZ4_byte*)buf; + size_t remain = size; + size_t chunk; + size_t ret; + + if (lz4fWrite == NULL || buf == NULL) + RETURN_ERROR(parameter_null); + while (remain) { + if (remain > lz4fWrite->maxWriteSize) + chunk = lz4fWrite->maxWriteSize; + else + chunk = remain; + + ret = LZ4F_compressUpdate(lz4fWrite->cctxPtr, + lz4fWrite->dstBuf, lz4fWrite->dstBufMaxSize, + p, chunk, + NULL); + if (LZ4F_isError(ret)) { + lz4fWrite->errCode = ret; + return ret; + } + + if (ret != fwrite(lz4fWrite->dstBuf, 1, ret, lz4fWrite->fp)) { + lz4fWrite->errCode = returnErrorCode(LZ4F_ERROR_io_write); + RETURN_ERROR(io_write); + } + + p += chunk; + remain -= chunk; + } + + return size; +} + +LZ4F_errorCode_t LZ4F_writeClose(LZ4_writeFile_t* lz4fWrite) +{ + LZ4F_errorCode_t ret = LZ4F_OK_NoError; + + if (lz4fWrite == NULL) { + RETURN_ERROR(parameter_null); + } + + if (lz4fWrite->errCode == LZ4F_OK_NoError) { + ret = LZ4F_compressEnd(lz4fWrite->cctxPtr, + lz4fWrite->dstBuf, lz4fWrite->dstBufMaxSize, + NULL); + if (LZ4F_isError(ret)) { + goto out; + } + + if (ret != fwrite(lz4fWrite->dstBuf, 1, ret, lz4fWrite->fp)) { + ret = returnErrorCode(LZ4F_ERROR_io_write); + } + } + +out: + LZ4F_freeWriteFile(lz4fWrite); + return ret; +} diff --git a/example/android/third_party/lz4/include/lz4file.h b/example/android/third_party/lz4/include/lz4file.h new file mode 100644 index 00000000..598ad705 --- /dev/null +++ b/example/android/third_party/lz4/include/lz4file.h @@ -0,0 +1,93 @@ +/* + LZ4 file library + Header File + Copyright (C) 2022, Xiaomi Inc. + BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) + + Redistribution and use in source and binary forms, with or without + modification, are permitted provided that the following conditions are + met: + + * Redistributions of source code must retain the above copyright + notice, this list of conditions and the following disclaimer. + * Redistributions in binary form must reproduce the above + copyright notice, this list of conditions and the following disclaimer + in the documentation and/or other materials provided with the + distribution. + + THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + + You can contact the author at : + - LZ4 source repository : https://github.com/lz4/lz4 + - LZ4 public forum : https://groups.google.com/forum/#!forum/lz4c +*/ +#if defined (__cplusplus) +extern "C" { +#endif + +#ifndef LZ4FILE_H +#define LZ4FILE_H + +#include /* FILE* */ +#include "lz4frame_static.h" + +typedef struct LZ4_readFile_s LZ4_readFile_t; +typedef struct LZ4_writeFile_s LZ4_writeFile_t; + +/*! LZ4F_readOpen() : + * Set read lz4file handle. + * `lz4f` will set a lz4file handle. + * `fp` must be the return value of the lz4 file opened by fopen. + */ +LZ4FLIB_STATIC_API LZ4F_errorCode_t LZ4F_readOpen(LZ4_readFile_t** lz4fRead, FILE* fp); + +/*! LZ4F_read() : + * Read lz4file content to buffer. + * `lz4f` must use LZ4_readOpen to set first. + * `buf` read data buffer. + * `size` read data buffer size. + */ +LZ4FLIB_STATIC_API size_t LZ4F_read(LZ4_readFile_t* lz4fRead, void* buf, size_t size); + +/*! LZ4F_readClose() : + * Close lz4file handle. + * `lz4f` must use LZ4_readOpen to set first. + */ +LZ4FLIB_STATIC_API LZ4F_errorCode_t LZ4F_readClose(LZ4_readFile_t* lz4fRead); + +/*! LZ4F_writeOpen() : + * Set write lz4file handle. + * `lz4f` will set a lz4file handle. + * `fp` must be the return value of the lz4 file opened by fopen. + */ +LZ4FLIB_STATIC_API LZ4F_errorCode_t LZ4F_writeOpen(LZ4_writeFile_t** lz4fWrite, FILE* fp, const LZ4F_preferences_t* prefsPtr); + +/*! LZ4F_write() : + * Write buffer to lz4file. + * `lz4f` must use LZ4F_writeOpen to set first. + * `buf` write data buffer. + * `size` write data buffer size. + */ +LZ4FLIB_STATIC_API size_t LZ4F_write(LZ4_writeFile_t* lz4fWrite, const void* buf, size_t size); + +/*! LZ4F_writeClose() : + * Close lz4file handle. + * `lz4f` must use LZ4F_writeOpen to set first. + */ +LZ4FLIB_STATIC_API LZ4F_errorCode_t LZ4F_writeClose(LZ4_writeFile_t* lz4fWrite); + +#endif /* LZ4FILE_H */ + +#if defined (__cplusplus) +} +#endif diff --git a/example/android/third_party/lz4/include/lz4frame.c b/example/android/third_party/lz4/include/lz4frame.c new file mode 100644 index 00000000..4eb9713f --- /dev/null +++ b/example/android/third_party/lz4/include/lz4frame.c @@ -0,0 +1,2134 @@ +/* + * LZ4 auto-framing library + * Copyright (C) 2011-2016, Yann Collet. + * + * BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions are + * met: + * + * - Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * - Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following disclaimer + * in the documentation and/or other materials provided with the + * distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * You can contact the author at : + * - LZ4 homepage : http://www.lz4.org + * - LZ4 source repository : https://github.com/lz4/lz4 + */ + +/* LZ4F is a stand-alone API to create LZ4-compressed Frames + * in full conformance with specification v1.6.1 . + * This library rely upon memory management capabilities (malloc, free) + * provided either by , + * or redirected towards another library of user's choice + * (see Memory Routines below). + */ + + +/*-************************************ +* Compiler Options +**************************************/ +#include +#ifdef _MSC_VER /* Visual Studio */ +# pragma warning(disable : 4127) /* disable: C4127: conditional expression is constant */ +#endif + + +/*-************************************ +* Tuning parameters +**************************************/ +/* + * LZ4F_HEAPMODE : + * Control how LZ4F_compressFrame allocates the Compression State, + * either on stack (0:default, fastest), or in memory heap (1:requires malloc()). + */ +#ifndef LZ4F_HEAPMODE +# define LZ4F_HEAPMODE 0 +#endif + + +/*-************************************ +* Library declarations +**************************************/ +#define LZ4F_STATIC_LINKING_ONLY +#include "lz4frame.h" +#define LZ4_STATIC_LINKING_ONLY +#include "lz4.h" +#define LZ4_HC_STATIC_LINKING_ONLY +#include "lz4hc.h" +#define XXH_STATIC_LINKING_ONLY +#include "xxhash.h" + + +/*-************************************ +* Memory routines +**************************************/ +/* + * User may redirect invocations of + * malloc(), calloc() and free() + * towards another library or solution of their choice + * by modifying below section. +**/ + +#include /* memset, memcpy, memmove */ +#ifndef LZ4_SRC_INCLUDED /* avoid redefinition when sources are coalesced */ +# define MEM_INIT(p,v,s) memset((p),(v),(s)) +#endif + +#ifndef LZ4_SRC_INCLUDED /* avoid redefinition when sources are coalesced */ +# include /* malloc, calloc, free */ +# define ALLOC(s) malloc(s) +# define ALLOC_AND_ZERO(s) calloc(1,(s)) +# define FREEMEM(p) free(p) +#endif + +static void* LZ4F_calloc(size_t s, LZ4F_CustomMem cmem) +{ + /* custom calloc defined : use it */ + if (cmem.customCalloc != NULL) { + return cmem.customCalloc(cmem.opaqueState, s); + } + /* nothing defined : use default 's calloc() */ + if (cmem.customAlloc == NULL) { + return ALLOC_AND_ZERO(s); + } + /* only custom alloc defined : use it, and combine it with memset() */ + { void* const p = cmem.customAlloc(cmem.opaqueState, s); + if (p != NULL) MEM_INIT(p, 0, s); + return p; +} } + +static void* LZ4F_malloc(size_t s, LZ4F_CustomMem cmem) +{ + /* custom malloc defined : use it */ + if (cmem.customAlloc != NULL) { + return cmem.customAlloc(cmem.opaqueState, s); + } + /* nothing defined : use default 's malloc() */ + return ALLOC(s); +} + +static void LZ4F_free(void* p, LZ4F_CustomMem cmem) +{ + /* custom malloc defined : use it */ + if (cmem.customFree != NULL) { + cmem.customFree(cmem.opaqueState, p); + return; + } + /* nothing defined : use default 's free() */ + FREEMEM(p); +} + + +/*-************************************ +* Debug +**************************************/ +#if defined(LZ4_DEBUG) && (LZ4_DEBUG>=1) +# include +#else +# ifndef assert +# define assert(condition) ((void)0) +# endif +#endif + +#define LZ4F_STATIC_ASSERT(c) { enum { LZ4F_static_assert = 1/(int)(!!(c)) }; } /* use only *after* variable declarations */ + +#if defined(LZ4_DEBUG) && (LZ4_DEBUG>=2) && !defined(DEBUGLOG) +# include +static int g_debuglog_enable = 1; +# define DEBUGLOG(l, ...) { \ + if ((g_debuglog_enable) && (l<=LZ4_DEBUG)) { \ + fprintf(stderr, __FILE__ " (%i): ", __LINE__ ); \ + fprintf(stderr, __VA_ARGS__); \ + fprintf(stderr, " \n"); \ + } } +#else +# define DEBUGLOG(l, ...) {} /* disabled */ +#endif + + +/*-************************************ +* Basic Types +**************************************/ +#if !defined (__VMS) && (defined (__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) ) +# include + typedef uint8_t BYTE; + typedef uint16_t U16; + typedef uint32_t U32; + typedef int32_t S32; + typedef uint64_t U64; +#else + typedef unsigned char BYTE; + typedef unsigned short U16; + typedef unsigned int U32; + typedef signed int S32; + typedef unsigned long long U64; +#endif + + +/* unoptimized version; solves endianness & alignment issues */ +static U32 LZ4F_readLE32 (const void* src) +{ + const BYTE* const srcPtr = (const BYTE*)src; + U32 value32 = srcPtr[0]; + value32 += ((U32)srcPtr[1])<< 8; + value32 += ((U32)srcPtr[2])<<16; + value32 += ((U32)srcPtr[3])<<24; + return value32; +} + +static void LZ4F_writeLE32 (void* dst, U32 value32) +{ + BYTE* const dstPtr = (BYTE*)dst; + dstPtr[0] = (BYTE)value32; + dstPtr[1] = (BYTE)(value32 >> 8); + dstPtr[2] = (BYTE)(value32 >> 16); + dstPtr[3] = (BYTE)(value32 >> 24); +} + +static U64 LZ4F_readLE64 (const void* src) +{ + const BYTE* const srcPtr = (const BYTE*)src; + U64 value64 = srcPtr[0]; + value64 += ((U64)srcPtr[1]<<8); + value64 += ((U64)srcPtr[2]<<16); + value64 += ((U64)srcPtr[3]<<24); + value64 += ((U64)srcPtr[4]<<32); + value64 += ((U64)srcPtr[5]<<40); + value64 += ((U64)srcPtr[6]<<48); + value64 += ((U64)srcPtr[7]<<56); + return value64; +} + +static void LZ4F_writeLE64 (void* dst, U64 value64) +{ + BYTE* const dstPtr = (BYTE*)dst; + dstPtr[0] = (BYTE)value64; + dstPtr[1] = (BYTE)(value64 >> 8); + dstPtr[2] = (BYTE)(value64 >> 16); + dstPtr[3] = (BYTE)(value64 >> 24); + dstPtr[4] = (BYTE)(value64 >> 32); + dstPtr[5] = (BYTE)(value64 >> 40); + dstPtr[6] = (BYTE)(value64 >> 48); + dstPtr[7] = (BYTE)(value64 >> 56); +} + + +/*-************************************ +* Constants +**************************************/ +#ifndef LZ4_SRC_INCLUDED /* avoid double definition */ +# define KB *(1<<10) +# define MB *(1<<20) +# define GB *(1<<30) +#endif + +#define _1BIT 0x01 +#define _2BITS 0x03 +#define _3BITS 0x07 +#define _4BITS 0x0F +#define _8BITS 0xFF + +#define LZ4F_BLOCKUNCOMPRESSED_FLAG 0x80000000U +#define LZ4F_BLOCKSIZEID_DEFAULT LZ4F_max64KB + +static const size_t minFHSize = LZ4F_HEADER_SIZE_MIN; /* 7 */ +static const size_t maxFHSize = LZ4F_HEADER_SIZE_MAX; /* 19 */ +static const size_t BHSize = LZ4F_BLOCK_HEADER_SIZE; /* block header : size, and compress flag */ +static const size_t BFSize = LZ4F_BLOCK_CHECKSUM_SIZE; /* block footer : checksum (optional) */ + + +/*-************************************ +* Structures and local types +**************************************/ + +typedef enum { LZ4B_COMPRESSED, LZ4B_UNCOMPRESSED} LZ4F_BlockCompressMode_e; +typedef enum { ctxNone, ctxFast, ctxHC } LZ4F_CtxType_e; + +typedef struct LZ4F_cctx_s +{ + LZ4F_CustomMem cmem; + LZ4F_preferences_t prefs; + U32 version; + U32 cStage; /* 0 : compression uninitialized ; 1 : initialized, can compress */ + const LZ4F_CDict* cdict; + size_t maxBlockSize; + size_t maxBufferSize; + BYTE* tmpBuff; /* internal buffer, for streaming */ + BYTE* tmpIn; /* starting position of data compress within internal buffer (>= tmpBuff) */ + size_t tmpInSize; /* amount of data to compress after tmpIn */ + U64 totalInSize; + XXH32_state_t xxh; + void* lz4CtxPtr; + U16 lz4CtxAlloc; /* sized for: 0 = none, 1 = lz4 ctx, 2 = lz4hc ctx */ + U16 lz4CtxType; /* in use as: 0 = none, 1 = lz4 ctx, 2 = lz4hc ctx */ + LZ4F_BlockCompressMode_e blockCompressMode; +} LZ4F_cctx_t; + + +/*-************************************ +* Error management +**************************************/ +#define LZ4F_GENERATE_STRING(STRING) #STRING, +static const char* LZ4F_errorStrings[] = { LZ4F_LIST_ERRORS(LZ4F_GENERATE_STRING) }; + + +unsigned LZ4F_isError(LZ4F_errorCode_t code) +{ + return (code > (LZ4F_errorCode_t)(-LZ4F_ERROR_maxCode)); +} + +const char* LZ4F_getErrorName(LZ4F_errorCode_t code) +{ + static const char* codeError = "Unspecified error code"; + if (LZ4F_isError(code)) return LZ4F_errorStrings[-(int)(code)]; + return codeError; +} + +LZ4F_errorCodes LZ4F_getErrorCode(size_t functionResult) +{ + if (!LZ4F_isError(functionResult)) return LZ4F_OK_NoError; + return (LZ4F_errorCodes)(-(ptrdiff_t)functionResult); +} + +static LZ4F_errorCode_t LZ4F_returnErrorCode(LZ4F_errorCodes code) +{ + /* A compilation error here means sizeof(ptrdiff_t) is not large enough */ + LZ4F_STATIC_ASSERT(sizeof(ptrdiff_t) >= sizeof(size_t)); + return (LZ4F_errorCode_t)-(ptrdiff_t)code; +} + +#define RETURN_ERROR(e) return LZ4F_returnErrorCode(LZ4F_ERROR_ ## e) + +#define RETURN_ERROR_IF(c,e) do { \ + if (c) { \ + DEBUGLOG(3, "Error: " #c); \ + RETURN_ERROR(e); \ + } \ + } while (0) + +#define FORWARD_IF_ERROR(r) do { if (LZ4F_isError(r)) return (r); } while (0) + +unsigned LZ4F_getVersion(void) { return LZ4F_VERSION; } + +int LZ4F_compressionLevel_max(void) { return LZ4HC_CLEVEL_MAX; } + +size_t LZ4F_getBlockSize(LZ4F_blockSizeID_t blockSizeID) +{ + static const size_t blockSizes[4] = { 64 KB, 256 KB, 1 MB, 4 MB }; + + if (blockSizeID == 0) blockSizeID = LZ4F_BLOCKSIZEID_DEFAULT; + if (blockSizeID < LZ4F_max64KB || blockSizeID > LZ4F_max4MB) + RETURN_ERROR(maxBlockSize_invalid); + { int const blockSizeIdx = (int)blockSizeID - (int)LZ4F_max64KB; + return blockSizes[blockSizeIdx]; +} } + +/*-************************************ +* Private functions +**************************************/ +#define MIN(a,b) ( (a) < (b) ? (a) : (b) ) + +static BYTE LZ4F_headerChecksum (const void* header, size_t length) +{ + U32 const xxh = XXH32(header, length, 0); + return (BYTE)(xxh >> 8); +} + + +/*-************************************ +* Simple-pass compression functions +**************************************/ +static LZ4F_blockSizeID_t LZ4F_optimalBSID(const LZ4F_blockSizeID_t requestedBSID, + const size_t srcSize) +{ + LZ4F_blockSizeID_t proposedBSID = LZ4F_max64KB; + size_t maxBlockSize = 64 KB; + while (requestedBSID > proposedBSID) { + if (srcSize <= maxBlockSize) + return proposedBSID; + proposedBSID = (LZ4F_blockSizeID_t)((int)proposedBSID + 1); + maxBlockSize <<= 2; + } + return requestedBSID; +} + +/*! LZ4F_compressBound_internal() : + * Provides dstCapacity given a srcSize to guarantee operation success in worst case situations. + * prefsPtr is optional : if NULL is provided, preferences will be set to cover worst case scenario. + * @return is always the same for a srcSize and prefsPtr, so it can be relied upon to size reusable buffers. + * When srcSize==0, LZ4F_compressBound() provides an upper bound for LZ4F_flush() and LZ4F_compressEnd() operations. + */ +static size_t LZ4F_compressBound_internal(size_t srcSize, + const LZ4F_preferences_t* preferencesPtr, + size_t alreadyBuffered) +{ + LZ4F_preferences_t prefsNull = LZ4F_INIT_PREFERENCES; + prefsNull.frameInfo.contentChecksumFlag = LZ4F_contentChecksumEnabled; /* worst case */ + prefsNull.frameInfo.blockChecksumFlag = LZ4F_blockChecksumEnabled; /* worst case */ + { const LZ4F_preferences_t* const prefsPtr = (preferencesPtr==NULL) ? &prefsNull : preferencesPtr; + U32 const flush = prefsPtr->autoFlush | (srcSize==0); + LZ4F_blockSizeID_t const blockID = prefsPtr->frameInfo.blockSizeID; + size_t const blockSize = LZ4F_getBlockSize(blockID); + size_t const maxBuffered = blockSize - 1; + size_t const bufferedSize = MIN(alreadyBuffered, maxBuffered); + size_t const maxSrcSize = srcSize + bufferedSize; + unsigned const nbFullBlocks = (unsigned)(maxSrcSize / blockSize); + size_t const partialBlockSize = maxSrcSize & (blockSize-1); + size_t const lastBlockSize = flush ? partialBlockSize : 0; + unsigned const nbBlocks = nbFullBlocks + (lastBlockSize>0); + + size_t const blockCRCSize = BFSize * prefsPtr->frameInfo.blockChecksumFlag; + size_t const frameEnd = BHSize + (prefsPtr->frameInfo.contentChecksumFlag*BFSize); + + return ((BHSize + blockCRCSize) * nbBlocks) + + (blockSize * nbFullBlocks) + lastBlockSize + frameEnd; + } +} + +size_t LZ4F_compressFrameBound(size_t srcSize, const LZ4F_preferences_t* preferencesPtr) +{ + LZ4F_preferences_t prefs; + size_t const headerSize = maxFHSize; /* max header size, including optional fields */ + + if (preferencesPtr!=NULL) prefs = *preferencesPtr; + else MEM_INIT(&prefs, 0, sizeof(prefs)); + prefs.autoFlush = 1; + + return headerSize + LZ4F_compressBound_internal(srcSize, &prefs, 0);; +} + + +/*! LZ4F_compressFrame_usingCDict() : + * Compress srcBuffer using a dictionary, in a single step. + * cdict can be NULL, in which case, no dictionary is used. + * dstBuffer MUST be >= LZ4F_compressFrameBound(srcSize, preferencesPtr). + * The LZ4F_preferences_t structure is optional : you may provide NULL as argument, + * however, it's the only way to provide a dictID, so it's not recommended. + * @return : number of bytes written into dstBuffer, + * or an error code if it fails (can be tested using LZ4F_isError()) + */ +size_t LZ4F_compressFrame_usingCDict(LZ4F_cctx* cctx, + void* dstBuffer, size_t dstCapacity, + const void* srcBuffer, size_t srcSize, + const LZ4F_CDict* cdict, + const LZ4F_preferences_t* preferencesPtr) +{ + LZ4F_preferences_t prefs; + LZ4F_compressOptions_t options; + BYTE* const dstStart = (BYTE*) dstBuffer; + BYTE* dstPtr = dstStart; + BYTE* const dstEnd = dstStart + dstCapacity; + + DEBUGLOG(4, "LZ4F_compressFrame_usingCDict (srcSize=%u)", (unsigned)srcSize); + if (preferencesPtr!=NULL) + prefs = *preferencesPtr; + else + MEM_INIT(&prefs, 0, sizeof(prefs)); + if (prefs.frameInfo.contentSize != 0) + prefs.frameInfo.contentSize = (U64)srcSize; /* auto-correct content size if selected (!=0) */ + + prefs.frameInfo.blockSizeID = LZ4F_optimalBSID(prefs.frameInfo.blockSizeID, srcSize); + prefs.autoFlush = 1; + if (srcSize <= LZ4F_getBlockSize(prefs.frameInfo.blockSizeID)) + prefs.frameInfo.blockMode = LZ4F_blockIndependent; /* only one block => no need for inter-block link */ + + MEM_INIT(&options, 0, sizeof(options)); + options.stableSrc = 1; + + RETURN_ERROR_IF(dstCapacity < LZ4F_compressFrameBound(srcSize, &prefs), dstMaxSize_tooSmall); + + { size_t const headerSize = LZ4F_compressBegin_usingCDict(cctx, dstBuffer, dstCapacity, cdict, &prefs); /* write header */ + FORWARD_IF_ERROR(headerSize); + dstPtr += headerSize; /* header size */ } + + assert(dstEnd >= dstPtr); + { size_t const cSize = LZ4F_compressUpdate(cctx, dstPtr, (size_t)(dstEnd-dstPtr), srcBuffer, srcSize, &options); + FORWARD_IF_ERROR(cSize); + dstPtr += cSize; } + + assert(dstEnd >= dstPtr); + { size_t const tailSize = LZ4F_compressEnd(cctx, dstPtr, (size_t)(dstEnd-dstPtr), &options); /* flush last block, and generate suffix */ + FORWARD_IF_ERROR(tailSize); + dstPtr += tailSize; } + + assert(dstEnd >= dstStart); + return (size_t)(dstPtr - dstStart); +} + + +/*! LZ4F_compressFrame() : + * Compress an entire srcBuffer into a valid LZ4 frame, in a single step. + * dstBuffer MUST be >= LZ4F_compressFrameBound(srcSize, preferencesPtr). + * The LZ4F_preferences_t structure is optional : you can provide NULL as argument. All preferences will be set to default. + * @return : number of bytes written into dstBuffer. + * or an error code if it fails (can be tested using LZ4F_isError()) + */ +size_t LZ4F_compressFrame(void* dstBuffer, size_t dstCapacity, + const void* srcBuffer, size_t srcSize, + const LZ4F_preferences_t* preferencesPtr) +{ + size_t result; +#if (LZ4F_HEAPMODE) + LZ4F_cctx_t* cctxPtr; + result = LZ4F_createCompressionContext(&cctxPtr, LZ4F_VERSION); + FORWARD_IF_ERROR(result); +#else + LZ4F_cctx_t cctx; + LZ4_stream_t lz4ctx; + LZ4F_cctx_t* const cctxPtr = &cctx; + + MEM_INIT(&cctx, 0, sizeof(cctx)); + cctx.version = LZ4F_VERSION; + cctx.maxBufferSize = 5 MB; /* mess with real buffer size to prevent dynamic allocation; works only because autoflush==1 & stableSrc==1 */ + if ( preferencesPtr == NULL + || preferencesPtr->compressionLevel < LZ4HC_CLEVEL_MIN ) { + LZ4_initStream(&lz4ctx, sizeof(lz4ctx)); + cctxPtr->lz4CtxPtr = &lz4ctx; + cctxPtr->lz4CtxAlloc = 1; + cctxPtr->lz4CtxType = ctxFast; + } +#endif + DEBUGLOG(4, "LZ4F_compressFrame"); + + result = LZ4F_compressFrame_usingCDict(cctxPtr, dstBuffer, dstCapacity, + srcBuffer, srcSize, + NULL, preferencesPtr); + +#if (LZ4F_HEAPMODE) + LZ4F_freeCompressionContext(cctxPtr); +#else + if ( preferencesPtr != NULL + && preferencesPtr->compressionLevel >= LZ4HC_CLEVEL_MIN ) { + LZ4F_free(cctxPtr->lz4CtxPtr, cctxPtr->cmem); + } +#endif + return result; +} + + +/*-*************************************************** +* Dictionary compression +*****************************************************/ + +struct LZ4F_CDict_s { + LZ4F_CustomMem cmem; + void* dictContent; + LZ4_stream_t* fastCtx; + LZ4_streamHC_t* HCCtx; +}; /* typedef'd to LZ4F_CDict within lz4frame_static.h */ + +LZ4F_CDict* +LZ4F_createCDict_advanced(LZ4F_CustomMem cmem, const void* dictBuffer, size_t dictSize) +{ + const char* dictStart = (const char*)dictBuffer; + LZ4F_CDict* const cdict = (LZ4F_CDict*)LZ4F_malloc(sizeof(*cdict), cmem); + DEBUGLOG(4, "LZ4F_createCDict_advanced"); + if (!cdict) return NULL; + cdict->cmem = cmem; + if (dictSize > 64 KB) { + dictStart += dictSize - 64 KB; + dictSize = 64 KB; + } + cdict->dictContent = LZ4F_malloc(dictSize, cmem); + cdict->fastCtx = (LZ4_stream_t*)LZ4F_malloc(sizeof(LZ4_stream_t), cmem); + if (cdict->fastCtx) + LZ4_initStream(cdict->fastCtx, sizeof(LZ4_stream_t)); + cdict->HCCtx = (LZ4_streamHC_t*)LZ4F_malloc(sizeof(LZ4_streamHC_t), cmem); + if (cdict->HCCtx) + LZ4_initStreamHC(cdict->HCCtx, sizeof(LZ4_streamHC_t)); + if (!cdict->dictContent || !cdict->fastCtx || !cdict->HCCtx) { + LZ4F_freeCDict(cdict); + return NULL; + } + memcpy(cdict->dictContent, dictStart, dictSize); + LZ4_loadDictSlow(cdict->fastCtx, (const char*)cdict->dictContent, (int)dictSize); + LZ4_setCompressionLevel(cdict->HCCtx, LZ4HC_CLEVEL_DEFAULT); + LZ4_loadDictHC(cdict->HCCtx, (const char*)cdict->dictContent, (int)dictSize); + return cdict; +} + +/*! LZ4F_createCDict() : + * When compressing multiple messages / blocks with the same dictionary, it's recommended to load it just once. + * LZ4F_createCDict() will create a digested dictionary, ready to start future compression operations without startup delay. + * LZ4F_CDict can be created once and shared by multiple threads concurrently, since its usage is read-only. + * @dictBuffer can be released after LZ4F_CDict creation, since its content is copied within CDict + * @return : digested dictionary for compression, or NULL if failed */ +LZ4F_CDict* LZ4F_createCDict(const void* dictBuffer, size_t dictSize) +{ + DEBUGLOG(4, "LZ4F_createCDict"); + return LZ4F_createCDict_advanced(LZ4F_defaultCMem, dictBuffer, dictSize); +} + +void LZ4F_freeCDict(LZ4F_CDict* cdict) +{ + if (cdict==NULL) return; /* support free on NULL */ + LZ4F_free(cdict->dictContent, cdict->cmem); + LZ4F_free(cdict->fastCtx, cdict->cmem); + LZ4F_free(cdict->HCCtx, cdict->cmem); + LZ4F_free(cdict, cdict->cmem); +} + + +/*-********************************* +* Advanced compression functions +***********************************/ + +LZ4F_cctx* +LZ4F_createCompressionContext_advanced(LZ4F_CustomMem customMem, unsigned version) +{ + LZ4F_cctx* const cctxPtr = + (LZ4F_cctx*)LZ4F_calloc(sizeof(LZ4F_cctx), customMem); + if (cctxPtr==NULL) return NULL; + + cctxPtr->cmem = customMem; + cctxPtr->version = version; + cctxPtr->cStage = 0; /* Uninitialized. Next stage : init cctx */ + + return cctxPtr; +} + +/*! LZ4F_createCompressionContext() : + * The first thing to do is to create a compressionContext object, which will be used in all compression operations. + * This is achieved using LZ4F_createCompressionContext(), which takes as argument a version and an LZ4F_preferences_t structure. + * The version provided MUST be LZ4F_VERSION. It is intended to track potential incompatible differences between different binaries. + * The function will provide a pointer to an allocated LZ4F_compressionContext_t object. + * If the result LZ4F_errorCode_t is not OK_NoError, there was an error during context creation. + * Object can release its memory using LZ4F_freeCompressionContext(); +**/ +LZ4F_errorCode_t +LZ4F_createCompressionContext(LZ4F_cctx** LZ4F_compressionContextPtr, unsigned version) +{ + assert(LZ4F_compressionContextPtr != NULL); /* considered a violation of narrow contract */ + /* in case it nonetheless happen in production */ + RETURN_ERROR_IF(LZ4F_compressionContextPtr == NULL, parameter_null); + + *LZ4F_compressionContextPtr = LZ4F_createCompressionContext_advanced(LZ4F_defaultCMem, version); + RETURN_ERROR_IF(*LZ4F_compressionContextPtr==NULL, allocation_failed); + return LZ4F_OK_NoError; +} + +LZ4F_errorCode_t LZ4F_freeCompressionContext(LZ4F_cctx* cctxPtr) +{ + if (cctxPtr != NULL) { /* support free on NULL */ + LZ4F_free(cctxPtr->lz4CtxPtr, cctxPtr->cmem); /* note: LZ4_streamHC_t and LZ4_stream_t are simple POD types */ + LZ4F_free(cctxPtr->tmpBuff, cctxPtr->cmem); + LZ4F_free(cctxPtr, cctxPtr->cmem); + } + return LZ4F_OK_NoError; +} + + +/** + * This function prepares the internal LZ4(HC) stream for a new compression, + * resetting the context and attaching the dictionary, if there is one. + * + * It needs to be called at the beginning of each independent compression + * stream (i.e., at the beginning of a frame in blockLinked mode, or at the + * beginning of each block in blockIndependent mode). + */ +static void LZ4F_initStream(void* ctx, + const LZ4F_CDict* cdict, + int level, + LZ4F_blockMode_t blockMode) { + if (level < LZ4HC_CLEVEL_MIN) { + if (cdict || blockMode == LZ4F_blockLinked) { + /* In these cases, we will call LZ4_compress_fast_continue(), + * which needs an already reset context. Otherwise, we'll call a + * one-shot API. The non-continued APIs internally perform their own + * resets at the beginning of their calls, where they know what + * tableType they need the context to be in. So in that case this + * would be misguided / wasted work. */ + LZ4_resetStream_fast((LZ4_stream_t*)ctx); + if (cdict) + LZ4_attach_dictionary((LZ4_stream_t*)ctx, cdict->fastCtx); + } + /* In these cases, we'll call a one-shot API. + * The non-continued APIs internally perform their own resets + * at the beginning of their calls, where they know + * which tableType they need the context to be in. + * Therefore, a reset here would be wasted work. */ + } else { + LZ4_resetStreamHC_fast((LZ4_streamHC_t*)ctx, level); + if (cdict) + LZ4_attach_HC_dictionary((LZ4_streamHC_t*)ctx, cdict->HCCtx); + } +} + +static int ctxTypeID_to_size(int ctxTypeID) { + switch(ctxTypeID) { + case 1: + return LZ4_sizeofState(); + case 2: + return LZ4_sizeofStateHC(); + default: + return 0; + } +} + +/* LZ4F_compressBegin_internal() + * Note: only accepts @cdict _or_ @dictBuffer as non NULL. + */ +size_t LZ4F_compressBegin_internal(LZ4F_cctx* cctx, + void* dstBuffer, size_t dstCapacity, + const void* dictBuffer, size_t dictSize, + const LZ4F_CDict* cdict, + const LZ4F_preferences_t* preferencesPtr) +{ + LZ4F_preferences_t const prefNull = LZ4F_INIT_PREFERENCES; + BYTE* const dstStart = (BYTE*)dstBuffer; + BYTE* dstPtr = dstStart; + + RETURN_ERROR_IF(dstCapacity < maxFHSize, dstMaxSize_tooSmall); + if (preferencesPtr == NULL) preferencesPtr = &prefNull; + cctx->prefs = *preferencesPtr; + + /* cctx Management */ + { U16 const ctxTypeID = (cctx->prefs.compressionLevel < LZ4HC_CLEVEL_MIN) ? 1 : 2; + int requiredSize = ctxTypeID_to_size(ctxTypeID); + int allocatedSize = ctxTypeID_to_size(cctx->lz4CtxAlloc); + if (allocatedSize < requiredSize) { + /* not enough space allocated */ + LZ4F_free(cctx->lz4CtxPtr, cctx->cmem); + if (cctx->prefs.compressionLevel < LZ4HC_CLEVEL_MIN) { + /* must take ownership of memory allocation, + * in order to respect custom allocator contract */ + cctx->lz4CtxPtr = LZ4F_malloc(sizeof(LZ4_stream_t), cctx->cmem); + if (cctx->lz4CtxPtr) + LZ4_initStream(cctx->lz4CtxPtr, sizeof(LZ4_stream_t)); + } else { + cctx->lz4CtxPtr = LZ4F_malloc(sizeof(LZ4_streamHC_t), cctx->cmem); + if (cctx->lz4CtxPtr) + LZ4_initStreamHC(cctx->lz4CtxPtr, sizeof(LZ4_streamHC_t)); + } + RETURN_ERROR_IF(cctx->lz4CtxPtr == NULL, allocation_failed); + cctx->lz4CtxAlloc = ctxTypeID; + cctx->lz4CtxType = ctxTypeID; + } else if (cctx->lz4CtxType != ctxTypeID) { + /* otherwise, a sufficient buffer is already allocated, + * but we need to reset it to the correct context type */ + if (cctx->prefs.compressionLevel < LZ4HC_CLEVEL_MIN) { + LZ4_initStream((LZ4_stream_t*)cctx->lz4CtxPtr, sizeof(LZ4_stream_t)); + } else { + LZ4_initStreamHC((LZ4_streamHC_t*)cctx->lz4CtxPtr, sizeof(LZ4_streamHC_t)); + LZ4_setCompressionLevel((LZ4_streamHC_t*)cctx->lz4CtxPtr, cctx->prefs.compressionLevel); + } + cctx->lz4CtxType = ctxTypeID; + } } + + /* Buffer Management */ + if (cctx->prefs.frameInfo.blockSizeID == 0) + cctx->prefs.frameInfo.blockSizeID = LZ4F_BLOCKSIZEID_DEFAULT; + cctx->maxBlockSize = LZ4F_getBlockSize(cctx->prefs.frameInfo.blockSizeID); + + { size_t const requiredBuffSize = preferencesPtr->autoFlush ? + ((cctx->prefs.frameInfo.blockMode == LZ4F_blockLinked) ? 64 KB : 0) : /* only needs past data up to window size */ + cctx->maxBlockSize + ((cctx->prefs.frameInfo.blockMode == LZ4F_blockLinked) ? 128 KB : 0); + + if (cctx->maxBufferSize < requiredBuffSize) { + cctx->maxBufferSize = 0; + LZ4F_free(cctx->tmpBuff, cctx->cmem); + cctx->tmpBuff = (BYTE*)LZ4F_malloc(requiredBuffSize, cctx->cmem); + RETURN_ERROR_IF(cctx->tmpBuff == NULL, allocation_failed); + cctx->maxBufferSize = requiredBuffSize; + } } + cctx->tmpIn = cctx->tmpBuff; + cctx->tmpInSize = 0; + (void)XXH32_reset(&(cctx->xxh), 0); + + /* context init */ + cctx->cdict = cdict; + if (cctx->prefs.frameInfo.blockMode == LZ4F_blockLinked) { + /* frame init only for blockLinked : blockIndependent will be init at each block */ + LZ4F_initStream(cctx->lz4CtxPtr, cdict, cctx->prefs.compressionLevel, LZ4F_blockLinked); + } + if (preferencesPtr->compressionLevel >= LZ4HC_CLEVEL_MIN) { + LZ4_favorDecompressionSpeed((LZ4_streamHC_t*)cctx->lz4CtxPtr, (int)preferencesPtr->favorDecSpeed); + } + if (dictBuffer) { + assert(cdict == NULL); + RETURN_ERROR_IF(dictSize > INT_MAX, parameter_invalid); + if (cctx->lz4CtxType == ctxFast) { + /* lz4 fast*/ + LZ4_loadDict((LZ4_stream_t*)cctx->lz4CtxPtr, (const char*)dictBuffer, (int)dictSize); + } else { + /* lz4hc */ + assert(cctx->lz4CtxType == ctxHC); + LZ4_loadDictHC((LZ4_streamHC_t*)cctx->lz4CtxPtr, (const char*)dictBuffer, (int)dictSize); + } + } + + /* Stage 2 : Write Frame Header */ + + /* Magic Number */ + LZ4F_writeLE32(dstPtr, LZ4F_MAGICNUMBER); + dstPtr += 4; + { BYTE* const headerStart = dstPtr; + + /* FLG Byte */ + *dstPtr++ = (BYTE)(((1 & _2BITS) << 6) /* Version('01') */ + + ((cctx->prefs.frameInfo.blockMode & _1BIT ) << 5) + + ((cctx->prefs.frameInfo.blockChecksumFlag & _1BIT ) << 4) + + ((unsigned)(cctx->prefs.frameInfo.contentSize > 0) << 3) + + ((cctx->prefs.frameInfo.contentChecksumFlag & _1BIT ) << 2) + + (cctx->prefs.frameInfo.dictID > 0) ); + /* BD Byte */ + *dstPtr++ = (BYTE)((cctx->prefs.frameInfo.blockSizeID & _3BITS) << 4); + /* Optional Frame content size field */ + if (cctx->prefs.frameInfo.contentSize) { + LZ4F_writeLE64(dstPtr, cctx->prefs.frameInfo.contentSize); + dstPtr += 8; + cctx->totalInSize = 0; + } + /* Optional dictionary ID field */ + if (cctx->prefs.frameInfo.dictID) { + LZ4F_writeLE32(dstPtr, cctx->prefs.frameInfo.dictID); + dstPtr += 4; + } + /* Header CRC Byte */ + *dstPtr = LZ4F_headerChecksum(headerStart, (size_t)(dstPtr - headerStart)); + dstPtr++; + } + + cctx->cStage = 1; /* header written, now request input data block */ + return (size_t)(dstPtr - dstStart); +} + +size_t LZ4F_compressBegin(LZ4F_cctx* cctx, + void* dstBuffer, size_t dstCapacity, + const LZ4F_preferences_t* preferencesPtr) +{ + return LZ4F_compressBegin_internal(cctx, dstBuffer, dstCapacity, + NULL, 0, + NULL, preferencesPtr); +} + +/* LZ4F_compressBegin_usingDictOnce: + * Hidden implementation, + * employed for multi-threaded compression + * when frame defines linked blocks */ +size_t LZ4F_compressBegin_usingDictOnce(LZ4F_cctx* cctx, + void* dstBuffer, size_t dstCapacity, + const void* dict, size_t dictSize, + const LZ4F_preferences_t* preferencesPtr) +{ + return LZ4F_compressBegin_internal(cctx, dstBuffer, dstCapacity, + dict, dictSize, + NULL, preferencesPtr); +} + +size_t LZ4F_compressBegin_usingDict(LZ4F_cctx* cctx, + void* dstBuffer, size_t dstCapacity, + const void* dict, size_t dictSize, + const LZ4F_preferences_t* preferencesPtr) +{ + /* note : incorrect implementation : + * this will only use the dictionary once, + * instead of once *per* block when frames defines independent blocks */ + return LZ4F_compressBegin_usingDictOnce(cctx, dstBuffer, dstCapacity, + dict, dictSize, + preferencesPtr); +} + +size_t LZ4F_compressBegin_usingCDict(LZ4F_cctx* cctx, + void* dstBuffer, size_t dstCapacity, + const LZ4F_CDict* cdict, + const LZ4F_preferences_t* preferencesPtr) +{ + return LZ4F_compressBegin_internal(cctx, dstBuffer, dstCapacity, + NULL, 0, + cdict, preferencesPtr); +} + + +/* LZ4F_compressBound() : + * @return minimum capacity of dstBuffer for a given srcSize to handle worst case scenario. + * LZ4F_preferences_t structure is optional : if NULL, preferences will be set to cover worst case scenario. + * This function cannot fail. + */ +size_t LZ4F_compressBound(size_t srcSize, const LZ4F_preferences_t* preferencesPtr) +{ + if (preferencesPtr && preferencesPtr->autoFlush) { + return LZ4F_compressBound_internal(srcSize, preferencesPtr, 0); + } + return LZ4F_compressBound_internal(srcSize, preferencesPtr, (size_t)-1); +} + + +typedef int (*compressFunc_t)(void* ctx, const char* src, char* dst, int srcSize, int dstSize, int level, const LZ4F_CDict* cdict); + + +/*! LZ4F_makeBlock(): + * compress a single block, add header and optional checksum. + * assumption : dst buffer capacity is >= BHSize + srcSize + crcSize + */ +static size_t LZ4F_makeBlock(void* dst, + const void* src, size_t srcSize, + compressFunc_t compress, void* lz4ctx, int level, + const LZ4F_CDict* cdict, + LZ4F_blockChecksum_t crcFlag) +{ + BYTE* const cSizePtr = (BYTE*)dst; + U32 cSize; + assert(compress != NULL); + cSize = (U32)compress(lz4ctx, (const char*)src, (char*)(cSizePtr+BHSize), + (int)(srcSize), (int)(srcSize-1), + level, cdict); + + if (cSize == 0 || cSize >= srcSize) { + cSize = (U32)srcSize; + LZ4F_writeLE32(cSizePtr, cSize | LZ4F_BLOCKUNCOMPRESSED_FLAG); + memcpy(cSizePtr+BHSize, src, srcSize); + } else { + LZ4F_writeLE32(cSizePtr, cSize); + } + if (crcFlag) { + U32 const crc32 = XXH32(cSizePtr+BHSize, cSize, 0); /* checksum of compressed data */ + LZ4F_writeLE32(cSizePtr+BHSize+cSize, crc32); + } + return BHSize + cSize + ((U32)crcFlag)*BFSize; +} + + +static int LZ4F_compressBlock(void* ctx, const char* src, char* dst, int srcSize, int dstCapacity, int level, const LZ4F_CDict* cdict) +{ + int const acceleration = (level < 0) ? -level + 1 : 1; + DEBUGLOG(5, "LZ4F_compressBlock (srcSize=%i)", srcSize); + LZ4F_initStream(ctx, cdict, level, LZ4F_blockIndependent); + if (cdict) { + return LZ4_compress_fast_continue((LZ4_stream_t*)ctx, src, dst, srcSize, dstCapacity, acceleration); + } else { + return LZ4_compress_fast_extState_fastReset(ctx, src, dst, srcSize, dstCapacity, acceleration); + } +} + +static int LZ4F_compressBlock_continue(void* ctx, const char* src, char* dst, int srcSize, int dstCapacity, int level, const LZ4F_CDict* cdict) +{ + int const acceleration = (level < 0) ? -level + 1 : 1; + (void)cdict; /* init once at beginning of frame */ + DEBUGLOG(5, "LZ4F_compressBlock_continue (srcSize=%i)", srcSize); + return LZ4_compress_fast_continue((LZ4_stream_t*)ctx, src, dst, srcSize, dstCapacity, acceleration); +} + +static int LZ4F_compressBlockHC(void* ctx, const char* src, char* dst, int srcSize, int dstCapacity, int level, const LZ4F_CDict* cdict) +{ + LZ4F_initStream(ctx, cdict, level, LZ4F_blockIndependent); + if (cdict) { + return LZ4_compress_HC_continue((LZ4_streamHC_t*)ctx, src, dst, srcSize, dstCapacity); + } + return LZ4_compress_HC_extStateHC_fastReset(ctx, src, dst, srcSize, dstCapacity, level); +} + +static int LZ4F_compressBlockHC_continue(void* ctx, const char* src, char* dst, int srcSize, int dstCapacity, int level, const LZ4F_CDict* cdict) +{ + (void)level; (void)cdict; /* init once at beginning of frame */ + return LZ4_compress_HC_continue((LZ4_streamHC_t*)ctx, src, dst, srcSize, dstCapacity); +} + +static int LZ4F_doNotCompressBlock(void* ctx, const char* src, char* dst, int srcSize, int dstCapacity, int level, const LZ4F_CDict* cdict) +{ + (void)ctx; (void)src; (void)dst; (void)srcSize; (void)dstCapacity; (void)level; (void)cdict; + return 0; +} + +static compressFunc_t LZ4F_selectCompression(LZ4F_blockMode_t blockMode, int level, LZ4F_BlockCompressMode_e compressMode) +{ + if (compressMode == LZ4B_UNCOMPRESSED) + return LZ4F_doNotCompressBlock; + if (level < LZ4HC_CLEVEL_MIN) { + if (blockMode == LZ4F_blockIndependent) return LZ4F_compressBlock; + return LZ4F_compressBlock_continue; + } + if (blockMode == LZ4F_blockIndependent) return LZ4F_compressBlockHC; + return LZ4F_compressBlockHC_continue; +} + +/* Save history (up to 64KB) into @tmpBuff */ +static int LZ4F_localSaveDict(LZ4F_cctx_t* cctxPtr) +{ + if (cctxPtr->prefs.compressionLevel < LZ4HC_CLEVEL_MIN) + return LZ4_saveDict ((LZ4_stream_t*)(cctxPtr->lz4CtxPtr), (char*)(cctxPtr->tmpBuff), 64 KB); + return LZ4_saveDictHC ((LZ4_streamHC_t*)(cctxPtr->lz4CtxPtr), (char*)(cctxPtr->tmpBuff), 64 KB); +} + +typedef enum { notDone, fromTmpBuffer, fromSrcBuffer } LZ4F_lastBlockStatus; + +static const LZ4F_compressOptions_t k_cOptionsNull = { 0, { 0, 0, 0 } }; + + + /*! LZ4F_compressUpdateImpl() : + * LZ4F_compressUpdate() can be called repetitively to compress as much data as necessary. + * When successful, the function always entirely consumes @srcBuffer. + * src data is either buffered or compressed into @dstBuffer. + * If the block compression does not match the compression of the previous block, the old data is flushed + * and operations continue with the new compression mode. + * @dstCapacity MUST be >= LZ4F_compressBound(srcSize, preferencesPtr) when block compression is turned on. + * @compressOptionsPtr is optional : provide NULL to mean "default". + * @return : the number of bytes written into dstBuffer. It can be zero, meaning input data was just buffered. + * or an error code if it fails (which can be tested using LZ4F_isError()) + * After an error, the state is left in a UB state, and must be re-initialized. + */ +static size_t LZ4F_compressUpdateImpl(LZ4F_cctx* cctxPtr, + void* dstBuffer, size_t dstCapacity, + const void* srcBuffer, size_t srcSize, + const LZ4F_compressOptions_t* compressOptionsPtr, + LZ4F_BlockCompressMode_e blockCompression) + { + size_t const blockSize = cctxPtr->maxBlockSize; + const BYTE* srcPtr = (const BYTE*)srcBuffer; + const BYTE* const srcEnd = srcPtr + srcSize; + BYTE* const dstStart = (BYTE*)dstBuffer; + BYTE* dstPtr = dstStart; + LZ4F_lastBlockStatus lastBlockCompressed = notDone; + compressFunc_t const compress = LZ4F_selectCompression(cctxPtr->prefs.frameInfo.blockMode, cctxPtr->prefs.compressionLevel, blockCompression); + size_t bytesWritten; + DEBUGLOG(4, "LZ4F_compressUpdate (srcSize=%zu)", srcSize); + + RETURN_ERROR_IF(cctxPtr->cStage != 1, compressionState_uninitialized); /* state must be initialized and waiting for next block */ + if (dstCapacity < LZ4F_compressBound_internal(srcSize, &(cctxPtr->prefs), cctxPtr->tmpInSize)) + RETURN_ERROR(dstMaxSize_tooSmall); + + if (blockCompression == LZ4B_UNCOMPRESSED && dstCapacity < srcSize) + RETURN_ERROR(dstMaxSize_tooSmall); + + /* flush currently written block, to continue with new block compression */ + if (cctxPtr->blockCompressMode != blockCompression) { + bytesWritten = LZ4F_flush(cctxPtr, dstBuffer, dstCapacity, compressOptionsPtr); + dstPtr += bytesWritten; + cctxPtr->blockCompressMode = blockCompression; + } + + if (compressOptionsPtr == NULL) compressOptionsPtr = &k_cOptionsNull; + + /* complete tmp buffer */ + if (cctxPtr->tmpInSize > 0) { /* some data already within tmp buffer */ + size_t const sizeToCopy = blockSize - cctxPtr->tmpInSize; + assert(blockSize > cctxPtr->tmpInSize); + if (sizeToCopy > srcSize) { + /* add src to tmpIn buffer */ + memcpy(cctxPtr->tmpIn + cctxPtr->tmpInSize, srcBuffer, srcSize); + srcPtr = srcEnd; + cctxPtr->tmpInSize += srcSize; + /* still needs some CRC */ + } else { + /* complete tmpIn block and then compress it */ + lastBlockCompressed = fromTmpBuffer; + memcpy(cctxPtr->tmpIn + cctxPtr->tmpInSize, srcBuffer, sizeToCopy); + srcPtr += sizeToCopy; + + dstPtr += LZ4F_makeBlock(dstPtr, + cctxPtr->tmpIn, blockSize, + compress, cctxPtr->lz4CtxPtr, cctxPtr->prefs.compressionLevel, + cctxPtr->cdict, + cctxPtr->prefs.frameInfo.blockChecksumFlag); + if (cctxPtr->prefs.frameInfo.blockMode==LZ4F_blockLinked) cctxPtr->tmpIn += blockSize; + cctxPtr->tmpInSize = 0; + } } + + while ((size_t)(srcEnd - srcPtr) >= blockSize) { + /* compress full blocks */ + lastBlockCompressed = fromSrcBuffer; + dstPtr += LZ4F_makeBlock(dstPtr, + srcPtr, blockSize, + compress, cctxPtr->lz4CtxPtr, cctxPtr->prefs.compressionLevel, + cctxPtr->cdict, + cctxPtr->prefs.frameInfo.blockChecksumFlag); + srcPtr += blockSize; + } + + if ((cctxPtr->prefs.autoFlush) && (srcPtr < srcEnd)) { + /* autoFlush : remaining input (< blockSize) is compressed */ + lastBlockCompressed = fromSrcBuffer; + dstPtr += LZ4F_makeBlock(dstPtr, + srcPtr, (size_t)(srcEnd - srcPtr), + compress, cctxPtr->lz4CtxPtr, cctxPtr->prefs.compressionLevel, + cctxPtr->cdict, + cctxPtr->prefs.frameInfo.blockChecksumFlag); + srcPtr = srcEnd; + } + + /* preserve dictionary within @tmpBuff whenever necessary */ + if ((cctxPtr->prefs.frameInfo.blockMode==LZ4F_blockLinked) && (lastBlockCompressed==fromSrcBuffer)) { + /* linked blocks are only supported in compressed mode, see LZ4F_uncompressedUpdate */ + assert(blockCompression == LZ4B_COMPRESSED); + if (compressOptionsPtr->stableSrc) { + cctxPtr->tmpIn = cctxPtr->tmpBuff; /* src is stable : dictionary remains in src across invocations */ + } else { + int const realDictSize = LZ4F_localSaveDict(cctxPtr); + assert(0 <= realDictSize && realDictSize <= 64 KB); + cctxPtr->tmpIn = cctxPtr->tmpBuff + realDictSize; + } + } + + /* keep tmpIn within limits */ + if (!(cctxPtr->prefs.autoFlush) /* no autoflush : there may be some data left within internal buffer */ + && (cctxPtr->tmpIn + blockSize) > (cctxPtr->tmpBuff + cctxPtr->maxBufferSize) ) /* not enough room to store next block */ + { + /* only preserve 64KB within internal buffer. Ensures there is enough room for next block. + * note: this situation necessarily implies lastBlockCompressed==fromTmpBuffer */ + int const realDictSize = LZ4F_localSaveDict(cctxPtr); + cctxPtr->tmpIn = cctxPtr->tmpBuff + realDictSize; + assert((cctxPtr->tmpIn + blockSize) <= (cctxPtr->tmpBuff + cctxPtr->maxBufferSize)); + } + + /* some input data left, necessarily < blockSize */ + if (srcPtr < srcEnd) { + /* fill tmp buffer */ + size_t const sizeToCopy = (size_t)(srcEnd - srcPtr); + memcpy(cctxPtr->tmpIn, srcPtr, sizeToCopy); + cctxPtr->tmpInSize = sizeToCopy; + } + + if (cctxPtr->prefs.frameInfo.contentChecksumFlag == LZ4F_contentChecksumEnabled) + (void)XXH32_update(&(cctxPtr->xxh), srcBuffer, srcSize); + + cctxPtr->totalInSize += srcSize; + return (size_t)(dstPtr - dstStart); +} + +/*! LZ4F_compressUpdate() : + * LZ4F_compressUpdate() can be called repetitively to compress as much data as necessary. + * When successful, the function always entirely consumes @srcBuffer. + * src data is either buffered or compressed into @dstBuffer. + * If previously an uncompressed block was written, buffered data is flushed + * before appending compressed data is continued. + * @dstCapacity MUST be >= LZ4F_compressBound(srcSize, preferencesPtr). + * @compressOptionsPtr is optional : provide NULL to mean "default". + * @return : the number of bytes written into dstBuffer. It can be zero, meaning input data was just buffered. + * or an error code if it fails (which can be tested using LZ4F_isError()) + * After an error, the state is left in a UB state, and must be re-initialized. + */ +size_t LZ4F_compressUpdate(LZ4F_cctx* cctxPtr, + void* dstBuffer, size_t dstCapacity, + const void* srcBuffer, size_t srcSize, + const LZ4F_compressOptions_t* compressOptionsPtr) +{ + return LZ4F_compressUpdateImpl(cctxPtr, + dstBuffer, dstCapacity, + srcBuffer, srcSize, + compressOptionsPtr, LZ4B_COMPRESSED); +} + +/*! LZ4F_uncompressedUpdate() : + * Same as LZ4F_compressUpdate(), but requests blocks to be sent uncompressed. + * This symbol is only supported when LZ4F_blockIndependent is used + * @dstCapacity MUST be >= LZ4F_compressBound(srcSize, preferencesPtr). + * @compressOptionsPtr is optional : provide NULL to mean "default". + * @return : the number of bytes written into dstBuffer. It can be zero, meaning input data was just buffered. + * or an error code if it fails (which can be tested using LZ4F_isError()) + * After an error, the state is left in a UB state, and must be re-initialized. + */ +size_t LZ4F_uncompressedUpdate(LZ4F_cctx* cctxPtr, + void* dstBuffer, size_t dstCapacity, + const void* srcBuffer, size_t srcSize, + const LZ4F_compressOptions_t* compressOptionsPtr) +{ + return LZ4F_compressUpdateImpl(cctxPtr, + dstBuffer, dstCapacity, + srcBuffer, srcSize, + compressOptionsPtr, LZ4B_UNCOMPRESSED); +} + + +/*! LZ4F_flush() : + * When compressed data must be sent immediately, without waiting for a block to be filled, + * invoke LZ4_flush(), which will immediately compress any remaining data stored within LZ4F_cctx. + * The result of the function is the number of bytes written into dstBuffer. + * It can be zero, this means there was no data left within LZ4F_cctx. + * The function outputs an error code if it fails (can be tested using LZ4F_isError()) + * LZ4F_compressOptions_t* is optional. NULL is a valid argument. + */ +size_t LZ4F_flush(LZ4F_cctx* cctxPtr, + void* dstBuffer, size_t dstCapacity, + const LZ4F_compressOptions_t* compressOptionsPtr) +{ + BYTE* const dstStart = (BYTE*)dstBuffer; + BYTE* dstPtr = dstStart; + compressFunc_t compress; + + if (cctxPtr->tmpInSize == 0) return 0; /* nothing to flush */ + RETURN_ERROR_IF(cctxPtr->cStage != 1, compressionState_uninitialized); + RETURN_ERROR_IF(dstCapacity < (cctxPtr->tmpInSize + BHSize + BFSize), dstMaxSize_tooSmall); + (void)compressOptionsPtr; /* not useful (yet) */ + + /* select compression function */ + compress = LZ4F_selectCompression(cctxPtr->prefs.frameInfo.blockMode, cctxPtr->prefs.compressionLevel, cctxPtr->blockCompressMode); + + /* compress tmp buffer */ + dstPtr += LZ4F_makeBlock(dstPtr, + cctxPtr->tmpIn, cctxPtr->tmpInSize, + compress, cctxPtr->lz4CtxPtr, cctxPtr->prefs.compressionLevel, + cctxPtr->cdict, + cctxPtr->prefs.frameInfo.blockChecksumFlag); + assert(((void)"flush overflows dstBuffer!", (size_t)(dstPtr - dstStart) <= dstCapacity)); + + if (cctxPtr->prefs.frameInfo.blockMode == LZ4F_blockLinked) + cctxPtr->tmpIn += cctxPtr->tmpInSize; + cctxPtr->tmpInSize = 0; + + /* keep tmpIn within limits */ + if ((cctxPtr->tmpIn + cctxPtr->maxBlockSize) > (cctxPtr->tmpBuff + cctxPtr->maxBufferSize)) { /* necessarily LZ4F_blockLinked */ + int const realDictSize = LZ4F_localSaveDict(cctxPtr); + cctxPtr->tmpIn = cctxPtr->tmpBuff + realDictSize; + } + + return (size_t)(dstPtr - dstStart); +} + + +/*! LZ4F_compressEnd() : + * When you want to properly finish the compressed frame, just call LZ4F_compressEnd(). + * It will flush whatever data remained within compressionContext (like LZ4_flush()) + * but also properly finalize the frame, with an endMark and an (optional) checksum. + * LZ4F_compressOptions_t structure is optional : you can provide NULL as argument. + * @return: the number of bytes written into dstBuffer (necessarily >= 4 (endMark size)) + * or an error code if it fails (can be tested using LZ4F_isError()) + * The context can then be used again to compress a new frame, starting with LZ4F_compressBegin(). + */ +size_t LZ4F_compressEnd(LZ4F_cctx* cctxPtr, + void* dstBuffer, size_t dstCapacity, + const LZ4F_compressOptions_t* compressOptionsPtr) +{ + BYTE* const dstStart = (BYTE*)dstBuffer; + BYTE* dstPtr = dstStart; + + size_t const flushSize = LZ4F_flush(cctxPtr, dstBuffer, dstCapacity, compressOptionsPtr); + DEBUGLOG(5,"LZ4F_compressEnd: dstCapacity=%u", (unsigned)dstCapacity); + FORWARD_IF_ERROR(flushSize); + dstPtr += flushSize; + + assert(flushSize <= dstCapacity); + dstCapacity -= flushSize; + + RETURN_ERROR_IF(dstCapacity < 4, dstMaxSize_tooSmall); + LZ4F_writeLE32(dstPtr, 0); + dstPtr += 4; /* endMark */ + + if (cctxPtr->prefs.frameInfo.contentChecksumFlag == LZ4F_contentChecksumEnabled) { + U32 const xxh = XXH32_digest(&(cctxPtr->xxh)); + RETURN_ERROR_IF(dstCapacity < 8, dstMaxSize_tooSmall); + DEBUGLOG(5,"Writing 32-bit content checksum (0x%0X)", xxh); + LZ4F_writeLE32(dstPtr, xxh); + dstPtr+=4; /* content Checksum */ + } + + cctxPtr->cStage = 0; /* state is now re-usable (with identical preferences) */ + + if (cctxPtr->prefs.frameInfo.contentSize) { + if (cctxPtr->prefs.frameInfo.contentSize != cctxPtr->totalInSize) + RETURN_ERROR(frameSize_wrong); + } + + return (size_t)(dstPtr - dstStart); +} + + +/*-*************************************************** +* Frame Decompression +*****************************************************/ + +typedef enum { + dstage_getFrameHeader=0, dstage_storeFrameHeader, + dstage_init, + dstage_getBlockHeader, dstage_storeBlockHeader, + dstage_copyDirect, dstage_getBlockChecksum, + dstage_getCBlock, dstage_storeCBlock, + dstage_flushOut, + dstage_getSuffix, dstage_storeSuffix, + dstage_getSFrameSize, dstage_storeSFrameSize, + dstage_skipSkippable +} dStage_t; + +struct LZ4F_dctx_s { + LZ4F_CustomMem cmem; + LZ4F_frameInfo_t frameInfo; + U32 version; + dStage_t dStage; + U64 frameRemainingSize; + size_t maxBlockSize; + size_t maxBufferSize; + BYTE* tmpIn; + size_t tmpInSize; + size_t tmpInTarget; + BYTE* tmpOutBuffer; + const BYTE* dict; + size_t dictSize; + BYTE* tmpOut; + size_t tmpOutSize; + size_t tmpOutStart; + XXH32_state_t xxh; + XXH32_state_t blockChecksum; + int skipChecksum; + BYTE header[LZ4F_HEADER_SIZE_MAX]; +}; /* typedef'd to LZ4F_dctx in lz4frame.h */ + + +LZ4F_dctx* LZ4F_createDecompressionContext_advanced(LZ4F_CustomMem customMem, unsigned version) +{ + LZ4F_dctx* const dctx = (LZ4F_dctx*)LZ4F_calloc(sizeof(LZ4F_dctx), customMem); + if (dctx == NULL) return NULL; + + dctx->cmem = customMem; + dctx->version = version; + return dctx; +} + +/*! LZ4F_createDecompressionContext() : + * Create a decompressionContext object, which will track all decompression operations. + * Provides a pointer to a fully allocated and initialized LZ4F_decompressionContext object. + * Object can later be released using LZ4F_freeDecompressionContext(). + * @return : if != 0, there was an error during context creation. + */ +LZ4F_errorCode_t +LZ4F_createDecompressionContext(LZ4F_dctx** LZ4F_decompressionContextPtr, unsigned versionNumber) +{ + assert(LZ4F_decompressionContextPtr != NULL); /* violation of narrow contract */ + RETURN_ERROR_IF(LZ4F_decompressionContextPtr == NULL, parameter_null); /* in case it nonetheless happen in production */ + + *LZ4F_decompressionContextPtr = LZ4F_createDecompressionContext_advanced(LZ4F_defaultCMem, versionNumber); + if (*LZ4F_decompressionContextPtr == NULL) { /* failed allocation */ + RETURN_ERROR(allocation_failed); + } + return LZ4F_OK_NoError; +} + +LZ4F_errorCode_t LZ4F_freeDecompressionContext(LZ4F_dctx* dctx) +{ + LZ4F_errorCode_t result = LZ4F_OK_NoError; + if (dctx != NULL) { /* can accept NULL input, like free() */ + result = (LZ4F_errorCode_t)dctx->dStage; + LZ4F_free(dctx->tmpIn, dctx->cmem); + LZ4F_free(dctx->tmpOutBuffer, dctx->cmem); + LZ4F_free(dctx, dctx->cmem); + } + return result; +} + + +/*==--- Streaming Decompression operations ---==*/ +void LZ4F_resetDecompressionContext(LZ4F_dctx* dctx) +{ + DEBUGLOG(5, "LZ4F_resetDecompressionContext"); + dctx->dStage = dstage_getFrameHeader; + dctx->dict = NULL; + dctx->dictSize = 0; + dctx->skipChecksum = 0; + dctx->frameRemainingSize = 0; +} + + +/*! LZ4F_decodeHeader() : + * input : `src` points at the **beginning of the frame** + * output : set internal values of dctx, such as + * dctx->frameInfo and dctx->dStage. + * Also allocates internal buffers. + * @return : nb Bytes read from src (necessarily <= srcSize) + * or an error code (testable with LZ4F_isError()) + */ +static size_t LZ4F_decodeHeader(LZ4F_dctx* dctx, const void* src, size_t srcSize) +{ + unsigned blockMode, blockChecksumFlag, contentSizeFlag, contentChecksumFlag, dictIDFlag, blockSizeID; + size_t frameHeaderSize; + const BYTE* srcPtr = (const BYTE*)src; + + DEBUGLOG(5, "LZ4F_decodeHeader"); + /* need to decode header to get frameInfo */ + RETURN_ERROR_IF(srcSize < minFHSize, frameHeader_incomplete); /* minimal frame header size */ + MEM_INIT(&(dctx->frameInfo), 0, sizeof(dctx->frameInfo)); + + /* special case : skippable frames */ + if ((LZ4F_readLE32(srcPtr) & 0xFFFFFFF0U) == LZ4F_MAGIC_SKIPPABLE_START) { + dctx->frameInfo.frameType = LZ4F_skippableFrame; + if (src == (void*)(dctx->header)) { + dctx->tmpInSize = srcSize; + dctx->tmpInTarget = 8; + dctx->dStage = dstage_storeSFrameSize; + return srcSize; + } else { + dctx->dStage = dstage_getSFrameSize; + return 4; + } } + + /* control magic number */ +#ifndef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION + if (LZ4F_readLE32(srcPtr) != LZ4F_MAGICNUMBER) { + DEBUGLOG(4, "frame header error : unknown magic number"); + RETURN_ERROR(frameType_unknown); + } +#endif + dctx->frameInfo.frameType = LZ4F_frame; + + /* Flags */ + { U32 const FLG = srcPtr[4]; + U32 const version = (FLG>>6) & _2BITS; + blockChecksumFlag = (FLG>>4) & _1BIT; + blockMode = (FLG>>5) & _1BIT; + contentSizeFlag = (FLG>>3) & _1BIT; + contentChecksumFlag = (FLG>>2) & _1BIT; + dictIDFlag = FLG & _1BIT; + /* validate */ + if (((FLG>>1)&_1BIT) != 0) RETURN_ERROR(reservedFlag_set); /* Reserved bit */ + if (version != 1) RETURN_ERROR(headerVersion_wrong); /* Version Number, only supported value */ + } + DEBUGLOG(6, "contentSizeFlag: %u", contentSizeFlag); + + /* Frame Header Size */ + frameHeaderSize = minFHSize + (contentSizeFlag?8:0) + (dictIDFlag?4:0); + + if (srcSize < frameHeaderSize) { + /* not enough input to fully decode frame header */ + if (srcPtr != dctx->header) + memcpy(dctx->header, srcPtr, srcSize); + dctx->tmpInSize = srcSize; + dctx->tmpInTarget = frameHeaderSize; + dctx->dStage = dstage_storeFrameHeader; + return srcSize; + } + + { U32 const BD = srcPtr[5]; + blockSizeID = (BD>>4) & _3BITS; + /* validate */ + if (((BD>>7)&_1BIT) != 0) RETURN_ERROR(reservedFlag_set); /* Reserved bit */ + if (blockSizeID < 4) RETURN_ERROR(maxBlockSize_invalid); /* 4-7 only supported values for the time being */ + if (((BD>>0)&_4BITS) != 0) RETURN_ERROR(reservedFlag_set); /* Reserved bits */ + } + + /* check header */ + assert(frameHeaderSize > 5); +#ifndef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION + { BYTE const HC = LZ4F_headerChecksum(srcPtr+4, frameHeaderSize-5); + RETURN_ERROR_IF(HC != srcPtr[frameHeaderSize-1], headerChecksum_invalid); + } +#endif + + /* save */ + dctx->frameInfo.blockMode = (LZ4F_blockMode_t)blockMode; + dctx->frameInfo.blockChecksumFlag = (LZ4F_blockChecksum_t)blockChecksumFlag; + dctx->frameInfo.contentChecksumFlag = (LZ4F_contentChecksum_t)contentChecksumFlag; + dctx->frameInfo.blockSizeID = (LZ4F_blockSizeID_t)blockSizeID; + dctx->maxBlockSize = LZ4F_getBlockSize((LZ4F_blockSizeID_t)blockSizeID); + if (contentSizeFlag) { + dctx->frameRemainingSize = dctx->frameInfo.contentSize = LZ4F_readLE64(srcPtr+6); + } + if (dictIDFlag) + dctx->frameInfo.dictID = LZ4F_readLE32(srcPtr + frameHeaderSize - 5); + + dctx->dStage = dstage_init; + + return frameHeaderSize; +} + + +/*! LZ4F_headerSize() : + * @return : size of frame header + * or an error code, which can be tested using LZ4F_isError() + */ +size_t LZ4F_headerSize(const void* src, size_t srcSize) +{ + RETURN_ERROR_IF(src == NULL, srcPtr_wrong); + + /* minimal srcSize to determine header size */ + if (srcSize < LZ4F_MIN_SIZE_TO_KNOW_HEADER_LENGTH) + RETURN_ERROR(frameHeader_incomplete); + + /* special case : skippable frames */ + if ((LZ4F_readLE32(src) & 0xFFFFFFF0U) == LZ4F_MAGIC_SKIPPABLE_START) + return 8; + + /* control magic number */ +#ifndef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION + if (LZ4F_readLE32(src) != LZ4F_MAGICNUMBER) + RETURN_ERROR(frameType_unknown); +#endif + + /* Frame Header Size */ + { BYTE const FLG = ((const BYTE*)src)[4]; + U32 const contentSizeFlag = (FLG>>3) & _1BIT; + U32 const dictIDFlag = FLG & _1BIT; + return minFHSize + (contentSizeFlag?8:0) + (dictIDFlag?4:0); + } +} + +/*! LZ4F_getFrameInfo() : + * This function extracts frame parameters (max blockSize, frame checksum, etc.). + * Usage is optional. Objective is to provide relevant information for allocation purposes. + * This function works in 2 situations : + * - At the beginning of a new frame, in which case it will decode this information from `srcBuffer`, and start the decoding process. + * Amount of input data provided must be large enough to successfully decode the frame header. + * A header size is variable, but is guaranteed to be <= LZ4F_HEADER_SIZE_MAX bytes. It's possible to provide more input data than this minimum. + * - After decoding has been started. In which case, no input is read, frame parameters are extracted from dctx. + * The number of bytes consumed from srcBuffer will be updated within *srcSizePtr (necessarily <= original value). + * Decompression must resume from (srcBuffer + *srcSizePtr). + * @return : an hint about how many srcSize bytes LZ4F_decompress() expects for next call, + * or an error code which can be tested using LZ4F_isError() + * note 1 : in case of error, dctx is not modified. Decoding operations can resume from where they stopped. + * note 2 : frame parameters are *copied into* an already allocated LZ4F_frameInfo_t structure. + */ +LZ4F_errorCode_t LZ4F_getFrameInfo(LZ4F_dctx* dctx, + LZ4F_frameInfo_t* frameInfoPtr, + const void* srcBuffer, size_t* srcSizePtr) +{ + LZ4F_STATIC_ASSERT(dstage_getFrameHeader < dstage_storeFrameHeader); + if (dctx->dStage > dstage_storeFrameHeader) { + /* frameInfo already decoded */ + size_t o=0, i=0; + *srcSizePtr = 0; + *frameInfoPtr = dctx->frameInfo; + /* returns : recommended nb of bytes for LZ4F_decompress() */ + return LZ4F_decompress(dctx, NULL, &o, NULL, &i, NULL); + } else { + if (dctx->dStage == dstage_storeFrameHeader) { + /* frame decoding already started, in the middle of header => automatic fail */ + *srcSizePtr = 0; + RETURN_ERROR(frameDecoding_alreadyStarted); + } else { + size_t const hSize = LZ4F_headerSize(srcBuffer, *srcSizePtr); + if (LZ4F_isError(hSize)) { *srcSizePtr=0; return hSize; } + if (*srcSizePtr < hSize) { + *srcSizePtr=0; + RETURN_ERROR(frameHeader_incomplete); + } + + { size_t decodeResult = LZ4F_decodeHeader(dctx, srcBuffer, hSize); + if (LZ4F_isError(decodeResult)) { + *srcSizePtr = 0; + } else { + *srcSizePtr = decodeResult; + decodeResult = BHSize; /* block header size */ + } + *frameInfoPtr = dctx->frameInfo; + return decodeResult; + } } } +} + + +/* LZ4F_updateDict() : + * only used for LZ4F_blockLinked mode + * Condition : @dstPtr != NULL + */ +static void LZ4F_updateDict(LZ4F_dctx* dctx, + const BYTE* dstPtr, size_t dstSize, const BYTE* dstBufferStart, + unsigned withinTmp) +{ + assert(dstPtr != NULL); + if (dctx->dictSize==0) dctx->dict = (const BYTE*)dstPtr; /* will lead to prefix mode */ + assert(dctx->dict != NULL); + + if (dctx->dict + dctx->dictSize == dstPtr) { /* prefix mode, everything within dstBuffer */ + dctx->dictSize += dstSize; + return; + } + + assert(dstPtr >= dstBufferStart); + if ((size_t)(dstPtr - dstBufferStart) + dstSize >= 64 KB) { /* history in dstBuffer becomes large enough to become dictionary */ + dctx->dict = (const BYTE*)dstBufferStart; + dctx->dictSize = (size_t)(dstPtr - dstBufferStart) + dstSize; + return; + } + + assert(dstSize < 64 KB); /* if dstSize >= 64 KB, dictionary would be set into dstBuffer directly */ + + /* dstBuffer does not contain whole useful history (64 KB), so it must be saved within tmpOutBuffer */ + assert(dctx->tmpOutBuffer != NULL); + + if (withinTmp && (dctx->dict == dctx->tmpOutBuffer)) { /* continue history within tmpOutBuffer */ + /* withinTmp expectation : content of [dstPtr,dstSize] is same as [dict+dictSize,dstSize], so we just extend it */ + assert(dctx->dict + dctx->dictSize == dctx->tmpOut + dctx->tmpOutStart); + dctx->dictSize += dstSize; + return; + } + + if (withinTmp) { /* copy relevant dict portion in front of tmpOut within tmpOutBuffer */ + size_t const preserveSize = (size_t)(dctx->tmpOut - dctx->tmpOutBuffer); + size_t copySize = 64 KB - dctx->tmpOutSize; + const BYTE* const oldDictEnd = dctx->dict + dctx->dictSize - dctx->tmpOutStart; + if (dctx->tmpOutSize > 64 KB) copySize = 0; + if (copySize > preserveSize) copySize = preserveSize; + + memcpy(dctx->tmpOutBuffer + preserveSize - copySize, oldDictEnd - copySize, copySize); + + dctx->dict = dctx->tmpOutBuffer; + dctx->dictSize = preserveSize + dctx->tmpOutStart + dstSize; + return; + } + + if (dctx->dict == dctx->tmpOutBuffer) { /* copy dst into tmp to complete dict */ + if (dctx->dictSize + dstSize > dctx->maxBufferSize) { /* tmp buffer not large enough */ + size_t const preserveSize = 64 KB - dstSize; + memcpy(dctx->tmpOutBuffer, dctx->dict + dctx->dictSize - preserveSize, preserveSize); + dctx->dictSize = preserveSize; + } + memcpy(dctx->tmpOutBuffer + dctx->dictSize, dstPtr, dstSize); + dctx->dictSize += dstSize; + return; + } + + /* join dict & dest into tmp */ + { size_t preserveSize = 64 KB - dstSize; + if (preserveSize > dctx->dictSize) preserveSize = dctx->dictSize; + memcpy(dctx->tmpOutBuffer, dctx->dict + dctx->dictSize - preserveSize, preserveSize); + memcpy(dctx->tmpOutBuffer + preserveSize, dstPtr, dstSize); + dctx->dict = dctx->tmpOutBuffer; + dctx->dictSize = preserveSize + dstSize; + } +} + + +/*! LZ4F_decompress() : + * Call this function repetitively to regenerate compressed data in srcBuffer. + * The function will attempt to decode up to *srcSizePtr bytes from srcBuffer + * into dstBuffer of capacity *dstSizePtr. + * + * The number of bytes regenerated into dstBuffer will be provided within *dstSizePtr (necessarily <= original value). + * + * The number of bytes effectively read from srcBuffer will be provided within *srcSizePtr (necessarily <= original value). + * If number of bytes read is < number of bytes provided, then decompression operation is not complete. + * Remaining data will have to be presented again in a subsequent invocation. + * + * The function result is an hint of the better srcSize to use for next call to LZ4F_decompress. + * Schematically, it's the size of the current (or remaining) compressed block + header of next block. + * Respecting the hint provides a small boost to performance, since it allows less buffer shuffling. + * Note that this is just a hint, and it's always possible to any srcSize value. + * When a frame is fully decoded, @return will be 0. + * If decompression failed, @return is an error code which can be tested using LZ4F_isError(). + */ +size_t LZ4F_decompress(LZ4F_dctx* dctx, + void* dstBuffer, size_t* dstSizePtr, + const void* srcBuffer, size_t* srcSizePtr, + const LZ4F_decompressOptions_t* decompressOptionsPtr) +{ + LZ4F_decompressOptions_t optionsNull; + const BYTE* const srcStart = (const BYTE*)srcBuffer; + const BYTE* const srcEnd = srcStart + *srcSizePtr; + const BYTE* srcPtr = srcStart; + BYTE* const dstStart = (BYTE*)dstBuffer; + BYTE* const dstEnd = dstStart ? dstStart + *dstSizePtr : NULL; + BYTE* dstPtr = dstStart; + const BYTE* selectedIn = NULL; + unsigned doAnotherStage = 1; + size_t nextSrcSizeHint = 1; + + + DEBUGLOG(5, "LZ4F_decompress: src[%p](%u) => dst[%p](%u)", + srcBuffer, (unsigned)*srcSizePtr, dstBuffer, (unsigned)*dstSizePtr); + if (dstBuffer == NULL) assert(*dstSizePtr == 0); + MEM_INIT(&optionsNull, 0, sizeof(optionsNull)); + if (decompressOptionsPtr==NULL) decompressOptionsPtr = &optionsNull; + *srcSizePtr = 0; + *dstSizePtr = 0; + assert(dctx != NULL); + dctx->skipChecksum |= (decompressOptionsPtr->skipChecksums != 0); /* once set, disable for the remainder of the frame */ + + /* behaves as a state machine */ + + while (doAnotherStage) { + + switch(dctx->dStage) + { + + case dstage_getFrameHeader: + DEBUGLOG(6, "dstage_getFrameHeader"); + if ((size_t)(srcEnd-srcPtr) >= maxFHSize) { /* enough to decode - shortcut */ + size_t const hSize = LZ4F_decodeHeader(dctx, srcPtr, (size_t)(srcEnd-srcPtr)); /* will update dStage appropriately */ + FORWARD_IF_ERROR(hSize); + srcPtr += hSize; + break; + } + dctx->tmpInSize = 0; + if (srcEnd-srcPtr == 0) return minFHSize; /* 0-size input */ + dctx->tmpInTarget = minFHSize; /* minimum size to decode header */ + dctx->dStage = dstage_storeFrameHeader; + /* fall-through */ + + case dstage_storeFrameHeader: + DEBUGLOG(6, "dstage_storeFrameHeader"); + { size_t const sizeToCopy = MIN(dctx->tmpInTarget - dctx->tmpInSize, (size_t)(srcEnd - srcPtr)); + memcpy(dctx->header + dctx->tmpInSize, srcPtr, sizeToCopy); + dctx->tmpInSize += sizeToCopy; + srcPtr += sizeToCopy; + } + if (dctx->tmpInSize < dctx->tmpInTarget) { + nextSrcSizeHint = (dctx->tmpInTarget - dctx->tmpInSize) + BHSize; /* rest of header + nextBlockHeader */ + doAnotherStage = 0; /* not enough src data, ask for some more */ + break; + } + FORWARD_IF_ERROR( LZ4F_decodeHeader(dctx, dctx->header, dctx->tmpInTarget) ); /* will update dStage appropriately */ + break; + + case dstage_init: + DEBUGLOG(6, "dstage_init"); + if (dctx->frameInfo.contentChecksumFlag) (void)XXH32_reset(&(dctx->xxh), 0); + /* internal buffers allocation */ + { size_t const bufferNeeded = dctx->maxBlockSize + + ((dctx->frameInfo.blockMode==LZ4F_blockLinked) ? 