minor simplification for dependency generation

also : fix zstd-nomt exclusion and test
define LIB_SRCDIR and LIB_BINDIR
2023-09-12 13:46:03 -07:00 · 2023-09-12 13:46:03 -07:00 · 2023-09-12 13:46:03 -07:00 · 2023-09-12 13:46:03 -07:00 · 2023-09-12 13:46:03 -07:00 · 2023-09-12 13:43:59 -07:00
201 changed files with 8003 additions and 4804 deletions
@@ -1,4 +1,4 @@
-FROM circleci/buildpack-deps:focal
+FROM circleci/buildpack-deps@sha256:f6f10c11b7b8ccfd4f4a5b830c3256803604ce61292b60cb22e26b12f62b0e8c

 RUN sudo dpkg --add-architecture i386
 RUN sudo apt-get -y -qq update
@@ -2,8 +2,8 @@ task:
  name: FreeBSD (shortest)
  freebsd_instance:
    matrix:
-      image_family: freebsd-13-0
-      image_family: freebsd-12-2
+      image_family: freebsd-13-2
+      image_family: freebsd-12-4
  install_script: pkg install -y gmake coreutils
  script: |
    MOREFLAGS="-Werror" gmake -j all
@@ -9,11 +9,13 @@ on:
  pull_request:
    branches: [ dev, release, actionsTest ]

+permissions: read-all
+
 jobs:
  make-all:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: make all
      run: make all

@@ -24,7 +26,7 @@ jobs:
      DEVNULLRIGHTS: 1
      READFROMBLOCKDEVICE: 1
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: make test
      run: make test

@@ -32,28 +34,42 @@ jobs:
  make-test-osx:
    runs-on: macos-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: OS-X test
      run: make test # make -c lib all doesn't work because of the fact that it's not a tty

+  # lasts ~24mn
+  make-test-32bit:
+    runs-on: ubuntu-latest
+    env:
+      DEVNULLRIGHTS: 1
+      READFROMBLOCKDEVICE: 1
+    steps:
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
+    - name: make test
+      run: |
+        sudo apt-get -qqq update
+        make libc6install
+        CFLAGS="-m32" make test
+
  no-intrinsics-fuzztest:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: no intrinsics fuzztest
      run: MOREFLAGS="-DZSTD_NO_INTRINSICS" make -C tests fuzztest

  tsan-zstreamtest:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: thread sanitizer zstreamtest
      run: CC=clang ZSTREAM_TESTTIME=-T3mn make tsan-test-zstream

  ubsan-zstreamtest:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: undefined behavior sanitizer zstreamtest
      run: CC=clang make uasan-test-zstream

@@ -61,15 +77,26 @@ jobs:
  tsan-fuzztest:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: thread sanitizer fuzztest
      run: CC=clang make tsan-fuzztest

+
+  big-tests-zstreamtest32:
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
+    - name: zstream tests in 32bit mode, with big tests
+      run: |
+        sudo apt-get -qqq update
+        make libc6install
+        CC=clang make -C tests test-zstream32 FUZZER_FLAGS="--big-tests"
+
  # lasts ~23mn
  gcc-8-asan-ubsan-testzstd:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: gcc-8 + ASan + UBSan + Test Zstd
      # See https://askubuntu.com/a/1428822
      run: |
@@ -81,14 +108,14 @@ jobs:
  clang-asan-ubsan-testzstd:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: clang + ASan + UBSan + Test Zstd
      run: CC=clang make -j uasan-test-zstd </dev/null V=1

  gcc-asan-ubsan-testzstd-32bit:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: ASan + UBSan + Test Zstd, 32bit mode
      run: |
        sudo apt-get -qqq update
@@ -102,7 +129,7 @@ jobs:
  gcc-8-asan-ubsan-fuzz:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: gcc-8 + ASan + UBSan + Fuzz Test
      # See https://askubuntu.com/a/1428822
      run: |
@@ -114,14 +141,14 @@ jobs:
  clang-asan-ubsan-fuzz:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: clang + ASan + UBSan + Fuzz Test
      run: CC=clang FUZZER_FLAGS="--long-tests" make clean uasan-fuzztest

  gcc-asan-ubsan-fuzz32:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: ASan + UBSan + Fuzz Test 32bit
      run: |
        sudo apt-get -qqq update
@@ -131,7 +158,7 @@ jobs:
  clang-asan-ubsan-fuzz32:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: clang + ASan + UBSan + Fuzz Test 32bit
      run: |
        sudo apt-get -qqq update
@@ -141,28 +168,38 @@ jobs:
  asan-ubsan-regression:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: ASan + UBSan + Regression Test
      run: make -j uasanregressiontest

  clang-ubsan-regression:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: clang + ASan + UBSan + Regression Test
      run: CC=clang make -j uasanregressiontest

  msan-regression:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: MSan + Regression Test
      run: make -j msanregressiontest

+  clang-msan-fuzz-unoptimized:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
+      - name: clang + MSan + Fuzz Test
+        run: |
+          sudo apt-get -qqq update
+          sudo apt-get install clang
+          CC=clang MOREFLAGS="-O0" make clean msan-fuzztest
+
  clang-msan-fuzz:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: clang + MSan + Fuzz Test
      run: |
        sudo apt-get -qqq update
@@ -173,7 +210,7 @@ jobs:
  clang-msan-testzstd:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: clang + MSan + Test Zstd
      run: |
        sudo apt-get update
@@ -183,7 +220,7 @@ jobs:
  armfuzz:
      runs-on: ubuntu-latest
      steps:
-      - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+      - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
      - name: Qemu ARM emulation + Fuzz Test
        run: |
          sudo apt-get -qqq update
@@ -193,7 +230,7 @@ jobs:
  valgrind-fuzz-test:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: valgrind + fuzz test stack mode    # ~ 7mn
      shell: 'script -q -e -c "bash {0}"'
      run: |
@@ -204,57 +241,31 @@ jobs:
        make -C tests test-fuzzer-stackmode

  mingw-long-test:
-    runs-on: windows-2019
-    strategy:
-      fail-fast: false
-      matrix:
-        include: [
-          { compiler: clang, platform: x64, action: build, script: "MOREFLAGS='--target=x86_64-w64-mingw32 -Werror -Wconversion -Wno-sign-conversion -Wno-unused-command-line-argument -Wno-implicit-int-float-conversion' make -j allzstd V=1"},
-          { compiler: gcc, platform: x64, action: test, script: ""},
-        ]
+    runs-on: windows-latest
+    defaults:
+      run:
+        shell: msys2 {0}
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
-    - name: Mingw long test
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
+    - uses: msys2/setup-msys2@5beef6d11f48bba68b9eb503e3adc60b23c0cc36 # tag=v2
+      with:
+        msystem: MINGW64
+        install: make
+        update: true
+    # Based on https://ariya.io/2020/07/on-github-actions-with-msys2
+    - name: install mingw gcc
+      run: pacman --noconfirm -S gcc
+    - name: MINGW64 gcc fuzztest
      run: |
-        $env:PATH_ORIGINAL = $env:PATH
-        $env:PATH_MINGW32 = "C:\msys64\mingw32\bin"
-        $env:PATH_MINGW64 = "C:\msys64\mingw64\bin"
-        COPY C:\msys64\usr\bin\make.exe C:\msys64\mingw32\bin\make.exe
-        COPY C:\msys64\usr\bin\make.exe C:\msys64\mingw64\bin\make.exe
-        IF ("${{matrix.platform}}" -eq "x64")
-        {
-            $env:PATH = $env:PATH_MINGW64 + ";" + $env:PATH_ORIGINAL
-        }
-        ELSEIF ("${{matrix.platform}}" -eq "x86")
-        {
-            $env:PATH = $env:PATH_MINGW32 + ";" + $env:PATH_ORIGINAL
-        }
-        IF ("${{matrix.action}}" -eq "build")
-        {
-          make -v
-          sh -c "${{matrix.compiler}} -v"
-          ECHO "Building zlib to static link"
-          $env:CC = "${{matrix.compiler}}"
-          sh -c "cd .. && git clone --depth 1 --branch v1.2.11 https://github.com/madler/zlib"
-          sh -c "cd ../zlib && make -f win32/Makefile.gcc libz.a"
-          ECHO "Building zstd"
-          $env:CPPFLAGS = "-I../../zlib"
-          $env:LDFLAGS = "../../zlib/libz.a"
-          sh -c "${{matrix.script}}"
-        }
-        ELSEIF ("${{matrix.action}}" -eq "test")
-        {
-            ECHO "Testing ${{matrix.compiler}} ${{matrix.platform}}"
-            $env:CC = "gcc"
-            $env:CXX = "g++"
-            MKDIR build\cmake\build
-            CD build\cmake\build
-            $env:FUZZERTEST = "-T2mn"
-            $env:ZSTREAM_TESTTIME = "-T2mn"
-            cmake -G "Visual Studio 14 2015 Win64" ..
-            cd ..\..\..
-            make clean
-        }
+        export CC="gcc"
+        export CXX="g++"
+        export FUZZERTEST="-T2mn"
+        export ZSTREAM_TESTTIME="-T2mn"
+        echo "Testing $CC $CXX MINGW64"
+        make -v
+        $CC --version
+        $CXX --version
+        make -C tests fuzztest

  # lasts ~20mn
  oss-fuzz:
@@ -279,7 +290,7 @@ jobs:
        dry-run: false
        sanitizer: ${{ matrix.sanitizer }}
    - name: Upload Crash
-      uses: actions/upload-artifact@83fd05a356d7e2593de66fc9913b3002723633cb # tag=v3.1.1
+      uses: actions/upload-artifact@a8a3f3ad30e3422c9c7b888a15615d19a852ae32 # tag=v3.1.3
      if: failure() && steps.build.outcome == 'success'
      with:
        name: ${{ matrix.sanitizer }}-artifacts
@@ -10,25 +10,27 @@ on:
  pull_request:
    branches: [ dev, release, actionsTest ]

+permissions: read-all
+
 jobs:
  linux-kernel:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: linux kernel, library + build + test
      run: make -C contrib/linux-kernel test CFLAGS="-Werror -Wunused-const-variable -Wunused-but-set-variable"

  benchmarking:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: make benchmarking
      run: make benchmarking

  check-32bit: # designed to catch https://github.com/facebook/zstd/issues/2428
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: make check on 32-bit
      run: |
        sudo apt update
@@ -38,7 +40,7 @@ jobs:
  check-x32:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: make check on x32 ABI # https://en.wikipedia.org/wiki/X32_ABI
      env:
        CHECK_CONSTRAINED_MEM: true
@@ -47,10 +49,27 @@ jobs:
        APT_PACKAGES="gcc-multilib" make apt-install
        CFLAGS="-mx32 -O1 -fstack-protector" make check V=1

+  build-c89:
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
+    - name: ensure zstd can be build with c89/c90 compilers (+ long long support + variadic macros)
+      run: |
+        make c89build V=1
+
+  build-zstd-dll:
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
+    - name: build zstd bin against a dynamic lib (debuglevel for more dependencies)
+      run: |
+        make -C lib lib-mt-release 
+        DEBUGLEVEL=2 make -C programs zstd-dll
+
  gcc-7-libzstd:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: gcc-7 + libzstdmt compilation
      # See https://askubuntu.com/a/1428822
      run: |
@@ -67,7 +86,7 @@ jobs:
  cmake-build-and-test-check:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: cmake build and test check
      run: |
        FUZZERTEST=-T1mn ZSTREAM_TESTTIME=-T1mn make cmakebuild
@@ -78,7 +97,7 @@ jobs:
  cpp-gnu90-c99-compatibility:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: C++, gnu90 and c99 compatibility
      run: |
        make cxxtest
@@ -92,7 +111,7 @@ jobs:
  mingw-cross-compilation:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: mingw cross-compilation
      run: |
        # sudo update-alternatives --set x86_64-w64-mingw32-g++ /usr/bin/x86_64-w64-mingw32-g++-posix; (doesn't work)
@@ -103,7 +122,7 @@ jobs:
  armbuild:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: ARM Build Test
      run: |
        sudo apt-get -qqq update
@@ -113,7 +132,7 @@ jobs:
  bourne-shell:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: Bourne shell compatibility (shellcheck)
      run: |
        wget https://github.com/koalaman/shellcheck/releases/download/v0.7.1/shellcheck-v0.7.1.linux.x86_64.tar.xz
@@ -123,7 +142,7 @@ jobs:
  zlib-wrapper:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: zlib wrapper test
      run: |
        sudo apt-get -qqq update
@@ -134,7 +153,7 @@ jobs:
  lz4-threadpool-libs:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: LZ4, thread pool, and libs build testslib wrapper test
      run: |
        make lz4install
@@ -148,7 +167,7 @@ jobs:
  gcc-make-tests-32bit:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: Make all, 32bit mode
      run: |
        sudo apt-get -qqq update
@@ -158,7 +177,7 @@ jobs:
  gcc-8-make:
    runs-on: ubuntu-latest
    steps:
-      - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+      - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
      - name: gcc-8 build
        # See https://askubuntu.com/a/1428822
        run: |
@@ -167,10 +186,29 @@ jobs:
          make gcc8install
          CC=gcc-8 CFLAGS="-Werror" make -j all

+  make-external-compressors:
+    strategy:
+      matrix:
+        include:
+          - name: "no external compressors"
+            flags: "HAVE_ZLIB=0 HAVE_LZ4=0 HAVE_LZMA=0"
+          - name: "only zlib"
+            flags: "HAVE_ZLIB=1 HAVE_LZ4=0 HAVE_LZMA=0"
+          - name: "only lz4"
+            flags: "HAVE_ZLIB=0 HAVE_LZ4=1 HAVE_LZMA=0"
+          - name: "only lzma"
+            flags: "HAVE_ZLIB=0 HAVE_LZ4=0 HAVE_LZMA=1"
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
+      - name: Build with ${{matrix.name}}
+        run: ${{matrix.flags}} make zstd
+
+
  implicit-fall-through:
    runs-on: ubuntu-latest
    steps:
-      - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+      - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
      - name: -Wimplicit-fallthrough build
        run: |
          make clean
@@ -181,7 +219,7 @@ jobs:
  meson-linux:
    runs-on: ubuntu-latest
    steps:
-      - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+      - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
      - name: Install packages
        run: |
          sudo apt-get update
@@ -205,11 +243,11 @@ jobs:
  meson-windows:
    runs-on: windows-latest
    steps:
-      - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+      - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
      - name: Install packages
        run: pip install --pre meson
      - name: Initialize the MSVC dev command prompt
-        uses: ilammy/msvc-dev-cmd@7315a94840631165970262a99c72cfb48a65d25d
+        uses: ilammy/msvc-dev-cmd@cec98b9d092141f74527d0afa6feb2af698cfe89
      - name: Configure with Meson
        run: |
          meson setup build/meson/ builddir -Dbin_tests=true -Dbin_programs=true -Dbin_contrib=true
@@ -223,20 +261,22 @@ jobs:
        run: |
          meson install -C builddir --destdir staging/

-  cmake-visual-2019:
-    runs-on: windows-2019
+  cmake-visual-2022:
    strategy:
      matrix:
        include:
-          - generator: "Visual Studio 16 2019"
+          - generator: "Visual Studio 17 2022"
            flags: "-A x64"
-          - generator: "Visual Studio 16 2019"
+          - generator: "Visual Studio 17 2022"
            flags: "-A Win32"
          - generator: "MinGW Makefiles"
+          - generator: "Visual Studio 17 2022"
+            flags: "-T ClangCL"
+    runs-on: windows-2022
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: Add MSBuild to PATH
-      uses: microsoft/setup-msbuild@34cfbaee7f672c76950673338facd8a73f637506 # tag=v1.1.3
+      uses: microsoft/setup-msbuild@1ff57057b5cfdc39105cd07a01d78e9b0ea0c14c # tag=v1.3
    - name: Build
      working-directory: ${{env.GITHUB_WORKSPACE}}
      run: |
@@ -246,48 +286,47 @@ jobs:
        cmake.exe -G "${{matrix.generator}}" ${{matrix.flags}} ..
        cmake.exe --build .