128 KB : 0); + if (bufferNeeded > dctx->maxBufferSize) { /* tmp buffers too small */ + dctx->maxBufferSize = 0; /* ensure allocation will be re-attempted on next entry*/ + LZ4F_free(dctx->tmpIn, dctx->cmem); + dctx->tmpIn = (BYTE*)LZ4F_malloc(dctx->maxBlockSize + BFSize /* block checksum */, dctx->cmem); + RETURN_ERROR_IF(dctx->tmpIn == NULL, allocation_failed); + LZ4F_free(dctx->tmpOutBuffer, dctx->cmem); + dctx->tmpOutBuffer= (BYTE*)LZ4F_malloc(bufferNeeded, dctx->cmem); + RETURN_ERROR_IF(dctx->tmpOutBuffer== NULL, allocation_failed); + dctx->maxBufferSize = bufferNeeded; + } } + dctx->tmpInSize = 0; + dctx->tmpInTarget = 0; + dctx->tmpOut = dctx->tmpOutBuffer; + dctx->tmpOutStart = 0; + dctx->tmpOutSize = 0; + + dctx->dStage = dstage_getBlockHeader; + /* fall-through */ + + case dstage_getBlockHeader: + if ((size_t)(srcEnd - srcPtr) >= BHSize) { + selectedIn = srcPtr; + srcPtr += BHSize; + } else { + /* not enough input to read cBlockSize field */ + dctx->tmpInSize = 0; + dctx->dStage = dstage_storeBlockHeader; + } + + if (dctx->dStage == dstage_storeBlockHeader) /* can be skipped */ + case dstage_storeBlockHeader: + { size_t const remainingInput = (size_t)(srcEnd - srcPtr); + size_t const wantedData = BHSize - dctx->tmpInSize; + size_t const sizeToCopy = MIN(wantedData, remainingInput); + memcpy(dctx->tmpIn + dctx->tmpInSize, srcPtr, sizeToCopy); + srcPtr += sizeToCopy; + dctx->tmpInSize += sizeToCopy; + + if (dctx->tmpInSize < BHSize) { /* not enough input for cBlockSize */ + nextSrcSizeHint = BHSize - dctx->tmpInSize; + doAnotherStage = 0; + break; + } + selectedIn = dctx->tmpIn; + } /* if (dctx->dStage == dstage_storeBlockHeader) */ + + /* decode block header */ + { U32 const blockHeader = LZ4F_readLE32(selectedIn); + size_t const nextCBlockSize = blockHeader & 0x7FFFFFFFU; + size_t const crcSize = dctx->frameInfo.blockChecksumFlag * BFSize; + if (blockHeader==0) { /* frameEnd signal, no more block */ + DEBUGLOG(5, "end of frame"); + dctx->dStage = dstage_getSuffix; + break; + } + if (nextCBlockSize > dctx->maxBlockSize) { + RETURN_ERROR(maxBlockSize_invalid); + } + if (blockHeader & LZ4F_BLOCKUNCOMPRESSED_FLAG) { + /* next block is uncompressed */ + dctx->tmpInTarget = nextCBlockSize; + DEBUGLOG(5, "next block is uncompressed (size %u)", (U32)nextCBlockSize); + if (dctx->frameInfo.blockChecksumFlag) { + (void)XXH32_reset(&dctx->blockChecksum, 0); + } + dctx->dStage = dstage_copyDirect; + break; + } + /* next block is a compressed block */ + dctx->tmpInTarget = nextCBlockSize + crcSize; + dctx->dStage = dstage_getCBlock; + if (dstPtr==dstEnd || srcPtr==srcEnd) { + nextSrcSizeHint = BHSize + nextCBlockSize + crcSize; + doAnotherStage = 0; + } + break; + } + + case dstage_copyDirect: /* uncompressed block */ + DEBUGLOG(6, "dstage_copyDirect"); + { size_t sizeToCopy; + if (dstPtr == NULL) { + sizeToCopy = 0; + } else { + size_t const minBuffSize = MIN((size_t)(srcEnd-srcPtr), (size_t)(dstEnd-dstPtr)); + sizeToCopy = MIN(dctx->tmpInTarget, minBuffSize); + memcpy(dstPtr, srcPtr, sizeToCopy); + if (!dctx->skipChecksum) { + if (dctx->frameInfo.blockChecksumFlag) { + (void)XXH32_update(&dctx->blockChecksum, srcPtr, sizeToCopy); + } + if (dctx->frameInfo.contentChecksumFlag) + (void)XXH32_update(&dctx->xxh, srcPtr, sizeToCopy); + } + if (dctx->frameInfo.contentSize) + dctx->frameRemainingSize -= sizeToCopy; + + /* history management (linked blocks only)*/ + if (dctx->frameInfo.blockMode == LZ4F_blockLinked) { + LZ4F_updateDict(dctx, dstPtr, sizeToCopy, dstStart, 0); + } + srcPtr += sizeToCopy; + dstPtr += sizeToCopy; + } + if (sizeToCopy == dctx->tmpInTarget) { /* all done */ + if (dctx->frameInfo.blockChecksumFlag) { + dctx->tmpInSize = 0; + dctx->dStage = dstage_getBlockChecksum; + } else + dctx->dStage = dstage_getBlockHeader; /* new block */ + break; + } + dctx->tmpInTarget -= sizeToCopy; /* need to copy more */ + } + nextSrcSizeHint = dctx->tmpInTarget + + +(dctx->frameInfo.blockChecksumFlag ? BFSize : 0) + + BHSize /* next header size */; + doAnotherStage = 0; + break; + + /* check block checksum for recently transferred uncompressed block */ + case dstage_getBlockChecksum: + DEBUGLOG(6, "dstage_getBlockChecksum"); + { const void* crcSrc; + if ((srcEnd-srcPtr >= 4) && (dctx->tmpInSize==0)) { + crcSrc = srcPtr; + srcPtr += 4; + } else { + size_t const stillToCopy = 4 - dctx->tmpInSize; + size_t const sizeToCopy = MIN(stillToCopy, (size_t)(srcEnd-srcPtr)); + memcpy(dctx->header + dctx->tmpInSize, srcPtr, sizeToCopy); + dctx->tmpInSize += sizeToCopy; + srcPtr += sizeToCopy; + if (dctx->tmpInSize < 4) { /* all input consumed */ + doAnotherStage = 0; + break; + } + crcSrc = dctx->header; + } + if (!dctx->skipChecksum) { + U32 const readCRC = LZ4F_readLE32(crcSrc); + U32 const calcCRC = XXH32_digest(&dctx->blockChecksum); +#ifndef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION + DEBUGLOG(6, "compare block checksum"); + if (readCRC != calcCRC) { + DEBUGLOG(4, "incorrect block checksum: %08X != %08X", + readCRC, calcCRC); + RETURN_ERROR(blockChecksum_invalid); + } +#else + (void)readCRC; + (void)calcCRC; +#endif + } } + dctx->dStage = dstage_getBlockHeader; /* new block */ + break; + + case dstage_getCBlock: + DEBUGLOG(6, "dstage_getCBlock"); + if ((size_t)(srcEnd-srcPtr) < dctx->tmpInTarget) { + dctx->tmpInSize = 0; + dctx->dStage = dstage_storeCBlock; + break; + } + /* input large enough to read full block directly */ + selectedIn = srcPtr; + srcPtr += dctx->tmpInTarget; + + if (0) /* always jump over next block */ + case dstage_storeCBlock: + { size_t const wantedData = dctx->tmpInTarget - dctx->tmpInSize; + size_t const inputLeft = (size_t)(srcEnd-srcPtr); + size_t const sizeToCopy = MIN(wantedData, inputLeft); + memcpy(dctx->tmpIn + dctx->tmpInSize, srcPtr, sizeToCopy); + dctx->tmpInSize += sizeToCopy; + srcPtr += sizeToCopy; + if (dctx->tmpInSize < dctx->tmpInTarget) { /* need more input */ + nextSrcSizeHint = (dctx->tmpInTarget - dctx->tmpInSize) + + (dctx->frameInfo.blockChecksumFlag ? BFSize : 0) + + BHSize /* next header size */; + doAnotherStage = 0; + break; + } + selectedIn = dctx->tmpIn; + } + + /* At this stage, input is large enough to decode a block */ + + /* First, decode and control block checksum if it exists */ + if (dctx->frameInfo.blockChecksumFlag) { + assert(dctx->tmpInTarget >= 4); + dctx->tmpInTarget -= 4; + assert(selectedIn != NULL); /* selectedIn is defined at this stage (either srcPtr, or dctx->tmpIn) */ + { U32 const readBlockCrc = LZ4F_readLE32(selectedIn + dctx->tmpInTarget); + U32 const calcBlockCrc = XXH32(selectedIn, dctx->tmpInTarget, 0); +#ifndef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION + RETURN_ERROR_IF(readBlockCrc != calcBlockCrc, blockChecksum_invalid); +#else + (void)readBlockCrc; + (void)calcBlockCrc; +#endif + } } + + /* decode directly into destination buffer if there is enough room */ + if ( ((size_t)(dstEnd-dstPtr) >= dctx->maxBlockSize) + /* unless the dictionary is stored in tmpOut: + * in which case it's faster to decode within tmpOut + * to benefit from prefix speedup */ + && !(dctx->dict!= NULL && (const BYTE*)dctx->dict + dctx->dictSize == dctx->tmpOut) ) + { + const char* dict = (const char*)dctx->dict; + size_t dictSize = dctx->dictSize; + int decodedSize; + assert(dstPtr != NULL); + if (dict && dictSize > 1 GB) { + /* overflow control : dctx->dictSize is an int, avoid truncation / sign issues */ + dict += dictSize - 64 KB; + dictSize = 64 KB; + } + decodedSize = LZ4_decompress_safe_usingDict( + (const char*)selectedIn, (char*)dstPtr, + (int)dctx->tmpInTarget, (int)dctx->maxBlockSize, + dict, (int)dictSize); + RETURN_ERROR_IF(decodedSize < 0, decompressionFailed); + if ((dctx->frameInfo.contentChecksumFlag) && (!dctx->skipChecksum)) + XXH32_update(&(dctx->xxh), dstPtr, (size_t)decodedSize); + if (dctx->frameInfo.contentSize) + dctx->frameRemainingSize -= (size_t)decodedSize; + + /* dictionary management */ + if (dctx->frameInfo.blockMode==LZ4F_blockLinked) { + LZ4F_updateDict(dctx, dstPtr, (size_t)decodedSize, dstStart, 0); + } + + dstPtr += decodedSize; + dctx->dStage = dstage_getBlockHeader; /* end of block, let's get another one */ + break; + } + + /* not enough place into dst : decode into tmpOut */ + + /* manage dictionary */ + if (dctx->frameInfo.blockMode == LZ4F_blockLinked) { + if (dctx->dict == dctx->tmpOutBuffer) { + /* truncate dictionary to 64 KB if too big */ + if (dctx->dictSize > 128 KB) { + memcpy(dctx->tmpOutBuffer, dctx->dict + dctx->dictSize - 64 KB, 64 KB); + dctx->dictSize = 64 KB; + } + dctx->tmpOut = dctx->tmpOutBuffer + dctx->dictSize; + } else { /* dict not within tmpOut */ + size_t const reservedDictSpace = MIN(dctx->dictSize, 64 KB); + dctx->tmpOut = dctx->tmpOutBuffer + reservedDictSpace; + } } + + /* Decode block into tmpOut */ + { const char* dict = (const char*)dctx->dict; + size_t dictSize = dctx->dictSize; + int decodedSize; + if (dict && dictSize > 1 GB) { + /* the dictSize param is an int, avoid truncation / sign issues */ + dict += dictSize - 64 KB; + dictSize = 64 KB; + } + decodedSize = LZ4_decompress_safe_usingDict( + (const char*)selectedIn, (char*)dctx->tmpOut, + (int)dctx->tmpInTarget, (int)dctx->maxBlockSize, + dict, (int)dictSize); + RETURN_ERROR_IF(decodedSize < 0, decompressionFailed); + if (dctx->frameInfo.contentChecksumFlag && !dctx->skipChecksum) + XXH32_update(&(dctx->xxh), dctx->tmpOut, (size_t)decodedSize); + if (dctx->frameInfo.contentSize) + dctx->frameRemainingSize -= (size_t)decodedSize; + dctx->tmpOutSize = (size_t)decodedSize; + dctx->tmpOutStart = 0; + dctx->dStage = dstage_flushOut; + } + /* fall-through */ + + case dstage_flushOut: /* flush decoded data from tmpOut to dstBuffer */ + DEBUGLOG(6, "dstage_flushOut"); + if (dstPtr != NULL) { + size_t const sizeToCopy = MIN(dctx->tmpOutSize - dctx->tmpOutStart, (size_t)(dstEnd-dstPtr)); + memcpy(dstPtr, dctx->tmpOut + dctx->tmpOutStart, sizeToCopy); + + /* dictionary management */ + if (dctx->frameInfo.blockMode == LZ4F_blockLinked) + LZ4F_updateDict(dctx, dstPtr, sizeToCopy, dstStart, 1 /*withinTmp*/); + + dctx->tmpOutStart += sizeToCopy; + dstPtr += sizeToCopy; + } + if (dctx->tmpOutStart == dctx->tmpOutSize) { /* all flushed */ + dctx->dStage = dstage_getBlockHeader; /* get next block */ + break; + } + /* could not flush everything : stop there, just request a block header */ + doAnotherStage = 0; + nextSrcSizeHint = BHSize; + break; + + case dstage_getSuffix: + RETURN_ERROR_IF(dctx->frameRemainingSize, frameSize_wrong); /* incorrect frame size decoded */ + if (!dctx->frameInfo.contentChecksumFlag) { /* no checksum, frame is completed */ + nextSrcSizeHint = 0; + LZ4F_resetDecompressionContext(dctx); + doAnotherStage = 0; + break; + } + if ((srcEnd - srcPtr) < 4) { /* not enough size for entire CRC */ + dctx->tmpInSize = 0; + dctx->dStage = dstage_storeSuffix; + } else { + selectedIn = srcPtr; + srcPtr += 4; + } + + if (dctx->dStage == dstage_storeSuffix) /* can be skipped */ + case dstage_storeSuffix: + { size_t const remainingInput = (size_t)(srcEnd - srcPtr); + size_t const wantedData = 4 - dctx->tmpInSize; + size_t const sizeToCopy = MIN(wantedData, remainingInput); + memcpy(dctx->tmpIn + dctx->tmpInSize, srcPtr, sizeToCopy); + srcPtr += sizeToCopy; + dctx->tmpInSize += sizeToCopy; + if (dctx->tmpInSize < 4) { /* not enough input to read complete suffix */ + nextSrcSizeHint = 4 - dctx->tmpInSize; + doAnotherStage=0; + break; + } + selectedIn = dctx->tmpIn; + } /* if (dctx->dStage == dstage_storeSuffix) */ + + /* case dstage_checkSuffix: */ /* no direct entry, avoid initialization risks */ + if (!dctx->skipChecksum) { + U32 const readCRC = LZ4F_readLE32(selectedIn); + U32 const resultCRC = XXH32_digest(&(dctx->xxh)); + DEBUGLOG(4, "frame checksum: stored 0x%0X vs 0x%0X processed", readCRC, resultCRC); +#ifndef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION + RETURN_ERROR_IF(readCRC != resultCRC, contentChecksum_invalid); +#else + (void)readCRC; + (void)resultCRC; +#endif + } + nextSrcSizeHint = 0; + LZ4F_resetDecompressionContext(dctx); + doAnotherStage = 0; + break; + + case dstage_getSFrameSize: + if ((srcEnd - srcPtr) >= 4) { + selectedIn = srcPtr; + srcPtr += 4; + } else { + /* not enough input to read cBlockSize field */ + dctx->tmpInSize = 4; + dctx->tmpInTarget = 8; + dctx->dStage = dstage_storeSFrameSize; + } + + if (dctx->dStage == dstage_storeSFrameSize) + case dstage_storeSFrameSize: + { size_t const sizeToCopy = MIN(dctx->tmpInTarget - dctx->tmpInSize, + (size_t)(srcEnd - srcPtr) ); + memcpy(dctx->header + dctx->tmpInSize, srcPtr, sizeToCopy); + srcPtr += sizeToCopy; + dctx->tmpInSize += sizeToCopy; + if (dctx->tmpInSize < dctx->tmpInTarget) { + /* not enough input to get full sBlockSize; wait for more */ + nextSrcSizeHint = dctx->tmpInTarget - dctx->tmpInSize; + doAnotherStage = 0; + break; + } + selectedIn = dctx->header + 4; + } /* if (dctx->dStage == dstage_storeSFrameSize) */ + + /* case dstage_decodeSFrameSize: */ /* no direct entry */ + { size_t const SFrameSize = LZ4F_readLE32(selectedIn); + dctx->frameInfo.contentSize = SFrameSize; + dctx->tmpInTarget = SFrameSize; + dctx->dStage = dstage_skipSkippable; + break; + } + + case dstage_skipSkippable: + { size_t const skipSize = MIN(dctx->tmpInTarget, (size_t)(srcEnd-srcPtr)); + srcPtr += skipSize; + dctx->tmpInTarget -= skipSize; + doAnotherStage = 0; + nextSrcSizeHint = dctx->tmpInTarget; + if (nextSrcSizeHint) break; /* still more to skip */ + /* frame fully skipped : prepare context for a new frame */ + LZ4F_resetDecompressionContext(dctx); + break; + } + } /* switch (dctx->dStage) */ + } /* while (doAnotherStage) */ + + /* preserve history within tmpOut whenever necessary */ + LZ4F_STATIC_ASSERT((unsigned)dstage_init == 2); + if ( (dctx->frameInfo.blockMode==LZ4F_blockLinked) /* next block will use up to 64KB from previous ones */ + && (dctx->dict != dctx->tmpOutBuffer) /* dictionary is not already within tmp */ + && (dctx->dict != NULL) /* dictionary exists */ + && (!decompressOptionsPtr->stableDst) /* cannot rely on dst data to remain there for next call */ + && ((unsigned)(dctx->dStage)-2 < (unsigned)(dstage_getSuffix)-2) ) /* valid stages : [init ... getSuffix[ */ + { + if (dctx->dStage == dstage_flushOut) { + size_t const preserveSize = (size_t)(dctx->tmpOut - dctx->tmpOutBuffer); + size_t copySize = 64 KB - dctx->tmpOutSize; + const BYTE* oldDictEnd = dctx->dict + dctx->dictSize - dctx->tmpOutStart; + if (dctx->tmpOutSize > 64 KB) copySize = 0; + if (copySize > preserveSize) copySize = preserveSize; + assert(dctx->tmpOutBuffer != NULL); + + memcpy(dctx->tmpOutBuffer + preserveSize - copySize, oldDictEnd - copySize, copySize); + + dctx->dict = dctx->tmpOutBuffer; + dctx->dictSize = preserveSize + dctx->tmpOutStart; + } else { + const BYTE* const oldDictEnd = dctx->dict + dctx->dictSize; + size_t const newDictSize = MIN(dctx->dictSize, 64 KB); + + memcpy(dctx->tmpOutBuffer, oldDictEnd - newDictSize, newDictSize); + + dctx->dict = dctx->tmpOutBuffer; + dctx->dictSize = newDictSize; + dctx->tmpOut = dctx->tmpOutBuffer + newDictSize; + } + } + + *srcSizePtr = (size_t)(srcPtr - srcStart); + *dstSizePtr = (size_t)(dstPtr - dstStart); + return nextSrcSizeHint; +} + +/*! LZ4F_decompress_usingDict() : + * Same as LZ4F_decompress(), using a predefined dictionary. + * Dictionary is used "in place", without any preprocessing. + * It must remain accessible throughout the entire frame decoding. + */ +size_t LZ4F_decompress_usingDict(LZ4F_dctx* dctx, + void* dstBuffer, size_t* dstSizePtr, + const void* srcBuffer, size_t* srcSizePtr, + const void* dict, size_t dictSize, + const LZ4F_decompressOptions_t* decompressOptionsPtr) +{ + if (dctx->dStage <= dstage_init) { + dctx->dict = (const BYTE*)dict; + dctx->dictSize = dictSize; + } + return LZ4F_decompress(dctx, dstBuffer, dstSizePtr, + srcBuffer, srcSizePtr, + decompressOptionsPtr); +} diff --git a/example/android/third_party/lz4/include/lz4frame.h b/example/android/third_party/lz4/include/lz4frame.h new file mode 100644 index 00000000..88d59df1 --- /dev/null +++ b/example/android/third_party/lz4/include/lz4frame.h @@ -0,0 +1,746 @@ +/* + LZ4F - LZ4-Frame library + Header File + Copyright (C) 2011-2020, Yann Collet. + BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) + + Redistribution and use in source and binary forms, with or without + modification, are permitted provided that the following conditions are + met: + + * Redistributions of source code must retain the above copyright + notice, this list of conditions and the following disclaimer. + * Redistributions in binary form must reproduce the above + copyright notice, this list of conditions and the following disclaimer + in the documentation and/or other materials provided with the + distribution. + + THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + + You can contact the author at : + - LZ4 source repository : https://github.com/lz4/lz4 + - LZ4 public forum : https://groups.google.com/forum/#!forum/lz4c +*/ + +/* LZ4F is a stand-alone API able to create and decode LZ4 frames + * conformant with specification v1.6.1 in doc/lz4_Frame_format.md . + * Generated frames are compatible with `lz4` CLI. + * + * LZ4F also offers streaming capabilities. + * + * lz4.h is not required when using lz4frame.h, + * except to extract common constants such as LZ4_VERSION_NUMBER. + * */ + +#ifndef LZ4F_H_09782039843 +#define LZ4F_H_09782039843 + +#if defined (__cplusplus) +extern "C" { +#endif + +/* --- Dependency --- */ +#include /* size_t */ + + +/** + * Introduction + * + * lz4frame.h implements LZ4 frame specification: see doc/lz4_Frame_format.md . + * LZ4 Frames are compatible with `lz4` CLI, + * and designed to be interoperable with any system. +**/ + +/*-*************************************************************** + * Compiler specifics + *****************************************************************/ +/* LZ4_DLL_EXPORT : + * Enable exporting of functions when building a Windows DLL + * LZ4FLIB_VISIBILITY : + * Control library symbols visibility. + */ +#ifndef LZ4FLIB_VISIBILITY +# if defined(__GNUC__) && (__GNUC__ >= 4) +# define LZ4FLIB_VISIBILITY __attribute__ ((visibility ("default"))) +# else +# define LZ4FLIB_VISIBILITY +# endif +#endif +#if defined(LZ4_DLL_EXPORT) && (LZ4_DLL_EXPORT==1) +# define LZ4FLIB_API __declspec(dllexport) LZ4FLIB_VISIBILITY +#elif defined(LZ4_DLL_IMPORT) && (LZ4_DLL_IMPORT==1) +# define LZ4FLIB_API __declspec(dllimport) LZ4FLIB_VISIBILITY +#else +# define LZ4FLIB_API LZ4FLIB_VISIBILITY +#endif + +#ifdef LZ4F_DISABLE_DEPRECATE_WARNINGS +# define LZ4F_DEPRECATE(x) x +#else +# if defined(_MSC_VER) +# define LZ4F_DEPRECATE(x) x /* __declspec(deprecated) x - only works with C++ */ +# elif defined(__clang__) || (defined(__GNUC__) && (__GNUC__ >= 6)) +# define LZ4F_DEPRECATE(x) x __attribute__((deprecated)) +# else +# define LZ4F_DEPRECATE(x) x /* no deprecation warning for this compiler */ +# endif +#endif + + +/*-************************************ + * Error management + **************************************/ +typedef size_t LZ4F_errorCode_t; + +LZ4FLIB_API unsigned LZ4F_isError(LZ4F_errorCode_t code); /**< tells when a function result is an error code */ +LZ4FLIB_API const char* LZ4F_getErrorName(LZ4F_errorCode_t code); /**< return error code string; for debugging */ + + +/*-************************************ + * Frame compression types + ************************************* */ +/* #define LZ4F_ENABLE_OBSOLETE_ENUMS // uncomment to enable obsolete enums */ +#ifdef LZ4F_ENABLE_OBSOLETE_ENUMS +# define LZ4F_OBSOLETE_ENUM(x) , LZ4F_DEPRECATE(x) = LZ4F_##x +#else +# define LZ4F_OBSOLETE_ENUM(x) +#endif + +/* The larger the block size, the (slightly) better the compression ratio, + * though there are diminishing returns. + * Larger blocks also increase memory usage on both compression and decompression sides. + */ +typedef enum { + LZ4F_default=0, + LZ4F_max64KB=4, + LZ4F_max256KB=5, + LZ4F_max1MB=6, + LZ4F_max4MB=7 + LZ4F_OBSOLETE_ENUM(max64KB) + LZ4F_OBSOLETE_ENUM(max256KB) + LZ4F_OBSOLETE_ENUM(max1MB) + LZ4F_OBSOLETE_ENUM(max4MB) +} LZ4F_blockSizeID_t; + +/* Linked blocks sharply reduce inefficiencies when using small blocks, + * they compress better. + * However, some LZ4 decoders are only compatible with independent blocks */ +typedef enum { + LZ4F_blockLinked=0, + LZ4F_blockIndependent + LZ4F_OBSOLETE_ENUM(blockLinked) + LZ4F_OBSOLETE_ENUM(blockIndependent) +} LZ4F_blockMode_t; + +typedef enum { + LZ4F_noContentChecksum=0, + LZ4F_contentChecksumEnabled + LZ4F_OBSOLETE_ENUM(noContentChecksum) + LZ4F_OBSOLETE_ENUM(contentChecksumEnabled) +} LZ4F_contentChecksum_t; + +typedef enum { + LZ4F_noBlockChecksum=0, + LZ4F_blockChecksumEnabled +} LZ4F_blockChecksum_t; + +typedef enum { + LZ4F_frame=0, + LZ4F_skippableFrame + LZ4F_OBSOLETE_ENUM(skippableFrame) +} LZ4F_frameType_t; + +#ifdef LZ4F_ENABLE_OBSOLETE_ENUMS +typedef LZ4F_blockSizeID_t blockSizeID_t; +typedef LZ4F_blockMode_t blockMode_t; +typedef LZ4F_frameType_t frameType_t; +typedef LZ4F_contentChecksum_t contentChecksum_t; +#endif + +/*! LZ4F_frameInfo_t : + * makes it possible to set or read frame parameters. + * Structure must be first init to 0, using memset() or LZ4F_INIT_FRAMEINFO, + * setting all parameters to default. + * It's then possible to update selectively some parameters */ +typedef struct { + LZ4F_blockSizeID_t blockSizeID; /* max64KB, max256KB, max1MB, max4MB; 0 == default (LZ4F_max64KB) */ + LZ4F_blockMode_t blockMode; /* LZ4F_blockLinked, LZ4F_blockIndependent; 0 == default (LZ4F_blockLinked) */ + LZ4F_contentChecksum_t contentChecksumFlag; /* 1: add a 32-bit checksum of frame's decompressed data; 0 == default (disabled) */ + LZ4F_frameType_t frameType; /* read-only field : LZ4F_frame or LZ4F_skippableFrame */ + unsigned long long contentSize; /* Size of uncompressed content ; 0 == unknown */ + unsigned dictID; /* Dictionary ID, sent by compressor to help decoder select correct dictionary; 0 == no dictID provided */ + LZ4F_blockChecksum_t blockChecksumFlag; /* 1: each block followed by a checksum of block's compressed data; 0 == default (disabled) */ +} LZ4F_frameInfo_t; + +#define LZ4F_INIT_FRAMEINFO { LZ4F_max64KB, LZ4F_blockLinked, LZ4F_noContentChecksum, LZ4F_frame, 0ULL, 0U, LZ4F_noBlockChecksum } /* v1.8.3+ */ + +/*! LZ4F_preferences_t : + * makes it possible to supply advanced compression instructions to streaming interface. + * Structure must be first init to 0, using memset() or LZ4F_INIT_PREFERENCES, + * setting all parameters to default. + * All reserved fields must be set to zero. */ +typedef struct { + LZ4F_frameInfo_t frameInfo; + int compressionLevel; /* 0: default (fast mode); values > LZ4HC_CLEVEL_MAX count as LZ4HC_CLEVEL_MAX; values < 0 trigger "fast acceleration" */ + unsigned autoFlush; /* 1: always flush; reduces usage of internal buffers */ + unsigned favorDecSpeed; /* 1: parser favors decompression speed vs compression ratio. Only works for high compression modes (>= LZ4HC_CLEVEL_OPT_MIN) */ /* v1.8.2+ */ + unsigned reserved[3]; /* must be zero for forward compatibility */ +} LZ4F_preferences_t; + +#define LZ4F_INIT_PREFERENCES { LZ4F_INIT_FRAMEINFO, 0, 0u, 0u, { 0u, 0u, 0u } } /* v1.8.3+ */ + + +/*-********************************* +* Simple compression function +***********************************/ + +/*! LZ4F_compressFrame() : + * Compress srcBuffer content into an LZ4-compressed frame. + * It's a one shot operation, all input content is consumed, and all output is generated. + * + * Note : it's a stateless operation (no LZ4F_cctx state needed). + * In order to reduce load on the allocator, LZ4F_compressFrame(), by default, + * uses the stack to allocate space for the compression state and some table. + * If this usage of the stack is too much for your application, + * consider compiling `lz4frame.c` with compile-time macro LZ4F_HEAPMODE set to 1 instead. + * All state allocations will use the Heap. + * It also means each invocation of LZ4F_compressFrame() will trigger several internal alloc/free invocations. + * + * @dstCapacity MUST be >= LZ4F_compressFrameBound(srcSize, preferencesPtr). + * @preferencesPtr is optional : one can provide NULL, in which case all preferences are set to default. + * @return : number of bytes written into dstBuffer. + * or an error code if it fails (can be tested using LZ4F_isError()) + */ +LZ4FLIB_API size_t LZ4F_compressFrame(void* dstBuffer, size_t dstCapacity, + const void* srcBuffer, size_t srcSize, + const LZ4F_preferences_t* preferencesPtr); + +/*! LZ4F_compressFrameBound() : + * Returns the maximum possible compressed size with LZ4F_compressFrame() given srcSize and preferences. + * `preferencesPtr` is optional. It can be replaced by NULL, in which case, the function will assume default preferences. + * Note : this result is only usable with LZ4F_compressFrame(). + * It may also be relevant to LZ4F_compressUpdate() _only if_ no flush() operation is ever performed. + */ +LZ4FLIB_API size_t LZ4F_compressFrameBound(size_t srcSize, const LZ4F_preferences_t* preferencesPtr); + + +/*! LZ4F_compressionLevel_max() : + * @return maximum allowed compression level (currently: 12) + */ +LZ4FLIB_API int LZ4F_compressionLevel_max(void); /* v1.8.0+ */ + + +/*-*********************************** +* Advanced compression functions +*************************************/ +typedef struct LZ4F_cctx_s LZ4F_cctx; /* incomplete type */ +typedef LZ4F_cctx* LZ4F_compressionContext_t; /* for compatibility with older APIs, prefer using LZ4F_cctx */ + +typedef struct { + unsigned stableSrc; /* 1 == src content will remain present on future calls to LZ4F_compress(); skip copying src content within tmp buffer */ + unsigned reserved[3]; +} LZ4F_compressOptions_t; + +/*--- Resource Management ---*/ + +#define LZ4F_VERSION 100 /* This number can be used to check for an incompatible API breaking change */ +LZ4FLIB_API unsigned LZ4F_getVersion(void); + +/*! LZ4F_createCompressionContext() : + * The first thing to do is to create a compressionContext object, + * which will keep track of operation state during streaming compression. + * This is achieved using LZ4F_createCompressionContext(), which takes as argument a version, + * and a pointer to LZ4F_cctx*, to write the resulting pointer into. + * @version provided MUST be LZ4F_VERSION. It is intended to track potential version mismatch, notably when using DLL. + * The function provides a pointer to a fully allocated LZ4F_cctx object. + * @cctxPtr MUST be != NULL. + * If @return != zero, context creation failed. + * A created compression context can be employed multiple times for consecutive streaming operations. + * Once all streaming compression jobs are completed, + * the state object can be released using LZ4F_freeCompressionContext(). + * Note1 : LZ4F_freeCompressionContext() is always successful. Its return value can be ignored. + * Note2 : LZ4F_freeCompressionContext() works fine with NULL input pointers (do nothing). +**/ +LZ4FLIB_API LZ4F_errorCode_t LZ4F_createCompressionContext(LZ4F_cctx** cctxPtr, unsigned version); +LZ4FLIB_API LZ4F_errorCode_t LZ4F_freeCompressionContext(LZ4F_cctx* cctx); + + +/*---- Compression ----*/ + +#define LZ4F_HEADER_SIZE_MIN 7 /* LZ4 Frame header size can vary, depending on selected parameters */ +#define LZ4F_HEADER_SIZE_MAX 19 + +/* Size in bytes of a block header in little-endian format. Highest bit indicates if block data is uncompressed */ +#define LZ4F_BLOCK_HEADER_SIZE 4 + +/* Size in bytes of a block checksum footer in little-endian format. */ +#define LZ4F_BLOCK_CHECKSUM_SIZE 4 + +/* Size in bytes of the content checksum. */ +#define LZ4F_CONTENT_CHECKSUM_SIZE 4 + +/*! LZ4F_compressBegin() : + * will write the frame header into dstBuffer. + * dstCapacity must be >= LZ4F_HEADER_SIZE_MAX bytes. + * `prefsPtr` is optional : NULL can be provided to set all preferences to default. + * @return : number of bytes written into dstBuffer for the header + * or an error code (which can be tested using LZ4F_isError()) + */ +LZ4FLIB_API size_t LZ4F_compressBegin(LZ4F_cctx* cctx, + void* dstBuffer, size_t dstCapacity, + const LZ4F_preferences_t* prefsPtr); + +/*! LZ4F_compressBound() : + * Provides minimum dstCapacity required to guarantee success of + * LZ4F_compressUpdate(), given a srcSize and preferences, for a worst case scenario. + * When srcSize==0, LZ4F_compressBound() provides an upper bound for LZ4F_flush() and LZ4F_compressEnd() instead. + * Note that the result is only valid for a single invocation of LZ4F_compressUpdate(). + * When invoking LZ4F_compressUpdate() multiple times, + * if the output buffer is gradually filled up instead of emptied and re-used from its start, + * one must check if there is enough remaining capacity before each invocation, using LZ4F_compressBound(). + * @return is always the same for a srcSize and prefsPtr. + * prefsPtr is optional : when NULL is provided, preferences will be set to cover worst case scenario. + * tech details : + * @return if automatic flushing is not enabled, includes the possibility that internal buffer might already be filled by up to (blockSize-1) bytes. + * It also includes frame footer (ending + checksum), since it might be generated by LZ4F_compressEnd(). + * @return doesn't include frame header, as it was already generated by LZ4F_compressBegin(). + */ +LZ4FLIB_API size_t LZ4F_compressBound(size_t srcSize, const LZ4F_preferences_t* prefsPtr); + +/*! LZ4F_compressUpdate() : + * LZ4F_compressUpdate() can be called repetitively to compress as much data as necessary. + * Important rule: dstCapacity MUST be large enough to ensure operation success even in worst case situations. + * This value is provided by LZ4F_compressBound(). + * If this condition is not respected, LZ4F_compress() will fail (result is an errorCode). + * After an error, the state is left in a UB state, and must be re-initialized or freed. + * If previously an uncompressed block was written, buffered data is flushed + * before appending compressed data is continued. + * `cOptPtr` is optional : NULL can be provided, in which case all options are set to default. + * @return : number of bytes written into `dstBuffer` (it can be zero, meaning input data was just buffered). + * or an error code if it fails (which can be tested using LZ4F_isError()) + */ +LZ4FLIB_API size_t LZ4F_compressUpdate(LZ4F_cctx* cctx, + void* dstBuffer, size_t dstCapacity, + const void* srcBuffer, size_t srcSize, + const LZ4F_compressOptions_t* cOptPtr); + +/*! LZ4F_flush() : + * When data must be generated and sent immediately, without waiting for a block to be completely filled, + * it's possible to call LZ4_flush(). It will immediately compress any data buffered within cctx. + * `dstCapacity` must be large enough to ensure the operation will be successful. + * `cOptPtr` is optional : it's possible to provide NULL, all options will be set to default. + * @return : nb of bytes written into dstBuffer (can be zero, when there is no data stored within cctx) + * or an error code if it fails (which can be tested using LZ4F_isError()) + * Note : LZ4F_flush() is guaranteed to be successful when dstCapacity >= LZ4F_compressBound(0, prefsPtr). + */ +LZ4FLIB_API size_t LZ4F_flush(LZ4F_cctx* cctx, + void* dstBuffer, size_t dstCapacity, + const LZ4F_compressOptions_t* cOptPtr); + +/*! LZ4F_compressEnd() : + * To properly finish an LZ4 frame, invoke LZ4F_compressEnd(). + * It will flush whatever data remained within `cctx` (like LZ4_flush()) + * and properly finalize the frame, with an endMark and a checksum. + * `cOptPtr` is optional : NULL can be provided, in which case all options will be set to default. + * @return : nb of bytes written into dstBuffer, necessarily >= 4 (endMark), + * or an error code if it fails (which can be tested using LZ4F_isError()) + * Note : LZ4F_compressEnd() is guaranteed to be successful when dstCapacity >= LZ4F_compressBound(0, prefsPtr). + * A successful call to LZ4F_compressEnd() makes `cctx` available again for another compression task. + */ +LZ4FLIB_API size_t LZ4F_compressEnd(LZ4F_cctx* cctx, + void* dstBuffer, size_t dstCapacity, + const LZ4F_compressOptions_t* cOptPtr); + + +/*-********************************* +* Decompression functions +***********************************/ +typedef struct LZ4F_dctx_s LZ4F_dctx; /* incomplete type */ +typedef LZ4F_dctx* LZ4F_decompressionContext_t; /* compatibility with previous API versions */ + +typedef struct { + unsigned stableDst; /* pledges that last 64KB decompressed data is present right before @dstBuffer pointer. + * This optimization skips internal storage operations. + * Once set, this pledge must remain valid up to the end of current frame. */ + unsigned skipChecksums; /* disable checksum calculation and verification, even when one is present in frame, to save CPU time. + * Setting this option to 1 once disables all checksums for the rest of the frame. */ + unsigned reserved1; /* must be set to zero for forward compatibility */ + unsigned reserved0; /* idem */ +} LZ4F_decompressOptions_t; + + +/* Resource management */ + +/*! LZ4F_createDecompressionContext() : + * Create an LZ4F_dctx object, to track all decompression operations. + * @version provided MUST be LZ4F_VERSION. + * @dctxPtr MUST be valid. + * The function fills @dctxPtr with the value of a pointer to an allocated and initialized LZ4F_dctx object. + * The @return is an errorCode, which can be tested using LZ4F_isError(). + * dctx memory can be released using LZ4F_freeDecompressionContext(); + * Result of LZ4F_freeDecompressionContext() indicates current state of decompressionContext when being released. + * That is, it should be == 0 if decompression has been completed fully and correctly. + */ +LZ4FLIB_API LZ4F_errorCode_t LZ4F_createDecompressionContext(LZ4F_dctx** dctxPtr, unsigned version); +LZ4FLIB_API LZ4F_errorCode_t LZ4F_freeDecompressionContext(LZ4F_dctx* dctx); + + +/*-*********************************** +* Streaming decompression functions +*************************************/ + +#define LZ4F_MAGICNUMBER 0x184D2204U +#define LZ4F_MAGIC_SKIPPABLE_START 0x184D2A50U +#define LZ4F_MIN_SIZE_TO_KNOW_HEADER_LENGTH 5 + +/*! LZ4F_headerSize() : v1.9.0+ + * Provide the header size of a frame starting at `src`. + * `srcSize` must be >= LZ4F_MIN_SIZE_TO_KNOW_HEADER_LENGTH, + * which is enough to decode the header length. + * @return : size of frame header + * or an error code, which can be tested using LZ4F_isError() + * note : Frame header size is variable, but is guaranteed to be + * >= LZ4F_HEADER_SIZE_MIN bytes, and <= LZ4F_HEADER_SIZE_MAX bytes. + */ +LZ4FLIB_API size_t LZ4F_headerSize(const void* src, size_t srcSize); + +/*! LZ4F_getFrameInfo() : + * This function extracts frame parameters (max blockSize, dictID, etc.). + * Its usage is optional: user can also invoke LZ4F_decompress() directly. + * + * Extracted information will fill an existing LZ4F_frameInfo_t structure. + * This can be useful for allocation and dictionary identification purposes. + * + * LZ4F_getFrameInfo() can work in the following situations : + * + * 1) At the beginning of a new frame, before any invocation of LZ4F_decompress(). + * It will decode header from `srcBuffer`, + * consuming the header and starting the decoding process. + * + * Input size must be large enough to contain the full frame header. + * Frame header size can be known beforehand by LZ4F_headerSize(). + * Frame header size is variable, but is guaranteed to be >= LZ4F_HEADER_SIZE_MIN bytes, + * and not more than <= LZ4F_HEADER_SIZE_MAX bytes. + * Hence, blindly providing LZ4F_HEADER_SIZE_MAX bytes or more will always work. + * It's allowed to provide more input data than the header size, + * LZ4F_getFrameInfo() will only consume the header. + * + * If input size is not large enough, + * aka if it's smaller than header size, + * function will fail and return an error code. + * + * 2) After decoding has been started, + * it's possible to invoke LZ4F_getFrameInfo() anytime + * to extract already decoded frame parameters stored within dctx. + * + * Note that, if decoding has barely started, + * and not yet read enough information to decode the header, + * LZ4F_getFrameInfo() will fail. + * + * The number of bytes consumed from srcBuffer will be updated in *srcSizePtr (necessarily <= original value). + * LZ4F_getFrameInfo() only consumes bytes when decoding has not yet started, + * and when decoding the header has been successful. + * Decompression must then resume from (srcBuffer + *srcSizePtr). + * + * @return : a hint about how many srcSize bytes LZ4F_decompress() expects for next call, + * or an error code which can be tested using LZ4F_isError(). + * note 1 : in case of error, dctx is not modified. Decoding operation can resume from beginning safely. + * note 2 : frame parameters are *copied into* an already allocated LZ4F_frameInfo_t structure. + */ +LZ4FLIB_API size_t +LZ4F_getFrameInfo(LZ4F_dctx* dctx, + LZ4F_frameInfo_t* frameInfoPtr, + const void* srcBuffer, size_t* srcSizePtr); + +/*! LZ4F_decompress() : + * Call this function repetitively to regenerate data compressed in `srcBuffer`. + * + * The function requires a valid dctx state. + * It will read up to *srcSizePtr bytes from srcBuffer, + * and decompress data into dstBuffer, of capacity *dstSizePtr. + * + * The nb of bytes consumed from srcBuffer will be written into *srcSizePtr (necessarily <= original value). + * The nb of bytes decompressed into dstBuffer will be written into *dstSizePtr (necessarily <= original value). + * + * The function does not necessarily read all input bytes, so always check value in *srcSizePtr. + * Unconsumed source data must be presented again in subsequent invocations. + * + * `dstBuffer` can freely change between each consecutive function invocation. + * `dstBuffer` content will be overwritten. + * + * Note: if `LZ4F_getFrameInfo()` is called before `LZ4F_decompress()`, srcBuffer must be updated to reflect + * the number of bytes consumed after reading the frame header. Failure to update srcBuffer before calling + * `LZ4F_decompress()` will cause decompression failure or, even worse, successful but incorrect decompression. + * See the `LZ4F_getFrameInfo()` docs for details. + * + * @return : an hint of how many `srcSize` bytes LZ4F_decompress() expects for next call. + * Schematically, it's the size of the current (or remaining) compressed block + header of next block. + * Respecting the hint provides some small speed benefit, because it skips intermediate buffers. + * This is just a hint though, it's always possible to provide any srcSize. + * + * When a frame is fully decoded, @return will be 0 (no more data expected). + * When provided with more bytes than necessary to decode a frame, + * LZ4F_decompress() will stop reading exactly at end of current frame, and @return 0. + * + * If decompression failed, @return is an error code, which can be tested using LZ4F_isError(). + * After a decompression error, the `dctx` context is not resumable. + * Use LZ4F_resetDecompressionContext() to return to clean state. + * + * After a frame is fully decoded, dctx can be used again to decompress another frame. + */ +LZ4FLIB_API size_t +LZ4F_decompress(LZ4F_dctx* dctx, + void* dstBuffer, size_t* dstSizePtr, + const void* srcBuffer, size_t* srcSizePtr, + const LZ4F_decompressOptions_t* dOptPtr); + + +/*! LZ4F_resetDecompressionContext() : added in v1.8.0 + * In case of an error, the context is left in "undefined" state. + * In which case, it's necessary to reset it, before re-using it. + * This method can also be used to abruptly stop any unfinished decompression, + * and start a new one using same context resources. */ +LZ4FLIB_API void LZ4F_resetDecompressionContext(LZ4F_dctx* dctx); /* always successful */ + + + +#if defined (__cplusplus) +} +#endif + +#endif /* LZ4F_H_09782039843 */ + +#if defined(LZ4F_STATIC_LINKING_ONLY) && !defined(LZ4F_H_STATIC_09782039843) +#define LZ4F_H_STATIC_09782039843 + +/* Note : + * The below declarations are not stable and may change in the future. + * They are therefore only safe to depend on + * when the caller is statically linked against the library. + * To access their declarations, define LZ4F_STATIC_LINKING_ONLY. + * + * By default, these symbols aren't published into shared/dynamic libraries. + * You can override this behavior and force them to be published + * by defining LZ4F_PUBLISH_STATIC_FUNCTIONS. + * Use at your own risk. + */ + +#if defined (__cplusplus) +extern "C" { +#endif + +#ifdef LZ4F_PUBLISH_STATIC_FUNCTIONS +# define LZ4FLIB_STATIC_API LZ4FLIB_API +#else +# define LZ4FLIB_STATIC_API +#endif + + +/* --- Error List --- */ +#define LZ4F_LIST_ERRORS(ITEM) \ + ITEM(OK_NoError) \ + ITEM(ERROR_GENERIC) \ + ITEM(ERROR_maxBlockSize_invalid) \ + ITEM(ERROR_blockMode_invalid) \ + ITEM(ERROR_parameter_invalid) \ + ITEM(ERROR_compressionLevel_invalid) \ + ITEM(ERROR_headerVersion_wrong) \ + ITEM(ERROR_blockChecksum_invalid) \ + ITEM(ERROR_reservedFlag_set) \ + ITEM(ERROR_allocation_failed) \ + ITEM(ERROR_srcSize_tooLarge) \ + ITEM(ERROR_dstMaxSize_tooSmall) \ + ITEM(ERROR_frameHeader_incomplete) \ + ITEM(ERROR_frameType_unknown) \ + ITEM(ERROR_frameSize_wrong) \ + ITEM(ERROR_srcPtr_wrong) \ + ITEM(ERROR_decompressionFailed) \ + ITEM(ERROR_headerChecksum_invalid) \ + ITEM(ERROR_contentChecksum_invalid) \ + ITEM(ERROR_frameDecoding_alreadyStarted) \ + ITEM(ERROR_compressionState_uninitialized) \ + ITEM(ERROR_parameter_null) \ + ITEM(ERROR_io_write) \ + ITEM(ERROR_io_read) \ + ITEM(ERROR_maxCode) + +#define LZ4F_GENERATE_ENUM(ENUM) LZ4F_##ENUM, + +/* enum list is exposed, to handle specific errors */ +typedef enum { LZ4F_LIST_ERRORS(LZ4F_GENERATE_ENUM) + _LZ4F_dummy_error_enum_for_c89_never_used } LZ4F_errorCodes; + +LZ4FLIB_STATIC_API LZ4F_errorCodes LZ4F_getErrorCode(size_t functionResult); + +/********************************** + * Advanced compression operations + *********************************/ + +/*! LZ4F_getBlockSize() : + * @return, in scalar format (size_t), + * the maximum block size associated with @blockSizeID, + * or an error code (can be tested using LZ4F_isError()) if @blockSizeID is invalid. +**/ +LZ4FLIB_STATIC_API size_t LZ4F_getBlockSize(LZ4F_blockSizeID_t blockSizeID); + +/*! LZ4F_uncompressedUpdate() : + * LZ4F_uncompressedUpdate() can be called repetitively to add data stored as uncompressed blocks. + * Important rule: dstCapacity MUST be large enough to store the entire source buffer as + * no compression is done for this operation + * If this condition is not respected, LZ4F_uncompressedUpdate() will fail (result is an errorCode). + * After an error, the state is left in a UB state, and must be re-initialized or freed. + * If previously a compressed block was written, buffered data is flushed first, + * before appending uncompressed data is continued. + * This operation is only supported when LZ4F_blockIndependent is used. + * `cOptPtr` is optional : NULL can be provided, in which case all options are set to default. + * @return : number of bytes written into `dstBuffer` (it can be zero, meaning input data was just buffered). + * or an error code if it fails (which can be tested using LZ4F_isError()) + */ +LZ4FLIB_STATIC_API size_t +LZ4F_uncompressedUpdate(LZ4F_cctx* cctx, + void* dstBuffer, size_t dstCapacity, + const void* srcBuffer, size_t srcSize, + const LZ4F_compressOptions_t* cOptPtr); + +/********************************** + * Dictionary compression API + *********************************/ + +/* A Dictionary is useful for the compression of small messages (KB range). + * It dramatically improves compression efficiency. + * + * LZ4 can ingest any input as dictionary, though only the last 64 KB are useful. + * Better results are generally achieved by using Zstandard's Dictionary Builder + * to generate a high-quality dictionary from a set of samples. + * + * The same dictionary will have to be used on the decompression side + * for decoding to be successful. + * To help identify the correct dictionary at decoding stage, + * the frame header allows optional embedding of a dictID field. + */ + +/*! LZ4F_compressBegin_usingDict() : + * Inits dictionary compression streaming, and writes the frame header into dstBuffer. + * `dstCapacity` must be >= LZ4F_HEADER_SIZE_MAX bytes. + * `prefsPtr` is optional : you may provide NULL as argument, + * however, it's the only way to provide dictID in the frame header. + * `dictBuffer` must outlive the compression session. + * @return : number of bytes written into dstBuffer for the header, + * or an error code (which can be tested using LZ4F_isError()) + * NOTE: this entry point doesn't fully exploit the spec, + * which allows each independent block to be compressed with the dictionary. + * Currently, only the first block uses the dictionary. + * This is still technically compliant, but less efficient for large inputs. + */ +LZ4FLIB_STATIC_API size_t +LZ4F_compressBegin_usingDict(LZ4F_cctx* cctx, + void* dstBuffer, size_t dstCapacity, + const void* dictBuffer, size_t dictSize, + const LZ4F_preferences_t* prefsPtr); + +/*! LZ4F_decompress_usingDict() : + * Same as LZ4F_decompress(), using a predefined dictionary. + * Dictionary is used "in place", without any preprocessing. +** It must remain accessible throughout the entire frame decoding. */ +LZ4FLIB_STATIC_API size_t +LZ4F_decompress_usingDict(LZ4F_dctx* dctxPtr, + void* dstBuffer, size_t* dstSizePtr, + const void* srcBuffer, size_t* srcSizePtr, + const void* dict, size_t dictSize, + const LZ4F_decompressOptions_t* decompressOptionsPtr); + +/********************************** + * Bulk processing dictionary API + *********************************/ + +/* Loading a dictionary has a cost, since it involves construction of tables. + * The Bulk processing dictionary API makes it possible to share this cost + * over an arbitrary number of compression jobs, even concurrently, + * markedly improving compression latency for these cases. + */ +typedef struct LZ4F_CDict_s LZ4F_CDict; + +/*! LZ4_createCDict() : + * When compressing multiple messages / blocks using the same dictionary, it's recommended to load it just once. + * LZ4_createCDict() will create a digested dictionary, ready to start future compression operations without startup delay. + * LZ4_CDict can be created once and shared by multiple threads concurrently, since its usage is read-only. + * `dictBuffer` can be released after LZ4_CDict creation, since its content is copied within CDict */ +LZ4FLIB_STATIC_API LZ4F_CDict* LZ4F_createCDict(const void* dictBuffer, size_t dictSize); +LZ4FLIB_STATIC_API void LZ4F_freeCDict(LZ4F_CDict* CDict); + +/*! LZ4_compressFrame_usingCDict() : + * Compress an entire srcBuffer into a valid LZ4 frame using a digested Dictionary. + * cctx must point to a context created by LZ4F_createCompressionContext(). + * If cdict==NULL, compress without a dictionary. + * dstBuffer MUST be >= LZ4F_compressFrameBound(srcSize, preferencesPtr). + * If this condition is not respected, function will fail (@return an errorCode). + * The LZ4F_preferences_t structure is optional : you may provide NULL as argument, + * but it's not recommended, as it's the only way to provide dictID in the frame header. + * @return : number of bytes written into dstBuffer. + * or an error code if it fails (can be tested using LZ4F_isError()) */ +LZ4FLIB_STATIC_API size_t +LZ4F_compressFrame_usingCDict(LZ4F_cctx* cctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + const LZ4F_CDict* cdict, + const LZ4F_preferences_t* preferencesPtr); + +/*! LZ4F_compressBegin_usingCDict() : + * Inits streaming dictionary compression, and writes the frame header into dstBuffer. + * `dstCapacity` must be >= LZ4F_HEADER_SIZE_MAX bytes. + * `prefsPtr` is optional : you may provide NULL as argument, + * however, it's the only way to provide dictID in the frame header. + * `cdict` must outlive the compression session. + * @return : number of bytes written into dstBuffer for the header, + * or an error code (which can be tested using LZ4F_isError()) */ +LZ4FLIB_STATIC_API size_t +LZ4F_compressBegin_usingCDict(LZ4F_cctx* cctx, + void* dstBuffer, size_t dstCapacity, + const LZ4F_CDict* cdict, + const LZ4F_preferences_t* prefsPtr); + + +/********************************** + * Custom memory allocation + *********************************/ + +/*! Custom memory allocation : v1.9.4+ + * These prototypes make it possible to pass custom allocation/free functions. + * LZ4F_customMem is provided at state creation time, using LZ4F_create*_advanced() listed below. + * All allocation/free operations will be completed using these custom variants instead of regular ones. + */ +typedef void* (*LZ4F_AllocFunction) (void* opaqueState, size_t size); +typedef void* (*LZ4F_CallocFunction) (void* opaqueState, size_t size); +typedef void (*LZ4F_FreeFunction) (void* opaqueState, void* address); +typedef struct { + LZ4F_AllocFunction customAlloc; + LZ4F_CallocFunction customCalloc; /* optional; when not defined, uses customAlloc + memset */ + LZ4F_FreeFunction customFree; + void* opaqueState; +} LZ4F_CustomMem; +static +#ifdef __GNUC__ +__attribute__((__unused__)) +#endif +LZ4F_CustomMem const LZ4F_defaultCMem = { NULL, NULL, NULL, NULL }; /**< this constant defers to stdlib's functions */ + +LZ4FLIB_STATIC_API LZ4F_cctx* LZ4F_createCompressionContext_advanced(LZ4F_CustomMem customMem, unsigned version); +LZ4FLIB_STATIC_API LZ4F_dctx* LZ4F_createDecompressionContext_advanced(LZ4F_CustomMem customMem, unsigned version); +LZ4FLIB_STATIC_API LZ4F_CDict* LZ4F_createCDict_advanced(LZ4F_CustomMem customMem, const void* dictBuffer, size_t dictSize); + + +#if defined (__cplusplus) +} +#endif + +#endif /* defined(LZ4F_STATIC_LINKING_ONLY) && !defined(LZ4F_H_STATIC_09782039843) */ diff --git a/example/android/third_party/lz4/include/lz4frame_static.h b/example/android/third_party/lz4/include/lz4frame_static.h new file mode 100644 index 00000000..2b44a631 --- /dev/null +++ b/example/android/third_party/lz4/include/lz4frame_static.h @@ -0,0 +1,47 @@ +/* + LZ4 auto-framing library + Header File for static linking only + Copyright (C) 2011-2020, Yann Collet. + + BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) + + Redistribution and use in source and binary forms, with or without + modification, are permitted provided that the following conditions are + met: + + * Redistributions of source code must retain the above copyright + notice, this list of conditions and the following disclaimer. + * Redistributions in binary form must reproduce the above + copyright notice, this list of conditions and the following disclaimer + in the documentation and/or other materials provided with the + distribution. + + THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + + You can contact the author at : + - LZ4 source repository : https://github.com/lz4/lz4 + - LZ4 public forum : https://groups.google.com/forum/#!forum/lz4c +*/ + +#ifndef LZ4FRAME_STATIC_H_0398209384 +#define LZ4FRAME_STATIC_H_0398209384 + +/* The declarations that formerly were made here have been merged into + * lz4frame.h, protected by the LZ4F_STATIC_LINKING_ONLY macro. Going forward, + * it is recommended to simply include that header directly. + */ + +#define LZ4F_STATIC_LINKING_ONLY +#include "lz4frame.h" + +#endif /* LZ4FRAME_STATIC_H_0398209384 */ diff --git a/example/android/third_party/lz4/include/lz4hc.c b/example/android/third_party/lz4/include/lz4hc.c new file mode 100644 index 00000000..a5bdbb5e --- /dev/null +++ b/example/android/third_party/lz4/include/lz4hc.c @@ -0,0 +1,2041 @@ +/* + LZ4 HC - High Compression Mode of LZ4 + Copyright (C) 2011-2020, Yann Collet. + + BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) + + Redistribution and use in source and binary forms, with or without + modification, are permitted provided that the following conditions are + met: + + * Redistributions of source code must retain the above copyright + notice, this list of conditions and the following disclaimer. + * Redistributions in binary form must reproduce the above + copyright notice, this list of conditions and the following disclaimer + in the documentation and/or other materials provided with the + distribution. + + THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + + You can contact the author at : + - LZ4 source repository : https://github.com/lz4/lz4 + - LZ4 public forum : https://groups.google.com/forum/#!forum/lz4c +*/ +/* note : lz4hc is not an independent module, it requires lz4.h/lz4.c for proper compilation */ + + +/* ************************************* +* Tuning Parameter +***************************************/ + +/*! HEAPMODE : + * Select how stateless HC compression functions like `LZ4_compress_HC()` + * allocate memory for their workspace: + * in stack (0:fastest), or in heap (1:default, requires malloc()). + * Since workspace is rather large, heap mode is recommended. +**/ +#ifndef LZ4HC_HEAPMODE +# define LZ4HC_HEAPMODE 1 +#endif + + +/*=== Dependency ===*/ +#define LZ4_HC_STATIC_LINKING_ONLY +#include "lz4hc.h" +#include + + +/*=== Shared lz4.c code ===*/ +#ifndef LZ4_SRC_INCLUDED +# if defined(__GNUC__) +# pragma GCC diagnostic ignored "-Wunused-function" +# endif +# if defined (__clang__) +# pragma clang diagnostic ignored "-Wunused-function" +# endif +# define LZ4_COMMONDEFS_ONLY +# include "lz4.c" /* LZ4_count, constants, mem */ +#endif + + +/*=== Enums ===*/ +typedef enum { noDictCtx, usingDictCtxHc } dictCtx_directive; + + +/*=== Constants ===*/ +#define OPTIMAL_ML (int)((ML_MASK-1)+MINMATCH) +#define LZ4_OPT_NUM (1<<12) + + +/*=== Macros ===*/ +#define MIN(a,b) ( (a) < (b) ? (a) : (b) ) +#define MAX(a,b) ( (a) > (b) ? (a) : (b) ) +#define DELTANEXTU16(table, pos) table[(U16)(pos)] /* faster */ +/* Make fields passed to, and updated by LZ4HC_encodeSequence explicit */ +#define UPDATABLE(ip, op, anchor) &ip, &op, &anchor + + +/*=== Hashing ===*/ +#define LZ4HC_HASHSIZE 4 +#define HASH_FUNCTION(i) (((i) * 2654435761U) >> ((MINMATCH*8)-LZ4HC_HASH_LOG)) +static U32 LZ4HC_hashPtr(const void* ptr) { return HASH_FUNCTION(LZ4_read32(ptr)); } + +#if defined(LZ4_FORCE_MEMORY_ACCESS) && (LZ4_FORCE_MEMORY_ACCESS==2) +/* lie to the compiler about data alignment; use with caution */ +static U64 LZ4_read64(const void* memPtr) { return *(const U64*) memPtr; } + +#elif defined(LZ4_FORCE_MEMORY_ACCESS) && (LZ4_FORCE_MEMORY_ACCESS==1) +/* __pack instructions are safer, but compiler specific */ +LZ4_PACK(typedef struct { U64 u64; }) LZ4_unalign64; +static U64 LZ4_read64(const void* ptr) { return ((const LZ4_unalign64*)ptr)->u64; } + +#else /* safe and portable access using memcpy() */ +static U64 LZ4_read64(const void* memPtr) +{ + U64 val; LZ4_memcpy(&val, memPtr, sizeof(val)); return val; +} + +#endif /* LZ4_FORCE_MEMORY_ACCESS */ + +#define LZ4MID_HASHSIZE 8 +#define LZ4MID_HASHLOG (LZ4HC_HASH_LOG-1) +#define LZ4MID_HASHTABLESIZE (1 << LZ4MID_HASHLOG) + +static U32 LZ4MID_hash4(U32 v) { return (v * 2654435761U) >> (32-LZ4MID_HASHLOG); } +static U32 LZ4MID_hash4Ptr(const void* ptr) { return LZ4MID_hash4(LZ4_read32(ptr)); } +/* note: hash7 hashes the lower 56-bits. + * It presumes input was read using little endian.*/ +static U32 LZ4MID_hash7(U64 v) { return (U32)(((v << (64-56)) * 58295818150454627ULL) >> (64-LZ4MID_HASHLOG)) ; } +static U64 LZ4_readLE64(const void* memPtr); +static U32 LZ4MID_hash8Ptr(const void* ptr) { return LZ4MID_hash7(LZ4_readLE64(ptr)); } + +static U64 LZ4_readLE64(const void* memPtr) +{ + if (LZ4_isLittleEndian()) { + return LZ4_read64(memPtr); + } else { + const BYTE* p = (const BYTE*)memPtr; + /* note: relies on the compiler to simplify this expression */ + return (U64)p[0] + ((U64)p[1]<<8) + ((U64)p[2]<<16) + ((U64)p[3]<<24) + + ((U64)p[4]<<32) + ((U64)p[5]<<40) + ((U64)p[6]<<48) + ((U64)p[7]<<56); + } +} + + +/*=== Count match length ===*/ +LZ4_FORCE_INLINE +unsigned LZ4HC_NbCommonBytes32(U32 val) +{ + assert(val != 0); + if (LZ4_isLittleEndian()) { +# if defined(_MSC_VER) && (_MSC_VER >= 1400) && !defined(LZ4_FORCE_SW_BITCOUNT) + unsigned long r; + _BitScanReverse(&r, val); + return (unsigned)((31 - r) >> 3); +# elif (defined(__clang__) || (defined(__GNUC__) && ((__GNUC__ > 3) || \ + ((__GNUC__ == 3) && (__GNUC_MINOR__ >= 4))))) && \ + !defined(LZ4_FORCE_SW_BITCOUNT) + return (unsigned)__builtin_clz(val) >> 3; +# else + val >>= 8; + val = ((((val + 0x00FFFF00) | 0x00FFFFFF) + val) | + (val + 0x00FF0000)) >> 24; + return (unsigned)val ^ 3; +# endif + } else { +# if defined(_MSC_VER) && (_MSC_VER >= 1400) && !defined(LZ4_FORCE_SW_BITCOUNT) + unsigned long r; + _BitScanForward(&r, val); + return (unsigned)(r >> 3); +# elif (defined(__clang__) || (defined(__GNUC__) && ((__GNUC__ > 3) || \ + ((__GNUC__ == 3) && (__GNUC_MINOR__ >= 4))))) && \ + !defined(LZ4_FORCE_SW_BITCOUNT) + return (unsigned)__builtin_ctz(val) >> 3; +# else + const U32 m = 0x01010101; + return (unsigned)((((val - 1) ^ val) & (m - 1)) * m) >> 24; +# endif + } +} + +/** LZ4HC_countBack() : + * @return : negative value, nb of common bytes before ip/match */ +LZ4_FORCE_INLINE +int LZ4HC_countBack(const BYTE* const ip, const BYTE* const match, + const BYTE* const iMin, const BYTE* const mMin) +{ + int back = 0; + int const min = (int)MAX(iMin - ip, mMin - match); + assert(min <= 0); + assert(ip >= iMin); assert((size_t)(ip-iMin) < (1U<<31)); + assert(match >= mMin); assert((size_t)(match - mMin) < (1U<<31)); + + while ((back - min) > 3) { + U32 const v = LZ4_read32(ip + back - 4) ^ LZ4_read32(match + back - 4); + if (v) { + return (back - (int)LZ4HC_NbCommonBytes32(v)); + } else back -= 4; /* 4-byte step */ + } + /* check remainder if any */ + while ( (back > min) + && (ip[back-1] == match[back-1]) ) + back--; + return back; +} + + +/************************************** +* Init +**************************************/ +static void LZ4HC_clearTables (LZ4HC_CCtx_internal* hc4) +{ + MEM_INIT(hc4->hashTable, 0, sizeof(hc4->hashTable)); + MEM_INIT(hc4->chainTable, 0xFF, sizeof(hc4->chainTable)); +} + +static void LZ4HC_init_internal (LZ4HC_CCtx_internal* hc4, const BYTE* start) +{ + size_t const bufferSize = (size_t)(hc4->end - hc4->prefixStart); + size_t newStartingOffset = bufferSize + hc4->dictLimit; + DEBUGLOG(5, "LZ4HC_init_internal"); + assert(newStartingOffset >= bufferSize); /* check overflow */ + if (newStartingOffset > 1 GB) { + LZ4HC_clearTables(hc4); + newStartingOffset = 0; + } + newStartingOffset += 64 KB; + hc4->nextToUpdate = (U32)newStartingOffset; + hc4->prefixStart = start; + hc4->end = start; + hc4->dictStart = start; + hc4->dictLimit = (U32)newStartingOffset; + hc4->lowLimit = (U32)newStartingOffset; +} + + +/************************************** +* Encode +**************************************/ +/* LZ4HC_encodeSequence() : + * @return : 0 if ok, + * 1 if buffer issue detected */ +LZ4_FORCE_INLINE int LZ4HC_encodeSequence ( + const BYTE** _ip, + BYTE** _op, + const BYTE** _anchor, + int matchLength, + int offset, + limitedOutput_directive limit, + BYTE* oend) +{ +#define ip (*_ip) +#define op (*_op) +#define anchor (*_anchor) + + size_t length; + BYTE* const token = op++; + +#if defined(LZ4_DEBUG) && (LZ4_DEBUG >= 6) + static const BYTE* start = NULL; + static U32 totalCost = 0; + U32 const pos = (start==NULL) ? 0 : (U32)(anchor - start); + U32 const ll = (U32)(ip - anchor); + U32 const llAdd = (ll>=15) ? ((ll-15) / 255) + 1 : 0; + U32 const mlAdd = (matchLength>=19) ? ((matchLength-19) / 255) + 1 : 0; + U32 const cost = 1 + llAdd + ll + 2 + mlAdd; + if (start==NULL) start = anchor; /* only works for single segment */ + /* g_debuglog_enable = (pos >= 2228) & (pos <= 2262); */ + DEBUGLOG(6, "pos:%7u -- literals:%4u, match:%4i, offset:%5i, cost:%4u + %5u", + pos, + (U32)(ip - anchor), matchLength, offset, + cost, totalCost); + totalCost += cost; +#endif + + /* Encode Literal length */ + length = (size_t)(ip - anchor); + LZ4_STATIC_ASSERT(notLimited == 0); + /* Check output limit */ + if (limit && ((op + (length / 255) + length + (2 + 1 + LASTLITERALS)) > oend)) { + DEBUGLOG(6, "Not enough room to write %i literals (%i bytes remaining)", + (int)length, (int)(oend - op)); + return 1; + } + if (length >= RUN_MASK) { + size_t len = length - RUN_MASK; + *token = (RUN_MASK << ML_BITS); + for(; len >= 255 ; len -= 255) *op++ = 255; + *op++ = (BYTE)len; + } else { + *token = (BYTE)(length << ML_BITS); + } + + /* Copy Literals */ + LZ4_wildCopy8(op, anchor, op + length); + op += length; + + /* Encode Offset */ + assert(offset <= LZ4_DISTANCE_MAX ); + assert(offset > 0); + LZ4_writeLE16(op, (U16)(offset)); op += 2; + + /* Encode MatchLength */ + assert(matchLength >= MINMATCH); + length = (size_t)matchLength - MINMATCH; + if (limit && (op + (length / 255) + (1 + LASTLITERALS) > oend)) { + DEBUGLOG(6, "Not enough room to write match length"); + return 1; /* Check output limit */ + } + if (length >= ML_MASK) { + *token += ML_MASK; + length -= ML_MASK; + for(; length >= 510 ; length -= 510) { *op++ = 255; *op++ = 255; } + if (length >= 255) { length -= 255; *op++ = 255; } + *op++ = (BYTE)length; + } else { + *token += (BYTE)(length); + } + + /* Prepare next loop */ + ip += matchLength; + anchor = ip; + + return 0; + +#undef ip +#undef op +#undef anchor +} + + +typedef struct { + int off; + int len; + int back; /* negative value */ +} LZ4HC_match_t; + +LZ4HC_match_t LZ4HC_searchExtDict(const BYTE* ip, U32 ipIndex, + const BYTE* const iLowLimit, const BYTE* const iHighLimit, + const LZ4HC_CCtx_internal* dictCtx, U32 gDictEndIndex, + int currentBestML, int nbAttempts) +{ + size_t const lDictEndIndex = (size_t)(dictCtx->end - dictCtx->prefixStart) + dictCtx->dictLimit; + U32 lDictMatchIndex = dictCtx->hashTable[LZ4HC_hashPtr(ip)]; + U32 matchIndex = lDictMatchIndex + gDictEndIndex - (U32)lDictEndIndex; + int offset = 0, sBack = 0; + assert(lDictEndIndex <= 1 GB); + if (lDictMatchIndex>0) + DEBUGLOG(7, "lDictEndIndex = %zu, lDictMatchIndex = %u", lDictEndIndex, lDictMatchIndex); + while (ipIndex - matchIndex <= LZ4_DISTANCE_MAX && nbAttempts--) { + const BYTE* const matchPtr = dictCtx->prefixStart - dictCtx->dictLimit + lDictMatchIndex; + + if (LZ4_read32(matchPtr) == LZ4_read32(ip)) { + int mlt; + int back = 0; + const BYTE* vLimit = ip + (lDictEndIndex - lDictMatchIndex); + if (vLimit > iHighLimit) vLimit = iHighLimit; + mlt = (int)LZ4_count(ip+MINMATCH, matchPtr+MINMATCH, vLimit) + MINMATCH; + back = (ip > iLowLimit) ? LZ4HC_countBack(ip, matchPtr, iLowLimit, dictCtx->prefixStart) : 0; + mlt -= back; + if (mlt > currentBestML) { + currentBestML = mlt; + offset = (int)(ipIndex - matchIndex); + sBack = back; + DEBUGLOG(7, "found match of length %i within extDictCtx", currentBestML); + } } + + { U32 const nextOffset = DELTANEXTU16(dictCtx->chainTable, lDictMatchIndex); + lDictMatchIndex -= nextOffset; + matchIndex -= nextOffset; + } } + + { LZ4HC_match_t md; + md.len = currentBestML; + md.off = offset; + md.back = sBack; + return md; + } +} + +/************************************** +* Mid Compression (level 2) +**************************************/ + +LZ4_FORCE_INLINE void +LZ4MID_addPosition(U32* hTable, U32 hValue, U32 index) +{ + hTable[hValue] = index; +} + +#define ADDPOS8(_p, _idx) LZ4MID_addPosition(hash8Table, LZ4MID_hash8Ptr(_p), _idx) +#define ADDPOS4(_p, _idx) LZ4MID_addPosition(hash4Table, LZ4MID_hash4Ptr(_p), _idx) + +static int LZ4HC_compress_2hashes ( + LZ4HC_CCtx_internal* const ctx, + const char* const src, + char* const dst, + int* srcSizePtr, + int const maxOutputSize, + const limitedOutput_directive limit, + const dictCtx_directive dict + ) +{ + U32* const hash4Table = ctx->hashTable; + U32* const hash8Table = hash4Table + LZ4MID_HASHTABLESIZE; + const BYTE* ip = (const BYTE*)src; + const BYTE* anchor = ip; + const BYTE* const iend = ip + *srcSizePtr; + const BYTE* const mflimit = iend - MFLIMIT; + const BYTE* const matchlimit = (iend - LASTLITERALS); + const BYTE* const ilimit = (iend - LZ4MID_HASHSIZE); + BYTE* op = (BYTE*)dst; + BYTE* oend = op + maxOutputSize; + + const BYTE* const prefixPtr = ctx->prefixStart; + const U32 prefixIdx = ctx->dictLimit; + const U32 ilimitIdx = (U32)(ilimit - prefixPtr) + prefixIdx; + const U32 gDictEndIndex = ctx->lowLimit; + unsigned matchLength; + unsigned matchDistance; + + /* input sanitization */ + DEBUGLOG(5, "LZ4HC_compress_2hashes (%i bytes)", *srcSizePtr); + assert(*srcSizePtr >= 0); + if (*srcSizePtr) assert(src != NULL); + if (maxOutputSize) assert(dst != NULL); + if (*srcSizePtr < 0) return 0; /* invalid */ + if (maxOutputSize < 0) return 0; /* invalid */ + if (*srcSizePtr > LZ4_MAX_INPUT_SIZE) { + /* forbidden: no input is allowed to be that large */ + return 0; + } + if (limit == fillOutput) oend -= LASTLITERALS; /* Hack for support LZ4 format restriction */ + if (*srcSizePtr < LZ4_minLength) + goto _lz4mid_last_literals; /* Input too small, no compression (all literals) */ + + /* main loop */ + while (ip <= mflimit) { + const U32 ipIndex = (U32)(ip - prefixPtr) + prefixIdx; + /* search long match */ + { U32 h8 = LZ4MID_hash8Ptr(ip); + U32 pos8 = hash8Table[h8]; + assert(h8 < LZ4MID_HASHTABLESIZE); + assert(h8 < ipIndex); + LZ4MID_addPosition(hash8Table, h8, ipIndex); + if ( ipIndex - pos8 <= LZ4_DISTANCE_MAX + && pos8 >= prefixIdx /* note: currently only search within prefix */ + ) { + /* match candidate found */ + const BYTE* matchPtr = prefixPtr + pos8 - prefixIdx; + assert(matchPtr < ip); + matchLength = LZ4_count(ip, matchPtr, matchlimit); + if (matchLength >= MINMATCH) { + DEBUGLOG(7, "found candidate match at pos %u (len=%u)", pos8, matchLength); + matchDistance = ipIndex - pos8; + goto _lz4mid_encode_sequence; + } + } } + /* search short match */ + { U32 h4 = LZ4MID_hash4Ptr(ip); + U32 pos4 = hash4Table[h4]; + assert(h4 < LZ4MID_HASHTABLESIZE); + assert(pos4 < ipIndex); + LZ4MID_addPosition(hash4Table, h4, ipIndex); + if (ipIndex - pos4 <= LZ4_DISTANCE_MAX + && pos4 >= prefixIdx /* only search within prefix */ + ) { + /* match candidate found */ + const BYTE* const matchPtr = prefixPtr + (pos4 - prefixIdx); + assert(matchPtr < ip); + assert(matchPtr >= prefixPtr); + matchLength = LZ4_count(ip, matchPtr, matchlimit); + if (matchLength >= MINMATCH) { + /* short match found, let's just check ip+1 for longer */ + U32 const h8 = LZ4MID_hash8Ptr(ip+1); + U32 const pos8 = hash8Table[h8]; + U32 const m2Distance = ipIndex + 1 - pos8; + matchDistance = ipIndex - pos4; + if ( m2Distance <= LZ4_DISTANCE_MAX + && pos8 >= prefixIdx /* only search within prefix */ + && likely(ip < mflimit) + ) { + const BYTE* const m2Ptr = prefixPtr + (pos8 - prefixIdx); + unsigned ml2 = LZ4_count(ip+1, m2Ptr, matchlimit); + if (ml2 > matchLength) { + LZ4MID_addPosition(hash8Table, h8, ipIndex+1); + ip++; + matchLength = ml2; + matchDistance = m2Distance; + } } + goto _lz4mid_encode_sequence; + } + } } + /* no match found in prefix */ + if ( (dict == usingDictCtxHc) + && (ipIndex - gDictEndIndex < LZ4_DISTANCE_MAX - 8) ) { + /* search a match in dictionary */ + LZ4HC_match_t dMatch = LZ4HC_searchExtDict(ip, ipIndex, + anchor, matchlimit, + ctx->dictCtx, gDictEndIndex, + 0, 2); + if (dMatch.len >= MINMATCH) { + DEBUGLOG(7, "found Dictionary match (offset=%i)", dMatch.off); + ip += dMatch.back; + assert(ip >= anchor); + matchLength = (unsigned)dMatch.len; + matchDistance = (unsigned)dMatch.off; + goto _lz4mid_encode_sequence; + } + } + /* no match found */ + ip += 1 + ((ip-anchor) >> 9); /* skip faster over incompressible data */ + continue; + +_lz4mid_encode_sequence: + /* catch back */ + while (((ip > anchor) & ((U32)(ip-prefixPtr) > matchDistance)) && (unlikely(ip[-1] == ip[-(int)matchDistance-1]))) { + ip--; matchLength++; + }; + + /* fill table with beginning of match */ + ADDPOS8(ip+1, ipIndex+1); + ADDPOS8(ip+2, ipIndex+2); + ADDPOS4(ip+1, ipIndex+1); + + /* encode */ + { BYTE* const saved_op = op; + /* LZ4HC_encodeSequence always updates @op; on success, it updates @ip and @anchor */ + if (LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor), + (int)matchLength, (int)matchDistance, + limit, oend) ) { + op = saved_op; /* restore @op value before failed LZ4HC_encodeSequence */ + goto _lz4mid_dest_overflow; + } + } + + /* fill table with end of match */ + { U32 endMatchIdx = (U32)(ip-prefixPtr) + prefixIdx; + U32 pos_m2 = endMatchIdx - 2; + if (pos_m2 < ilimitIdx) { + if (likely(ip - prefixPtr > 5)) { + ADDPOS8(ip-5, endMatchIdx - 5); + } + ADDPOS8(ip-3, endMatchIdx - 3); + ADDPOS8(ip-2, endMatchIdx - 2); + ADDPOS4(ip-2, endMatchIdx - 2); + ADDPOS4(ip-1, endMatchIdx - 1); + } + } + } + +_lz4mid_last_literals: + /* Encode Last Literals */ + { size_t lastRunSize = (size_t)(iend - anchor); /* literals */ + size_t llAdd = (lastRunSize + 255 - RUN_MASK) / 255; + size_t const totalSize = 1 + llAdd + lastRunSize; + if (limit == fillOutput) oend += LASTLITERALS; /* restore correct value */ + if (limit && (op + totalSize > oend)) { + if (limit == limitedOutput) return 0; /* not enough space in @dst */ + /* adapt lastRunSize to fill 'dest' */ + lastRunSize = (size_t)(oend - op) - 1 /*token*/; + llAdd = (lastRunSize + 256 - RUN_MASK) / 256; + lastRunSize -= llAdd; + } + DEBUGLOG(6, "Final literal run : %i literals", (int)lastRunSize); + ip = anchor + lastRunSize; /* can be != iend if limit==fillOutput */ + + if (lastRunSize >= RUN_MASK) { + size_t accumulator = lastRunSize - RUN_MASK; + *op++ = (RUN_MASK << ML_BITS); + for(; accumulator >= 255 ; accumulator -= 255) + *op++ = 255; + *op++ = (BYTE) accumulator; + } else { + *op++ = (BYTE)(lastRunSize << ML_BITS); + } + assert(lastRunSize <= (size_t)(oend - op)); + LZ4_memcpy(op, anchor, lastRunSize); + op += lastRunSize; + } + + /* End */ + DEBUGLOG(5, "compressed %i bytes into %i bytes", *srcSizePtr, (int)((char*)op - dst)); + assert(ip >= (const BYTE*)src); + assert(ip <= iend); + *srcSizePtr = (int)(ip - (const BYTE*)src); + assert((char*)op >= dst); + assert(op <= oend); + assert((char*)op - dst < INT_MAX); + return (int)((char*)op - dst); + +_lz4mid_dest_overflow: + if (limit == fillOutput) { + /* Assumption : @ip, @anchor, @optr and @matchLength must be set correctly */ + size_t const ll = (size_t)(ip - anchor); + size_t const ll_addbytes = (ll + 240) / 255; + size_t const ll_totalCost = 1 + ll_addbytes + ll; + BYTE* const maxLitPos = oend - 3; /* 2 for offset, 1 for token */ + DEBUGLOG(6, "Last sequence is overflowing : %u literals, %u remaining space", + (unsigned)ll, (unsigned)(oend-op)); + if (op + ll_totalCost <= maxLitPos) { + /* ll validated; now adjust match length */ + size_t const bytesLeftForMl = (size_t)(maxLitPos - (op+ll_totalCost)); + size_t const maxMlSize = MINMATCH + (ML_MASK-1) + (bytesLeftForMl * 255); + assert(maxMlSize < INT_MAX); + if ((size_t)matchLength > maxMlSize) matchLength= (unsigned)maxMlSize; + if ((oend + LASTLITERALS) - (op + ll_totalCost + 2) - 1 + matchLength >= MFLIMIT) { + DEBUGLOG(6, "Let's encode a last sequence (ll=%u, ml=%u)", (unsigned)ll, matchLength); + LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor), + (int)matchLength, (int)matchDistance, + notLimited, oend); + } } + DEBUGLOG(6, "Let's finish with a run of literals (%u bytes left)", (unsigned)(oend-op)); + goto _lz4mid_last_literals; + } + /* compression failed */ + return 0; +} + + +/************************************** +* HC Compression - Search +**************************************/ + +/* Update chains up to ip (excluded) */ +LZ4_FORCE_INLINE void LZ4HC_Insert (LZ4HC_CCtx_internal* hc4, const BYTE* ip) +{ + U16* const chainTable = hc4->chainTable; + U32* const hashTable = hc4->hashTable; + const BYTE* const prefixPtr = hc4->prefixStart; + U32 const prefixIdx = hc4->dictLimit; + U32 const target = (U32)(ip - prefixPtr) + prefixIdx; + U32 idx = hc4->nextToUpdate; + assert(ip >= prefixPtr); + assert(target >= prefixIdx); + + while (idx < target) { + U32 const h = LZ4HC_hashPtr(prefixPtr+idx-prefixIdx); + size_t delta = idx - hashTable[h]; + if (delta>LZ4_DISTANCE_MAX) delta = LZ4_DISTANCE_MAX; + DELTANEXTU16(chainTable, idx) = (U16)delta; + hashTable[h] = idx; + idx++; + } + + hc4->nextToUpdate = target; +} + +#if defined(_MSC_VER) +# define LZ4HC_rotl32(x,r) _rotl(x,r) +#else +# define LZ4HC_rotl32(x,r) ((x << r) | (x >> (32 - r))) +#endif + + +static U32 LZ4HC_rotatePattern(size_t const rotate, U32 const pattern) +{ + size_t const bitsToRotate = (rotate & (sizeof(pattern) - 1)) << 3; + if (bitsToRotate == 0) return pattern; + return LZ4HC_rotl32(pattern, (int)bitsToRotate); +} + +/* LZ4HC_countPattern() : + * pattern32 must be a sample of repetitive pattern of length 1, 2 or 4 (but not 3!) */ +static unsigned +LZ4HC_countPattern(const BYTE* ip, const BYTE* const iEnd, U32 const pattern32) +{ + const BYTE* const iStart = ip; + reg_t const pattern = (sizeof(pattern)==8) ? + (reg_t)pattern32 + (((reg_t)pattern32) << (sizeof(pattern)*4)) : pattern32; + + while (likely(ip < iEnd-(sizeof(pattern)-1))) { + reg_t const diff = LZ4_read_ARCH(ip) ^ pattern; + if (!diff) { ip+=sizeof(pattern); continue; } + ip += LZ4_NbCommonBytes(diff); + return (unsigned)(ip - iStart); + } + + if (LZ4_isLittleEndian()) { + reg_t patternByte = pattern; + while ((ip>= 8; + } + } else { /* big endian */ + U32 bitOffset = (sizeof(pattern)*8) - 8; + while (ip < iEnd) { + BYTE const byte = (BYTE)(pattern >> bitOffset); + if (*ip != byte) break; + ip ++; bitOffset -= 8; + } } + + return (unsigned)(ip - iStart); +} + +/* LZ4HC_reverseCountPattern() : + * pattern must be a sample of repetitive pattern of length 1, 2 or 4 (but not 3!) + * read using natural platform endianness */ +static unsigned +LZ4HC_reverseCountPattern(const BYTE* ip, const BYTE* const iLow, U32 pattern) +{ + const BYTE* const iStart = ip; + + while (likely(ip >= iLow+4)) { + if (LZ4_read32(ip-4) != pattern) break; + ip -= 4; + } + { const BYTE* bytePtr = (const BYTE*)(&pattern) + 3; /* works for any endianness */ + while (likely(ip>iLow)) { + if (ip[-1] != *bytePtr) break; + ip--; bytePtr--; + } } + return (unsigned)(iStart - ip); +} + +/* LZ4HC_protectDictEnd() : + * Checks if the match is in the last 3 bytes of the dictionary, so reading the + * 4 byte MINMATCH would overflow. + * @returns true if the match index is okay. + */ +static int LZ4HC_protectDictEnd(U32 const dictLimit, U32 const matchIndex) +{ + return ((U32)((dictLimit - 1) - matchIndex) >= 3); +} + +typedef enum { rep_untested, rep_not, rep_confirmed } repeat_state_e; +typedef enum { favorCompressionRatio=0, favorDecompressionSpeed } HCfavor_e; + + +LZ4_FORCE_INLINE LZ4HC_match_t +LZ4HC_InsertAndGetWiderMatch ( + LZ4HC_CCtx_internal* const hc4, + const BYTE* const ip, + const BYTE* const iLowLimit, const BYTE* const iHighLimit, + int longest, + const int maxNbAttempts, + const int patternAnalysis, const int chainSwap, + const dictCtx_directive dict, + const HCfavor_e favorDecSpeed) +{ + U16* const chainTable = hc4->chainTable; + U32* const hashTable = hc4->hashTable; + const LZ4HC_CCtx_internal* const dictCtx = hc4->dictCtx; + const BYTE* const prefixPtr = hc4->prefixStart; + const U32 prefixIdx = hc4->dictLimit; + const U32 ipIndex = (U32)(ip - prefixPtr) + prefixIdx; + const int withinStartDistance = (hc4->lowLimit + (LZ4_DISTANCE_MAX + 1) > ipIndex); + const U32 lowestMatchIndex = (withinStartDistance) ? hc4->lowLimit : ipIndex - LZ4_DISTANCE_MAX; + const BYTE* const dictStart = hc4->dictStart; + const U32 dictIdx = hc4->lowLimit; + const BYTE* const dictEnd = dictStart + prefixIdx - dictIdx; + int const lookBackLength = (int)(ip-iLowLimit); + int nbAttempts = maxNbAttempts; + U32 matchChainPos = 0; + U32 const pattern = LZ4_read32(ip); + U32 matchIndex; + repeat_state_e repeat = rep_untested; + size_t srcPatternLength = 0; + int offset = 0, sBack = 0; + + DEBUGLOG(7, "LZ4HC_InsertAndGetWiderMatch"); + /* First Match */ + LZ4HC_Insert(hc4, ip); /* insert all prior positions up to ip (excluded) */ + matchIndex = hashTable[LZ4HC_hashPtr(ip)]; + DEBUGLOG(7, "First candidate match for pos %u found at index %u / %u (lowestMatchIndex)", + ipIndex, matchIndex, lowestMatchIndex); + + while ((matchIndex>=lowestMatchIndex) && (nbAttempts>0)) { + int matchLength=0; + nbAttempts--; + assert(matchIndex < ipIndex); + if (favorDecSpeed && (ipIndex - matchIndex < 8)) { + /* do nothing: + * favorDecSpeed intentionally skips matches with offset < 8 */ + } else if (matchIndex >= prefixIdx) { /* within current Prefix */ + const BYTE* const matchPtr = prefixPtr + (matchIndex - prefixIdx); + assert(matchPtr < ip); + assert(longest >= 1); + if (LZ4_read16(iLowLimit + longest - 1) == LZ4_read16(matchPtr - lookBackLength + longest - 1)) { + if (LZ4_read32(matchPtr) == pattern) { + int const back = lookBackLength ? LZ4HC_countBack(ip, matchPtr, iLowLimit, prefixPtr) : 0; + matchLength = MINMATCH + (int)LZ4_count(ip+MINMATCH, matchPtr+MINMATCH, iHighLimit); + matchLength -= back; + if (matchLength > longest) { + longest = matchLength; + offset = (int)(ipIndex - matchIndex); + sBack = back; + DEBUGLOG(7, "Found match of len=%i within prefix, offset=%i, back=%i", longest, offset, -back); + } } } + } else { /* lowestMatchIndex <= matchIndex < dictLimit : within Ext Dict */ + const BYTE* const matchPtr = dictStart + (matchIndex - dictIdx); + assert(matchIndex >= dictIdx); + if ( likely(matchIndex <= prefixIdx - 4) + && (LZ4_read32(matchPtr) == pattern) ) { + int back = 0; + const BYTE* vLimit = ip + (prefixIdx - matchIndex); + if (vLimit > iHighLimit) vLimit = iHighLimit; + matchLength = (int)LZ4_count(ip+MINMATCH, matchPtr+MINMATCH, vLimit) + MINMATCH; + if ((ip+matchLength == vLimit) && (vLimit < iHighLimit)) + matchLength += LZ4_count(ip+matchLength, prefixPtr, iHighLimit); + back = lookBackLength ? LZ4HC_countBack(ip, matchPtr, iLowLimit, dictStart) : 0; + matchLength -= back; + if (matchLength > longest) { + longest = matchLength; + offset = (int)(ipIndex - matchIndex); + sBack = back; + DEBUGLOG(7, "Found match of len=%i within dict, offset=%i, back=%i", longest, offset, -back); + } } } + + if (chainSwap && matchLength==longest) { /* better match => select a better chain */ + assert(lookBackLength==0); /* search forward only */ + if (matchIndex + (U32)longest <= ipIndex) { + int const kTrigger = 4; + U32 distanceToNextMatch = 1; + int const end = longest - MINMATCH + 1; + int step = 1; + int accel = 1 << kTrigger; + int pos; + for (pos = 0; pos < end; pos += step) { + U32 const candidateDist = DELTANEXTU16(chainTable, matchIndex + (U32)pos); + step = (accel++ >> kTrigger); + if (candidateDist > distanceToNextMatch) { + distanceToNextMatch = candidateDist; + matchChainPos = (U32)pos; + accel = 1 << kTrigger; + } } + if (distanceToNextMatch > 1) { + if (distanceToNextMatch > matchIndex) break; /* avoid overflow */ + matchIndex -= distanceToNextMatch; + continue; + } } } + + { U32 const distNextMatch = DELTANEXTU16(chainTable, matchIndex); + if (patternAnalysis && distNextMatch==1 && matchChainPos==0) { + U32 const matchCandidateIdx = matchIndex-1; + /* may be a repeated pattern */ + if (repeat == rep_untested) { + if ( ((pattern & 0xFFFF) == (pattern >> 16)) + & ((pattern & 0xFF) == (pattern >> 24)) ) { + DEBUGLOG(7, "Repeat pattern detected, char %02X", pattern >> 24); + repeat = rep_confirmed; + srcPatternLength = LZ4HC_countPattern(ip+sizeof(pattern), iHighLimit, pattern) + sizeof(pattern); + } else { + repeat = rep_not; + } } + if ( (repeat == rep_confirmed) && (matchCandidateIdx >= lowestMatchIndex) + && LZ4HC_protectDictEnd(prefixIdx, matchCandidateIdx) ) { + const int extDict = matchCandidateIdx < prefixIdx; + const BYTE* const matchPtr = extDict ? dictStart + (matchCandidateIdx - dictIdx) : prefixPtr + (matchCandidateIdx - prefixIdx); + if (LZ4_read32(matchPtr) == pattern) { /* good candidate */ + const BYTE* const iLimit = extDict ? dictEnd : iHighLimit; + size_t forwardPatternLength = LZ4HC_countPattern(matchPtr+sizeof(pattern), iLimit, pattern) + sizeof(pattern); + if (extDict && matchPtr + forwardPatternLength == iLimit) { + U32 const rotatedPattern = LZ4HC_rotatePattern(forwardPatternLength, pattern); + forwardPatternLength += LZ4HC_countPattern(prefixPtr, iHighLimit, rotatedPattern); + } + { const BYTE* const lowestMatchPtr = extDict ? dictStart : prefixPtr; + size_t backLength = LZ4HC_reverseCountPattern(matchPtr, lowestMatchPtr, pattern); + size_t currentSegmentLength; + if (!extDict + && matchPtr - backLength == prefixPtr + && dictIdx < prefixIdx) { + U32 const rotatedPattern = LZ4HC_rotatePattern((U32)(-(int)backLength), pattern); + backLength += LZ4HC_reverseCountPattern(dictEnd, dictStart, rotatedPattern); + } + /* Limit backLength not go further than lowestMatchIndex */ + backLength = matchCandidateIdx - MAX(matchCandidateIdx - (U32)backLength, lowestMatchIndex); + assert(matchCandidateIdx - backLength >= lowestMatchIndex); + currentSegmentLength = backLength + forwardPatternLength; + /* Adjust to end of pattern if the source pattern fits, otherwise the beginning of the pattern */ + if ( (currentSegmentLength >= srcPatternLength) /* current pattern segment large enough to contain full srcPatternLength */ + && (forwardPatternLength <= srcPatternLength) ) { /* haven't reached this position yet */ + U32 const newMatchIndex = matchCandidateIdx + (U32)forwardPatternLength - (U32)srcPatternLength; /* best position, full pattern, might be followed by more match */ + if (LZ4HC_protectDictEnd(prefixIdx, newMatchIndex)) + matchIndex = newMatchIndex; + else { + /* Can only happen if started in the prefix */ + assert(newMatchIndex >= prefixIdx - 3 && newMatchIndex < prefixIdx && !extDict); + matchIndex = prefixIdx; + } + } else { + U32 const newMatchIndex = matchCandidateIdx - (U32)backLength; /* farthest position in current segment, will find a match of length currentSegmentLength + maybe some back */ + if (!LZ4HC_protectDictEnd(prefixIdx, newMatchIndex)) { + assert(newMatchIndex >= prefixIdx - 3 && newMatchIndex < prefixIdx && !extDict); + matchIndex = prefixIdx; + } else { + matchIndex = newMatchIndex; + if (lookBackLength==0) { /* no back possible */ + size_t const maxML = MIN(currentSegmentLength, srcPatternLength); + if ((size_t)longest < maxML) { + assert(prefixPtr - prefixIdx + matchIndex != ip); + if ((size_t)(ip - prefixPtr) + prefixIdx - matchIndex > LZ4_DISTANCE_MAX) break; + assert(maxML < 2 GB); + longest = (int)maxML; + offset = (int)(ipIndex - matchIndex); + assert(sBack == 0); + DEBUGLOG(7, "Found repeat pattern match of len=%i, offset=%i", longest, offset); + } + { U32 const distToNextPattern = DELTANEXTU16(chainTable, matchIndex); + if (distToNextPattern > matchIndex) break; /* avoid overflow */ + matchIndex -= distToNextPattern; + } } } } } + continue; + } } + } } /* PA optimization */ + + /* follow current chain */ + matchIndex -= DELTANEXTU16(chainTable, matchIndex + matchChainPos); + + } /* while ((matchIndex>=lowestMatchIndex) && (nbAttempts)) */ + + if ( dict == usingDictCtxHc + && nbAttempts > 0 + && withinStartDistance) { + size_t const dictEndOffset = (size_t)(dictCtx->end - dictCtx->prefixStart) + dictCtx->dictLimit; + U32 dictMatchIndex = dictCtx->hashTable[LZ4HC_hashPtr(ip)]; + assert(dictEndOffset <= 1 GB); + matchIndex = dictMatchIndex + lowestMatchIndex - (U32)dictEndOffset; + if (dictMatchIndex>0) DEBUGLOG(7, "dictEndOffset = %zu, dictMatchIndex = %u => relative matchIndex = %i", dictEndOffset, dictMatchIndex, (int)dictMatchIndex - (int)dictEndOffset); + while (ipIndex - matchIndex <= LZ4_DISTANCE_MAX && nbAttempts--) { + const BYTE* const matchPtr = dictCtx->prefixStart - dictCtx->dictLimit + dictMatchIndex; + + if (LZ4_read32(matchPtr) == pattern) { + int mlt; + int back = 0; + const BYTE* vLimit = ip + (dictEndOffset - dictMatchIndex); + if (vLimit > iHighLimit) vLimit = iHighLimit; + mlt = (int)LZ4_count(ip+MINMATCH, matchPtr+MINMATCH, vLimit) + MINMATCH; + back = lookBackLength ? LZ4HC_countBack(ip, matchPtr, iLowLimit, dictCtx->prefixStart) : 0; + mlt -= back; + if (mlt > longest) { + longest = mlt; + offset = (int)(ipIndex - matchIndex); + sBack = back; + DEBUGLOG(7, "found match of length %i within extDictCtx", longest); + } } + + { U32 const nextOffset = DELTANEXTU16(dictCtx->chainTable, dictMatchIndex); + dictMatchIndex -= nextOffset; + matchIndex -= nextOffset; + } } } + + { LZ4HC_match_t md; + assert(longest >= 0); + md.len = longest; + md.off = offset; + md.back = sBack; + return md; + } +} + +LZ4_FORCE_INLINE LZ4HC_match_t +LZ4HC_InsertAndFindBestMatch(LZ4HC_CCtx_internal* const hc4, /* Index table will be updated */ + const BYTE* const ip, const BYTE* const iLimit, + const int maxNbAttempts, + const int patternAnalysis, + const dictCtx_directive dict) +{ + DEBUGLOG(7, "LZ4HC_InsertAndFindBestMatch"); + /* note : LZ4HC_InsertAndGetWiderMatch() is able to modify the starting position of a match (*startpos), + * but this won't be the case here, as we define iLowLimit==ip, + * so LZ4HC_InsertAndGetWiderMatch() won't be allowed to search past ip */ + return LZ4HC_InsertAndGetWiderMatch(hc4, ip, ip, iLimit, MINMATCH-1, maxNbAttempts, patternAnalysis, 0 /*chainSwap*/, dict, favorCompressionRatio); +} + + +LZ4_FORCE_INLINE int LZ4HC_compress_hashChain ( + LZ4HC_CCtx_internal* const ctx, + const char* const source, + char* const dest, + int* srcSizePtr, + int const maxOutputSize, + int maxNbAttempts, + const limitedOutput_directive limit, + const dictCtx_directive dict + ) +{ + const int inputSize = *srcSizePtr; + const int patternAnalysis = (maxNbAttempts > 128); /* levels 9+ */ + + const BYTE* ip = (const BYTE*) source; + const BYTE* anchor = ip; + const BYTE* const iend = ip + inputSize; + const BYTE* const mflimit = iend - MFLIMIT; + const BYTE* const matchlimit = (iend - LASTLITERALS); + + BYTE* optr = (BYTE*) dest; + BYTE* op = (BYTE*) dest; + BYTE* oend = op + maxOutputSize; + + const BYTE* start0; + const BYTE* start2 = NULL; + const BYTE* start3 = NULL; + LZ4HC_match_t m0, m1, m2, m3; + const LZ4HC_match_t nomatch = {0, 0, 0}; + + /* init */ + DEBUGLOG(5, "LZ4HC_compress_hashChain (dict?=>%i)", dict); + *srcSizePtr = 0; + if (limit == fillOutput) oend -= LASTLITERALS; /* Hack for support LZ4 format restriction */ + if (inputSize < LZ4_minLength) goto _last_literals; /* Input too small, no compression (all literals) */ + + /* Main Loop */ + while (ip <= mflimit) { + m1 = LZ4HC_InsertAndFindBestMatch(ctx, ip, matchlimit, maxNbAttempts, patternAnalysis, dict); + if (m1.len encode ML1 immediately */ + optr = op; + if (LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor), + m1.len, m1.off, + limit, oend) ) + goto _dest_overflow; + continue; + } + + if (start0 < ip) { /* first match was skipped at least once */ + if (start2 < ip + m0.len) { /* squeezing ML1 between ML0(original ML1) and ML2 */ + ip = start0; m1 = m0; /* restore initial Match1 */ + } } + + /* Here, start0==ip */ + if ((start2 - ip) < 3) { /* First Match too small : removed */ + ip = start2; + m1 = m2; + goto _Search2; + } + +_Search3: + if ((start2 - ip) < OPTIMAL_ML) { + int correction; + int new_ml = m1.len; + if (new_ml > OPTIMAL_ML) new_ml = OPTIMAL_ML; + if (ip+new_ml > start2 + m2.len - MINMATCH) + new_ml = (int)(start2 - ip) + m2.len - MINMATCH; + correction = new_ml - (int)(start2 - ip); + if (correction > 0) { + start2 += correction; + m2.len -= correction; + } + } + + if (start2 + m2.len <= mflimit) { + start3 = start2 + m2.len - 3; + m3 = LZ4HC_InsertAndGetWiderMatch(ctx, + start3, start2, matchlimit, m2.len, + maxNbAttempts, patternAnalysis, 0, dict, favorCompressionRatio); + start3 += m3.back; + } else { + m3 = nomatch; /* do not search further */ + } + + if (m3.len <= m2.len) { /* No better match => encode ML1 and ML2 */ + /* ip & ref are known; Now for ml */ + if (start2 < ip+m1.len) m1.len = (int)(start2 - ip); + /* Now, encode 2 sequences */ + optr = op; + if (LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor), + m1.len, m1.off, + limit, oend) ) + goto _dest_overflow; + ip = start2; + optr = op; + if (LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor), + m2.len, m2.off, + limit, oend) ) { + m1 = m2; + goto _dest_overflow; + } + continue; + } + + if (start3 < ip+m1.len+3) { /* Not enough space for match 2 : remove it */ + if (start3 >= (ip+m1.len)) { /* can write Seq1 immediately ==> Seq2 is removed, so Seq3 becomes Seq1 */ + if (start2 < ip+m1.len) { + int correction = (int)(ip+m1.len - start2); + start2 += correction; + m2.len -= correction; + if (m2.len < MINMATCH) { + start2 = start3; + m2 = m3; + } + } + + optr = op; + if (LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor), + m1.len, m1.off, + limit, oend) ) + goto _dest_overflow; + ip = start3; + m1 = m3; + + start0 = start2; + m0 = m2; + goto _Search2; + } + + start2 = start3; + m2 = m3; + goto _Search3; + } + + /* + * OK, now we have 3 ascending matches; + * let's write the first one ML1. + * ip & ref are known; Now decide ml. + */ + if (start2 < ip+m1.len) { + if ((start2 - ip) < OPTIMAL_ML) { + int correction; + if (m1.len > OPTIMAL_ML) m1.len = OPTIMAL_ML; + if (ip + m1.len > start2 + m2.len - MINMATCH) + m1.len = (int)(start2 - ip) + m2.len - MINMATCH; + correction = m1.len - (int)(start2 - ip); + if (correction > 0) { + start2 += correction; + m2.len -= correction; + } + } else { + m1.len = (int)(start2 - ip); + } + } + optr = op; + if ( LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor), + m1.len, m1.off, + limit, oend) ) + goto _dest_overflow; + + /* ML2 becomes ML1 */ + ip = start2; m1 = m2; + + /* ML3 becomes ML2 */ + start2 = start3; m2 = m3; + + /* let's find a new ML3 */ + goto _Search3; + } + +_last_literals: + /* Encode Last Literals */ + { size_t lastRunSize = (size_t)(iend - anchor); /* literals */ + size_t llAdd = (lastRunSize + 255 - RUN_MASK) / 255; + size_t const totalSize = 1 + llAdd + lastRunSize; + if (limit == fillOutput) oend += LASTLITERALS; /* restore correct value */ + if (limit && (op + totalSize > oend)) { + if (limit == limitedOutput) return 0; + /* adapt lastRunSize to fill 'dest' */ + lastRunSize = (size_t)(oend - op) - 1 /*token*/; + llAdd = (lastRunSize + 256 - RUN_MASK) / 256; + lastRunSize -= llAdd; + } + DEBUGLOG(6, "Final literal run : %i literals", (int)lastRunSize); + ip = anchor + lastRunSize; /* can be != iend if limit==fillOutput */ + + if (lastRunSize >= RUN_MASK) { + size_t accumulator = lastRunSize - RUN_MASK; + *op++ = (RUN_MASK << ML_BITS); + for(; accumulator >= 255 ; accumulator -= 255) *op++ = 255; + *op++ = (BYTE) accumulator; + } else { + *op++ = (BYTE)(lastRunSize << ML_BITS); + } + LZ4_memcpy(op, anchor, lastRunSize); + op += lastRunSize; + } + + /* End */ + *srcSizePtr = (int) (((const char*)ip) - source); + return (int) (((char*)op)-dest); + +_dest_overflow: + if (limit == fillOutput) { + /* Assumption : @ip, @anchor, @optr and @m1 must be set correctly */ + size_t const ll = (size_t)(ip - anchor); + size_t const ll_addbytes = (ll + 240) / 255; + size_t const ll_totalCost = 1 + ll_addbytes + ll; + BYTE* const maxLitPos = oend - 3; /* 2 for offset, 1 for token */ + DEBUGLOG(6, "Last sequence overflowing"); + op = optr; /* restore correct out pointer */ + if (op + ll_totalCost <= maxLitPos) { + /* ll validated; now adjust match length */ + size_t const bytesLeftForMl = (size_t)(maxLitPos - (op+ll_totalCost)); + size_t const maxMlSize = MINMATCH + (ML_MASK-1) + (bytesLeftForMl * 255); + assert(maxMlSize < INT_MAX); assert(m1.len >= 0); + if ((size_t)m1.len > maxMlSize) m1.len = (int)maxMlSize; + if ((oend + LASTLITERALS) - (op + ll_totalCost + 2) - 1 + m1.len >= MFLIMIT) { + LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor), m1.len, m1.off, notLimited, oend); + } } + goto _last_literals; + } + /* compression failed */ + return 0; +} + + +static int LZ4HC_compress_optimal( LZ4HC_CCtx_internal* ctx, + const char* const source, char* dst, + int* srcSizePtr, int dstCapacity, + int const nbSearches, size_t sufficient_len, + const limitedOutput_directive limit, int const fullUpdate, + const dictCtx_directive dict, + const HCfavor_e favorDecSpeed); + + +LZ4_FORCE_INLINE int +LZ4HC_compress_generic_internal ( + LZ4HC_CCtx_internal* const ctx, + const char* const src, + char* const dst, + int* const srcSizePtr, + int const dstCapacity, + int cLevel, + const limitedOutput_directive limit, + const dictCtx_directive dict + ) +{ + typedef enum { lz4mid, lz4hc, lz4opt } lz4hc_strat_e; + typedef struct { + lz4hc_strat_e strat; + int nbSearches; + U32 targetLength; + } cParams_t; + static const cParams_t clTable[LZ4HC_CLEVEL_MAX+1] = { + { lz4mid, 2, 16 }, /* 0, unused */ + { lz4mid, 2, 16 }, /* 1, unused */ + { lz4mid, 2, 16 }, /* 2 */ + { lz4hc, 4, 16 }, /* 3 */ + { lz4hc, 8, 16 }, /* 4 */ + { lz4hc, 16, 16 }, /* 5 */ + { lz4hc, 32, 16 }, /* 6 */ + { lz4hc, 64, 16 }, /* 7 */ + { lz4hc, 128, 16 }, /* 8 */ + { lz4hc, 256, 16 }, /* 9 */ + { lz4opt, 96, 64 }, /*10==LZ4HC_CLEVEL_OPT_MIN*/ + { lz4opt, 512,128 }, /*11 */ + { lz4opt,16384,LZ4_OPT_NUM }, /* 12==LZ4HC_CLEVEL_MAX */ + }; + + DEBUGLOG(5, "LZ4HC_compress_generic_internal(src=%p, srcSize=%d)", + src, *srcSizePtr); + + if (limit == fillOutput && dstCapacity < 1) return 0; /* Impossible to store anything */ + if ((U32)*srcSizePtr > (U32)LZ4_MAX_INPUT_SIZE) return 0; /* Unsupported input size (too large or negative) */ + + ctx->end += *srcSizePtr; + /* note : clevel convention is a bit different from lz4frame, + * possibly something worth revisiting for consistency */ + if (cLevel < 1) + cLevel = LZ4HC_CLEVEL_DEFAULT; + cLevel = MIN(LZ4HC_CLEVEL_MAX, cLevel); + { cParams_t const cParam = clTable[cLevel]; + HCfavor_e const favor = ctx->favorDecSpeed ? favorDecompressionSpeed : favorCompressionRatio; + int result; + + if (cParam.strat == lz4mid) { + result = LZ4HC_compress_2hashes(ctx, + src, dst, srcSizePtr, dstCapacity, + limit, dict); + } else if (cParam.strat == lz4hc) { + result = LZ4HC_compress_hashChain(ctx, + src, dst, srcSizePtr, dstCapacity, + cParam.nbSearches, limit, dict); + } else { + assert(cParam.strat == lz4opt); + result = LZ4HC_compress_optimal(ctx, + src, dst, srcSizePtr, dstCapacity, + cParam.nbSearches, cParam.targetLength, limit, + cLevel == LZ4HC_CLEVEL_MAX, /* ultra mode */ + dict, favor); + } + if (result <= 0) ctx->dirty = 1; + return result; + } +} + +static void LZ4HC_setExternalDict(LZ4HC_CCtx_internal* ctxPtr, const BYTE* newBlock); + +static int +LZ4HC_compress_generic_noDictCtx ( + LZ4HC_CCtx_internal* const ctx, + const char* const src, + char* const dst, + int* const srcSizePtr, + int const dstCapacity, + int cLevel, + limitedOutput_directive limit + ) +{ + assert(ctx->dictCtx == NULL); + return LZ4HC_compress_generic_internal(ctx, src, dst, srcSizePtr, dstCapacity, cLevel, limit, noDictCtx); +} + +static int +LZ4HC_compress_generic_dictCtx ( + LZ4HC_CCtx_internal* const ctx, + const char* const src, + char* const dst, + int* const srcSizePtr, + int const dstCapacity, + int cLevel, + limitedOutput_directive limit + ) +{ + const size_t position = (size_t)(ctx->end - ctx->prefixStart) + (ctx->dictLimit - ctx->lowLimit); + assert(ctx->dictCtx != NULL); + if (position >= 64 KB) { + ctx->dictCtx = NULL; + return LZ4HC_compress_generic_noDictCtx(ctx, src, dst, srcSizePtr, dstCapacity, cLevel, limit); + } else if (position == 0 && *srcSizePtr > 4 KB) { + LZ4_memcpy(ctx, ctx->dictCtx, sizeof(LZ4HC_CCtx_internal)); + LZ4HC_setExternalDict(ctx, (const BYTE *)src); + ctx->compressionLevel = (short)cLevel; + return LZ4HC_compress_generic_noDictCtx(ctx, src, dst, srcSizePtr, dstCapacity, cLevel, limit); + } else { + return LZ4HC_compress_generic_internal(ctx, src, dst, srcSizePtr, dstCapacity, cLevel, limit, usingDictCtxHc); + } +} + +static int +LZ4HC_compress_generic ( + LZ4HC_CCtx_internal* const ctx, + const char* const src, + char* const dst, + int* const srcSizePtr, + int const dstCapacity, + int cLevel, + limitedOutput_directive limit + ) +{ + if (ctx->dictCtx == NULL) { + return LZ4HC_compress_generic_noDictCtx(ctx, src, dst, srcSizePtr, dstCapacity, cLevel, limit); + } else { + return LZ4HC_compress_generic_dictCtx(ctx, src, dst, srcSizePtr, dstCapacity, cLevel, limit); + } +} + + +int LZ4_sizeofStateHC(void) { return (int)sizeof(LZ4_streamHC_t); } + +static size_t LZ4_streamHC_t_alignment(void) +{ +#if LZ4_ALIGN_TEST + typedef struct { char c; LZ4_streamHC_t t; } t_a; + return sizeof(t_a) - sizeof(LZ4_streamHC_t); +#else + return 1; /* effectively disabled */ +#endif +} + +/* state is presumed correctly initialized, + * in which case its size and alignment have already been validate */ +int LZ4_compress_HC_extStateHC_fastReset (void* state, const char* src, char* dst, int srcSize, int dstCapacity, int compressionLevel) +{ + LZ4HC_CCtx_internal* const ctx = &((LZ4_streamHC_t*)state)->internal_donotuse; + if (!LZ4_isAligned(state, LZ4_streamHC_t_alignment())) return 0; + LZ4_resetStreamHC_fast((LZ4_streamHC_t*)state, compressionLevel); + LZ4HC_init_internal (ctx, (const BYTE*)src); + if (dstCapacity < LZ4_compressBound(srcSize)) + return LZ4HC_compress_generic (ctx, src, dst, &srcSize, dstCapacity, compressionLevel, limitedOutput); + else + return LZ4HC_compress_generic (ctx, src, dst, &srcSize, dstCapacity, compressionLevel, notLimited); +} + +int LZ4_compress_HC_extStateHC (void* state, const char* src, char* dst, int srcSize, int dstCapacity, int compressionLevel) +{ + LZ4_streamHC_t* const ctx = LZ4_initStreamHC(state, sizeof(*ctx)); + if (ctx==NULL) return 0; /* init failure */ + return LZ4_compress_HC_extStateHC_fastReset(state, src, dst, srcSize, dstCapacity, compressionLevel); +} + +int LZ4_compress_HC(const char* src, char* dst, int srcSize, int dstCapacity, int compressionLevel) +{ + int cSize; +#if defined(LZ4HC_HEAPMODE) && LZ4HC_HEAPMODE==1 + LZ4_streamHC_t* const statePtr = (LZ4_streamHC_t*)ALLOC(sizeof(LZ4_streamHC_t)); + if (statePtr==NULL) return 0; +#else + LZ4_streamHC_t state; + LZ4_streamHC_t* const statePtr = &state; +#endif + DEBUGLOG(5, "LZ4_compress_HC") + cSize = LZ4_compress_HC_extStateHC(statePtr, src, dst, srcSize, dstCapacity, compressionLevel); +#if defined(LZ4HC_HEAPMODE) && LZ4HC_HEAPMODE==1 + FREEMEM(statePtr); +#endif + return cSize; +} + +/* state is presumed sized correctly (>= sizeof(LZ4_streamHC_t)) */ +int LZ4_compress_HC_destSize(void* state, const char* source, char* dest, int* sourceSizePtr, int targetDestSize, int cLevel) +{ + LZ4_streamHC_t* const ctx = LZ4_initStreamHC(state, sizeof(*ctx)); + if (ctx==NULL) return 0; /* init failure */ + LZ4HC_init_internal(&ctx->internal_donotuse, (const BYTE*) source); + LZ4_setCompressionLevel(ctx, cLevel); + return LZ4HC_compress_generic(&ctx->internal_donotuse, source, dest, sourceSizePtr, targetDestSize, cLevel, fillOutput); +} + + + +/************************************** +* Streaming Functions +**************************************/ +/* allocation */ +#if !defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION) +LZ4_streamHC_t* LZ4_createStreamHC(void) +{ + LZ4_streamHC_t* const state = + (LZ4_streamHC_t*)ALLOC_AND_ZERO(sizeof(LZ4_streamHC_t)); + if (state == NULL) return NULL; + LZ4_setCompressionLevel(state, LZ4HC_CLEVEL_DEFAULT); + return state; +} + +int LZ4_freeStreamHC (LZ4_streamHC_t* LZ4_streamHCPtr) +{ + DEBUGLOG(4, "LZ4_freeStreamHC(%p)", LZ4_streamHCPtr); + if (!LZ4_streamHCPtr) return 0; /* support free on NULL */ + FREEMEM(LZ4_streamHCPtr); + return 0; +} +#endif + + +LZ4_streamHC_t* LZ4_initStreamHC (void* buffer, size_t size) +{ + LZ4_streamHC_t* const LZ4_streamHCPtr = (LZ4_streamHC_t*)buffer; + DEBUGLOG(4, "LZ4_initStreamHC(%p, %u)", buffer, (unsigned)size); + /* check conditions */ + if (buffer == NULL) return NULL; + if (size < sizeof(LZ4_streamHC_t)) return NULL; + if (!LZ4_isAligned(buffer, LZ4_streamHC_t_alignment())) return NULL; + /* init */ + { LZ4HC_CCtx_internal* const hcstate = &(LZ4_streamHCPtr->internal_donotuse); + MEM_INIT(hcstate, 0, sizeof(*hcstate)); } + LZ4_setCompressionLevel(LZ4_streamHCPtr, LZ4HC_CLEVEL_DEFAULT); + return LZ4_streamHCPtr; +} + +/* just a stub */ +void LZ4_resetStreamHC (LZ4_streamHC_t* LZ4_streamHCPtr, int compressionLevel) +{ + LZ4_initStreamHC(LZ4_streamHCPtr, sizeof(*LZ4_streamHCPtr)); + LZ4_setCompressionLevel(LZ4_streamHCPtr, compressionLevel); +} + +void LZ4_resetStreamHC_fast (LZ4_streamHC_t* LZ4_streamHCPtr, int compressionLevel) +{ + LZ4HC_CCtx_internal* const s = &LZ4_streamHCPtr->internal_donotuse; + DEBUGLOG(5, "LZ4_resetStreamHC_fast(%p, %d)", LZ4_streamHCPtr, compressionLevel); + if (s->dirty) { + LZ4_initStreamHC(LZ4_streamHCPtr, sizeof(*LZ4_streamHCPtr)); + } else { + assert(s->end >= s->prefixStart); + s->dictLimit += (U32)(s->end - s->prefixStart); + s->prefixStart = NULL; + s->end = NULL; + s->dictCtx = NULL; + } + LZ4_setCompressionLevel(LZ4_streamHCPtr, compressionLevel); +} + +void LZ4_setCompressionLevel(LZ4_streamHC_t* LZ4_streamHCPtr, int compressionLevel) +{ + DEBUGLOG(5, "LZ4_setCompressionLevel(%p, %d)", LZ4_streamHCPtr, compressionLevel); + if (compressionLevel < 1) compressionLevel = LZ4HC_CLEVEL_DEFAULT; + if (compressionLevel > LZ4HC_CLEVEL_MAX) compressionLevel = LZ4HC_CLEVEL_MAX; + LZ4_streamHCPtr->internal_donotuse.compressionLevel = (short)compressionLevel; +} + +void LZ4_favorDecompressionSpeed(LZ4_streamHC_t* LZ4_streamHCPtr, int favor) +{ + LZ4_streamHCPtr->internal_donotuse.favorDecSpeed = (favor!=0); +} + +/* LZ4_loadDictHC() : + * LZ4_streamHCPtr is presumed properly initialized */ +int LZ4_loadDictHC (LZ4_streamHC_t* LZ4_streamHCPtr, + const char* dictionary, int dictSize) +{ + LZ4HC_CCtx_internal* const ctxPtr = &LZ4_streamHCPtr->internal_donotuse; + DEBUGLOG(4, "LZ4_loadDictHC(ctx:%p, dict:%p, dictSize:%d)", LZ4_streamHCPtr, dictionary, dictSize); + assert(LZ4_streamHCPtr != NULL); + if (dictSize > 64 KB) { + dictionary += (size_t)dictSize - 64 KB; + dictSize = 64 KB; + } + /* need a full initialization, there are bad side-effects when using resetFast() */ + { int const cLevel = ctxPtr->compressionLevel; + LZ4_initStreamHC(LZ4_streamHCPtr, sizeof(*LZ4_streamHCPtr)); + LZ4_setCompressionLevel(LZ4_streamHCPtr, cLevel); + } + LZ4HC_init_internal (ctxPtr, (const BYTE*)dictionary); + ctxPtr->end = (const BYTE*)dictionary + dictSize; + if (dictSize >= LZ4HC_HASHSIZE) LZ4HC_Insert (ctxPtr, ctxPtr->end-3); + return dictSize; +} + +void LZ4_attach_HC_dictionary(LZ4_streamHC_t *working_stream, const LZ4_streamHC_t *dictionary_stream) { + working_stream->internal_donotuse.dictCtx = dictionary_stream != NULL ? &(dictionary_stream->internal_donotuse) : NULL; +} + +/* compression */ + +static void LZ4HC_setExternalDict(LZ4HC_CCtx_internal* ctxPtr, const BYTE* newBlock) +{ + DEBUGLOG(4, "LZ4HC_setExternalDict(%p, %p)", ctxPtr, newBlock); + if (ctxPtr->end >= ctxPtr->prefixStart + 4) + LZ4HC_Insert (ctxPtr, ctxPtr->end-3); /* Referencing remaining dictionary content */ + + /* Only one memory segment for extDict, so any previous extDict is lost at this stage */ + ctxPtr->lowLimit = ctxPtr->dictLimit; + ctxPtr->dictStart = ctxPtr->prefixStart; + ctxPtr->dictLimit += (U32)(ctxPtr->end - ctxPtr->prefixStart); + ctxPtr->prefixStart = newBlock; + ctxPtr->end = newBlock; + ctxPtr->nextToUpdate = ctxPtr->dictLimit; /* match referencing will resume from there */ + + /* cannot reference an extDict and a dictCtx at the same time */ + ctxPtr->dictCtx = NULL; +} + +static int +LZ4_compressHC_continue_generic (LZ4_streamHC_t* LZ4_streamHCPtr, + const char* src, char* dst, + int* srcSizePtr, int dstCapacity, + limitedOutput_directive limit) +{ + LZ4HC_CCtx_internal* const ctxPtr = &LZ4_streamHCPtr->internal_donotuse; + DEBUGLOG(5, "LZ4_compressHC_continue_generic(ctx=%p, src=%p, srcSize=%d, limit=%d)", + LZ4_streamHCPtr, src, *srcSizePtr, limit); + assert(ctxPtr != NULL); + /* auto-init if forgotten */ + if (ctxPtr->prefixStart == NULL) LZ4HC_init_internal (ctxPtr, (const BYTE*) src); + + /* Check overflow */ + if ((size_t)(ctxPtr->end - ctxPtr->prefixStart) + ctxPtr->dictLimit > 2 GB) { + size_t dictSize = (size_t)(ctxPtr->end - ctxPtr->prefixStart); + if (dictSize > 64 KB) dictSize = 64 KB; + LZ4_loadDictHC(LZ4_streamHCPtr, (const char*)(ctxPtr->end) - dictSize, (int)dictSize); + } + + /* Check if blocks follow each other */ + if ((const BYTE*)src != ctxPtr->end) + LZ4HC_setExternalDict(ctxPtr, (const BYTE*)src); + + /* Check overlapping input/dictionary space */ + { const BYTE* sourceEnd = (const BYTE*) src + *srcSizePtr; + const BYTE* const dictBegin = ctxPtr->dictStart; + const BYTE* const dictEnd = ctxPtr->dictStart + (ctxPtr->dictLimit - ctxPtr->lowLimit); + if ((sourceEnd > dictBegin) && ((const BYTE*)src < dictEnd)) { + if (sourceEnd > dictEnd) sourceEnd = dictEnd; + ctxPtr->lowLimit += (U32)(sourceEnd - ctxPtr->dictStart); + ctxPtr->dictStart += (U32)(sourceEnd - ctxPtr->dictStart); + /* invalidate dictionary is it's too small */ + if (ctxPtr->dictLimit - ctxPtr->lowLimit < LZ4HC_HASHSIZE) { + ctxPtr->lowLimit = ctxPtr->dictLimit; + ctxPtr->dictStart = ctxPtr->prefixStart; + } } } + + return LZ4HC_compress_generic (ctxPtr, src, dst, srcSizePtr, dstCapacity, ctxPtr->compressionLevel, limit); +} + +int LZ4_compress_HC_continue (LZ4_streamHC_t* LZ4_streamHCPtr, const char* src, char* dst, int srcSize, int dstCapacity) +{ + DEBUGLOG(5, "LZ4_compress_HC_continue"); + if (dstCapacity < LZ4_compressBound(srcSize)) + return LZ4_compressHC_continue_generic (LZ4_streamHCPtr, src, dst, &srcSize, dstCapacity, limitedOutput); + else + return LZ4_compressHC_continue_generic (LZ4_streamHCPtr, src, dst, &srcSize, dstCapacity, notLimited); +} + +int LZ4_compress_HC_continue_destSize (LZ4_streamHC_t* LZ4_streamHCPtr, const char* src, char* dst, int* srcSizePtr, int targetDestSize) +{ + return LZ4_compressHC_continue_generic(LZ4_streamHCPtr, src, dst, srcSizePtr, targetDestSize, fillOutput); +} + + + +/* LZ4_saveDictHC : + * save history content + * into a user-provided buffer + * which is then used to continue compression + */ +int LZ4_saveDictHC (LZ4_streamHC_t* LZ4_streamHCPtr, char* safeBuffer, int dictSize) +{ + LZ4HC_CCtx_internal* const streamPtr = &LZ4_streamHCPtr->internal_donotuse; + int const prefixSize = (int)(streamPtr->end - streamPtr->prefixStart); + DEBUGLOG(5, "LZ4_saveDictHC(%p, %p, %d)", LZ4_streamHCPtr, safeBuffer, dictSize); + assert(prefixSize >= 0); + if (dictSize > 64 KB) dictSize = 64 KB; + if (dictSize < 4) dictSize = 0; + if (dictSize > prefixSize) dictSize = prefixSize; + if (safeBuffer == NULL) assert(dictSize == 0); + if (dictSize > 0) + LZ4_memmove(safeBuffer, streamPtr->end - dictSize, (size_t)dictSize); + { U32 const endIndex = (U32)(streamPtr->end - streamPtr->prefixStart) + streamPtr->dictLimit; + streamPtr->end = (safeBuffer == NULL) ? NULL : (const BYTE*)safeBuffer + dictSize; + streamPtr->prefixStart = (const BYTE*)safeBuffer; + streamPtr->dictLimit = endIndex - (U32)dictSize; + streamPtr->lowLimit = endIndex - (U32)dictSize; + streamPtr->dictStart = streamPtr->prefixStart; + if (streamPtr->nextToUpdate < streamPtr->dictLimit) + streamPtr->nextToUpdate = streamPtr->dictLimit; + } + return dictSize; +} + + +/*************************************************** +* Deprecated Functions +***************************************************/ + +/* These functions currently generate deprecation warnings */ + +/* Wrappers for deprecated compression functions */ +int LZ4_compressHC(const char* src, char* dst, int srcSize) { return LZ4_compress_HC (src, dst, srcSize, LZ4_compressBound(srcSize), 0); } +int LZ4_compressHC_limitedOutput(const char* src, char* dst, int srcSize, int maxDstSize) { return LZ4_compress_HC(src, dst, srcSize, maxDstSize, 0); } +int LZ4_compressHC2(const char* src, char* dst, int srcSize, int cLevel) { return LZ4_compress_HC (src, dst, srcSize, LZ4_compressBound(srcSize), cLevel); } +int LZ4_compressHC2_limitedOutput(const char* src, char* dst, int srcSize, int maxDstSize, int cLevel) { return LZ4_compress_HC(src, dst, srcSize, maxDstSize, cLevel); } +int LZ4_compressHC_withStateHC (void* state, const char* src, char* dst, int srcSize) { return LZ4_compress_HC_extStateHC (state, src, dst, srcSize, LZ4_compressBound(srcSize), 0); } +int LZ4_compressHC_limitedOutput_withStateHC (void* state, const char* src, char* dst, int srcSize, int maxDstSize) { return LZ4_compress_HC_extStateHC (state, src, dst, srcSize, maxDstSize, 0); } +int LZ4_compressHC2_withStateHC (void* state, const char* src, char* dst, int srcSize, int cLevel) { return LZ4_compress_HC_extStateHC(state, src, dst, srcSize, LZ4_compressBound(srcSize), cLevel); } +int LZ4_compressHC2_limitedOutput_withStateHC (void* state, const char* src, char* dst, int srcSize, int maxDstSize, int cLevel) { return LZ4_compress_HC_extStateHC(state, src, dst, srcSize, maxDstSize, cLevel); } +int LZ4_compressHC_continue (LZ4_streamHC_t* ctx, const char* src, char* dst, int srcSize) { return LZ4_compress_HC_continue (ctx, src, dst, srcSize, LZ4_compressBound(srcSize)); } +int LZ4_compressHC_limitedOutput_continue (LZ4_streamHC_t* ctx, const char* src, char* dst, int srcSize, int maxDstSize) { return LZ4_compress_HC_continue (ctx, src, dst, srcSize, maxDstSize); } + + +/* Deprecated streaming functions */ +int LZ4_sizeofStreamStateHC(void) { return sizeof(LZ4_streamHC_t); } + +/* state is presumed correctly sized, aka >= sizeof(LZ4_streamHC_t) + * @return : 0 on success, !=0 if error */ +int LZ4_resetStreamStateHC(void* state, char* inputBuffer) +{ + LZ4_streamHC_t* const hc4 = LZ4_initStreamHC(state, sizeof(*hc4)); + if (hc4 == NULL) return 1; /* init failed */ + LZ4HC_init_internal (&hc4->internal_donotuse, (const BYTE*)inputBuffer); + return 0; +} + +#if !defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION) +void* LZ4_createHC (const char* inputBuffer) +{ + LZ4_streamHC_t* const hc4 = LZ4_createStreamHC(); + if (hc4 == NULL) return NULL; /* not enough memory */ + LZ4HC_init_internal (&hc4->internal_donotuse, (const BYTE*)inputBuffer); + return hc4; +} + +int LZ4_freeHC (void* LZ4HC_Data) +{ + if (!LZ4HC_Data) return 0; /* support free on NULL */ + FREEMEM(LZ4HC_Data); + return 0; +} +#endif + +int LZ4_compressHC2_continue (void* LZ4HC_Data, const char* src, char* dst, int srcSize, int cLevel) +{ + return LZ4HC_compress_generic (&((LZ4_streamHC_t*)LZ4HC_Data)->internal_donotuse, src, dst, &srcSize, 0, cLevel, notLimited); +} + +int LZ4_compressHC2_limitedOutput_continue (void* LZ4HC_Data, const char* src, char* dst, int srcSize, int dstCapacity, int cLevel) +{ + return LZ4HC_compress_generic (&((LZ4_streamHC_t*)LZ4HC_Data)->internal_donotuse, src, dst, &srcSize, dstCapacity, cLevel, limitedOutput); +} + +char* LZ4_slideInputBufferHC(void* LZ4HC_Data) +{ + LZ4HC_CCtx_internal* const s = &((LZ4_streamHC_t*)LZ4HC_Data)->internal_donotuse; + const BYTE* const bufferStart = s->prefixStart - s->dictLimit + s->lowLimit; + LZ4_resetStreamHC_fast((LZ4_streamHC_t*)LZ4HC_Data, s->compressionLevel); + /* ugly conversion trick, required to evade (const char*) -> (char*) cast-qual warning :( */ + return (char*)(uptrval)bufferStart; +} + + +/* ================================================ + * LZ4 Optimal parser (levels [LZ4HC_CLEVEL_OPT_MIN - LZ4HC_CLEVEL_MAX]) + * ===============================================*/ +typedef struct { + int price; + int off; + int mlen; + int litlen; +} LZ4HC_optimal_t; + +/* price in bytes */ +LZ4_FORCE_INLINE int LZ4HC_literalsPrice(int const litlen) +{ + int price = litlen; + assert(litlen >= 0); + if (litlen >= (int)RUN_MASK) + price += 1 + ((litlen-(int)RUN_MASK) / 255); + return price; +} + + +/* requires mlen >= MINMATCH */ +LZ4_FORCE_INLINE int LZ4HC_sequencePrice(int litlen, int mlen) +{ + int price = 1 + 2 ; /* token + 16-bit offset */ + assert(litlen >= 0); + assert(mlen >= MINMATCH); + + price += LZ4HC_literalsPrice(litlen); + + if (mlen >= (int)(ML_MASK+MINMATCH)) + price += 1 + ((mlen-(int)(ML_MASK+MINMATCH)) / 255); + + return price; +} + + + +LZ4_FORCE_INLINE LZ4HC_match_t +LZ4HC_FindLongerMatch(LZ4HC_CCtx_internal* const ctx, + const BYTE* ip, const BYTE* const iHighLimit, + int minLen, int nbSearches, + const dictCtx_directive dict, + const HCfavor_e favorDecSpeed) +{ + LZ4HC_match_t const match0 = { 0 , 0, 0 }; + /* note : LZ4HC_InsertAndGetWiderMatch() is able to modify the starting position of a match (*startpos), + * but this won't be the case here, as we define iLowLimit==ip, + ** so LZ4HC_InsertAndGetWiderMatch() won't be allowed to search past ip */ + LZ4HC_match_t md = LZ4HC_InsertAndGetWiderMatch(ctx, ip, ip, iHighLimit, minLen, nbSearches, 1 /*patternAnalysis*/, 1 /*chainSwap*/, dict, favorDecSpeed); + assert(md.back == 0); + if (md.len <= minLen) return match0; + if (favorDecSpeed) { + if ((md.len>18) & (md.len<=36)) md.len=18; /* favor dec.speed (shortcut) */ + } + return md; +} + + +static int LZ4HC_compress_optimal ( LZ4HC_CCtx_internal* ctx, + const char* const source, + char* dst, + int* srcSizePtr, + int dstCapacity, + int const nbSearches, + size_t sufficient_len, + const limitedOutput_directive limit, + int const fullUpdate, + const dictCtx_directive dict, + const HCfavor_e favorDecSpeed) +{ + int retval = 0; +#define TRAILING_LITERALS 3 +#if defined(LZ4HC_HEAPMODE) && LZ4HC_HEAPMODE==1 + LZ4HC_optimal_t* const opt = (LZ4HC_optimal_t*)ALLOC(sizeof(LZ4HC_optimal_t) * (LZ4_OPT_NUM + TRAILING_LITERALS)); +#else + LZ4HC_optimal_t opt[LZ4_OPT_NUM + TRAILING_LITERALS]; /* ~64 KB, which is a bit large for stack... */ +#endif + + const BYTE* ip = (const BYTE*) source; + const BYTE* anchor = ip; + const BYTE* const iend = ip + *srcSizePtr; + const BYTE* const mflimit = iend - MFLIMIT; + const BYTE* const matchlimit = iend - LASTLITERALS; + BYTE* op = (BYTE*) dst; + BYTE* opSaved = (BYTE*) dst; + BYTE* oend = op + dstCapacity; + int ovml = MINMATCH; /* overflow - last sequence */ + int ovoff = 0; + + /* init */ +#if defined(LZ4HC_HEAPMODE) && LZ4HC_HEAPMODE==1 + if (opt == NULL) goto _return_label; +#endif + DEBUGLOG(5, "LZ4HC_compress_optimal(dst=%p, dstCapa=%u)", dst, (unsigned)dstCapacity); + *srcSizePtr = 0; + if (limit == fillOutput) oend -= LASTLITERALS; /* Hack for support LZ4 format restriction */ + if (sufficient_len >= LZ4_OPT_NUM) sufficient_len = LZ4_OPT_NUM-1; + + /* Main Loop */ + while (ip <= mflimit) { + int const llen = (int)(ip - anchor); + int best_mlen, best_off; + int cur, last_match_pos = 0; + + LZ4HC_match_t const firstMatch = LZ4HC_FindLongerMatch(ctx, ip, matchlimit, MINMATCH-1, nbSearches, dict, favorDecSpeed); + if (firstMatch.len==0) { ip++; continue; } + + if ((size_t)firstMatch.len > sufficient_len) { + /* good enough solution : immediate encoding */ + int const firstML = firstMatch.len; + opSaved = op; + if ( LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor), firstML, firstMatch.off, limit, oend) ) { /* updates ip, op and anchor */ + ovml = firstML; + ovoff = firstMatch.off; + goto _dest_overflow; + } + continue; + } + + /* set prices for first positions (literals) */ + { int rPos; + for (rPos = 0 ; rPos < MINMATCH ; rPos++) { + int const cost = LZ4HC_literalsPrice(llen + rPos); + opt[rPos].mlen = 1; + opt[rPos].off = 0; + opt[rPos].litlen = llen + rPos; + opt[rPos].price = cost; + DEBUGLOG(7, "rPos:%3i => price:%3i (litlen=%i) -- initial setup", + rPos, cost, opt[rPos].litlen); + } } + /* set prices using initial match */ + { int mlen = MINMATCH; + int const matchML = firstMatch.len; /* necessarily < sufficient_len < LZ4_OPT_NUM */ + int const offset = firstMatch.off; + assert(matchML < LZ4_OPT_NUM); + for ( ; mlen <= matchML ; mlen++) { + int const cost = LZ4HC_sequencePrice(llen, mlen); + opt[mlen].mlen = mlen; + opt[mlen].off = offset; + opt[mlen].litlen = llen; + opt[mlen].price = cost; + DEBUGLOG(7, "rPos:%3i => price:%3i (matchlen=%i) -- initial setup", + mlen, cost, mlen); + } } + last_match_pos = firstMatch.len; + { int addLit; + for (addLit = 1; addLit <= TRAILING_LITERALS; addLit ++) { + opt[last_match_pos+addLit].mlen = 1; /* literal */ + opt[last_match_pos+addLit].off = 0; + opt[last_match_pos+addLit].litlen = addLit; + opt[last_match_pos+addLit].price = opt[last_match_pos].price + LZ4HC_literalsPrice(addLit); + DEBUGLOG(7, "rPos:%3i => price:%3i (litlen=%i) -- initial setup", + last_match_pos+addLit, opt[last_match_pos+addLit].price, addLit); + } } + + /* check further positions */ + for (cur = 1; cur < last_match_pos; cur++) { + const BYTE* const curPtr = ip + cur; + LZ4HC_match_t newMatch; + + if (curPtr > mflimit) break; + DEBUGLOG(7, "rPos:%u[%u] vs [%u]%u", + cur, opt[cur].price, opt[cur+1].price, cur+1); + if (fullUpdate) { + /* not useful to search here if next position has same (or lower) cost */ + if ( (opt[cur+1].price <= opt[cur].price) + /* in some cases, next position has same cost, but cost rises sharply after, so a small match would still be beneficial */ + && (opt[cur+MINMATCH].price < opt[cur].price + 3/*min seq price*/) ) + continue; + } else { + /* not useful to search here if next position has same (or lower) cost */ + if (opt[cur+1].price <= opt[cur].price) continue; + } + + DEBUGLOG(7, "search at rPos:%u", cur); + if (fullUpdate) + newMatch = LZ4HC_FindLongerMatch(ctx, curPtr, matchlimit, MINMATCH-1, nbSearches, dict, favorDecSpeed); + else + /* only test matches of minimum length; slightly faster, but misses a few bytes */ + newMatch = LZ4HC_FindLongerMatch(ctx, curPtr, matchlimit, last_match_pos - cur, nbSearches, dict, favorDecSpeed); + if (!newMatch.len) continue; + + if ( ((size_t)newMatch.len > sufficient_len) + || (newMatch.len + cur >= LZ4_OPT_NUM) ) { + /* immediate encoding */ + best_mlen = newMatch.len; + best_off = newMatch.off; + last_match_pos = cur + 1; + goto encode; + } + + /* before match : set price with literals at beginning */ + { int const baseLitlen = opt[cur].litlen; + int litlen; + for (litlen = 1; litlen < MINMATCH; litlen++) { + int const price = opt[cur].price - LZ4HC_literalsPrice(baseLitlen) + LZ4HC_literalsPrice(baseLitlen+litlen); + int const pos = cur + litlen; + if (price < opt[pos].price) { + opt[pos].mlen = 1; /* literal */ + opt[pos].off = 0; + opt[pos].litlen = baseLitlen+litlen; + opt[pos].price = price; + DEBUGLOG(7, "rPos:%3i => price:%3i (litlen=%i)", + pos, price, opt[pos].litlen); + } } } + + /* set prices using match at position = cur */ + { int const matchML = newMatch.len; + int ml = MINMATCH; + + assert(cur + newMatch.len < LZ4_OPT_NUM); + for ( ; ml <= matchML ; ml++) { + int const pos = cur + ml; + int const offset = newMatch.off; + int price; + int ll; + DEBUGLOG(7, "testing price rPos %i (last_match_pos=%i)", + pos, last_match_pos); + if (opt[cur].mlen == 1) { + ll = opt[cur].litlen; + price = ((cur > ll) ? opt[cur - ll].price : 0) + + LZ4HC_sequencePrice(ll, ml); + } else { + ll = 0; + price = opt[cur].price + LZ4HC_sequencePrice(0, ml); + } + + assert((U32)favorDecSpeed <= 1); + if (pos > last_match_pos+TRAILING_LITERALS + || price <= opt[pos].price - (int)favorDecSpeed) { + DEBUGLOG(7, "rPos:%3i => price:%3i (matchlen=%i)", + pos, price, ml); + assert(pos < LZ4_OPT_NUM); + if ( (ml == matchML) /* last pos of last match */ + && (last_match_pos < pos) ) + last_match_pos = pos; + opt[pos].mlen = ml; + opt[pos].off = offset; + opt[pos].litlen = ll; + opt[pos].price = price; + } } } + /* complete following positions with literals */ + { int addLit; + for (addLit = 1; addLit <= TRAILING_LITERALS; addLit ++) { + opt[last_match_pos+addLit].mlen = 1; /* literal */ + opt[last_match_pos+addLit].off = 0; + opt[last_match_pos+addLit].litlen = addLit; + opt[last_match_pos+addLit].price = opt[last_match_pos].price + LZ4HC_literalsPrice(addLit); + DEBUGLOG(7, "rPos:%3i => price:%3i (litlen=%i)", last_match_pos+addLit, opt[last_match_pos+addLit].price, addLit); + } } + } /* for (cur = 1; cur <= last_match_pos; cur++) */ + + assert(last_match_pos < LZ4_OPT_NUM + TRAILING_LITERALS); + best_mlen = opt[last_match_pos].mlen; + best_off = opt[last_match_pos].off; + cur = last_match_pos - best_mlen; + +encode: /* cur, last_match_pos, best_mlen, best_off must be set */ + assert(cur < LZ4_OPT_NUM); + assert(last_match_pos >= 1); /* == 1 when only one candidate */ + DEBUGLOG(6, "reverse traversal, looking for shortest path (last_match_pos=%i)", last_match_pos); + { int candidate_pos = cur; + int selected_matchLength = best_mlen; + int selected_offset = best_off; + while (1) { /* from end to beginning */ + int const next_matchLength = opt[candidate_pos].mlen; /* can be 1, means literal */ + int const next_offset = opt[candidate_pos].off; + DEBUGLOG(7, "pos %i: sequence length %i", candidate_pos, selected_matchLength); + opt[candidate_pos].mlen = selected_matchLength; + opt[candidate_pos].off = selected_offset; + selected_matchLength = next_matchLength; + selected_offset = next_offset; + if (next_matchLength > candidate_pos) break; /* last match elected, first match to encode */ + assert(next_matchLength > 0); /* can be 1, means literal */ + candidate_pos -= next_matchLength; + } } + + /* encode all recorded sequences in order */ + { int rPos = 0; /* relative position (to ip) */ + while (rPos < last_match_pos) { + int const ml = opt[rPos].mlen; + int const offset = opt[rPos].off; + if (ml == 1) { ip++; rPos++; continue; } /* literal; note: can end up with several literals, in which case, skip them */ + rPos += ml; + assert(ml >= MINMATCH); + assert((offset >= 1) && (offset <= LZ4_DISTANCE_MAX)); + opSaved = op; + if ( LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor), ml, offset, limit, oend) ) { /* updates ip, op and anchor */ + ovml = ml; + ovoff = offset; + goto _dest_overflow; + } } } + } /* while (ip <= mflimit) */ + +_last_literals: + /* Encode Last Literals */ + { size_t lastRunSize = (size_t)(iend - anchor); /* literals */ + size_t llAdd = (lastRunSize + 255 - RUN_MASK) / 255; + size_t const totalSize = 1 + llAdd + lastRunSize; + if (limit == fillOutput) oend += LASTLITERALS; /* restore correct value */ + if (limit && (op + totalSize > oend)) { + if (limit == limitedOutput) { /* Check output limit */ + retval = 0; + goto _return_label; + } + /* adapt lastRunSize to fill 'dst' */ + lastRunSize = (size_t)(oend - op) - 1 /*token*/; + llAdd = (lastRunSize + 256 - RUN_MASK) / 256; + lastRunSize -= llAdd; + } + DEBUGLOG(6, "Final literal run : %i literals", (int)lastRunSize); + ip = anchor + lastRunSize; /* can be != iend if limit==fillOutput */ + + if (lastRunSize >= RUN_MASK) { + size_t accumulator = lastRunSize - RUN_MASK; + *op++ = (RUN_MASK << ML_BITS); + for(; accumulator >= 255 ; accumulator -= 255) *op++ = 255; + *op++ = (BYTE) accumulator; + } else { + *op++ = (BYTE)(lastRunSize << ML_BITS); + } + LZ4_memcpy(op, anchor, lastRunSize); + op += lastRunSize; + } + + /* End */ + *srcSizePtr = (int) (((const char*)ip) - source); + retval = (int) ((char*)op-dst); + goto _return_label; + +_dest_overflow: +if (limit == fillOutput) { + /* Assumption : ip, anchor, ovml and ovref must be set correctly */ + size_t const ll = (size_t)(ip - anchor); + size_t const ll_addbytes = (ll + 240) / 255; + size_t const ll_totalCost = 1 + ll_addbytes + ll; + BYTE* const maxLitPos = oend - 3; /* 2 for offset, 1 for token */ + DEBUGLOG(6, "Last sequence overflowing (only %i bytes remaining)", (int)(oend-1-opSaved)); + op = opSaved; /* restore correct out pointer */ + if (op + ll_totalCost <= maxLitPos) { + /* ll validated; now adjust match length */ + size_t const bytesLeftForMl = (size_t)(maxLitPos - (op+ll_totalCost)); + size_t const maxMlSize = MINMATCH + (ML_MASK-1) + (bytesLeftForMl * 255); + assert(maxMlSize < INT_MAX); assert(ovml >= 0); + if ((size_t)ovml > maxMlSize) ovml = (int)maxMlSize; + if ((oend + LASTLITERALS) - (op + ll_totalCost + 2) - 1 + ovml >= MFLIMIT) { + DEBUGLOG(6, "Space to end : %i + ml (%i)", (int)((oend + LASTLITERALS) - (op + ll_totalCost + 2) - 1), ovml); + DEBUGLOG(6, "Before : ip = %p, anchor = %p", ip, anchor); + LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor), ovml, ovoff, notLimited, oend); + DEBUGLOG(6, "After : ip = %p, anchor = %p", ip, anchor); + } } + goto _last_literals; +} +_return_label: +#if defined(LZ4HC_HEAPMODE) && LZ4HC_HEAPMODE==1 + if (opt) FREEMEM(opt); +#endif + return retval; +} diff --git a/example/android/third_party/lz4/include/lz4hc.h b/example/android/third_party/lz4/include/lz4hc.h new file mode 100644 index 00000000..f23db5f6 --- /dev/null +++ b/example/android/third_party/lz4/include/lz4hc.h @@ -0,0 +1,413 @@ +/* + LZ4 HC - High Compression Mode of LZ4 + Header File + Copyright (C) 2011-2020, Yann Collet. + BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) + + Redistribution and use in source and binary forms, with or without + modification, are permitted provided that the following conditions are + met: + + * Redistributions of source code must retain the above copyright + notice, this list of conditions and the following disclaimer. + * Redistributions in binary form must reproduce the above + copyright notice, this list of conditions and the following disclaimer + in the documentation and/or other materials provided with the + distribution. + + THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + + You can contact the author at : + - LZ4 source repository : https://github.com/lz4/lz4 + - LZ4 public forum : https://groups.google.com/forum/#!forum/lz4c +*/ +#ifndef LZ4_HC_H_19834876238432 +#define LZ4_HC_H_19834876238432 + +#if defined (__cplusplus) +extern "C" { +#endif + +/* --- Dependency --- */ +/* note : lz4hc requires lz4.h/lz4.c for compilation */ +#include "lz4.h" /* stddef, LZ4LIB_API, LZ4_DEPRECATED */ + + +/* --- Useful constants --- */ +#define LZ4HC_CLEVEL_MIN 2 +#define LZ4HC_CLEVEL_DEFAULT 9 +#define LZ4HC_CLEVEL_OPT_MIN 10 +#define LZ4HC_CLEVEL_MAX 12 + + +/*-************************************ + * Block Compression + **************************************/ +/*! LZ4_compress_HC() : + * Compress data from `src` into `dst`, using the powerful but slower "HC" algorithm. + * `dst` must be already allocated. + * Compression is guaranteed to succeed if `dstCapacity >= LZ4_compressBound(srcSize)` (see "lz4.h") + * Max supported `srcSize` value is LZ4_MAX_INPUT_SIZE (see "lz4.h") + * `compressionLevel` : any value between 1 and LZ4HC_CLEVEL_MAX will work. + * Values > LZ4HC_CLEVEL_MAX behave the same as LZ4HC_CLEVEL_MAX. + * @return : the number of bytes written into 'dst' + * or 0 if compression fails. + */ +LZ4LIB_API int LZ4_compress_HC (const char* src, char* dst, int srcSize, int dstCapacity, int compressionLevel); + + +/* Note : + * Decompression functions are provided within "lz4.h" (BSD license) + */ + + +/*! LZ4_compress_HC_extStateHC() : + * Same as LZ4_compress_HC(), but using an externally allocated memory segment for `state`. + * `state` size is provided by LZ4_sizeofStateHC(). + * Memory segment must be aligned on 8-bytes boundaries (which a normal malloc() should do properly). + */ +LZ4LIB_API int LZ4_sizeofStateHC(void); +LZ4LIB_API int LZ4_compress_HC_extStateHC(void* stateHC, const char* src, char* dst, int srcSize, int maxDstSize, int compressionLevel); + + +/*! LZ4_compress_HC_destSize() : v1.9.0+ + * Will compress as much data as possible from `src` + * to fit into `targetDstSize` budget. + * Result is provided in 2 parts : + * @return : the number of bytes written into 'dst' (necessarily <= targetDstSize) + * or 0 if compression fails. + * `srcSizePtr` : on success, *srcSizePtr is updated to indicate how much bytes were read from `src` + */ +LZ4LIB_API int LZ4_compress_HC_destSize(void* stateHC, + const char* src, char* dst, + int* srcSizePtr, int targetDstSize, + int compressionLevel); + + +/*-************************************ + * Streaming Compression + * Bufferless synchronous API + **************************************/ + typedef union LZ4_streamHC_u LZ4_streamHC_t; /* incomplete type (defined later) */ + +/*! LZ4_createStreamHC() and LZ4_freeStreamHC() : + * These functions create and release memory for LZ4 HC streaming state. + * Newly created states are automatically initialized. + * A same state can be used multiple times consecutively, + * starting with LZ4_resetStreamHC_fast() to start a new stream of blocks. + */ +LZ4LIB_API LZ4_streamHC_t* LZ4_createStreamHC(void); +LZ4LIB_API int LZ4_freeStreamHC (LZ4_streamHC_t* streamHCPtr); + +/* + These functions compress data in successive blocks of any size, + using previous blocks as dictionary, to improve compression ratio. + One key assumption is that previous blocks (up to 64 KB) remain read-accessible while compressing next blocks. + There is an exception for ring buffers, which can be smaller than 64 KB. + Ring-buffer scenario is automatically detected and handled within LZ4_compress_HC_continue(). + + Before starting compression, state must be allocated and properly initialized. + LZ4_createStreamHC() does both, though compression level is set to LZ4HC_CLEVEL_DEFAULT. + + Selecting the compression level can be done with LZ4_resetStreamHC_fast() (starts a new stream) + or LZ4_setCompressionLevel() (anytime, between blocks in the same stream) (experimental). + LZ4_resetStreamHC_fast() only works on states which have been properly initialized at least once, + which is automatically the case when state is created using LZ4_createStreamHC(). + + After reset, a first "fictional block" can be designated as initial dictionary, + using LZ4_loadDictHC() (Optional). + + Invoke LZ4_compress_HC_continue() to compress each successive block. + The number of blocks is unlimited. + Previous input blocks, including initial dictionary when present, + must remain accessible and unmodified during compression. + + It's allowed to update compression level anytime between blocks, + using LZ4_setCompressionLevel() (experimental). + + 'dst' buffer should be sized to handle worst case scenarios + (see LZ4_compressBound(), it ensures compression success). + In case of failure, the API does not guarantee recovery, + so the state _must_ be reset. + To ensure compression success + whenever `dst` buffer size cannot be made >= LZ4_compressBound(), + consider using LZ4_compress_HC_continue_destSize(). + + Whenever previous input blocks can't be preserved unmodified in-place during compression of next blocks, + it's possible to copy the last blocks into a more stable memory space, using LZ4_saveDictHC(). + Return value of LZ4_saveDictHC() is the size of dictionary effectively saved into 'safeBuffer' (<= 64 KB) + + After completing a streaming compression, + it's possible to start a new stream of blocks, using the same LZ4_streamHC_t state, + just by resetting it, using LZ4_resetStreamHC_fast(). +*/ + +LZ4LIB_API void LZ4_resetStreamHC_fast(LZ4_streamHC_t* streamHCPtr, int compressionLevel); /* v1.9.0+ */ +LZ4LIB_API int LZ4_loadDictHC (LZ4_streamHC_t* streamHCPtr, const char* dictionary, int dictSize); + +LZ4LIB_API int LZ4_compress_HC_continue (LZ4_streamHC_t* streamHCPtr, + const char* src, char* dst, + int srcSize, int maxDstSize); + +/*! LZ4_compress_HC_continue_destSize() : v1.9.0+ + * Similar to LZ4_compress_HC_continue(), + * but will read as much data as possible from `src` + * to fit into `targetDstSize` budget. + * Result is provided into 2 parts : + * @return : the number of bytes written into 'dst' (necessarily <= targetDstSize) + * or 0 if compression fails. + * `srcSizePtr` : on success, *srcSizePtr will be updated to indicate how much bytes were read from `src`. + * Note that this function may not consume the entire input. + */ +LZ4LIB_API int LZ4_compress_HC_continue_destSize(LZ4_streamHC_t* LZ4_streamHCPtr, + const char* src, char* dst, + int* srcSizePtr, int targetDstSize); + +LZ4LIB_API int LZ4_saveDictHC (LZ4_streamHC_t* streamHCPtr, char* safeBuffer, int maxDictSize); + + + +/*^********************************************** + * !!!!!! STATIC LINKING ONLY !!!!!! + ***********************************************/ + +/*-****************************************************************** + * PRIVATE DEFINITIONS : + * Do not use these definitions directly. + * They are merely exposed to allow static allocation of `LZ4_streamHC_t`. + * Declare an `LZ4_streamHC_t` directly, rather than any type below. + * Even then, only do so in the context of static linking, as definitions may change between versions. + ********************************************************************/ + +#define LZ4HC_DICTIONARY_LOGSIZE 16 +#define LZ4HC_MAXD (1<= LZ4HC_CLEVEL_OPT_MIN. + */ +LZ4LIB_STATIC_API void LZ4_favorDecompressionSpeed( + LZ4_streamHC_t* LZ4_streamHCPtr, int favor); + +/*! LZ4_resetStreamHC_fast() : v1.9.0+ + * When an LZ4_streamHC_t is known to be in a internally coherent state, + * it can often be prepared for a new compression with almost no work, only + * sometimes falling back to the full, expensive reset that is always required + * when the stream is in an