-  visual-2019:
-    runs-on: windows-2019
+  msbuild-visual-studio:
    strategy:
+      fail-fast: false  # 'false' means Don't stop matrix workflows even if some matrix failed.
      matrix:
-        platform: [x64, Win32]
-        configuration: [Debug, Release]
+        include: [
+          { name: "VS 2022 x64 Debug", platform: x64, configuration: Debug, toolset: v143, runner: "windows-2022", arch: "" },
+          { name: "VS 2022 Win32 Debug", platform: Win32, configuration: Debug, toolset: v143, runner: "windows-2022", arch: "" },
+          { name: "VS 2022 x64 Release", platform: x64, configuration: Release, toolset: v143, runner: "windows-2022", arch: ""},
+          { name: "VS 2022 Win32 Release", platform: Win32, configuration: Release, toolset: v143, runner: "windows-2022", arch: ""},
+          { name: "VS 2019 x64 Release", platform: Win32, configuration: Release, toolset: v142, runner: "windows-2019", arch: ""},
+          { name: "VS 2019 Win32 Release", platform: x64, configuration: Release, toolset: v142, runner: "windows-2019", arch: ""},
+          { name: "VS 2022 x64 Release AVX2", platform: x64, configuration: Release, toolset: v143, runner: "windows-2022", arch: "AdvancedVectorExtensions2" },
+        ]
+    runs-on: ${{matrix.runner}}
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: Add MSBuild to PATH
-      uses: microsoft/setup-msbuild@34cfbaee7f672c76950673338facd8a73f637506 # tag=v1.1.3
-    - name: Build
+      uses: microsoft/setup-msbuild@1ff57057b5cfdc39105cd07a01d78e9b0ea0c14c # tag=v1.3
+    - name: Build ${{matrix.name}}
      working-directory: ${{env.GITHUB_WORKSPACE}}
      # See https://docs.microsoft.com/visualstudio/msbuild/msbuild-command-line-reference
+      if: ${{ matrix.arch == '' }}
      run: >
-        msbuild "build\VS2010\zstd.sln" /m /verbosity:minimal /property:PlatformToolset=v142
-        /t:Clean,Build /p:Platform=${{matrix.platform}} /p:Configuration=${{matrix.configuration}}
+        msbuild "build\VS2010\zstd.sln" /m /verbosity:minimal /property:PlatformToolset=${{matrix.toolset}}
+        /t:Clean,Build /p:Platform=${{matrix.platform}} /p:Configuration=${{matrix.configuration}} /warnaserror
+    - name: Build ${{matrix.name}}
+      working-directory: ${{env.GITHUB_WORKSPACE}}
+      # See https://docs.microsoft.com/visualstudio/msbuild/msbuild-command-line-reference
+      if: ${{ matrix.arch != '' }}
+      run: >
+        msbuild "build\VS2010\zstd.sln" /m /verbosity:minimal /property:PlatformToolset=${{matrix.toolset}}
+        /t:Clean,Build /p:Platform=${{matrix.platform}} /p:Configuration=${{matrix.configuration}} /warnaserror
+        /p:InstructionSet=${{matrix.arch}}

-# TODO: fix as part of https://github.com/facebook/zstd/issues/3064
-#  visual-2015:
-#    # only GH actions windows-2016 contains VS 2015
-#    runs-on: windows-2016
-#    strategy:
-#      matrix:
-#        platform: [x64, Win32]
-#        configuration: [Debug, Release]
-#    steps:
-#    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
-#    - name: Add MSBuild to PATH
-#      uses: microsoft/setup-msbuild@34cfbaee7f672c76950673338facd8a73f637506 # tag=v1.1.3
-#    - name: Build
-#      working-directory: ${{env.GITHUB_WORKSPACE}}
-#      run: >
-#        msbuild "build\VS2010\zstd.sln" /m /verbosity:minimal /property:PlatformToolset=v140
-#        /t:Clean,Build /p:Platform=${{matrix.platform}} /p:Configuration=${{matrix.configuration}}
-
-  # This tests that we don't accidently grow the size too much.
+  # This tests that we don't accidentally grow the size too much.
  # If the size grows intentionally, you can raise these numbers.
  # But we do need to think about binary size, since it is a concern.
  libzstd-size:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: libzstd size test
      run: |
        make clean && make -j -C lib libzstd && ./tests/check_size.py lib/libzstd.so 1100000
@@ -298,7 +337,7 @@ jobs:
  minimal-decompressor-macros:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: minimal decompressor macros
      run: |
        make clean && make -j all ZSTD_LIB_MINIFY=1 MOREFLAGS="-Werror"
@@ -309,11 +348,13 @@ jobs:
        make clean && make check MOREFLAGS="-Werror -DHUF_FORCE_DECOMPRESS_X2 -DZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG"
        make clean && make -j all MOREFLAGS="-Werror -DZSTD_NO_INLINE -DZSTD_STRIP_ERROR_STRINGS"
        make clean && make check MOREFLAGS="-Werror -DZSTD_NO_INLINE -DZSTD_STRIP_ERROR_STRINGS"
+        make clean && make check ZSTD_LIB_EXCLUDE_COMPRESSORS_DFAST_AND_UP=1 MOREFLAGS="-Werror"
+        make clean && make check ZSTD_LIB_EXCLUDE_COMPRESSORS_GREEDY_AND_UP=1 MOREFLAGS="-Werror"

  dynamic-bmi2:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: dynamic bmi2 tests
      run: |
        make clean && make -j check MOREFLAGS="-O0 -Werror -mbmi2"
@@ -325,13 +366,12 @@ jobs:
  test-variants:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: make all variants & validate
      run: |
        make -j -C programs allVariants MOREFLAGS=-O0
        ./tests/test-variants.sh

-
  qemu-consistency:
    name: QEMU ${{ matrix.name }}
    runs-on: ubuntu-20.04
@@ -351,7 +391,7 @@ jobs:
      XCC: ${{ matrix.xcc }}
      XEMU: ${{ matrix.xemu }}
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: apt update & install
      run: |
        sudo apt-get update
@@ -394,52 +434,58 @@ jobs:
        LDFLAGS="-static" CC=$XCC QEMU_SYS=$XEMU make clean check

  mingw-short-test:
-    runs-on: windows-2019
+    runs-on: windows-latest
    strategy:
-      fail-fast: false
+      fail-fast: false  # 'false' means Don't stop matrix workflows even if some matrix failed.
      matrix:
        include: [
-          { compiler: gcc, platform: x64, script: "CFLAGS=-Werror make -j allzstd DEBUGLEVEL=2"},
-          { compiler: gcc, platform: x86, script: "CFLAGS=-Werror make -j allzstd"},
-          { compiler: clang, platform: x64, script: "CFLAGS='--target=x86_64-w64-mingw32 -Werror -Wconversion -Wno-sign-conversion' make -j allzstd V=1"},
+          { compiler: gcc, msystem: MINGW32, cflags: "-Werror"},
+          { compiler: gcc, msystem: MINGW64, cflags: "-Werror"},
+          { compiler: clang, msystem: MINGW64, cflags: "--target=x86_64-w64-mingw32 -Werror -Wconversion -Wno-sign-conversion -Wno-unused-command-line-argument"},
        ]
+    defaults:
+      run:
+        shell: msys2 {0}
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
-    - name: Mingw short test
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
+    - uses: msys2/setup-msys2@5beef6d11f48bba68b9eb503e3adc60b23c0cc36 # tag=v2
+      with:
+        msystem: ${{ matrix.msystem }}
+        install: make diffutils
+        update: true
+    # Based on https://ariya.io/2020/07/on-github-actions-with-msys2
+    - name: install mingw gcc i686
+      if: ${{ (matrix.msystem == 'MINGW32') && (matrix.compiler == 'gcc') }}
+      run: pacman --noconfirm -S mingw-w64-i686-gcc
+    - name: install mingw gcc x86_64
+      if: ${{ (matrix.msystem == 'MINGW64') && (matrix.compiler == 'gcc') }}
+      run: pacman --noconfirm -S mingw-w64-x86_64-gcc
+    - name: install mingw clang i686
+      if: ${{ (matrix.msystem == 'MINGW32') && (matrix.compiler == 'clang') }}
+      run: pacman --noconfirm -S mingw-w64-i686-clang
+    - name: install mingw clang x86_64
+      if: ${{ (matrix.msystem == 'MINGW64') && (matrix.compiler == 'clang') }}
+      run: pacman --noconfirm -S mingw-w64-x86_64-clang
+    - name: run mingw tests
      run: |
-        ECHO "Building ${{matrix.compiler}} ${{matrix.platform}}"
-        $env:PATH_ORIGINAL = $env:PATH
-        $env:PATH_MINGW32 = "C:\msys64\mingw32\bin"
-        $env:PATH_MINGW64 = "C:\msys64\mingw64\bin"
-        COPY C:\msys64\usr\bin\make.exe C:\msys64\mingw32\bin\make.exe
-        COPY C:\msys64\usr\bin\make.exe C:\msys64\mingw64\bin\make.exe
-        IF ("${{matrix.platform}}" -eq "x64")
-        {
-            $env:PATH = $env:PATH_MINGW64 + ";" + $env:PATH_ORIGINAL
-        }
-        ELSEIF ("${{matrix.platform}}" -eq "x86")
-        {
-            $env:PATH = $env:PATH_MINGW32 + ";" + $env:PATH_ORIGINAL
-        }
        make -v
-        sh -c "${{matrix.compiler}} -v"
-        $env:CC = "${{matrix.compiler}}"
-        sh -c "${{matrix.script}}"
-        ECHO "Testing ${{matrix.compiler}} ${{matrix.platform}}"
+        export CC=${{ matrix.compiler }}
+        $CC --version
+        CFLAGS="${{ matrix.cflags }}" make -j allzstd
+        echo "Testing $CC ${{ matrix.msystem }}"
        make clean
-        make check
-
+        MSYS="" make check

  visual-runtime-tests:
-    runs-on: windows-2019
+    runs-on: windows-latest
    strategy:
      matrix:
        platform: [x64, Win32]
        configuration: [Release]
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: Add MSBuild to PATH
-      uses: microsoft/setup-msbuild@34cfbaee7f672c76950673338facd8a73f637506 # tag=v1.1.3
+      uses: microsoft/setup-msbuild@1ff57057b5cfdc39105cd07a01d78e9b0ea0c14c # tag=v1.3
    - name: Build and run tests
      working-directory: ${{env.GITHUB_WORKSPACE}}
      env:
@@ -451,13 +497,39 @@ jobs:
        COPY build\VS2010\bin\${{matrix.platform}}_${{matrix.configuration}}\*.exe tests\
        CD tests
        sh -e playTests.sh
-        DIR
        .\fuzzer.exe -T2m

+  # Following instructions at: https://github.com/marketplace/actions/install-cygwin-action
+  cygwin-tests:
+    runs-on: windows-latest
+    steps:
+    - run: git config --global core.autocrlf input
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
+    - uses: cygwin/cygwin-install-action@006ad0b0946ca6d0a3ea2d4437677fa767392401 # tag=master
+      with:
+        platform: x86_64
+        packages: >-
+          autoconf,
+          automake,
+          gcc-g++,
+          make,
+          mingw64-x86_64-gcc-g++,
+          patch,
+          perl
+    - name: cygwin tests
+      shell: C:\cygwin\bin\bash.exe --noprofile --norc -eo pipefail '{0}'
+      run: >-
+        export PATH=/usr/bin:$(cygpath ${SYSTEMROOT})/system32 &&
+        export CFLAGS="-Werror -O1" &&
+        ls &&
+        make -j allzstd &&
+        make -C tests fuzzer &&
+        ./tests/fuzzer.exe -v -T1m
+
  intel-cet-compatibility:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: Build Zstd
      run: |
        make -j zstd V=1
@@ -478,7 +550,7 @@ jobs:
    container:
      image: debian:testing
    steps:
-      - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+      - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
      - name: Install dependencies
        run: |
          apt -y update
@@ -493,7 +565,7 @@ jobs:
  versions-compatibility:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: Versions Compatibility Test
      run: |
        make -C tests versionsTest
@@ -501,11 +573,34 @@ jobs:
  clangbuild:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
    - name: make clangbuild
      run: |
        make clangbuild

+  clang-pgo:
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
+    - name: Build PGO Zstd with Clang
+      env:
+        CC: clang-14
+        LLVM_PROFDATA: llvm-profdata-14
+      run: |
+        make -C programs zstd-pgo
+        ./programs/zstd -b
+
+  gcc-pgo:
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
+    - name: Build PGO Zstd with GCC
+      env:
+        CC: gcc
+      run: |
+        make -C programs zstd-pgo
+        ./programs/zstd -b
+

 # For reference : icc tests
 # icc tests are currently failing on Github Actions, likely to issues during installation stage
@@ -525,7 +620,7 @@ jobs:
 #        sudo add-apt-repository "deb https://apt.repos.intel.com/oneapi all main"
 #        sudo apt-get update
 #        sudo apt-get install -y intel-basekit intel-hpckit
-#    - uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+#    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
 #    - name: make check
 #      run: |
 #        make CC=/opt/intel/oneapi/compiler/latest/linux/bin/intel64/icc check
@@ -5,21 +5,19 @@ on:
    types:
      - published