indeterminate state (i.e., the reset performed by + * LZ4_resetStreamHC()). + * + * LZ4_streamHCs are guaranteed to be in a valid state when: + * - returned from LZ4_createStreamHC() + * - reset by LZ4_resetStreamHC() + * - memset(stream, 0, sizeof(LZ4_streamHC_t)) + * - the stream was in a valid state and was reset by LZ4_resetStreamHC_fast() + * - the stream was in a valid state and was then used in any compression call + * that returned success + * - the stream was in an indeterminate state and was used in a compression + * call that fully reset the state (LZ4_compress_HC_extStateHC()) and that + * returned success + * + * Note: + * A stream that was last used in a compression call that returned an error + * may be passed to this function. However, it will be fully reset, which will + * clear any existing history and settings from the context. + */ +LZ4LIB_STATIC_API void LZ4_resetStreamHC_fast( + LZ4_streamHC_t* LZ4_streamHCPtr, int compressionLevel); + +/*! LZ4_compress_HC_extStateHC_fastReset() : + * A variant of LZ4_compress_HC_extStateHC(). + * + * Using this variant avoids an expensive initialization step. It is only safe + * to call if the state buffer is known to be correctly initialized already + * (see above comment on LZ4_resetStreamHC_fast() for a definition of + * "correctly initialized"). From a high level, the difference is that this + * function initializes the provided state with a call to + * LZ4_resetStreamHC_fast() while LZ4_compress_HC_extStateHC() starts with a + * call to LZ4_resetStreamHC(). + */ +LZ4LIB_STATIC_API int LZ4_compress_HC_extStateHC_fastReset ( + void* state, + const char* src, char* dst, + int srcSize, int dstCapacity, + int compressionLevel); + +/*! LZ4_attach_HC_dictionary() : + * This is an experimental API that allows for the efficient use of a + * static dictionary many times. + * + * Rather than re-loading the dictionary buffer into a working context before + * each compression, or copying a pre-loaded dictionary's LZ4_streamHC_t into a + * working LZ4_streamHC_t, this function introduces a no-copy setup mechanism, + * in which the working stream references the dictionary stream in-place. + * + * Several assumptions are made about the state of the dictionary stream. + * Currently, only streams which have been prepared by LZ4_loadDictHC() should + * be expected to work. + * + * Alternatively, the provided dictionary stream pointer may be NULL, in which + * case any existing dictionary stream is unset. + * + * A dictionary should only be attached to a stream without any history (i.e., + * a stream that has just been reset). + * + * The dictionary will remain attached to the working stream only for the + * current stream session. Calls to LZ4_resetStreamHC(_fast) will remove the + * dictionary context association from the working stream. The dictionary + * stream (and source buffer) must remain in-place / accessible / unchanged + * through the lifetime of the stream session. + */ +LZ4LIB_STATIC_API void LZ4_attach_HC_dictionary( + LZ4_streamHC_t *working_stream, + const LZ4_streamHC_t *dictionary_stream); + +#if defined (__cplusplus) +} +#endif + +#endif /* LZ4_HC_SLO_098092834 */ +#endif /* LZ4_HC_STATIC_LINKING_ONLY */ diff --git a/example/android/third_party/lz4/include/xxhash.c b/example/android/third_party/lz4/include/xxhash.c new file mode 100644 index 00000000..08d8b360 --- /dev/null +++ b/example/android/third_party/lz4/include/xxhash.c @@ -0,0 +1,1030 @@ +/* +* xxHash - Fast Hash algorithm +* Copyright (C) 2012-2016, Yann Collet +* +* BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) +* +* Redistribution and use in source and binary forms, with or without +* modification, are permitted provided that the following conditions are +* met: +* +* * Redistributions of source code must retain the above copyright +* notice, this list of conditions and the following disclaimer. +* * Redistributions in binary form must reproduce the above +* copyright notice, this list of conditions and the following disclaimer +* in the documentation and/or other materials provided with the +* distribution. +* +* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +* "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +* LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +* A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +* OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +* SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +* LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +* DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY +* THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. +* +* You can contact the author at : +* - xxHash homepage: http://www.xxhash.com +* - xxHash source repository : https://github.com/Cyan4973/xxHash +*/ + + +/* ************************************* +* Tuning parameters +***************************************/ +/*!XXH_FORCE_MEMORY_ACCESS : + * By default, access to unaligned memory is controlled by `memcpy()`, which is safe and portable. + * Unfortunately, on some target/compiler combinations, the generated assembly is sub-optimal. + * The below switch allow to select different access method for improved performance. + * Method 0 (default) : use `memcpy()`. Safe and portable. + * Method 1 : `__packed` statement. It depends on compiler extension (ie, not portable). + * This method is safe if your compiler supports it, and *generally* as fast or faster than `memcpy`. + * Method 2 : direct access. This method doesn't depend on compiler but violate C standard. + * It can generate buggy code on targets which do not support unaligned memory accesses. + * But in some circumstances, it's the only known way to get the most performance (ie GCC + ARMv6) + * See http://stackoverflow.com/a/32095106/646947 for details. + * Prefer these methods in priority order (0 > 1 > 2) + */ +#ifndef XXH_FORCE_MEMORY_ACCESS /* can be defined externally, on command line for example */ +# if defined(__GNUC__) && ( defined(__ARM_ARCH_6__) || defined(__ARM_ARCH_6J__) \ + || defined(__ARM_ARCH_6K__) || defined(__ARM_ARCH_6Z__) \ + || defined(__ARM_ARCH_6ZK__) || defined(__ARM_ARCH_6T2__) ) +# define XXH_FORCE_MEMORY_ACCESS 2 +# elif (defined(__INTEL_COMPILER) && !defined(_WIN32)) || \ + (defined(__GNUC__) && ( defined(__ARM_ARCH_7__) || defined(__ARM_ARCH_7A__) \ + || defined(__ARM_ARCH_7R__) || defined(__ARM_ARCH_7M__) \ + || defined(__ARM_ARCH_7S__) )) +# define XXH_FORCE_MEMORY_ACCESS 1 +# endif +#endif + +/*!XXH_ACCEPT_NULL_INPUT_POINTER : + * If input pointer is NULL, xxHash default behavior is to dereference it, triggering a segfault. + * When this macro is enabled, xxHash actively checks input for null pointer. + * It it is, result for null input pointers is the same as a null-length input. + */ +#ifndef XXH_ACCEPT_NULL_INPUT_POINTER /* can be defined externally */ +# define XXH_ACCEPT_NULL_INPUT_POINTER 0 +#endif + +/*!XXH_FORCE_NATIVE_FORMAT : + * By default, xxHash library provides endian-independent Hash values, based on little-endian convention. + * Results are therefore identical for little-endian and big-endian CPU. + * This comes at a performance cost for big-endian CPU, since some swapping is required to emulate little-endian format. + * Should endian-independence be of no importance for your application, you may set the #define below to 1, + * to improve speed for Big-endian CPU. + * This option has no impact on Little_Endian CPU. + */ +#ifndef XXH_FORCE_NATIVE_FORMAT /* can be defined externally */ +# define XXH_FORCE_NATIVE_FORMAT 0 +#endif + +/*!XXH_FORCE_ALIGN_CHECK : + * This is a minor performance trick, only useful with lots of very small keys. + * It means : check for aligned/unaligned input. + * The check costs one initial branch per hash; + * set it to 0 when the input is guaranteed to be aligned, + * or when alignment doesn't matter for performance. + */ +#ifndef XXH_FORCE_ALIGN_CHECK /* can be defined externally */ +# if defined(__i386) || defined(_M_IX86) || defined(__x86_64__) || defined(_M_X64) +# define XXH_FORCE_ALIGN_CHECK 0 +# else +# define XXH_FORCE_ALIGN_CHECK 1 +# endif +#endif + + +/* ************************************* +* Includes & Memory related functions +***************************************/ +/*! Modify the local functions below should you wish to use some other memory routines +* for malloc(), free() */ +#include +static void* XXH_malloc(size_t s) { return malloc(s); } +static void XXH_free (void* p) { free(p); } +/*! and for memcpy() */ +#include +static void* XXH_memcpy(void* dest, const void* src, size_t size) { return memcpy(dest,src,size); } + +#include /* assert */ + +#define XXH_STATIC_LINKING_ONLY +#include "xxhash.h" + + +/* ************************************* +* Compiler Specific Options +***************************************/ +#if defined (_MSC_VER) && !defined (__clang__) /* MSVC */ +# pragma warning(disable : 4127) /* disable: C4127: conditional expression is constant */ +# define FORCE_INLINE static __forceinline +#else +# if defined (__cplusplus) || defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L /* C99 */ +# if defined (__GNUC__) || defined (__clang__) +# define FORCE_INLINE static inline __attribute__((always_inline)) +# else +# define FORCE_INLINE static inline +# endif +# else +# define FORCE_INLINE static +# endif /* __STDC_VERSION__ */ +#endif + + +/* ************************************* +* Basic Types +***************************************/ +#ifndef MEM_MODULE +# if !defined (__VMS) \ + && (defined (__cplusplus) \ + || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) ) +# include + typedef uint8_t BYTE; + typedef uint16_t U16; + typedef uint32_t U32; +# else + typedef unsigned char BYTE; + typedef unsigned short U16; + typedef unsigned int U32; +# endif +#endif + +#if (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==2)) + +/* Force direct memory access. Only works on CPU which support unaligned memory access in hardware */ +static U32 XXH_read32(const void* memPtr) { return *(const U32*) memPtr; } + +#elif (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==1)) + +/* __pack instructions are safer, but compiler specific, hence potentially problematic for some compilers */ +/* currently only defined for gcc and icc */ +typedef union { U32 u32; } __attribute__((packed)) unalign; +static U32 XXH_read32(const void* ptr) { return ((const unalign*)ptr)->u32; } + +#else + +/* portable and safe solution. Generally efficient. + * see : http://stackoverflow.com/a/32095106/646947 + */ +static U32 XXH_read32(const void* memPtr) +{ + U32 val; + memcpy(&val, memPtr, sizeof(val)); + return val; +} + +#endif /* XXH_FORCE_DIRECT_MEMORY_ACCESS */ + + +/* **************************************** +* Compiler-specific Functions and Macros +******************************************/ +#define XXH_GCC_VERSION (__GNUC__ * 100 + __GNUC_MINOR__) + +/* Note : although _rotl exists for minGW (GCC under windows), performance seems poor */ +#if defined(_MSC_VER) +# define XXH_rotl32(x,r) _rotl(x,r) +# define XXH_rotl64(x,r) _rotl64(x,r) +#else +# define XXH_rotl32(x,r) ((x << r) | (x >> (32 - r))) +# define XXH_rotl64(x,r) ((x << r) | (x >> (64 - r))) +#endif + +#if defined(_MSC_VER) /* Visual Studio */ +# define XXH_swap32 _byteswap_ulong +#elif XXH_GCC_VERSION >= 403 +# define XXH_swap32 __builtin_bswap32 +#else +static U32 XXH_swap32 (U32 x) +{ + return ((x << 24) & 0xff000000 ) | + ((x << 8) & 0x00ff0000 ) | + ((x >> 8) & 0x0000ff00 ) | + ((x >> 24) & 0x000000ff ); +} +#endif + + +/* ************************************* +* Architecture Macros +***************************************/ +typedef enum { XXH_bigEndian=0, XXH_littleEndian=1 } XXH_endianess; + +/* XXH_CPU_LITTLE_ENDIAN can be defined externally, for example on the compiler command line */ +#ifndef XXH_CPU_LITTLE_ENDIAN +static int XXH_isLittleEndian(void) +{ + const union { U32 u; BYTE c[4]; } one = { 1 }; /* don't use static : performance detrimental */ + return one.c[0]; +} +# define XXH_CPU_LITTLE_ENDIAN XXH_isLittleEndian() +#endif + + +/* *************************** +* Memory reads +*****************************/ +typedef enum { XXH_aligned, XXH_unaligned } XXH_alignment; + +FORCE_INLINE U32 XXH_readLE32_align(const void* ptr, XXH_endianess endian, XXH_alignment align) +{ + if (align==XXH_unaligned) + return endian==XXH_littleEndian ? XXH_read32(ptr) : XXH_swap32(XXH_read32(ptr)); + else + return endian==XXH_littleEndian ? *(const U32*)ptr : XXH_swap32(*(const U32*)ptr); +} + +FORCE_INLINE U32 XXH_readLE32(const void* ptr, XXH_endianess endian) +{ + return XXH_readLE32_align(ptr, endian, XXH_unaligned); +} + +static U32 XXH_readBE32(const void* ptr) +{ + return XXH_CPU_LITTLE_ENDIAN ? XXH_swap32(XXH_read32(ptr)) : XXH_read32(ptr); +} + + +/* ************************************* +* Macros +***************************************/ +#define XXH_STATIC_ASSERT(c) { enum { XXH_sa = 1/(int)(!!(c)) }; } /* use after variable declarations */ +XXH_PUBLIC_API unsigned XXH_versionNumber (void) { return XXH_VERSION_NUMBER; } + + +/* ******************************************************************* +* 32-bit hash functions +*********************************************************************/ +static const U32 PRIME32_1 = 2654435761U; +static const U32 PRIME32_2 = 2246822519U; +static const U32 PRIME32_3 = 3266489917U; +static const U32 PRIME32_4 = 668265263U; +static const U32 PRIME32_5 = 374761393U; + +static U32 XXH32_round(U32 seed, U32 input) +{ + seed += input * PRIME32_2; + seed = XXH_rotl32(seed, 13); + seed *= PRIME32_1; + return seed; +} + +/* mix all bits */ +static U32 XXH32_avalanche(U32 h32) +{ + h32 ^= h32 >> 15; + h32 *= PRIME32_2; + h32 ^= h32 >> 13; + h32 *= PRIME32_3; + h32 ^= h32 >> 16; + return(h32); +} + +#define XXH_get32bits(p) XXH_readLE32_align(p, endian, align) + +static U32 +XXH32_finalize(U32 h32, const void* ptr, size_t len, + XXH_endianess endian, XXH_alignment align) + +{ + const BYTE* p = (const BYTE*)ptr; + +#define PROCESS1 \ + h32 += (*p++) * PRIME32_5; \ + h32 = XXH_rotl32(h32, 11) * PRIME32_1 ; + +#define PROCESS4 \ + h32 += XXH_get32bits(p) * PRIME32_3; \ + p+=4; \ + h32 = XXH_rotl32(h32, 17) * PRIME32_4 ; + + switch(len&15) /* or switch(bEnd - p) */ + { + case 12: PROCESS4; + /* fallthrough */ + case 8: PROCESS4; + /* fallthrough */ + case 4: PROCESS4; + return XXH32_avalanche(h32); + + case 13: PROCESS4; + /* fallthrough */ + case 9: PROCESS4; + /* fallthrough */ + case 5: PROCESS4; + PROCESS1; + return XXH32_avalanche(h32); + + case 14: PROCESS4; + /* fallthrough */ + case 10: PROCESS4; + /* fallthrough */ + case 6: PROCESS4; + PROCESS1; + PROCESS1; + return XXH32_avalanche(h32); + + case 15: PROCESS4; + /* fallthrough */ + case 11: PROCESS4; + /* fallthrough */ + case 7: PROCESS4; + /* fallthrough */ + case 3: PROCESS1; + /* fallthrough */ + case 2: PROCESS1; + /* fallthrough */ + case 1: PROCESS1; + /* fallthrough */ + case 0: return XXH32_avalanche(h32); + } + assert(0); + return h32; /* reaching this point is deemed impossible */ +} + + +FORCE_INLINE U32 +XXH32_endian_align(const void* input, size_t len, U32 seed, + XXH_endianess endian, XXH_alignment align) +{ + const BYTE* p = (const BYTE*)input; + const BYTE* bEnd = p + len; + U32 h32; + +#if defined(XXH_ACCEPT_NULL_INPUT_POINTER) && (XXH_ACCEPT_NULL_INPUT_POINTER>=1) + if (p==NULL) { + len=0; + bEnd=p=(const BYTE*)(size_t)16; + } +#endif + + if (len>=16) { + const BYTE* const limit = bEnd - 15; + U32 v1 = seed + PRIME32_1 + PRIME32_2; + U32 v2 = seed + PRIME32_2; + U32 v3 = seed + 0; + U32 v4 = seed - PRIME32_1; + + do { + v1 = XXH32_round(v1, XXH_get32bits(p)); p+=4; + v2 = XXH32_round(v2, XXH_get32bits(p)); p+=4; + v3 = XXH32_round(v3, XXH_get32bits(p)); p+=4; + v4 = XXH32_round(v4, XXH_get32bits(p)); p+=4; + } while (p < limit); + + h32 = XXH_rotl32(v1, 1) + XXH_rotl32(v2, 7) + + XXH_rotl32(v3, 12) + XXH_rotl32(v4, 18); + } else { + h32 = seed + PRIME32_5; + } + + h32 += (U32)len; + + return XXH32_finalize(h32, p, len&15, endian, align); +} + + +XXH_PUBLIC_API unsigned int XXH32 (const void* input, size_t len, unsigned int seed) +{ +#if 0 + /* Simple version, good for code maintenance, but unfortunately slow for small inputs */ + XXH32_state_t state; + XXH32_reset(&state, seed); + XXH32_update(&state, input, len); + return XXH32_digest(&state); +#else + XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN; + + if (XXH_FORCE_ALIGN_CHECK) { + if ((((size_t)input) & 3) == 0) { /* Input is 4-bytes aligned, leverage the speed benefit */ + if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT) + return XXH32_endian_align(input, len, seed, XXH_littleEndian, XXH_aligned); + else + return XXH32_endian_align(input, len, seed, XXH_bigEndian, XXH_aligned); + } } + + if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT) + return XXH32_endian_align(input, len, seed, XXH_littleEndian, XXH_unaligned); + else + return XXH32_endian_align(input, len, seed, XXH_bigEndian, XXH_unaligned); +#endif +} + + + +/*====== Hash streaming ======*/ + +XXH_PUBLIC_API XXH32_state_t* XXH32_createState(void) +{ + return (XXH32_state_t*)XXH_malloc(sizeof(XXH32_state_t)); +} +XXH_PUBLIC_API XXH_errorcode XXH32_freeState(XXH32_state_t* statePtr) +{ + XXH_free(statePtr); + return XXH_OK; +} + +XXH_PUBLIC_API void XXH32_copyState(XXH32_state_t* dstState, const XXH32_state_t* srcState) +{ + memcpy(dstState, srcState, sizeof(*dstState)); +} + +XXH_PUBLIC_API XXH_errorcode XXH32_reset(XXH32_state_t* statePtr, unsigned int seed) +{ + XXH32_state_t state; /* using a local state to memcpy() in order to avoid strict-aliasing warnings */ + memset(&state, 0, sizeof(state)); + state.v1 = seed + PRIME32_1 + PRIME32_2; + state.v2 = seed + PRIME32_2; + state.v3 = seed + 0; + state.v4 = seed - PRIME32_1; + /* do not write into reserved, planned to be removed in a future version */ + memcpy(statePtr, &state, sizeof(state) - sizeof(state.reserved)); + return XXH_OK; +} + + +FORCE_INLINE XXH_errorcode +XXH32_update_endian(XXH32_state_t* state, const void* input, size_t len, XXH_endianess endian) +{ + if (input==NULL) +#if defined(XXH_ACCEPT_NULL_INPUT_POINTER) && (XXH_ACCEPT_NULL_INPUT_POINTER>=1) + return XXH_OK; +#else + return XXH_ERROR; +#endif + + { const BYTE* p = (const BYTE*)input; + const BYTE* const bEnd = p + len; + + state->total_len_32 += (unsigned)len; + state->large_len |= (len>=16) | (state->total_len_32>=16); + + if (state->memsize + len < 16) { /* fill in tmp buffer */ + XXH_memcpy((BYTE*)(state->mem32) + state->memsize, input, len); + state->memsize += (unsigned)len; + return XXH_OK; + } + + if (state->memsize) { /* some data left from previous update */ + XXH_memcpy((BYTE*)(state->mem32) + state->memsize, input, 16-state->memsize); + { const U32* p32 = state->mem32; + state->v1 = XXH32_round(state->v1, XXH_readLE32(p32, endian)); p32++; + state->v2 = XXH32_round(state->v2, XXH_readLE32(p32, endian)); p32++; + state->v3 = XXH32_round(state->v3, XXH_readLE32(p32, endian)); p32++; + state->v4 = XXH32_round(state->v4, XXH_readLE32(p32, endian)); + } + p += 16-state->memsize; + state->memsize = 0; + } + + if (p <= bEnd-16) { + const BYTE* const limit = bEnd - 16; + U32 v1 = state->v1; + U32 v2 = state->v2; + U32 v3 = state->v3; + U32 v4 = state->v4; + + do { + v1 = XXH32_round(v1, XXH_readLE32(p, endian)); p+=4; + v2 = XXH32_round(v2, XXH_readLE32(p, endian)); p+=4; + v3 = XXH32_round(v3, XXH_readLE32(p, endian)); p+=4; + v4 = XXH32_round(v4, XXH_readLE32(p, endian)); p+=4; + } while (p<=limit); + + state->v1 = v1; + state->v2 = v2; + state->v3 = v3; + state->v4 = v4; + } + + if (p < bEnd) { + XXH_memcpy(state->mem32, p, (size_t)(bEnd-p)); + state->memsize = (unsigned)(bEnd-p); + } + } + + return XXH_OK; +} + + +XXH_PUBLIC_API XXH_errorcode XXH32_update (XXH32_state_t* state_in, const void* input, size_t len) +{ + XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN; + + if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT) + return XXH32_update_endian(state_in, input, len, XXH_littleEndian); + else + return XXH32_update_endian(state_in, input, len, XXH_bigEndian); +} + + +FORCE_INLINE U32 +XXH32_digest_endian (const XXH32_state_t* state, XXH_endianess endian) +{ + U32 h32; + + if (state->large_len) { + h32 = XXH_rotl32(state->v1, 1) + + XXH_rotl32(state->v2, 7) + + XXH_rotl32(state->v3, 12) + + XXH_rotl32(state->v4, 18); + } else { + h32 = state->v3 /* == seed */ + PRIME32_5; + } + + h32 += state->total_len_32; + + return XXH32_finalize(h32, state->mem32, state->memsize, endian, XXH_aligned); +} + + +XXH_PUBLIC_API unsigned int XXH32_digest (const XXH32_state_t* state_in) +{ + XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN; + + if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT) + return XXH32_digest_endian(state_in, XXH_littleEndian); + else + return XXH32_digest_endian(state_in, XXH_bigEndian); +} + + +/*====== Canonical representation ======*/ + +/*! Default XXH result types are basic unsigned 32 and 64 bits. +* The canonical representation follows human-readable write convention, aka big-endian (large digits first). +* These functions allow transformation of hash result into and from its canonical format. +* This way, hash values can be written into a file or buffer, remaining comparable across different systems. +*/ + +XXH_PUBLIC_API void XXH32_canonicalFromHash(XXH32_canonical_t* dst, XXH32_hash_t hash) +{ + XXH_STATIC_ASSERT(sizeof(XXH32_canonical_t) == sizeof(XXH32_hash_t)); + if (XXH_CPU_LITTLE_ENDIAN) hash = XXH_swap32(hash); + memcpy(dst, &hash, sizeof(*dst)); +} + +XXH_PUBLIC_API XXH32_hash_t XXH32_hashFromCanonical(const XXH32_canonical_t* src) +{ + return XXH_readBE32(src); +} + + +#ifndef XXH_NO_LONG_LONG + +/* ******************************************************************* +* 64-bit hash functions +*********************************************************************/ + +/*====== Memory access ======*/ + +#ifndef MEM_MODULE +# define MEM_MODULE +# if !defined (__VMS) \ + && (defined (__cplusplus) \ + || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) ) +# include + typedef uint64_t U64; +# else + /* if compiler doesn't support unsigned long long, replace by another 64-bit type */ + typedef unsigned long long U64; +# endif +#endif + + +#if (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==2)) + +/* Force direct memory access. Only works on CPU which support unaligned memory access in hardware */ +static U64 XXH_read64(const void* memPtr) { return *(const U64*) memPtr; } + +#elif (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==1)) + +/* __pack instructions are safer, but compiler specific, hence potentially problematic for some compilers */ +/* currently only defined for gcc and icc */ +typedef union { U32 u32; U64 u64; } __attribute__((packed)) unalign64; +static U64 XXH_read64(const void* ptr) { return ((const unalign64*)ptr)->u64; } + +#else + +/* portable and safe solution. Generally efficient. + * see : http://stackoverflow.com/a/32095106/646947 + */ + +static U64 XXH_read64(const void* memPtr) +{ + U64 val; + memcpy(&val, memPtr, sizeof(val)); + return val; +} + +#endif /* XXH_FORCE_DIRECT_MEMORY_ACCESS */ + +#if defined(_MSC_VER) /* Visual Studio */ +# define XXH_swap64 _byteswap_uint64 +#elif XXH_GCC_VERSION >= 403 +# define XXH_swap64 __builtin_bswap64 +#else +static U64 XXH_swap64 (U64 x) +{ + return ((x << 56) & 0xff00000000000000ULL) | + ((x << 40) & 0x00ff000000000000ULL) | + ((x << 24) & 0x0000ff0000000000ULL) | + ((x << 8) & 0x000000ff00000000ULL) | + ((x >> 8) & 0x00000000ff000000ULL) | + ((x >> 24) & 0x0000000000ff0000ULL) | + ((x >> 40) & 0x000000000000ff00ULL) | + ((x >> 56) & 0x00000000000000ffULL); +} +#endif + +FORCE_INLINE U64 XXH_readLE64_align(const void* ptr, XXH_endianess endian, XXH_alignment align) +{ + if (align==XXH_unaligned) + return endian==XXH_littleEndian ? XXH_read64(ptr) : XXH_swap64(XXH_read64(ptr)); + else + return endian==XXH_littleEndian ? *(const U64*)ptr : XXH_swap64(*(const U64*)ptr); +} + +FORCE_INLINE U64 XXH_readLE64(const void* ptr, XXH_endianess endian) +{ + return XXH_readLE64_align(ptr, endian, XXH_unaligned); +} + +static U64 XXH_readBE64(const void* ptr) +{ + return XXH_CPU_LITTLE_ENDIAN ? XXH_swap64(XXH_read64(ptr)) : XXH_read64(ptr); +} + + +/*====== xxh64 ======*/ + +static const U64 PRIME64_1 = 11400714785074694791ULL; +static const U64 PRIME64_2 = 14029467366897019727ULL; +static const U64 PRIME64_3 = 1609587929392839161ULL; +static const U64 PRIME64_4 = 9650029242287828579ULL; +static const U64 PRIME64_5 = 2870177450012600261ULL; + +static U64 XXH64_round(U64 acc, U64 input) +{ + acc += input * PRIME64_2; + acc = XXH_rotl64(acc, 31); + acc *= PRIME64_1; + return acc; +} + +static U64 XXH64_mergeRound(U64 acc, U64 val) +{ + val = XXH64_round(0, val); + acc ^= val; + acc = acc * PRIME64_1 + PRIME64_4; + return acc; +} + +static U64 XXH64_avalanche(U64 h64) +{ + h64 ^= h64 >> 33; + h64 *= PRIME64_2; + h64 ^= h64 >> 29; + h64 *= PRIME64_3; + h64 ^= h64 >> 32; + return h64; +} + + +#define XXH_get64bits(p) XXH_readLE64_align(p, endian, align) + +static U64 +XXH64_finalize(U64 h64, const void* ptr, size_t len, + XXH_endianess endian, XXH_alignment align) +{ + const BYTE* p = (const BYTE*)ptr; + +#define PROCESS1_64 \ + h64 ^= (*p++) * PRIME64_5; \ + h64 = XXH_rotl64(h64, 11) * PRIME64_1; + +#define PROCESS4_64 \ + h64 ^= (U64)(XXH_get32bits(p)) * PRIME64_1; \ + p+=4; \ + h64 = XXH_rotl64(h64, 23) * PRIME64_2 + PRIME64_3; + +#define PROCESS8_64 { \ + U64 const k1 = XXH64_round(0, XXH_get64bits(p)); \ + p+=8; \ + h64 ^= k1; \ + h64 = XXH_rotl64(h64,27) * PRIME64_1 + PRIME64_4; \ +} + + switch(len&31) { + case 24: PROCESS8_64; + /* fallthrough */ + case 16: PROCESS8_64; + /* fallthrough */ + case 8: PROCESS8_64; + return XXH64_avalanche(h64); + + case 28: PROCESS8_64; + /* fallthrough */ + case 20: PROCESS8_64; + /* fallthrough */ + case 12: PROCESS8_64; + /* fallthrough */ + case 4: PROCESS4_64; + return XXH64_avalanche(h64); + + case 25: PROCESS8_64; + /* fallthrough */ + case 17: PROCESS8_64; + /* fallthrough */ + case 9: PROCESS8_64; + PROCESS1_64; + return XXH64_avalanche(h64); + + case 29: PROCESS8_64; + /* fallthrough */ + case 21: PROCESS8_64; + /* fallthrough */ + case 13: PROCESS8_64; + /* fallthrough */ + case 5: PROCESS4_64; + PROCESS1_64; + return XXH64_avalanche(h64); + + case 26: PROCESS8_64; + /* fallthrough */ + case 18: PROCESS8_64; + /* fallthrough */ + case 10: PROCESS8_64; + PROCESS1_64; + PROCESS1_64; + return XXH64_avalanche(h64); + + case 30: PROCESS8_64; + /* fallthrough */ + case 22: PROCESS8_64; + /* fallthrough */ + case 14: PROCESS8_64; + /* fallthrough */ + case 6: PROCESS4_64; + PROCESS1_64; + PROCESS1_64; + return XXH64_avalanche(h64); + + case 27: PROCESS8_64; + /* fallthrough */ + case 19: PROCESS8_64; + /* fallthrough */ + case 11: PROCESS8_64; + PROCESS1_64; + PROCESS1_64; + PROCESS1_64; + return XXH64_avalanche(h64); + + case 31: PROCESS8_64; + /* fallthrough */ + case 23: PROCESS8_64; + /* fallthrough */ + case 15: PROCESS8_64; + /* fallthrough */ + case 7: PROCESS4_64; + /* fallthrough */ + case 3: PROCESS1_64; + /* fallthrough */ + case 2: PROCESS1_64; + /* fallthrough */ + case 1: PROCESS1_64; + /* fallthrough */ + case 0: return XXH64_avalanche(h64); + } + + /* impossible to reach */ + assert(0); + return 0; /* unreachable, but some compilers complain without it */ +} + +FORCE_INLINE U64 +XXH64_endian_align(const void* input, size_t len, U64 seed, + XXH_endianess endian, XXH_alignment align) +{ + const BYTE* p = (const BYTE*)input; + const BYTE* bEnd = p + len; + U64 h64; + +#if defined(XXH_ACCEPT_NULL_INPUT_POINTER) && (XXH_ACCEPT_NULL_INPUT_POINTER>=1) + if (p==NULL) { + len=0; + bEnd=p=(const BYTE*)(size_t)32; + } +#endif + + if (len>=32) { + const BYTE* const limit = bEnd - 32; + U64 v1 = seed + PRIME64_1 + PRIME64_2; + U64 v2 = seed + PRIME64_2; + U64 v3 = seed + 0; + U64 v4 = seed - PRIME64_1; + + do { + v1 = XXH64_round(v1, XXH_get64bits(p)); p+=8; + v2 = XXH64_round(v2, XXH_get64bits(p)); p+=8; + v3 = XXH64_round(v3, XXH_get64bits(p)); p+=8; + v4 = XXH64_round(v4, XXH_get64bits(p)); p+=8; + } while (p<=limit); + + h64 = XXH_rotl64(v1, 1) + XXH_rotl64(v2, 7) + XXH_rotl64(v3, 12) + XXH_rotl64(v4, 18); + h64 = XXH64_mergeRound(h64, v1); + h64 = XXH64_mergeRound(h64, v2); + h64 = XXH64_mergeRound(h64, v3); + h64 = XXH64_mergeRound(h64, v4); + + } else { + h64 = seed + PRIME64_5; + } + + h64 += (U64) len; + + return XXH64_finalize(h64, p, len, endian, align); +} + + +XXH_PUBLIC_API unsigned long long XXH64 (const void* input, size_t len, unsigned long long seed) +{ +#if 0 + /* Simple version, good for code maintenance, but unfortunately slow for small inputs */ + XXH64_state_t state; + XXH64_reset(&state, seed); + XXH64_update(&state, input, len); + return XXH64_digest(&state); +#else + XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN; + + if (XXH_FORCE_ALIGN_CHECK) { + if ((((size_t)input) & 7)==0) { /* Input is aligned, let's leverage the speed advantage */ + if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT) + return XXH64_endian_align(input, len, seed, XXH_littleEndian, XXH_aligned); + else + return XXH64_endian_align(input, len, seed, XXH_bigEndian, XXH_aligned); + } } + + if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT) + return XXH64_endian_align(input, len, seed, XXH_littleEndian, XXH_unaligned); + else + return XXH64_endian_align(input, len, seed, XXH_bigEndian, XXH_unaligned); +#endif +} + +/*====== Hash Streaming ======*/ + +XXH_PUBLIC_API XXH64_state_t* XXH64_createState(void) +{ + return (XXH64_state_t*)XXH_malloc(sizeof(XXH64_state_t)); +} +XXH_PUBLIC_API XXH_errorcode XXH64_freeState(XXH64_state_t* statePtr) +{ + XXH_free(statePtr); + return XXH_OK; +} + +XXH_PUBLIC_API void XXH64_copyState(XXH64_state_t* dstState, const XXH64_state_t* srcState) +{ + memcpy(dstState, srcState, sizeof(*dstState)); +} + +XXH_PUBLIC_API XXH_errorcode XXH64_reset(XXH64_state_t* statePtr, unsigned long long seed) +{ + XXH64_state_t state; /* using a local state to memcpy() in order to avoid strict-aliasing warnings */ + memset(&state, 0, sizeof(state)); + state.v1 = seed + PRIME64_1 + PRIME64_2; + state.v2 = seed + PRIME64_2; + state.v3 = seed + 0; + state.v4 = seed - PRIME64_1; + /* do not write into reserved, planned to be removed in a future version */ + memcpy(statePtr, &state, sizeof(state) - sizeof(state.reserved)); + return XXH_OK; +} + +FORCE_INLINE XXH_errorcode +XXH64_update_endian (XXH64_state_t* state, const void* input, size_t len, XXH_endianess endian) +{ + if (input==NULL) +#if defined(XXH_ACCEPT_NULL_INPUT_POINTER) && (XXH_ACCEPT_NULL_INPUT_POINTER>=1) + return XXH_OK; +#else + return XXH_ERROR; +#endif + + { const BYTE* p = (const BYTE*)input; + const BYTE* const bEnd = p + len; + + state->total_len += len; + + if (state->memsize + len < 32) { /* fill in tmp buffer */ + XXH_memcpy(((BYTE*)state->mem64) + state->memsize, input, len); + state->memsize += (U32)len; + return XXH_OK; + } + + if (state->memsize) { /* tmp buffer is full */ + XXH_memcpy(((BYTE*)state->mem64) + state->memsize, input, 32-state->memsize); + state->v1 = XXH64_round(state->v1, XXH_readLE64(state->mem64+0, endian)); + state->v2 = XXH64_round(state->v2, XXH_readLE64(state->mem64+1, endian)); + state->v3 = XXH64_round(state->v3, XXH_readLE64(state->mem64+2, endian)); + state->v4 = XXH64_round(state->v4, XXH_readLE64(state->mem64+3, endian)); + p += 32-state->memsize; + state->memsize = 0; + } + + if (p+32 <= bEnd) { + const BYTE* const limit = bEnd - 32; + U64 v1 = state->v1; + U64 v2 = state->v2; + U64 v3 = state->v3; + U64 v4 = state->v4; + + do { + v1 = XXH64_round(v1, XXH_readLE64(p, endian)); p+=8; + v2 = XXH64_round(v2, XXH_readLE64(p, endian)); p+=8; + v3 = XXH64_round(v3, XXH_readLE64(p, endian)); p+=8; + v4 = XXH64_round(v4, XXH_readLE64(p, endian)); p+=8; + } while (p<=limit); + + state->v1 = v1; + state->v2 = v2; + state->v3 = v3; + state->v4 = v4; + } + + if (p < bEnd) { + XXH_memcpy(state->mem64, p, (size_t)(bEnd-p)); + state->memsize = (unsigned)(bEnd-p); + } + } + + return XXH_OK; +} + +XXH_PUBLIC_API XXH_errorcode XXH64_update (XXH64_state_t* state_in, const void* input, size_t len) +{ + XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN; + + if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT) + return XXH64_update_endian(state_in, input, len, XXH_littleEndian); + else + return XXH64_update_endian(state_in, input, len, XXH_bigEndian); +} + +FORCE_INLINE U64 XXH64_digest_endian (const XXH64_state_t* state, XXH_endianess endian) +{ + U64 h64; + + if (state->total_len >= 32) { + U64 const v1 = state->v1; + U64 const v2 = state->v2; + U64 const v3 = state->v3; + U64 const v4 = state->v4; + + h64 = XXH_rotl64(v1, 1) + XXH_rotl64(v2, 7) + XXH_rotl64(v3, 12) + XXH_rotl64(v4, 18); + h64 = XXH64_mergeRound(h64, v1); + h64 = XXH64_mergeRound(h64, v2); + h64 = XXH64_mergeRound(h64, v3); + h64 = XXH64_mergeRound(h64, v4); + } else { + h64 = state->v3 /*seed*/ + PRIME64_5; + } + + h64 += (U64) state->total_len; + + return XXH64_finalize(h64, state->mem64, (size_t)state->total_len, endian, XXH_aligned); +} + +XXH_PUBLIC_API unsigned long long XXH64_digest (const XXH64_state_t* state_in) +{ + XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN; + + if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT) + return XXH64_digest_endian(state_in, XXH_littleEndian); + else + return XXH64_digest_endian(state_in, XXH_bigEndian); +} + + +/*====== Canonical representation ======*/ + +XXH_PUBLIC_API void XXH64_canonicalFromHash(XXH64_canonical_t* dst, XXH64_hash_t hash) +{ + XXH_STATIC_ASSERT(sizeof(XXH64_canonical_t) == sizeof(XXH64_hash_t)); + if (XXH_CPU_LITTLE_ENDIAN) hash = XXH_swap64(hash); + memcpy(dst, &hash, sizeof(*dst)); +} + +XXH_PUBLIC_API XXH64_hash_t XXH64_hashFromCanonical(const XXH64_canonical_t* src) +{ + return XXH_readBE64(src); +} + +#endif /* XXH_NO_LONG_LONG */ diff --git a/example/android/third_party/lz4/include/xxhash.h b/example/android/third_party/lz4/include/xxhash.h new file mode 100644 index 00000000..d6bad943 --- /dev/null +++ b/example/android/third_party/lz4/include/xxhash.h @@ -0,0 +1,328 @@ +/* + xxHash - Extremely Fast Hash algorithm + Header File + Copyright (C) 2012-2016, Yann Collet. + + BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) + + Redistribution and use in source and binary forms, with or without + modification, are permitted provided that the following conditions are + met: + + * Redistributions of source code must retain the above copyright + notice, this list of conditions and the following disclaimer. + * Redistributions in binary form must reproduce the above + copyright notice, this list of conditions and the following disclaimer + in the documentation and/or other materials provided with the + distribution. + + THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + + You can contact the author at : + - xxHash source repository : https://github.com/Cyan4973/xxHash +*/ + +/* Notice extracted from xxHash homepage : + +xxHash is an extremely fast Hash algorithm, running at RAM speed limits. +It also successfully passes all tests from the SMHasher suite. + +Comparison (single thread, Windows Seven 32 bits, using SMHasher on a Core 2 Duo @3GHz) + +Name Speed Q.Score Author +xxHash 5.4 GB/s 10 +CrapWow 3.2 GB/s 2 Andrew +MumurHash 3a 2.7 GB/s 10 Austin Appleby +SpookyHash 2.0 GB/s 10 Bob Jenkins +SBox 1.4 GB/s 9 Bret Mulvey +Lookup3 1.2 GB/s 9 Bob Jenkins +SuperFastHash 1.2 GB/s 1 Paul Hsieh +CityHash64 1.05 GB/s 10 Pike & Alakuijala +FNV 0.55 GB/s 5 Fowler, Noll, Vo +CRC32 0.43 GB/s 9 +MD5-32 0.33 GB/s 10 Ronald L. Rivest +SHA1-32 0.28 GB/s 10 + +Q.Score is a measure of quality of the hash function. +It depends on successfully passing SMHasher test set. +10 is a perfect score. + +A 64-bit version, named XXH64, is available since r35. +It offers much better speed, but for 64-bit applications only. +Name Speed on 64 bits Speed on 32 bits +XXH64 13.8 GB/s 1.9 GB/s +XXH32 6.8 GB/s 6.0 GB/s +*/ + +#ifndef XXHASH_H_5627135585666179 +#define XXHASH_H_5627135585666179 1 + +#if defined (__cplusplus) +extern "C" { +#endif + + +/* **************************** +* Definitions +******************************/ +#include /* size_t */ +typedef enum { XXH_OK=0, XXH_ERROR } XXH_errorcode; + + +/* **************************** + * API modifier + ******************************/ +/** XXH_INLINE_ALL (and XXH_PRIVATE_API) + * This is useful to include xxhash functions in `static` mode + * in order to inline them, and remove their symbol from the public list. + * Inlining can offer dramatic performance improvement on small keys. + * Methodology : + * #define XXH_INLINE_ALL + * #include "xxhash.h" + * `xxhash.c` is automatically included. + * It's not useful to compile and link it as a separate module. + */ +#if defined(XXH_INLINE_ALL) || defined(XXH_PRIVATE_API) +# ifndef XXH_STATIC_LINKING_ONLY +# define XXH_STATIC_LINKING_ONLY +# endif +# if defined(__GNUC__) +# define XXH_PUBLIC_API static __inline __attribute__((unused)) +# elif defined (__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) +# define XXH_PUBLIC_API static inline +# elif defined(_MSC_VER) +# define XXH_PUBLIC_API static __inline +# else + /* this version may generate warnings for unused static functions */ +# define XXH_PUBLIC_API static +# endif +#else +# define XXH_PUBLIC_API /* do nothing */ +#endif /* XXH_INLINE_ALL || XXH_PRIVATE_API */ + +/*! XXH_NAMESPACE, aka Namespace Emulation : + * + * If you want to include _and expose_ xxHash functions from within your own library, + * but also want to avoid symbol collisions with other libraries which may also include xxHash, + * + * you can use XXH_NAMESPACE, to automatically prefix any public symbol from xxhash library + * with the value of XXH_NAMESPACE (therefore, avoid NULL and numeric values). + * + * Note that no change is required within the calling program as long as it includes `xxhash.h` : + * regular symbol name will be automatically translated by this header. + */ +#ifdef XXH_NAMESPACE +# define XXH_CAT(A,B) A##B +# define XXH_NAME2(A,B) XXH_CAT(A,B) +# define XXH_versionNumber XXH_NAME2(XXH_NAMESPACE, XXH_versionNumber) +# define XXH32 XXH_NAME2(XXH_NAMESPACE, XXH32) +# define XXH32_createState XXH_NAME2(XXH_NAMESPACE, XXH32_createState) +# define XXH32_freeState XXH_NAME2(XXH_NAMESPACE, XXH32_freeState) +# define XXH32_reset XXH_NAME2(XXH_NAMESPACE, XXH32_reset) +# define XXH32_update XXH_NAME2(XXH_NAMESPACE, XXH32_update) +# define XXH32_digest XXH_NAME2(XXH_NAMESPACE, XXH32_digest) +# define XXH32_copyState XXH_NAME2(XXH_NAMESPACE, XXH32_copyState) +# define XXH32_canonicalFromHash XXH_NAME2(XXH_NAMESPACE, XXH32_canonicalFromHash) +# define XXH32_hashFromCanonical XXH_NAME2(XXH_NAMESPACE, XXH32_hashFromCanonical) +# define XXH64 XXH_NAME2(XXH_NAMESPACE, XXH64) +# define XXH64_createState XXH_NAME2(XXH_NAMESPACE, XXH64_createState) +# define XXH64_freeState XXH_NAME2(XXH_NAMESPACE, XXH64_freeState) +# define XXH64_reset XXH_NAME2(XXH_NAMESPACE, XXH64_reset) +# define XXH64_update XXH_NAME2(XXH_NAMESPACE, XXH64_update) +# define XXH64_digest XXH_NAME2(XXH_NAMESPACE, XXH64_digest) +# define XXH64_copyState XXH_NAME2(XXH_NAMESPACE, XXH64_copyState) +# define XXH64_canonicalFromHash XXH_NAME2(XXH_NAMESPACE, XXH64_canonicalFromHash) +# define XXH64_hashFromCanonical XXH_NAME2(XXH_NAMESPACE, XXH64_hashFromCanonical) +#endif + + +/* ************************************* +* Version +***************************************/ +#define XXH_VERSION_MAJOR 0 +#define XXH_VERSION_MINOR 6 +#define XXH_VERSION_RELEASE 5 +#define XXH_VERSION_NUMBER (XXH_VERSION_MAJOR *100*100 + XXH_VERSION_MINOR *100 + XXH_VERSION_RELEASE) +XXH_PUBLIC_API unsigned XXH_versionNumber (void); + + +/*-********************************************************************** +* 32-bit hash +************************************************************************/ +typedef unsigned int XXH32_hash_t; + +/*! XXH32() : + Calculate the 32-bit hash of sequence "length" bytes stored at memory address "input". + The memory between input & input+length must be valid (allocated and read-accessible). + "seed" can be used to alter the result predictably. + Speed on Core 2 Duo @ 3 GHz (single thread, SMHasher benchmark) : 5.4 GB/s */ +XXH_PUBLIC_API XXH32_hash_t XXH32 (const void* input, size_t length, unsigned int seed); + +/*====== Streaming ======*/ +typedef struct XXH32_state_s XXH32_state_t; /* incomplete type */ +XXH_PUBLIC_API XXH32_state_t* XXH32_createState(void); +XXH_PUBLIC_API XXH_errorcode XXH32_freeState(XXH32_state_t* statePtr); +XXH_PUBLIC_API void XXH32_copyState(XXH32_state_t* dst_state, const XXH32_state_t* src_state); + +XXH_PUBLIC_API XXH_errorcode XXH32_reset (XXH32_state_t* statePtr, unsigned int seed); +XXH_PUBLIC_API XXH_errorcode XXH32_update (XXH32_state_t* statePtr, const void* input, size_t length); +XXH_PUBLIC_API XXH32_hash_t XXH32_digest (const XXH32_state_t* statePtr); + +/* + * Streaming functions generate the xxHash of an input provided in multiple segments. + * Note that, for small input, they are slower than single-call functions, due to state management. + * For small inputs, prefer `XXH32()` and `XXH64()`, which are better optimized. + * + * XXH state must first be allocated, using XXH*_createState() . + * + * Start a new hash by initializing state with a seed, using XXH*_reset(). + * + * Then, feed the hash state by calling XXH*_update() as many times as necessary. + * The function returns an error code, with 0 meaning OK, and any other value meaning there is an error. + * + * Finally, a hash value can be produced anytime, by using XXH*_digest(). + * This function returns the nn-bits hash as an int or long long. + * + * It's still possible to continue inserting input into the hash state after a digest, + * and generate some new hashes later on, by calling again XXH*_digest(). + * + * When done, free XXH state space if it was allocated dynamically. + */ + +/*====== Canonical representation ======*/ + +typedef struct { unsigned char digest[4]; } XXH32_canonical_t; +XXH_PUBLIC_API void XXH32_canonicalFromHash(XXH32_canonical_t* dst, XXH32_hash_t hash); +XXH_PUBLIC_API XXH32_hash_t XXH32_hashFromCanonical(const XXH32_canonical_t* src); + +/* Default result type for XXH functions are primitive unsigned 32 and 64 bits. + * The canonical representation uses human-readable write convention, aka big-endian (large digits first). + * These functions allow transformation of hash result into and from its canonical format. + * This way, hash values can be written into a file / memory, and remain comparable on different systems and programs. + */ + + +#ifndef XXH_NO_LONG_LONG +/*-********************************************************************** +* 64-bit hash +************************************************************************/ +typedef unsigned long long XXH64_hash_t; + +/*! XXH64() : + Calculate the 64-bit hash of sequence of length "len" stored at memory address "input". + "seed" can be used to alter the result predictably. + This function runs faster on 64-bit systems, but slower on 32-bit systems (see benchmark). +*/ +XXH_PUBLIC_API XXH64_hash_t XXH64 (const void* input, size_t length, unsigned long long seed); + +/*====== Streaming ======*/ +typedef struct XXH64_state_s XXH64_state_t; /* incomplete type */ +XXH_PUBLIC_API XXH64_state_t* XXH64_createState(void); +XXH_PUBLIC_API XXH_errorcode XXH64_freeState(XXH64_state_t* statePtr); +XXH_PUBLIC_API void XXH64_copyState(XXH64_state_t* dst_state, const XXH64_state_t* src_state); + +XXH_PUBLIC_API XXH_errorcode XXH64_reset (XXH64_state_t* statePtr, unsigned long long seed); +XXH_PUBLIC_API XXH_errorcode XXH64_update (XXH64_state_t* statePtr, const void* input, size_t length); +XXH_PUBLIC_API XXH64_hash_t XXH64_digest (const XXH64_state_t* statePtr); + +/*====== Canonical representation ======*/ +typedef struct { unsigned char digest[8]; } XXH64_canonical_t; +XXH_PUBLIC_API void XXH64_canonicalFromHash(XXH64_canonical_t* dst, XXH64_hash_t hash); +XXH_PUBLIC_API XXH64_hash_t XXH64_hashFromCanonical(const XXH64_canonical_t* src); +#endif /* XXH_NO_LONG_LONG */ + + + +#ifdef XXH_STATIC_LINKING_ONLY + +/* ================================================================================================ + This section contains declarations which are not guaranteed to remain stable. + They may change in future versions, becoming incompatible with a different version of the library. + These declarations should only be used with static linking. + Never use them in association with dynamic linking ! +=================================================================================================== */ + +/* These definitions are only present to allow + * static allocation of XXH state, on stack or in a struct for example. + * Never **ever** use members directly. */ + +#if !defined (__VMS) \ + && (defined (__cplusplus) \ + || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) ) +# include + +struct XXH32_state_s { + uint32_t total_len_32; + uint32_t large_len; + uint32_t v1; + uint32_t v2; + uint32_t v3; + uint32_t v4; + uint32_t mem32[4]; + uint32_t memsize; + uint32_t reserved; /* never read nor write, might be removed in a future version */ +}; /* typedef'd to XXH32_state_t */ + +struct XXH64_state_s { + uint64_t total_len; + uint64_t v1; + uint64_t v2; + uint64_t v3; + uint64_t v4; + uint64_t mem64[4]; + uint32_t memsize; + uint32_t reserved[2]; /* never read nor write, might be removed in a future version */ +}; /* typedef'd to XXH64_state_t */ + +# else + +struct XXH32_state_s { + unsigned total_len_32; + unsigned large_len; + unsigned v1; + unsigned v2; + unsigned v3; + unsigned v4; + unsigned mem32[4]; + unsigned memsize; + unsigned reserved; /* never read nor write, might be removed in a future version */ +}; /* typedef'd to XXH32_state_t */ + +# ifndef XXH_NO_LONG_LONG /* remove 64-bit support */ +struct XXH64_state_s { + unsigned long long total_len; + unsigned long long v1; + unsigned long long v2; + unsigned long long v3; + unsigned long long v4; + unsigned long long mem64[4]; + unsigned memsize; + unsigned reserved[2]; /* never read nor write, might be removed in a future version */ +}; /* typedef'd to XXH64_state_t */ +# endif + +# endif + + +#if defined(XXH_INLINE_ALL) || defined(XXH_PRIVATE_API) +# include "xxhash.c" /* include xxhash function bodies as `static`, for inlining */ +#endif + +#endif /* XXH_STATIC_LINKING_ONLY */ + + +#if defined (__cplusplus) +} +#endif + +#endif /* XXHASH_H_5627135585666179 */ diff --git a/example/android/third_party/lz4/x86-64/liblz4.a b/example/android/third_party/lz4/x86-64/liblz4.a new file mode 100644 index 00000000..d4764fe8 Binary files /dev/null and b/example/android/third_party/lz4/x86-64/liblz4.a differ diff --git a/example/android/third_party/lz4/x86-64/liblz4.so b/example/android/third_party/lz4/x86-64/liblz4.so new file mode 100644 index 00000000..91290e20 Binary files /dev/null and b/example/android/third_party/lz4/x86-64/liblz4.so differ diff --git a/lite-client/lite-client.cpp b/lite-client/lite-client.cpp index cb0a7dec..55d46ad1 100644 --- a/lite-client/lite-client.cpp +++ b/lite-client/lite-client.cpp @@ -973,6 +973,7 @@ bool TestNode::show_help(std::string command) { "savecomplaints \tSaves all complaints registered for specified validator set id " "into files .boc\n" "complaintprice \tComputes the price (in nanograms) for creating a complaint\n" + "msgqueuesizes\tShows current sizes of outbound message queues in all shards\n" "known\tShows the list of all known block ids\n" "knowncells\tShows the list of hashes of all known (cached) cells\n" "dumpcell \nDumps a cached cell by a prefix of its hash\n" @@ -1108,6 +1109,8 @@ bool TestNode::do_parse_line() { std::string filename; return parse_uint32(expire_in) && get_word_to(filename) && seekeoln() && set_error(get_complaint_price(expire_in, filename)); + } else if (word == "msgqueuesizes") { + return get_msg_queue_sizes(); } else if (word == "known") { return eoln() && show_new_blkids(true); } else if (word == "knowncells") { @@ -1611,6 +1614,30 @@ void TestNode::send_compute_complaint_price_query(ton::StdSmcAddress elector_add std::move(P)); } +bool TestNode::get_msg_queue_sizes() { + auto q = ton::serialize_tl_object(ton::create_tl_object(0, 0, 0), true); + return envelope_send_query(std::move(q), [Self = actor_id(this)](td::Result res) -> void { + if (res.is_error()) { + LOG(ERROR) << "liteServer.getOutMsgQueueSizes error: " << res.move_as_error(); + return; + } + auto F = ton::fetch_tl_object(res.move_as_ok(), true); + if (F.is_error()) { + LOG(ERROR) << "cannot parse answer to liteServer.getOutMsgQueueSizes"; + return; + } + td::actor::send_closure_later(Self, &TestNode::got_msg_queue_sizes, F.move_as_ok()); + }); +} + +void TestNode::got_msg_queue_sizes(ton::tl_object_ptr f) { + td::TerminalIO::out() << "Outbound message queue sizes:" << std::endl; + for (auto &x : f->shards_) { + td::TerminalIO::out() << ton::create_block_id(x->id_).id.to_str() << " " << x->size_ << std::endl; + } + td::TerminalIO::out() << "External message queue size limit: " << f->ext_msg_queue_size_limit_ << std::endl; +} + bool TestNode::dns_resolve_start(ton::WorkchainId workchain, ton::StdSmcAddress addr, ton::BlockIdExt blkid, std::string domain, td::Bits256 cat, int mode) { if (domain.size() >= 2 && domain[0] == '"' && domain.back() == '"') { diff --git a/lite-client/lite-client.h b/lite-client/lite-client.h index 2602a0a7..219ba7d5 100644 --- a/lite-client/lite-client.h +++ b/lite-client/lite-client.h @@ -35,6 +35,7 @@ #include "block/block.h" #include "block/mc-config.h" #include "td/utils/filesystem.h" +#include "auto/tl/lite_api.h" using td::Ref; @@ -302,6 +303,8 @@ class TestNode : public td::actor::Actor { td::Bits256 chash = td::Bits256::zero(), std::string filename = ""); void send_compute_complaint_price_query(ton::StdSmcAddress elector_addr, unsigned expires_in, unsigned bits, unsigned refs, td::Bits256 chash, std::string filename); + bool get_msg_queue_sizes(); + void got_msg_queue_sizes(ton::tl_object_ptr f); bool cache_cell(Ref cell); bool list_cached_cells() const; bool dump_cached_cell(td::Slice hash_pfx, td::Slice type_name = {}); diff --git a/overlay/overlay-manager.cpp b/overlay/overlay-manager.cpp index 3c5f5eab..b9eb95b9 100644 --- a/overlay/overlay-manager.cpp +++ b/overlay/overlay-manager.cpp @@ -108,11 +108,12 @@ void OverlayManager::create_public_overlay_ex(adnl::AdnlNodeIdShort local_id, Ov void OverlayManager::create_private_overlay(adnl::AdnlNodeIdShort local_id, OverlayIdFull overlay_id, std::vector nodes, - std::unique_ptr callback, OverlayPrivacyRules rules) { + std::unique_ptr callback, OverlayPrivacyRules rules, + std::string scope) { auto id = overlay_id.compute_short_id(); register_overlay(local_id, id, Overlay::create(keyring_, adnl_, actor_id(this), dht_node_, local_id, std::move(overlay_id), - std::move(nodes), std::move(callback), std::move(rules))); + std::move(nodes), std::move(callback), std::move(rules), std::move(scope))); } void OverlayManager::receive_message(adnl::AdnlNodeIdShort src, adnl::AdnlNodeIdShort dst, td::BufferSlice data) { diff --git a/overlay/overlay-manager.h b/overlay/overlay-manager.h index 035ef3e8..1b9c75a4 100644 --- a/overlay/overlay-manager.h +++ b/overlay/overlay-manager.h @@ -57,7 +57,7 @@ class OverlayManager : public Overlays { OverlayOptions opts) override; void create_private_overlay(adnl::AdnlNodeIdShort local_id, OverlayIdFull overlay_id, std::vector nodes, std::unique_ptr callback, - OverlayPrivacyRules rules) override; + OverlayPrivacyRules rules, std::string scope) override; void delete_overlay(adnl::AdnlNodeIdShort local_id, OverlayIdShort overlay_id) override; void send_query(adnl::AdnlNodeIdShort dst, adnl::AdnlNodeIdShort src, OverlayIdShort overlay_id, std::string name, td::Promise promise, td::Timestamp timeout, td::BufferSlice query) override { diff --git a/overlay/overlay.cpp b/overlay/overlay.cpp index af01e045..f964ccc5 100644 --- a/overlay/overlay.cpp +++ b/overlay/overlay.cpp @@ -49,10 +49,11 @@ td::actor::ActorOwn Overlay::create(td::actor::ActorId manager, td::actor::ActorId dht_node, adnl::AdnlNodeIdShort local_id, OverlayIdFull overlay_id, std::vector nodes, - std::unique_ptr callback, OverlayPrivacyRules rules) { - auto R = - td::actor::create_actor("overlay", keyring, adnl, manager, dht_node, local_id, std::move(overlay_id), - false, std::move(nodes), std::move(callback), std::move(rules)); + std::unique_ptr callback, OverlayPrivacyRules rules, + std::string scope) { + auto R = td::actor::create_actor("overlay", keyring, adnl, manager, dht_node, local_id, + std::move(overlay_id), false, std::move(nodes), std::move(callback), + std::move(rules), std::move(scope)); return td::actor::ActorOwn(std::move(R)); } diff --git a/overlay/overlay.h b/overlay/overlay.h index da41a247..c1fe9643 100644 --- a/overlay/overlay.h +++ b/overlay/overlay.h @@ -48,7 +48,8 @@ class Overlay : public td::actor::Actor { td::actor::ActorId manager, td::actor::ActorId dht_node, adnl::AdnlNodeIdShort local_id, OverlayIdFull overlay_id, std::vector nodes, - std::unique_ptr callback, OverlayPrivacyRules rules); + std::unique_ptr callback, OverlayPrivacyRules rules, + std::string scope); virtual void update_dht_node(td::actor::ActorId dht) = 0; diff --git a/overlay/overlays.h b/overlay/overlays.h index 79551e05..6bf5852f 100644 --- a/overlay/overlays.h +++ b/overlay/overlays.h @@ -205,7 +205,7 @@ class Overlays : public td::actor::Actor { td::string scope, OverlayOptions opts) = 0; virtual void create_private_overlay(adnl::AdnlNodeIdShort local_id, OverlayIdFull overlay_id, std::vector nodes, std::unique_ptr callback, - OverlayPrivacyRules rules) = 0; + OverlayPrivacyRules rules, std::string scope) = 0; virtual void delete_overlay(adnl::AdnlNodeIdShort local_id, OverlayIdShort overlay_id) = 0; virtual void send_query(adnl::AdnlNodeIdShort dst, adnl::AdnlNodeIdShort src, OverlayIdShort overlay_id, diff --git a/recent_changelog.md b/recent_changelog.md index e741a328..dac0c651 100644 --- a/recent_changelog.md +++ b/recent_changelog.md @@ -1,6 +1,15 @@ -## 2024.03 Update +## 2024.04 Update -1. Preparatory (not enabled yet) code for pre-compiled smart-contract. -2. Minor fixes for fee-related opcodes. +1. Emulator: Single call optimized runGetMethod added +2. Tonlib: a series of proof improvements, also breaking Change in `liteServer.getAllShardsInfo` method (see below) +3. DB: usage statistics now collected, outdated persistent states are not serialized +4. LS: fast `getOutMsgQueueSizes` added, preliminary support of non-final block requests +5. Network: lz4 compression of block candidates (disabled by default). + + + +--- + +* `liteServer.getAllShardsInfo` method was updated for better efficiency. Previously, field proof contained BoC with two roots: one for BlockState from block's root and another for ShardHashes from BlockState. Now, it returns a single-root proof BoC, specifically the merkle proof of ShardHashes directly from the block's root, streamlining data access and integrity. Checking of the proof requires to check that ShardHashes in the `data` correspond to ShardHashes from the block. diff --git a/storage/Torrent.cpp b/storage/Torrent.cpp index 6807519c..545b6d10 100644 --- a/storage/Torrent.cpp +++ b/storage/Torrent.cpp @@ -324,7 +324,7 @@ td::Status Torrent::add_piece(td::uint64 piece_i, td::Slice data, td::Ref Torrent::read_file(td::Slice name) { @@ -488,7 +488,7 @@ Torrent::Torrent(Info info, td::optional header, ton::MerkleTree , info_(info) , root_dir_(std::move(root_dir)) , header_(std::move(header)) - , enabled_wirte_to_files_(true) + , enabled_write_to_files_(true) , merkle_tree_(std::move(tree)) , piece_is_ready_(info_.pieces_count(), true) , ready_parts_count_{info_.pieces_count()} @@ -586,7 +586,7 @@ void Torrent::set_file_excluded(size_t i, bool excluded) { included_ready_size_ += chunk.ready_size; } chunk.excluded = excluded; - if (!enabled_wirte_to_files_ || excluded) { + if (!enabled_write_to_files_ || excluded) { return; } auto range = get_file_parts_range(i); diff --git a/storage/Torrent.h b/storage/Torrent.h index 6401e405..1ae30fc2 100644 --- a/storage/Torrent.h +++ b/storage/Torrent.h @@ -173,7 +173,7 @@ class Torrent { size_t not_ready_pending_piece_count_{0}; size_t header_pieces_count_{0}; std::map pending_pieces_; - bool enabled_wirte_to_files_ = false; + bool enabled_write_to_files_ = false; struct InMemoryPiece { std::string data; std::set pending_chunks; diff --git a/tddb/td/db/RocksDb.cpp b/tddb/td/db/RocksDb.cpp index 500985e2..a84a804b 100644 --- a/tddb/td/db/RocksDb.cpp +++ b/tddb/td/db/RocksDb.cpp @@ -59,9 +59,8 @@ RocksDb RocksDb::clone() const { return RocksDb{db_, statistics_}; } -Result RocksDb::open(std::string path) { +Result RocksDb::open(std::string path, std::shared_ptr statistics) { rocksdb::OptimisticTransactionDB *db; - auto statistics = rocksdb::CreateDBStatistics(); { rocksdb::Options options; @@ -94,6 +93,18 @@ Result RocksDb::open(std::string path) { return RocksDb(std::shared_ptr(db), std::move(statistics)); } +std::shared_ptr RocksDb::create_statistics() { + return rocksdb::CreateDBStatistics(); +} + +std::string RocksDb::statistics_to_string(const std::shared_ptr statistics) { + return statistics->ToString(); +} + +void RocksDb::reset_statistics(const std::shared_ptr statistics) { + statistics->Reset(); +} + std::unique_ptr RocksDb::snapshot() { auto res = std::make_unique(clone()); res->begin_snapshot().ensure(); diff --git a/tddb/td/db/RocksDb.h b/tddb/td/db/RocksDb.h index b8bfaf9d..1afba4cc 100644 --- a/tddb/td/db/RocksDb.h +++ b/tddb/td/db/RocksDb.h @@ -38,7 +38,7 @@ class RocksDb : public KeyValue { public: static Status destroy(Slice path); RocksDb clone() const; - static Result open(std::string path); + static Result open(std::string path, std::shared_ptr statistics = nullptr); Result get(Slice key, std::string &value) override; Status set(Slice key, Slice value) override; @@ -60,6 +60,10 @@ class RocksDb : public KeyValue { std::unique_ptr snapshot() override; std::string stats() const override; + static std::shared_ptr create_statistics(); + static std::string statistics_to_string(const std::shared_ptr statistics); + static void reset_statistics(const std::shared_ptr statistics); + RocksDb(RocksDb &&); RocksDb &operator=(RocksDb &&); ~RocksDb(); diff --git a/tdutils/CMakeLists.txt b/tdutils/CMakeLists.txt index f1e4b1ea..4c713247 100644 --- a/tdutils/CMakeLists.txt +++ b/tdutils/CMakeLists.txt @@ -280,6 +280,19 @@ if (TDUTILS_MIME_TYPE) ) endif() +if (NOT LZ4_FOUND) + pkg_check_modules(LZ4 REQUIRED liblz4) +endif() + +if (LZ4_FOUND) + set(TD_HAVE_LZ4 1) + set(TDUTILS_SOURCE + ${TDUTILS_SOURCE} + td/utils/lz4.cpp + td/utils/lz4.h + ) +endif() + set(TDUTILS_TEST_SOURCE ${CMAKE_CURRENT_SOURCE_DIR}/test/buffer.cpp ${CMAKE_CURRENT_SOURCE_DIR}/test/ConcurrentHashMap.cpp @@ -338,6 +351,13 @@ endif() if (CRC32C_FOUND) target_link_libraries(tdutils PRIVATE crc32c) endif() + +if (LZ4_FOUND) + message(STATUS "Found LZ4 ${LZ4_LIBRARIES} ${LZ4_INCLUDE_DIRS}") + target_link_libraries(tdutils PRIVATE ${LZ4_LIBRARIES}) + target_include_directories(tdutils SYSTEM PRIVATE ${LZ4_INCLUDE_DIRS}) +endif() + if (ABSL_FOUND) target_link_libraries_system(tdutils absl::flat_hash_map absl::flat_hash_set absl::hash) endif() diff --git a/tdutils/td/utils/config.h.in b/tdutils/td/utils/config.h.in index f8b89aeb..3f4e1bf2 100644 --- a/tdutils/td/utils/config.h.in +++ b/tdutils/td/utils/config.h.in @@ -3,6 +3,7 @@ #cmakedefine01 TD_HAVE_OPENSSL #cmakedefine01 TD_HAVE_ZLIB #cmakedefine01 TD_HAVE_CRC32C +#cmakedefine01 TD_HAVE_LZ4 #cmakedefine01 TD_HAVE_COROUTINES #cmakedefine01 TD_HAVE_ABSL #cmakedefine01 TD_FD_DEBUG diff --git a/tdutils/td/utils/lz4.cpp b/tdutils/td/utils/lz4.cpp new file mode 100644 index 00000000..ebf456aa --- /dev/null +++ b/tdutils/td/utils/lz4.cpp @@ -0,0 +1,48 @@ +/* + This file is part of TON Blockchain Library. + + TON Blockchain Library is free software: you can redistribute it and/or modify + it under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation, either version 2 of the License, or + (at your option) any later version. + + TON Blockchain Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with TON Blockchain Library. If not, see . +*/ +#include "td/utils/buffer.h" +#include "td/utils/misc.h" +#include + +namespace td { + +td::BufferSlice lz4_compress(td::Slice data) { + int size = narrow_cast(data.size()); + int buf_size = LZ4_compressBound(size); + td::BufferSlice compressed(buf_size); + int compressed_size = LZ4_compress_default(data.data(), compressed.data(), size, buf_size); + CHECK(compressed_size > 0); + return td::BufferSlice{compressed.as_slice().substr(0, compressed_size)}; +} + +td::Result lz4_decompress(td::Slice data, int max_decompressed_size) { + TRY_RESULT(size, narrow_cast_safe(data.size())); + if (max_decompressed_size < 0) { + return td::Status::Error("invalid max_decompressed_size"); + } + td::BufferSlice decompressed(max_decompressed_size); + int result = LZ4_decompress_safe(data.data(), decompressed.data(), size, max_decompressed_size); + if (result < 0) { + return td::Status::Error(PSTRING() << "lz4 decompression failed, error code: " << result); + } + if (result == max_decompressed_size) { + return decompressed; + } + return td::BufferSlice{decompressed.as_slice().substr(0, result)}; +} + +} // namespace td diff --git a/tonlib/tonlib/ClientActor.cpp b/tdutils/td/utils/lz4.h similarity index 75% rename from tonlib/tonlib/ClientActor.cpp rename to tdutils/td/utils/lz4.h index ab2bc125..fbbc470f 100644 --- a/tonlib/tonlib/ClientActor.cpp +++ b/tdutils/td/utils/lz4.h @@ -13,6 +13,15 @@ You should have received a copy of the GNU Lesser General Public License along with TON Blockchain Library. If not, see . - - Copyright 2017-2020 Telegram Systems LLP */ +#pragma once + +#include "td/utils/buffer.h" +#include "td/utils/Status.h" + +namespace td { + +td::BufferSlice lz4_compress(td::Slice data); +td::Result lz4_decompress(td::Slice data, int max_decompressed_size); + +} // namespace td diff --git a/tdutils/td/utils/misc.cpp b/tdutils/td/utils/misc.cpp index d592593a..caff44e3 100644 --- a/tdutils/td/utils/misc.cpp +++ b/tdutils/td/utils/misc.cpp @@ -153,9 +153,9 @@ string buffer_to_hex(Slice buffer) { const char *hex = "0123456789ABCDEF"; string res(2 * buffer.size(), '\0'); for (std::size_t i = 0; i < buffer.size(); i++) { - auto c = buffer.ubegin()[i]; - res[2 * i] = hex[c & 15]; - res[2 * i + 1] = hex[c >> 4]; + unsigned char c = buffer[i]; + res[2 * i] = hex[c >> 4]; + res[2 * i + 1] = hex[c & 15]; } return res; } diff --git a/test/regression-tests.ans b/test/regression-tests.ans index 96117ee9..bfea7089 100644 --- a/test/regression-tests.ans +++ b/test/regression-tests.ans @@ -1,7 +1,7 @@ abce Test_Bigint_main_default 76f38492ec19464a1d0eac51d389023a31ce10396b3894061361d159567ce8cd Test_Bitstrings_main_default a8b08af3116923c4c2a14e138d168375abd0c059f2f780d3267b294929a1110e -Test_Cells_simple_default 832502642fe4fe5db70de82681aedb7d54d7f3530e0069861fff405fe6f6cf23 +Test_Cells_simple_default 414f68da0a2f6fa09b2bdb99c453cdf919db48b5b4ca1c6ac1dfcd837e2d8170 Test_Fift_bug_div_default 1ac42861ce96b2896001c587f65e9afe1617db48859f19c2f4e3063a20ea60b0 Test_Fift_bug_newlize_default e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855 Test_Fift_bug_ufits_default 51bf5a9f1ed7633a193f6fdd17a7a3af8e032dfe72a9669c85e8639aa8a7c195 diff --git a/test/test-ton-collator.cpp b/test/test-ton-collator.cpp index b76dd5eb..286a5fea 100644 --- a/test/test-ton-collator.cpp +++ b/test/test-ton-collator.cpp @@ -300,7 +300,7 @@ class TestNode : public td::actor::Actor { shard_top_block_id_, db_root_); for (auto &msg : ext_msgs_) { td::actor::send_closure(validator_manager_, &ton::validator::ValidatorManager::new_external_message, - std::move(msg)); + std::move(msg), 0); } for (auto &topmsg : top_shard_descrs_) { td::actor::send_closure(validator_manager_, &ton::validator::ValidatorManager::new_shard_block, ton::BlockIdExt{}, diff --git a/tl-utils/lite-utils.cpp b/tl-utils/lite-utils.cpp index 427852cb..9ea7756a 100644 --- a/tl-utils/lite-utils.cpp +++ b/tl-utils/lite-utils.cpp @@ -148,6 +148,7 @@ std::string lite_query_name_by_id(int id) { {lite_api::liteServer_getOneTransaction::ID, "getOneTransaction"}, {lite_api::liteServer_getTransactions::ID, "getTransactions"}, {lite_api::liteServer_lookupBlock::ID, "lookupBlock"}, + {lite_api::liteServer_lookupBlockWithProof::ID, "lookupBlockWithProof"}, {lite_api::liteServer_listBlockTransactions::ID, "listBlockTransactions"}, {lite_api::liteServer_listBlockTransactionsExt::ID, "listBlockTransactionsExt"}, {lite_api::liteServer_getBlockProof::ID, "getBlockProof"}, @@ -155,7 +156,11 @@ std::string lite_query_name_by_id(int id) { {lite_api::liteServer_getConfigParams::ID, "getConfigParams"}, {lite_api::liteServer_getValidatorStats::ID, "getValidatorStats"}, {lite_api::liteServer_getLibraries::ID, "getLibraries"}, - {lite_api::liteServer_getShardBlockProof::ID, "getShardBlockProof"}}; + {lite_api::liteServer_getLibrariesWithProof::ID, "getLibrariesWithProof"}, + {lite_api::liteServer_getShardBlockProof::ID, "getShardBlockProof"}, + {lite_api::liteServer_getOutMsgQueueSizes::ID, "getOutMsgQueueSizes"}, + {lite_api::liteServer_nonfinal_getCandidate::ID, "nonfinal.getCandidate"}, + {lite_api::liteServer_nonfinal_getValidatorGroups::ID, "nonfinal.getValidatorGroups"}}; auto it = names.find(id); if (it == names.end()) { return "unknown"; diff --git a/tl/generate/scheme/lite_api.tl b/tl/generate/scheme/lite_api.tl index ce4d2731..d74e8953 100644 --- a/tl/generate/scheme/lite_api.tl +++ b/tl/generate/scheme/lite_api.tl @@ -53,11 +53,21 @@ liteServer.partialBlockProof complete:Bool from:tonNode.blockIdExt to:tonNode.bl liteServer.configInfo mode:# id:tonNode.blockIdExt state_proof:bytes config_proof:bytes = liteServer.ConfigInfo; liteServer.validatorStats mode:# id:tonNode.blockIdExt count:int complete:Bool state_proof:bytes data_proof:bytes = liteServer.ValidatorStats; liteServer.libraryResult result:(vector liteServer.libraryEntry) = liteServer.LibraryResult; +liteServer.libraryResultWithProof id:tonNode.blockIdExt mode:# result:(vector liteServer.libraryEntry) state_proof:bytes data_proof:bytes = liteServer.LibraryResultWithProof; liteServer.shardBlockLink id:tonNode.blockIdExt proof:bytes = liteServer.ShardBlockLink; liteServer.shardBlockProof masterchain_id:tonNode.blockIdExt links:(vector liteServer.shardBlockLink) = liteServer.ShardBlockProof; +liteServer.lookupBlockResult id:tonNode.blockIdExt mode:# mc_block_id:tonNode.blockIdExt client_mc_state_proof:bytes mc_block_proof:bytes shard_links:(vector liteServer.shardBlockLink) header:bytes prev_header:bytes = liteServer.LookupBlockResult; +liteServer.outMsgQueueSize id:tonNode.blockIdExt size:int = liteServer.OutMsgQueueSize; +liteServer.outMsgQueueSizes shards:(vector liteServer.outMsgQueueSize) ext_msg_queue_size_limit:int = liteServer.OutMsgQueueSizes; liteServer.debug.verbosity value:int = liteServer.debug.Verbosity; +liteServer.nonfinal.candidateId block_id:tonNode.blockIdExt creator:int256 collated_data_hash:int256 = liteServer.nonfinal.CandidateId; +liteServer.nonfinal.candidate id:liteServer.nonfinal.candidateId data:bytes collated_data:bytes = liteServer.nonfinal.Candidate; +liteServer.nonfinal.candidateInfo id:liteServer.nonfinal.candidateId available:Bool approved_weight:long signed_weight:long total_weight:long = liteServer.nonfinal.CandidateInfo; +liteServer.nonfinal.validatorGroupInfo next_block_id:tonNode.blockId cc_seqno:int prev:(vector tonNode.blockIdExt) candidates:(vector liteServer.nonfinal.candidateInfo) = liteServer.nonfinal.ValidatorGroupInfo; +liteServer.nonfinal.validatorGroups groups:(vector liteServer.nonfinal.validatorGroupInfo) = liteServer.nonfinal.ValidatorGroups; + ---functions--- liteServer.getMasterchainInfo = liteServer.MasterchainInfo; @@ -76,6 +86,7 @@ liteServer.getAllShardsInfo id:tonNode.blockIdExt = liteServer.AllShardsInfo; liteServer.getOneTransaction id:tonNode.blockIdExt account:liteServer.accountId lt:long = liteServer.TransactionInfo; liteServer.getTransactions count:# account:liteServer.accountId lt:long hash:int256 = liteServer.TransactionList; liteServer.lookupBlock mode:# id:tonNode.blockId lt:mode.1?long utime:mode.2?int = liteServer.BlockHeader; +liteServer.lookupBlockWithProof mode:# id:tonNode.blockId mc_block_id:tonNode.blockIdExt lt:mode.1?long utime:mode.2?int = liteServer.LookupBlockResult; liteServer.listBlockTransactions id:tonNode.blockIdExt mode:# count:# after:mode.7?liteServer.transactionId3 reverse_order:mode.6?true want_proof:mode.5?true = liteServer.BlockTransactions; liteServer.listBlockTransactionsExt id:tonNode.blockIdExt mode:# count:# after:mode.7?liteServer.transactionId3 reverse_order:mode.6?true want_proof:mode.5?true = liteServer.BlockTransactionsExt; liteServer.getBlockProof mode:# known_block:tonNode.blockIdExt target_block:mode.0?tonNode.blockIdExt = liteServer.PartialBlockProof; @@ -83,7 +94,12 @@ liteServer.getConfigAll mode:# id:tonNode.blockIdExt = liteServer.ConfigInfo; liteServer.getConfigParams mode:# id:tonNode.blockIdExt param_list:(vector int) = liteServer.ConfigInfo; liteServer.getValidatorStats#091a58bc mode:# id:tonNode.blockIdExt limit:int start_after:mode.0?int256 modified_after:mode.2?int = liteServer.ValidatorStats; liteServer.getLibraries library_list:(vector int256) = liteServer.LibraryResult; +liteServer.getLibrariesWithProof id:tonNode.blockIdExt mode:# library_list:(vector int256) = liteServer.LibraryResultWithProof; liteServer.getShardBlockProof id:tonNode.blockIdExt = liteServer.ShardBlockProof; +liteServer.getOutMsgQueueSizes mode:# wc:mode.0?int shard:mode.0?long = liteServer.OutMsgQueueSizes; + +liteServer.nonfinal.getValidatorGroups mode:# wc:mode.0?int shard:mode.1?long = liteServer.nonfinal.ValidatorGroups; +liteServer.nonfinal.getCandidate id:liteServer.nonfinal.candidateId = liteServer.nonfinal.Candidate; liteServer.queryPrefix = Object; liteServer.query data:bytes = Object; diff --git a/tl/generate/scheme/lite_api.tlo b/tl/generate/scheme/lite_api.tlo index da64ac53..ccae66f3 100644 Binary files a/tl/generate/scheme/lite_api.tlo and b/tl/generate/scheme/lite_api.tlo differ diff --git a/tl/generate/scheme/ton_api.tl b/tl/generate/scheme/ton_api.tl index 0f94e7f1..c61bad85 100644 --- a/tl/generate/scheme/ton_api.tl +++ b/tl/generate/scheme/ton_api.tl @@ -309,6 +309,7 @@ validatorSession.candidateId src:int256 root_hash:int256 file_hash:int256 collat validatorSession.blockUpdate ts:long actions:(vector validatorSession.round.Message) state:int = validatorSession.BlockUpdate; validatorSession.candidate src:int256 round:int root_hash:int256 data:bytes collated_data:bytes = validatorSession.Candidate; +validatorSession.compressedCandidate flags:# src:int256 round:int root_hash:int256 decompressed_size:int data:bytes = validatorSession.Candidate; validatorSession.config catchain_idle_timeout:double catchain_max_deps:int round_candidates:int next_candidate_delay:double round_attempt_duration:int max_round_attempts:int max_block_size:int max_collated_data_size:int = validatorSession.Config; @@ -385,9 +386,13 @@ tonNode.externalMessage data:bytes = tonNode.ExternalMessage; tonNode.newShardBlock block:tonNode.blockIdExt cc_seqno:int data:bytes = tonNode.NewShardBlock; +tonNode.blockBroadcastCompressed.data signatures:(vector tonNode.blockSignature) proof_data:bytes = tonNode.blockBroadcaseCompressed.Data; + tonNode.blockBroadcast id:tonNode.blockIdExt catchain_seqno:int validator_set_hash:int signatures:(vector tonNode.blockSignature) proof:bytes data:bytes = tonNode.Broadcast; +tonNode.blockBroadcastCompressed id:tonNode.blockIdExt catchain_seqno:int validator_set_hash:int + flags:# compressed:bytes = tonNode.Broadcast; tonNode.ihrMessageBroadcast message:tonNode.ihrMessage = tonNode.Broadcast; tonNode.externalMessageBroadcast message:tonNode.externalMessage = tonNode.Broadcast; tonNode.newShardBlockBroadcast block:tonNode.newShardBlock = tonNode.Broadcast; @@ -395,15 +400,15 @@ tonNode.newShardBlockBroadcast block:tonNode.newShardBlock = tonNode.Broadcast; tonNode.shardPublicOverlayId workchain:int shard:long zero_state_file_hash:int256 = tonNode.ShardPublicOverlayId; tonNode.privateBlockOverlayId zero_state_file_hash:int256 nodes:(vector int256) = tonNode.PrivateBlockOverlayId; +tonNode.customOverlayId zero_state_file_hash:int256 name:string nodes:(vector int256) = tonNode.CustomOverlayId; tonNode.keyBlocks blocks:(vector tonNode.blockIdExt) incomplete:Bool error:Bool = tonNode.KeyBlocks; ton.blockId root_cell_hash:int256 file_hash:int256 = ton.BlockId; ton.blockIdApprove root_cell_hash:int256 file_hash:int256 = ton.BlockId; -tonNode.dataList data:(vector bytes) = tonNode.DataList; - tonNode.dataFull id:tonNode.blockIdExt proof:bytes block:bytes is_link:Bool = tonNode.DataFull; +tonNode.dataFullCompressed id:tonNode.blockIdExt flags:# compressed:bytes is_link:Bool = tonNode.DataFull; tonNode.dataFullEmpty = tonNode.DataFull; tonNode.capabilities version:int capabilities:long = tonNode.Capabilities; @@ -430,18 +435,13 @@ tonNode.getNextKeyBlockIds block:tonNode.blockIdExt max_size:int = tonNode.KeyBl tonNode.downloadNextBlockFull prev_block:tonNode.blockIdExt = tonNode.DataFull; tonNode.downloadBlockFull block:tonNode.blockIdExt = tonNode.DataFull; tonNode.downloadBlock block:tonNode.blockIdExt = tonNode.Data; -tonNode.downloadBlocks blocks:(vector tonNode.blockIdExt) = tonNode.DataList; tonNode.downloadPersistentState block:tonNode.blockIdExt masterchain_block:tonNode.blockIdExt = tonNode.Data; tonNode.downloadPersistentStateSlice block:tonNode.blockIdExt masterchain_block:tonNode.blockIdExt offset:long max_size:long = tonNode.Data; tonNode.downloadZeroState block:tonNode.blockIdExt = tonNode.Data; tonNode.downloadBlockProof block:tonNode.blockIdExt = tonNode.Data; tonNode.downloadKeyBlockProof block:tonNode.blockIdExt = tonNode.Data; -tonNode.downloadBlockProofs blocks:(vector tonNode.blockIdExt) = tonNode.DataList; -tonNode.downloadKeyBlockProofs blocks:(vector tonNode.blockIdExt) = tonNode.DataList; tonNode.downloadBlockProofLink block:tonNode.blockIdExt = tonNode.Data; tonNode.downloadKeyBlockProofLink block:tonNode.blockIdExt = tonNode.Data; -tonNode.downloadBlockProofLinks blocks:(vector tonNode.blockIdExt) = tonNode.DataList; -tonNode.downloadKeyBlockProofLinks blocks:(vector tonNode.blockIdExt) = tonNode.DataList; tonNode.getArchiveInfo masterchain_seqno:int = tonNode.ArchiveInfo; tonNode.getArchiveSlice archive_id:long offset:long max_size:int = tonNode.Data; @@ -595,6 +595,10 @@ engine.validator.config out_port:int addrs:(vector engine.Addr) adnl:(vector eng liteservers:(vector engine.liteServer) control:(vector engine.controlInterface) gc:engine.gc = engine.validator.Config; +engine.validator.customOverlayNode adnl_id:int256 msg_sender:Bool msg_sender_priority:int = engine.validator.CustomOverlayNode; +engine.validator.customOverlay name:string nodes:(vector engine.validator.customOverlayNode) = engine.validator.CustomOverlay; +engine.validator.customOverlaysConfig overlays:(vector engine.validator.customOverlay) = engine.validator.CustomOverlaysConfig; + ---functions--- ---types--- @@ -698,6 +702,10 @@ engine.validator.getPerfTimerStats name:string = engine.validator.PerfTimerStats engine.validator.getShardOutQueueSize flags:# block_id:tonNode.blockId dest_wc:flags.0?int dest_shard:flags.0?long = engine.validator.ShardOutQueueSize; engine.validator.setExtMessagesBroadcastDisabled disabled:Bool = engine.validator.Success; +engine.validator.addCustomOverlay overlay:engine.validator.customOverlay = engine.validator.Success; +engine.validator.delCustomOverlay name:string = engine.validator.Success; +engine.validator.showCustomOverlays = engine.validator.CustomOverlaysConfig; + ---types--- storage.pong = storage.Pong; diff --git a/tl/generate/scheme/ton_api.tlo b/tl/generate/scheme/ton_api.tlo index e71fb0f7..b8f5d5a4 100644 Binary files a/tl/generate/scheme/ton_api.tlo and b/tl/generate/scheme/ton_api.tlo differ diff --git a/tonlib/test/offline.cpp b/tonlib/test/offline.cpp index 7b4342dd..b7423853 100644 --- a/tonlib/test/offline.cpp +++ b/tonlib/test/offline.cpp @@ -83,7 +83,7 @@ using namespace tonlib; TEST(Tonlib, PublicKey) { block::PublicKey::parse("pubjns2gp7DGCnEH7EOWeCnb6Lw1akm538YYaz6sdLVHfRB2").ensure_error(); auto key = block::PublicKey::parse("Pubjns2gp7DGCnEH7EOWeCnb6Lw1akm538YYaz6sdLVHfRB2").move_as_ok(); - CHECK(td::buffer_to_hex(key.key) == "3EE9DC0A7A0B6CA01770CE34698792BD8ECB53A6949BFD6C81B6E3CA475B74D7"); + CHECK(td::buffer_to_hex(key.key) == "E39ECDA0A7B0C60A7107EC43967829DBE8BC356A49B9DFC6186B3EAC74B5477D"); CHECK(key.serialize() == "Pubjns2gp7DGCnEH7EOWeCnb6Lw1akm538YYaz6sdLVHfRB2"); } diff --git a/tonlib/tonlib/CellString.cpp b/tonlib/tonlib/CellString.cpp deleted file mode 100644 index ad2cbf5f..00000000 --- a/tonlib/tonlib/CellString.cpp +++ /dev/null @@ -1,64 +0,0 @@ -#include "CellString.h" -#include "td/utils/misc.h" - -#include "vm/cells/CellSlice.h" - -namespace vm { -td::Status CellString::store(CellBuilder &cb, td::Slice slice, unsigned int top_bits) { - td::uint32 size = td::narrow_cast(slice.size() * 8); - return store(cb, td::BitSlice(slice.ubegin(), size), top_bits); -} - -td::Status CellString::store(CellBuilder &cb, td::BitSlice slice, unsigned int top_bits) { - if (slice.size() > max_bytes * 8) { - return td::Status::Error("String is too long (1)"); - } - unsigned int head = td::min(slice.size(), td::min(cb.remaining_bits(), top_bits)) / 8 * 8; - auto max_bits = vm::Cell::max_bits / 8 * 8; - auto depth = 1 + (slice.size() - head + max_bits - 1) / max_bits; - if (depth > max_chain_length) { - return td::Status::Error("String is too long (2)"); - } - cb.append_bitslice(slice.subslice(0, head)); - slice.advance(head); - if (slice.size() == 0) { - return td::Status::OK(); - } - CellBuilder child_cb; - store(child_cb, std::move(slice)); - cb.store_ref(child_cb.finalize()); - return td::Status::OK(); -} - -template -void CellString::for_each(F &&f, CellSlice &cs, unsigned int top_bits) { - unsigned int head = td::min(cs.size(), top_bits); - f(cs.prefetch_bits(head)); - if (!cs.have_refs()) { - return; - } - auto ref = cs.prefetch_ref(); - while (true) { - auto cs = vm::load_cell_slice(ref); - f(cs.prefetch_bits(cs.size())); - if (!cs.have_refs()) { - return; - } - ref = cs.prefetch_ref(); - } -} - -td::Result CellString::load(CellSlice &cs, unsigned int top_bits) { - unsigned int size = 0; - for_each([&](auto slice) { size += slice.size(); }, cs, top_bits); - if (size % 8 != 0) { - return td::Status::Error("Size is not divisible by 8"); - } - std::string res(size / 8, 0); - - td::BitPtr to(td::MutableSlice(res).ubegin()); - for_each([&](auto slice) { to.concat(slice); }, cs, top_bits); - CHECK(to.offs == (int)size); - return res; -} -} // namespace vm diff --git a/tonlib/tonlib/CellString.h b/tonlib/tonlib/CellString.h deleted file mode 100644 index 89c933d8..00000000 --- a/tonlib/tonlib/CellString.h +++ /dev/null @@ -1,22 +0,0 @@ -#pragma once - -#include "td/utils/Status.h" - -#include "vm/cells/CellBuilder.h" - -namespace vm { -class CellString { - public: - static constexpr unsigned int max_bytes = 1024; - static constexpr unsigned int max_chain_length = 16; - - static td::Status store(CellBuilder &cb, td::Slice slice, unsigned int top_bits = Cell::max_bits); - static td::Status store(CellBuilder &cb, td::BitSlice slice, unsigned int top_bits = Cell::max_bits); - static td::Result load(CellSlice &cs, unsigned int top_bits = Cell::max_bits); - - private: - template - static void for_each(F &&f, CellSlice &cs, unsigned int top_bits = Cell::max_bits); -}; - -} // namespace vm diff --git a/tonlib/tonlib/ClientActor.h b/tonlib/tonlib/ClientActor.h deleted file mode 100644 index 592589cd..00000000 --- a/tonlib/tonlib/ClientActor.h +++ /dev/null @@ -1,41 +0,0 @@ -/* - This file is part of TON Blockchain Library. - - TON Blockchain Library is free software: you can redistribute it and/or modify - it under the terms of the GNU Lesser General Public License as published by - the Free Software Foundation, either version 2 of the License, or - (at your option) any later version. - - TON Blockchain Library is distributed in the hope that it will be useful, - but WITHOUT ANY WARRANTY; without even the implied warranty of - MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - GNU Lesser General Public License for more details. - - You should have received a copy of the GNU Lesser General Public License - along with TON Blockchain Library. If not, see . - - Copyright 2017-2020 Telegram Systems LLP -*/ -#pragma once - -#include "tonlib/TonlibCallback.h" -#include "tonlib/TonlibClient.h" -#include "auto/tl/tonlib_api.h" - -namespace tonlinb { -class ClientActor : public td::actor::Actor { - public: - explicit ClientActor(td::unique_ptr callback); - void request(td::uint64 id, tonlib_api::object_ptr request); - static tonlib_api::object_ptr execute(tonlib_api::object_ptr request); - ~ClientActor(); - ClientActor(ClientActor&& other); - ClientActor& operator=(ClientActor&& other); - - ClientActor(const ClientActor& other) = delete; - ClientActor& operator=(const ClientActor& other) = delete; - - private: - td::actor::ActorOwn tonlib_; -}; -} // namespace tonlinb diff --git a/tonlib/tonlib/GenericAccount.cpp b/tonlib/tonlib/GenericAccount.cpp deleted file mode 100644 index 9f44ede2..00000000 --- a/tonlib/tonlib/GenericAccount.cpp +++ /dev/null @@ -1,80 +0,0 @@ -/* - This file is part of TON Blockchain Library. - - TON Blockchain Library is free software: you can redistribute it and/or modify - it under the terms of the GNU Lesser General Public License as published by - the Free Software Foundation, either version 2 of the License, or - (at your option) any later version. - - TON Blockchain Library is distributed in the hope that it will be useful, - but WITHOUT ANY WARRANTY; without even the implied warranty of - MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - GNU Lesser General Public License for more details. - - You should have received a copy of the GNU Lesser General Public License - along with TON Blockchain Library. If not, see . - - Copyright 2017-2019 Telegram Systems LLP -*/ -#include "tonlib/GenericAccount.h" -#include "tonlib/utils.h" -#include "block/block-auto.h" -namespace tonlib { -td::Ref GenericAccount::get_init_state(td::Ref code, td::Ref data) noexcept { - return vm::CellBuilder() - .append_cellslice(binary_bitstring_to_cellslice("b{00110}").move_as_ok()) - .store_ref(std::move(code)) - .store_ref(std::move(data)) - .finalize(); -} -block::StdAddress GenericAccount::get_address(ton::WorkchainId workchain_id, - const td::Ref& init_state) noexcept { - return block::StdAddress(workchain_id, init_state->get_hash().bits(), true /*bounce*/); -} -td::Ref GenericAccount::create_ext_message(const block::StdAddress& address, td::Ref new_state, - td::Ref body) noexcept { - block::gen::Message::Record message; - /*info*/ { - block::gen::CommonMsgInfo::Record_ext_in_msg_info info; - /* src */ - tlb::csr_pack(info.src, block::gen::MsgAddressExt::Record_addr_none{}); - /* dest */ { - block::gen::MsgAddressInt::Record_addr_std dest; - dest.anycast = vm::CellBuilder().store_zeroes(1).as_cellslice_ref(); - dest.workchain_id = address.workchain; - dest.address = address.addr; - - tlb::csr_pack(info.dest, dest); - } - /* import_fee */ { - vm::CellBuilder cb; - block::tlb::t_Grams.store_integer_value(cb, td::BigInt256(0)); - info.import_fee = cb.as_cellslice_ref(); - } - - tlb::csr_pack(message.info, info); - } - /* init */ { - if (new_state.not_null()) { - // Just(Left(new_state)) - message.init = vm::CellBuilder() - .store_ones(1) - .store_zeroes(1) - .append_cellslice(vm::load_cell_slice(new_state)) - .as_cellslice_ref(); - } else { - message.init = vm::CellBuilder().store_zeroes(1).as_cellslice_ref(); - CHECK(message.init.not_null()); - } - } - /* body */ { - message.body = vm::CellBuilder().store_zeroes(1).append_cellslice(vm::load_cell_slice_ref(body)).as_cellslice_ref(); - } - - td::Ref res; - tlb::type_pack_cell(res, block::gen::t_Message_Any, message); - CHECK(res.not_null()); - - return res; -} -} // namespace tonlib diff --git a/tonlib/tonlib/GenericAccount.h b/tonlib/tonlib/GenericAccount.h deleted file mode 100644 index 4a36d78a..00000000 --- a/tonlib/tonlib/GenericAccount.h +++ /dev/null @@ -1,30 +0,0 @@ -/* - This file is part of TON Blockchain Library. - - TON Blockchain Library is free software: you can redistribute it and/or modify - it under the terms of the GNU Lesser General Public License as published by - the Free Software Foundation, either version 2 of the License, or - (at your option) any later version. - - TON Blockchain Library is distributed in the hope that it will be useful, - but WITHOUT ANY WARRANTY; without even the implied warranty of - MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - GNU Lesser General Public License for more details. - - You should have received a copy of the GNU Lesser General Public License - along with TON Blockchain Library. If not, see . - - Copyright 2017-2019 Telegram Systems LLP -*/ -#pragma once -#include "vm/cells.h" -#include "block/block.h" -namespace tonlib { -class GenericAccount { - public: - static td::Ref get_init_state(td::Ref code, td::Ref data) noexcept; - static block::StdAddress get_address(ton::WorkchainId workchain_id, const td::Ref& init_state) noexcept; - static td::Ref create_ext_message(const block::StdAddress& address, td::Ref new_state, - td::Ref body) noexcept; -}; -} // namespace tonlib diff --git a/tonlib/tonlib/KeyValue.cpp b/tonlib/tonlib/KeyValue.cpp index cdb0f5ee..b62ce7c0 100644 --- a/tonlib/tonlib/KeyValue.cpp +++ b/tonlib/tonlib/KeyValue.cpp @@ -22,6 +22,7 @@ #include "td/utils/port/path.h" #include "td/utils/PathView.h" +#include #include #include @@ -42,7 +43,11 @@ class KeyValueDir : public KeyValue { } td::Status add(td::Slice key, td::Slice value) override { - auto path = to_file_path(key.str()); + auto key_str = key.str(); + if (!is_valid_key(key_str)) { + return td::Status::Error("Invalid key"); + } + auto path = to_file_path(key_str); if (td::stat(path).is_ok()) { return td::Status::Error(PSLICE() << "File " << path << "already exists"); } @@ -50,15 +55,27 @@ class KeyValueDir : public KeyValue { } td::Status set(td::Slice key, td::Slice value) override { - return td::atomic_write_file(to_file_path(key.str()), value); + auto key_str = key.str(); + if (!is_valid_key(key_str)) { + return td::Status::Error("Invalid key"); + } + return td::atomic_write_file(to_file_path(key_str), value); } td::Result get(td::Slice key) override { - return td::read_file_secure(to_file_path(key.str())); + auto key_str = key.str(); + if (!is_valid_key(key_str)) { + return td::Status::Error("Invalid key"); + } + return td::read_file_secure(to_file_path(key_str)); } td::Status erase(td::Slice key) override { - return td::unlink(to_file_path(key.str())); + auto key_str = key.str(); + if (!is_valid_key(key_str)) { + return td::Status::Error("Invalid key"); + } + return td::unlink(to_file_path(key_str)); } void foreach_key(std::function f) override { @@ -83,6 +100,20 @@ class KeyValueDir : public KeyValue { std::string to_file_path(std::string key) { return directory_ + TD_DIR_SLASH + key; } + + bool is_valid_key(const std::string& key) { + if (key.empty()) { + return false; + } + + if (key.find(TD_DIR_SLASH) != std::string::npos || key.find("..") != std::string::npos) { + return false; + } + + return std::all_of(key.begin(), key.end(), [](char c) { + return std::isalnum(c) || c == '_' || c == '-' || c == '.'; + }); + } }; class KeyValueInmemory : public KeyValue { diff --git a/tonlib/tonlib/LastBlockStorage.cpp b/tonlib/tonlib/LastBlockStorage.cpp index 53c4456e..6e5b3f0f 100644 --- a/tonlib/tonlib/LastBlockStorage.cpp +++ b/tonlib/tonlib/LastBlockStorage.cpp @@ -32,13 +32,43 @@ void LastBlockStorage::set_key_value(std::shared_ptr kv) { } namespace { +std::string buffer_to_hex_nibbles_reversed(td::Slice buffer) { + const char *hex = "0123456789ABCDEF"; + std::string res(2 * buffer.size(), '\0'); + for (std::size_t i = 0; i < buffer.size(); i++) { + unsigned char c = buffer[i]; + res[2 * i + 1] = hex[c >> 4]; + res[2 * i] = hex[c & 15]; + } + return res; +} + +std::string get_file_name_depr(td::Slice name) { + return buffer_to_hex_nibbles_reversed(name) + ".blkstate"; +} + std::string get_file_name(td::Slice name) { return td::buffer_to_hex(name) + ".blkstate"; } } // namespace td::Result LastBlockStorage::get_state(td::Slice name) { - TRY_RESULT(data, kv_->get(get_file_name(name))); + // This migration addresses an issue in the old version of Tonlib, where the td::buffer_to_hex + // incorrectly reversed the order of nibbles in hex representation. + auto data_r = kv_->get(get_file_name(name)); + if (data_r.is_error()) { + auto key_depr = get_file_name_depr(name); + auto data_depr = kv_->get(key_depr); + if (data_depr.is_ok()) { + kv_->set(get_file_name(name), data_depr.move_as_ok()); + kv_->erase(key_depr); + data_r = std::move(data_depr); + } else { + return td::Status::Error("not found"); + } + } + + auto data = data_r.move_as_ok(); if (data.size() < 8) { return td::Status::Error("too short"); } diff --git a/tonlib/tonlib/TestGiver.cpp b/tonlib/tonlib/TestGiver.cpp deleted file mode 100644 index b906193d..00000000 --- a/tonlib/tonlib/TestGiver.cpp +++ /dev/null @@ -1,52 +0,0 @@ -/* - This file is part of TON Blockchain Library. - - TON Blockchain Library is free software: you can redistribute it and/or modify - it under the terms of the GNU Lesser General Public License as published by - the Free Software Foundation, either version 2 of the License, or - (at your option) any later version. - - TON Blockchain Library is distributed in the hope that it will be useful, - but WITHOUT ANY WARRANTY; without even the implied warranty of - MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - GNU Lesser General Public License for more details. - - You should have received a copy of the GNU Lesser General Public License - along with TON Blockchain Library. If not, see . - - Copyright 2017-2019 Telegram Systems LLP -*/ -#include "tonlib/TestGiver.h" -#include "tonlib/utils.h" - -#include "td/utils/base64.h" - -namespace tonlib { -const block::StdAddress& TestGiver::address() noexcept { - static block::StdAddress res = - block::StdAddress::parse("kf_8uRo6OBbQ97jCx2EIuKm8Wmt6Vb15-KsQHFLbKSMiYIny").move_as_ok(); - return res; -} - -vm::CellHash TestGiver::get_init_code_hash() noexcept { - return vm::CellHash::from_slice(td::base64_decode("wDkZp0yR4xo+9+BnuAPfGVjBzK6FPzqdv2DwRq3z3KE=").move_as_ok()); -} - -td::Ref TestGiver::make_a_gift_message(td::uint32 seqno, td::uint64 gramms, td::Slice message, - const block::StdAddress& dest_address) noexcept { - td::BigInt256 dest_addr; - dest_addr.import_bits(dest_address.addr.as_bitslice()); - vm::CellBuilder cb; - cb.append_cellslice(binary_bitstring_to_cellslice("b{01}").move_as_ok()) - .store_long(dest_address.bounceable, 1) - .append_cellslice(binary_bitstring_to_cellslice("b{000100}").move_as_ok()) - .store_long(dest_address.workchain, 8) - .store_int256(dest_addr, 256); - block::tlb::t_Grams.store_integer_value(cb, td::BigInt256(gramms)); - - cb.store_zeroes(9 + 64 + 32 + 1 + 1).store_bytes("\0\0\0\0", 4); - vm::CellString::store(cb, message, 35 * 8).ensure(); - auto message_inner = cb.finalize(); - return vm::CellBuilder().store_long(seqno, 32).store_long(1, 8).store_ref(message_inner).finalize(); -} -} // namespace tonlib diff --git a/tonlib/tonlib/TestWallet.cpp b/tonlib/tonlib/TestWallet.cpp deleted file mode 100644 index 8bdf78c9..00000000 --- a/tonlib/tonlib/TestWallet.cpp +++ /dev/null @@ -1,84 +0,0 @@ -/* - This file is part of TON Blockchain Library. - - TON Blockchain Library is free software: you can redistribute it and/or modify - it under the terms of the GNU Lesser General Public License as published by - the Free Software Foundation, either version 2 of the License, or - (at your option) any later version. - - TON Blockchain Library is distributed in the hope that it will be useful, - but WITHOUT ANY WARRANTY; without even the implied warranty of - MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - GNU Lesser General Public License for more details. - - You should have received a copy of the GNU Lesser General Public License - along with TON Blockchain Library. If not, see . - - Copyright 2017-2019 Telegram Systems LLP -*/ -#include "tonlib/TestWallet.h" -#include "tonlib/GenericAccount.h" -#include "tonlib/utils.h" - -#include "vm/boc.h" -#include "td/utils/base64.h" - -namespace tonlib { -td::Ref TestWallet::get_init_state(const td::Ed25519::PublicKey& public_key) noexcept { - auto code = get_init_code(); - auto data = get_init_data(public_key); - return GenericAccount::get_init_state(std::move(code), std::move(data)); -} - -td::Ref TestWallet::get_init_message(const td::Ed25519::PrivateKey& private_key) noexcept { - std::string seq_no(4, 0); - auto signature = - private_key.sign(vm::CellBuilder().store_bytes(seq_no).finalize()->get_hash().as_slice()).move_as_ok(); - return vm::CellBuilder().store_bytes(signature).store_bytes(seq_no).finalize(); -} - -td::Ref TestWallet::make_a_gift_message(const td::Ed25519::PrivateKey& private_key, td::uint32 seqno, - td::int64 gramms, td::Slice message, - const block::StdAddress& dest_address) noexcept { - td::BigInt256 dest_addr; - dest_addr.import_bits(dest_address.addr.as_bitslice()); - vm::CellBuilder cb; - cb.append_cellslice(binary_bitstring_to_cellslice("b{01}").move_as_ok()) - .store_long(dest_address.bounceable, 1) - .append_cellslice(binary_bitstring_to_cellslice("b{000100}").move_as_ok()) - .store_long(dest_address.workchain, 8) - .store_int256(dest_addr, 256); - td::int32 send_mode = 3; - if (gramms == -1) { - gramms = 0; - send_mode += 128; - } - block::tlb::t_Grams.store_integer_value(cb, td::BigInt256(gramms)); - cb.store_zeroes(9 + 64 + 32 + 1 + 1).store_bytes("\0\0\0\0", 4); - vm::CellString::store(cb, message, 35 * 8).ensure(); - auto message_inner = cb.finalize(); - auto message_outer = - vm::CellBuilder().store_long(seqno, 32).store_long(send_mode, 8).store_ref(message_inner).finalize(); - auto signature = private_key.sign(message_outer->get_hash().as_slice()).move_as_ok(); - return vm::CellBuilder().store_bytes(signature).append_cellslice(vm::load_cell_slice(message_outer)).finalize(); -} - -td::Ref TestWallet::get_init_code() noexcept { - static auto res = [] { - auto serialized_code = td::base64_decode( - "te6ccgEEAQEAAAAAUwAAov8AIN0gggFMl7qXMO1E0NcLH+Ck8mCBAgDXGCDXCx/tRNDTH9P/" - "0VESuvKhIvkBVBBE+RDyovgAAdMfMSDXSpbTB9QC+wDe0aTIyx/L/8ntVA==") - .move_as_ok(); - return vm::std_boc_deserialize(serialized_code).move_as_ok(); - }(); - return res; -} - -vm::CellHash TestWallet::get_init_code_hash() noexcept { - return get_init_code()->get_hash(); -} - -td::Ref TestWallet::get_init_data(const td::Ed25519::PublicKey& public_key) noexcept { - return vm::CellBuilder().store_long(0, 32).store_bytes(public_key.as_octet_string()).finalize(); -} -} // namespace tonlib diff --git a/tonlib/tonlib/TestWallet.h b/tonlib/tonlib/TestWallet.h deleted file mode 100644 index ef726b55..00000000 --- a/tonlib/tonlib/TestWallet.h +++ /dev/null @@ -1,40 +0,0 @@ -/* - This file is part of TON Blockchain Library. - - TON Blockchain Library is free software: you can redistribute it and/or modify - it under the terms of the GNU Lesser General Public License as published by - the Free Software Foundation, either version 2 of the License, or - (at your option) any later version. - - TON Blockchain Library is distributed in the hope that it will be useful, - but WITHOUT ANY WARRANTY; without even the implied warranty of - MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - GNU Lesser General Public License for more details. - - You should have received a copy of the GNU Lesser General Public License - along with TON Blockchain Library. If not, see . - - Copyright 2017-2019 Telegram Systems LLP -*/ -#pragma once - -#include "vm/cells.h" -#include "Ed25519.h" -#include "block/block.h" -#include "CellString.h" - -namespace tonlib { -class TestWallet { - public: - static constexpr unsigned max_message_size = vm::CellString::max_bytes; - static td::Ref get_init_state(const td::Ed25519::PublicKey& public_key) noexcept; - static td::Ref get_init_message(const td::Ed25519::PrivateKey& private_key) noexcept; - static td::Ref make_a_gift_message(const td::Ed25519::PrivateKey& private_key, td::uint32 seqno, - td::int64 gramms, td::Slice message, - const block::StdAddress& dest_address) noexcept; - - static td::Ref get_init_code() noexcept; - static vm::CellHash get_init_code_hash() noexcept; - static td::Ref get_init_data(const td::Ed25519::PublicKey& public_key) noexcept; -}; -} // namespace tonlib diff --git a/tonlib/tonlib/TonlibClient.cpp b/tonlib/tonlib/TonlibClient.cpp index d450bc1f..86acbcc3 100644 --- a/tonlib/tonlib/TonlibClient.cpp +++ b/tonlib/tonlib/TonlibClient.cpp @@ -1227,8 +1227,7 @@ class RemoteRunSmcMethod : public td::actor::Actor { td::Status do_with_last_block(td::Result r_last_block) { TRY_RESULT(last_block, std::move(r_last_block)); query_.block_id = std::move(last_block.last_block_id); - with_block_id(); - return td::Status::OK(); + return with_block_id(); } void start_up() override { @@ -1729,6 +1728,9 @@ class GetShardBlockProof : public td::actor::Actor { auto to_lite_api(const tonlib_api::ton_blockIdExt& blk) -> td::Result>; auto to_tonlib_api(const ton::lite_api::liteServer_transactionId& txid) -> tonlib_api_ptr; +td::Status check_block_transactions_proof(lite_api_ptr& bTxes, int32_t mode, + ton::LogicalTime start_lt, td::Bits256 start_addr, td::Bits256 root_hash, int req_count); + class RunEmulator : public TonlibQueryActor { public: RunEmulator(ExtClientRef ext_client_ref, int_api::GetAccountStateByTransaction request, @@ -1778,7 +1780,6 @@ class RunEmulator : public TonlibQueryActor { return td::Status::Error("block header proof is not a valid Merkle proof"); } - ton::RootHash vhash{virt_root->get_hash().bits()}; if (ton::RootHash{virt_root->get_hash().bits()} != block_id.root_hash) { return td::Status::Error("block header has incorrect root hash"); } @@ -1837,14 +1838,18 @@ class RunEmulator : public TonlibQueryActor { td::Status get_transactions(std::int64_t lt) { TRY_RESULT(lite_block, to_lite_api(*to_tonlib_api(block_id_.id))); auto after = ton::lite_api::make_object(request_.address.addr, lt); - auto query = ton::lite_api::liteServer_listBlockTransactions(std::move(lite_block), 0b10100111, 256, std::move(after), false, false); + auto mode = 0b10100111; + constexpr int req_count = 256; + auto query = ton::lite_api::liteServer_listBlockTransactions(std::move(lite_block), mode, req_count, std::move(after), false, true); - client_.send_query(std::move(query), [self = this](lite_api_ptr&& bTxes) { + client_.send_query(std::move(query), [self = this, mode, lt, root_hash = block_id_.id.root_hash, req_count](lite_api_ptr&& bTxes) { if (!bTxes) { self->check(td::Status::Error("liteServer.blockTransactions is null")); return; } + self->check(check_block_transactions_proof(bTxes, mode, lt, self->request_.address.addr, root_hash, req_count)); + std::int64_t last_lt = 0; for (auto& id : bTxes->ids_) { last_lt = id->lt_; @@ -2971,7 +2976,7 @@ struct ToRawTransactions { tonlib_api::object_ptr data; if (try_decode_messages_ && body->size() >= 32) { auto type = static_cast(body.write().fetch_long(32)); - if (type == 0 || type == 0x2167da4b) { + if (type == 0 || type == ton::WalletInterface::EncryptedCommentOp) { td::Status status; auto r_body_message = TRY_VM(vm::CellString::load(body.write())); @@ -4241,8 +4246,7 @@ td::Status TonlibClient::do_request(const tonlib_api::query_send& request, td::Status TonlibClient::do_request(tonlib_api::query_forget& request, td::Promise>&& promise) { - auto it = queries_.find(request.id_); - if (it == queries_.end()) { + if (queries_.erase(request.id_) == 0) { return TonlibError::InvalidQueryId(); } promise.set_value(tonlib_api::make_object()); @@ -4490,10 +4494,31 @@ void deep_library_search(std::set& set, std::set& v td::Status TonlibClient::do_request(const tonlib_api::smc_getLibraries& request, td::Promise>&& promise) { + if (request.library_list_.size() > 16) { + promise.set_error(TonlibError::InvalidField("library_list", ": too many libraries requested, 16 maximum")); + } + if (query_context_.block_id) { + get_libraries(query_context_.block_id.value(), request.library_list_, std::move(promise)); + } else { + client_.with_last_block([this, promise = std::move(promise), library_list = request.library_list_](td::Result r_last_block) mutable { + if (r_last_block.is_error()) { + promise.set_error(r_last_block.move_as_error_prefix(TonlibError::Internal("get last block failed "))); + } else { + this->get_libraries(r_last_block.move_as_ok().last_block_id, library_list, std::move(promise)); + } + }); + } + return td::Status::OK(); +} + +void TonlibClient::get_libraries(ton::BlockIdExt blkid, std::vector library_list, td::Promise>&& promise) { + sort(library_list.begin(), library_list.end()); + library_list.erase(unique(library_list.begin(), library_list.end()), library_list.end()); + std::vector> result_entries; - result_entries.reserve(request.library_list_.size()); + result_entries.reserve(library_list.size()); std::vector not_cached_hashes; - for (auto& library_hash : request.library_list_) { + for (auto& library_hash : library_list) { if (libraries.key_exists(library_hash)) { auto library_content = vm::std_boc_serialize(libraries.lookup_ref(library_hash)).move_as_ok().as_slice().str(); result_entries.push_back(tonlib_api::make_object(library_hash, library_content)); @@ -4504,40 +4529,80 @@ td::Status TonlibClient::do_request(const tonlib_api::smc_getLibraries& request, if (not_cached_hashes.empty()) { promise.set_value(tonlib_api::make_object(std::move(result_entries))); - return td::Status::OK(); + return; } - client_.send_query(ton::lite_api::liteServer_getLibraries(std::move(not_cached_hashes)), - promise.wrap([self=this, result_entries = std::move(result_entries)] - (td::Result> r_libraries) mutable - { - if (r_libraries.is_error()) { - LOG(WARNING) << "cannot obtain found libraries: " << r_libraries.move_as_error().to_string(); - } else { - auto libraries = r_libraries.move_as_ok(); - bool updated = false; - for (auto& lr : libraries->result_) { - auto contents = vm::std_boc_deserialize(lr->data_); - if (contents.is_ok() && contents.ok().not_null()) { - if (contents.ok()->get_hash().bits().compare(lr->hash_.cbits(), 256)) { - LOG(WARNING) << "hash mismatch for library " << lr->hash_.to_hex(); - continue; - } - result_entries.push_back(tonlib_api::make_object(lr->hash_, lr->data_.as_slice().str())); - self->libraries.set_ref(lr->hash_, contents.move_as_ok()); - updated = true; - LOG(DEBUG) << "registered library " << lr->hash_.to_hex(); - } else { - LOG(WARNING) << "failed to deserialize library: " << lr->hash_.to_hex(); - } - if (updated) { - self->store_libs_to_disk(); - } - } + client_.send_query(ton::lite_api::liteServer_getLibrariesWithProof(ton::create_tl_lite_block_id(blkid), 1, std::move(not_cached_hashes)), + promise.wrap([self=this, blkid, result_entries = std::move(result_entries), not_cached_hashes] + (td::Result> r_libraries) mutable + -> td::Result> { + if (r_libraries.is_error()) { + LOG(WARNING) << "cannot obtain found libraries: " << r_libraries.move_as_error().to_string(); + return r_libraries.move_as_error(); + } + + auto libraries = r_libraries.move_as_ok(); + auto state = block::check_extract_state_proof(blkid, libraries->state_proof_.as_slice(), + libraries->data_proof_.as_slice()); + if (state.is_error()) { + LOG(WARNING) << "cannot check state proof: " << state.move_as_error().to_string(); + return state.move_as_error(); + } + auto state_root = state.move_as_ok(); + + try { + block::gen::ShardStateUnsplit::Record state_record; + if (!tlb::unpack_cell(state_root, state_record)) { + return td::Status::Error("cannot unpack shardchain state"); } + auto libraries_dict = vm::Dictionary(state_record.r1.libraries->prefetch_ref(), 256); + + for (auto& hash : not_cached_hashes) { + auto csr = libraries_dict.lookup(hash.bits(), 256); + if (csr.is_null()) { + LOG(WARNING) << "library " << hash.to_hex() << " not found in config"; + if (std::any_of(libraries->result_.begin(), libraries->result_.end(), + [&hash](const auto& lib) { return lib->hash_.bits().equals(hash.cbits(), 256); })) { + return TonlibError::Internal("library is included in response but it's not found in proof"); + } + continue; + } + block::gen::LibDescr::Record libdescr; + if (!tlb::csr_unpack(csr, libdescr)) { + return TonlibError::Internal("cannot unpack LibDescr record"); + } + + auto lib_it = std::find_if(libraries->result_.begin(), libraries->result_.end(), + [&hash](const auto& lib) { return lib->hash_.bits().equals(hash.cbits(), 256); }); + if (lib_it == libraries->result_.end()) { + return TonlibError::Internal("library is found in proof but not in response"); + } + auto& lib = *lib_it; + auto contents = vm::std_boc_deserialize(lib->data_); + if (!contents.is_ok() || contents.ok().is_null()) { + return TonlibError::Internal(PSLICE() << "cannot deserialize library cell " << lib->hash_.to_hex()); + } + + if (!contents.ok()->get_hash().bits().equals(hash.cbits(), 256)) { + return TonlibError::Internal(PSLICE() << "library hash mismatch data " << contents.ok()->get_hash().to_hex() << " != requested " << hash.to_hex()); + } + + if (contents.ok()->get_hash() != libdescr.lib->get_hash()) { + return TonlibError::Internal(PSLICE() << "library hash mismatch data " << lib->hash_.to_hex() << " != proof " << libdescr.lib->get_hash().to_hex()); + } + + result_entries.push_back(tonlib_api::make_object(lib->hash_, lib->data_.as_slice().str())); + self->libraries.set_ref(lib->hash_, contents.move_as_ok()); + LOG(DEBUG) << "registered library " << lib->hash_.to_hex(); + } + self->store_libs_to_disk(); return tonlib_api::make_object(std::move(result_entries)); - })); - return td::Status::OK(); + } catch (vm::VmError& err) { + return TonlibError::Internal(PSLICE() << "error while checking getLibrariesWithProof proof: " << err.get_msg()); + } catch (vm::VmVirtError& err) { + return TonlibError::Internal(PSLICE() << "virtualization error while checking getLibrariesWithProof proof: " << err.get_msg()); + } + })); } td::Status TonlibClient::do_request(const tonlib_api::smc_getLibrariesExt& request, @@ -5413,53 +5478,247 @@ td::Status TonlibClient::do_request(const tonlib_api::blocks_getMasterchainInfo& td::Status TonlibClient::do_request(const tonlib_api::blocks_getShards& request, td::Promise>&& promise) { TRY_RESULT(block, to_lite_api(*request.id_)) + TRY_RESULT(req_blk_id, to_block_id(*request.id_)); client_.send_query(ton::lite_api::liteServer_getAllShardsInfo(std::move(block)), - promise.wrap([](lite_api_ptr&& all_shards_info) + promise.wrap([req_blk_id](lite_api_ptr&& all_shards_info) -> td::Result> { - td::BufferSlice proof = std::move((*all_shards_info).proof_); - td::BufferSlice data = std::move((*all_shards_info).data_); - if (data.empty()) { - return td::Status::Error("shard configuration is empty"); - } else { - auto R = vm::std_boc_deserialize(data.clone()); - if (R.is_error()) { - return R.move_as_error_prefix("cannot deserialize shard configuration: "); - } - auto root = R.move_as_ok(); - block::ShardConfig sh_conf; - if (!sh_conf.unpack(vm::load_cell_slice_ref(root))) { - return td::Status::Error("cannot extract shard block list from shard configuration"); - } else { - auto ids = sh_conf.get_shard_hash_ids(true); - tonlib_api::blocks_shards shards; - for (auto id : ids) { - auto ref = sh_conf.get_shard_hash(ton::ShardIdFull(id)); - if (ref.not_null()) { - shards.shards_.push_back(to_tonlib_api(ref->top_block_id())); - } - } - return tonlib_api::make_object(std::move(shards)); - } - } + auto blk_id = ton::create_block_id(all_shards_info->id_); + if (blk_id != req_blk_id) { + return td::Status::Error("Liteserver responded with wrong block"); + } + td::BufferSlice proof = std::move((*all_shards_info).proof_); + td::BufferSlice data = std::move((*all_shards_info).data_); + if (data.empty() || proof.empty()) { + return td::Status::Error("Shard configuration or proof is empty"); + } + auto proof_cell = vm::std_boc_deserialize(std::move(proof)); + if (proof_cell.is_error()) { + return proof_cell.move_as_error_prefix("Couldn't deserialize shards proof: "); + } + auto data_cell = vm::std_boc_deserialize(std::move(data)); + if (data_cell.is_error()) { + return data_cell.move_as_error_prefix("Couldn't deserialize shards data: "); + } + try { + auto virt_root = vm::MerkleProof::virtualize(proof_cell.move_as_ok(), 1); + if (virt_root.is_null()) { + return td::Status::Error("Virt root is null"); + } + if (ton::RootHash{virt_root->get_hash().bits()} != blk_id.root_hash) { + return td::Status::Error("Block shards merkle proof has incorrect root hash"); + } + + block::gen::Block::Record blk; + block::gen::BlockExtra::Record extra; + block::gen::McBlockExtra::Record mc_extra; + if (!tlb::unpack_cell(virt_root, blk) || !tlb::unpack_cell(blk.extra, extra) || !extra.custom->have_refs() || + !tlb::unpack_cell(extra.custom->prefetch_ref(), mc_extra)) { + return td::Status::Error("cannot unpack block extra of block " + blk_id.to_str()); + } + auto data_csr = vm::load_cell_slice_ref(data_cell.move_as_ok()); + if (data_csr->prefetch_ref()->get_hash() != mc_extra.shard_hashes->prefetch_ref()->get_hash()) { + return td::Status::Error("Block shards data and proof hashes don't match"); + } + + block::ShardConfig sh_conf; + if (!sh_conf.unpack(mc_extra.shard_hashes)) { + return td::Status::Error("cannot extract shard block list from shard configuration"); + } + auto ids = sh_conf.get_shard_hash_ids(true); + tonlib_api::blocks_shards shards; + for (auto& id : ids) { + auto ref = sh_conf.get_shard_hash(ton::ShardIdFull(id)); + if (ref.not_null()) { + shards.shards_.push_back(to_tonlib_api(ref->top_block_id())); + } + } + return tonlib_api::make_object(std::move(shards)); + } catch (vm::VmError& err) { + return err.as_status("Couldn't verify proof: "); + } catch (vm::VmVirtError& err) { + return err.as_status("Couldn't verify proof: "); + } catch (...) { + return td::Status::Error("Unknown exception raised while verifying proof"); + } })); return td::Status::OK(); } +td::Status check_lookup_block_proof(lite_api_ptr& result, int mode, ton::BlockId blkid, ton::BlockIdExt client_mc_blkid, td::uint64 lt, td::uint32 utime); td::Status TonlibClient::do_request(const tonlib_api::blocks_lookupBlock& request, td::Promise>&& promise) { - client_.send_query(ton::lite_api::liteServer_lookupBlock( - request.mode_, - ton::lite_api::make_object((*request.id_).workchain_, (*request.id_).shard_, (*request.id_).seqno_), - (td::uint64)(request.lt_), - (td::uint32)(request.utime_)), - promise.wrap([](lite_api_ptr&& header) { - const auto& id = header->id_; - return to_tonlib_api(*id); - //tonlib_api::make_object( - // ton::tonlib_api::ton_blockIdExt(id->workchain_, id->) - //); - })); + auto lite_block = ton::lite_api::make_object((*request.id_).workchain_, (*request.id_).shard_, (*request.id_).seqno_); + auto blkid = ton::BlockId(request.id_->workchain_, request.id_->shard_, request.id_->seqno_); + client_.with_last_block( + [self = this, blkid, lite_block = std::move(lite_block), mode = request.mode_, lt = (td::uint64)request.lt_, + utime = (td::uint32)request.utime_, promise = std::move(promise)](td::Result r_last_block) mutable { + self->client_.send_query(ton::lite_api::liteServer_lookupBlockWithProof(mode, std::move(lite_block), ton::create_tl_lite_block_id(r_last_block.ok().last_block_id), lt, utime), + promise.wrap([blkid, mode, utime, lt, last_block = r_last_block.ok().last_block_id](lite_api_ptr&& result) + -> td::Result> { + TRY_STATUS(check_lookup_block_proof(result, mode, blkid, last_block, lt, utime)); + return to_tonlib_api(*result->id_); + }) + ); + }); + return td::Status::OK(); +} + +td::Status check_lookup_block_proof(lite_api_ptr& result, int mode, ton::BlockId blkid, ton::BlockIdExt client_mc_blkid, td::uint64 lt, td::uint32 utime) { + try { + ton::BlockIdExt cur_id = ton::create_block_id(result->mc_block_id_); + try { + for (auto& link : result->shard_links_) { + ton::BlockIdExt prev_id = create_block_id(link->id_); + td::BufferSlice proof = std::move(link->proof_); + auto R = vm::std_boc_deserialize(proof); + if (R.is_error()) { + return TonlibError::InvalidBagOfCells("proof"); + } + auto block_root = vm::MerkleProof::virtualize(R.move_as_ok(), 1); + if (cur_id.root_hash != block_root->get_hash().bits()) { + return td::Status::Error("invalid block hash in proof"); + } + if (cur_id.is_masterchain()) { + if (client_mc_blkid != cur_id) { + auto state = block::check_extract_state_proof(client_mc_blkid, result->client_mc_state_proof_.as_slice(), + result->mc_block_proof_.as_slice()); + if (state.is_error()) { + LOG(WARNING) << "cannot check state proof: " << state.move_as_error().to_string(); + return state.move_as_error(); + } + auto state_root = state.move_as_ok(); + auto prev_blocks_dict = block::get_prev_blocks_dict(state_root); + if (!prev_blocks_dict) { + return td::Status::Error("cannot extract prev blocks dict from state"); + } + + if (!block::check_old_mc_block_id(*prev_blocks_dict, cur_id)) { + return td::Status::Error("couldn't check old mc block id"); + } + } + block::gen::Block::Record blk; + block::gen::BlockExtra::Record extra; + block::gen::McBlockExtra::Record mc_extra; + if (!tlb::unpack_cell(block_root, blk) || !tlb::unpack_cell(blk.extra, extra) || !extra.custom->have_refs() || + !tlb::unpack_cell(extra.custom->prefetch_ref(), mc_extra)) { + return td::Status::Error("cannot unpack block header"); + } + block::ShardConfig shards(mc_extra.shard_hashes->prefetch_ref()); + td::Ref shard_hash = shards.get_shard_hash(prev_id.shard_full(), true); + if (shard_hash.is_null() || shard_hash->top_block_id() != prev_id) { + return td::Status::Error("invalid proof chain: prev block is not in mc shard list"); + } + } else { + std::vector prev; + ton::BlockIdExt mc_blkid; + bool after_split; + td::Status S = block::unpack_block_prev_blk_try(block_root, cur_id, prev, mc_blkid, after_split); + if (S.is_error()) { + return S; + } + CHECK(prev.size() == 1 || prev.size() == 2); + bool found = prev_id == prev[0] || (prev.size() == 2 && prev_id == prev[1]); + if (!found) { + return td::Status::Error("invalid proof chain: prev block is not in prev blocks list"); + } + } + cur_id = prev_id; + } + } catch (vm::VmVirtError& err) { + return err.as_status(); + } + if (cur_id.id.workchain != blkid.workchain || !ton::shard_contains(cur_id.id.shard, blkid.shard)) { + return td::Status::Error("response block has incorrect workchain/shard"); + } + + auto header_r = vm::std_boc_deserialize(std::move(result->header_)); + if (header_r.is_error()) { + return TonlibError::InvalidBagOfCells("header"); + } + auto header_root = vm::MerkleProof::virtualize(header_r.move_as_ok(), 1); + if (header_root.is_null()) { + return td::Status::Error("header_root is null"); + } + if (cur_id.root_hash != header_root->get_hash().bits()) { + return td::Status::Error("invalid header hash in proof"); + } + + std::vector prev; + ton::BlockIdExt mc_blkid; + bool after_split; + auto R = block::unpack_block_prev_blk_try(header_root, cur_id, prev, mc_blkid, after_split); + if (R.is_error()) { + return R; + } + if (cur_id != ton::create_block_id(result->id_)) { + return td::Status::Error("response blkid doesn't match header"); + } + + block::gen::Block::Record blk; + block::gen::BlockInfo::Record info; + if (!(tlb::unpack_cell(header_root, blk) && tlb::unpack_cell(blk.info, info))) { + return td::Status::Error("block header unpack failed"); + } + + if (mode & 1) { + if (cur_id.seqno() != blkid.seqno) { + return td::Status::Error("invalid seqno in proof"); + } + } else if (mode & 6) { + auto prev_header_r = vm::std_boc_deserialize(std::move(result->prev_header_)); + if (prev_header_r.is_error()) { + return TonlibError::InvalidBagOfCells("prev_headers"); + } + auto prev_header = prev_header_r.move_as_ok(); + auto prev_root = vm::MerkleProof::virtualize(prev_header, 1); + if (prev_root.is_null()) { + return td::Status::Error("prev_root is null"); + } + + bool prev_valid = false; + int prev_idx = -1; + for (size_t i = 0; i < prev.size(); i++) { + if (prev[i].root_hash == prev_root->get_hash().bits()) { + prev_valid = true; + prev_idx = i; + } + } + if (!prev_valid) { + return td::Status::Error("invalid prev header hash in proof"); + } + if (!ton::shard_contains(prev[prev_idx].id.shard, blkid.shard)) { + return td::Status::Error("invalid prev header shard in proof"); + } + + block::gen::Block::Record prev_blk; + block::gen::BlockInfo::Record prev_info; + if (!(tlb::unpack_cell(prev_root, prev_blk) && tlb::unpack_cell(prev_blk.info, prev_info))) { + return td::Status::Error("prev header unpack failed"); + } + + if (mode & 2) { + if (prev_info.end_lt > lt) { + return td::Status::Error("prev header end_lt > lt"); + } + if (info.end_lt < lt) { + return td::Status::Error("header end_lt < lt"); + } + } else if (mode & 4) { + if (prev_info.gen_utime > utime) { + return td::Status::Error("prev header end_lt > lt"); + } + if (info.gen_utime < utime) { + return td::Status::Error("header end_lt < lt"); + } + } + } + } catch (vm::VmError& err) { + return td::Status::Error(PSLICE() << "error while checking lookupBlock proof: " << err.get_msg()); + } catch (vm::VmVirtError& err) { + return td::Status::Error(PSLICE() << "virtualization error while checking lookupBlock proof: " << err.get_msg()); + } + return td::Status::OK(); } @@ -5469,13 +5728,95 @@ auto to_tonlib_api(const ton::lite_api::liteServer_transactionId& txid) txid.mode_, txid.account_.as_slice().str(), txid.lt_, txid.hash_.as_slice().str()); } +td::Status check_block_transactions_proof(lite_api_ptr& bTxes, int32_t mode, + ton::LogicalTime start_lt, td::Bits256 start_addr, td::Bits256 root_hash, int req_count) { + if (mode & ton::lite_api::liteServer_listBlockTransactions::WANT_PROOF_MASK == 0) { + return td::Status::OK(); + } + constexpr int max_answer_transactions = 256; + bool reverse_mode = mode & ton::lite_api::liteServer_listBlockTransactions::REVERSE_ORDER_MASK; + + try { + TRY_RESULT(proof_cell, vm::std_boc_deserialize(std::move(bTxes->proof_))); + auto virt_root = vm::MerkleProof::virtualize(proof_cell, 1); + + if (root_hash != virt_root->get_hash().bits()) { + return td::Status::Error("Invalid block proof root hash"); + } + block::gen::Block::Record blk; + block::gen::BlockExtra::Record extra; + if (!(tlb::unpack_cell(virt_root, blk) && tlb::unpack_cell(std::move(blk.extra), extra))) { + return td::Status::Error("Error unpacking proof cell"); + } + vm::AugmentedDictionary acc_dict{vm::load_cell_slice_ref(extra.account_blocks), 256, + block::tlb::aug_ShardAccountBlocks}; + + bool eof = false; + ton::LogicalTime reverse = reverse_mode ? ~0ULL : 0; + ton::LogicalTime trans_lt = static_cast(start_lt); + td::Bits256 cur_addr = start_addr; + bool allow_same = true; + int count = 0; + while (!eof && count < req_count && count < max_answer_transactions) { + auto value = acc_dict.extract_value( + acc_dict.vm::DictionaryFixed::lookup_nearest_key(cur_addr.bits(), 256, !reverse, allow_same)); + if (value.is_null()) { + eof = true; + break; + } + allow_same = false; + if (cur_addr != start_addr) { + trans_lt = reverse; + } + + block::gen::AccountBlock::Record acc_blk; + if (!tlb::csr_unpack(std::move(value), acc_blk) || acc_blk.account_addr != cur_addr) { + return td::Status::Error("Error unpacking proof account block"); + } + vm::AugmentedDictionary trans_dict{vm::DictNonEmpty(), std::move(acc_blk.transactions), 64, + block::tlb::aug_AccountTransactions}; + td::BitArray<64> cur_trans{(long long)trans_lt}; + while (count < req_count && count < max_answer_transactions) { + auto tvalue = trans_dict.extract_value_ref( + trans_dict.vm::DictionaryFixed::lookup_nearest_key(cur_trans.bits(), 64, !reverse)); + if (tvalue.is_null()) { + trans_lt = reverse; + break; + } + if (static_cast(count) < bTxes->ids_.size()) { + if (mode & 4 && !tvalue->get_hash().bits().equals(bTxes->ids_[count]->hash_.bits(), 256)) { + return td::Status::Error("Couldn't verify proof (hash)"); + } + if (mode & 2 && cur_trans != td::BitArray<64>(bTxes->ids_[count]->lt_)) { + return td::Status::Error("Couldn't verify proof (lt)"); + } + if (mode & 1 && cur_addr != bTxes->ids_[count]->account_) { + return td::Status::Error("Couldn't verify proof (account)"); + } + } + count++; + } + } + if (static_cast(count) != bTxes->ids_.size()) { + return td::Status::Error(PSLICE() << "Txs count mismatch in proof (" << count << ") and response (" << bTxes->ids_.size() << ")"); + } + } catch (vm::VmError& err) { + return err.as_status("Couldn't verify proof: "); + } catch (vm::VmVirtError& err) { + return err.as_status("Couldn't verify proof: "); + } catch (...) { + return td::Status::Error("Unknown exception raised while verifying proof"); + } + return td::Status::OK(); +} + td::Status TonlibClient::do_request(const tonlib_api::blocks_getTransactions& request, td::Promise>&& promise) { TRY_RESULT(block, to_lite_api(*request.id_)) auto root_hash = block->root_hash_; - bool check_proof = request.mode_ & 32; - bool reverse_mode = request.mode_ & 64; - bool has_starting_tx = request.mode_ & 128; + bool check_proof = request.mode_ & ton::lite_api::liteServer_listBlockTransactions::WANT_PROOF_MASK; + bool reverse_mode = request.mode_ & ton::lite_api::liteServer_listBlockTransactions::REVERSE_ORDER_MASK; + bool has_starting_tx = request.mode_ & ton::lite_api::liteServer_listBlockTransactions::AFTER_MASK; td::Bits256 start_addr; ton::LogicalTime start_lt; @@ -5502,80 +5843,7 @@ td::Status TonlibClient::do_request(const tonlib_api::blocks_getTransactions& re check_proof), promise.wrap([check_proof, reverse_mode, root_hash, req_count = request.count_, start_addr, start_lt, mode = request.mode_] (lite_api_ptr&& bTxes) -> td::Result> { - if (check_proof) { - try { - constexpr int max_answer_transactions = 256; - TRY_RESULT(proof_cell, vm::std_boc_deserialize(std::move(bTxes->proof_))); - auto virt_root = vm::MerkleProof::virtualize(proof_cell, 1); - - if (root_hash != virt_root->get_hash().bits()) { - return td::Status::Error("Invalid block proof root hash"); - } - block::gen::Block::Record blk; - block::gen::BlockExtra::Record extra; - if (!(tlb::unpack_cell(virt_root, blk) && tlb::unpack_cell(std::move(blk.extra), extra))) { - return td::Status::Error("Error unpacking proof cell"); - } - vm::AugmentedDictionary acc_dict{vm::load_cell_slice_ref(extra.account_blocks), 256, - block::tlb::aug_ShardAccountBlocks}; - - bool eof = false; - ton::LogicalTime reverse = reverse_mode ? ~0ULL : 0; - ton::LogicalTime trans_lt = static_cast(start_lt); - td::Bits256 cur_addr = start_addr; - bool allow_same = true; - int count = 0; - while (!eof && count < req_count && count < max_answer_transactions) { - auto value = acc_dict.extract_value( - acc_dict.vm::DictionaryFixed::lookup_nearest_key(cur_addr.bits(), 256, !reverse, allow_same)); - if (value.is_null()) { - eof = true; - break; - } - allow_same = false; - if (cur_addr != start_addr) { - trans_lt = reverse; - } - - block::gen::AccountBlock::Record acc_blk; - if (!tlb::csr_unpack(std::move(value), acc_blk) || acc_blk.account_addr != cur_addr) { - return td::Status::Error("Error unpacking proof account block"); - } - vm::AugmentedDictionary trans_dict{vm::DictNonEmpty(), std::move(acc_blk.transactions), 64, - block::tlb::aug_AccountTransactions}; - td::BitArray<64> cur_trans{(long long)trans_lt}; - while (count < req_count && count < max_answer_transactions) { - auto tvalue = trans_dict.extract_value_ref( - trans_dict.vm::DictionaryFixed::lookup_nearest_key(cur_trans.bits(), 64, !reverse)); - if (tvalue.is_null()) { - trans_lt = reverse; - break; - } - if (static_cast(count) < bTxes->ids_.size()) { - if (mode & 4 && !tvalue->get_hash().bits().equals(bTxes->ids_[count]->hash_.bits(), 256)) { - return td::Status::Error("Couldn't verify proof (hash)"); - } - if (mode & 2 && cur_trans != td::BitArray<64>(bTxes->ids_[count]->lt_)) { - return td::Status::Error("Couldn't verify proof (lt)"); - } - if (mode & 1 && cur_addr != bTxes->ids_[count]->account_) { - return td::Status::Error("Couldn't verify proof (account)"); - } - } - count++; - } - } - if (static_cast(count) != bTxes->ids_.size()) { - return td::Status::Error(PSLICE() << "Txs count mismatch in proof (" << count << ") and response (" << bTxes->ids_.size() << ")"); - } - } catch (vm::VmError& err) { - return err.as_status("Couldn't verify proof: "); - } catch (vm::VmVirtError& err) { - return err.as_status("Couldn't verify proof: "); - } catch (...) { - return td::Status::Error("Unknown exception raised while verifying proof"); - } - } + TRY_STATUS(check_block_transactions_proof(bTxes, mode, start_lt, start_addr, root_hash, req_count)); tonlib_api::blocks_transactions r; r.id_ = to_tonlib_api(*bTxes->id_); @@ -5592,9 +5860,9 @@ td::Status TonlibClient::do_request(const tonlib_api::blocks_getTransactions& re td::Status TonlibClient::do_request(const tonlib_api::blocks_getTransactionsExt& request, td::Promise>&& promise) { TRY_RESULT(block, to_lite_api(*request.id_)) - bool check_proof = request.mode_ & 32; - bool reverse_mode = request.mode_ & 64; - bool has_starting_tx = request.mode_ & 128; + bool check_proof = request.mode_ & ton::lite_api::liteServer_listBlockTransactionsExt::WANT_PROOF_MASK; + bool reverse_mode = request.mode_ & ton::lite_api::liteServer_listBlockTransactionsExt::REVERSE_ORDER_MASK; + bool has_starting_tx = request.mode_ & ton::lite_api::liteServer_listBlockTransactionsExt::AFTER_MASK; td::Bits256 start_addr; ton::LogicalTime start_lt; @@ -5655,77 +5923,78 @@ td::Status TonlibClient::do_request(const tonlib_api::blocks_getTransactionsExt& td::Status TonlibClient::do_request(const tonlib_api::blocks_getBlockHeader& request, td::Promise>&& promise) { - TRY_RESULT(block, to_lite_api(*request.id_)) + TRY_RESULT(lite_block, to_lite_api(*request.id_)) + TRY_RESULT(req_blk_id, to_block_id(*request.id_)); client_.send_query(ton::lite_api::liteServer_getBlockHeader( - std::move(block), + std::move(lite_block), 0xffff), - promise.wrap([](lite_api_ptr&& hdr) { + promise.wrap([req_blk_id](lite_api_ptr&& hdr) -> td::Result> { auto blk_id = ton::create_block_id(hdr->id_); - auto R = vm::std_boc_deserialize(std::move(hdr->header_proof_)); - tonlib_api::blocks_header header; - if (R.is_error()) { - LOG(WARNING) << "R.is_error() "; - } else { - auto root = R.move_as_ok(); - try { - ton::RootHash vhash{root->get_hash().bits()}; - auto virt_root = vm::MerkleProof::virtualize(root, 1); - if (virt_root.is_null()) { - LOG(WARNING) << "virt root is null"; - } else { - std::vector prev; - ton::BlockIdExt mc_blkid; - bool after_split; - auto res = block::unpack_block_prev_blk_ext(virt_root, blk_id, prev, mc_blkid, after_split); - if (res.is_error()) { - LOG(WARNING) << "res.is_error() "; - } else { - block::gen::Block::Record blk; - block::gen::BlockInfo::Record info; - if (!(tlb::unpack_cell(virt_root, blk) && tlb::unpack_cell(blk.info, info))) { - LOG(WARNING) << "unpack failed"; - } else { - header.id_ = to_tonlib_api(blk_id); - header.global_id_ = blk.global_id; - header.version_ = info.version; - header.flags_ = info.flags; - header.after_merge_ = info.after_merge; - header.after_split_ = info.after_split; - header.before_split_ = info.before_split; - header.want_merge_ = info.want_merge; - header.want_split_ = info.want_split; - header.validator_list_hash_short_ = info.gen_validator_list_hash_short; - header.catchain_seqno_ = info.gen_catchain_seqno; - header.min_ref_mc_seqno_ = info.min_ref_mc_seqno; - header.start_lt_ = info.start_lt; - header.end_lt_ = info.end_lt; - header.gen_utime_ = info.gen_utime; - header.is_key_block_ = info.key_block; - header.vert_seqno_ = info.vert_seq_no; - if(!info.not_master) { - header.prev_key_block_seqno_ = info.prev_key_block_seqno; - } - for (auto id : prev) { - header.prev_blocks_.push_back(to_tonlib_api(id)); - } - //if(info.before_split) { - //} else { - //} - return tonlib_api::make_object(std::move(header)); - } - } - } - } catch (vm::VmError& err) { - auto E = err.as_status(PSLICE() << "error processing header for " << blk_id.to_str() << " :"); - LOG(ERROR) << std::move(E); - } catch (vm::VmVirtError& err) { - auto E = err.as_status(PSLICE() << "error processing header for " << blk_id.to_str() << " :"); - LOG(ERROR) << std::move(E); - } catch (...) { - LOG(WARNING) << "exception catched "; - } + if (blk_id != req_blk_id) { + return td::Status::Error("Liteserver responded with wrong block"); + } + auto R = vm::std_boc_deserialize(std::move(hdr->header_proof_)); + if (R.is_error()) { + return R.move_as_error_prefix("Couldn't deserialize header proof: "); + } else { + auto root = R.move_as_ok(); + try { + auto virt_root = vm::MerkleProof::virtualize(root, 1); + if (virt_root.is_null()) { + return td::Status::Error("Virt root is null"); + } else { + if (ton::RootHash{virt_root->get_hash().bits()} != blk_id.root_hash) { + return td::Status::Error("Block header merkle proof has incorrect root hash"); + } + std::vector prev; + ton::BlockIdExt mc_blkid; + bool after_split; + auto res = + block::unpack_block_prev_blk_ext(virt_root, blk_id, prev, mc_blkid, after_split); + if (res.is_error()) { + return td::Status::Error("Unpack failed"); + } else { + block::gen::Block::Record blk; + block::gen::BlockInfo::Record info; + if (!