-permissions:
-  contents: read
+permissions: read-all

 jobs:
  publish-release-artifacts:
    permissions:
-      contents: read # to fetch code (actions/checkout)
-      actions: write # to attach binaries to release artifacts (skx/github-action-publish-binaries)
+      contents: write # to fetch code and upload artifacts

    runs-on: ubuntu-latest
    if: startsWith(github.ref, 'refs/tags/')

    steps:
      - name: Checkout
-        uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+        uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3

      - name: Archive
        env:
@@ -68,7 +66,7 @@ jobs:
          fi

      - name: Publish
-        uses: skx/github-action-publish-binaries@release-2.0
+        uses: skx/github-action-publish-binaries@b9ca5643b2f1d7371a6cba7f35333f1461bbc703 # tag=release-2.0
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
        with:
@@ -27,12 +27,12 @@ jobs:

    steps:
      - name: "Checkout code"
-        uses: actions/checkout@ac593985615ec2ede58e132d2e21d2b1cbd6127c # tag=v3
+        uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
        with:
          persist-credentials: false

      - name: "Run analysis"
-        uses: ossf/scorecard-action@937ffa90d79c7d720498178154ad4c7ba1e4ad8c # tag=v2.1.0
+        uses: ossf/scorecard-action@08b4669551908b1024bb425080c797723083c031 # tag=v2.2.0
        with:
          results_file: results.sarif
          results_format: sarif
@@ -51,7 +51,7 @@ jobs:
      # Upload the results as artifacts (optional). Commenting out will disable uploads of run results in SARIF
      # format to the repository Actions tab.
      - name: "Upload artifact"
-        uses: actions/upload-artifact@83fd05a356d7e2593de66fc9913b3002723633cb # tag=v3.1.1
+        uses: actions/upload-artifact@a8a3f3ad30e3422c9c7b888a15615d19a852ae32 # tag=v3.1.3
        with:
          name: SARIF file
          path: results.sarif
@@ -59,6 +59,6 @@ jobs:

      # Upload the results to GitHub's code scanning dashboard.
      - name: "Upload to code-scanning"
-        uses: github/codeql-action/upload-sarif@959cbb7472c4d4ad70cdfe6f4976053fe48ab394 # tag=v2.1.37
+        uses: github/codeql-action/upload-sarif@a09933a12a80f87b87005513f0abb1494c27a716 # tag=v2.21.4
        with:
          sarif_file: results.sarif
@@ -0,0 +1,58 @@
+name: windows-artifacts
+
+on:
+  push:
+    branches: [ test_artifacts, win_artifacts ]
+  release:
+    types:
+      - published
+
+permissions: read-all
+
+jobs:
+  windows-artifacts:
+    # see https://ariya.io/2020/07/on-github-actions-with-msys2
+    runs-on: windows-latest
+    # see https://github.com/msys2/setup-msys2
+    strategy:
+      matrix:
+        include:
+          - { msystem: mingw64, env: x86_64, ziparch: win64 }
+          - { msystem: mingw32, env: i686, ziparch: win32 }
+    defaults:
+      run:
+        shell: msys2 {0}
+    steps:
+    - uses: actions/checkout@c85c95e3d7251135ab7dc9ce3241c5835cc595a9 # tag=v3
+    - uses: msys2/setup-msys2@5beef6d11f48bba68b9eb503e3adc60b23c0cc36 # tag=v2
+      with:
+        msystem: ${{ matrix.msystem }}
+        install: make zlib git p7zip mingw-w64-${{matrix.env}}-gcc
+        update: true
+
+    - name: display versions
+      run: |
+        make -v
+        cc -v
+
+    - name: Building zlib to static link
+      run: |
+        git clone --depth 1 --branch v1.2.11 https://github.com/madler/zlib
+        make -C zlib -f win32/Makefile.gcc libz.a
+
+    - name: Building zstd programs
+      run: |
+        CPPFLAGS=-I../zlib LDFLAGS=../zlib/libz.a make -j allzstd MOREFLAGS=-static V=1
+
+    - name: Create artifacts
+      run: |
+        ./lib/dll/example/build_package.bat
+        mv bin/ zstd-${{ github.ref_name }}-${{matrix.ziparch}}/
+        7z a -tzip -mx9 zstd-${{ github.ref_name }}-${{matrix.ziparch}}.zip zstd-${{ github.ref_name }}-${{matrix.ziparch}}/
+        cd ..
+
+    - name: Publish zstd-$VERSION-${{matrix.ziparch}}.zip
+      uses: actions/upload-artifact@a8a3f3ad30e3422c9c7b888a15615d19a852ae32 # tag=v3
+      with:
+        path: ${{ github.workspace }}/zstd-${{ github.ref_name }}-${{matrix.ziparch}}.zip
+        name: zstd-${{ github.ref_name }}-${{matrix.ziparch}}.zip
@@ -42,6 +42,7 @@ build-*
 _codelite/
 _zstdbench/
 .clang_complete
+compile_flags.txt
 *.idea
 *.swp
 .DS_Store
@@ -1,128 +0,0 @@
-# Travis CI is used to test platforms that github-actions currently doesn't support
-# without either self-hosting or some finnicky work-around. Also, some tests
-# are troublesome to migrate since GH Actions runs tests not in a tty.
-language: c
-
-git:
-  depth: 1
-
-branches:
-  only:
-  - dev
-  - release
-  - master
-  - travisTest
-
-addons:
-  apt:
-    update: true
-
-env:
-  global:
-    - FUZZERTEST=-T1mn
-      ZSTREAM_TESTTIME=-T1mn
-      DECODECORPUS_TESTTIME=-T1mn
-
-matrix:
-  fast_finish: true
-  include:
-    - name: S390X (big endian) + Fuzz test
-      dist: trusty
-      arch: s390x
-      script:
-        - FUZZER_FLAGS=--no-big-tests make -C tests fuzztest
-
-    - name: S390X (big endian) + Fuzz test + no intrinsics
-      dist: trusty
-      arch: s390x
-      script:
-        - MOREFLAGS="-DZSTD_NO_INTRINSICS" FUZZER_FLAGS=--no-big-tests make -C tests fuzztest
-
-    - name: arm64    # ~2.5 mn
-      os: linux
-      arch: arm64
-      script:
-        - make check
-
-    - name: arm64fuzz
-      os: linux
-      arch: arm64
-      script:
-        - make -C tests fuzztest
-
-    # TODO: migrate to GH Actions once newest clang staticanalyze warnings are fixed
-    - name: static analyzer scanbuild    # ~8mn
-      dist: trusty  # note : it's important to pin down a version of static analyzer, since different versions report different false positives
-      script:
-        - make staticAnalyze
-
-    # GH actions can't run this command on OS-X, non-tty issues
-    - name: OS-X make all lib
-      os: osx
-      script:
-        - make -C lib all
-
-    # Introduced to check compat with old toolchains, to prevent e.g. #1872
-    - name: ARM Build Test (on Trusty)
-      dist: trusty
-      script:
-        - make arminstall
-        - make armbuild
-
-    # check release number (release/new tag only)
-    - name: Tag-Specific Test
-      if: tag =~ ^v[0-9]\.[0-9]
-      script:
-        - make -C tests checkTag
-        - tests/checkTag "$TRAVIS_BRANCH"
-
-    - name: PPC64LE + Fuzz test  # ~13mn
-      arch: ppc64le
-      env:
-        - FUZZER_FLAGS=--no-big-tests
-        - MOREFLAGS="-static"
-      script:
-        - cat /proc/cpuinfo
-        - make -C tests fuzztest
-
-    # This test currently fails on GA specifically, for no obvious reason
-    # (it works fine on travisCI, and on local test platforms).
-    - name: Versions Compatibility Test   # ~6mn
-      script:
-        - make -C tests versionsTest
-
-    # meson dedicated test
-    - name: Focal (Meson + clang)    # ~15mn
-      dist: focal
-      language: cpp
-      compiler: clang
-      install:
-        - sudo apt-get install -qq liblz4-dev valgrind tree
-        - |
-          travis_retry curl -o ~/ninja.zip -L 'https://github.com/ninja-build/ninja/releases/download/v1.9.0/ninja-linux.zip' &&
-          unzip ~/ninja.zip -d ~/.local/bin
-        - |
-          travis_retry curl -o ~/get-pip.py -L 'https://bootstrap.pypa.io/pip/3.6/get-pip.py' &&
-          python3 ~/get-pip.py --user &&
-          pip3 install --user meson
-      script:
-        - |
-          meson setup \
-            --buildtype=debugoptimized \
-            -Db_lundef=false \
-            -Dauto_features=enabled \
-            -Dbin_programs=true \
-            -Dbin_tests=true \
-            -Dbin_contrib=true \
-            -Ddefault_library=both \
-             build/meson builddir
-        - pushd builddir
-        - ninja
-        - meson test --verbose --no-rebuild
-        - DESTDIR=./staging ninja install
-        - tree ./staging
-      after_failure:
-        - cat "$TRAVIS_BUILD_DIR"/builddir/meson-logs/testlog.txt
-
-  allow_failures:
-    - env: ALLOW_FAILURES=true
@@ -1,37 +1,81 @@
-v1.5.3 (June, 2022)
-perf: 5-30% faster dictionary compression at levels 1-4 (#3086, #3114, #3152, @embg)
-perf: 5-10% faster streaming compression at levels 1-2 (#3114, @embg)
-perf: Remove branch in ZSTD_fast_noDict (#3129, @felixhandte)
-perf: Add option to prefetch CDict tables (#3177, @embg)
-perf: Minor compression ratio improvement (#2983, @Cyan4973)
-perf: Minor speed improvement for Huffman decoding (#3013, @WojciechMula)
-perf: Enable STATIC_BMI2 for gcc/clang (#3080, @TocarIP)
-perf: Optimize ZSTD_row_getMatchMask for ARM levels 8-10 (#3139, #3160, @danlark1)
-perf: aarch64 performance improvements (#3145, #3141, @JunHe77)
-perf: Lazy parameters adaptation (#2974, @Cyan4973)
-cli: Async write for decompression (#2975, @yoniko)
-cli: Use buffered output (#2985, @yoniko)
-cli: Change zstdless behavior to align with zless (#2909, @binhdvo)
-cli: AsyncIO compression (#3021, #3022, @yoniko)
-cli: Print zlib/lz4/lzma library versions in verbose version output (#3030, @terrelln)
-cli: Fix for -r on empty directory (#3027, @brailovich)
-cli: Fix required decompression memory usage reported by -vv + --long (#3042, @u1f35c)
+v1.5.5 (Apr 2023)
+fix: fix rare corruption bug affecting the high compression mode, reported by @danlark1 (#3517, @terrelln)
+perf: improve mid-level compression speed (#3529, #3533, #3543, @yoniko and #3552, @terrelln)
+lib: deprecated bufferless block-level API (#3534) by @terrelln
+cli: mmap large dictionaries to save memory, by @daniellerozenblit
+cli: improve speed of --patch-from mode (~+50%) (#3545) by @daniellerozenblit
+cli: improve i/o speed (~+10%) when processing lots of small files (#3479) by @felixhandte
+cli: zstd no longer crashes when requested to write into write-protected directory (#3541) by @felixhandte
+cli: fix decompression into block device using -o, reported by @georgmu (#3583)
+build: fix zstd CLI compiled with lzma support but not zlib support (#3494) by @Hello71
+build: fix cmake does no longer require 3.18 as minimum version (#3510) by @kou
+build: fix MSVC+ClangCL linking issue (#3569) by @tru
+build: fix zstd-dll, version of zstd CLI that links to the dynamic library (#3496) by @yoniko
+build: fix MSVC warnings (#3495) by @embg
+doc: updated zstd specification to clarify corner cases, by @Cyan4973
+doc: document how to create fat binaries for macos (#3568) by @rickmark
+misc: improve seekable format ingestion speed (~+100%) for very small chunk sizes (#3544) by @Cyan4973
+misc: tests/fullbench can benchmark multiple files (#3516) by @dloidolt
+
+v1.5.4 (Feb 2023)
+perf: +20% faster huffman decompression for targets that can't compile x64 assembly (#3449, @terrelln)
+perf: up to +10% faster streaming compression at levels 1-2 (#3114, @embg)
+perf: +4-13% for levels 5-12 by optimizing function generation (#3295, @terrelln)
+pref: +3-11% compression speed for `arm` target (#3199, #3164, #3145, #3141, #3138, @JunHe77 and #3139, #3160, @danlark1)
+perf: +5-30% faster dictionary compression at levels 1-4 (#3086, #3114, #3152, @embg)
+perf: +10-20% cold dict compression speed by prefetching CDict tables (#3177, @embg)
+perf: +1% faster compression by removing a branch in ZSTD_fast_noDict (#3129, @felixhandte)
+perf: Small compression ratio improvements in high compression mode (#2983, #3391, @Cyan4973 and #3285, #3302, @daniellerozenblit)
+perf: small speed improvement by better detecting `STATIC_BMI2` for `clang` (#3080, @TocarIP)
+perf: Improved streaming performance when `ZSTD_c_stableInBuffer` is set (#2974, @Cyan4973)
+cli: Asynchronous I/O for improved cli speed (#2975, #2985, #3021, #3022, @yoniko)
+cli: Change `zstdless` behavior to align with `zless` (#2909, @binhdvo)
+cli: Keep original file if `-c` or `--stdout` is given (#3052, @dirkmueller)
+cli: Keep original files when result is concatenated into a single output with `-o` (#3450, @Cyan4973)
+cli: Preserve Permissions and Ownership of regular files (#3432, @felixhandte)
+cli: Print zlib/lz4/lzma library versions with `-vv` (#3030, @terrelln)
+cli: Print checksum value for single frame files with `-lv`  (#3332, @Cyan4973)
+cli: Print `dictID` when present with `-lv` (#3184, @htnhan)
+cli: when `stderr` is *not* the console, disable status updates, but preserve final summary (#3458, @Cyan4973)
+cli: support `--best` and `--no-name` in `gzip` compatibility mode (#3059, @dirkmueller)
+cli: support for `posix` high resolution timer `clock_gettime()`, for improved benchmark accuracy (#3423, @Cyan4973)
+cli: improved help/usage (`-h`,  `-H`) formatting (#3094, @dirkmueller and #3385, @jonpalmisc)
+cli: Fix better handling of bogus numeric values (#3268, @ctkhanhly)
+cli: Fix input consists of multiple files _and_ `stdin` (#3222, @yoniko)
+cli: Fix tiny files passthrough (#3215, @cgbur)
+cli: Fix for `-r` on empty directory (#3027, @brailovich)
+cli: Fix empty string as argument for `--output-dir-*` (#3220, @embg)
+cli: Fix decompression memory usage reported by `-vv --long` (#3042, @u1f35c, and #3232, @zengyijing)
 cli: Fix infinite loop when empty input is passed to trainer (#3081, @terrelln)
-cli: Implement more gzip compatibility (#3059, @dirkmueller)
-cli: Keep original file if -c or --stdout is given (#3052, @dirkmueller)
+cli: Fix `--adapt` doesn't work when `--no-progress` is also set (#3354, @terrelln)
+api: Support for Block-Level Sequence Producer (#3333, @embg)
+api: Support for in-place decompression (#3432, @terrelln)
+api: New  `ZSTD_CCtx_setCParams()`  function, set all parameters defined in a  `ZSTD_compressionParameters`  structure (#3403, @Cyan4973)
+api: Streaming decompression detects incorrect header ID sooner (#3175, @Cyan4973)
+api: Window size resizing optimization for edge case (#3345, @daniellerozenblit)
+api: More accurate error codes for busy-loop scenarios (#3413, #3455, @Cyan4973)
+api: Fix limit overflow in `compressBound` and `decompressBound` (#3362, #3373, Cyan4973) reported by @nigeltao
+api: Deprecate several advanced experimental functions: streaming (#3408, @embg), copy (#3196, @mileshu)
+bug: Fix corruption that rarely occurs in 32-bit mode with wlog=25 (#3361, @terrelln)
 bug: Fix for block-splitter (#3033, @Cyan4973)
 bug: Fixes for Sequence Compression API (#3023, #3040, @Cyan4973)
 bug: Fix leaking thread handles on Windows (#3147, @animalize)
 bug: Fix timing issues with cmake/meson builds (#3166, #3167, #3170, @Cyan4973)
 build: Allow user to select legacy level for cmake (#3050, @shadchin)
 build: Enable legacy support by default in cmake (#3079, @niamster)
-build: Meson improvements (#3039, #3122, @eli-schwartz)
+build: Meson build script improvements (#3039, #3120, #3122, #3327, #3357, @eli-schwartz and #3276, @neheb)
 build: Add aarch64 to supported architectures for zstd_trace (#3054, @ooosssososos)
-doc: Split help in long and short version, cleanup formatting (#3094, @dirkmueller)
-doc: Updated man page, providing more details for --train mode (#3112, @Cyan4973)
+build: support AIX architecture (#3219, @qiongsiwu)
+build: Fix `ZSTD_LIB_MINIFY` build macro, which now reduces static library size by half (#3366, @terrelln)
+build: Fix Windows issues with Multithreading translation layer (#3364, #3380, @yoniko) and ARM64 target (#3320, @cwoffenden)
+build: Fix `cmake` script (#3382, #3392, @terrelln and #3252 @Tachi107 and #3167 @Cyan4973)
+doc: Updated man page, providing more details for `--train` mode (#3112, @Cyan4973)
 doc: Add decompressor errata document (#3092, @terrelln)
 misc: Enable Intel CET (#2992, #2994, @hjl-tools)
-misc: Streaming decompression can detect incorrect header ID sooner (#3175, @Cyan4973)
+misc: Fix `contrib/` seekable format (#3058, @yhoogstrate and #3346, @daniellerozenblit)
+misc: Improve speed of the one-file library generator (#3241, @wahern and #3005, @cwoffenden)
+
+v1.5.3 (dev version, unpublished)

 v1.5.2 (Jan, 2022)
 perf: Regain Minimal memset()-ing During Reuse of Compression Contexts (@Cyan4973, #2969)
@@ -171,8 +171,8 @@ who want earlier signal.
 | Cirrus CI | Used for testing on FreeBSD                                                                                | https://github.com/marketplace/cirrus-ci/                                                                                                              | `.cirrus.yml`          |
 | Circle CI | Historically was used to provide faster signal,<br/> but we may be able to migrate these to Github Actions | https://circleci.com/docs/2.0/getting-started/#setting-up-circleci <br> https://youtu.be/Js3hMUsSZ2c <br> https://circleci.com/docs/2.0/enable-checks/ | `.circleci/config.yml` |

-Note: the instructions linked above mostly cover how to set up a repository with CI from scratch. 
-The general idea should be the same for setting up CI on your fork of zstd, but you may have to 
+Note: the instructions linked above mostly cover how to set up a repository with CI from scratch.
+The general idea should be the same for setting up CI on your fork of zstd, but you may have to
 follow slightly different steps. In particular, please ignore any instructions related to setting up
 config files (since zstd already has configs for each of these services).

@@ -216,7 +216,7 @@ will typically not be stable enough to obtain reliable benchmark results. If you
 hands on a desktop, this is usually a better scenario.

 Of course, benchmarking can be done on non-hyper-stable machines as well. You will just have to
-do a little more work to ensure that you are in fact measuring the changes you've made not and
+do a little more work to ensure that you are in fact measuring the changes you've made and not
 noise. Here are some things you can do to make your benchmarks more stable:

 1. The most simple thing you can do to drastically improve the stability of your benchmark is
@@ -123,7 +123,7 @@ contrib: lib
 	$(MAKE) -C contrib/seekable_format/examples all
 	$(MAKE) -C contrib/seekable_format/tests test
 	$(MAKE) -C contrib/largeNbDicts all
-	$(MAKE) -C contrib/externalMatchfinder all
+	$(MAKE) -C contrib/externalSequenceProducer all
 	cd build/single_file_libs/ ; ./build_decoder_test.sh
 	cd build/single_file_libs/ ; ./build_library_test.sh

@@ -143,7 +143,7 @@ clean:
 	$(Q)$(MAKE) -C contrib/seekable_format/examples $@ > $(VOID)
 	$(Q)$(MAKE) -C contrib/seekable_format/tests $@ > $(VOID)
 	$(Q)$(MAKE) -C contrib/largeNbDicts $@ > $(VOID)
-	$(Q)$(MAKE) -C contrib/externalMatchfinder $@ > $(VOID)
+	$(Q)$(MAKE) -C contrib/externalSequenceProducer $@ > $(VOID)
 	$(Q)$(RM) zstd$(EXT) zstdmt$(EXT) tmp*
 	$(Q)$(RM) -r lz4
 	@echo Cleaning completed
@@ -151,7 +151,7 @@ clean:
 #------------------------------------------------------------------------------
 # make install is validated only for Linux, macOS, Hurd and some BSD targets
 #------------------------------------------------------------------------------
-ifneq (,$(filter $(shell uname),Linux Darwin GNU/kFreeBSD GNU OpenBSD FreeBSD DragonFly NetBSD MSYS_NT Haiku AIX))
+ifneq (,$(filter $(shell uname),Linux Darwin GNU/kFreeBSD GNU OpenBSD FreeBSD DragonFly NetBSD MSYS_NT CYGWIN_NT Haiku AIX))

 HOST_OS = POSIX

@@ -197,18 +197,27 @@ uninstall:
 travis-install:
 	$(MAKE) install PREFIX=~/install_test_dir

+.PHONY: clangbuild-darwin-fat
+clangbuild-darwin-fat: clean
+	clang -v
+	CXX=clang++ CC=clang CFLAGS="-Werror -Wconversion -Wno-sign-conversion -Wdocumentation -arch arm64" $(MAKE) zstd-release
+	mv programs/zstd programs/zstd_arm64
+	CXX=clang++ CC=clang CFLAGS="-Werror -Wconversion -Wno-sign-conversion -Wdocumentation -arch x86_64" $(MAKE) zstd-release
+	mv programs/zstd programs/zstd_x64
+	lipo -create programs/zstd_x64 programs/zstd_arm64 -output programs/zstd
+
 .PHONY: gcc5build gcc6build gcc7build clangbuild m32build armbuild aarch64build ppcbuild ppc64build
 gcc5build: clean
 	gcc-5 -v
-	CC=gcc-5 $(MAKE) all MOREFLAGS="-Werror"
+	CC=gcc-5 $(MAKE) all MOREFLAGS="-Werror $(MOREFLAGS)"

 gcc6build: clean
 	gcc-6 -v
-	CC=gcc-6 $(MAKE) all MOREFLAGS="-Werror"
+	CC=gcc-6 $(MAKE) all MOREFLAGS="-Werror $(MOREFLAGS)"

 gcc7build: clean
 	gcc-7 -v
-	CC=gcc-7 $(MAKE) all MOREFLAGS="-Werror"
+	CC=gcc-7 $(MAKE) all MOREFLAGS="-Werror $(MOREFLAGS)"

 clangbuild: clean
 	clang -v
@@ -232,17 +241,17 @@ ppc64build: clean

 .PHONY: armfuzz aarch64fuzz ppcfuzz ppc64fuzz
 armfuzz: clean
-	CC=arm-linux-gnueabi-gcc QEMU_SYS=qemu-arm-static MOREFLAGS="-static" FUZZER_FLAGS=--no-big-tests $(MAKE) -C $(TESTDIR) fuzztest
+	CC=arm-linux-gnueabi-gcc QEMU_SYS=qemu-arm-static MOREFLAGS="-static $(MOREFLAGS)" FUZZER_FLAGS="--no-big-tests $(FUZZER_FLAGS)" $(MAKE) -C $(TESTDIR) fuzztest

 aarch64fuzz: clean
 	ld -v
-	CC=aarch64-linux-gnu-gcc QEMU_SYS=qemu-aarch64-static MOREFLAGS="-static" FUZZER_FLAGS=--no-big-tests $(MAKE) -C $(TESTDIR) fuzztest
+	CC=aarch64-linux-gnu-gcc QEMU_SYS=qemu-aarch64-static MOREFLAGS="-static $(MOREFLAGS)" FUZZER_FLAGS="--no-big-tests $(FUZZER_FLAGS)" $(MAKE) -C $(TESTDIR) fuzztest

 ppcfuzz: clean
-	CC=powerpc-linux-gnu-gcc QEMU_SYS=qemu-ppc-static MOREFLAGS="-static" FUZZER_FLAGS=--no-big-tests $(MAKE) -C $(TESTDIR) fuzztest
+	CC=powerpc-linux-gnu-gcc QEMU_SYS=qemu-ppc-static MOREFLAGS="-static $(MOREFLAGS)" FUZZER_FLAGS="--no-big-tests $(FUZZER_FLAGS)" $(MAKE) -C $(TESTDIR) fuzztest

 ppc64fuzz: clean
-	CC=powerpc-linux-gnu-gcc QEMU_SYS=qemu-ppc64-static MOREFLAGS="-m64 -static" FUZZER_FLAGS=--no-big-tests $(MAKE) -C $(TESTDIR) fuzztest
+	CC=powerpc-linux-gnu-gcc QEMU_SYS=qemu-ppc64-static MOREFLAGS="-m64 -static $(MOREFLAGS)" FUZZER_FLAGS="--no-big-tests $(FUZZER_FLAGS)" $(MAKE) -C $(TESTDIR) fuzztest

 .PHONY: cxxtest gcc5test gcc6test armtest aarch64test ppctest ppc64test
 cxxtest: CXXFLAGS += -Wall -Wextra -Wundef -Wshadow -Wcast-align -Werror
@@ -251,34 +260,34 @@ cxxtest: clean

 gcc5test: clean
 	gcc-5 -v
-	$(MAKE) all CC=gcc-5 MOREFLAGS="-Werror"
+	$(MAKE) all CC=gcc-5 MOREFLAGS="-Werror $(MOREFLAGS)"

 gcc6test: clean
 	gcc-6 -v
-	$(MAKE) all CC=gcc-6 MOREFLAGS="-Werror"
+	$(MAKE) all CC=gcc-6 MOREFLAGS="-Werror $(MOREFLAGS)"

 armtest: clean
 	$(MAKE) -C $(TESTDIR) datagen   # use native, faster
-	$(MAKE) -C $(TESTDIR) test CC=arm-linux-gnueabi-gcc QEMU_SYS=qemu-arm-static ZSTDRTTEST= MOREFLAGS="-Werror -static" FUZZER_FLAGS=--no-big-tests
+	$(MAKE) -C $(TESTDIR) test CC=arm-linux-gnueabi-gcc QEMU_SYS=qemu-arm-static ZSTDRTTEST= MOREFLAGS="-Werror -static $(MOREFLAGS)" FUZZER_FLAGS="--no-big-tests $(FUZZER_FLAGS)"

 aarch64test:
 	$(MAKE) -C $(TESTDIR) datagen   # use native, faster
-	$(MAKE) -C $(TESTDIR) test CC=aarch64-linux-gnu-gcc QEMU_SYS=qemu-aarch64-static ZSTDRTTEST= MOREFLAGS="-Werror -static" FUZZER_FLAGS=--no-big-tests
+	$(MAKE) -C $(TESTDIR) test CC=aarch64-linux-gnu-gcc QEMU_SYS=qemu-aarch64-static ZSTDRTTEST= MOREFLAGS="-Werror -static $(MOREFLAGS)" FUZZER_FLAGS="--no-big-tests $(FUZZER_FLAGS)"

 ppctest: clean
 	$(MAKE) -C $(TESTDIR) datagen   # use native, faster
-	$(MAKE) -C $(TESTDIR) test CC=powerpc-linux-gnu-gcc QEMU_SYS=qemu-ppc-static ZSTDRTTEST= MOREFLAGS="-Werror -Wno-attributes -static" FUZZER_FLAGS=--no-big-tests
+	$(MAKE) -C $(TESTDIR) test CC=powerpc-linux-gnu-gcc QEMU_SYS=qemu-ppc-static ZSTDRTTEST= MOREFLAGS="-Werror -Wno-attributes -static $(MOREFLAGS)" FUZZER_FLAGS="--no-big-tests $(FUZZER_FLAGS)"

 ppc64test: clean
 	$(MAKE) -C $(TESTDIR) datagen   # use native, faster
-	$(MAKE) -C $(TESTDIR) test CC=powerpc-linux-gnu-gcc QEMU_SYS=qemu-ppc64-static ZSTDRTTEST= MOREFLAGS="-m64 -static" FUZZER_FLAGS=--no-big-tests
+	$(MAKE) -C $(TESTDIR) test CC=powerpc-linux-gnu-gcc QEMU_SYS=qemu-ppc64-static ZSTDRTTEST= MOREFLAGS="-m64 -static $(MOREFLAGS)" FUZZER_FLAGS="--no-big-tests $(FUZZER_FLAGS)"