(tlb::unpack_cell(virt_root, blk) && tlb::unpack_cell(blk.info, info))) { + return td::Status::Error("Unpack failed"); + } else { + tonlib_api::blocks_header header; + header.id_ = to_tonlib_api(blk_id); + header.global_id_ = blk.global_id; + header.version_ = info.version; + header.flags_ = info.flags; + header.after_merge_ = info.after_merge; + header.after_split_ = info.after_split; + header.before_split_ = info.before_split; + header.want_merge_ = info.want_merge; + header.want_split_ = info.want_split; + header.validator_list_hash_short_ = info.gen_validator_list_hash_short; + header.catchain_seqno_ = info.gen_catchain_seqno; + header.min_ref_mc_seqno_ = info.min_ref_mc_seqno; + header.start_lt_ = info.start_lt; + header.end_lt_ = info.end_lt; + header.gen_utime_ = info.gen_utime; + header.is_key_block_ = info.key_block; + header.vert_seqno_ = info.vert_seq_no; + if (!info.not_master) { + header.prev_key_block_seqno_ = info.prev_key_block_seqno; + } + for (auto& id : prev) { + header.prev_blocks_.push_back(to_tonlib_api(id)); + } + return tonlib_api::make_object(std::move(header)); + } + } + } + } catch (vm::VmError& err) { + return err.as_status(PSLICE() << "error processing header for " << blk_id.to_str() << " :"); + } catch (vm::VmVirtError& err) { + return err.as_status(PSLICE() << "error processing header for " << blk_id.to_str() << " :"); + } catch (...) { + return td::Status::Error("Unhandled exception catched while processing header"); + } } - return tonlib_api::make_object(std::move(header)); })); return td::Status::OK(); } @@ -5764,14 +6033,14 @@ void TonlibClient::load_libs_from_disk() { } libraries = vm::Dictionary(vm::load_cell_slice(vm::CellBuilder().append_cellslice(vm::load_cell_slice( r_dict.move_as_ok())).finalize()), 256); - // int n = 0; for (auto&& lr : libraries) n++; + LOG(DEBUG) << "loaded libraries from disk cache"; } void TonlibClient::store_libs_to_disk() { // NB: Dictionary.get_root_cell does not compute_root, and it is protected kv_->set("tonlib.libcache", vm::std_boc_serialize(vm::CellBuilder().append_cellslice(libraries.get_root()) .finalize()).move_as_ok().as_slice()); - // int n = 0; for (auto&& lr : libraries) n++; + LOG(DEBUG) << "stored libraries to disk cache"; } diff --git a/tonlib/tonlib/TonlibClient.h b/tonlib/tonlib/TonlibClient.h index f4651b9d..001df748 100644 --- a/tonlib/tonlib/TonlibClient.h +++ b/tonlib/tonlib/TonlibClient.h @@ -330,6 +330,7 @@ class TonlibClient : public td::actor::Actor { td::Status do_request(const tonlib_api::smc_getLibraries& request, td::Promise>&& promise); + void get_libraries(ton::BlockIdExt blkid, std::vector library_list_, td::Promise>&& promise); td::Status do_request(const tonlib_api::smc_getLibrariesExt& request, td::Promise>&& promise); diff --git a/tonlib/tonlib/Wallet.cpp b/tonlib/tonlib/Wallet.cpp deleted file mode 100644 index b4682823..00000000 --- a/tonlib/tonlib/Wallet.cpp +++ /dev/null @@ -1,95 +0,0 @@ -/* - This file is part of TON Blockchain Library. - - TON Blockchain Library is free software: you can redistribute it and/or modify - it under the terms of the GNU Lesser General Public License as published by - the Free Software Foundation, either version 2 of the License, or - (at your option) any later version. - - TON Blockchain Library is distributed in the hope that it will be useful, - but WITHOUT ANY WARRANTY; without even the implied warranty of - MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - GNU Lesser General Public License for more details. - - You should have received a copy of the GNU Lesser General Public License - along with TON Blockchain Library. If not, see . - - Copyright 2017-2019 Telegram Systems LLP -*/ -#include "tonlib/Wallet.h" -#include "tonlib/CellString.h" -#include "tonlib/GenericAccount.h" -#include "tonlib/utils.h" - -#include "vm/boc.h" -#include "td/utils/base64.h" - -#include - -namespace tonlib { -td::Ref Wallet::get_init_state(const td::Ed25519::PublicKey& public_key) noexcept { - auto code = get_init_code(); - auto data = get_init_data(public_key); - return GenericAccount::get_init_state(std::move(code), std::move(data)); -} - -td::Ref Wallet::get_init_message(const td::Ed25519::PrivateKey& private_key) noexcept { - td::uint32 seqno = 0; - td::uint32 valid_until = std::numeric_limits::max(); - auto signature = - private_key - .sign(vm::CellBuilder().store_long(seqno, 32).store_long(valid_until, 32).finalize()->get_hash().as_slice()) - .move_as_ok(); - return vm::CellBuilder().store_bytes(signature).store_long(seqno, 32).store_long(valid_until, 32).finalize(); -} - -td::Ref Wallet::make_a_gift_message(const td::Ed25519::PrivateKey& private_key, td::uint32 seqno, - td::uint32 valid_until, td::int64 gramms, td::Slice message, - const block::StdAddress& dest_address) noexcept { - td::BigInt256 dest_addr; - dest_addr.import_bits(dest_address.addr.as_bitslice()); - vm::CellBuilder cb; - cb.append_cellslice(binary_bitstring_to_cellslice("b{01}").move_as_ok()) - .store_long(dest_address.bounceable, 1) - .append_cellslice(binary_bitstring_to_cellslice("b{000100}").move_as_ok()) - .store_long(dest_address.workchain, 8) - .store_int256(dest_addr, 256); - td::int32 send_mode = 3; - if (gramms == -1) { - gramms = 0; - send_mode += 128; - } - block::tlb::t_Grams.store_integer_value(cb, td::BigInt256(gramms)); - cb.store_zeroes(9 + 64 + 32 + 1 + 1).store_bytes("\0\0\0\0", 4); - vm::CellString::store(cb, message, 35 * 8).ensure(); - auto message_inner = cb.finalize(); - auto message_outer = vm::CellBuilder() - .store_long(seqno, 32) - .store_long(valid_until, 32) - .store_long(send_mode, 8) - .store_ref(message_inner) - .finalize(); - std::string seq_no(4, 0); - auto signature = private_key.sign(message_outer->get_hash().as_slice()).move_as_ok(); - return vm::CellBuilder().store_bytes(signature).append_cellslice(vm::load_cell_slice(message_outer)).finalize(); -} - -td::Ref Wallet::get_init_code() noexcept { - static auto res = [] { - auto serialized_code = td::base64_decode( - "te6ccgEEAQEAAAAAVwAAqv8AIN0gggFMl7qXMO1E0NcLH+Ck8mCDCNcYINMf0x8B+CO78mPtRNDTH9P/" - "0VExuvKhA/kBVBBC+RDyovgAApMg10qW0wfUAvsA6NGkyMsfy//J7VQ=") - .move_as_ok(); - return vm::std_boc_deserialize(serialized_code).move_as_ok(); - }(); - return res; -} - -vm::CellHash Wallet::get_init_code_hash() noexcept { - return get_init_code()->get_hash(); -} - -td::Ref Wallet::get_init_data(const td::Ed25519::PublicKey& public_key) noexcept { - return vm::CellBuilder().store_long(0, 32).store_bytes(public_key.as_octet_string()).finalize(); -} -} // namespace tonlib diff --git a/tonlib/tonlib/Wallet.h b/tonlib/tonlib/Wallet.h deleted file mode 100644 index dd114cce..00000000 --- a/tonlib/tonlib/Wallet.h +++ /dev/null @@ -1,40 +0,0 @@ -/* - This file is part of TON Blockchain Library. - - TON Blockchain Library is free software: you can redistribute it and/or modify - it under the terms of the GNU Lesser General Public License as published by - the Free Software Foundation, either version 2 of the License, or - (at your option) any later version. - - TON Blockchain Library is distributed in the hope that it will be useful, - but WITHOUT ANY WARRANTY; without even the implied warranty of - MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - GNU Lesser General Public License for more details. - - You should have received a copy of the GNU Lesser General Public License - along with TON Blockchain Library. If not, see . - - Copyright 2017-2019 Telegram Systems LLP -*/ -#pragma once - -#include "vm/cells.h" -#include "Ed25519.h" -#include "block/block.h" -#include "CellString.h" - -namespace tonlib { -class Wallet { - public: - static constexpr unsigned max_message_size = vm::CellString::max_bytes; - static td::Ref get_init_state(const td::Ed25519::PublicKey& public_key) noexcept; - static td::Ref get_init_message(const td::Ed25519::PrivateKey& private_key) noexcept; - static td::Ref make_a_gift_message(const td::Ed25519::PrivateKey& private_key, td::uint32 seqno, - td::uint32 valid_until, td::int64 gramms, td::Slice message, - const block::StdAddress& dest_address) noexcept; - - static td::Ref get_init_code() noexcept; - static vm::CellHash get_init_code_hash() noexcept; - static td::Ref get_init_data(const td::Ed25519::PublicKey& public_key) noexcept; -}; -} // namespace tonlib diff --git a/validator-engine-console/validator-engine-console-query.cpp b/validator-engine-console/validator-engine-console-query.cpp index bd13225a..956c23aa 100644 --- a/validator-engine-console/validator-engine-console-query.cpp +++ b/validator-engine-console/validator-engine-console-query.cpp @@ -33,6 +33,8 @@ #include "td/utils/filesystem.h" #include "overlay/overlays.h" #include "ton/ton-tl.hpp" +#include "td/utils/JsonBuilder.h" +#include "auto/tl/ton_api_json.h" #include #include @@ -1107,3 +1109,73 @@ td::Status SetExtMessagesBroadcastDisabledQuery::receive(td::BufferSlice data) { td::TerminalIO::out() << "success\n"; return td::Status::OK(); } + +td::Status AddCustomOverlayQuery::run() { + TRY_RESULT_ASSIGN(file_name_, tokenizer_.get_token()); + TRY_STATUS(tokenizer_.check_endl()); + return td::Status::OK(); +} + +td::Status AddCustomOverlayQuery::send() { + TRY_RESULT(data, td::read_file(file_name_)); + TRY_RESULT(json, td::json_decode(data.as_slice())); + auto overlay = ton::create_tl_object(); + TRY_STATUS(ton::ton_api::from_json(*overlay, json.get_object())); + auto b = ton::create_serialize_tl_object(std::move(overlay)); + td::actor::send_closure(console_, &ValidatorEngineConsole::envelope_send_query, std::move(b), create_promise()); + return td::Status::OK(); +} + +td::Status AddCustomOverlayQuery::receive(td::BufferSlice data) { + TRY_RESULT_PREFIX(f, ton::fetch_tl_object(data.as_slice(), true), + "received incorrect answer: "); + td::TerminalIO::out() << "success\n"; + return td::Status::OK(); +} + +td::Status DelCustomOverlayQuery::run() { + TRY_RESULT_ASSIGN(name_, tokenizer_.get_token()); + TRY_STATUS(tokenizer_.check_endl()); + return td::Status::OK(); +} + +td::Status DelCustomOverlayQuery::send() { + auto b = ton::create_serialize_tl_object(name_); + td::actor::send_closure(console_, &ValidatorEngineConsole::envelope_send_query, std::move(b), create_promise()); + return td::Status::OK(); +} + +td::Status DelCustomOverlayQuery::receive(td::BufferSlice data) { + TRY_RESULT_PREFIX(f, ton::fetch_tl_object(data.as_slice(), true), + "received incorrect answer: "); + td::TerminalIO::out() << "success\n"; + return td::Status::OK(); +} + +td::Status ShowCustomOverlaysQuery::run() { + TRY_STATUS(tokenizer_.check_endl()); + return td::Status::OK(); +} + +td::Status ShowCustomOverlaysQuery::send() { + auto b = ton::create_serialize_tl_object(); + td::actor::send_closure(console_, &ValidatorEngineConsole::envelope_send_query, std::move(b), create_promise()); + return td::Status::OK(); +} + +td::Status ShowCustomOverlaysQuery::receive(td::BufferSlice data) { + TRY_RESULT_PREFIX(f, ton::fetch_tl_object(data.as_slice(), true), + "received incorrect answer: "); + td::TerminalIO::out() << f->overlays_.size() << " custom overlays:\n\n"; + for (const auto &overlay : f->overlays_) { + td::TerminalIO::out() << "Overlay \"" << overlay->name_ << "\": " << overlay->nodes_.size() << " nodes\n"; + for (const auto &node : overlay->nodes_) { + td::TerminalIO::out() << " " << node->adnl_id_ + << (node->msg_sender_ ? (PSTRING() << " (sender, p=" << node->msg_sender_priority_ << ")") + : "") + << "\n"; + } + td::TerminalIO::out() << "\n"; + } + return td::Status::OK(); +} diff --git a/validator-engine-console/validator-engine-console-query.h b/validator-engine-console/validator-engine-console-query.h index b1bdac7c..34b516d6 100644 --- a/validator-engine-console/validator-engine-console-query.h +++ b/validator-engine-console/validator-engine-console-query.h @@ -1144,3 +1144,66 @@ class SetExtMessagesBroadcastDisabledQuery : public Query { private: bool value; }; + +class AddCustomOverlayQuery : public Query { + public: + AddCustomOverlayQuery(td::actor::ActorId console, Tokenizer tokenizer) + : Query(console, std::move(tokenizer)) { + } + td::Status run() override; + td::Status send() override; + td::Status receive(td::BufferSlice data) override; + static std::string get_name() { + return "addcustomoverlay"; + } + static std::string get_help() { + return "addcustomoverlay \tadd custom overlay with config from file "; + } + std::string name() const override { + return get_name(); + } + + private: + std::string file_name_; +}; + +class DelCustomOverlayQuery : public Query { + public: + DelCustomOverlayQuery(td::actor::ActorId console, Tokenizer tokenizer) + : Query(console, std::move(tokenizer)) { + } + td::Status run() override; + td::Status send() override; + td::Status receive(td::BufferSlice data) override; + static std::string get_name() { + return "delcustomoverlay"; + } + static std::string get_help() { + return "delcustomoverlay \tdelete custom overlay with name "; + } + std::string name() const override { + return get_name(); + } + + private: + std::string name_; +}; + +class ShowCustomOverlaysQuery : public Query { + public: + ShowCustomOverlaysQuery(td::actor::ActorId console, Tokenizer tokenizer) + : Query(console, std::move(tokenizer)) { + } + td::Status run() override; + td::Status send() override; + td::Status receive(td::BufferSlice data) override; + static std::string get_name() { + return "showcustomoverlays"; + } + static std::string get_help() { + return "showcustomoverlays\tshow all custom overlays"; + } + std::string name() const override { + return get_name(); + } +}; diff --git a/validator-engine-console/validator-engine-console.cpp b/validator-engine-console/validator-engine-console.cpp index 01acced9..263bca3c 100644 --- a/validator-engine-console/validator-engine-console.cpp +++ b/validator-engine-console/validator-engine-console.cpp @@ -143,6 +143,9 @@ void ValidatorEngineConsole::run() { add_query_runner(std::make_unique>()); add_query_runner(std::make_unique>()); add_query_runner(std::make_unique>()); + add_query_runner(std::make_unique>()); + add_query_runner(std::make_unique>()); + add_query_runner(std::make_unique>()); } bool ValidatorEngineConsole::envelope_send_query(td::BufferSlice query, td::Promise promise) { diff --git a/validator-engine/validator-engine.cpp b/validator-engine/validator-engine.cpp index 96103cd0..8ba99178 100644 --- a/validator-engine/validator-engine.cpp +++ b/validator-engine/validator-engine.cpp @@ -1367,6 +1367,8 @@ td::Status ValidatorEngine::load_global_config() { validator_options_.write().set_celldb_compress_depth(celldb_compress_depth_); validator_options_.write().set_max_open_archive_files(max_open_archive_files_); validator_options_.write().set_archive_preload_period(archive_preload_period_); + validator_options_.write().set_disable_rocksdb_stats(disable_rocksdb_stats_); + validator_options_.write().set_nonfinal_ls_queries_enabled(nonfinal_ls_queries_enabled_); std::vector h; for (auto &x : conf.validator_->hardforks_) { @@ -1841,6 +1843,7 @@ void ValidatorEngine::start_full_node() { config_.full_node_config, keyring_.get(), adnl_.get(), rldp_.get(), rldp2_.get(), default_dht_node_.is_zero() ? td::actor::ActorId{} : dht_nodes_[default_dht_node_].get(), overlay_manager_.get(), validator_manager_.get(), full_node_client_.get(), db_root_); + load_custom_overlays_config(); } for (auto &v : config_.validators) { @@ -2333,6 +2336,66 @@ void ValidatorEngine::try_del_proxy(td::uint32 ip, td::int32 port, std::vector(); + auto data_R = td::read_file(custom_overlays_config_file()); + if (data_R.is_error()) { + return; + } + auto data = data_R.move_as_ok(); + auto json_R = td::json_decode(data.as_slice()); + if (json_R.is_error()) { + LOG(ERROR) << "Failed to parse custom overlays config: " << json_R.move_as_error(); + return; + } + auto json = json_R.move_as_ok(); + auto S = ton::ton_api::from_json(*custom_overlays_config_, json.get_object()); + if (S.is_error()) { + LOG(ERROR) << "Failed to parse custom overlays config: " << S; + return; + } + + for (auto &overlay : custom_overlays_config_->overlays_) { + std::vector nodes; + std::map senders; + for (const auto &node : overlay->nodes_) { + nodes.emplace_back(node->adnl_id_); + if (node->msg_sender_) { + senders[ton::adnl::AdnlNodeIdShort{node->adnl_id_}] = node->msg_sender_priority_; + } + } + td::actor::send_closure(full_node_, &ton::validator::fullnode::FullNode::add_ext_msg_overlay, std::move(nodes), + std::move(senders), overlay->name_, [](td::Result R) { R.ensure(); }); + } +} + +td::Status ValidatorEngine::write_custom_overlays_config() { + auto s = td::json_encode(td::ToJson(*custom_overlays_config_), true); + TRY_STATUS_PREFIX(td::write_file(custom_overlays_config_file(), s), "failed to write config: "); + return td::Status::OK(); +} + +void ValidatorEngine::add_custom_overlay_to_config( + ton::tl_object_ptr overlay, td::Promise promise) { + custom_overlays_config_->overlays_.push_back(std::move(overlay)); + TRY_STATUS_PROMISE(promise, write_custom_overlays_config()); + promise.set_result(td::Unit()); +} + +void ValidatorEngine::del_custom_overlay_from_config(std::string name, td::Promise promise) { + auto &overlays = custom_overlays_config_->overlays_; + for (size_t i = 0; i < overlays.size(); ++i) { + if (overlays[i]->name_ == name) { + overlays.erase(overlays.begin() + i); + TRY_STATUS_PROMISE(promise, write_custom_overlays_config()); + promise.set_result(td::Unit()); + return; + } + } + promise.set_error(td::Status::Error(PSTRING() << "no overlay \"" << name << "\" in config")); +} + void ValidatorEngine::check_key(ton::PublicKeyHash id, td::Promise promise) { if (keys_.count(id) == 1) { promise.set_value(td::Unit()); @@ -3460,7 +3523,8 @@ void ValidatorEngine::run_control_query(ton::ton_api::engine_validator_getShardO }); } -void ValidatorEngine::run_control_query(ton::ton_api::engine_validator_setExtMessagesBroadcastDisabled &query, td::BufferSlice data, ton::PublicKeyHash src, td::uint32 perm, +void ValidatorEngine::run_control_query(ton::ton_api::engine_validator_setExtMessagesBroadcastDisabled &query, + td::BufferSlice data, ton::PublicKeyHash src, td::uint32 perm, td::Promise promise) { if (!(perm & ValidatorEnginePermissions::vep_modify)) { promise.set_value(create_control_query_error(td::Status::Error(ton::ErrorCode::error, "not authorized"))); @@ -3486,6 +3550,95 @@ void ValidatorEngine::run_control_query(ton::ton_api::engine_validator_setExtMes }); } +void ValidatorEngine::run_control_query(ton::ton_api::engine_validator_addCustomOverlay &query, + td::BufferSlice data, ton::PublicKeyHash src, td::uint32 perm, + td::Promise promise) { + if (!(perm & ValidatorEnginePermissions::vep_modify)) { + promise.set_value(create_control_query_error(td::Status::Error(ton::ErrorCode::error, "not authorized"))); + return; + } + if (!started_ || full_node_.empty()) { + promise.set_value(create_control_query_error(td::Status::Error(ton::ErrorCode::notready, "not started"))); + return; + } + + auto &overlay = query.overlay_; + std::vector nodes; + std::map senders; + for (const auto &node : overlay->nodes_) { + nodes.emplace_back(node->adnl_id_); + if (node->msg_sender_) { + senders[ton::adnl::AdnlNodeIdShort{node->adnl_id_}] = node->msg_sender_priority_; + } + } + std::string name = overlay->name_; + td::actor::send_closure( + full_node_, &ton::validator::fullnode::FullNode::add_ext_msg_overlay, std::move(nodes), std::move(senders), + std::move(name), + [SelfId = actor_id(this), overlay = std::move(overlay), + promise = std::move(promise)](td::Result R) mutable { + if (R.is_error()) { + promise.set_value(create_control_query_error(R.move_as_error())); + return; + } + td::actor::send_closure( + SelfId, &ValidatorEngine::add_custom_overlay_to_config, std::move(overlay), + [promise = std::move(promise)](td::Result R) mutable { + if (R.is_error()) { + promise.set_value(create_control_query_error(R.move_as_error())); + return; + } + promise.set_value(ton::create_serialize_tl_object()); + }); + }); +} + +void ValidatorEngine::run_control_query(ton::ton_api::engine_validator_delCustomOverlay &query, + td::BufferSlice data, ton::PublicKeyHash src, td::uint32 perm, + td::Promise promise) { + if (!(perm & ValidatorEnginePermissions::vep_modify)) { + promise.set_value(create_control_query_error(td::Status::Error(ton::ErrorCode::error, "not authorized"))); + return; + } + if (!started_ || full_node_.empty()) { + promise.set_value(create_control_query_error(td::Status::Error(ton::ErrorCode::notready, "not started"))); + return; + } + td::actor::send_closure( + full_node_, &ton::validator::fullnode::FullNode::del_ext_msg_overlay, query.name_, + [SelfId = actor_id(this), name = query.name_, promise = std::move(promise)](td::Result R) mutable { + if (R.is_error()) { + promise.set_value(create_control_query_error(R.move_as_error())); + return; + } + td::actor::send_closure( + SelfId, &ValidatorEngine::del_custom_overlay_from_config, std::move(name), + [promise = std::move(promise)](td::Result R) mutable { + if (R.is_error()) { + promise.set_value(create_control_query_error(R.move_as_error())); + return; + } + promise.set_value(ton::create_serialize_tl_object()); + }); + }); +} + +void ValidatorEngine::run_control_query(ton::ton_api::engine_validator_showCustomOverlays &query, + td::BufferSlice data, ton::PublicKeyHash src, td::uint32 perm, + td::Promise promise) { + if (!(perm & ValidatorEnginePermissions::vep_default)) { + promise.set_value(create_control_query_error(td::Status::Error(ton::ErrorCode::error, "not authorized"))); + return; + } + if (!started_ || full_node_.empty()) { + promise.set_value(create_control_query_error(td::Status::Error(ton::ErrorCode::notready, "not started"))); + return; + } + + promise.set_value(ton::serialize_tl_object( + custom_overlays_config_, true)); +} + void ValidatorEngine::process_control_query(td::uint16 port, ton::adnl::AdnlNodeIdShort src, ton::adnl::AdnlNodeIdShort dst, td::BufferSlice data, td::Promise promise) { @@ -3816,6 +3969,12 @@ int main(int argc, char *argv[]) { p.add_option('\0', "enable-precompiled-smc", "enable exectuion of precompiled contracts (experimental, disabled by default)", []() { block::precompiled::set_precompiled_execution_enabled(true); }); + p.add_option('\0', "disable-rocksdb-stats", "disable gathering rocksdb statistics (enabled by default)", [&]() { + acts.push_back([&x]() { td::actor::send_closure(x, &ValidatorEngine::set_disable_rocksdb_stats, true); }); + }); + p.add_option('\0', "nonfinal-ls", "enable special LS queries to non-finalized blocks", [&]() { + acts.push_back([&x]() { td::actor::send_closure(x, &ValidatorEngine::set_nonfinal_ls_queries_enabled); }); + }); auto S = p.run(argc, argv); if (S.is_error()) { LOG(ERROR) << "failed to parse options: " << S.move_as_error(); diff --git a/validator-engine/validator-engine.hpp b/validator-engine/validator-engine.hpp index da50c71f..0a99dfcc 100644 --- a/validator-engine/validator-engine.hpp +++ b/validator-engine/validator-engine.hpp @@ -167,6 +167,7 @@ class ValidatorEngine : public td::actor::Actor { std::shared_ptr dht_config_; td::Ref validator_options_; Config config_; + ton::tl_object_ptr custom_overlays_config_; std::set running_gc_; @@ -206,6 +207,8 @@ class ValidatorEngine : public td::actor::Actor { td::uint32 celldb_compress_depth_ = 0; size_t max_open_archive_files_ = 0; double archive_preload_period_ = 0.0; + bool disable_rocksdb_stats_ = false; + bool nonfinal_ls_queries_enabled_ = false; bool read_config_ = false; bool started_keyring_ = false; bool started_ = false; @@ -272,6 +275,12 @@ class ValidatorEngine : public td::actor::Actor { void set_archive_preload_period(double value) { archive_preload_period_ = value; } + void set_disable_rocksdb_stats(bool value) { + disable_rocksdb_stats_ = value; + } + void set_nonfinal_ls_queries_enabled() { + nonfinal_ls_queries_enabled_ = true; + } void start_up() override; ValidatorEngine() { } @@ -354,6 +363,16 @@ class ValidatorEngine : public td::actor::Actor { void try_del_proxy(td::uint32 ip, td::int32 port, std::vector cats, std::vector prio_cats, td::Promise promise); + std::string custom_overlays_config_file() const { + return db_root_ + "/custom-overlays.json"; + } + + void load_custom_overlays_config(); + td::Status write_custom_overlays_config(); + void add_custom_overlay_to_config( + ton::tl_object_ptr overlay, td::Promise promise); + void del_custom_overlay_from_config(std::string name, td::Promise promise); + void check_key(ton::PublicKeyHash id, td::Promise promise); static td::BufferSlice create_control_query_error(td::Status error); @@ -432,6 +451,12 @@ class ValidatorEngine : public td::actor::Actor { ton::PublicKeyHash src, td::uint32 perm, td::Promise promise); void run_control_query(ton::ton_api::engine_validator_setExtMessagesBroadcastDisabled &query, td::BufferSlice data, ton::PublicKeyHash src, td::uint32 perm, td::Promise promise); + void run_control_query(ton::ton_api::engine_validator_addCustomOverlay &query, td::BufferSlice data, + ton::PublicKeyHash src, td::uint32 perm, td::Promise promise); + void run_control_query(ton::ton_api::engine_validator_delCustomOverlay &query, td::BufferSlice data, + ton::PublicKeyHash src, td::uint32 perm, td::Promise promise); + void run_control_query(ton::ton_api::engine_validator_showCustomOverlays &query, td::BufferSlice data, + ton::PublicKeyHash src, td::uint32 perm, td::Promise promise); template void run_control_query(T &query, td::BufferSlice data, ton::PublicKeyHash src, td::uint32 perm, td::Promise promise) { diff --git a/validator-session/CMakeLists.txt b/validator-session/CMakeLists.txt index c769f4d8..41911a8b 100644 --- a/validator-session/CMakeLists.txt +++ b/validator-session/CMakeLists.txt @@ -5,12 +5,14 @@ if (NOT OPENSSL_FOUND) endif() set(VALIDATOR_SESSION_SOURCE + candidate-serializer.cpp persistent-vector.cpp validator-session-description.cpp validator-session-state.cpp validator-session.cpp validator-session-round-attempt-state.cpp + candidate-serializer.h persistent-vector.h validator-session-description.h validator-session-description.hpp diff --git a/validator-session/candidate-serializer.cpp b/validator-session/candidate-serializer.cpp new file mode 100644 index 00000000..c220b4a9 --- /dev/null +++ b/validator-session/candidate-serializer.cpp @@ -0,0 +1,76 @@ +/* + This file is part of TON Blockchain Library. + + TON Blockchain Library is free software: you can redistribute it and/or modify + it under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation, either version 2 of the License, or + (at your option) any later version. + + TON Blockchain Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with TON Blockchain Library. If not, see . +*/ +#pragma once +#include "candidate-serializer.h" +#include "tl-utils/tl-utils.hpp" +#include "vm/boc.h" +#include "td/utils/lz4.h" +#include "validator-session-types.h" + +namespace ton::validatorsession { + +td::Result serialize_candidate(const tl_object_ptr &block, + bool compression_enabled) { + if (!compression_enabled) { + return serialize_tl_object(block, true); + } + vm::BagOfCells boc1, boc2; + TRY_STATUS(boc1.deserialize(block->data_)); + if (boc1.get_root_count() != 1) { + return td::Status::Error("block candidate should have exactly one root"); + } + std::vector> roots = {boc1.get_root_cell()}; + TRY_STATUS(boc2.deserialize(block->collated_data_)); + for (int i = 0; i < boc2.get_root_count(); ++i) { + roots.push_back(boc2.get_root_cell(i)); + } + TRY_RESULT(data, vm::std_boc_serialize_multi(std::move(roots), 2)); + td::BufferSlice compressed = td::lz4_compress(data); + LOG(VALIDATOR_SESSION_DEBUG) << "Compressing block candidate: " << block->data_.size() + block->collated_data_.size() + << " -> " << compressed.size(); + return create_serialize_tl_object( + 0, block->src_, block->round_, block->root_hash_, (int)data.size(), std::move(compressed)); +} + +td::Result> deserialize_candidate(td::Slice data, + bool compression_enabled, + int max_decompressed_data_size) { + if (!compression_enabled) { + return fetch_tl_object(data, true); + } + TRY_RESULT(f, fetch_tl_object(data, true)); + if (f->decompressed_size_ > max_decompressed_data_size) { + return td::Status::Error("decompressed size is too big"); + } + TRY_RESULT(decompressed, td::lz4_decompress(f->data_, f->decompressed_size_)); + if (decompressed.size() != (size_t)f->decompressed_size_) { + return td::Status::Error("decompressed size mismatch"); + } + TRY_RESULT(roots, vm::std_boc_deserialize_multi(decompressed)); + if (roots.empty()) { + return td::Status::Error("boc is empty"); + } + TRY_RESULT(block_data, vm::std_boc_serialize(roots[0], 31)); + roots.erase(roots.begin()); + TRY_RESULT(collated_data, vm::std_boc_serialize_multi(std::move(roots), 31)); + LOG(VALIDATOR_SESSION_DEBUG) << "Decompressing block candidate: " << f->data_.size() << " -> " + << block_data.size() + collated_data.size(); + return create_tl_object(f->src_, f->round_, f->root_hash_, std::move(block_data), + std::move(collated_data)); +} + +} // namespace ton::validatorsession diff --git a/tonlib/tonlib/TestGiver.h b/validator-session/candidate-serializer.h similarity index 54% rename from tonlib/tonlib/TestGiver.h rename to validator-session/candidate-serializer.h index f8b62599..030a412c 100644 --- a/tonlib/tonlib/TestGiver.h +++ b/validator-session/candidate-serializer.h @@ -13,19 +13,17 @@ You should have received a copy of the GNU Lesser General Public License along with TON Blockchain Library. If not, see . - - Copyright 2017-2019 Telegram Systems LLP */ #pragma once -#include "block/block.h" -#include "CellString.h" -namespace tonlib { -class TestGiver { - public: - static constexpr unsigned max_message_size = vm::CellString::max_bytes; - static const block::StdAddress& address() noexcept; - static vm::CellHash get_init_code_hash() noexcept; - static td::Ref make_a_gift_message(td::uint32 seqno, td::uint64 gramms, td::Slice message, - const block::StdAddress& dest_address) noexcept; -}; -} // namespace tonlib +#include "ton/ton-types.h" +#include "auto/tl/ton_api.h" + +namespace ton::validatorsession { + +td::Result serialize_candidate(const tl_object_ptr &block, + bool compression_enabled); +td::Result> deserialize_candidate(td::Slice data, + bool compression_enabled, + int max_decompressed_data_size); + +} // namespace ton::validatorsession diff --git a/validator-session/validator-session-state.h b/validator-session/validator-session-state.h index f827b3fc..35910535 100644 --- a/validator-session/validator-session-state.h +++ b/validator-session/validator-session-state.h @@ -376,6 +376,15 @@ class ValidatorSessionRoundState : public ValidatorSessionDescription::RootObjec void dump(ValidatorSessionDescription& desc, td::StringBuilder& sb, td::uint32 att) const; void dump_cur_attempt(ValidatorSessionDescription& desc, td::StringBuilder& sb) const; + void for_each_sent_block(std::function foo) const { + if (!sent_blocks_) { + return; + } + for (td::uint32 i = 0; i < sent_blocks_->size(); ++i) { + foo(sent_blocks_->at(i)); + } + } + private: const SentBlock* precommitted_block_; const td::uint32 seqno_; @@ -516,6 +525,19 @@ class ValidatorSessionState : public ValidatorSessionDescription::RootObject { cur_round_->dump_cur_attempt(desc, sb); } + void for_each_cur_round_sent_block(std::function foo) const { + cur_round_->for_each_sent_block(std::move(foo)); + } + + const SentBlock* get_cur_round_precommitted_block() const { + bool found; + return cur_round_->get_precommitted_block(found); + } + + const CntVector* get_cur_round_signatures() const { + return cur_round_->get_signatures(); + } + static const ValidatorSessionState* make_one(ValidatorSessionDescription& desc, const ValidatorSessionState* state, td::uint32 src_idx, td::uint32 att, bool& made); static const ValidatorSessionState* make_all(ValidatorSessionDescription& desc, const ValidatorSessionState* state, diff --git a/validator-session/validator-session.cpp b/validator-session/validator-session.cpp index 88de0fa5..6b328b2b 100644 --- a/validator-session/validator-session.cpp +++ b/validator-session/validator-session.cpp @@ -19,6 +19,7 @@ #include "validator-session.hpp" #include "td/utils/Random.h" #include "td/utils/crypto.h" +#include "candidate-serializer.h" namespace ton { @@ -221,7 +222,9 @@ void ValidatorSessionImpl::process_broadcast(PublicKeyHash src, td::BufferSlice // Note: src is not necessarily equal to the sender of this message: // If requested using get_broadcast_p2p, src is the creator of the block, sender possibly is some other node. auto src_idx = description().get_source_idx(src); - auto R = fetch_tl_object(data.clone(), true); + auto R = + deserialize_candidate(data, compress_block_candidates_, + description().opts().max_block_size + description().opts().max_collated_data_size + 1024); if (R.is_error()) { VLOG(VALIDATOR_SESSION_WARNING) << this << "[node " << src << "][broadcast " << sha256_bits256(data.as_slice()) << "]: failed to parse: " << R.move_as_error(); @@ -343,17 +346,17 @@ void ValidatorSessionImpl::process_query(PublicKeyHash src, td::BufferSlice data } CHECK(block); - auto P = td::PromiseCreator::lambda( - [promise = std::move(promise), src = f->id_->src_, round_id](td::Result R) mutable { - if (R.is_error()) { - promise.set_error(R.move_as_error_prefix("failed to get candidate: ")); - } else { - auto c = R.move_as_ok(); - auto obj = create_tl_object( - src, round_id, c.id.root_hash, std::move(c.data), std::move(c.collated_data)); - promise.set_value(serialize_tl_object(obj, true)); - } - }); + auto P = td::PromiseCreator::lambda([promise = std::move(promise), src = f->id_->src_, round_id, + compress = compress_block_candidates_](td::Result R) mutable { + if (R.is_error()) { + promise.set_error(R.move_as_error_prefix("failed to get candidate: ")); + } else { + auto c = R.move_as_ok(); + auto obj = create_tl_object(src, round_id, c.id.root_hash, std::move(c.data), + std::move(c.collated_data)); + promise.set_result(serialize_candidate(obj, compress)); + } + }); callback_->get_approved_candidate(description().get_source_public_key(block->get_src_idx()), f->id_->root_hash_, f->id_->file_hash_, f->id_->collated_data_file_hash_, std::move(P)); @@ -431,7 +434,7 @@ void ValidatorSessionImpl::generated_block(td::uint32 round, ValidatorSessionCan auto b = create_tl_object(local_id().tl(), round, root_hash, std::move(data), std::move(collated_data)); - auto B = serialize_tl_object(b, true); + auto B = serialize_candidate(b, compress_block_candidates_).move_as_ok(); auto block_id = description().candidate_id(local_idx(), root_hash, file_hash, collated_data_file_hash); @@ -862,7 +865,8 @@ void ValidatorSessionImpl::on_catchain_started() { if (x) { auto P = td::PromiseCreator::lambda([SelfId = actor_id(this), round = virtual_state_->cur_round_seqno(), src = description().get_source_id(x->get_src_idx()), - root_hash = x->get_root_hash()](td::Result R) { + root_hash = x->get_root_hash(), + compress = compress_block_candidates_](td::Result R) { if (R.is_error()) { LOG(ERROR) << "failed to get candidate: " << R.move_as_error(); } else { @@ -870,7 +874,7 @@ void ValidatorSessionImpl::on_catchain_started() { auto broadcast = create_tl_object( src.tl(), round, root_hash, std::move(B.data), std::move(B.collated_data)); td::actor::send_closure(SelfId, &ValidatorSessionImpl::process_broadcast, src, - serialize_tl_object(broadcast, true), td::optional(), + serialize_candidate(broadcast, compress).move_as_ok(), td::optional(), false); } }); @@ -898,6 +902,7 @@ ValidatorSessionImpl::ValidatorSessionImpl(catchain::CatChainSessionId session_i , rldp_(rldp) , overlay_manager_(overlays) , allow_unsafe_self_blocks_resync_(allow_unsafe_self_blocks_resync) { + compress_block_candidates_ = opts.proto_version >= 4; description_ = ValidatorSessionDescription::create(std::move(opts), nodes, local_id); src_round_candidate_.resize(description_->get_total_nodes()); } @@ -929,6 +934,53 @@ void ValidatorSessionImpl::get_current_stats(td::Promise promise.set_result(std::move(stats)); } +void ValidatorSessionImpl::get_validator_group_info_for_litequery( + td::uint32 cur_round, + td::Promise>> promise) { + if (cur_round != cur_round_ || real_state_->cur_round_seqno() != cur_round) { + promise.set_value({}); + return; + } + std::vector> result; + real_state_->for_each_cur_round_sent_block([&](const SessionBlockCandidate *block) { + if (block->get_block() == nullptr) { + return; + } + auto candidate = create_tl_object(); + + candidate->id_ = create_tl_object(); + candidate->id_->block_id_ = create_tl_object(); + candidate->id_->block_id_->root_hash_ = + block->get_block()->get_root_hash(); // other fields will be filled in validator-group.cpp + candidate->id_->block_id_->file_hash_ = block->get_block()->get_file_hash(); + candidate->id_->creator_ = + description().get_source_public_key(block->get_block()->get_src_idx()).ed25519_value().raw(); + candidate->id_->collated_data_hash_ = block->get_block()->get_collated_data_file_hash(); + + candidate->total_weight_ = description().get_total_weight(); + candidate->approved_weight_ = 0; + candidate->signed_weight_ = 0; + for (td::uint32 i = 0; i < description().get_total_nodes(); ++i) { + if (real_state_->check_block_is_approved_by(description(), i, block->get_id())) { + candidate->approved_weight_ += description().get_node_weight(i); + } + } + auto precommited = real_state_->get_cur_round_precommitted_block(); + if (SentBlock::get_block_id(precommited) == SentBlock::get_block_id(block->get_block())) { + auto signatures = real_state_->get_cur_round_signatures(); + if (signatures) { + for (td::uint32 i = 0; i < description().get_total_nodes(); ++i) { + if (signatures->at(i)) { + candidate->signed_weight_ += description().get_node_weight(i); + } + } + } + } + result.push_back(std::move(candidate)); + }); + promise.set_result(std::move(result)); +} + void ValidatorSessionImpl::start_up() { CHECK(!rldp_.empty()); cur_round_ = 0; diff --git a/validator-session/validator-session.h b/validator-session/validator-session.h index 3f3b7ab9..6300eaa4 100644 --- a/validator-session/validator-session.h +++ b/validator-session/validator-session.h @@ -28,6 +28,7 @@ #include "catchain/catchain-types.h" #include "validator-session-types.h" +#include "auto/tl/lite_api.h" namespace ton { @@ -92,6 +93,9 @@ class ValidatorSession : public td::actor::Actor { virtual void start() = 0; virtual void destroy() = 0; virtual void get_current_stats(td::Promise promise) = 0; + virtual void get_validator_group_info_for_litequery( + td::uint32 cur_round, + td::Promise>> promise) = 0; static td::actor::ActorOwn create( catchain::CatChainSessionId session_id, ValidatorSessionOptions opts, PublicKeyHash local_id, diff --git a/validator-session/validator-session.hpp b/validator-session/validator-session.hpp index 2dcbb46c..619f7178 100644 --- a/validator-session/validator-session.hpp +++ b/validator-session/validator-session.hpp @@ -156,6 +156,7 @@ class ValidatorSessionImpl : public ValidatorSession { bool started_ = false; bool catchain_started_ = false; bool allow_unsafe_self_blocks_resync_; + bool compress_block_candidates_ = false; ValidatorSessionStats cur_stats_; void stats_init(); @@ -175,6 +176,9 @@ class ValidatorSessionImpl : public ValidatorSession { void start() override; void destroy() override; void get_current_stats(td::Promise promise) override; + void get_validator_group_info_for_litequery( + td::uint32 cur_round, + td::Promise>> promise) override; void process_blocks(std::vector blocks); void finished_processing(); diff --git a/validator/CMakeLists.txt b/validator/CMakeLists.txt index 832374c6..a25ac7e1 100644 --- a/validator/CMakeLists.txt +++ b/validator/CMakeLists.txt @@ -146,6 +146,8 @@ set(FULL_NODE_SOURCE full-node-master.cpp full-node-private-overlay.hpp full-node-private-overlay.cpp + full-node-serializer.hpp + full-node-serializer.cpp net/download-block.hpp net/download-block.cpp diff --git a/validator/db/archive-manager.cpp b/validator/db/archive-manager.cpp index 2c0c82e5..c806f4e5 100644 --- a/validator/db/archive-manager.cpp +++ b/validator/db/archive-manager.cpp @@ -601,7 +601,7 @@ void ArchiveManager::load_package(PackageId id) { } desc.file = - td::actor::create_actor("slice", id.id, id.key, id.temp, false, db_root_, archive_lru_.get()); + td::actor::create_actor("slice", id.id, id.key, id.temp, false, db_root_, archive_lru_.get(), statistics_); m.emplace(id, std::move(desc)); update_permanent_slices(); @@ -636,7 +636,7 @@ const ArchiveManager::FileDescription *ArchiveManager::add_file_desc(ShardIdFull td::mkdir(db_root_ + id.path()).ensure(); std::string prefix = PSTRING() << db_root_ << id.path() << id.name(); new_desc.file = - td::actor::create_actor("slice", id.id, id.key, id.temp, false, db_root_, archive_lru_.get()); + td::actor::create_actor("slice", id.id, id.key, id.temp, false, db_root_, archive_lru_.get(), statistics_); const FileDescription &desc = f.emplace(id, std::move(new_desc)); if (!id.temp) { update_desc(f, desc, shard, seqno, ts, lt); @@ -829,7 +829,10 @@ void ArchiveManager::start_up() { if (opts_->get_max_open_archive_files() > 0) { archive_lru_ = td::actor::create_actor("archive_lru", opts_->get_max_open_archive_files()); } - index_ = std::make_shared(td::RocksDb::open(db_root_ + "/files/globalindex").move_as_ok()); + if (!opts_->get_disable_rocksdb_stats()) { + statistics_.init(); + } + index_ = std::make_shared(td::RocksDb::open(db_root_ + "/files/globalindex", statistics_.rocksdb_statistics).move_as_ok()); std::string value; auto v = index_->get(create_serialize_tl_object().as_slice(), value); v.ensure(); @@ -903,10 +906,31 @@ void ArchiveManager::start_up() { break; } } + + if (!opts_->get_disable_rocksdb_stats()) { + alarm_timestamp() = td::Timestamp::in(60.0); + } +} + +void ArchiveManager::alarm() { + alarm_timestamp() = td::Timestamp::in(60.0); + auto stats = statistics_.to_string_and_reset(); + auto to_file_r = td::FileFd::open(db_root_ + "/db_stats.txt", td::FileFd::Truncate | td::FileFd::Create | td::FileFd::Write, 0644); + if (to_file_r.is_error()) { + LOG(ERROR) << "Failed to open db_stats.txt: " << to_file_r.move_as_error(); + return; + } + auto to_file = to_file_r.move_as_ok(); + auto res = to_file.write(stats); + to_file.close(); + if (res.is_error()) { + LOG(ERROR) << "Failed to write to db_stats.txt: " << res.move_as_error(); + return; + } } void ArchiveManager::run_gc(UnixTime mc_ts, UnixTime gc_ts, UnixTime archive_ttl) { - auto p = get_temp_package_id_by_unixtime(std::max(gc_ts, mc_ts - TEMP_PACKAGES_TTL)); + auto p = get_temp_package_id_by_unixtime(mc_ts - TEMP_PACKAGES_TTL); std::vector vec; for (auto &x : temp_files_) { if (x.first < p) { diff --git a/validator/db/archive-manager.hpp b/validator/db/archive-manager.hpp index 1c5deaf8..a1ed9702 100644 --- a/validator/db/archive-manager.hpp +++ b/validator/db/archive-manager.hpp @@ -70,6 +70,7 @@ class ArchiveManager : public td::actor::Actor { td::Promise promise); void start_up() override; + void alarm() override; void commit_transaction(); void set_async_mode(bool mode, td::Promise promise); @@ -173,6 +174,8 @@ class ArchiveManager : public td::actor::Actor { bool huge_transaction_started_ = false; td::uint32 huge_transaction_size_ = 0; + DbStatistics statistics_; + FileMap &get_file_map(const PackageId &p) { return p.key ? key_files_ : p.temp ? temp_files_ : files_; } @@ -223,7 +226,7 @@ class ArchiveManager : public td::actor::Actor { void update_permanent_slices(); - static const td::uint32 TEMP_PACKAGES_TTL = 86400 * 7; + static const td::uint32 TEMP_PACKAGES_TTL = 3600; }; } // namespace validator diff --git a/validator/db/archive-slice.cpp b/validator/db/archive-slice.cpp index 52abc008..86c2ee93 100644 --- a/validator/db/archive-slice.cpp +++ b/validator/db/archive-slice.cpp @@ -29,29 +29,151 @@ namespace ton { namespace validator { +class PackageStatistics { + public: + void record_open(uint64_t count = 1) { + open_count.fetch_add(count, std::memory_order_relaxed); + } + + void record_close(uint64_t count = 1) { + close_count.fetch_add(count, std::memory_order_relaxed); + } + + void record_read(double time, uint64_t bytes) { + read_bytes.fetch_add(bytes, std::memory_order_relaxed); + std::lock_guard guard(read_mutex); + read_time.insert(time); + read_time_sum += time; + } + + void record_write(double time, uint64_t bytes) { + write_bytes.fetch_add(bytes, std::memory_order_relaxed); + std::lock_guard guard(write_mutex); + write_time.insert(time); + write_time_sum += time; + } + + std::string to_string_and_reset() { + std::stringstream ss; + ss.setf(std::ios::fixed); + ss.precision(6); + + ss << "ton.pack.open COUNT : " << open_count.exchange(0, std::memory_order_relaxed) << "\n"; + ss << "ton.pack.close COUNT : " << close_count.exchange(0, std::memory_order_relaxed) << "\n"; + + ss << "ton.pack.read.bytes COUNT : " << read_bytes.exchange(0, std::memory_order_relaxed) << "\n"; + ss << "ton.pack.write.bytes COUNT : " << write_bytes.exchange(0, std::memory_order_relaxed) << "\n"; + + std::multiset temp_read_time; + double temp_read_time_sum; + { + std::lock_guard guard(read_mutex); + temp_read_time = std::move(read_time); + read_time.clear(); + temp_read_time_sum = read_time_sum; + read_time_sum = 0; + } + auto read_stats = calculate_statistics(temp_read_time); + ss << "ton.pack.read.micros P50 : " << read_stats[0] << + " P95 : " << read_stats[1] << + " P99 : " << read_stats[2] << + " P100 : " << read_stats[3] << + " COUNT : " << temp_read_time.size() << + " SUM : " << temp_read_time_sum << "\n"; + + std::multiset temp_write_time; + double temp_write_time_sum; + { + std::lock_guard guard(write_mutex); + temp_write_time = std::move(write_time); + write_time.clear(); + temp_write_time_sum = write_time_sum; + write_time_sum = 0; + } + auto write_stats = calculate_statistics(temp_write_time); + ss << "ton.pack.write.micros P50 : " << write_stats[0] << + " P95 : " << write_stats[1] << + " P99 : " << write_stats[2] << + " P100 : " << write_stats[3] << + " COUNT : " << temp_write_time.size() << + " SUM : " << temp_write_time_sum << "\n"; + + return ss.str(); + } + + private: + std::atomic_uint64_t open_count; + std::atomic_uint64_t close_count; + std::multiset read_time; + std::atomic_uint64_t read_bytes; + std::multiset write_time; + std::atomic_uint64_t write_bytes; + double read_time_sum; + double write_time_sum; + + mutable std::mutex read_mutex; + mutable std::mutex write_mutex; + + std::vector calculate_statistics(const std::multiset& data) const { + if (data.empty()) return {0, 0, 0, 0}; + + auto size = data.size(); + auto calc_percentile = [&](double p) -> double { + auto it = data.begin(); + std::advance(it, static_cast(std::ceil(p * double(size)) - 1)); + return *it; + }; + + return {calc_percentile(0.5), calc_percentile(0.95), calc_percentile(0.99), *data.rbegin()}; + } +}; + +void DbStatistics::init() { + rocksdb_statistics = td::RocksDb::create_statistics(); + pack_statistics = std::make_shared(); +} + +std::string DbStatistics::to_string_and_reset() { + std::stringstream ss; + ss << td::RocksDb::statistics_to_string(rocksdb_statistics) << pack_statistics->to_string_and_reset(); + td::RocksDb::reset_statistics(rocksdb_statistics); + return ss.str(); +} + void PackageWriter::append(std::string filename, td::BufferSlice data, td::Promise> promise) { td::uint64 offset, size; + auto data_size = data.size(); + td::Timestamp start, end; { auto p = package_.lock(); if (!p) { promise.set_error(td::Status::Error("Package is closed")); return; } - offset = p->append(std::move(filename), std::move(data), !async_mode_); + start = td::Timestamp::now(); + offset = p->append(std::move(filename), std::move(data), !async_mode_); + end = td::Timestamp::now(); size = p->size(); } + if (statistics_) { + statistics_->record_write((end.at() - start.at()) * 1e6, data_size); + } promise.set_value(std::pair{offset, size}); } class PackageReader : public td::actor::Actor { public: PackageReader(std::shared_ptr package, td::uint64 offset, - td::Promise> promise) - : package_(std::move(package)), offset_(offset), promise_(std::move(promise)) { + td::Promise> promise, std::shared_ptr statistics) + : package_(std::move(package)), offset_(offset), promise_(std::move(promise)), statistics_(std::move(statistics)) { } void start_up() override { + auto start = td::Timestamp::now(); auto result = package_->read(offset_); + if (statistics_ && result.is_ok()) { + statistics_->record_read((td::Timestamp::now().at() - start.at()) * 1e6, result.ok_ref().second.size()); + } package_ = {}; promise_.set_result(std::move(result)); stop(); @@ -61,6 +183,7 @@ class PackageReader : public td::actor::Actor { std::shared_ptr package_; td::uint64 offset_; td::Promise> promise_; + std::shared_ptr statistics_; }; void ArchiveSlice::add_handle(BlockHandle handle, td::Promise promise) { @@ -297,7 +420,7 @@ void ArchiveSlice::get_file(ConstBlockHandle handle, FileReference ref_id, td::P promise.set_value(std::move(R.move_as_ok().second)); } }); - td::actor::create_actor("reader", p->package, offset, std::move(P)).release(); + td::actor::create_actor("reader", p->package, offset, std::move(P), statistics_.pack_statistics).release(); } void ArchiveSlice::get_block_common(AccountIdPrefixFull account_id, @@ -465,7 +588,7 @@ void ArchiveSlice::get_archive_id(BlockSeqno masterchain_seqno, td::Promise(td::RocksDb::open(db_path_).move_as_ok()); + kv_ = std::make_unique(td::RocksDb::open(db_path_, statistics_.rocksdb_statistics).move_as_ok()); std::string value; auto R2 = kv_->get("status", value); R2.ensure(); @@ -547,6 +670,9 @@ void ArchiveSlice::do_close() { LOG(DEBUG) << "Closing archive slice " << db_path_; status_ = st_closed; kv_ = {}; + if (statistics_.pack_statistics) { + statistics_.pack_statistics->record_close(packages_.size()); + } packages_.clear(); } @@ -604,14 +730,15 @@ void ArchiveSlice::set_async_mode(bool mode, td::Promise promise) { } ArchiveSlice::ArchiveSlice(td::uint32 archive_id, bool key_blocks_only, bool temp, bool finalized, std::string db_root, - td::actor::ActorId archive_lru) + td::actor::ActorId archive_lru, DbStatistics statistics) : archive_id_(archive_id) , key_blocks_only_(key_blocks_only) , temp_(temp) , finalized_(finalized) , p_id_(archive_id_, key_blocks_only_, temp_) , db_root_(std::move(db_root)) - , archive_lru_(std::move(archive_lru)) { + , archive_lru_(std::move(archive_lru)) + , statistics_(statistics) { db_path_ = PSTRING() << db_root_ << p_id_.path() << p_id_.name() << ".index"; } @@ -649,6 +776,9 @@ void ArchiveSlice::add_package(td::uint32 seqno, td::uint64 size, td::uint32 ver LOG(FATAL) << "failed to open/create archive '" << path << "': " << R.move_as_error(); return; } + if (statistics_.pack_statistics) { + statistics_.pack_statistics->record_open(); + } auto idx = td::narrow_cast(packages_.size()); if (finalized_) { packages_.emplace_back(nullptr, td::actor::ActorOwn(), seqno, path, idx, version); @@ -658,7 +788,7 @@ void ArchiveSlice::add_package(td::uint32 seqno, td::uint64 size, td::uint32 ver if (version >= 1) { pack->truncate(size).ensure(); } - auto writer = td::actor::create_actor("writer", pack, async_mode_); + auto writer = td::actor::create_actor("writer", pack, async_mode_, statistics_.pack_statistics); packages_.emplace_back(std::move(pack), std::move(writer), seqno, path, idx, version); } @@ -691,7 +821,9 @@ void ArchiveSlice::destroy(td::Promise promise) { for (auto &p : packages_) { td::unlink(p.path).ensure(); } - + if (statistics_.pack_statistics) { + statistics_.pack_statistics->record_close(packages_.size()); + } packages_.clear(); kv_ = nullptr; @@ -897,7 +1029,10 @@ void ArchiveSlice::truncate(BlockSeqno masterchain_seqno, ConstBlockHandle handl for (auto idx = pack->idx + 1; idx < packages_.size(); idx++) { td::unlink(packages_[idx].path).ensure(); } - packages_.erase(packages_.begin() + pack->idx + 1); + if (statistics_.pack_statistics) { + statistics_.pack_statistics->record_close(packages_.size() - pack->idx - 1); + } + packages_.erase(packages_.begin() + pack->idx + 1, packages_.end()); kv_->commit_transaction().ensure(); diff --git a/validator/db/archive-slice.hpp b/validator/db/archive-slice.hpp index f178a9b8..faec2fb8 100644 --- a/validator/db/archive-slice.hpp +++ b/validator/db/archive-slice.hpp @@ -21,8 +21,13 @@ #include "validator/interfaces/db.h" #include "package.hpp" #include "fileref.hpp" +#include "td/db/RocksDb.h" #include +namespace rocksdb { +class Statistics; +} + namespace ton { namespace validator { @@ -53,10 +58,20 @@ struct PackageId { } }; +class PackageStatistics; + +struct DbStatistics { + void init(); + std::string to_string_and_reset(); + + std::shared_ptr pack_statistics; + std::shared_ptr rocksdb_statistics; +}; + class PackageWriter : public td::actor::Actor { public: - PackageWriter(std::weak_ptr package, bool async_mode = false) - : package_(std::move(package)), async_mode_(async_mode) { + PackageWriter(std::weak_ptr package, bool async_mode = false, std::shared_ptr statistics = nullptr) + : package_(std::move(package)), async_mode_(async_mode), statistics_(std::move(statistics)) { } void append(std::string filename, td::BufferSlice data, td::Promise> promise); @@ -74,6 +89,7 @@ class PackageWriter : public td::actor::Actor { private: std::weak_ptr package_; bool async_mode_ = false; + std::shared_ptr statistics_; }; class ArchiveLru; @@ -81,7 +97,7 @@ class ArchiveLru; class ArchiveSlice : public td::actor::Actor { public: ArchiveSlice(td::uint32 archive_id, bool key_blocks_only, bool temp, bool finalized, std::string db_root, - td::actor::ActorId archive_lru); + td::actor::ActorId archive_lru, DbStatistics statistics = {}); void get_archive_id(BlockSeqno masterchain_seqno, td::Promise promise); @@ -151,6 +167,7 @@ class ArchiveSlice : public td::actor::Actor { std::string db_root_; td::actor::ActorId archive_lru_; + DbStatistics statistics_; std::unique_ptr kv_; struct PackageInfo { diff --git a/validator/db/celldb.cpp b/validator/db/celldb.cpp index d29126ce..d9803cbe 100644 --- a/validator/db/celldb.cpp +++ b/validator/db/celldb.cpp @@ -20,6 +20,7 @@ #include "rootdb.hpp" #include "td/db/RocksDb.h" +#include "td/utils/filesystem.h" #include "ton/ton-tl.hpp" #include "ton/ton-io.hpp" @@ -83,7 +84,12 @@ void CellDbIn::start_up() { }; CellDbBase::start_up(); - cell_db_ = std::make_shared(td::RocksDb::open(path_).move_as_ok()); + if (!opts_->get_disable_rocksdb_stats()) { + statistics_ = td::RocksDb::create_statistics(); + statistics_flush_at_ = td::Timestamp::in(60.0); + } + cell_db_ = std::make_shared(td::RocksDb::open(path_, statistics_).move_as_ok()); + boc_ = vm::DynamicBagOfCellsDb::create(); boc_->set_celldb_compress_depth(opts_->get_celldb_compress_depth()); @@ -155,7 +161,29 @@ void CellDbIn::get_cell_db_reader(td::Promise> promise.set_result(boc_->get_cell_db_reader()); } +void CellDbIn::flush_db_stats() { + auto stats = td::RocksDb::statistics_to_string(statistics_); + auto to_file_r = td::FileFd::open(path_ + "/db_stats.txt", td::FileFd::Truncate | td::FileFd::Create | td::FileFd::Write, 0644); + if (to_file_r.is_error()) { + LOG(ERROR) << "Failed to open db_stats.txt: " << to_file_r.move_as_error(); + return; + } + auto to_file = to_file_r.move_as_ok(); + auto res = to_file.write(stats); + to_file.close(); + if (res.is_error()) { + LOG(ERROR) << "Failed to write to db_stats.txt: " << res.move_as_error(); + return; + } + td::RocksDb::reset_statistics(statistics_); +} + void CellDbIn::alarm() { + if (statistics_flush_at_ && statistics_flush_at_.is_in_past()) { + statistics_flush_at_ = td::Timestamp::in(60.0); + flush_db_stats(); + } + if (migrate_after_ && migrate_after_.is_in_past()) { migrate_cells(); } diff --git a/validator/db/celldb.hpp b/validator/db/celldb.hpp index a2a84ab4..573d4b99 100644 --- a/validator/db/celldb.hpp +++ b/validator/db/celldb.hpp @@ -27,6 +27,10 @@ #include "auto/tl/ton_api.h" #include "validator.h" +namespace rocksdb { +class Statistics; +} + namespace ton { namespace validator { @@ -56,6 +60,8 @@ class CellDbIn : public CellDbBase { void migrate_cell(td::Bits256 hash); + void flush_db_stats(); + CellDbIn(td::actor::ActorId root_db, td::actor::ActorId parent, std::string path, td::Ref opts); @@ -103,6 +109,8 @@ class CellDbIn : public CellDbBase { std::unique_ptr boc_; std::shared_ptr cell_db_; + std::shared_ptr statistics_; + td::Timestamp statistics_flush_at_ = td::Timestamp::never(); std::function on_load_callback_; std::set cells_to_migrate_; diff --git a/validator/full-node-private-overlay.cpp b/validator/full-node-private-overlay.cpp index ea72230b..a64c0e9b 100644 --- a/validator/full-node-private-overlay.cpp +++ b/validator/full-node-private-overlay.cpp @@ -14,32 +14,30 @@ You should have received a copy of the GNU Lesser General Public License along with TON Blockchain Library. If not, see . */ -#pragma once - #include "full-node-private-overlay.hpp" #include "ton/ton-tl.hpp" #include "common/delay.h" +#include "full-node-serializer.hpp" -namespace ton { +namespace ton::validator::fullnode { -namespace validator { +void FullNodePrivateBlockOverlay::process_broadcast(PublicKeyHash src, ton_api::tonNode_blockBroadcast &query) { + process_block_broadcast(src, query); +} -namespace fullnode { +void FullNodePrivateBlockOverlay::process_broadcast(PublicKeyHash src, + ton_api::tonNode_blockBroadcastCompressed &query) { + process_block_broadcast(src, query); +} -void FullNodePrivateOverlay::process_broadcast(PublicKeyHash, ton_api::tonNode_blockBroadcast &query) { - std::vector signatures; - for (auto &sig : query.signatures_) { - signatures.emplace_back(BlockSignature{sig->who_, std::move(sig->signature_)}); +void FullNodePrivateBlockOverlay::process_block_broadcast(PublicKeyHash src, ton_api::tonNode_Broadcast &query) { + auto B = deserialize_block_broadcast(query, overlay::Overlays::max_fec_broadcast_size()); + if (B.is_error()) { + LOG(DEBUG) << "dropped broadcast: " << B.move_as_error(); + return; } - - BlockIdExt block_id = create_block_id(query.id_); - BlockBroadcast B{block_id, - std::move(signatures), - static_cast(query.catchain_seqno_), - static_cast(query.validator_set_hash_), - std::move(query.data_), - std::move(query.proof_)}; - + VLOG(FULL_NODE_DEBUG) << "Received block broadcast in private overlay from " << src << ": " + << B.ok().block_id.to_str(); auto P = td::PromiseCreator::lambda([](td::Result R) { if (R.is_error()) { if (R.error().code() == ErrorCode::notready) { @@ -49,29 +47,32 @@ void FullNodePrivateOverlay::process_broadcast(PublicKeyHash, ton_api::tonNode_b } } }); - td::actor::send_closure(validator_manager_, &ValidatorManagerInterface::prevalidate_block, std::move(B), + td::actor::send_closure(validator_manager_, &ValidatorManagerInterface::prevalidate_block, B.move_as_ok(), std::move(P)); } -void FullNodePrivateOverlay::process_broadcast(PublicKeyHash, ton_api::tonNode_newShardBlockBroadcast &query) { - td::actor::send_closure(validator_manager_, &ValidatorManagerInterface::new_shard_block, - create_block_id(query.block_->block_), query.block_->cc_seqno_, - std::move(query.block_->data_)); +void FullNodePrivateBlockOverlay::process_broadcast(PublicKeyHash src, ton_api::tonNode_newShardBlockBroadcast &query) { + BlockIdExt block_id = create_block_id(query.block_->block_); + VLOG(FULL_NODE_DEBUG) << "Received newShardBlockBroadcast in private overlay from " << src << ": " + << block_id.to_str(); + td::actor::send_closure(validator_manager_, &ValidatorManagerInterface::new_shard_block, block_id, + query.block_->cc_seqno_, std::move(query.block_->data_)); } -void FullNodePrivateOverlay::receive_broadcast(PublicKeyHash src, td::BufferSlice broadcast) { +void FullNodePrivateBlockOverlay::receive_broadcast(PublicKeyHash src, td::BufferSlice broadcast) { auto B = fetch_tl_object(std::move(broadcast), true); if (B.is_error()) { return; } - ton_api::downcast_call(*B.move_as_ok(), [src, Self = this](auto &obj) { Self->process_broadcast(src, obj); }); } -void FullNodePrivateOverlay::send_shard_block_info(BlockIdExt block_id, CatchainSeqno cc_seqno, td::BufferSlice data) { +void FullNodePrivateBlockOverlay::send_shard_block_info(BlockIdExt block_id, CatchainSeqno cc_seqno, + td::BufferSlice data) { if (!inited_) { return; } + VLOG(FULL_NODE_DEBUG) << "Sending newShardBlockBroadcast in private overlay: " << block_id.to_str(); auto B = create_serialize_tl_object( create_tl_object(create_tl_block_id(block_id), cc_seqno, std::move(data))); if (B.size() <= overlay::Overlays::max_simple_broadcast_size()) { @@ -83,22 +84,22 @@ void FullNodePrivateOverlay::send_shard_block_info(BlockIdExt block_id, Catchain } } -void FullNodePrivateOverlay::send_broadcast(BlockBroadcast broadcast) { +void FullNodePrivateBlockOverlay::send_broadcast(BlockBroadcast broadcast) { if (!inited_) { return; } - std::vector> sigs; - for (auto &sig : broadcast.signatures) { - sigs.emplace_back(create_tl_object(sig.node, sig.signature.clone())); + VLOG(FULL_NODE_DEBUG) << "Sending block broadcast in private overlay" + << (enable_compression_ ? " (with compression)" : "") << ": " << broadcast.block_id.to_str(); + auto B = serialize_block_broadcast(broadcast, enable_compression_); + if (B.is_error()) { + VLOG(FULL_NODE_WARNING) << "failed to serialize block broadcast: " << B.move_as_error(); + return; } - auto B = create_serialize_tl_object( - create_tl_block_id(broadcast.block_id), broadcast.catchain_seqno, broadcast.validator_set_hash, std::move(sigs), - broadcast.proof.clone(), broadcast.data.clone()); td::actor::send_closure(overlays_, &overlay::Overlays::send_broadcast_fec_ex, local_id_, overlay_id_, - local_id_.pubkey_hash(), overlay::Overlays::BroadcastFlagAnySender(), std::move(B)); + local_id_.pubkey_hash(), overlay::Overlays::BroadcastFlagAnySender(), B.move_as_ok()); } -void FullNodePrivateOverlay::start_up() { +void FullNodePrivateBlockOverlay::start_up() { std::sort(nodes_.begin(), nodes_.end()); nodes_.erase(std::unique(nodes_.begin(), nodes_.end()), nodes_.end()); @@ -115,22 +116,22 @@ void FullNodePrivateOverlay::start_up() { try_init(); } -void FullNodePrivateOverlay::try_init() { +void FullNodePrivateBlockOverlay::try_init() { // Sometimes adnl id is added to validator engine later (or not at all) td::actor::send_closure( adnl_, &adnl::Adnl::check_id_exists, local_id_, [SelfId = actor_id(this)](td::Result R) { if (R.is_ok() && R.ok()) { - td::actor::send_closure(SelfId, &FullNodePrivateOverlay::init); + td::actor::send_closure(SelfId, &FullNodePrivateBlockOverlay::init); } else { - delay_action([SelfId]() { td::actor::send_closure(SelfId, &FullNodePrivateOverlay::try_init); }, + delay_action([SelfId]() { td::actor::send_closure(SelfId, &FullNodePrivateBlockOverlay::try_init); }, td::Timestamp::in(30.0)); } }); } -void FullNodePrivateOverlay::init() { - LOG(FULL_NODE_INFO) << "Creating private block overlay for adnl id " << local_id_ << " : " << nodes_.size() - << " nodes"; +void FullNodePrivateBlockOverlay::init() { + LOG(FULL_NODE_WARNING) << "Creating private block overlay for adnl id " << local_id_ << " : " << nodes_.size() + << " nodes, overlay_id=" << overlay_id_; class Callback : public overlay::Overlays::Callback { public: void receive_message(adnl::AdnlNodeIdShort src, overlay::OverlayIdShort overlay_id, td::BufferSlice data) override { @@ -139,37 +140,139 @@ void FullNodePrivateOverlay::init() { td::Promise promise) override { } void receive_broadcast(PublicKeyHash src, overlay::OverlayIdShort overlay_id, td::BufferSlice data) override { - td::actor::send_closure(node_, &FullNodePrivateOverlay::receive_broadcast, src, std::move(data)); + td::actor::send_closure(node_, &FullNodePrivateBlockOverlay::receive_broadcast, src, std::move(data)); } void check_broadcast(PublicKeyHash src, overlay::OverlayIdShort overlay_id, td::BufferSlice data, td::Promise promise) override { } - Callback(td::actor::ActorId node) : node_(node) { + Callback(td::actor::ActorId node) : node_(node) { } private: - td::actor::ActorId node_; + td::actor::ActorId node_; }; overlay::OverlayPrivacyRules rules{overlay::Overlays::max_fec_broadcast_size(), overlay::CertificateFlags::AllowFec | overlay::CertificateFlags::Trusted, {}}; td::actor::send_closure(overlays_, &overlay::Overlays::create_private_overlay, local_id_, overlay_id_full_.clone(), - nodes_, std::make_unique(actor_id(this)), rules); + nodes_, std::make_unique(actor_id(this)), rules, R"({ "type": "private-blocks" })"); td::actor::send_closure(rldp_, &rldp::Rldp::add_id, local_id_); td::actor::send_closure(rldp2_, &rldp2::Rldp::add_id, local_id_); inited_ = true; } -void FullNodePrivateOverlay::tear_down() { +void FullNodePrivateBlockOverlay::tear_down() { if (inited_) { td::actor::send_closure(overlays_, &ton::overlay::Overlays::delete_overlay, local_id_, overlay_id_); } } -} // namespace fullnode +void FullNodeCustomOverlay::process_broadcast(PublicKeyHash src, ton_api::tonNode_externalMessageBroadcast &query) { + auto it = senders_.find(adnl::AdnlNodeIdShort{src}); + if (it == senders_.end()) { + return; + } + LOG(FULL_NODE_DEBUG) << "Got external message in private overlay \"" << name_ << "\" from " << src + << " (priority=" << it->second << ")"; + td::actor::send_closure(validator_manager_, &ValidatorManagerInterface::new_external_message, + std::move(query.message_->data_), it->second); +} -} // namespace validator +void FullNodeCustomOverlay::receive_broadcast(PublicKeyHash src, td::BufferSlice broadcast) { + auto B = fetch_tl_object(std::move(broadcast), true); + if (B.is_error()) { + return; + } + ton_api::downcast_call(*B.move_as_ok(), [src, Self = this](auto &obj) { Self->process_broadcast(src, obj); }); +} -} // namespace ton +void FullNodeCustomOverlay::send_external_message(td::BufferSlice data) { + if (!inited_ || config_.ext_messages_broadcast_disabled_) { + return; + } + LOG(FULL_NODE_DEBUG) << "Sending external message to private overlay \"" << name_ << "\""; + auto B = create_serialize_tl_object( + create_tl_object(std::move(data))); + if (B.size() <= overlay::Overlays::max_simple_broadcast_size()) { + td::actor::send_closure(overlays_, &overlay::Overlays::send_broadcast_ex, local_id_, overlay_id_, + local_id_.pubkey_hash(), 0, std::move(B)); + } else { + td::actor::send_closure(overlays_, &overlay::Overlays::send_broadcast_fec_ex, local_id_, overlay_id_, + local_id_.pubkey_hash(), 0, std::move(B)); + } +} + +void FullNodeCustomOverlay::start_up() { + std::sort(nodes_.begin(), nodes_.end()); + nodes_.erase(std::unique(nodes_.begin(), nodes_.end()), nodes_.end()); + std::vector nodes; + for (const adnl::AdnlNodeIdShort &id : nodes_) { + nodes.push_back(id.bits256_value()); + } + auto X = create_hash_tl_object(zero_state_file_hash_, name_, std::move(nodes)); + td::BufferSlice b{32}; + b.as_slice().copy_from(as_slice(X)); + overlay_id_full_ = overlay::OverlayIdFull{std::move(b)}; + overlay_id_ = overlay_id_full_.compute_short_id(); + try_init(); +} + +void FullNodeCustomOverlay::try_init() { + // Sometimes adnl id is added to validator engine later (or not at all) + td::actor::send_closure( + adnl_, &adnl::Adnl::check_id_exists, local_id_, [SelfId = actor_id(this)](td::Result R) { + if (R.is_ok() && R.ok()) { + td::actor::send_closure(SelfId, &FullNodeCustomOverlay::init); + } else { + delay_action([SelfId]() { td::actor::send_closure(SelfId, &FullNodeCustomOverlay::try_init); }, + td::Timestamp::in(30.0)); + } + }); +} + +void FullNodeCustomOverlay::init() { + LOG(FULL_NODE_WARNING) << "Creating custom overlay \"" << name_ << "\" for adnl id " << local_id_ << " : " + << nodes_.size() << " nodes, overlay_id=" << overlay_id_; + class Callback : public overlay::Overlays::Callback { + public: + void receive_message(adnl::AdnlNodeIdShort src, overlay::OverlayIdShort overlay_id, td::BufferSlice data) override { + } + void receive_query(adnl::AdnlNodeIdShort src, overlay::OverlayIdShort overlay_id, td::BufferSlice data, + td::Promise promise) override { + } + void receive_broadcast(PublicKeyHash src, overlay::OverlayIdShort overlay_id, td::BufferSlice data) override { + td::actor::send_closure(node_, &FullNodeCustomOverlay::receive_broadcast, src, std::move(data)); + } + void check_broadcast(PublicKeyHash src, overlay::OverlayIdShort overlay_id, td::BufferSlice data, + td::Promise promise) override { + } + Callback(td::actor::ActorId node) : node_(node) { + } + + private: + td::actor::ActorId node_; + }; + + std::map authorized_keys; + for (const auto &sender : senders_) { + authorized_keys[sender.first.pubkey_hash()] = overlay::Overlays::max_fec_broadcast_size(); + } + overlay::OverlayPrivacyRules rules{overlay::Overlays::max_fec_broadcast_size(), 0, std::move(authorized_keys)}; + td::actor::send_closure( + overlays_, &overlay::Overlays::create_private_overlay, local_id_, overlay_id_full_.clone(), nodes_, + std::make_unique(actor_id(this)), rules, + PSTRING() << R"({ "type": "custom-overlay", "name": ")" << td::format::Escaped{name_} << R"(" })"); + + td::actor::send_closure(rldp_, &rldp::Rldp::add_id, local_id_); + td::actor::send_closure(rldp2_, &rldp2::Rldp::add_id, local_id_); + inited_ = true; +} + +void FullNodeCustomOverlay::tear_down() { + LOG(FULL_NODE_WARNING) << "Destroying custom overlay \"" << name_ << "\" for adnl id " << local_id_; + td::actor::send_closure(overlays_, &ton::overlay::Overlays::delete_overlay, local_id_, overlay_id_); +} + +} // namespace ton::validator::fullnode diff --git a/validator/full-node-private-overlay.hpp b/validator/full-node-private-overlay.hpp index 6463fda2..c651acef 100644 --- a/validator/full-node-private-overlay.hpp +++ b/validator/full-node-private-overlay.hpp @@ -18,15 +18,14 @@ #include "full-node.h" -namespace ton { +namespace ton::validator::fullnode { -namespace validator { - -namespace fullnode { - -class FullNodePrivateOverlay : public td::actor::Actor { +class FullNodePrivateBlockOverlay : public td::actor::Actor { public: void process_broadcast(PublicKeyHash src, ton_api::tonNode_blockBroadcast &query); + void process_broadcast(PublicKeyHash src, ton_api::tonNode_blockBroadcastCompressed &query); + void process_block_broadcast(PublicKeyHash src, ton_api::tonNode_Broadcast &query); + void process_broadcast(PublicKeyHash src, ton_api::tonNode_newShardBlockBroadcast &query); template void process_broadcast(PublicKeyHash, T &) { @@ -37,19 +36,88 @@ class FullNodePrivateOverlay : public td::actor::Actor { void send_shard_block_info(BlockIdExt block_id, CatchainSeqno cc_seqno, td::BufferSlice data); void send_broadcast(BlockBroadcast broadcast); + void set_config(FullNodeConfig config) { + config_ = std::move(config); + } + + void set_enable_compression(bool value) { + enable_compression_ = value; + } + void start_up() override; void tear_down() override; - FullNodePrivateOverlay(adnl::AdnlNodeIdShort local_id, std::vector nodes, - FileHash zero_state_file_hash, FullNodeConfig config, - td::actor::ActorId keyring, td::actor::ActorId adnl, - td::actor::ActorId rldp, td::actor::ActorId rldp2, - td::actor::ActorId overlays, - td::actor::ActorId validator_manager) + FullNodePrivateBlockOverlay(adnl::AdnlNodeIdShort local_id, std::vector nodes, + FileHash zero_state_file_hash, FullNodeConfig config, bool enable_compression, + td::actor::ActorId keyring, td::actor::ActorId adnl, + td::actor::ActorId rldp, td::actor::ActorId rldp2, + td::actor::ActorId overlays, + td::actor::ActorId validator_manager) : local_id_(local_id) , nodes_(std::move(nodes)) , zero_state_file_hash_(zero_state_file_hash) , config_(config) + , enable_compression_(enable_compression) + , keyring_(keyring) + , adnl_(adnl) + , rldp_(rldp) + , rldp2_(rldp2) + , overlays_(overlays) + , validator_manager_(validator_manager) { + } + + private: + adnl::AdnlNodeIdShort local_id_; + std::vector nodes_; + FileHash zero_state_file_hash_; + FullNodeConfig config_; + bool enable_compression_; + + td::actor::ActorId keyring_; + td::actor::ActorId adnl_; + td::actor::ActorId rldp_; + td::actor::ActorId rldp2_; + td::actor::ActorId overlays_; + td::actor::ActorId validator_manager_; + + bool inited_ = false; + overlay::OverlayIdFull overlay_id_full_; + overlay::OverlayIdShort overlay_id_; + + void try_init(); + void init(); +}; + +class FullNodeCustomOverlay : public td::actor::Actor { + public: + void process_broadcast(PublicKeyHash src, ton_api::tonNode_externalMessageBroadcast &query); + template + void process_broadcast(PublicKeyHash, T &) { + VLOG(FULL_NODE_WARNING) << "dropping unknown broadcast"; + } + void receive_broadcast(PublicKeyHash src, td::BufferSlice query); + + void send_external_message(td::BufferSlice data); + + void set_config(FullNodeConfig config) { + config_ = std::move(config); + } + + void start_up() override; + void tear_down() override; + + FullNodeCustomOverlay(adnl::AdnlNodeIdShort local_id, std::vector nodes, + std::map senders, std::string name, FileHash zero_state_file_hash, + FullNodeConfig config, td::actor::ActorId keyring, + td::actor::ActorId adnl, td::actor::ActorId rldp, + td::actor::ActorId rldp2, td::actor::ActorId overlays, + td::actor::ActorId validator_manager) + : local_id_(local_id) + , nodes_(std::move(nodes)) + , senders_(std::move(senders)) + , name_(std::move(name)) + , zero_state_file_hash_(zero_state_file_hash) + , config_(config) , keyring_(keyring) , adnl_(adnl) , rldp_(rldp) @@ -61,6 +129,8 @@ class FullNodePrivateOverlay : public td::actor::Actor { private: adnl::AdnlNodeIdShort local_id_; std::vector nodes_; + std::map senders_; + std::string name_; FileHash zero_state_file_hash_; FullNodeConfig config_; @@ -79,8 +149,4 @@ class FullNodePrivateOverlay : public td::actor::Actor { void init(); }; -} // namespace fullnode - -} // namespace validator - -} // namespace ton +} // namespace ton::validator::fullnode diff --git a/validator/full-node-serializer.cpp b/validator/full-node-serializer.cpp new file mode 100644 index 00000000..42e68286 --- /dev/null +++ b/validator/full-node-serializer.cpp @@ -0,0 +1,155 @@ +/* + This file is part of TON Blockchain Library. + + TON Blockchain Library is free software: you can redistribute it and/or modify + it under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation, either version 2 of the License, or + (at your option) any later version. + + TON Blockchain Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with TON Blockchain Library. If not, see . +*/ +#include "full-node-serializer.hpp" +#include "ton/ton-tl.hpp" +#include "tl-utils/common-utils.hpp" +#include "auto/tl/ton_api.hpp" +#include "tl-utils/tl-utils.hpp" +#include "vm/boc.h" +#include "td/utils/lz4.h" +#include "full-node.h" +#include "td/utils/overloaded.h" + +namespace ton::validator::fullnode { + +td::Result serialize_block_broadcast(const BlockBroadcast& broadcast, bool compression_enabled) { + std::vector> sigs; + for (auto& sig : broadcast.signatures) { + sigs.emplace_back(create_tl_object(sig.node, sig.signature.clone())); + } + if (!compression_enabled) { + return create_serialize_tl_object( + create_tl_block_id(broadcast.block_id), broadcast.catchain_seqno, broadcast.validator_set_hash, std::move(sigs), + broadcast.proof.clone(), broadcast.data.clone()); + } + + TRY_RESULT(proof_root, vm::std_boc_deserialize(broadcast.proof)); + TRY_RESULT(data_root, vm::std_boc_deserialize(broadcast.data)); + TRY_RESULT(boc, vm::std_boc_serialize_multi({proof_root, data_root}, 2)); + td::BufferSlice data = + create_serialize_tl_object(std::move(sigs), std::move(boc)); + td::BufferSlice compressed = td::lz4_compress(data); + VLOG(FULL_NODE_DEBUG) << "Compressing block broadcast: " + << broadcast.data.size() + broadcast.proof.size() + broadcast.signatures.size() * 96 << " -> " + << compressed.size(); + return create_serialize_tl_object( + create_tl_block_id(broadcast.block_id), broadcast.catchain_seqno, broadcast.validator_set_hash, 0, + std::move(compressed)); +} + +static td::Result deserialize_block_broadcast(ton_api::tonNode_blockBroadcast& f) { + std::vector signatures; + for (auto& sig : f.signatures_) { + signatures.emplace_back(BlockSignature{sig->who_, std::move(sig->signature_)}); + } + return BlockBroadcast{create_block_id(f.id_), + std::move(signatures), + static_cast(f.catchain_seqno_), + static_cast(f.validator_set_hash_), + std::move(f.data_), + std::move(f.proof_)}; +} + +static td::Result deserialize_block_broadcast(ton_api::tonNode_blockBroadcastCompressed& f, + int max_decompressed_size) { + TRY_RESULT(decompressed, td::lz4_decompress(f.compressed_, max_decompressed_size)); + TRY_RESULT(f2, fetch_tl_object(decompressed, true)); + std::vector signatures; + for (auto& sig : f2->signatures_) { + signatures.emplace_back(BlockSignature{sig->who_, std::move(sig->signature_)}); + } + TRY_RESULT(roots, vm::std_boc_deserialize_multi(f2->proof_data_, 2)); + if (roots.size() != 2) { + return td::Status::Error("expected 2 roots in boc"); + } + TRY_RESULT(proof, vm::std_boc_serialize(roots[0], 0)); + TRY_RESULT(data, vm::std_boc_serialize(roots[1], 31)); + VLOG(FULL_NODE_DEBUG) << "Decompressing block broadcast: " << f.compressed_.size() << " -> " + << data.size() + proof.size() + signatures.size() * 96; + return BlockBroadcast{create_block_id(f.id_), + std::move(signatures), + static_cast(f.catchain_seqno_), + static_cast(f.validator_set_hash_), + std::move(data), + std::move(proof)}; +} + +td::Result deserialize_block_broadcast(ton_api::tonNode_Broadcast& obj, + int max_decompressed_data_size) { + td::Result B; + ton_api::downcast_call(obj, + td::overloaded([&](ton_api::tonNode_blockBroadcast& f) { B = deserialize_block_broadcast(f); }, + [&](ton_api::tonNode_blockBroadcastCompressed& f) { + B = deserialize_block_broadcast(f, max_decompressed_data_size); + }, + [&](auto&) { B = td::Status::Error("unknown broadcast type"); })); + return B; +} + +td::Result serialize_block_full(const BlockIdExt& id, td::Slice proof, td::Slice data, + bool is_proof_link, bool compression_enabled) { + if (!compression_enabled) { + return create_serialize_tl_object(create_tl_block_id(id), td::BufferSlice(proof), + td::BufferSlice(data), is_proof_link); + } + TRY_RESULT(proof_root, vm::std_boc_deserialize(proof)); + TRY_RESULT(data_root, vm::std_boc_deserialize(data)); + TRY_RESULT(boc, vm::std_boc_serialize_multi({proof_root, data_root}, 2)); + td::BufferSlice compressed = td::lz4_compress(boc); + VLOG(FULL_NODE_DEBUG) << "Compressing block full: " << data.size() + proof.size() << " -> " << compressed.size(); + return create_serialize_tl_object(create_tl_block_id(id), 0, + std::move(compressed), is_proof_link); +} + +static td::Status deserialize_block_full(ton_api::tonNode_dataFull& f, BlockIdExt& id, td::BufferSlice& proof, + td::BufferSlice& data, bool& is_proof_link) { + id = create_block_id(f.id_); + proof = std::move(f.proof_); + data = std::move(f.block_); + is_proof_link = f.is_link_; + return td::Status::OK(); +} + +static td::Status deserialize_block_full(ton_api::tonNode_dataFullCompressed& f, BlockIdExt& id, td::BufferSlice& proof, + td::BufferSlice& data, bool& is_proof_link, int max_decompressed_size) { + TRY_RESULT(decompressed, td::lz4_decompress(f.compressed_, max_decompressed_size)); + TRY_RESULT(roots, vm::std_boc_deserialize_multi(decompressed, 2)); + if (roots.size() != 2) { + return td::Status::Error("expected 2 roots in boc"); + } + TRY_RESULT_ASSIGN(proof, vm::std_boc_serialize(roots[0], 0)); + TRY_RESULT_ASSIGN(data, vm::std_boc_serialize(roots[1], 31)); + VLOG(FULL_NODE_DEBUG) << "Decompressing block full: " << f.compressed_.size() << " -> " << data.size() + proof.size(); + id = create_block_id(f.id_); + is_proof_link = f.is_link_; + return td::Status::OK(); +} + +td::Status deserialize_block_full(ton_api::tonNode_DataFull& obj, BlockIdExt& id, td::BufferSlice& proof, + td::BufferSlice& data, bool& is_proof_link, int max_decompressed_data_size) { + td::Status S; + ton_api::downcast_call( + obj, td::overloaded( + [&](ton_api::tonNode_dataFull& f) { S = deserialize_block_full(f, id, proof, data, is_proof_link); }, + [&](ton_api::tonNode_dataFullCompressed& f) { + S = deserialize_block_full(f, id, proof, data, is_proof_link, max_decompressed_data_size); + }, + [&](auto&) { S = td::Status::Error("unknown data type"); })); + return S; +} + +} // namespace ton::validator::fullnode diff --git a/validator/full-node-serializer.hpp b/validator/full-node-serializer.hpp new file mode 100644 index 00000000..a5c73cbc --- /dev/null +++ b/validator/full-node-serializer.hpp @@ -0,0 +1,31 @@ +/* + This file is part of TON Blockchain Library. + + TON Blockchain Library is free software: you can redistribute it and/or modify + it under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation, either version 2 of the License, or + (at your option) any later version. + + TON Blockchain Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with TON Blockchain Library. If not, see . +*/ +#pragma once +#include "ton/ton-types.h" +#include "auto/tl/ton_api.h" + +namespace ton::validator::fullnode { + +td::Result serialize_block_broadcast(const BlockBroadcast& broadcast, bool compression_enabled); +td::Result deserialize_block_broadcast(ton_api::tonNode_Broadcast& obj, int max_decompressed_data_size); + +td::Result serialize_block_full(const BlockIdExt& id, td::Slice proof, td::Slice data, + bool is_proof_link, bool compression_enabled); +td::Status deserialize_block_full(ton_api::tonNode_DataFull& obj, BlockIdExt& id, td::BufferSlice& proof, + td::BufferSlice& data, bool& is_proof_link, int max_decompressed_data_size); + +} // namespace ton::validator::fullnode diff --git a/validator/full-node-shard-queries.hpp b/validator/full-node-shard-queries.hpp index 1a2cd716..17068229 100644 --- a/validator/full-node-shard-queries.hpp +++ b/validator/full-node-shard-queries.hpp @@ -20,6 +20,7 @@ #include "validator/validator.h" #include "ton/ton-tl.hpp" +#include "full-node-serializer.hpp" namespace ton { @@ -38,8 +39,8 @@ class BlockFullSender : public td::actor::Actor { stop(); } void finish_query() { - promise_.set_value(create_serialize_tl_object( - create_tl_block_id(block_id_), std::move(proof_), std::move(data_), is_proof_link_)); + promise_.set_result( + serialize_block_full(block_id_, proof_, data_, is_proof_link_, false)); // compression_enabled = false stop(); } void start_up() override { diff --git a/validator/full-node-shard.cpp b/validator/full-node-shard.cpp index 8e42716b..433d9469 100644 --- a/validator/full-node-shard.cpp +++ b/validator/full-node-shard.cpp @@ -21,6 +21,7 @@ #include "td/utils/SharedSlice.h" #include "full-node-shard.hpp" #include "full-node-shard-queries.hpp" +#include "full-node-serializer.hpp" #include "ton/ton-shard.h" #include "ton/ton-tl.hpp" @@ -117,7 +118,7 @@ void FullNodeShardImpl::check_broadcast(PublicKeyHash src, td::BufferSlice broad promise.set_error(td::Status::Error("rebroadcasting external messages is disabled")); promise = [manager = validator_manager_, message = q->message_->data_.clone()](td::Result R) mutable { if (R.is_ok()) { - td::actor::send_closure(manager, &ValidatorManagerInterface::new_external_message, std::move(message)); + td::actor::send_closure(manager, &ValidatorManagerInterface::new_external_message, std::move(message), 0); } }; } @@ -635,29 +636,31 @@ void FullNodeShardImpl::process_broadcast(PublicKeyHash src, ton_api::tonNode_ih void FullNodeShardImpl::process_broadcast(PublicKeyHash src, ton_api::tonNode_externalMessageBroadcast &query) { td::actor::send_closure(validator_manager_, &ValidatorManagerInterface::new_external_message, - std::move(query.message_->data_)); + std::move(query.message_->data_), 0); } void FullNodeShardImpl::process_broadcast(PublicKeyHash src, ton_api::tonNode_newShardBlockBroadcast &query) { - td::actor::send_closure(validator_manager_, &ValidatorManagerInterface::new_shard_block, - create_block_id(query.block_->block_), query.block_->cc_seqno_, - std::move(query.block_->data_)); + BlockIdExt block_id = create_block_id(query.block_->block_); + VLOG(FULL_NODE_DEBUG) << "Received newShardBlockBroadcast from " << src << ": " << block_id.to_str(); + td::actor::send_closure(validator_manager_, &ValidatorManagerInterface::new_shard_block, block_id, + query.block_->cc_seqno_, std::move(query.block_->data_)); } void FullNodeShardImpl::process_broadcast(PublicKeyHash src, ton_api::tonNode_blockBroadcast &query) { - std::vector signatures; - for (auto &sig : query.signatures_) { - signatures.emplace_back(BlockSignature{sig->who_, std::move(sig->signature_)}); + process_block_broadcast(src, query); +} + +void FullNodeShardImpl::process_broadcast(PublicKeyHash src, ton_api::tonNode_blockBroadcastCompressed &query) { + process_block_broadcast(src, query); +} + +void FullNodeShardImpl::process_block_broadcast(PublicKeyHash src, ton_api::tonNode_Broadcast &query) { + auto B = deserialize_block_broadcast(query, overlay::Overlays::max_fec_broadcast_size()); + if (B.is_error()) { + LOG(DEBUG) << "dropped broadcast: " << B.move_as_error(); + return; } - - BlockIdExt block_id = create_block_id(query.id_); - BlockBroadcast B{block_id, - std::move(signatures), - static_cast(query.catchain_seqno_), - static_cast(query.validator_set_hash_), - std::move(query.data_), - std::move(query.proof_)}; - + VLOG(FULL_NODE_DEBUG) << "Received block broadcast from " << src << ": " << B.ok().block_id.to_str(); auto P = td::PromiseCreator::lambda([](td::Result R) { if (R.is_error()) { if (R.error().code() == ErrorCode::notready) { @@ -667,7 +670,7 @@ void FullNodeShardImpl::process_broadcast(PublicKeyHash src, ton_api::tonNode_bl } } }); - td::actor::send_closure(validator_manager_, &ValidatorManagerInterface::prevalidate_block, std::move(B), + td::actor::send_closure(validator_manager_, &ValidatorManagerInterface::prevalidate_block, B.move_as_ok(), std::move(P)); } @@ -733,6 +736,7 @@ void FullNodeShardImpl::send_shard_block_info(BlockIdExt block_id, CatchainSeqno UNREACHABLE(); return; } + VLOG(FULL_NODE_DEBUG) << "Sending newShardBlockBroadcast: " << block_id.to_str(); auto B = create_serialize_tl_object( create_tl_object(create_tl_block_id(block_id), cc_seqno, std::move(data))); if (B.size() <= overlay::Overlays::max_simple_broadcast_size()) { @@ -749,15 +753,14 @@ void FullNodeShardImpl::send_broadcast(BlockBroadcast broadcast) { UNREACHABLE(); return; } - std::vector> sigs; - for (auto &sig : broadcast.signatures) { - sigs.emplace_back(create_tl_object(sig.node, sig.signature.clone())); + VLOG(FULL_NODE_DEBUG) << "Sending block broadcast in private overlay: " << broadcast.block_id.to_str(); + auto B = serialize_block_broadcast(broadcast, false); // compression_enabled = false + if (B.is_error()) { + VLOG(FULL_NODE_WARNING) << "failed to serialize block broadcast: " << B.move_as_error(); + return; } - auto B = create_serialize_tl_object( - create_tl_block_id(broadcast.block_id), broadcast.catchain_seqno, broadcast.validator_set_hash, std::move(sigs), - broadcast.proof.clone(), broadcast.data.clone()); td::actor::send_closure(overlays_, &overlay::Overlays::send_broadcast_fec_ex, adnl_id_, overlay_id_, local_id_, - overlay::Overlays::BroadcastFlagAnySender(), std::move(B)); + overlay::Overlays::BroadcastFlagAnySender(), B.move_as_ok()); } void FullNodeShardImpl::download_block(BlockIdExt id, td::uint32 priority, td::Timestamp timeout, diff --git a/validator/full-node-shard.hpp b/validator/full-node-shard.hpp index 0525474e..c8b5301a 100644 --- a/validator/full-node-shard.hpp +++ b/validator/full-node-shard.hpp @@ -71,7 +71,7 @@ class FullNodeShardImpl : public FullNodeShard { return 2; } static constexpr td::uint64 proto_capabilities() { - return 2; + return 3; } static constexpr td::uint32 max_neighbours() { return 16; @@ -146,6 +146,9 @@ class FullNodeShardImpl : public FullNodeShard { void receive_query(adnl::AdnlNodeIdShort src, td::BufferSlice query, td::Promise promise); void process_broadcast(PublicKeyHash src, ton_api::tonNode_blockBroadcast &query); + void process_broadcast(PublicKeyHash src, ton_api::tonNode_blockBroadcastCompressed &query); + void process_block_broadcast(PublicKeyHash src, ton_api::tonNode_Broadcast &query); + void process_broadcast(PublicKeyHash src, ton_api::tonNode_ihrMessageBroadcast &query); void process_broadcast(PublicKeyHash src, ton_api::tonNode_externalMessageBroadcast &query); void process_broadcast(PublicKeyHash src, ton_api::tonNode_newShardBlockBroadcast &query); diff --git a/validator/full-node.cpp b/validator/full-node.cpp index 5a822b26..affb8bcc 100644 --- a/validator/full-node.cpp +++ b/validator/full-node.cpp @@ -35,6 +35,10 @@ void FullNodeImpl::add_permanent_key(PublicKeyHash key, td::Promise pr } local_keys_.insert(key); + create_private_block_overlay(key); + for (auto &p : private_custom_overlays_) { + update_ext_msg_overlay(p.first, p.second); + } if (!sign_cert_by_.is_zero()) { promise.set_value(td::Unit()); @@ -50,7 +54,6 @@ void FullNodeImpl::add_permanent_key(PublicKeyHash key, td::Promise pr for (auto &shard : shards_) { td::actor::send_closure(shard.second, &FullNodeShard::update_validators, all_validators_, sign_cert_by_); } - create_private_block_overlay(key); promise.set_value(td::Unit()); } @@ -60,6 +63,11 @@ void FullNodeImpl::del_permanent_key(PublicKeyHash key, td::Promise pr return; } local_keys_.erase(key); + private_block_overlays_.erase(key); + for (auto &p : private_custom_overlays_) { + update_ext_msg_overlay(p.first, p.second); + } + if (sign_cert_by_ != key) { promise.set_value(td::Unit()); return; @@ -75,7 +83,6 @@ void FullNodeImpl::del_permanent_key(PublicKeyHash key, td::Promise pr for (auto &shard : shards_) { td::actor::send_closure(shard.second, &FullNodeShard::update_validators, all_validators_, sign_cert_by_); } - private_block_overlays_.erase(key); promise.set_value(td::Unit()); } @@ -111,6 +118,10 @@ void FullNodeImpl::update_adnl_id(adnl::AdnlNodeIdShort adnl_id, td::Promise nodes, + std::map senders, std::string name, + td::Promise promise) { + if (nodes.empty()) { + promise.set_error(td::Status::Error("list of nodes is empty")); + return; + } + if (private_custom_overlays_.count(name)) { + promise.set_error(td::Status::Error(PSTRING() << "duplicate overlay name \"" << name << "\"")); + return; + } + VLOG(FULL_NODE_WARNING) << "Adding private overlay for external messages \"" << name << "\", " << nodes.size() + << " nodes"; + auto &p = private_custom_overlays_[name]; + p.nodes_ = nodes; + p.senders_ = senders; + update_ext_msg_overlay(name, p); + promise.set_result(td::Unit()); +} + +void FullNodeImpl::del_ext_msg_overlay(std::string name, td::Promise promise) { + auto it = private_custom_overlays_.find(name); + if (it == private_custom_overlays_.end()) { + promise.set_error(td::Status::Error(PSTRING() << "no such overlay \"" << name << "\"")); + return; + } + private_custom_overlays_.erase(it); + promise.set_result(td::Unit()); } void FullNodeImpl::initial_read_complete(BlockHandle top_handle) { @@ -172,6 +221,14 @@ void FullNodeImpl::send_ext_message(AccountIdPrefixFull dst, td::BufferSlice dat VLOG(FULL_NODE_WARNING) << "dropping OUT ext message to unknown shard"; return; } + for (auto &private_overlay : private_custom_overlays_) { + for (auto &actor : private_overlay.second.actors_) { + auto local_id = actor.first; + if (private_overlay.second.senders_.count(local_id)) { + td::actor::send_closure(actor.second, &FullNodeCustomOverlay::send_external_message, data.clone()); + } + } + } td::actor::send_closure(shard, &FullNodeShard::send_external_message, std::move(data)); } @@ -182,8 +239,8 @@ void FullNodeImpl::send_shard_block_info(BlockIdExt block_id, CatchainSeqno cc_s return; } if (!private_block_overlays_.empty()) { - td::actor::send_closure(private_block_overlays_.begin()->second, &FullNodePrivateOverlay::send_shard_block_info, - block_id, cc_seqno, data.clone()); + td::actor::send_closure(private_block_overlays_.begin()->second, + &FullNodePrivateBlockOverlay::send_shard_block_info, block_id, cc_seqno, data.clone()); } td::actor::send_closure(shard, &FullNodeShard::send_shard_block_info, block_id, cc_seqno, std::move(data)); } @@ -195,7 +252,7 @@ void FullNodeImpl::send_broadcast(BlockBroadcast broadcast) { return; } if (!private_block_overlays_.empty()) { - td::actor::send_closure(private_block_overlays_.begin()->second, &FullNodePrivateOverlay::send_broadcast, + td::actor::send_closure(private_block_overlays_.begin()->second, &FullNodePrivateBlockOverlay::send_broadcast, broadcast.clone()); } td::actor::send_closure(shard, &FullNodeShard::send_broadcast, std::move(broadcast)); @@ -292,50 +349,7 @@ td::actor::ActorId FullNodeImpl::get_shard(AccountIdPrefixFull ds return get_shard(shard_prefix(dst, 60)); } -void FullNodeImpl::got_key_block_proof(td::Ref proof) { - auto R = proof->get_key_block_config(); - R.ensure(); - auto config = R.move_as_ok(); - - PublicKeyHash l = PublicKeyHash::zero(); - std::vector keys; - std::map current_validators; - for (td::int32 i = -1; i <= 1; i++) { - auto r = config->get_total_validator_set(i < 0 ? i : 1 - i); - if (r.not_null()) { - auto vec = r->export_vector(); - for (auto &el : vec) { - auto key = ValidatorFullId{el.key}.compute_short_id(); - keys.push_back(key); - if (local_keys_.count(key)) { - l = key; - } - if (i == 1) { - current_validators[key] = adnl::AdnlNodeIdShort{el.addr.is_zero() ? key.bits256_value() : el.addr}; - } - } - } - } - - if (current_validators != current_validators_) { - current_validators_ = std::move(current_validators); - update_private_block_overlays(); - } - - if (keys == all_validators_) { - return; - } - - all_validators_ = keys; - sign_cert_by_ = l; - CHECK(all_validators_.size() > 0); - - for (auto &shard : shards_) { - td::actor::send_closure(shard.second, &FullNodeShard::update_validators, all_validators_, sign_cert_by_); - } -} - -void FullNodeImpl::got_zero_block_state(td::Ref state) { +void FullNodeImpl::got_key_block_state(td::Ref state) { auto m = td::Ref{std::move(state)}; PublicKeyHash l = PublicKeyHash::zero(); @@ -358,9 +372,11 @@ void FullNodeImpl::got_zero_block_state(td::Ref state) { } } + set_private_block_overlays_enable_compression(m->get_consensus_config().proto_version >= 3); + if (current_validators != current_validators_) { current_validators_ = std::move(current_validators); - update_private_block_overlays(); + update_private_overlays(); } if (keys == all_validators_) { @@ -377,28 +393,15 @@ void FullNodeImpl::got_zero_block_state(td::Ref state) { } void FullNodeImpl::new_key_block(BlockHandle handle) { - if (handle->id().seqno() == 0) { - auto P = td::PromiseCreator::lambda([SelfId = actor_id(this)](td::Result> R) { - if (R.is_error()) { - VLOG(FULL_NODE_WARNING) << "failed to get zero state: " << R.move_as_error(); - } else { - td::actor::send_closure(SelfId, &FullNodeImpl::got_zero_block_state, R.move_as_ok()); - } - }); - td::actor::send_closure(validator_manager_, &ValidatorManagerInterface::get_shard_state_from_db, handle, - std::move(P)); - } else { - CHECK(handle->is_key_block()); - auto P = td::PromiseCreator::lambda([SelfId = actor_id(this)](td::Result> R) { - if (R.is_error()) { - VLOG(FULL_NODE_WARNING) << "failed to get key block proof: " << R.move_as_error(); - } else { - td::actor::send_closure(SelfId, &FullNodeImpl::got_key_block_proof, R.move_as_ok()); - } - }); - td::actor::send_closure(validator_manager_, &ValidatorManagerInterface::get_block_proof_link_from_db, handle, - std::move(P)); - } + auto P = td::PromiseCreator::lambda([SelfId = actor_id(this)](td::Result> R) { + if (R.is_error()) { + VLOG(FULL_NODE_WARNING) << "failed to get key block state: " << R.move_as_error(); + } else { + td::actor::send_closure(SelfId, &FullNodeImpl::got_key_block_state, R.move_as_ok()); + } + }); + td::actor::send_closure(validator_manager_, &ValidatorManagerInterface::get_shard_state_from_db, handle, + std::move(P)); } void FullNodeImpl::start_up() { @@ -484,7 +487,11 @@ void FullNodeImpl::start_up() { std::make_unique(actor_id(this)), std::move(P)); } -void FullNodeImpl::update_private_block_overlays() { +void FullNodeImpl::update_private_overlays() { + for (auto &p : private_custom_overlays_) { + update_ext_msg_overlay(p.first, p.second); + } + private_block_overlays_.clear(); if (local_keys_.empty()) { return; @@ -494,6 +501,16 @@ void FullNodeImpl::update_private_block_overlays() { } } +void FullNodeImpl::set_private_block_overlays_enable_compression(bool value) { + if (private_block_overlays_enable_compression_ == value) { + return; + } + private_block_overlays_enable_compression_ = true; + for (auto &p : private_block_overlays_) { + td::actor::send_closure(p.second, &FullNodePrivateBlockOverlay::set_enable_compression, value); + } +} + void FullNodeImpl::create_private_block_overlay(PublicKeyHash key) { CHECK(local_keys_.count(key)); if (current_validators_.count(key)) { @@ -501,9 +518,34 @@ void FullNodeImpl::create_private_block_overlay(PublicKeyHash key) { for (const auto &p : current_validators_) { nodes.push_back(p.second); } - private_block_overlays_[key] = td::actor::create_actor( - "BlocksPrivateOverlay", current_validators_[key], std::move(nodes), zero_state_file_hash_, config_, keyring_, - adnl_, rldp_, rldp2_, overlays_, validator_manager_); + private_block_overlays_[key] = td::actor::create_actor( + "BlocksPrivateOverlay", current_validators_[key], std::move(nodes), zero_state_file_hash_, config_, + private_block_overlays_enable_compression_, keyring_, adnl_, rldp_, rldp2_, overlays_, validator_manager_); + } +} + +void FullNodeImpl::update_ext_msg_overlay(const std::string &name, ExtMsgOverlayInfo &overlay) { + auto old_actors = std::move(overlay.actors_); + overlay.actors_.clear(); + auto try_local_id = [&](const adnl::AdnlNodeIdShort &local_id) { + if (std::find(overlay.nodes_.begin(), overlay.nodes_.end(), local_id) != overlay.nodes_.end()) { + auto it = old_actors.find(local_id); + if (it != old_actors.end()) { + overlay.actors_[local_id] = std::move(it->second); + old_actors.erase(it); + } else { + overlay.actors_[local_id] = td::actor::create_actor( + "CustomOverlay", local_id, overlay.nodes_, overlay.senders_, name, zero_state_file_hash_, config_, + keyring_, adnl_, rldp_, rldp2_, overlays_, validator_manager_); + } + } + }; + try_local_id(adnl_id_); + for (const PublicKeyHash &local_key: local_keys_) { + auto it = current_validators_.find(local_key); + if (it != current_validators_.end()) { + try_local_id(it->second); + } } } diff --git a/validator/full-node.h b/validator/full-node.h index 15d54b55..82e0dd5c 100644 --- a/validator/full-node.h +++ b/validator/full-node.h @@ -74,6 +74,11 @@ class FullNode : public td::actor::Actor { virtual void update_adnl_id(adnl::AdnlNodeIdShort adnl_id, td::Promise promise) = 0; virtual void set_config(FullNodeConfig config) = 0; + virtual void add_ext_msg_overlay(std::vector nodes, + std::map senders, std::string name, + td::Promise promise) = 0; + virtual void del_ext_msg_overlay(std::string name, td::Promise promise) = 0; + static constexpr td::uint32 max_block_size() { return 4 << 20; } diff --git a/validator/full-node.hpp b/validator/full-node.hpp index 838700b5..86664a3b 100644 --- a/validator/full-node.hpp +++ b/validator/full-node.hpp @@ -53,6 +53,10 @@ class FullNodeImpl : public FullNode { void update_adnl_id(adnl::AdnlNodeIdShort adnl_id, td::Promise promise) override; void set_config(FullNodeConfig config) override; + void add_ext_msg_overlay(std::vector nodes, std::map senders, + std::string name, td::Promise promise) override; + void del_ext_msg_overlay(std::string name, td::Promise promise) override; + void add_shard(ShardIdFull shard); void del_shard(ShardIdFull shard); @@ -76,8 +80,7 @@ class FullNodeImpl : public FullNode { void download_archive(BlockSeqno masterchain_seqno, std::string tmp_dir, td::Timestamp timeout, td::Promise promise); - void got_key_block_proof(td::Ref proof); - void got_zero_block_state(td::Ref state); + void got_key_block_state(td::Ref state); void new_key_block(BlockHandle handle); void start_up() override; @@ -117,10 +120,21 @@ class FullNodeImpl : public FullNode { std::set local_keys_; FullNodeConfig config_; - std::map> private_block_overlays_; + std::map> private_block_overlays_; + bool private_block_overlays_enable_compression_ = false; - void update_private_block_overlays(); + struct ExtMsgOverlayInfo { + std::vector nodes_; + std::map senders_; + std::map> + actors_; // our local id -> actor + }; + std::map private_custom_overlays_; + + void update_private_overlays(); + void set_private_block_overlays_enable_compression(bool value); void create_private_block_overlay(PublicKeyHash key); + void update_ext_msg_overlay(const std::string& name, ExtMsgOverlayInfo& overlay); }; } // namespace fullnode diff --git a/validator/impl/CMakeLists.txt b/validator/impl/CMakeLists.txt index 7cd273f2..42dded31 100644 --- a/validator/impl/CMakeLists.txt +++ b/validator/impl/CMakeLists.txt @@ -7,6 +7,7 @@ endif() set(TON_VALIDATOR_SOURCE accept-block.cpp block.cpp + candidates-buffer.cpp check-proof.cpp collator.cpp config.cpp @@ -24,6 +25,7 @@ set(TON_VALIDATOR_SOURCE accept-block.hpp block.hpp + candidates-buffer.hpp check-proof.hpp collate-query-impl.h collator-impl.h diff --git a/validator/impl/candidates-buffer.cpp b/validator/impl/candidates-buffer.cpp new file mode 100644 index 00000000..e24913b6 --- /dev/null +++ b/validator/impl/candidates-buffer.cpp @@ -0,0 +1,213 @@ +/* + This file is part of TON Blockchain Library. + + TON Blockchain Library is free software: you can redistribute it and/or modify + it under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation, either version 2 of the License, or + (at your option) any later version. + + TON Blockchain Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with TON Blockchain Library. If not, see . +*/ +#include "candidates-buffer.hpp" +#include "fabric.h" + +namespace ton::validator { + +void CandidatesBuffer::start_up() { + alarm_timestamp() = td::Timestamp::in(60.0); +} + +void CandidatesBuffer::alarm() { + alarm_timestamp() = td::Timestamp::in(60.0); + for (auto it = candidates_.begin(); it != candidates_.end();) { + Candidate &entry = it->second; + if (entry.ttl_.is_in_past()) { + for (auto &p : entry.data_waiters_) { + p.set_error(td::Status::Error(ErrorCode::timeout, "timeout")); + } + for (auto &p : entry.state_waiters_) { + p.set_error(td::Status::Error(ErrorCode::timeout, "timeout")); + } + it = candidates_.erase(it); + } else { + ++it; + } + } +} + +void CandidatesBuffer::add_new_candidate(BlockIdExt id, PublicKey source, FileHash collated_data_file_hash) { + auto it = candidates_.emplace(id, Candidate{}); + Candidate &entry = it.first->second; + entry.ttl_ = td::Timestamp::in(120.0); + if (!it.second) { // not inserted + return; + } + LOG(DEBUG) << "New block candidate " << id.to_str(); + entry.source_ = source; + entry.collated_data_file_hash_ = collated_data_file_hash; +} + +void CandidatesBuffer::get_block_data(BlockIdExt id, td::Promise> promise) { + auto it = candidates_.find(id); + if (it == candidates_.end()) { + promise.set_error(td::Status::Error(ErrorCode::notready, "unknown block candidate")); + return; + } + Candidate &entry = it->second; + if (entry.data_.not_null()) { + promise.set_result(entry.data_); + return; + } + entry.data_waiters_.push_back(std::move(promise)); + if (entry.data_requested_) { + return; + } + entry.data_requested_ = true; + td::actor::send_closure(manager_, &ValidatorManager::get_block_candidate_from_db, entry.source_, id, + entry.collated_data_file_hash_, [SelfId = actor_id(this), id](td::Result R) { + td::actor::send_closure(SelfId, &CandidatesBuffer::got_block_candidate, id, std::move(R)); + }); +} + +void CandidatesBuffer::got_block_candidate(BlockIdExt id, td::Result R) { + if (R.is_error()) { + finish_get_block_data(id, R.move_as_error()); + return; + } + BlockCandidate cand = R.move_as_ok(); + CHECK(cand.id == id); + finish_get_block_data(id, create_block(id, std::move(cand.data))); +} + +void CandidatesBuffer::get_block_state(BlockIdExt id, td::Promise> promise) { + auto it = candidates_.find(id); + if (it == candidates_.end()) { + promise.set_error(td::Status::Error(ErrorCode::notready, "unknown block candidate")); + return; + } + Candidate &entry = it->second; + if (entry.state_.not_null()) { + promise.set_result(entry.state_); + return; + } + entry.state_waiters_.push_back(std::move(promise)); + if (entry.state_requested_) { + return; + } + entry.state_requested_ = true; + get_block_data(id, [SelfId = actor_id(this), id](td::Result> R) { + if (R.is_error()) { + td::actor::send_closure(SelfId, &CandidatesBuffer::finish_get_block_state, id, R.move_as_error()); + return; + } + td::actor::send_closure(SelfId, &CandidatesBuffer::get_block_state_cont, id, R.move_as_ok()); + }); +} + +void CandidatesBuffer::get_block_state_cont(BlockIdExt id, td::Ref data) { + CHECK(id == data->block_id()); + std::vector prev; + BlockIdExt mc_blkid; + bool after_split; + auto S = block::unpack_block_prev_blk_ext(data->root_cell(), id, prev, mc_blkid, after_split); + if (S.is_error()) { + finish_get_block_state(id, std::move(S)); + return; + } + get_block_state_cont2(std::move(data), std::move(prev), {}); +} + +void CandidatesBuffer::get_block_state_cont2(td::Ref block, std::vector prev, + std::vector> prev_states) { + if (prev_states.size() < prev.size()) { + BlockIdExt prev_id = prev[prev_states.size()]; + td::actor::send_closure(manager_, &ValidatorManager::get_shard_state_from_db_short, prev_id, + [SelfId = actor_id(this), block = std::move(block), prev = std::move(prev), + prev_states = std::move(prev_states)](td::Result> R) mutable { + if (R.is_error()) { + td::actor::send_closure(SelfId, &CandidatesBuffer::finish_get_block_state, + block->block_id(), R.move_as_error()); + return; + } + prev_states.push_back(R.move_as_ok()); + td::actor::send_closure(SelfId, &CandidatesBuffer::get_block_state_cont2, + std::move(block), std::move(prev), std::move(prev_states)); + }); + return; + } + + BlockIdExt id = block->block_id(); + td::Ref state; + CHECK(prev_states.size() == 1 || prev_states.size() == 2); + if (prev_states.size() == 2) { // after merge + auto R = prev_states[0]->merge_with(*prev_states[1]); + if (R.is_error()) { + finish_get_block_state(id, R.move_as_error()); + return; + } + state = R.move_as_ok(); + } else if (id.shard_full() != prev[0].shard_full()) { // after split + auto R = prev_states[0]->split(); + if (R.is_error()) { + finish_get_block_state(id, R.move_as_error()); + return; + } + auto s = R.move_as_ok(); + state = is_left_child(id.shard_full()) ? std::move(s.first) : std::move(s.second); + } else { // no split/merge + state = std::move(prev_states[0]); + } + + auto S = state.write().apply_block(id, std::move(block)); + if (S.is_error()) { + finish_get_block_state(id, std::move(S)); + return; + } + finish_get_block_state(id, std::move(state)); +} + +void CandidatesBuffer::finish_get_block_data(BlockIdExt id, td::Result> res) { + auto it = candidates_.find(id); + if (it == candidates_.end()) { + return; + } + Candidate &entry = it->second; + for (auto &p : entry.data_waiters_) { + p.set_result(res.clone()); + } + entry.data_waiters_.clear(); + entry.data_requested_ = false; + if (res.is_ok()) { + entry.data_ = res.move_as_ok(); + LOG(DEBUG) << "Loaded block data for " << id.to_str(); + } else { + LOG(DEBUG) << "Failed to load block data for " << id.to_str() << ": " << res.move_as_error(); + } +} + +void CandidatesBuffer::finish_get_block_state(BlockIdExt id, td::Result> res) { + auto it = candidates_.find(id); + if (it == candidates_.end()) { + return; + } + Candidate &entry = it->second; + for (auto &p : entry.state_waiters_) { + p.set_result(res.clone()); + } + entry.state_waiters_.clear(); + entry.state_requested_ = false; + if (res.is_ok()) { + entry.state_ = res.move_as_ok(); + LOG(DEBUG) << "Loaded block state for " << id.to_str(); + } else { + LOG(DEBUG) << "Failed to load block state for " << id.to_str() << ": " << res.move_as_error(); + } +} + +} // namespace ton::validator diff --git a/validator/impl/candidates-buffer.hpp b/validator/impl/candidates-buffer.hpp new file mode 100644 index 00000000..5db1e516 --- /dev/null +++ b/validator/impl/candidates-buffer.hpp @@ -0,0 +1,64 @@ +/* + This file is part of TON Blockchain Library. + + TON Blockchain Library is free software: you can redistribute it and/or modify + it under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation, either version 2 of the License, or + (at your option) any later version. + + TON Blockchain Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with TON Blockchain Library. If not, see . +*/ +#pragma once +#include "ton/ton-types.h" +#include "td/actor/actor.h" +#include "interfaces/validator-manager.h" + +namespace ton::validator { + +class CandidatesBuffer : public td::actor::Actor { + public: + explicit CandidatesBuffer(td::actor::ActorId manager) : manager_(std::move(manager)) { + } + + void start_up() override; + void alarm() override; + + void add_new_candidate(BlockIdExt id, PublicKey source, FileHash collated_data_file_hash); + void get_block_data(BlockIdExt id, td::Promise> promise); + void get_block_state(BlockIdExt id, td::Promise> promise); + + private: + td::actor::ActorId manager_; + + struct Candidate { + PublicKey source_; + FileHash collated_data_file_hash_; + td::Timestamp ttl_; + + td::Ref data_; + std::vector>> data_waiters_; + bool data_requested_{false}; + + td::Ref state_; + std::vector>> state_waiters_; + bool state_requested_{false}; + }; + std::map candidates_; + + void got_block_candidate(BlockIdExt id, td::Result R); + + void get_block_state_cont(BlockIdExt id, td::Ref data); + void get_block_state_cont2(td::Ref block, std::vector prev, + std::vector> prev_states); + + void finish_get_block_data(BlockIdExt id, td::Result> res); + void finish_get_block_state(BlockIdExt id, td::Result> res); +}; + +} // namespace ton::validator diff --git a/validator/impl/collator-impl.h b/validator/impl/collator-impl.h index c6ba27b4..055c1aed 100644 --- a/validator/impl/collator-impl.h +++ b/validator/impl/collator-impl.h @@ -183,7 +183,12 @@ class Collator final : public td::actor::Actor { block::ValueFlow value_flow_{block::ValueFlow::SetZero()}; std::unique_ptr fees_import_dict_; std::map ext_msg_map; - std::vector, ExtMessage::Hash>> ext_msg_list_; + struct ExtMsg { + Ref cell; + ExtMessage::Hash hash; + int priority; + }; + std::vector ext_msg_list_; std::priority_queue, std::greater> new_msgs; std::pair last_proc_int_msg_, first_unproc_int_msg_; std::unique_ptr in_msg_dict, out_msg_dict, out_msg_queue_, sibling_out_msg_queue_; @@ -268,8 +273,9 @@ class Collator final : public td::actor::Actor { bool is_our_address(Ref addr_ref) const; bool is_our_address(ton::AccountIdPrefixFull addr_prefix) const; bool is_our_address(const ton::StdSmcAddress& addr) const; - void after_get_external_messages(td::Result>> res); - td::Result register_external_message_cell(Ref ext_msg, const ExtMessage::Hash& ext_hash); + void after_get_external_messages(td::Result, int>>> res); + td::Result register_external_message_cell(Ref ext_msg, const ExtMessage::Hash& ext_hash, + int priority); // td::Result register_external_message(td::Slice ext_msg_boc); void register_new_msg(block::NewOutMsg msg); void register_new_msgs(block::transaction::Transaction& trans); @@ -322,6 +328,9 @@ class Collator final : public td::actor::Actor { bool create_block_candidate(); void return_block_candidate(td::Result saved); bool update_last_proc_int_msg(const std::pair& new_lt_hash); + + public: + static td::uint32 get_skip_externals_queue_size(); }; } // namespace validator diff --git a/validator/impl/collator.cpp b/validator/impl/collator.cpp index 413c5be9..fd4ddd34 100644 --- a/validator/impl/collator.cpp +++ b/validator/impl/collator.cpp @@ -49,6 +49,7 @@ static const td::uint32 FORCE_SPLIT_QUEUE_SIZE = 4096; static const td::uint32 SPLIT_MAX_QUEUE_SIZE = 100000; static const td::uint32 MERGE_MAX_QUEUE_SIZE = 2047; static const td::uint32 SKIP_EXTERNALS_QUEUE_SIZE = 8000; +static const int HIGH_PRIORITY_EXTERNAL = 10; // don't skip high priority externals when queue is big #define DBG(__n) dbg(__n)&& #define DSTART int __dcnt = 0; @@ -256,11 +257,10 @@ void Collator::start_up() { LOG(DEBUG) << "sending get_external_messages() query to Manager"; ++pending; td::actor::send_closure_later(manager, &ValidatorManager::get_external_messages, shard_, - [self = get_self()](td::Result>> res) -> void { - LOG(DEBUG) << "got answer to get_external_messages() query"; - td::actor::send_closure_later( - std::move(self), &Collator::after_get_external_messages, std::move(res)); - }); + [self = get_self()](td::Result, int>>> res) -> void { + LOG(DEBUG) << "got answer to get_external_messages() query"; + td::actor::send_closure_later(std::move(self), &Collator::after_get_external_messages, std::move(res)); + }); } if (is_masterchain() && !is_hardfork_) { // 5. load shard block info messages @@ -3413,12 +3413,15 @@ bool Collator::process_inbound_external_messages() { return true; } if (out_msg_queue_size_ > SKIP_EXTERNALS_QUEUE_SIZE) { - LOG(INFO) << "skipping processing of inbound external messages because out_msg_queue is too big (" + LOG(INFO) << "skipping processing of inbound external messages (except for high-priority) because out_msg_queue is " + "too big (" << out_msg_queue_size_ << " > " << SKIP_EXTERNALS_QUEUE_SIZE << ")"; - return true; } bool full = !block_limit_status_->fits(block::ParamLimits::cl_soft); - for (auto& ext_msg_pair : ext_msg_list_) { + for (auto& ext_msg_struct : ext_msg_list_) { + if (out_msg_queue_size_ > SKIP_EXTERNALS_QUEUE_SIZE && ext_msg_struct.priority < HIGH_PRIORITY_EXTERNAL) { + continue; + } if (full) { LOG(INFO) << "BLOCK FULL, stop processing external messages"; break; @@ -3427,15 +3430,15 @@ bool Collator::process_inbound_external_messages() { LOG(WARNING) << "medium timeout reached, stop processing inbound external messages"; break; } - auto ext_msg = ext_msg_pair.first; + auto ext_msg = ext_msg_struct.cell; ton::Bits256 hash{ext_msg->get_hash().bits()}; int r = process_external_message(std::move(ext_msg)); if (r < 0) { - bad_ext_msgs_.emplace_back(ext_msg_pair.second); + bad_ext_msgs_.emplace_back(ext_msg_struct.hash); return false; } if (!r) { - delay_ext_msgs_.emplace_back(ext_msg_pair.second); + delay_ext_msgs_.emplace_back(ext_msg_struct.hash); } if (r > 0) { full = !block_limit_status_->fits(block::ParamLimits::cl_soft); @@ -5062,7 +5065,8 @@ void Collator::return_block_candidate(td::Result saved) { * - If the external message has been previuosly registered and accepted, returns false. * - Otherwise returns true. */ -td::Result Collator::register_external_message_cell(Ref ext_msg, const ExtMessage::Hash& ext_hash) { +td::Result Collator::register_external_message_cell(Ref ext_msg, const ExtMessage::Hash& ext_hash, + int priority) { if (ext_msg->get_level() != 0) { return td::Status::Error("external message must have zero level"); } @@ -5106,7 +5110,7 @@ td::Result Collator::register_external_message_cell(Ref ext_msg, block::gen::t_Message_Any.print_ref(std::cerr, ext_msg); } ext_msg_map.emplace(hash, 1); - ext_msg_list_.emplace_back(std::move(ext_msg), ext_hash); + ext_msg_list_.push_back({std::move(ext_msg), ext_hash, priority}); return true; } @@ -5115,18 +5119,21 @@ td::Result Collator::register_external_message_cell(Ref ext_msg, * * @param res The result of the external message retrieval operation. */ -void Collator::after_get_external_messages(td::Result>> res) { +void Collator::after_get_external_messages(td::Result, int>>> res) { + // res: pair {ext msg, priority} --pending; if (res.is_error()) { fatal_error(res.move_as_error()); return; } auto vect = res.move_as_ok(); - for (auto&& ext_msg : vect) { + for (auto& p : vect) { + auto& ext_msg = p.first; + int priority = p.second; Ref ext_msg_cell = ext_msg->root_cell(); bool err = ext_msg_cell.is_null(); if (!err) { - auto reg_res = register_external_message_cell(std::move(ext_msg_cell), ext_msg->hash()); + auto reg_res = register_external_message_cell(std::move(ext_msg_cell), ext_msg->hash(), priority); if (reg_res.is_error() || !reg_res.move_as_ok()) { err = true; } @@ -5140,6 +5147,10 @@ void Collator::after_get_external_messages(td::Result send_message_cache_; size_t send_message_error_cnt_ = 0; - static const size_t MAX_CACHE_SIZE = 64 << 20; + static constexpr size_t MAX_CACHE_SIZE = 64 << 20; }; } // namespace ton::validator diff --git a/validator/impl/liteserver.cpp b/validator/impl/liteserver.cpp index 9b8cb300..7fa6e59e 100644 --- a/validator/impl/liteserver.cpp +++ b/validator/impl/liteserver.cpp @@ -42,6 +42,8 @@ #include "signature-set.hpp" #include "fabric.h" #include +#include "td/actor/MultiPromise.h" +#include "collator-impl.h" namespace ton { @@ -233,6 +235,9 @@ void LiteQuery::perform() { [&](lite_api::liteServer_lookupBlock& q) { this->perform_lookupBlock(ton::create_block_id_simple(q.id_), q.mode_, q.lt_, q.utime_); }, + [&](lite_api::liteServer_lookupBlockWithProof& q) { + this->perform_lookupBlockWithProof(ton::create_block_id_simple(q.id_), ton::create_block_id(q.mc_block_id_), q.mode_, q.lt_, q.utime_); + }, [&](lite_api::liteServer_listBlockTransactions& q) { this->perform_listBlockTransactions(ton::create_block_id(q.id_), q.mode_, q.count_, (q.mode_ & 128) ? q.after_->account_ : td::Bits256::zero(), @@ -266,9 +271,22 @@ void LiteQuery::perform() { [&](lite_api::liteServer_getLibraries& q) { this->perform_getLibraries(q.library_list_); }, + [&](lite_api::liteServer_getLibrariesWithProof& q) { + this->perform_getLibrariesWithProof(ton::create_block_id(q.id_), q.mode_, q.library_list_); + }, [&](lite_api::liteServer_getShardBlockProof& q) { this->perform_getShardBlockProof(create_block_id(q.id_)); }, + [&](lite_api::liteServer_nonfinal_getCandidate& q) { + this->perform_nonfinal_getCandidate(q.id_->creator_, create_block_id(q.id_->block_id_), + q.id_->collated_data_hash_); + }, + [&](lite_api::liteServer_nonfinal_getValidatorGroups& q) { + this->perform_nonfinal_getValidatorGroups(q.mode_, ShardIdFull{q.wc_, (ShardId)q.shard_}); + }, + [&](lite_api::liteServer_getOutMsgQueueSizes& q) { + this->perform_getOutMsgQueueSizes(q.mode_ & 1 ? ShardIdFull(q.wc_, q.shard_) : td::optional()); + }, [&](auto& obj) { this->abort_query(td::Status::Error(ErrorCode::protoviolation, "unknown query")); })); } @@ -337,21 +355,15 @@ void LiteQuery::perform_getBlock(BlockIdExt blkid) { fatal_error("invalid BlockIdExt"); return; } - get_block_handle_checked(blkid, [manager = manager_, Self = actor_id(this), blkid](td::Result R) { - if (R.is_error()) { - td::actor::send_closure(Self, &LiteQuery::abort_query, R.move_as_error()); - return; - } - td::actor::send_closure_later(manager, &ValidatorManager::get_block_data_from_db, R.move_as_ok(), - [=](td::Result> res) { - if (res.is_error()) { - td::actor::send_closure(Self, &LiteQuery::abort_query, res.move_as_error()); - } else { - td::actor::send_closure_later(Self, &LiteQuery::continue_getBlock, blkid, - res.move_as_ok()); - } - }); - }); + td::actor::send_closure(manager_, &ValidatorManager::get_block_data_for_litequery, blkid, + [Self = actor_id(this), blkid](td::Result> res) { + if (res.is_error()) { + td::actor::send_closure(Self, &LiteQuery::abort_query, res.move_as_error()); + } else { + td::actor::send_closure_later(Self, &LiteQuery::continue_getBlock, blkid, + res.move_as_ok()); + } + }); } void LiteQuery::continue_getBlock(BlockIdExt blkid, Ref block) { @@ -369,21 +381,15 @@ void LiteQuery::perform_getBlockHeader(BlockIdExt blkid, int mode) { fatal_error("invalid BlockIdExt"); return; } - get_block_handle_checked(blkid, [=, manager = manager_, Self = actor_id(this)](td::Result R) { - if (R.is_error()) { - td::actor::send_closure(Self, &LiteQuery::abort_query, R.move_as_error()); - return; - } - td::actor::send_closure_later(manager, &ValidatorManager::get_block_data_from_db, R.move_as_ok(), - [=](td::Result> res) { - if (res.is_error()) { - td::actor::send_closure(Self, &LiteQuery::abort_query, res.move_as_error()); - } else { - td::actor::send_closure_later(Self, &LiteQuery::continue_getBlockHeader, blkid, - mode, res.move_as_ok()); - } - }); - }); + td::actor::send_closure(manager_, &ValidatorManager::get_block_data_for_litequery, blkid, + [Self = actor_id(this), blkid, mode](td::Result> res) { + if (res.is_error()) { + td::actor::send_closure(Self, &LiteQuery::abort_query, res.move_as_error()); + } else { + td::actor::send_closure_later(Self, &LiteQuery::continue_getBlockHeader, blkid, mode, + res.move_as_ok()); + } + }); } static bool visit(Ref cell); @@ -489,33 +495,27 @@ void LiteQuery::perform_getState(BlockIdExt blkid) { fatal_error("cannot request total state: possibly too large"); return; } - get_block_handle_checked(blkid, [=, manager = manager_, Self = actor_id(this)](td::Result R) { - if (R.is_error()) { - td::actor::send_closure(Self, &LiteQuery::abort_query, R.move_as_error()); - return; - } - if (blkid.id.seqno) { - td::actor::send_closure_later(manager, &ValidatorManager::get_shard_state_from_db, R.move_as_ok(), - [=](td::Result> res) { - if (res.is_error()) { - td::actor::send_closure(Self, &LiteQuery::abort_query, res.move_as_error()); - } else { - td::actor::send_closure_later(Self, &LiteQuery::continue_getState, blkid, - res.move_as_ok()); - } - }); - } else { - td::actor::send_closure_later(manager, &ValidatorManager::get_zero_state, blkid, - [=](td::Result res) { - if (res.is_error()) { - td::actor::send_closure(Self, &LiteQuery::abort_query, res.move_as_error()); - } else { - td::actor::send_closure_later(Self, &LiteQuery::continue_getZeroState, blkid, - res.move_as_ok()); - } - }); - } - }); + if (blkid.id.seqno) { + td::actor::send_closure(manager_, &ValidatorManager::get_block_state_for_litequery, blkid, + [Self = actor_id(this), blkid](td::Result> res) { + if (res.is_error()) { + td::actor::send_closure(Self, &LiteQuery::abort_query, res.move_as_error()); + } else { + td::actor::send_closure_later(Self, &LiteQuery::continue_getState, blkid, + res.move_as_ok()); + } + }); + } else { + td::actor::send_closure_later(manager_, &ValidatorManager::get_zero_state, blkid, + [Self = actor_id(this), blkid](td::Result res) { + if (res.is_error()) { + td::actor::send_closure(Self, &LiteQuery::abort_query, res.move_as_error()); + } else { + td::actor::send_closure_later(Self, &LiteQuery::continue_getZeroState, blkid, + res.move_as_ok()); + } + }); + } } void LiteQuery::continue_getState(BlockIdExt blkid, Ref state) { @@ -577,7 +577,7 @@ bool LiteQuery::request_mc_block_data(BlockIdExt blkid) { base_blk_id_ = blkid; ++pending_; td::actor::send_closure_later( - manager_, &ValidatorManager::get_block_data_from_db_short, blkid, + manager_, &ValidatorManager::get_block_data_for_litequery, blkid, [Self = actor_id(this), blkid](td::Result> res) { if (res.is_error()) { td::actor::send_closure(Self, &LiteQuery::abort_query, @@ -635,7 +635,7 @@ bool LiteQuery::request_mc_block_state(BlockIdExt blkid) { base_blk_id_ = blkid; ++pending_; td::actor::send_closure_later( - manager_, &ValidatorManager::get_shard_state_from_db_short, blkid, + manager_, &ValidatorManager::get_block_state_for_litequery, blkid, [Self = actor_id(this), blkid](td::Result> res) { if (res.is_error()) { td::actor::send_closure(Self, &LiteQuery::abort_query, @@ -665,22 +665,16 @@ bool LiteQuery::request_block_state(BlockIdExt blkid) { } blk_id_ = blkid; ++pending_; - get_block_handle_checked(blkid, [=, manager = manager_, Self = actor_id(this)](td::Result R) { - if (R.is_error()) { - td::actor::send_closure(Self, &LiteQuery::abort_query, R.move_as_error()); - return; - } - td::actor::send_closure_later( - manager, &ValidatorManager::get_shard_state_from_db, R.move_as_ok(), - [=](td::Result> res) { - if (res.is_error()) { - td::actor::send_closure(Self, &LiteQuery::abort_query, - res.move_as_error_prefix("cannot load state for "s + blkid.to_str() + " : ")); - } else { - td::actor::send_closure_later(Self, &LiteQuery::got_block_state, blkid, res.move_as_ok()); - } - }); - }); + td::actor::send_closure(manager_, &ValidatorManager::get_block_state_for_litequery, blkid, + [Self = actor_id(this), blkid](td::Result> res) { + if (res.is_error()) { + td::actor::send_closure( + Self, &LiteQuery::abort_query, + res.move_as_error_prefix("cannot load state for "s + blkid.to_str() + " : ")); + } else { + td::actor::send_closure_later(Self, &LiteQuery::got_block_state, blkid, res.move_as_ok()); + } + }); return true; } @@ -693,22 +687,16 @@ bool LiteQuery::request_block_data(BlockIdExt blkid) { } blk_id_ = blkid; ++pending_; - get_block_handle_checked(blkid, [=, manager = manager_, Self = actor_id(this)](td::Result R) { - if (R.is_error()) { - td::actor::send_closure(Self, &LiteQuery::abort_query, R.move_as_error()); - return; - } - td::actor::send_closure_later( - manager, &ValidatorManager::get_block_data_from_db, R.move_as_ok(), - [=](td::Result> res) { - if (res.is_error()) { - td::actor::send_closure(Self, &LiteQuery::abort_query, - res.move_as_error_prefix("cannot load block "s + blkid.to_str() + " : ")); - } else { - td::actor::send_closure_later(Self, &LiteQuery::got_block_data, blkid, res.move_as_ok()); - } - }); - }); + td::actor::send_closure(manager_, &ValidatorManager::get_block_data_for_litequery, blkid, + [Self = actor_id(this), blkid](td::Result> res) { + if (res.is_error()) { + td::actor::send_closure( + Self, &LiteQuery::abort_query, + res.move_as_error_prefix("cannot load block "s + blkid.to_str() + " : ")); + } else { + td::actor::send_closure_later(Self, &LiteQuery::got_block_data, blkid, res.move_as_ok()); + } + }); return true; } @@ -964,6 +952,100 @@ void LiteQuery::continue_getLibraries(Ref mc_s finish_query(std::move(b)); } +void LiteQuery::perform_getLibrariesWithProof(BlockIdExt blkid, int mode, std::vector library_list) { + LOG(INFO) << "started a getLibrariesWithProof() liteserver query"; + if (library_list.size() > 16) { + LOG(INFO) << "too many libraries requested, returning only first 16"; + library_list.resize(16); + } + sort( library_list.begin(), library_list.end() ); + library_list.erase( unique( library_list.begin(), library_list.end() ), library_list.end() ); + + set_continuation([this, library_list, mode]() -> void { continue_getLibrariesWithProof(library_list, mode); }); + request_mc_block_data_state(blkid); +} + +void LiteQuery::continue_getLibrariesWithProof(std::vector library_list, int mode) { + LOG(INFO) << "obtained masterchain block = " << base_blk_id_.to_str(); + CHECK(mc_state_.not_null()); + + Ref state_proof, data_proof; + if (!make_mc_state_root_proof(state_proof)) { + return; + } + + vm::MerkleProofBuilder pb{mc_state_->root_cell()}; + block::gen::ShardStateUnsplit::Record state; + if (!tlb::unpack_cell(pb.root(), state)) { + fatal_error("cannot unpack header of shardchain state "s + base_blk_id_.to_str()); + } + auto libraries_dict = vm::Dictionary(state.r1.libraries->prefetch_ref(), 256); + + std::vector> result; + std::vector result_hashes; + for (const auto& hash : library_list) { + LOG(INFO) << "looking for library " << hash.to_hex(); + + auto csr = libraries_dict.lookup(hash.bits(), 256); + if (csr.is_null() || csr->prefetch_ulong(2) != 0 || !csr->have_refs()) { // shared_lib_descr$00 lib:^Cell + continue; + } + block::gen::LibDescr::Record libdescr; + if (!tlb::csr_unpack(csr, libdescr)) { + fatal_error("cannot unpack LibDescr record "s + hash.to_hex()); + return; + } + if (mode & 1) { + // include first 16 publishers in the proof + auto publishers_dict = vm::Dictionary{vm::DictNonEmpty(), libdescr.publishers, 256}; + auto iter = publishers_dict.begin(); + constexpr int max_publishers = 15; // set to 15 because publishers_dict.begin() counts as the first visit + for (int i = 0; i < max_publishers && iter != publishers_dict.end(); ++i, ++iter) {} + } + + result_hashes.push_back(hash); + } + + auto data_proof_boc = pb.extract_proof_boc(); + if (data_proof_boc.is_error()) { + fatal_error(data_proof_boc.move_as_error()); + return; + } + auto state_proof_boc = vm::std_boc_serialize(std::move(state_proof)); + if (state_proof_boc.is_error()) { + fatal_error(state_proof_boc.move_as_error()); + return; + } + + for (const auto& hash : result_hashes) { + auto csr = libraries_dict.lookup(hash.bits(), 256); + block::gen::LibDescr::Record libdescr; + if (!tlb::csr_unpack(csr, libdescr)) { + fatal_error("cannot unpack LibDescr record "s + hash.to_hex()); + return; + } + if (!libdescr.lib->get_hash().bits().equals(hash.bits(), 256)) { + LOG(ERROR) << "public library hash mismatch: expected " << hash.to_hex() << " , found " + << libdescr.lib->get_hash().to_hex(); + continue; + } + td::BufferSlice libdata; + if (!(mode & 2)) { + auto data = vm::std_boc_serialize(libdescr.lib); + if (data.is_error()) { + LOG(WARNING) << "library serialization failed: " << data.move_as_error().to_string(); + continue; + } + libdata = data.move_as_ok(); + } + result.push_back(ton::create_tl_object(hash, std::move(libdata))); + } + + auto b = ton::create_serialize_tl_object(ton::create_tl_lite_block_id(base_blk_id_), mode, std::move(result), + state_proof_boc.move_as_ok(), data_proof_boc.move_as_ok()); + finish_query(std::move(b)); +} + void LiteQuery::perform_getOneTransaction(BlockIdExt blkid, WorkchainId workchain, StdSmcAddress addr, LogicalTime lt) { LOG(INFO) << "started a getOneTransaction(" << blkid.to_str() << ", " << workchain << ", " << addr.to_hex() << "," << lt << ") liteserver query"; @@ -1333,11 +1415,11 @@ static td::Ref prepare_vm_c7(ton::UnixTime now, ton::LogicalTime lt, td::make_refint(lt), // trans_lt:Integer std::move(rand_seed_int), // rand_seed:Integer balance.as_vm_tuple(), // balance_remaining:[Integer (Maybe Cell)] - my_addr, // myself:MsgAddressInt - config ? config->get_root_cell() : vm::StackEntry() // global_config:(Maybe Cell) ] = SmartContractInfo; + my_addr, // myself:MsgAddressInt + config ? config->get_root_cell() : vm::StackEntry() // global_config:(Maybe Cell) ] = SmartContractInfo; }; if (config && config->get_global_version() >= 4) { - tuple.push_back(my_code); // code:Cell + tuple.push_back(vm::StackEntry::maybe(my_code)); // code:Cell tuple.push_back(block::CurrencyCollection::zero().as_vm_tuple()); // in_msg_value:[Integer (Maybe Cell)] tuple.push_back(td::zero_refint()); // storage_fees:Integer @@ -1348,8 +1430,14 @@ static td::Ref prepare_vm_c7(ton::UnixTime now, ton::LogicalTime lt, tuple.push_back(info.is_ok() ? info.move_as_ok() : vm::StackEntry()); } if (config && config->get_global_version() >= 6) { - tuple.push_back(config->get_unpacked_config_tuple(now)); // unpacked_config_tuple:[...] - tuple.push_back(due_payment); // due_payment:Integer + tuple.push_back(vm::StackEntry::maybe(config->get_unpacked_config_tuple(now))); // unpacked_config_tuple:[...] + tuple.push_back(due_payment); // due_payment:Integer + // precomiled_gas_usage:(Maybe Integer) + td::optional precompiled; + if (my_code.not_null()) { + precompiled = config->get_precompiled_contracts_config().get_contract(my_code->get_hash().bits()); + } + tuple.push_back(precompiled ? td::make_refint(precompiled.value().gas_usage) : vm::StackEntry()); } auto tuple_ref = td::make_cnt_ref>(std::move(tuple)); LOG(DEBUG) << "SmartContractInfo initialized with " << vm::StackEntry(tuple_ref).to_string(); @@ -1430,6 +1518,7 @@ void LiteQuery::finish_runSmcMethod(td::BufferSlice shard_proof, td::BufferSlice auto c7 = prepare_vm_c7(gen_utime, gen_lt, td::make_ref(acc.addr->clone()), balance, config.get(), std::move(code), due_payment); vm.set_c7(c7); // tuple with SmartContractInfo + vm.set_global_version(config->get_global_version()); // vm.incr_stack_trace(1); // enable stack dump after each step LOG(INFO) << "starting VM to run GET-method of smart contract " << acc_workchain_ << ":" << acc_addr_.to_hex(); // **** RUN VM **** @@ -1608,7 +1697,7 @@ void LiteQuery::continue_getTransactions(unsigned remaining, bool exact) { LOG(DEBUG) << "sending get_block_by_lt_from_db() query to manager for " << acc_workchain_ << ":" << acc_addr_.to_hex() << " " << trans_lt_; td::actor::send_closure_later( - manager_, &ValidatorManager::get_block_by_lt_from_db_for_litequery, ton::extract_addr_prefix(acc_workchain_, acc_addr_), + manager_, &ValidatorManager::get_block_by_lt_for_litequery, ton::extract_addr_prefix(acc_workchain_, acc_addr_), trans_lt_, [Self = actor_id(this), remaining, manager = manager_](td::Result res) { if (res.is_error()) { td::actor::send_closure(Self, &LiteQuery::abort_getTransactions, res.move_as_error(), ton::BlockIdExt{}); @@ -1849,7 +1938,7 @@ void LiteQuery::continue_getConfigParams(int mode, std::vector param_list) void LiteQuery::perform_getAllShardsInfo(BlockIdExt blkid) { LOG(INFO) << "started a getAllShardsInfo(" << blkid.to_str() << ") liteserver query"; set_continuation([&]() -> void { continue_getAllShardsInfo(); }); - request_mc_block_data_state(blkid); + request_mc_block_data(blkid); } void LiteQuery::continue_getShardInfo(ShardIdFull shard, bool exact) { @@ -1896,30 +1985,30 @@ void LiteQuery::continue_getShardInfo(ShardIdFull shard, bool exact) { void LiteQuery::continue_getAllShardsInfo() { LOG(INFO) << "completing getAllShardsInfo() query"; - Ref proof1, proof2; - if (!make_mc_state_root_proof(proof1)) { + vm::MerkleProofBuilder mpb{mc_block_->root_cell()}; + block::gen::Block::Record blk; + block::gen::BlockExtra::Record extra; + block::gen::McBlockExtra::Record mc_extra; + if (!tlb::unpack_cell(mpb.root(), blk) || !tlb::unpack_cell(blk.extra, extra) || !extra.custom->have_refs() || + !tlb::unpack_cell(extra.custom->prefetch_ref(), mc_extra)) { + fatal_error("cannot unpack header of block "s + mc_block_->block_id().to_str()); return; } - vm::MerkleProofBuilder mpb{mc_state_->root_cell()}; - auto shards_dict = block::ShardConfig::extract_shard_hashes_dict(mpb.root()); - if (!shards_dict) { - fatal_error("cannot extract ShardHashes from last mc state"); - return; - } - if (!mpb.extract_proof_to(proof2)) { + vm::Dictionary shards_dict(std::move(mc_extra.shard_hashes), 32); + Ref proof; + if (!mpb.extract_proof_to(proof)) { fatal_error("cannot construct Merkle proof for all shards dictionary"); return; } - shards_dict = block::ShardConfig::extract_shard_hashes_dict(mc_state_->root_cell()); - vm::CellBuilder cb; - Ref cell; - if (!(std::move(shards_dict)->append_dict_to_bool(cb) && cb.finalize_to(cell))) { - fatal_error("cannot store ShardHashes from last mc state into a new cell"); + auto proof_boc = vm::std_boc_serialize(std::move(proof)); + if (proof_boc.is_error()) { + fatal_error(proof_boc.move_as_error()); return; } - auto proof = vm::std_boc_serialize_multi({std::move(proof1), std::move(proof2)}); - if (proof.is_error()) { - fatal_error(proof.move_as_error()); + vm::CellBuilder cb; + Ref cell; + if (!(shards_dict.append_dict_to_bool(cb) && cb.finalize_to(cell))) { + fatal_error("cannot store ShardHashes from last mc block into a new cell"); return; } auto data = vm::std_boc_serialize(std::move(cell)); @@ -1929,10 +2018,309 @@ void LiteQuery::continue_getAllShardsInfo() { } LOG(INFO) << "getAllShardInfo() query completed"; auto b = ton::create_serialize_tl_object( - ton::create_tl_lite_block_id(base_blk_id_), proof.move_as_ok(), data.move_as_ok()); + ton::create_tl_lite_block_id(base_blk_id_), proof_boc.move_as_ok(), data.move_as_ok()); finish_query(std::move(b)); } +void LiteQuery::perform_lookupBlockWithProof(BlockId blkid, BlockIdExt mc_blkid, int mode, LogicalTime lt, UnixTime utime) { + if (!