 .PHONY: arm-ppc-compilation
 arm-ppc-compilation:
-	$(MAKE) -C $(PRGDIR) clean zstd CC=arm-linux-gnueabi-gcc QEMU_SYS=qemu-arm-static ZSTDRTTEST= MOREFLAGS="-Werror -static"
-	$(MAKE) -C $(PRGDIR) clean zstd CC=aarch64-linux-gnu-gcc QEMU_SYS=qemu-aarch64-static ZSTDRTTEST= MOREFLAGS="-Werror -static"
-	$(MAKE) -C $(PRGDIR) clean zstd CC=powerpc-linux-gnu-gcc QEMU_SYS=qemu-ppc-static ZSTDRTTEST= MOREFLAGS="-Werror -Wno-attributes -static"
-	$(MAKE) -C $(PRGDIR) clean zstd CC=powerpc-linux-gnu-gcc QEMU_SYS=qemu-ppc64-static ZSTDRTTEST= MOREFLAGS="-m64 -static"
+	$(MAKE) -C $(PRGDIR) clean zstd CC=arm-linux-gnueabi-gcc QEMU_SYS=qemu-arm-static ZSTDRTTEST= MOREFLAGS="-Werror -static $(MOREFLAGS)"
+	$(MAKE) -C $(PRGDIR) clean zstd CC=aarch64-linux-gnu-gcc QEMU_SYS=qemu-aarch64-static ZSTDRTTEST= MOREFLAGS="-Werror -static $(MOREFLAGS)"
+	$(MAKE) -C $(PRGDIR) clean zstd CC=powerpc-linux-gnu-gcc QEMU_SYS=qemu-ppc-static ZSTDRTTEST= MOREFLAGS="-Werror -Wno-attributes -static $(MOREFLAGS)"
+	$(MAKE) -C $(PRGDIR) clean zstd CC=powerpc-linux-gnu-gcc QEMU_SYS=qemu-ppc64-static ZSTDRTTEST= MOREFLAGS="-m64 -static $(MOREFLAGS)"

 regressiontest:
 	$(MAKE) -C $(FUZZDIR) regressiontest
@@ -308,31 +317,31 @@ update_regressionResults:
 # run UBsan with -fsanitize-recover=pointer-overflow
 # this only works with recent compilers such as gcc 8+
 usan: clean
-	$(MAKE) test CC=clang MOREFLAGS="-g -fno-sanitize-recover=all -fsanitize-recover=pointer-overflow -fsanitize=undefined -Werror"
+	$(MAKE) test CC=clang MOREFLAGS="-g -fno-sanitize-recover=all -fsanitize-recover=pointer-overflow -fsanitize=undefined -Werror $(MOREFLAGS)"

 asan: clean
-	$(MAKE) test CC=clang MOREFLAGS="-g -fsanitize=address -Werror"
+	$(MAKE) test CC=clang MOREFLAGS="-g -fsanitize=address -Werror $(MOREFLAGS)"

 asan-%: clean
-	LDFLAGS=-fuse-ld=gold MOREFLAGS="-g -fno-sanitize-recover=all -fsanitize=address -Werror" $(MAKE) -C $(TESTDIR) $*
+	LDFLAGS=-fuse-ld=gold MOREFLAGS="-g -fno-sanitize-recover=all -fsanitize=address -Werror $(MOREFLAGS)" $(MAKE) -C $(TESTDIR) $*

 msan: clean
-	$(MAKE) test CC=clang MOREFLAGS="-g -fsanitize=memory -fno-omit-frame-pointer -Werror" HAVE_LZMA=0   # datagen.c fails this test for no obvious reason
+	$(MAKE) test CC=clang MOREFLAGS="-g -fsanitize=memory -fno-omit-frame-pointer -Werror $(MOREFLAGS)" HAVE_LZMA=0   # datagen.c fails this test for no obvious reason

 msan-%: clean
-	LDFLAGS=-fuse-ld=gold MOREFLAGS="-g -fno-sanitize-recover=all -fsanitize=memory -fno-omit-frame-pointer -Werror" FUZZER_FLAGS=--no-big-tests $(MAKE) -C $(TESTDIR) HAVE_LZMA=0 $*
+	LDFLAGS=-fuse-ld=gold MOREFLAGS="-g -fno-sanitize-recover=all -fsanitize=memory -fno-omit-frame-pointer -Werror $(MOREFLAGS)" FUZZER_FLAGS="--no-big-tests $(FUZZER_FLAGS)" $(MAKE) -C $(TESTDIR) HAVE_LZMA=0 $*

 asan32: clean
-	$(MAKE) -C $(TESTDIR) test32 CC=clang MOREFLAGS="-g -fsanitize=address"
+	$(MAKE) -C $(TESTDIR) test32 CC=clang MOREFLAGS="-g -fsanitize=address $(MOREFLAGS)"

 uasan: clean
-	$(MAKE) test CC=clang MOREFLAGS="-g -fno-sanitize-recover=all -fsanitize-recover=pointer-overflow -fsanitize=address,undefined -Werror"
+	$(MAKE) test CC=clang MOREFLAGS="-g -fno-sanitize-recover=all -fsanitize-recover=pointer-overflow -fsanitize=address,undefined -Werror $(MOREFLAGS)"

 uasan-%: clean
-	LDFLAGS=-fuse-ld=gold MOREFLAGS="-g -fno-sanitize-recover=all -fsanitize-recover=pointer-overflow -fsanitize=address,undefined -Werror" $(MAKE) -C $(TESTDIR) $*
+	LDFLAGS=-fuse-ld=gold MOREFLAGS="-g -fno-sanitize-recover=all -fsanitize-recover=pointer-overflow -fsanitize=address,undefined -Werror $(MOREFLAGS)" $(MAKE) -C $(TESTDIR) $*

 tsan-%: clean
-	LDFLAGS=-fuse-ld=gold MOREFLAGS="-g -fno-sanitize-recover=all -fsanitize=thread -Werror" $(MAKE) -C $(TESTDIR) $* FUZZER_FLAGS=--no-big-tests
+	LDFLAGS=-fuse-ld=gold MOREFLAGS="-g -fno-sanitize-recover=all -fsanitize=thread -Werror $(MOREFLAGS)" $(MAKE) -C $(TESTDIR) $* FUZZER_FLAGS="--no-big-tests $(FUZZER_FLAGS)"

 .PHONY: apt-install
 apt-install:
@@ -404,7 +413,7 @@ cmakebuild:

 c89build: clean
 	$(CC) -v
-	CFLAGS="-std=c89 -Werror -O0" $(MAKE) allmost  # will fail, due to missing support for `long long`
+	CFLAGS="-std=c89 -Werror -Wno-attributes -Wpedantic -Wno-long-long -Wno-variadic-macros -O0" $(MAKE) lib zstd

 gnu90build: clean
 	$(CC) -v
@@ -5,7 +5,7 @@ targeting real-time compression scenarios at zlib-level and better compression r
 It's backed by a very fast entropy stage, provided by [Huff0 and FSE library](https://github.com/Cyan4973/FiniteStateEntropy).

 Zstandard's format is stable and documented in [RFC8878](https://datatracker.ietf.org/doc/html/rfc8878). Multiple independent implementations are already available.
-This repository represents the reference implementation, provided as an open-source dual [BSD](LICENSE) and [GPLv2](COPYING) licensed **C** library,
+This repository represents the reference implementation, provided as an open-source dual [BSD](LICENSE) OR [GPLv2](COPYING) licensed **C** library,
 and a command line utility producing and decoding `.zst`, `.gz`, `.xz` and `.lz4` files.
 Should your project require another programming language,
 a list of known ports and bindings is provided on [Zstandard homepage](https://facebook.github.io/zstd/#other-languages).
@@ -13,15 +13,12 @@ a list of known ports and bindings is provided on [Zstandard homepage](https://f
 **Development branch status:**

 [![Build Status][travisDevBadge]][travisLink]
-[![Build status][AppveyorDevBadge]][AppveyorLink]
 [![Build status][CircleDevBadge]][CircleLink]
 [![Build status][CirrusDevBadge]][CirrusLink]
 [![Fuzzing Status][OSSFuzzBadge]][OSSFuzzLink]

 [travisDevBadge]: https://api.travis-ci.com/facebook/zstd.svg?branch=dev "Continuous Integration test suite"
 [travisLink]: https://travis-ci.com/facebook/zstd
-[AppveyorDevBadge]: https://ci.appveyor.com/api/projects/status/xt38wbdxjk5mrbem/branch/dev?svg=true "Windows test suite"
-[AppveyorLink]: https://ci.appveyor.com/project/YannCollet/zstd-p0yf0
 [CircleDevBadge]: https://circleci.com/gh/facebook/zstd/tree/dev.svg?style=shield "Short test suite"
 [CircleLink]: https://circleci.com/gh/facebook/zstd
 [CirrusDevBadge]: https://api.cirrus-ci.com/github/facebook/zstd.svg?branch=dev
@@ -154,6 +151,18 @@ to create `zstd` binary, and `libzstd` dynamic and static libraries.

 By default, `CMAKE_BUILD_TYPE` is set to `Release`.

+#### Support for Fat (Universal2) Output
+
+`zstd` can be built and installed with support for both Apple Silicon (M1/M2) as well as Intel by using CMake's Universal2 support.
+To perform a Fat/Universal2 build and install use the following commands:
+
+```bash
+cmake -B build-cmake-debug -S build/cmake -G Ninja -DCMAKE_OSX_ARCHITECTURES="x86_64;x86_64h;arm64"
+cd build-cmake-debug
+ninja
+sudo ninja install
+```
+
 ### Meson

 A Meson project is provided within [`build/meson`](build/meson). Follow
@@ -191,18 +200,20 @@ The output binary will be in `buck-out/gen/programs/`.

 ## Testing

-You can run quick local smoke tests by executing the `playTest.sh` script from the `src/tests` directory.
-Two env variables `$ZSTD_BIN` and `$DATAGEN_BIN` are needed for the test script to locate the zstd and datagen binary.
-For information on CI testing, please refer to TESTING.md
+You can run quick local smoke tests by running `make check`.
+If you can't use `make`, execute the `playTest.sh` script from the `src/tests` directory.
+Two env variables `$ZSTD_BIN` and `$DATAGEN_BIN` are needed for the test script to locate the `zstd` and `datagen` binary.
+For information on CI testing, please refer to `TESTING.md`.

 ## Status

-Zstandard is currently deployed within Facebook. It is used continuously to compress large amounts of data in multiple formats and use cases.
+Zstandard is currently deployed within Facebook and many other large cloud infrastructures.
+It is run continuously to compress large amounts of data in multiple formats and use cases.
 Zstandard is considered safe for production environments.

 ## License

-Zstandard is dual-licensed under [BSD](LICENSE) and [GPLv2](COPYING).
+Zstandard is dual-licensed under [BSD](LICENSE) OR [GPLv2](COPYING).

 ## Contributing

@@ -1,205 +0,0 @@
-# Following tests are run _only_ on `release` branch
-# and on selected feature branch named `appveyorTest` or `visual*`
-
-
-  version: 1.0.{build}
-  branches:
-    only:
-    - release
-    - master
-    - /appveyor*/
-    - /visual*/
-  environment:
-    matrix:
-    - COMPILER: "gcc"
-      HOST:     "mingw"
-      PLATFORM: "x64"
-      SCRIPT:   "make allzstd MOREFLAGS=-static"
-      ARTIFACT: "true"
-      BUILD:    "true"
-    - COMPILER: "gcc"
-      HOST:     "mingw"
-      PLATFORM: "x86"
-      SCRIPT:   "make allzstd MOREFLAGS=-static"
-      ARTIFACT: "true"
-      BUILD:    "true"
-
-    - COMPILER: "clang-cl"
-      HOST:     "cmake-visual"
-      PLATFORM: "x64"
-      CONFIGURATION: "Release"
-      CMAKE_GENERATOR: "Visual Studio 15 2017"
-      CMAKE_GENERATOR_PLATFORM: "x64"
-      CMAKE_GENERATOR_TOOLSET: "LLVM"
-      APPVEYOR_BUILD_WORKER_IMAGE: "Visual Studio 2017"
-
-  install:
-  - ECHO Installing %COMPILER% %PLATFORM% %CONFIGURATION%
-  - SET PATH_ORIGINAL=%PATH%
-  - if [%HOST%]==[mingw] (
-      SET "PATH_MINGW32=C:\mingw-w64\i686-6.3.0-posix-dwarf-rt_v5-rev1\mingw32\bin" &&
-      SET "PATH_MINGW64=C:\mingw-w64\x86_64-6.3.0-posix-seh-rt_v5-rev1\mingw64\bin" &&
-      COPY C:\msys64\usr\bin\make.exe C:\mingw-w64\i686-6.3.0-posix-dwarf-rt_v5-rev1\mingw32\bin\make.exe &&
-      COPY C:\msys64\usr\bin\make.exe C:\mingw-w64\x86_64-6.3.0-posix-seh-rt_v5-rev1\mingw64\bin\make.exe
-    )
-  - IF [%HOST%]==[visual] IF [%PLATFORM%]==[x64] (
-      SET ADDITIONALPARAM=/p:LibraryPath="C:\Program Files\Microsoft SDKs\Windows\v7.1\lib\x64;c:\Program Files (x86)\Microsoft Visual Studio 10.0\VC\lib\amd64;C:\Program Files (x86)\Microsoft Visual Studio 10.0\;C:\Program Files (x86)\Microsoft Visual Studio 10.0\lib\amd64;"
-    )
-
-  build_script:
-  - if [%HOST%]==[mingw] (
-      ( if [%PLATFORM%]==[x64] (
-        SET "PATH=%PATH_MINGW64%;%PATH_ORIGINAL%"
-      ) else if [%PLATFORM%]==[x86] (
-        SET "PATH=%PATH_MINGW32%;%PATH_ORIGINAL%"
-      ) )
-    )
-  - if [%HOST%]==[mingw] if [%BUILD%]==[true] (
-      make -v &&
-      sh -c "%COMPILER% -v" &&
-      ECHO Building zlib to static link &&
-      SET "CC=%COMPILER%" &&
-      sh -c "cd .. && git clone --depth 1 --branch v1.2.11 https://github.com/madler/zlib" &&
-      sh -c "cd ../zlib && make -f win32/Makefile.gcc libz.a"
-      ECHO Building zstd &&
-      SET "CPPFLAGS=-I../../zlib" &&
-      SET "LDFLAGS=../../zlib/libz.a" &&
-      sh -c "%SCRIPT%" &&
-      ( if [%COMPILER%]==[gcc] if [%ARTIFACT%]==[true]
-          ECHO Creating artifacts &&
-          ECHO %cd% &&
-          lib\dll\example\build_package.bat &&
-          make -C programs DEBUGFLAGS= clean zstd &&
-          cd programs\ && 7z a -tzip -mx9 zstd-win-binary-%PLATFORM%.zip zstd.exe &&
-          appveyor PushArtifact zstd-win-binary-%PLATFORM%.zip &&
-          cp zstd.exe ..\bin\zstd.exe &&
-          git clone --depth 1 --branch release https://github.com/facebook/zstd &&
-          cd zstd &&
-          git archive --format=tar release -o zstd-src.tar &&
-          ..\zstd -19 zstd-src.tar &&
-          appveyor PushArtifact zstd-src.tar.zst &&
-          certUtil -hashfile zstd-src.tar.zst SHA256 > zstd-src.tar.zst.sha256.sig &&
-          appveyor PushArtifact zstd-src.tar.zst.sha256.sig &&
-          cd ..\..\bin\ &&
-          7z a -tzip -mx9 zstd-win-release-%PLATFORM%.zip * &&
-          appveyor PushArtifact zstd-win-release-%PLATFORM%.zip
-      )
-    )
-  - if [%HOST%]==[cmake-visual] (
-      ECHO *** &&
-      ECHO *** Building %CMAKE_GENERATOR% ^(%CMAKE_GENERATOR_TOOLSET%^) %PLATFORM%\%CONFIGURATION% &&
-      PUSHD build\cmake &&
-      cmake -DBUILD_TESTING=ON . &&
-      cmake --build . --config %CONFIGURATION% -j4 &&
-      POPD &&
-      ECHO ***
-    )
-
-  test_script:
-  - ECHO Testing %COMPILER% %PLATFORM% %CONFIGURATION%
-  - SET "CC=gcc"
-  - SET "CXX=g++"
-  - if [%TEST%]==[cmake] (
-      mkdir build\cmake\build &&
-      cd build\cmake\build &&
-      SET FUZZERTEST=-T2mn &&
-      SET ZSTREAM_TESTTIME=-T2mn &&
-      cmake -G "Visual Studio 14 2015 Win64" .. &&
-      cd ..\..\.. &&
-      make clean
-    )
-
-
-# The following tests are for regular pushes
-# into `dev` or some feature branch
-# There run less tests, for shorter feedback loop
-
-
-  version: 1.0.{build}
-  environment:
-    matrix:
-    - COMPILER: "visual"
-      HOST:     "visual"
-      PLATFORM: "x64"
-      CONFIGURATION: "Debug"
-    - COMPILER: "visual"
-      HOST:     "visual"
-      PLATFORM: "Win32"
-      CONFIGURATION: "Debug"
-    - COMPILER: "visual"
-      HOST:     "visual"
-      PLATFORM: "x64"
-      CONFIGURATION: "Release"
-    - COMPILER: "visual"
-      HOST:     "visual"
-      PLATFORM: "Win32"
-      CONFIGURATION: "Release"
-
-    - COMPILER: "gcc"
-      HOST:     "cygwin"
-      PLATFORM: "x64"
-
-    - COMPILER: "clang-cl"
-      HOST:     "cmake-visual"
-      PLATFORM: "x64"
-      CONFIGURATION: "Release"
-      CMAKE_GENERATOR: "Visual Studio 15 2017"
-      CMAKE_GENERATOR_PLATFORM: "x64"
-      CMAKE_GENERATOR_TOOLSET: "LLVM"
-      APPVEYOR_BUILD_WORKER_IMAGE: "Visual Studio 2017"
-
-  install:
-  - ECHO Installing %COMPILER% %PLATFORM% %CONFIGURATION%
-  - SET PATH_ORIGINAL=%PATH%
-  - if [%HOST%]==[cygwin] (
-      ECHO Installing Cygwin Packages &&
-      C:\cygwin64\setup-x86_64.exe -qnNdO -R "C:\cygwin64" -g -P ^
-        gcc,^
-        cmake,^
-        make
-    )
-  - IF [%HOST%]==[visual] IF [%PLATFORM%]==[x64] (
-      SET ADDITIONALPARAM=/p:LibraryPath="C:\Program Files\Microsoft SDKs\Windows\v7.1\lib\x64;c:\Program Files (x86)\Microsoft Visual Studio 10.0\VC\lib\amd64;C:\Program Files (x86)\Microsoft Visual Studio 10.0\;C:\Program Files (x86)\Microsoft Visual Studio 10.0\lib\amd64;"
-    )
-
-  build_script:
-  - ECHO Building %COMPILER% %PLATFORM% %CONFIGURATION%
-  - if [%HOST%]==[cygwin] (
-      set CHERE_INVOKING=yes &&
-      set CC=%COMPILER% &&
-      C:\cygwin64\bin\bash --login -c "
-        set -e;
-        cd build/cmake;
-        CFLAGS='-Werror' cmake -G 'Unix Makefiles' -DCMAKE_BUILD_TYPE=Debug -DZSTD_BUILD_TESTS:BOOL=ON -DZSTD_FUZZER_FLAGS=-T20s -DZSTD_ZSTREAM_FLAGS=-T20s -DZSTD_FULLBENCH_FLAGS=-i0 .;
-        make VERBOSE=1 -j;
-        ctest -V -L Medium;
-      "
-    )
-  - if [%HOST%]==[cmake-visual] (
-      ECHO *** &&
-      ECHO *** Building %CMAKE_GENERATOR% ^(%CMAKE_GENERATOR_TOOLSET%^) %PLATFORM%\%CONFIGURATION% &&
-      PUSHD build\cmake &&
-      cmake -DBUILD_TESTING=ON . &&
-      cmake --build . --config %CONFIGURATION% -j4 &&
-      POPD &&
-      ECHO ***
-    )
-  - if [%HOST%]==[visual] (
-      ECHO *** &&
-      ECHO *** Building Visual Studio 2012 %PLATFORM%\%CONFIGURATION% &&
-      ECHO *** &&
-      msbuild "build\VS2010\zstd.sln" /m /verbosity:minimal /property:PlatformToolset=v110 /p:ForceImportBeforeCppTargets=%APPVEYOR_BUILD_FOLDER%\build\VS2010\CompileAsCpp.props /t:Clean,Build /p:Platform=%PLATFORM% /p:Configuration=%CONFIGURATION% /logger:"C:\Program Files\AppVeyor\BuildAgent\Appveyor.MSBuildLogger.dll" &&
-      DIR build\VS2010\bin\%PLATFORM%_%CONFIGURATION%\*.exe &&
-      msbuild "build\VS2010\zstd.sln" /m /verbosity:minimal /property:PlatformToolset=v110 /t:Clean,Build /p:Platform=%PLATFORM% /p:Configuration=%CONFIGURATION% /logger:"C:\Program Files\AppVeyor\BuildAgent\Appveyor.MSBuildLogger.dll" &&
-      DIR build\VS2010\bin\%PLATFORM%_%CONFIGURATION%\*.exe
-    )
-
-
-  test_script:
-  - ECHO Testing %COMPILER% %PLATFORM% %CONFIGURATION%
-  - SET "FUZZERTEST=-T10s"
-  - if [%HOST%]==[mingw] (
-      set "CC=%COMPILER%" &&
-      make clean &&
-      make check
-    )
@@ -169,7 +169,6 @@
      <WarningLevel>Level4</WarningLevel>
      <Optimization>Disabled</Optimization>
      <PreprocessorDefinitions>ZSTD_DLL_EXPORT=1;ZSTD_MULTITHREAD=1;ZSTD_LEGACY_SUPPORT=5;WIN32;_DEBUG;_CONSOLE;_CRT_SECURE_NO_WARNINGS;%(PreprocessorDefinitions)</PreprocessorDefinitions>
-      <MinimalRebuild>true</MinimalRebuild>
      <BasicRuntimeChecks>EnableFastChecks</BasicRuntimeChecks>
      <RuntimeLibrary>MultiThreadedDebugDLL</RuntimeLibrary>
      <DebugInformationFormat>EditAndContinue</DebugInformationFormat>
@@ -162,7 +162,6 @@
      <WarningLevel>Level4</WarningLevel>
      <Optimization>Disabled</Optimization>
      <PreprocessorDefinitions>ZSTD_MULTITHREAD=1;ZSTD_LEGACY_SUPPORT=5;WIN32;_DEBUG;_CONSOLE;_CRT_SECURE_NO_WARNINGS;%(PreprocessorDefinitions)</PreprocessorDefinitions>
-      <MinimalRebuild>true</MinimalRebuild>
      <BasicRuntimeChecks>EnableFastChecks</BasicRuntimeChecks>
      <RuntimeLibrary>MultiThreadedDebugDLL</RuntimeLibrary>
      <DebugInformationFormat>EditAndContinue</DebugInformationFormat>
@@ -114,6 +114,7 @@
    <RootNamespace>zstd</RootNamespace>
    <OutDir>$(SolutionDir)bin\$(Platform)_$(Configuration)\</OutDir>
    <IntDir>$(SolutionDir)bin\obj\$(RootNamespace)_$(Platform)_$(Configuration)\</IntDir>
+    <InstructionSet>NotSet</InstructionSet>
  </PropertyGroup>
  <Import Project="$(VCTargetsPath)\Microsoft.Cpp.Default.props" />
  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'" Label="Configuration">
@@ -187,6 +188,7 @@
      <PreprocessorDefinitions>ZSTD_MULTITHREAD=1;ZSTD_LEGACY_SUPPORT=5;WIN32;_DEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions>
      <TreatWarningAsError>true</TreatWarningAsError>
      <EnablePREfast>false</EnablePREfast>
+      <EnableEnhancedInstructionSet>$(InstructionSet)</EnableEnhancedInstructionSet>
    </ClCompile>
    <Link>
      <SubSystem>Console</SubSystem>
@@ -203,6 +205,7 @@
      <PreprocessorDefinitions>ZSTD_MULTITHREAD=1;ZSTD_LEGACY_SUPPORT=5;WIN32;_DEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions>
      <TreatWarningAsError>true</TreatWarningAsError>
      <EnablePREfast>false</EnablePREfast>
+      <EnableEnhancedInstructionSet>$(InstructionSet)</EnableEnhancedInstructionSet>
    </ClCompile>
    <Link>
      <SubSystem>Console</SubSystem>
@@ -222,6 +225,7 @@
      <EnablePREfast>false</EnablePREfast>
      <TreatWarningAsError>false</TreatWarningAsError>
      <RuntimeLibrary>MultiThreaded</RuntimeLibrary>
+      <EnableEnhancedInstructionSet>$(InstructionSet)</EnableEnhancedInstructionSet>
    </ClCompile>
    <Link>
      <SubSystem>Console</SubSystem>
@@ -244,6 +248,7 @@
      <EnablePREfast>false</EnablePREfast>
      <RuntimeLibrary>MultiThreaded</RuntimeLibrary>
      <AdditionalOptions>/DZSTD_MULTITHREAD %(AdditionalOptions)</AdditionalOptions>
+      <EnableEnhancedInstructionSet>$(InstructionSet)</EnableEnhancedInstructionSet>
    </ClCompile>
    <Link>
      <SubSystem>Console</SubSystem>
@@ -1,6 +1,15 @@
 include(CheckCXXCompilerFlag)
 include(CheckCCompilerFlag)
-include(CheckLinkerFlag)
+# VERSION_GREATER_EQUAL requires CMake 3.7 or later.
+# https://cmake.org/cmake/help/latest/command/if.html#version-greater-equal
+if (CMAKE_VERSION VERSION_LESS 3.18)
+    set(ZSTD_HAVE_CHECK_LINKER_FLAG false)
+else ()
+    set(ZSTD_HAVE_CHECK_LINKER_FLAG true)
+endif ()
+if (ZSTD_HAVE_CHECK_LINKER_FLAG)
+    include(CheckLinkerFlag)
+endif()

 function(EnableCompilerFlag _flag _C _CXX _LD)
    string(REGEX REPLACE "\\+" "PLUS" varname "${_flag}")
@@ -20,7 +29,20 @@ function(EnableCompilerFlag _flag _C _CXX _LD)
        endif ()
    endif ()
    if (_LD)
-        CHECK_LINKER_FLAG(C ${_flag} LD_FLAG_${varname})
+        # We never add a linker flag with CMake < 3.18. We will
+        # implement CHECK_LINKER_FLAG() like feature for CMake < 3.18
+        # or require CMake >= 3.18 when we need to add a required
+        # linker flag in future.
+        #
+        # We also skip linker flags check for MSVC compilers (which includes
+        # clang-cl) since currently check_linker_flag() doesn't give correct
+        # results for this configuration,
+        # see: https://gitlab.kitware.com/cmake/cmake/-/issues/22023
+        if (ZSTD_HAVE_CHECK_LINKER_FLAG AND NOT MSVC)
+            CHECK_LINKER_FLAG(C ${_flag} LD_FLAG_${varname})
+        else ()
+            set(LD_FLAG_${varname} false)
+        endif ()
        if (LD_FLAG_${varname})
            set(CMAKE_EXE_LINKER_FLAGS "${CMAKE_EXE_LINKER_FLAGS} ${_flag}" PARENT_SCOPE)
            set(CMAKE_SHARED_LINKER_FLAGS "${CMAKE_SHARED_LINKER_FLAGS} ${_flag}" PARENT_SCOPE)
@@ -12,6 +12,22 @@ project(libzstd C ASM)
 set(CMAKE_INCLUDE_CURRENT_DIR TRUE)
 option(ZSTD_BUILD_STATIC "BUILD STATIC LIBRARIES" ON)
 option(ZSTD_BUILD_SHARED "BUILD SHARED LIBRARIES" ON)
+option(ZSTD_BUILD_COMPRESSION "BUILD COMPRESSION MODULE" ON)
+option(ZSTD_BUILD_DECOMPRESSION "BUILD DECOMPRESSION MODULE" ON)
+option(ZSTD_BUILD_DICTBUILDER "BUILD DICTBUILDER MODULE" ON)
+option(ZSTD_BUILD_DEPRECATED "BUILD DEPRECATED MODULE" OFF)
+
+set(ZSTDLIB_VISIBLE "" CACHE STRING "Visiblity for ZSTDLIB API")
+set(ZSTDERRORLIB_VISIBLE "" CACHE STRING "Visiblity for ZSTDERRORLIB_VISIBLE API")
+set(ZDICTLIB_VISIBLE "" CACHE STRING "Visiblity for ZDICTLIB_VISIBLE API")
+set(ZSTDLIB_STATIC_API "" CACHE STRING "Visiblity for ZSTDLIB_STATIC_API API")
+set(ZDICTLIB_STATIC_API "" CACHE STRING "Visiblity for ZDICTLIB_STATIC_API API")
+
+set_property(CACHE ZSTDLIB_VISIBLE PROPERTY STRINGS "" "hidden" "default" "protected" "internal")
+set_property(CACHE ZSTDERRORLIB_VISIBLE PROPERTY STRINGS "" "hidden" "default" "protected" "internal")
+set_property(CACHE ZDICTLIB_VISIBLE PROPERTY STRINGS "" "hidden" "default" "protected" "internal")
+set_property(CACHE ZSTDLIB_STATIC_API PROPERTY STRINGS "" "hidden" "default" "protected" "internal")
+set_property(CACHE ZDICTLIB_STATIC_API PROPERTY STRINGS "" "hidden" "default" "protected" "internal")

 if(NOT ZSTD_BUILD_SHARED AND NOT ZSTD_BUILD_STATIC)
    message(SEND_ERROR "You need to build at least one flavor of libzstd")
@@ -29,24 +45,32 @@ else ()
    file(GLOB DecompressSources ${LIBRARY_DIR}/decompress/*.c ${LIBRARY_DIR}/decompress/*.S)
 endif ()
 file(GLOB DictBuilderSources ${LIBRARY_DIR}/dictBuilder/*.c)
-
-set(Sources
-        ${CommonSources}
-        ${CompressSources}
-        ${DecompressSources}
-        ${DictBuilderSources})
+file(GLOB DeprecatedSources ${LIBRARY_DIR}/deprecated/*.c)

 file(GLOB CommonHeaders ${LIBRARY_DIR}/common/*.h)
 file(GLOB CompressHeaders ${LIBRARY_DIR}/compress/*.h)
 file(GLOB DecompressHeaders ${LIBRARY_DIR}/decompress/*.h)
 file(GLOB DictBuilderHeaders ${LIBRARY_DIR}/dictBuilder/*.h)
+file(GLOB DeprecatedHeaders ${LIBRARY_DIR}/deprecated/*.h)