((1 << (mode & 7)) & 0x16)) { + fatal_error("exactly one of mode.0, mode.1 and mode.2 bits must be set"); + return; + } + if (!mc_blkid.is_masterchain_ext()) { + fatal_error("masterchain block id must be specified"); + return; + } + if (!(mode & 1)) { + blkid.seqno = 0; + } + if (!(mode & 2)) { + lt = 0; + } + if (!(mode & 4)) { + utime = 0; + } + mode_ = mode; + base_blk_id_ = mc_blkid; + LOG(INFO) << "started a lookupBlockWithProof(" << blkid.to_str() << ", " << mc_blkid.to_str() << ", " << mode << ", " + << lt << ", " << utime << ") liteserver query"; + + ton::AccountIdPrefixFull pfx{blkid.workchain, blkid.shard}; + auto P = td::PromiseCreator::lambda( + [Self = actor_id(this), mc_blkid, manager = manager_, mode, pfx](td::Result res) { + if (res.is_error()) { + td::actor::send_closure(Self, &LiteQuery::abort_query, res.move_as_error()); + return; + } + auto handle = res.move_as_ok(); + if (!handle->inited_masterchain_ref_block()) { + td::actor::send_closure(Self, &LiteQuery::abort_query, td::Status::Error("block doesn't have masterchain ref")); + return; + } + if (handle->masterchain_ref_block() > mc_blkid.seqno()) { + td::actor::send_closure(Self, &LiteQuery::abort_query, td::Status::Error("specified mc block is older than block's masterchain ref")); + return; + } + LOG(DEBUG) << "requesting data for block " << handle->id().to_str(); + td::actor::send_closure_later(manager, &ValidatorManager::get_block_data_from_db, handle, + [Self, mc_ref_blkid = handle->masterchain_ref_block(), mc_blkid, pfx, mode](td::Result> res) { + if (res.is_error()) { + td::actor::send_closure(Self, &LiteQuery::abort_query, res.move_as_error()); + } else { + td::actor::send_closure_later(Self, &LiteQuery::continue_lookupBlockWithProof_getHeaderProof, res.move_as_ok(), pfx, mc_ref_blkid); + } + }); + }); + + if (mode & 2) { + td::actor::send_closure_later(manager_, &ValidatorManager::get_block_by_lt_for_litequery, pfx, lt, std::move(P)); + } else if (mode & 4) { + td::actor::send_closure_later(manager_, &ValidatorManager::get_block_by_unix_time_for_litequery, pfx, utime, + std::move(P)); + } else { + td::actor::send_closure_later(manager_, &ValidatorManager::get_block_by_seqno_for_litequery, pfx, blkid.seqno, + std::move(P)); + } +} + +void LiteQuery::continue_lookupBlockWithProof_getHeaderProof(Ref block, AccountIdPrefixFull req_prefix, BlockSeqno masterchain_ref_seqno) { + blk_id_ = block->block_id(); + LOG(INFO) << "obtained data for getBlockHeader(" << blk_id_.to_str() << ", " << mode_ << ")"; + CHECK(block.not_null()); + auto block_root = block->root_cell(); + if (block_root.is_null()) { + fatal_error("block has no valid root cell"); + return; + } + + vm::MerkleProofBuilder mpb{block_root}; + std::vector prev; + BlockIdExt mc_blkid; + bool after_split; + td::Status S = block::unpack_block_prev_blk_try(mpb.root(), blk_id_, prev, mc_blkid, after_split); + if (S.is_error()) { + fatal_error(std::move(S)); + return; + } + auto proof_data = mpb.extract_proof_boc(); + if (proof_data.is_error()) { + fatal_error(proof_data.move_as_error()); + return; + } + lookup_header_proof_ = proof_data.move_as_ok(); + + bool include_prev = mode_ & 6; + if (include_prev) { + BlockIdExt prev_blkid; + for (auto& p : prev) { + if (ton::shard_contains(p.shard_full(), req_prefix)) { + prev_blkid = p; + } + } + CHECK(prev_blkid.is_valid()); + get_block_handle_checked(prev_blkid, [Self = actor_id(this), masterchain_ref_seqno, manager = manager_](td::Result R) mutable { + if (R.is_error()) { + td::actor::send_closure(Self, &LiteQuery::abort_query, R.move_as_error()); + return; + } + td::actor::send_closure(manager, &ValidatorManager::get_block_data_from_db, R.move_as_ok(), + [Self, masterchain_ref_seqno](td::Result> res) mutable { + if (res.is_error()) { + td::actor::send_closure(Self, &LiteQuery::abort_query, res.move_as_error()); + return; + } + td::actor::send_closure(Self, &LiteQuery::continue_lookupBlockWithProof_gotPrevBlockData, res.move_as_ok(), masterchain_ref_seqno); + }); + }); + } else { + continue_lookupBlockWithProof_gotPrevBlockData(Ref(), masterchain_ref_seqno); + } +} + +void LiteQuery::continue_lookupBlockWithProof_gotPrevBlockData(Ref prev_block, BlockSeqno masterchain_ref_seqno) { + if (prev_block.not_null()) { + CHECK(prev_block.not_null()); + if (prev_block->root_cell().is_null()) { + fatal_error("block has no valid root cell"); + return; + } + vm::MerkleProofBuilder mpb{prev_block->root_cell()}; + block::gen::Block::Record blk; + block::gen::BlockInfo::Record info; + if (!(tlb::unpack_cell(mpb.root(), blk) && tlb::unpack_cell(blk.info, info))) { + fatal_error(td::Status::Error("cannot unpack prev block header")); + return; + } + auto proof_data = mpb.extract_proof_boc(); + if (proof_data.is_error()) { + fatal_error(proof_data.move_as_error()); + return; + } + lookup_prev_header_proof_ = proof_data.move_as_ok(); + } + + if (!blk_id_.is_masterchain()) { + ton::AccountIdPrefixFull pfx{ton::masterchainId, ton::shardIdAll}; + td::actor::send_closure_later(manager_, &ValidatorManager::get_block_by_seqno_from_db, pfx, masterchain_ref_seqno, + [manager = manager_, Self = actor_id(this)](td::Result R) mutable { + if (R.is_error()) { + td::actor::send_closure(Self, &LiteQuery::abort_query, R.move_as_error()); + return; + } + td::actor::send_closure(manager, &ValidatorManager::get_block_data_from_db, R.move_as_ok(), + [Self](td::Result> res) mutable { + if (res.is_error()) { + td::actor::send_closure(Self, &LiteQuery::abort_query, res.move_as_error()); + return; + } + td::actor::send_closure(Self, &LiteQuery::continue_lookupBlockWithProof_buildProofLinks, res.move_as_ok(), std::vector>>()); + }); + }); + } else { + base_blk_id_alt_ = blk_id_; + td::actor::send_closure(actor_id(this), &LiteQuery::continue_lookupBlockWithProof_getClientMcBlockDataState, std::vector>>()); + } +} + +void LiteQuery::continue_lookupBlockWithProof_buildProofLinks(td::Ref cur_block, + std::vector>> result) { + BlockIdExt cur_id = cur_block->block_id(); + BlockIdExt prev_id; + vm::MerkleProofBuilder mpb{cur_block->root_cell()}; + if (cur_id.is_masterchain()) { + base_blk_id_alt_ = cur_id; + block::gen::Block::Record blk; + block::gen::BlockExtra::Record extra; + block::gen::McBlockExtra::Record mc_extra; + if (!tlb::unpack_cell(mpb.root(), blk) || !tlb::unpack_cell(blk.extra, extra) || !extra.custom->have_refs() || + !tlb::unpack_cell(extra.custom->prefetch_ref(), mc_extra)) { + fatal_error("cannot unpack header of block "s + cur_id.to_str()); + return; + } + block::ShardConfig shards(mc_extra.shard_hashes->prefetch_ref()); + ShardIdFull shard_id = blk_id_.shard_full(); + shard_id.shard = (shard_id.shard & ~(1 << (63 - shard_id.pfx_len()))) | 1; + Ref shard_hash = shards.get_shard_hash(shard_id, false); + if (shard_hash.is_null()) { + fatal_error("shard not found"); + return; + } + prev_id = shard_hash->top_block_id(); + } else { + std::vector prev; + BlockIdExt mc_blkid; + bool after_split; + td::Status S = block::unpack_block_prev_blk_try(mpb.root(), cur_id, prev, mc_blkid, after_split); + if (S.is_error()) { + fatal_error(std::move(S)); + return; + } + bool found = false; + for (const BlockIdExt& id : prev) { + if (shard_intersects(id.shard_full(), blk_id_.shard_full())) { + found = true; + prev_id = id; + break; + } + } + if (!found) { + fatal_error("failed to find block chain"); + return; + } + } + auto proof = mpb.extract_proof(); + if (proof.is_error()) { + fatal_error(proof.move_as_error_prefix("cannot serialize Merkle proof : ")); + return; + } + result.emplace_back(prev_id, proof.move_as_ok()); + + if (prev_id == blk_id_) { + CHECK(base_blk_id_alt_.is_masterchain()); + if (base_blk_id_alt_ != base_blk_id_) { + continue_lookupBlockWithProof_getClientMcBlockDataState(std::move(result)); + } else { + continue_lookupBlockWithProof_getMcBlockPrev(std::move(result)); + } + return; + } + if (result.size() == 8) { + // Chains of shardblocks between masterchain blocks can't be longer than 8 (see collator.cpp:991) + fatal_error("proof chain is too long"); + return; + } + + td::actor::send_closure_later( + manager_, &ValidatorManager::get_block_data_from_db_short, prev_id, + [Self = actor_id(this), result = std::move(result)](td::Result> R) mutable { + if (R.is_error()) { + td::actor::send_closure(Self, &LiteQuery::abort_query, R.move_as_error()); + } else { + td::actor::send_closure_later(Self, &LiteQuery::continue_lookupBlockWithProof_buildProofLinks, R.move_as_ok(), + std::move(result)); + } + }); +} + +void LiteQuery::continue_lookupBlockWithProof_getClientMcBlockDataState(std::vector>> links) { + set_continuation([this, links = std::move(links)]() -> void { + continue_lookupBlockWithProof_getMcBlockPrev(std::move(links)); + }); + request_mc_block_data_state(base_blk_id_); +} + +void LiteQuery::continue_lookupBlockWithProof_getMcBlockPrev(std::vector>> links) { + td::BufferSlice mc_state_proof_buf, client_mc_blk_proof_buf; + + if (base_blk_id_alt_ != base_blk_id_) { + vm::MerkleProofBuilder mpb{mc_state_->root_cell()}; + auto prev_blocks_dict = block::get_prev_blocks_dict(mpb.root()); + if (!prev_blocks_dict) { + fatal_error(td::Status::Error("cannot extract prev_blocks from mc state")); + return; + } + if (!block::check_old_mc_block_id(*prev_blocks_dict, base_blk_id_alt_)) { + fatal_error(td::Status::Error("client mc blkid is not in prev_blocks")); + return; + } + auto client_mc_blk_proof = mpb.extract_proof_boc(); + if (client_mc_blk_proof.is_error()) { + fatal_error(client_mc_blk_proof.move_as_error()); + return; + } + client_mc_blk_proof_buf = client_mc_blk_proof.move_as_ok(); + + Ref mc_state_proof; + if (!make_mc_state_root_proof(mc_state_proof)) { + fatal_error(td::Status::Error("cannot create Merkle proof for mc state")); + return; + } + auto mc_state_proof_boc = vm::std_boc_serialize(std::move(mc_state_proof)); + if (mc_state_proof_boc.is_error()) { + fatal_error(mc_state_proof_boc.move_as_error()); + return; + } + mc_state_proof_buf = mc_state_proof_boc.move_as_ok(); + } + + std::vector> links_res; + for (auto& p : links) { + auto prev_block_proof = vm::std_boc_serialize(std::move(p.second)); + if (prev_block_proof.is_error()) { + fatal_error(prev_block_proof.move_as_error()); + return; + } + links_res.push_back( + create_tl_object(create_tl_lite_block_id(p.first), prev_block_proof.move_as_ok())); + } + + auto b = ton::create_serialize_tl_object(ton::create_tl_lite_block_id(blk_id_), + mode_, ton::create_tl_lite_block_id(base_blk_id_alt_), std::move(mc_state_proof_buf), std::move(client_mc_blk_proof_buf), + std::move(links_res), std::move(lookup_header_proof_), std::move(lookup_prev_header_proof_)); + finish_query(std::move(b)); +} + + void LiteQuery::perform_lookupBlock(BlockId blkid, int mode, LogicalTime lt, UnixTime utime) { if (!((1 << (mode & 7)) & 0x16)) { fatal_error("exactly one of mode.0, mode.1 and mode.2 bits must be set"); @@ -1967,13 +2355,13 @@ void LiteQuery::perform_lookupBlock(BlockId blkid, int mode, LogicalTime lt, Uni ton::AccountIdPrefixFull pfx{blkid.workchain, blkid.shard}; if (mode & 2) { - td::actor::send_closure_later(manager_, &ValidatorManager::get_block_by_lt_from_db_for_litequery, pfx, lt, + td::actor::send_closure_later(manager_, &ValidatorManager::get_block_by_lt_for_litequery, pfx, lt, std::move(P)); } else if (mode & 4) { - td::actor::send_closure_later(manager_, &ValidatorManager::get_block_by_unix_time_from_db_for_litequery, pfx, utime, + td::actor::send_closure_later(manager_, &ValidatorManager::get_block_by_unix_time_for_litequery, pfx, utime, std::move(P)); } else { - td::actor::send_closure_later(manager_, &ValidatorManager::get_block_by_seqno_from_db_for_litequery, pfx, + td::actor::send_closure_later(manager_, &ValidatorManager::get_block_by_seqno_for_litequery, pfx, blkid.seqno, std::move(P)); } } @@ -2732,7 +3120,7 @@ void LiteQuery::perform_getShardBlockProof(BlockIdExt blkid) { } AccountIdPrefixFull pfx{masterchainId, shardIdAll}; td::actor::send_closure_later( - manager, &ValidatorManager::get_block_by_seqno_from_db_for_litequery, pfx, handle->masterchain_ref_block(), + manager, &ValidatorManager::get_block_by_seqno_for_litequery, pfx, handle->masterchain_ref_block(), [Self, manager](td::Result R) { if (R.is_error()) { td::actor::send_closure(Self, &LiteQuery::abort_query, R.move_as_error()); @@ -2836,5 +3224,92 @@ void LiteQuery::continue_getShardBlockProof(Ref cur_block, }); } +void LiteQuery::perform_getOutMsgQueueSizes(td::optional shard) { + LOG(INFO) << "started a getOutMsgQueueSizes" << (shard ? shard.value().to_str() : "") << " liteserver query"; + td::actor::send_closure_later( + manager_, &ton::validator::ValidatorManager::get_last_liteserver_state_block, + [Self = actor_id(this), shard](td::Result, BlockIdExt>> res) { + if (res.is_error()) { + td::actor::send_closure(Self, &LiteQuery::abort_query, res.move_as_error()); + } else { + td::actor::send_closure_later(Self, &LiteQuery::continue_getOutMsgQueueSizes, shard, res.ok().first); + } + }); +} + +void LiteQuery::continue_getOutMsgQueueSizes(td::optional shard, Ref state) { + std::vector blocks; + if (!shard || shard_intersects(shard.value(), state->get_shard())) { + blocks.push_back(state->get_block_id()); + } + for (auto& x : state->get_shards()) { + if (!shard || shard_intersects(shard.value(), x->shard())) { + blocks.push_back(x->top_block_id()); + } + } + auto res = std::make_shared>>(blocks.size()); + td::MultiPromise mp; + auto ig = mp.init_guard(); + for (size_t i = 0; i < blocks.size(); ++i) { + td::actor::send_closure(manager_, &ValidatorManager::get_out_msg_queue_size, blocks[i], + [promise = ig.get_promise(), res, i, id = blocks[i]](td::Result R) mutable { + TRY_RESULT_PROMISE(promise, value, std::move(R)); + res->at(i) = create_tl_object( + create_tl_lite_block_id(id), value); + promise.set_value(td::Unit()); + }); + } + ig.add_promise([Self = actor_id(this), res](td::Result R) { + if (R.is_error()) { + td::actor::send_closure(Self, &LiteQuery::abort_query, R.move_as_error()); + return; + } + td::actor::send_closure(Self, &LiteQuery::finish_query, + create_serialize_tl_object( + std::move(*res), Collator::get_skip_externals_queue_size()), + false); + }); +} + +void LiteQuery::perform_nonfinal_getCandidate(td::Bits256 source, BlockIdExt blkid, td::Bits256 collated_data_hash) { + LOG(INFO) << "started a nonfinal.getCandidate liteserver query"; + td::actor::send_closure_later( + manager_, &ValidatorManager::get_block_candidate_for_litequery, PublicKey{pubkeys::Ed25519{source}}, blkid, collated_data_hash, + [Self = actor_id(this)](td::Result R) { + if (R.is_error()) { + td::actor::send_closure(Self, &LiteQuery::abort_query, R.move_as_error()); + } else { + BlockCandidate cand = R.move_as_ok(); + td::actor::send_closure_later( + Self, &LiteQuery::finish_query, + create_serialize_tl_object( + create_tl_object( + create_tl_lite_block_id(cand.id), cand.pubkey.as_bits256(), cand.collated_file_hash), + std::move(cand.data), std::move(cand.collated_data)), + false); + } + }); +} + +void LiteQuery::perform_nonfinal_getValidatorGroups(int mode, ShardIdFull shard) { + bool with_shard = mode & 1; + LOG(INFO) << "started a nonfinal.getValidatorGroups" << (with_shard ? shard.to_str() : "(all)") + << " liteserver query"; + td::optional maybe_shard; + if (with_shard) { + maybe_shard = shard; + } + td::actor::send_closure( + manager_, &ValidatorManager::get_validator_groups_info_for_litequery, maybe_shard, + [Self = actor_id(this)](td::Result> R) { + if (R.is_error()) { + td::actor::send_closure(Self, &LiteQuery::abort_query, R.move_as_error()); + } else { + td::actor::send_closure_later(Self, &LiteQuery::finish_query, serialize_tl_object(R.move_as_ok(), true), + false); + } + }); +} + } // namespace validator } // namespace ton diff --git a/validator/impl/liteserver.hpp b/validator/impl/liteserver.hpp index 145fff3d..34e569c9 100644 --- a/validator/impl/liteserver.hpp +++ b/validator/impl/liteserver.hpp @@ -69,6 +69,9 @@ class LiteQuery : public td::actor::Actor { std::unique_ptr chain_; Ref stack_; + td::BufferSlice lookup_header_proof_; + td::BufferSlice lookup_prev_header_proof_; + public: enum { default_timeout_msec = 4500, // 4.5 seconds @@ -124,6 +127,8 @@ class LiteQuery : public td::actor::Actor { UnixTime gen_utime, LogicalTime gen_lt); void perform_getLibraries(std::vector library_list); void continue_getLibraries(Ref mc_state, BlockIdExt blkid, std::vector library_list); + void perform_getLibrariesWithProof(BlockIdExt blkid, int mode, std::vector library_list); + void continue_getLibrariesWithProof(std::vector library_list, int mode); void perform_getOneTransaction(BlockIdExt blkid, WorkchainId workchain, StdSmcAddress addr, LogicalTime lt); void continue_getOneTransaction(); void perform_getTransactions(WorkchainId workchain, StdSmcAddress addr, LogicalTime lt, Bits256 hash, unsigned count); @@ -138,6 +143,12 @@ class LiteQuery : public td::actor::Actor { void perform_getConfigParams(BlockIdExt blkid, int mode, std::vector param_list = {}); void continue_getConfigParams(int mode, std::vector param_list); void perform_lookupBlock(BlockId blkid, int mode, LogicalTime lt, UnixTime utime); + void perform_lookupBlockWithProof(BlockId blkid, BlockIdExt client_mc_blkid, int mode, LogicalTime lt, UnixTime utime); + void continue_lookupBlockWithProof_getHeaderProof(Ref block, AccountIdPrefixFull req_prefix, BlockSeqno masterchain_ref_seqno); + void continue_lookupBlockWithProof_gotPrevBlockData(Ref prev_block, BlockSeqno masterchain_ref_seqno); + void continue_lookupBlockWithProof_buildProofLinks(td::Ref cur_block, std::vector>> result); + void continue_lookupBlockWithProof_getClientMcBlockDataState(std::vector>> links); + void continue_lookupBlockWithProof_getMcBlockPrev(std::vector>> links); void perform_listBlockTransactions(BlockIdExt blkid, int mode, int count, Bits256 account, LogicalTime lt); void finish_listBlockTransactions(int mode, int count); void perform_listBlockTransactionsExt(BlockIdExt blkid, int mode, int count, Bits256 account, LogicalTime lt); @@ -157,6 +168,11 @@ class LiteQuery : public td::actor::Actor { void perform_getShardBlockProof(BlockIdExt blkid); void continue_getShardBlockProof(Ref cur_block, std::vector> result); + void perform_getOutMsgQueueSizes(td::optional shard); + void continue_getOutMsgQueueSizes(td::optional shard, Ref state); + + void perform_nonfinal_getCandidate(td::Bits256 source, BlockIdExt blkid, td::Bits256 collated_data_hash); + void perform_nonfinal_getValidatorGroups(int mode, ShardIdFull shard); void load_prevKeyBlock(ton::BlockIdExt blkid, td::Promise>>); void continue_loadPrevKeyBlock(ton::BlockIdExt blkid, td::Result, BlockIdExt>> res, diff --git a/validator/impl/validate-query.cpp b/validator/impl/validate-query.cpp index d925fce6..8b2723c3 100644 --- a/validator/impl/validate-query.cpp +++ b/validator/impl/validate-query.cpp @@ -947,6 +947,7 @@ bool ValidateQuery::fetch_config_params() { compute_phase_cfg_.mc_gas_prices = mc_gas_prices.move_as_ok(); compute_phase_cfg_.special_gas_full = config_->get_global_version() >= 5; storage_phase_cfg_.enable_due_payment = config_->get_global_version() >= 4; + storage_phase_cfg_.global_version = config_->get_global_version(); compute_phase_cfg_.block_rand_seed = rand_seed_; compute_phase_cfg_.libraries = std::make_unique(config_->get_libraries_root(), 256); compute_phase_cfg_.max_vm_data_depth = size_limits.max_vm_data_depth; diff --git a/validator/interfaces/validator-manager.h b/validator/interfaces/validator-manager.h index f0e1bb18..41412fb1 100644 --- a/validator/interfaces/validator-manager.h +++ b/validator/interfaces/validator-manager.h @@ -29,6 +29,7 @@ #include "liteserver.h" #include "crypto/vm/db/DynamicBagOfCellsDb.h" #include "validator-session/validator-session-types.h" +#include "auto/tl/lite_api.h" namespace ton { @@ -103,7 +104,8 @@ class ValidatorManager : public ValidatorManagerInterface { td::Promise> promise) = 0; virtual void wait_block_message_queue_short(BlockIdExt id, td::uint32 priority, td::Timestamp timeout, td::Promise> promise) = 0; - virtual void get_external_messages(ShardIdFull shard, td::Promise>> promise) = 0; + virtual void get_external_messages(ShardIdFull shard, + td::Promise, int>>> promise) = 0; virtual void get_ihr_messages(ShardIdFull shard, td::Promise>> promise) = 0; virtual void get_shard_blocks(BlockIdExt masterchain_block_id, td::Promise>> promise) = 0; @@ -171,12 +173,19 @@ class ValidatorManager : public ValidatorManagerInterface { virtual void log_validator_session_stats(BlockIdExt block_id, validatorsession::ValidatorSessionStats stats) = 0; virtual void get_block_handle_for_litequery(BlockIdExt block_id, td::Promise promise) = 0; - virtual void get_block_by_lt_from_db_for_litequery(AccountIdPrefixFull account, LogicalTime lt, - td::Promise promise) = 0; - virtual void get_block_by_unix_time_from_db_for_litequery(AccountIdPrefixFull account, UnixTime ts, - td::Promise promise) = 0; - virtual void get_block_by_seqno_from_db_for_litequery(AccountIdPrefixFull account, BlockSeqno seqno, - td::Promise promise) = 0; + virtual void get_block_data_for_litequery(BlockIdExt block_id, td::Promise> promise) = 0; + virtual void get_block_state_for_litequery(BlockIdExt block_id, td::Promise> promise) = 0; + virtual void get_block_by_lt_for_litequery(AccountIdPrefixFull account, LogicalTime lt, + td::Promise promise) = 0; + virtual void get_block_by_unix_time_for_litequery(AccountIdPrefixFull account, UnixTime ts, + td::Promise promise) = 0; + virtual void get_block_by_seqno_for_litequery(AccountIdPrefixFull account, BlockSeqno seqno, + td::Promise promise) = 0; + virtual void get_block_candidate_for_litequery(PublicKey source, BlockIdExt block_id, FileHash collated_data_hash, + td::Promise promise) = 0; + virtual void get_validator_groups_info_for_litequery( + td::optional shard, + td::Promise> promise) = 0; virtual void add_lite_query_stats(int lite_query_id) { } diff --git a/validator/manager-disk.cpp b/validator/manager-disk.cpp index 3717d7b0..42e44081 100644 --- a/validator/manager-disk.cpp +++ b/validator/manager-disk.cpp @@ -259,7 +259,7 @@ void ValidatorManagerImpl::get_key_block_proof_link(BlockIdExt block_id, td::Pro td::actor::send_closure(db_, &Db::get_key_block_proof, block_id, std::move(P)); } -void ValidatorManagerImpl::new_external_message(td::BufferSlice data) { +void ValidatorManagerImpl::new_external_message(td::BufferSlice data, int priority) { if (last_masterchain_state_.is_null()) { return; } @@ -507,9 +507,13 @@ void ValidatorManagerImpl::wait_block_message_queue_short(BlockIdExt block_id, t get_block_handle(block_id, true, std::move(P)); } -void ValidatorManagerImpl::get_external_messages(ShardIdFull shard, - td::Promise>> promise) { - promise.set_result(ext_messages_); +void ValidatorManagerImpl::get_external_messages( + ShardIdFull shard, td::Promise, int>>> promise) { + std::vector, int>> res; + for (const auto& x : ext_messages_) { + res.emplace_back(x, 0); + } + promise.set_result(std::move(res)); } void ValidatorManagerImpl::get_ihr_messages(ShardIdFull shard, td::Promise>> promise) { diff --git a/validator/manager-disk.hpp b/validator/manager-disk.hpp index 4812d60a..d5a6e909 100644 --- a/validator/manager-disk.hpp +++ b/validator/manager-disk.hpp @@ -124,7 +124,7 @@ class ValidatorManagerImpl : public ValidatorManager { void get_key_block_proof_link(BlockIdExt block_id, td::Promise promise) override; //void get_block_description(BlockIdExt block_id, td::Promise promise) override; - void new_external_message(td::BufferSlice data) override; + void new_external_message(td::BufferSlice data, int priority) override; void check_external_message(td::BufferSlice data, td::Promise> promise) override { UNREACHABLE(); } @@ -188,7 +188,8 @@ class ValidatorManagerImpl : public ValidatorManager { td::Promise> promise) override; void wait_block_message_queue_short(BlockIdExt id, td::uint32 priority, td::Timestamp timeout, td::Promise> promise) override; - void get_external_messages(ShardIdFull shard, td::Promise>> promise) override; + void get_external_messages(ShardIdFull shard, + td::Promise, int>>> promise) override; void get_ihr_messages(ShardIdFull shard, td::Promise>> promise) override; void get_shard_blocks(BlockIdExt masterchain_block_id, td::Promise>> promise) override; @@ -244,7 +245,7 @@ class ValidatorManagerImpl : public ValidatorManager { UNREACHABLE(); } void send_external_message(td::Ref message) override { - new_external_message(message->serialize()); + new_external_message(message->serialize(), 0); } void send_ihr_message(td::Ref message) override { new_ihr_message(message->serialize()); @@ -387,18 +388,33 @@ class ValidatorManagerImpl : public ValidatorManager { void get_block_handle_for_litequery(BlockIdExt block_id, td::Promise promise) override { get_block_handle(block_id, false, promise.wrap([](BlockHandle &&handle) -> ConstBlockHandle { return handle; })); } - void get_block_by_lt_from_db_for_litequery(AccountIdPrefixFull account, LogicalTime lt, + void get_block_data_for_litequery(BlockIdExt block_id, td::Promise> promise) override { + get_block_data_from_db_short(block_id, std::move(promise)); + } + void get_block_state_for_litequery(BlockIdExt block_id, td::Promise> promise) override { + get_shard_state_from_db_short(block_id, std::move(promise)); + } + void get_block_by_lt_for_litequery(AccountIdPrefixFull account, LogicalTime lt, td::Promise promise) override { get_block_by_lt_from_db(account, lt, std::move(promise)); } - void get_block_by_unix_time_from_db_for_litequery(AccountIdPrefixFull account, UnixTime ts, + void get_block_by_unix_time_for_litequery(AccountIdPrefixFull account, UnixTime ts, td::Promise promise) override { get_block_by_unix_time_from_db(account, ts, std::move(promise)); } - void get_block_by_seqno_from_db_for_litequery(AccountIdPrefixFull account, BlockSeqno seqno, + void get_block_by_seqno_for_litequery(AccountIdPrefixFull account, BlockSeqno seqno, td::Promise promise) override { get_block_by_seqno_from_db(account, seqno, std::move(promise)); } + void get_block_candidate_for_litequery(PublicKey source, BlockIdExt block_id, FileHash collated_data_hash, + td::Promise promise) override { + promise.set_result(td::Status::Error("not implemented")); + } + void get_validator_groups_info_for_litequery( + td::optional shard, + td::Promise> promise) override { + promise.set_result(td::Status::Error("not implemented")); + } private: PublicKeyHash local_id_; diff --git a/validator/manager-hardfork.cpp b/validator/manager-hardfork.cpp index e290f635..49d27085 100644 --- a/validator/manager-hardfork.cpp +++ b/validator/manager-hardfork.cpp @@ -150,7 +150,7 @@ void ValidatorManagerImpl::get_key_block_proof_link(BlockIdExt block_id, td::Pro td::actor::send_closure(db_, &Db::get_key_block_proof_link, block_id, std::move(P)); } -void ValidatorManagerImpl::new_external_message(td::BufferSlice data) { +void ValidatorManagerImpl::new_external_message(td::BufferSlice data, int priority) { auto R = create_ext_message(std::move(data), block::SizeLimitsConfig::ExtMsgLimits()); if (R.is_ok()) { ext_messages_.emplace_back(R.move_as_ok()); @@ -357,9 +357,13 @@ void ValidatorManagerImpl::wait_block_message_queue_short(BlockIdExt block_id, t get_block_handle(block_id, true, std::move(P)); } -void ValidatorManagerImpl::get_external_messages(ShardIdFull shard, - td::Promise>> promise) { - promise.set_result(ext_messages_); +void ValidatorManagerImpl::get_external_messages( + ShardIdFull shard, td::Promise, int>>> promise) { + std::vector, int>> res; + for (const auto &x : ext_messages_) { + res.emplace_back(x, 0); + } + promise.set_result(std::move(res)); } void ValidatorManagerImpl::get_ihr_messages(ShardIdFull shard, td::Promise>> promise) { diff --git a/validator/manager-hardfork.hpp b/validator/manager-hardfork.hpp index c34ae5c7..7937729c 100644 --- a/validator/manager-hardfork.hpp +++ b/validator/manager-hardfork.hpp @@ -144,7 +144,7 @@ class ValidatorManagerImpl : public ValidatorManager { void get_key_block_proof(BlockIdExt block_id, td::Promise promise) override; void get_key_block_proof_link(BlockIdExt block_id, td::Promise promise) override; - void new_external_message(td::BufferSlice data) override; + void new_external_message(td::BufferSlice data, int priority) override; void check_external_message(td::BufferSlice data, td::Promise> promise) override { UNREACHABLE(); } @@ -228,7 +228,8 @@ class ValidatorManagerImpl : public ValidatorManager { td::Promise> promise) override; void wait_block_message_queue_short(BlockIdExt id, td::uint32 priority, td::Timestamp timeout, td::Promise> promise) override; - void get_external_messages(ShardIdFull shard, td::Promise>> promise) override; + void get_external_messages(ShardIdFull shard, + td::Promise, int>>> promise) override; void get_ihr_messages(ShardIdFull shard, td::Promise>> promise) override; void get_shard_blocks(BlockIdExt masterchain_block_id, td::Promise>> promise) override; @@ -308,7 +309,7 @@ class ValidatorManagerImpl : public ValidatorManager { UNREACHABLE(); } void send_external_message(td::Ref message) override { - new_external_message(message->serialize()); + new_external_message(message->serialize(), 0); } void send_ihr_message(td::Ref message) override { new_ihr_message(message->serialize()); @@ -448,18 +449,33 @@ class ValidatorManagerImpl : public ValidatorManager { void get_block_handle_for_litequery(BlockIdExt block_id, td::Promise promise) override { get_block_handle(block_id, false, promise.wrap([](BlockHandle &&handle) -> ConstBlockHandle { return handle; })); } - void get_block_by_lt_from_db_for_litequery(AccountIdPrefixFull account, LogicalTime lt, + void get_block_data_for_litequery(BlockIdExt block_id, td::Promise> promise) override { + get_block_data_from_db_short(block_id, std::move(promise)); + } + void get_block_state_for_litequery(BlockIdExt block_id, td::Promise> promise) override { + get_shard_state_from_db_short(block_id, std::move(promise)); + } + void get_block_by_lt_for_litequery(AccountIdPrefixFull account, LogicalTime lt, td::Promise promise) override { get_block_by_lt_from_db(account, lt, std::move(promise)); } - void get_block_by_unix_time_from_db_for_litequery(AccountIdPrefixFull account, UnixTime ts, + void get_block_by_unix_time_for_litequery(AccountIdPrefixFull account, UnixTime ts, td::Promise promise) override { get_block_by_unix_time_from_db(account, ts, std::move(promise)); } - void get_block_by_seqno_from_db_for_litequery(AccountIdPrefixFull account, BlockSeqno seqno, + void get_block_by_seqno_for_litequery(AccountIdPrefixFull account, BlockSeqno seqno, td::Promise promise) override { get_block_by_seqno_from_db(account, seqno, std::move(promise)); } + void get_block_candidate_for_litequery(PublicKey source, BlockIdExt block_id, FileHash collated_data_hash, + td::Promise promise) override { + promise.set_result(td::Status::Error("not implemented")); + } + void get_validator_groups_info_for_litequery( + td::optional shard, + td::Promise> promise) override { + promise.set_result(td::Status::Error("not implemented")); + } private: td::Ref opts_; diff --git a/validator/manager.cpp b/validator/manager.cpp index 55891ead..e186902a 100644 --- a/validator/manager.cpp +++ b/validator/manager.cpp @@ -225,13 +225,13 @@ void ValidatorManagerImpl::sync_complete(td::Promise promise) { VLOG(VALIDATOR_WARNING) << "completed sync. Validating " << validator_groups_.size() << " groups"; for (auto &v : validator_groups_) { - if (!v.second.empty()) { - td::actor::send_closure(v.second, &ValidatorGroup::create_session); + if (!v.second.actor.empty()) { + td::actor::send_closure(v.second.actor, &ValidatorGroup::create_session); } } for (auto &v : next_validator_groups_) { - if (!v.second.empty()) { - td::actor::send_closure(v.second, &ValidatorGroup::create_session); + if (!v.second.actor.empty()) { + td::actor::send_closure(v.second.actor, &ValidatorGroup::create_session); } } } @@ -368,7 +368,7 @@ void ValidatorManagerImpl::get_key_block_proof_link(BlockIdExt block_id, td::Pro td::actor::send_closure(db_, &Db::get_key_block_proof, block_id, std::move(P)); } -void ValidatorManagerImpl::new_external_message(td::BufferSlice data) { +void ValidatorManagerImpl::new_external_message(td::BufferSlice data, int priority) { if (!is_validator()) { return; } @@ -376,7 +376,7 @@ void ValidatorManagerImpl::new_external_message(td::BufferSlice data) { VLOG(VALIDATOR_NOTICE) << "dropping ext message: validator is not ready"; return; } - if (ext_messages_.size() > max_mempool_num()) { + if (ext_msgs_[priority].ext_messages_.size() > (size_t)max_mempool_num()) { return; } auto R = create_ext_message(std::move(data), last_masterchain_state_->get_ext_msg_limits()); @@ -384,21 +384,30 @@ void ValidatorManagerImpl::new_external_message(td::BufferSlice data) { VLOG(VALIDATOR_NOTICE) << "dropping bad ext message: " << R.move_as_error(); return; } - add_external_message(R.move_as_ok()); + add_external_message(R.move_as_ok(), priority); } -void ValidatorManagerImpl::add_external_message(td::Ref msg) { +void ValidatorManagerImpl::add_external_message(td::Ref msg, int priority) { + auto &msgs = ext_msgs_[priority]; auto message = std::make_unique>(msg); auto id = message->ext_id(); auto address = message->address(); unsigned long per_address_limit = 256; - if(ext_addr_messages_.count(address) < per_address_limit) { - if (ext_messages_hashes_.count(id.hash) == 0) { - ext_messages_.emplace(id, std::move(message)); - ext_messages_hashes_.emplace(id.hash, id); - ext_addr_messages_[address].emplace(id.hash, id); - } + auto it = msgs.ext_addr_messages_.find(address); + if (it != msgs.ext_addr_messages_.end() && it->second.size() >= per_address_limit) { + return; } + auto it2 = ext_messages_hashes_.find(id.hash); + if (it2 != ext_messages_hashes_.end()) { + int old_priority = it2->second.first; + if (old_priority >= priority) { + return; + } + ext_msgs_[old_priority].erase(id); + } + msgs.ext_messages_.emplace(id, std::move(message)); + msgs.ext_addr_messages_[address].emplace(id.hash, id); + ext_messages_hashes_[id.hash] = {priority, id}; } void ValidatorManagerImpl::check_external_message(td::BufferSlice data, td::Promise> promise) { ++ls_stats_check_ext_messages_; @@ -783,34 +792,44 @@ void ValidatorManagerImpl::wait_block_message_queue_short(BlockIdExt block_id, t get_block_handle(block_id, true, std::move(P)); } -void ValidatorManagerImpl::get_external_messages(ShardIdFull shard, - td::Promise>> promise) { +void ValidatorManagerImpl::get_external_messages( + ShardIdFull shard, td::Promise, int>>> promise) { td::Timer t; size_t processed = 0, deleted = 0; - std::vector> res; + std::vector, int>> res; MessageId left{AccountIdPrefixFull{shard.workchain, shard.shard & (shard.shard - 1)}, Bits256::zero()}; - auto it = ext_messages_.lower_bound(left); - while (it != ext_messages_.end()) { - auto s = it->first; - if (!shard_contains(shard, s.dst)) { - break; + size_t total_msgs = 0; + td::Random::Fast rnd; + for (auto iter = ext_msgs_.rbegin(); iter != ext_msgs_.rend(); ++iter) { + std::vector, int>> cur_res; + int priority = iter->first; + auto &msgs = iter->second; + auto it = msgs.ext_messages_.lower_bound(left); + while (it != msgs.ext_messages_.end()) { + auto s = it->first; + if (!shard_contains(shard, s.dst)) { + break; + } + ++processed; + if (it->second->expired()) { + msgs.ext_addr_messages_[it->second->address()].erase(it->first.hash); + ext_messages_hashes_.erase(it->first.hash); + it = msgs.ext_messages_.erase(it); + ++deleted; + continue; + } + if (it->second->is_active()) { + cur_res.emplace_back(it->second->message(), priority); + } + it++; } - ++processed; - if (it->second->expired()) { - ext_addr_messages_[it->second->address()].erase(it->first.hash); - ext_messages_hashes_.erase(it->first.hash); - it = ext_messages_.erase(it); - ++deleted; - continue; - } - if (it->second->is_active()) { - res.push_back(it->second->message()); - } - it++; + td::random_shuffle(td::as_mutable_span(cur_res), rnd); + res.insert(res.end(), cur_res.begin(), cur_res.end()); + total_msgs += msgs.ext_messages_.size(); } LOG(WARNING) << "get_external_messages to shard " << shard.to_str() << " : time=" << t.elapsed() << " result_size=" << res.size() << " processed=" << processed << " expired=" << deleted - << " total_size=" << ext_messages_.size(); + << " total_size=" << total_msgs; promise.set_value(std::move(res)); } @@ -850,8 +869,9 @@ void ValidatorManagerImpl::complete_external_messages(std::vectorsecond]->address()].erase(it->first); - CHECK(ext_messages_.erase(it->second)); + int priority = it->second.first; + auto msg_id = it->second.second; + ext_msgs_[priority].erase(msg_id); ext_messages_hashes_.erase(it); } } @@ -859,12 +879,14 @@ void ValidatorManagerImpl::complete_external_messages(std::vectorsecond); - if ((ext_messages_.size() < soft_mempool_limit) && it2->second->can_postpone()) { + int priority = it->second.first; + auto msg_id = it->second.second; + auto &msgs = ext_msgs_[priority]; + auto it2 = msgs.ext_messages_.find(msg_id); + if ((msgs.ext_messages_.size() < soft_mempool_limit) && it2->second->can_postpone()) { it2->second->postpone(); } else { - ext_addr_messages_[it2->second->address()].erase(it2->first.hash); - ext_messages_.erase(it2); + msgs.erase(msg_id); ext_messages_hashes_.erase(it); } } @@ -1168,6 +1190,10 @@ void ValidatorManagerImpl::set_next_block(BlockIdExt block_id, BlockIdExt next, } void ValidatorManagerImpl::set_block_candidate(BlockIdExt id, BlockCandidate candidate, td::Promise promise) { + if (!candidates_buffer_.empty()) { + td::actor::send_closure(candidates_buffer_, &CandidatesBuffer::add_new_candidate, id, + PublicKey{pubkeys::Ed25519{candidate.pubkey.as_bits256()}}, candidate.collated_file_hash); + } td::actor::send_closure(db_, &Db::store_block_candidate, std::move(candidate), std::move(promise)); } @@ -1456,7 +1482,7 @@ void ValidatorManagerImpl::send_get_next_key_blocks_request(BlockIdExt block_id, void ValidatorManagerImpl::send_external_message(td::Ref message) { callback_->send_ext_message(message->shard(), message->serialize()); - add_external_message(std::move(message)); + add_external_message(std::move(message), 0); } void ValidatorManagerImpl::send_ihr_message(td::Ref message) { @@ -1575,8 +1601,11 @@ void ValidatorManagerImpl::started(ValidatorManagerInitResult R) { td::actor::send_closure(SelfId, &ValidatorManagerImpl::read_gc_list, R.move_as_ok()); } }); - td::actor::send_closure(db_, &Db::get_destroyed_validator_sessions, std::move(P)); + + if (opts_->nonfinal_ls_queries_enabled()) { + candidates_buffer_ = td::actor::create_actor("candidates-buffer", actor_id(this)); + } } void ValidatorManagerImpl::read_gc_list(std::vector list) { @@ -1868,8 +1897,8 @@ void ValidatorManagerImpl::update_shards() { VLOG(VALIDATOR_DEBUG) << "total shards=" << new_shards.size() << " config shards=" << exp_vec.size(); - std::map> new_validator_groups_; - std::map> new_next_validator_groups_; + std::map new_validator_groups_; + std::map new_next_validator_groups_; bool force_recover = false; { @@ -1901,8 +1930,8 @@ void ValidatorManagerImpl::update_shards() { } else { auto it2 = next_validator_groups_.find(legacy_val_group_id); if (it2 != next_validator_groups_.end()) { - if (!it2->second.empty()) { - td::actor::send_closure(it2->second, &ValidatorGroup::start, prev, last_masterchain_block_id_, + if (!it2->second.actor.empty()) { + td::actor::send_closure(it2->second.actor, &ValidatorGroup::start, prev, last_masterchain_block_id_, last_masterchain_state_->get_unix_time()); } new_validator_groups_.emplace(val_group_id, std::move(it2->second)); @@ -1912,7 +1941,7 @@ void ValidatorManagerImpl::update_shards() { td::actor::send_closure(G, &ValidatorGroup::start, prev, last_masterchain_block_id_, last_masterchain_state_->get_unix_time()); } - new_validator_groups_.emplace(val_group_id, std::move(G)); + new_validator_groups_.emplace(val_group_id, ValidatorGroupEntry{std::move(G), shard}); } } } @@ -1957,8 +1986,8 @@ void ValidatorManagerImpl::update_shards() { } else { auto it2 = next_validator_groups_.find(val_group_id); if (it2 != next_validator_groups_.end()) { - if (!it2->second.empty()) { - td::actor::send_closure(it2->second, &ValidatorGroup::start, prev, last_masterchain_block_id_, + if (!it2->second.actor.empty()) { + td::actor::send_closure(it2->second.actor, &ValidatorGroup::start, prev, last_masterchain_block_id_, last_masterchain_state_->get_unix_time()); } new_validator_groups_.emplace(val_group_id, std::move(it2->second)); @@ -1968,7 +1997,7 @@ void ValidatorManagerImpl::update_shards() { td::actor::send_closure(G, &ValidatorGroup::start, prev, last_masterchain_block_id_, last_masterchain_state_->get_unix_time()); } - new_validator_groups_.emplace(val_group_id, std::move(G)); + new_validator_groups_.emplace(val_group_id, ValidatorGroupEntry{std::move(G), shard}); } } } @@ -1988,23 +2017,24 @@ void ValidatorManagerImpl::update_shards() { //CHECK(!it->second.empty()); new_next_validator_groups_.emplace(val_group_id, std::move(it->second)); } else { - new_next_validator_groups_.emplace(val_group_id, - create_validator_group(val_group_id, shard, val_set, opts, started_)); + new_next_validator_groups_.emplace( + val_group_id, + ValidatorGroupEntry{create_validator_group(val_group_id, shard, val_set, opts, started_), shard}); } } } std::vector> gc; for (auto &v : validator_groups_) { - if (!v.second.empty()) { + if (!v.second.actor.empty()) { gc_list_.push_back(v.first); - gc.push_back(v.second.release()); + gc.push_back(v.second.actor.release()); } } for (auto &v : next_validator_groups_) { - if (!v.second.empty()) { + if (!v.second.actor.empty()) { gc_list_.push_back(v.first); - gc.push_back(v.second.release()); + gc.push_back(v.second.actor.release()); } } @@ -2097,6 +2127,9 @@ td::actor::ActorOwn ValidatorManagerImpl::create_validator_group if (check_gc_list_.count(session_id) == 1) { return td::actor::ActorOwn{}; } else { + // Call get_external_messages to cleanup mempool for the shard + get_external_messages(shard, [](td::Result, int>>>) {}); + auto validator_id = get_validator(shard, validator_set); CHECK(!validator_id.is_zero()); auto G = td::actor::create_actor( @@ -2704,19 +2737,72 @@ void ValidatorManagerImpl::log_validator_session_stats(BlockIdExt block_id, } void ValidatorManagerImpl::get_block_handle_for_litequery(BlockIdExt block_id, td::Promise promise) { - get_block_handle( - block_id, false, - [SelfId = actor_id(this), block_id, promise = std::move(promise)](td::Result R) mutable { - if (R.is_ok() && R.ok()->is_applied()) { - promise.set_value(R.move_as_ok()); - } else { - td::actor::send_closure(SelfId, &ValidatorManagerImpl::process_block_handle_for_litequery_error, block_id, - std::move(R), std::move(promise)); - } - }); + get_block_handle(block_id, false, + [SelfId = actor_id(this), block_id, promise = std::move(promise), + allow_not_applied = opts_->nonfinal_ls_queries_enabled()](td::Result R) mutable { + if (R.is_ok() && (allow_not_applied || R.ok()->is_applied())) { + promise.set_value(R.move_as_ok()); + } else { + td::actor::send_closure(SelfId, &ValidatorManagerImpl::process_block_handle_for_litequery_error, + block_id, std::move(R), std::move(promise)); + } + }); } -void ValidatorManagerImpl::get_block_by_lt_from_db_for_litequery(AccountIdPrefixFull account, LogicalTime lt, +void ValidatorManagerImpl::get_block_data_for_litequery(BlockIdExt block_id, td::Promise> promise) { + if (candidates_buffer_.empty()) { + get_block_handle_for_litequery( + block_id, [manager = actor_id(this), promise = std::move(promise)](td::Result R) mutable { + TRY_RESULT_PROMISE(promise, handle, std::move(R)); + td::actor::send_closure_later(manager, &ValidatorManager::get_block_data_from_db, std::move(handle), + std::move(promise)); + }); + } else { + td::actor::send_closure( + candidates_buffer_, &CandidatesBuffer::get_block_data, block_id, + [manager = actor_id(this), promise = std::move(promise), block_id](td::Result> R) mutable { + if (R.is_ok()) { + promise.set_result(R.move_as_ok()); + return; + } + td::actor::send_closure(manager, &ValidatorManagerImpl::get_block_handle_for_litequery, block_id, + [manager, promise = std::move(promise)](td::Result R) mutable { + TRY_RESULT_PROMISE(promise, handle, std::move(R)); + td::actor::send_closure_later(manager, &ValidatorManager::get_block_data_from_db, + std::move(handle), std::move(promise)); + }); + }); + } +} + +void ValidatorManagerImpl::get_block_state_for_litequery(BlockIdExt block_id, + td::Promise> promise) { + if (candidates_buffer_.empty()) { + get_block_handle_for_litequery( + block_id, [manager = actor_id(this), promise = std::move(promise)](td::Result R) mutable { + TRY_RESULT_PROMISE(promise, handle, std::move(R)); + td::actor::send_closure_later(manager, &ValidatorManager::get_shard_state_from_db, std::move(handle), + std::move(promise)); + }); + } else { + td::actor::send_closure( + candidates_buffer_, &CandidatesBuffer::get_block_state, block_id, + [manager = actor_id(this), promise = std::move(promise), block_id](td::Result> R) mutable { + if (R.is_ok()) { + promise.set_result(R.move_as_ok()); + return; + } + td::actor::send_closure(manager, &ValidatorManagerImpl::get_block_handle_for_litequery, + block_id, [manager, promise = std::move(promise)](td::Result R) mutable { + TRY_RESULT_PROMISE(promise, handle, std::move(R)); + td::actor::send_closure_later(manager, &ValidatorManager::get_shard_state_from_db, std::move(handle), + std::move(promise)); + }); + }); + } +} + +void ValidatorManagerImpl::get_block_by_lt_for_litequery(AccountIdPrefixFull account, LogicalTime lt, td::Promise promise) { get_block_by_lt_from_db( account, lt, [=, SelfId = actor_id(this), promise = std::move(promise)](td::Result R) mutable { @@ -2729,7 +2815,7 @@ void ValidatorManagerImpl::get_block_by_lt_from_db_for_litequery(AccountIdPrefix }); } -void ValidatorManagerImpl::get_block_by_unix_time_from_db_for_litequery(AccountIdPrefixFull account, UnixTime ts, +void ValidatorManagerImpl::get_block_by_unix_time_for_litequery(AccountIdPrefixFull account, UnixTime ts, td::Promise promise) { get_block_by_unix_time_from_db( account, ts, [=, SelfId = actor_id(this), promise = std::move(promise)](td::Result R) mutable { @@ -2742,7 +2828,7 @@ void ValidatorManagerImpl::get_block_by_unix_time_from_db_for_litequery(AccountI }); } -void ValidatorManagerImpl::get_block_by_seqno_from_db_for_litequery(AccountIdPrefixFull account, BlockSeqno seqno, +void ValidatorManagerImpl::get_block_by_seqno_for_litequery(AccountIdPrefixFull account, BlockSeqno seqno, td::Promise promise) { get_block_by_seqno_from_db( account, seqno, @@ -2840,6 +2926,77 @@ void ValidatorManagerImpl::process_lookup_block_for_litequery_error(AccountIdPre promise.set_error(std::move(err)); } +void ValidatorManagerImpl::get_block_candidate_for_litequery(PublicKey source, BlockIdExt block_id, + FileHash collated_data_hash, + td::Promise promise) { + if (!opts_->nonfinal_ls_queries_enabled()) { + promise.set_error(td::Status::Error("query is not allowed")); + return; + } + get_block_candidate_from_db(source, block_id, collated_data_hash, std::move(promise)); +} + +void ValidatorManagerImpl::get_validator_groups_info_for_litequery( + td::optional shard, + td::Promise> promise) { + if (!opts_->nonfinal_ls_queries_enabled()) { + promise.set_error(td::Status::Error("query is not allowed")); + return; + } + class Actor : public td::actor::Actor { + public: + explicit Actor(std::vector> groups, + td::Promise> promise) + : groups_(std::move(groups)), promise_(std::move(promise)) { + } + + void start_up() override { + pending_ = groups_.size(); + if (pending_ == 0) { + promise_.set_result(std::move(result_)); + stop(); + return; + } + for (auto &x : groups_) { + td::actor::send_closure( + x, &ValidatorGroup::get_validator_group_info_for_litequery, + [SelfId = actor_id(this)](td::Result> R) { + td::actor::send_closure(SelfId, &Actor::on_result, R.is_ok() ? R.move_as_ok() : nullptr); + }); + } + } + + void on_result(tl_object_ptr r) { + if (r) { + result_->groups_.push_back(std::move(r)); + } + --pending_; + if (pending_ == 0) { + promise_.set_result(std::move(result_)); + stop(); + } + } + + private: + std::vector> groups_; + size_t pending_; + td::Promise> promise_; + tl_object_ptr result_ = + create_tl_object(); + }; + std::vector> groups; + for (auto &x : validator_groups_) { + if (x.second.actor.empty()) { + continue; + } + if (shard && shard.value() != x.second.shard) { + continue; + } + groups.push_back(x.second.actor.get()); + } + td::actor::create_actor("get-validator-groups-info", std::move(groups), std::move(promise)).release(); +} + td::actor::ActorOwn ValidatorManagerFactory::create( td::Ref opts, std::string db_root, td::actor::ActorId keyring, td::actor::ActorId adnl, td::actor::ActorId rldp, diff --git a/validator/manager.hpp b/validator/manager.hpp index 5a1fbc60..7e5930d3 100644 --- a/validator/manager.hpp +++ b/validator/manager.hpp @@ -29,6 +29,7 @@ #include "rldp/rldp.h" #include "token-manager.h" #include "queue-size-counter.hpp" +#include "impl/candidates-buffer.hpp" #include #include @@ -219,9 +220,19 @@ class ValidatorManagerImpl : public ValidatorManager { }; // DATA FOR COLLATOR std::map> shard_blocks_; - std::map, std::unique_ptr>> ext_messages_; - std::map, std::map>> ext_addr_messages_; - std::map> ext_messages_hashes_; + struct ExtMessages { + std::map, std::unique_ptr>> ext_messages_; + std::map, std::map>> + ext_addr_messages_; + void erase(const MessageId& id) { + auto it = ext_messages_.find(id); + CHECK(it != ext_messages_.end()); + ext_addr_messages_[it->second->address()].erase(id.hash); + ext_messages_.erase(it); + } + }; + std::map ext_msgs_; // priority -> messages + std::map>> ext_messages_hashes_; // hash -> priority // IHR ? std::map, std::unique_ptr>> ihr_messages_; std::map> ihr_messages_hashes_; @@ -235,8 +246,12 @@ class ValidatorManagerImpl : public ValidatorManager { td::Ref validator_set, validatorsession::ValidatorSessionOptions opts, bool create_catchain); - std::map> validator_groups_; - std::map> next_validator_groups_; + struct ValidatorGroupEntry { + td::actor::ActorOwn actor; + ShardIdFull shard; + }; + std::map validator_groups_; + std::map next_validator_groups_; std::set check_gc_list_; std::vector gc_list_; @@ -344,8 +359,8 @@ class ValidatorManagerImpl : public ValidatorManager { void get_key_block_proof_link(BlockIdExt block_id, td::Promise promise) override; //void get_block_description(BlockIdExt block_id, td::Promise promise) override; - void new_external_message(td::BufferSlice data) override; - void add_external_message(td::Ref message); + void new_external_message(td::BufferSlice data, int priority) override; + void add_external_message(td::Ref message, int priority); void check_external_message(td::BufferSlice data, td::Promise> promise) override; void new_ihr_message(td::BufferSlice data) override; @@ -405,7 +420,8 @@ class ValidatorManagerImpl : public ValidatorManager { td::Promise> promise) override; void wait_block_message_queue_short(BlockIdExt id, td::uint32 priority, td::Timestamp timeout, td::Promise> promise) override; - void get_external_messages(ShardIdFull shard, td::Promise>> promise) override; + void get_external_messages(ShardIdFull shard, + td::Promise, int>>> promise) override; void get_ihr_messages(ShardIdFull shard, td::Promise>> promise) override; void get_shard_blocks(BlockIdExt masterchain_block_id, td::Promise>> promise) override; @@ -563,17 +579,24 @@ class ValidatorManagerImpl : public ValidatorManager { } void get_block_handle_for_litequery(BlockIdExt block_id, td::Promise promise) override; - void get_block_by_lt_from_db_for_litequery(AccountIdPrefixFull account, LogicalTime lt, + void get_block_data_for_litequery(BlockIdExt block_id, td::Promise> promise) override; + void get_block_state_for_litequery(BlockIdExt block_id, td::Promise> promise) override; + void get_block_by_lt_for_litequery(AccountIdPrefixFull account, LogicalTime lt, td::Promise promise) override; - void get_block_by_unix_time_from_db_for_litequery(AccountIdPrefixFull account, UnixTime ts, + void get_block_by_unix_time_for_litequery(AccountIdPrefixFull account, UnixTime ts, td::Promise promise) override; - void get_block_by_seqno_from_db_for_litequery(AccountIdPrefixFull account, BlockSeqno seqno, + void get_block_by_seqno_for_litequery(AccountIdPrefixFull account, BlockSeqno seqno, td::Promise promise) override; void process_block_handle_for_litequery_error(BlockIdExt block_id, td::Result r_handle, td::Promise promise); void process_lookup_block_for_litequery_error(AccountIdPrefixFull account, int type, td::uint64 value, td::Result r_handle, td::Promise promise); + void get_block_candidate_for_litequery(PublicKey source, BlockIdExt block_id, FileHash collated_data_hash, + td::Promise promise) override; + void get_validator_groups_info_for_litequery( + td::optional shard, + td::Promise> promise) override; void add_lite_query_stats(int lite_query_id) override { ++ls_stats_[lite_query_id]; @@ -648,6 +671,8 @@ class ValidatorManagerImpl : public ValidatorManager { td::Timestamp log_ls_stats_at_; std::map ls_stats_; // lite_api ID -> count, 0 for unknown td::uint32 ls_stats_check_ext_messages_{0}; + + td::actor::ActorOwn candidates_buffer_; }; } // namespace validator diff --git a/validator/net/download-block-new.cpp b/validator/net/download-block-new.cpp index 14754f64..9ec36e33 100644 --- a/validator/net/download-block-new.cpp +++ b/validator/net/download-block-new.cpp @@ -23,6 +23,7 @@ #include "td/utils/overloaded.h" #include "ton/ton-io.hpp" #include "validator/full-node.h" +#include "full-node-serializer.hpp" namespace ton { @@ -219,52 +220,54 @@ void DownloadBlockNew::got_data(td::BufferSlice data) { } auto f = F.move_as_ok(); + if (f->get_id() == ton_api::tonNode_dataFullEmpty::ID) { + abort_query(td::Status::Error(ErrorCode::notready, "node doesn't have this block")); + return; + } + BlockIdExt id; + td::BufferSlice proof, block_data; + bool is_link; + td::Status S = deserialize_block_full(*f, id, proof, block_data, is_link, overlay::Overlays::max_fec_broadcast_size()); + if (S.is_error()) { + abort_query(S.move_as_error_prefix("cannot deserialize block: ")); + return; + } - ton_api::downcast_call( - *f.get(), - td::overloaded( - [&](ton_api::tonNode_dataFullEmpty &x) { - abort_query(td::Status::Error(ErrorCode::notready, "node doesn't have this block")); - }, - [&, self = this](ton_api::tonNode_dataFull &x) { - if (!allow_partial_proof_ && x.is_link_) { - abort_query(td::Status::Error(ErrorCode::notready, "node doesn't have proof for this block")); - return; - } - auto id = create_block_id(x.id_); - if (block_id_.is_valid() && id != block_id_) { - abort_query(td::Status::Error(ErrorCode::notready, "received data for wrong block")); - return; - } - block_.id = id; - block_.data = std::move(x.block_); - if (td::sha256_bits256(block_.data.as_slice()) != id.file_hash) { - abort_query(td::Status::Error(ErrorCode::notready, "received data with bad hash")); - return; - } + if (!allow_partial_proof_ && is_link) { + abort_query(td::Status::Error(ErrorCode::notready, "node doesn't have proof for this block")); + return; + } + if (block_id_.is_valid() && id != block_id_) { + abort_query(td::Status::Error(ErrorCode::notready, "received data for wrong block")); + return; + } + block_.id = id; + block_.data = std::move(block_data); + if (td::sha256_bits256(block_.data.as_slice()) != id.file_hash) { + abort_query(td::Status::Error(ErrorCode::notready, "received data with bad hash")); + return; + } - auto P = td::PromiseCreator::lambda([SelfId = actor_id(self)](td::Result R) { - if (R.is_error()) { - td::actor::send_closure(SelfId, &DownloadBlockNew::abort_query, - R.move_as_error_prefix("received bad proof: ")); - } else { - td::actor::send_closure(SelfId, &DownloadBlockNew::checked_block_proof); - } - }); - if (block_id_.is_valid()) { - if (x.is_link_) { - td::actor::send_closure(validator_manager_, &ValidatorManagerInterface::validate_block_proof_link, - block_id_, std::move(x.proof_), std::move(P)); - } else { - td::actor::send_closure(validator_manager_, &ValidatorManagerInterface::validate_block_proof, block_id_, - std::move(x.proof_), std::move(P)); - } - } else { - CHECK(!x.is_link_); - td::actor::send_closure(validator_manager_, &ValidatorManagerInterface::validate_block_is_next_proof, - prev_id_, id, std::move(x.proof_), std::move(P)); - } - })); + auto P = td::PromiseCreator::lambda([SelfId = actor_id(this)](td::Result R) { + if (R.is_error()) { + td::actor::send_closure(SelfId, &DownloadBlockNew::abort_query, R.move_as_error_prefix("received bad proof: ")); + } else { + td::actor::send_closure(SelfId, &DownloadBlockNew::checked_block_proof); + } + }); + if (block_id_.is_valid()) { + if (is_link) { + td::actor::send_closure(validator_manager_, &ValidatorManagerInterface::validate_block_proof_link, block_id_, + std::move(proof), std::move(P)); + } else { + td::actor::send_closure(validator_manager_, &ValidatorManagerInterface::validate_block_proof, block_id_, + std::move(proof), std::move(P)); + } + } else { + CHECK(!is_link); + td::actor::send_closure(validator_manager_, &ValidatorManagerInterface::validate_block_is_next_proof, prev_id_, id, + std::move(proof), std::move(P)); + } } void DownloadBlockNew::got_data_from_db(td::BufferSlice data) { diff --git a/validator/state-serializer.cpp b/validator/state-serializer.cpp index 4ac4fc70..8e1b1b57 100644 --- a/validator/state-serializer.cpp +++ b/validator/state-serializer.cpp @@ -287,7 +287,8 @@ bool AsyncStateSerializer::need_serialize(BlockHandle handle) { if (handle->id().id.seqno == 0 || !handle->is_key_block()) { return false; } - return ValidatorManager::is_persistent_state(handle->unix_time(), last_key_block_ts_); + return ValidatorManager::is_persistent_state(handle->unix_time(), last_key_block_ts_) && + ValidatorManager::persistent_state_ttl(handle->unix_time()) > (UnixTime)td::Clocks::system(); } } // namespace validator diff --git a/validator/validator-group.cpp b/validator/validator-group.cpp index 73ca22a6..05628ef5 100644 --- a/validator/validator-group.cpp +++ b/validator/validator-group.cpp @@ -21,6 +21,7 @@ #include "ton/ton-io.hpp" #include "td/utils/overloaded.h" #include "common/delay.h" +#include "ton/lite-tl.hpp" namespace ton { @@ -61,7 +62,9 @@ void ValidatorGroup::generated_block_candidate(std::shared_ptrresult = R.move_as_ok(); + auto candidate = R.move_as_ok(); + add_available_block_candidate(candidate.pubkey.as_bits256(), candidate.id, candidate.collated_file_hash); + cache->result = std::move(candidate); for (auto &p : cache->promises) { p.set_value(cache->result.value().clone()); } @@ -108,6 +111,8 @@ void ValidatorGroup::validate_block_candidate(td::uint32 round_id, BlockCandidat [&](UnixTime ts) { td::actor::send_closure(SelfId, &ValidatorGroup::update_approve_cache, block_to_cache_key(block), ts); + td::actor::send_closure(SelfId, &ValidatorGroup::add_available_block_candidate, block.pubkey.as_bits256(), + block.id, block.collated_file_hash); promise.set_result(ts); }, [&](CandidateReject reject) { @@ -224,6 +229,16 @@ BlockIdExt ValidatorGroup::create_next_block_id(RootHash root_hash, FileHash fil return BlockIdExt{shard_.workchain, shard_.shard, seqno + 1, root_hash, file_hash}; } +BlockId ValidatorGroup::create_next_block_id_simple() const { + BlockSeqno seqno = 0; + for (auto &p : prev_block_ids_) { + if (seqno < p.id.seqno) { + seqno = p.id.seqno; + } + } + return BlockId{shard_.workchain, shard_.shard, seqno + 1}; +} + std::unique_ptr ValidatorGroup::make_validator_session_callback() { class Callback : public validatorsession::ValidatorSession::Callback { public: @@ -377,6 +392,47 @@ void ValidatorGroup::destroy() { stop(); } +void ValidatorGroup::get_validator_group_info_for_litequery( + td::Promise> promise) { + if (session_.empty()) { + promise.set_error(td::Status::Error(ErrorCode::notready, "not started")); + return; + } + td::actor::send_closure( + session_, &validatorsession::ValidatorSession::get_validator_group_info_for_litequery, last_known_round_id_, + [SelfId = actor_id(this), promise = std::move(promise), round = last_known_round_id_]( + td::Result>> R) mutable { + TRY_RESULT_PROMISE(promise, result, std::move(R)); + td::actor::send_closure(SelfId, &ValidatorGroup::get_validator_group_info_for_litequery_cont, round, + std::move(result), std::move(promise)); + }); +} + +void ValidatorGroup::get_validator_group_info_for_litequery_cont( + td::uint32 expected_round, std::vector> candidates, + td::Promise> promise) { + if (expected_round != last_known_round_id_) { + candidates.clear(); + } + + BlockId next_block_id = create_next_block_id_simple(); + for (auto &candidate : candidates) { + BlockIdExt id{next_block_id, candidate->id_->block_id_->root_hash_, candidate->id_->block_id_->file_hash_}; + candidate->id_->block_id_ = create_tl_lite_block_id(id); + candidate->available_ = + available_block_candidates_.count({candidate->id_->creator_, id, candidate->id_->collated_data_hash_}); + } + + auto result = create_tl_object(); + result->next_block_id_ = create_tl_lite_block_id_simple(next_block_id); + for (const BlockIdExt& prev : prev_block_ids_) { + result->prev_.push_back(create_tl_lite_block_id(prev)); + } + result->cc_seqno_ = validator_set_->get_catchain_seqno(); + result->candidates_ = std::move(candidates); + promise.set_result(std::move(result)); +} + } // namespace validator } // namespace ton diff --git a/validator/validator-group.hpp b/validator/validator-group.hpp index 26818d43..2dbff8de 100644 --- a/validator/validator-group.hpp +++ b/validator/validator-group.hpp @@ -47,6 +47,7 @@ class ValidatorGroup : public td::actor::Actor { void get_approved_candidate(PublicKey source, RootHash root_hash, FileHash file_hash, FileHash collated_data_file_hash, td::Promise promise); BlockIdExt create_next_block_id(RootHash root_hash, FileHash file_hash) const; + BlockId create_next_block_id_simple() const; void start(std::vector prev, BlockIdExt min_masterchain_block_id, UnixTime min_ts); void create_session(); @@ -58,6 +59,9 @@ class ValidatorGroup : public td::actor::Actor { } } + void get_validator_group_info_for_litequery( + td::Promise> promise); + ValidatorGroup(ShardIdFull shard, PublicKeyHash local_id, ValidatorSessionId session_id, td::Ref validator_set, validatorsession::ValidatorSessionOptions config, td::actor::ActorId keyring, td::actor::ActorId adnl, @@ -135,6 +139,17 @@ class ValidatorGroup : public td::actor::Actor { static CacheKey block_to_cache_key(const BlockCandidate& block) { return std::make_tuple(block.pubkey.as_bits256(), block.id, sha256_bits256(block.data), block.collated_file_hash); } + + void get_validator_group_info_for_litequery_cont( + td::uint32 expected_round, + std::vector> candidates, + td::Promise> promise); + + std::set> available_block_candidates_; // source, id, collated hash + + void add_available_block_candidate(td::Bits256 source, BlockIdExt id, FileHash collated_data_hash) { + available_block_candidates_.emplace(source, id, collated_data_hash); + } }; } // namespace validator diff --git a/validator/validator-options.hpp b/validator/validator-options.hpp index 41019fa6..1b7c5b09 100644 --- a/validator/validator-options.hpp +++ b/validator/validator-options.hpp @@ -123,6 +123,12 @@ struct ValidatorManagerOptionsImpl : public ValidatorManagerOptions { double get_archive_preload_period() const override { return archive_preload_period_; } + bool get_disable_rocksdb_stats() const override { + return disable_rocksdb_stats_; + } + bool nonfinal_ls_queries_enabled() const override { + return nonfinal_ls_queries_enabled_; + } void set_zero_block_id(BlockIdExt block_id) override { zero_block_id_ = block_id; @@ -185,6 +191,12 @@ struct ValidatorManagerOptionsImpl : public ValidatorManagerOptions { void set_archive_preload_period(double value) override { archive_preload_period_ = value; } + void set_disable_rocksdb_stats(bool value) override { + disable_rocksdb_stats_ = value; + } + void set_nonfinal_ls_queries_enabled(bool value) override { + nonfinal_ls_queries_enabled_ = value; + } ValidatorManagerOptionsImpl *make_copy() const override { return new ValidatorManagerOptionsImpl(*this); @@ -230,6 +242,8 @@ struct ValidatorManagerOptionsImpl : public ValidatorManagerOptions { td::uint32 celldb_compress_depth_{0}; size_t max_open_archive_files_ = 0; double archive_preload_period_ = 0.0; + bool disable_rocksdb_stats_; + bool nonfinal_ls_queries_enabled_ = false; }; } // namespace validator diff --git a/validator/validator.h b/validator/validator.h index f2dc56c4..9ede0711 100644 --- a/validator/validator.h +++ b/validator/validator.h @@ -84,6 +84,8 @@ struct ValidatorManagerOptions : public td::CntObject { virtual td::uint32 get_celldb_compress_depth() const = 0; virtual size_t get_max_open_archive_files() const = 0; virtual double get_archive_preload_period() const = 0; + virtual bool get_disable_rocksdb_stats() const = 0; + virtual bool nonfinal_ls_queries_enabled() const = 0; virtual void set_zero_block_id(BlockIdExt block_id) = 0; virtual void set_init_block_id(BlockIdExt block_id) = 0; @@ -106,6 +108,8 @@ struct ValidatorManagerOptions : public td::CntObject { virtual void set_celldb_compress_depth(td::uint32 value) = 0; virtual void set_max_open_archive_files(size_t value) = 0; virtual void set_archive_preload_period(double value) = 0; + virtual void set_disable_rocksdb_stats(bool value) = 0; + virtual void set_nonfinal_ls_queries_enabled(bool value) = 0; static td::Ref create( BlockIdExt zero_block_id, BlockIdExt init_block_id, @@ -194,7 +198,7 @@ class ValidatorManagerInterface : public td::actor::Actor { virtual void get_next_block(BlockIdExt block_id, td::Promise promise) = 0; virtual void write_handle(BlockHandle handle, td::Promise promise) = 0; - virtual void new_external_message(td::BufferSlice data) = 0; + virtual void new_external_message(td::BufferSlice data, int priority) = 0; virtual void check_external_message(td::BufferSlice data, td::Promise> promise) = 0; virtual void new_ihr_message(td::BufferSlice data) = 0; virtual void new_shard_block(BlockIdExt block_id, CatchainSeqno cc_seqno, td::BufferSlice data) = 0;