-set(Headers
-        ${LIBRARY_DIR}/zstd.h
-        ${CommonHeaders}
-        ${CompressHeaders}
-        ${DecompressHeaders}
-        ${DictBuilderHeaders})
+set(Sources ${CommonSources})
+set(Headers ${LIBRARY_DIR}/zstd.h ${CommonHeaders})
+if (ZSTD_BUILD_COMPRESSION)
+    set(Sources ${Sources} ${CompressSources})
+    set(Headers ${Headers} ${CompressHeaders})
+endif()
+if (ZSTD_BUILD_DECOMPRESSION)
+    set(Sources ${Sources} ${DecompressSources})
+    set(Headers ${Headers} ${DecompressHeaders})
+endif()
+if (ZSTD_BUILD_DICTBUILDER)
+    set(Sources ${Sources} ${DictBuilderSources})
+    set(Headers ${Headers} ${DictBuilderHeaders})
+endif()
+if (ZSTD_BUILD_DEPRECATED)
+    set(Sources ${Sources} ${DeprecatedSources})
+    set(Headers ${Headers} ${DeprecatedHeaders})
+endif()

 if (ZSTD_LEGACY_SUPPORT)
    set(LIBRARY_LEGACY_DIR ${LIBRARY_DIR}/legacy)
@@ -81,7 +105,15 @@ endif ()
 # Our assembly expects to be compiled by a C compiler, and is only enabled for
 # __GNUC__ compatible compilers. Otherwise all the ASM code is disabled by
 # macros.
-set_source_files_properties(${Sources} PROPERTIES LANGUAGE C)
+if(NOT CMAKE_ASM_COMPILER STREQUAL CMAKE_C_COMPILER)
+    set_source_files_properties(${Sources} PROPERTIES LANGUAGE C)
+endif()
+
+macro (add_definition target var)
+    if (NOT ("${${var}}" STREQUAL ""))
+        set_property(TARGET ${target} APPEND PROPERTY COMPILE_DEFINITIONS "${var}=__attribute__((visibility(\"${${var}}\")))")
+    endif ()
+endmacro ()

 # Split project to static and shared libraries build
 set(library_targets)
@@ -93,7 +125,10 @@ if (ZSTD_BUILD_SHARED)
        if (UNIX)
            target_link_libraries(libzstd_shared ${THREADS_LIBS})
        endif ()
-    endif()
+    endif ()
+    add_definition(libzstd_shared ZSTDLIB_VISIBLE)
+    add_definition(libzstd_shared ZSTDERRORLIB_VISIBLE)
+    add_definition(libzstd_shared ZDICTLIB_VISIBLE)
 endif ()
 if (ZSTD_BUILD_STATIC)
    add_library(libzstd_static STATIC ${Sources} ${Headers})
@@ -104,6 +139,11 @@ if (ZSTD_BUILD_STATIC)
            target_link_libraries(libzstd_static ${THREADS_LIBS})
        endif ()
    endif ()
+    add_definition(libzstd_static ZSTDLIB_VISIBLE)
+    add_definition(libzstd_static ZSTDERRORLIB_VISIBLE)
+    add_definition(libzstd_static ZDICTLIB_VISIBLE)
+    add_definition(libzstd_static ZSTDLIB_STATIC_API)
+    add_definition(libzstd_static ZDICTLIB_STATIC_API)
 endif ()

 # Add specific compile definitions for MSVC project
@@ -32,7 +32,17 @@ if (MSVC)
    set(PlatformDependResources ${MSVC_RESOURCE_DIR}/zstd.rc)
 endif ()

-add_executable(zstd ${PROGRAMS_DIR}/zstdcli.c ${PROGRAMS_DIR}/util.c ${PROGRAMS_DIR}/timefn.c ${PROGRAMS_DIR}/fileio.c ${PROGRAMS_DIR}/fileio_asyncio.c ${PROGRAMS_DIR}/benchfn.c ${PROGRAMS_DIR}/benchzstd.c ${PROGRAMS_DIR}/datagen.c ${PROGRAMS_DIR}/dibio.c ${PROGRAMS_DIR}/zstdcli_trace.c ${PlatformDependResources})
+set(ZSTD_PROGRAM_SRCS ${PROGRAMS_DIR}/zstdcli.c ${PROGRAMS_DIR}/util.c
+    ${PROGRAMS_DIR}/timefn.c ${PROGRAMS_DIR}/fileio.c
+    ${PROGRAMS_DIR}/fileio_asyncio.c ${PROGRAMS_DIR}/benchfn.c
+    ${PROGRAMS_DIR}/benchzstd.c ${PROGRAMS_DIR}/datagen.c
+    ${PROGRAMS_DIR}/dibio.c ${PROGRAMS_DIR}/zstdcli_trace.c
+    ${PlatformDependResources})
+if (MSVC AND ZSTD_PROGRAMS_LINK_SHARED)
+    list(APPEND ZSTD_PROGRAM_SRCS ${LIBRARY_DIR}/common/pool.c ${LIBRARY_DIR}/common/threading.c)
+endif ()
+
+add_executable(zstd ${ZSTD_PROGRAM_SRCS})
 target_link_libraries(zstd ${PROGRAMS_ZSTD_LINK_TARGET})
 if (CMAKE_SYSTEM_NAME MATCHES "(Solaris|SunOS)")
    target_link_libraries(zstd rt)
@@ -75,7 +85,9 @@ if (UNIX)
        ${CMAKE_CURRENT_BINARY_DIR}/zstdless.1
        DESTINATION "${MAN_INSTALL_DIR}")

-    add_executable(zstd-frugal ${PROGRAMS_DIR}/zstdcli.c ${PROGRAMS_DIR}/util.c ${PROGRAMS_DIR}/timefn.c ${PROGRAMS_DIR}/fileio.c ${PROGRAMS_DIR}/fileio_asyncio.c)
+    add_executable(zstd-frugal ${PROGRAMS_DIR}/zstdcli.c
+        ${PROGRAMS_DIR}/util.c ${PROGRAMS_DIR}/timefn.c
+        ${PROGRAMS_DIR}/fileio.c ${PROGRAMS_DIR}/fileio_asyncio.c)
    target_link_libraries(zstd-frugal ${PROGRAMS_ZSTD_LINK_TARGET})
    set_property(TARGET zstd-frugal APPEND PROPERTY COMPILE_DEFINITIONS "ZSTD_NOBENCH;ZSTD_NODICT;ZSTD_NOTRACE")
 endif ()
@@ -132,7 +132,7 @@ endif

 subdir('lib')

-if bin_programs
+if bin_programs or bin_tests
  subdir('programs')
 endif

@@ -72,7 +72,14 @@ zstd = executable('zstd',
  c_args: zstd_c_args,
  dependencies: zstd_deps,
  export_dynamic: export_dynamic_on_windows, # Since Meson 0.45.0
-  install: true)
+  build_by_default: bin_programs,
+  install: bin_programs)
+
+if not bin_programs
+  # we generate rules to build the programs, but don't install anything
+  # so do not continue to installing scripts and manpages
+  subdir_done()
+endif

 zstd_frugal_sources = [join_paths(zstd_rootdir, 'programs/zstdcli.c'),
  join_paths(zstd_rootdir, 'programs/timefn.c'),
@@ -162,7 +162,7 @@ if host_machine_os != os_windows
      playTests_sh,
      args: opt,
      env: ['ZSTD_BIN=' + zstd.full_path(), 'DATAGEN_BIN=./datagen'],
-      depends: [datagen],
+      depends: [datagen, zstd],
      suite: suite,
      workdir: meson.current_build_dir(),
      timeout: 2800) # Timeout should work on HDD drive
@@ -1,13 +1,13 @@
 # Dockerfile
 # First image to build the binary
-FROM alpine as builder
+FROM alpine@sha256:69665d02cb32192e52e07644d76bc6f25abeb5410edc1c7a81a10ba3f0efb90a as builder

 RUN apk --no-cache add make gcc libc-dev
 COPY . /src
 RUN mkdir /pkg && cd /src && make && make DESTDIR=/pkg install

 # Second minimal image to only keep the built binary
-FROM alpine
+FROM alpine@sha256:69665d02cb32192e52e07644d76bc6f25abeb5410edc1c7a81a10ba3f0efb90a

 # Copy the built files
 COPY --from=builder /pkg /
@@ -1,2 +0,0 @@
-# build artifacts
-externalMatchfinder
@@ -1,14 +0,0 @@
-externalMatchfinder
-=====================
-
-`externalMatchfinder` is a test tool for the external matchfinder API.
-It demonstrates how to use the API to perform a simple round-trip test.
-
-A sample matchfinder is provided in matchfinder.c, but the user can swap
-this out with a different one if desired. The sample matchfinder implements
-LZ compression with a 1KB hashtable. Dictionary compression is not currently supported.
-
-Command line :
-```
-externalMatchfinder filename
-```
@@ -0,0 +1,2 @@
+# build artifacts
+externalSequenceProducer
@@ -23,11 +23,11 @@ DEBUGFLAGS= -Wall -Wextra -Wcast-qual -Wcast-align -Wshadow \
            -Wredundant-decls
 CFLAGS  += $(DEBUGFLAGS) $(MOREFLAGS)

-default: externalMatchfinder
+default: externalSequenceProducer

-all: externalMatchfinder
+all: externalSequenceProducer

-externalMatchfinder: matchfinder.c main.c $(LIBZSTD)
+externalSequenceProducer: sequence_producer.c main.c $(LIBZSTD)
 	$(CC) $(CPPFLAGS) $(CFLAGS) $^ $(LDFLAGS) -o $@

 .PHONY: $(LIBZSTD)
@@ -37,4 +37,4 @@ $(LIBZSTD):
 clean:
 	$(RM) *.o
 	$(MAKE) -C $(LIBDIR) clean > /dev/null
-	$(RM) externalMatchfinder
+	$(RM) externalSequenceProducer
@@ -0,0 +1,14 @@
+externalSequenceProducer
+=====================
+
+`externalSequenceProducer` is a test tool for the Block-Level Sequence Producer API.
+It demonstrates how to use the API to perform a simple round-trip test.
+
+A sample sequence producer is provided in sequence_producer.c, but the user can swap
+this out with a different one if desired. The sample sequence producer implements
+LZ parsing with a 1KB hashtable. Dictionary-based parsing is not currently supported.
+
+Command line :
+```
+externalSequenceProducer filename
+```
@@ -16,7 +16,7 @@
 #define ZSTD_STATIC_LINKING_ONLY
 #include "zstd.h"
 #include "zstd_errors.h"
-#include "matchfinder.h" // simpleExternalMatchFinder
+#include "sequence_producer.h" // simpleSequenceProducer

 #define CHECK(res)                                      \
 do {                                                    \
@@ -28,23 +28,23 @@ do {                                                    \

 int main(int argc, char *argv[]) {
    if (argc != 2) {
-        printf("Usage: exampleMatchfinder <file>\n");
+        printf("Usage: externalSequenceProducer <file>\n");
        return 1;
    }

    ZSTD_CCtx* const zc = ZSTD_createCCtx();

-    int simpleExternalMatchState = 0xdeadbeef;
+    int simpleSequenceProducerState = 0xdeadbeef;

    // Here is the crucial bit of code!
-    ZSTD_registerExternalMatchFinder(
+    ZSTD_registerSequenceProducer(
        zc,
-        &simpleExternalMatchState,
-        simpleExternalMatchFinder
+        &simpleSequenceProducerState,
+        simpleSequenceProducer
    );

    {
-        size_t const res = ZSTD_CCtx_setParameter(zc, ZSTD_c_enableMatchFinderFallback, 1);
+        size_t const res = ZSTD_CCtx_setParameter(zc, ZSTD_c_enableSeqProducerFallback, 1);
        CHECK(res);
    }

@@ -9,15 +9,15 @@
 */

 #include "zstd_compress_internal.h"
-#include "matchfinder.h"
+#include "sequence_producer.h"

 #define HSIZE 1024
 static U32 const HLOG = 10;
 static U32 const MLS = 4;
 static U32 const BADIDX = 0xffffffff;

-size_t simpleExternalMatchFinder(
-  void* externalMatchState,
+size_t simpleSequenceProducer(
+  void* sequenceProducerState,
  ZSTD_Sequence* outSeqs, size_t outSeqsCapacity,
  const void* src, size_t srcSize,
  const void* dict, size_t dictSize,
@@ -31,7 +31,7 @@ size_t simpleExternalMatchFinder(
    size_t seqCount = 0;
    U32 hashTable[HSIZE];

-    (void)externalMatchState;
+    (void)sequenceProducerState;
    (void)dict;
    (void)dictSize;
    (void)outSeqsCapacity;
@@ -14,8 +14,8 @@
 #define ZSTD_STATIC_LINKING_ONLY
 #include "zstd.h"

-size_t simpleExternalMatchFinder(
-  void* externalMatchState,
+size_t simpleSequenceProducer(
+  void* sequenceProducerState,
  ZSTD_Sequence* outSeqs, size_t outSeqsCapacity,
  const void* src, size_t srcSize,
  const void* dict, size_t dictSize,
@@ -985,7 +985,7 @@ int usage(const char* exeName)
    DISPLAY ("-#          : use compression level # (default: %u) \n", CLEVEL_DEFAULT);
    DISPLAY ("-D #        : use # as a dictionary (default: create one) \n");
    DISPLAY ("-i#         : nb benchmark rounds (default: %u) \n", BENCH_TIME_DEFAULT_S);
-    DISPLAY ("-p#         : print speed for all rounds 0=fastest 1=median (default: 0)");
+    DISPLAY ("-p#         : print speed for all rounds 0=fastest 1=median (default: 0) \n");
    DISPLAY ("--nbBlocks=#: use # blocks for bench (default: one per file) \n");
    DISPLAY ("--nbDicts=# : create # dictionaries for bench (default: one per block) \n");
    DISPLAY ("-h          : help (this text) \n");
@@ -32,7 +32,6 @@ libzstd:
 		-DZSTD_LEGACY_SUPPORT=0 \
 		-DZSTD_STATIC_LINKING_ONLY \
 		-DFSE_STATIC_LINKING_ONLY \
-		-DHUF_STATIC_LINKING_ONLY \
 		-DXXH_STATIC_LINKING_ONLY \
 		-D__GNUC__ \
 		-D__linux__=1 \
@@ -24,6 +24,7 @@
 /*-****************************************
 *  Compiler specifics
 ******************************************/
+#undef MEM_STATIC /* may be already defined from common/compiler.h */
 #define MEM_STATIC static inline

 /*-**************************************************************
@@ -24,9 +24,6 @@ EXPORT_SYMBOL_GPL(HUF_readStats_wksp);
 EXPORT_SYMBOL_GPL(ZSTD_isError);
 EXPORT_SYMBOL_GPL(ZSTD_getErrorName);
 EXPORT_SYMBOL_GPL(ZSTD_getErrorCode);
-EXPORT_SYMBOL_GPL(ZSTD_customMalloc);
-EXPORT_SYMBOL_GPL(ZSTD_customCalloc);
-EXPORT_SYMBOL_GPL(ZSTD_customFree);

 MODULE_LICENSE("Dual BSD/GPL");
 MODULE_DESCRIPTION("Zstd Common");
@@ -84,7 +84,7 @@ static uint64_t ZSTD_div64(uint64_t dividend, uint32_t divisor) {

 #include <linux/kernel.h>

-#define assert(x) WARN_ON((x))
+#define assert(x) WARN_ON(!(x))

 #endif /* ZSTD_DEPS_ASSERT */
 #endif /* ZSTD_DEPS_NEED_ASSERT */
@@ -37,10 +37,16 @@ CFLAGS   += -Wno-deprecated-declarations
 PZSTD_INC  = -I$(ZSTDDIR) -I$(ZSTDDIR)/common -I$(PROGDIR) -I.
 GTEST_INC  = -isystem googletest/googletest/include

+# If default C++ version is older than C++11, explicitly set C++11, which is the
+# minimum required by the code.
+ifeq ($(shell echo "\043if __cplusplus < 201103L\n\043error\n\043endif" | $(CXX) -x c++ -Werror -c - -o /dev/null 2>/dev/null && echo 1 || echo 0),0)
+PZSTD_CXX_STD := -std=c++11
+endif
+
 PZSTD_CPPFLAGS  = $(PZSTD_INC)
 PZSTD_CCXXFLAGS =
 PZSTD_CFLAGS    = $(PZSTD_CCXXFLAGS)
-PZSTD_CXXFLAGS  = $(PZSTD_CCXXFLAGS) -std=c++11
+PZSTD_CXXFLAGS  = $(PZSTD_CCXXFLAGS) $(PZSTD_CXX_STD)
 PZSTD_LDFLAGS   =
 EXTRA_FLAGS     =
 ALL_CFLAGS      = $(EXTRA_FLAGS) $(CPPFLAGS) $(PZSTD_CPPFLAGS) $(CFLAGS)   $(PZSTD_CFLAGS)
@@ -0,0 +1,42 @@
+# Zstandard Seekable Format
+
+The seekable format splits compressed data into a series of independent "frames",
+each compressed individually,
+so that decompression of a section in the middle of an archive
+only requires zstd to decompress at most a frame's worth of extra data,
+instead of the entire archive.
+
+The frames are appended, so that the decompression of the entire payload
+still regenerates the original content, using any compliant zstd decoder.
+
+On top of that, the seekable format generates a jump table,
+which makes it possible to jump directly to the position of the relevant frame
+when requesting only a segment of the data.
+The jump table is simply ignored by zstd decoders unaware of the seekable format.
+
+The format is delivered with an API to create seekable archives
+and to retrieve arbitrary segments inside the archive.
+
+### Maximum Frame Size parameter
+
+When creating a seekable archive, the main parameter is the maximum frame size.
+
+At compression time, user can manually select the boundaries between segments,
+but they don't have to: long segments will be automatically split
+when larger than selected maximum frame size.
+
+Small frame sizes reduce decompression cost when requesting small segments,
+because the decoder will nonetheless have to decompress an entire frame
+to recover just a single byte from it.
+
+A good rule of thumb is to select a maximum frame size roughly equivalent
+to the access pattern when it's known.
+For example, if the application tends to request 4KB blocks,
+then it's a good idea to set a maximum frame size in the vicinity of 4 KB.
+
+But small frame sizes also reduce compression ratio,
+and increase the cost for the jump table,
+so there is a balance to find.
+
+In general, try to avoid really tiny frame sizes (<1 KB),
+which would have a large negative impact on compression ratio.
@@ -25,7 +25,7 @@

 #include "pool.h"      // use zstd thread pool for demo

-#include "zstd_seekable.h"
+#include "../zstd_seekable.h"

 static void* malloc_orDie(size_t size)
 {
@@ -29,7 +29,7 @@

 #include "pool.h"      // use zstd thread pool for demo

-#include "zstd_seekable.h"
+#include "../zstd_seekable.h"

 #define MIN(a, b) ((a) < (b) ? (a) : (b))

@@ -13,7 +13,7 @@
 #define ZSTD_STATIC_LINKING_ONLY
 #include <zstd.h>      // presumes zstd library is installed

-#include "zstd_seekable.h"
+#include "../zstd_seekable.h"

 static void* malloc_orDie(size_t size)
 {
@@ -112,20 +112,23 @@ static char* createOutFilename_orDie(const char* filename)
    return (char*)outSpace;
 }

-int main(int argc, const char** argv) {
+#define CLEVEL_DEFAULT 5
+int main(int argc, const char** argv)
+{
    const char* const exeName = argv[0];
-    if (argc!=3) {
-        printf("wrong arguments\n");
-        printf("usage:\n");
-        printf("%s FILE FRAME_SIZE\n", exeName);
+    if (argc<3 || argc>4) {
+        printf("wrong arguments \n");
+        printf("usage: \n");
+        printf("%s FILE FRAME_SIZE [LEVEL] \n", exeName);
        return 1;
    }

    {   const char* const inFileName = argv[1];
        unsigned const frameSize = (unsigned)atoi(argv[2]);
+        int const cLevel = (argc==4) ? atoi(argv[3]) : CLEVEL_DEFAULT;

        char* const outFileName = createOutFilename_orDie(inFileName);
-        compressFile_orDie(inFileName, outFileName, 5, frameSize);
+        compressFile_orDie(inFileName, outFileName, cLevel, frameSize);
        free(outFileName);
    }

@@ -16,7 +16,7 @@
 #include <zstd.h>      // presumes zstd library is installed
 #include <zstd_errors.h>

-#include "zstd_seekable.h"
+#include "../zstd_seekable.h"

 #define MIN(a, b) ((a) < (b) ? (a) : (b))

@@ -3,8 +3,55 @@
 #include <stdlib.h>  // malloc
 #include <stdio.h>
 #include <assert.h>
+#include <string.h>

-#include "zstd_seekable.h"
+#include "../zstd_seekable.h"
+
+
+/* ZSTD_seekable_customFile implementation that reads/seeks a buffer while keeping track of total bytes read */
+typedef struct {
+    const void *ptr;
+    size_t size;
+    size_t pos;
+    size_t totalRead;
+} buffWrapperWithTotal_t;
+
+static int readBuffWithTotal(void* opaque, void* buffer, size_t n)
+{
+    buffWrapperWithTotal_t* const buff = (buffWrapperWithTotal_t*)opaque;
+    assert(buff != NULL);
+    if (buff->pos + n > buff->size) return -1;
+    memcpy(buffer, (const char*)buff->ptr + buff->pos, n);
+    buff->pos += n;
+    buff->totalRead += n;
+    return 0;
+}
+
+static int seekBuffWithTotal(void* opaque, long long offset, int origin)
+{
+    buffWrapperWithTotal_t* const buff = (buffWrapperWithTotal_t*) opaque;
+    unsigned long long newOffset;
+    assert(buff != NULL);
+    switch (origin) {
+    case SEEK_SET:
+        assert(offset >= 0);
+        newOffset = (unsigned long long)offset;
+        break;
+    case SEEK_CUR:
+        newOffset = (unsigned long long)((long long)buff->pos + offset);
+        break;
+    case SEEK_END:
+        newOffset = (unsigned long long)((long long)buff->size + offset);
+        break;
+    default:
+        assert(0);  /* not possible */
+    }
+    if (newOffset > buff->size) {
+        return -1;
+    }
+    buff->pos = newOffset;
+    return 0;
+}

 /* Basic unit tests for zstd seekable format */
 int main(int argc, const char** argv)
@@ -220,6 +267,92 @@ int main(int argc, const char** argv)
    }
    printf("Success!\n");

+
+    printf("Test %u - multiple decompress calls: ", testNb++);
+    {   char const inBuffer[] = "Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt";
+        size_t const inSize = sizeof(inBuffer);
+
+        size_t const seekCapacity = 5000;
+        void* const seekBuffer = malloc(seekCapacity);
+        assert(seekBuffer != NULL);
+        size_t seekSize;
+
+        size_t const outCapacity = inSize;
+        char* const outBuffer = malloc(outCapacity);
+        assert(outBuffer != NULL);
+
+        ZSTD_seekable_CStream* const zscs = ZSTD_seekable_createCStream();
+        assert(zscs != NULL);
+
+        /* compress test data with a small frame size to ensure multiple frames in the output */
+        unsigned const maxFrameSize = 40;
+        { size_t const initStatus = ZSTD_seekable_initCStream(zscs, 9, 0 /* checksumFlag */, maxFrameSize);
+          assert(!ZSTD_isError(initStatus));
+        }
+
+        {   ZSTD_outBuffer outb = { .dst=seekBuffer, .pos=0, .size=seekCapacity };
+            ZSTD_inBuffer inb = { .src=inBuffer, .pos=0, .size=inSize };
+
+            while (inb.pos < inb.size) {
+                size_t const cStatus = ZSTD_seekable_compressStream(zscs, &outb, &inb);
+                assert(!ZSTD_isError(cStatus));
+            }
+
+            size_t const endStatus = ZSTD_seekable_endStream(zscs, &outb);
+            assert(!ZSTD_isError(endStatus));
+            seekSize = outb.pos;
+        }
+
+        ZSTD_seekable* const stream = ZSTD_seekable_create();
+        assert(stream != NULL);
+        buffWrapperWithTotal_t buffWrapper = {seekBuffer, seekSize, 0, 0};
+        { ZSTD_seekable_customFile srcFile = {&buffWrapper, &readBuffWithTotal, &seekBuffWithTotal};
+          size_t const initStatus = ZSTD_seekable_initAdvanced(stream, srcFile);
+          assert(!ZSTD_isError(initStatus)); }
+
+        /* Perform a series of small reads and seeks (repeatedly read 1 byte and skip 1 byte)
+           and check that we didn't reread input data unnecessarily */
+        size_t pos;
+        for (pos = 0; pos < inSize; pos += 2) {
+            size_t const decStatus = ZSTD_seekable_decompress(stream, outBuffer, 1, pos);
+            if (decStatus != 1 || outBuffer[0] != inBuffer[pos]) {
+                goto _test_error;
+            }
+        }
+        if (buffWrapper.totalRead > seekSize) {
+            /* We read more than the compressed size, meaning there were some rereads.
+               This is unneeded because we only seeked forward. */
+            printf("Too much data read: %zu read, with compressed size %zu\n", buffWrapper.totalRead, seekSize);
+            goto _test_error;
+        }
+
+        /* Perform some reads and seeks to ensure correctness */
+        struct {
+            size_t offset;
+            size_t size;
+        } const tests[] = {  /* Assume the frame size is 40 */
+            {20, 40}, /* read partial data from two frames */
+            {60, 10}, /* continue reading from the same offset */
+            {50, 20}, /* seek backward within the same frame */
+            {10, 10}, /* seek backward to a different frame */
+            {25, 10}, /* seek forward within the same frame */
+            {60, 10}, /* seek forward to a different frame */
+        };
+        size_t idx;
+        for (idx = 0; idx < sizeof(tests) / sizeof(tests[0]); idx++) {
+            size_t const decStatus = ZSTD_seekable_decompress(stream, outBuffer, tests[idx].size, tests[idx].offset);
+            if (decStatus != tests[idx].size || memcmp(outBuffer, inBuffer + tests[idx].offset, tests[idx].size) != 0) {
+                goto _test_error;
+            }
+        }
+
+        free(seekBuffer);
+        free(outBuffer);
+        ZSTD_seekable_freeCStream(zscs);
+        ZSTD_seekable_free(stream);
+    }
+    printf("Success!\n");
+
    /* TODO: Add more tests */
    printf("Finished tests\n");
    return 0;
@@ -15,8 +15,8 @@ extern "C" {

 #define ZSTD_SEEKABLE_MAXFRAMES 0x8000000U

-/* Limit the maximum size to avoid any potential issues storing the compressed size */
-#define ZSTD_SEEKABLE_MAX_FRAME_DECOMPRESSED_SIZE 0x80000000U
+/* Limit maximum size to avoid potential issues storing the compressed size */
+#define ZSTD_SEEKABLE_MAX_FRAME_DECOMPRESSED_SIZE 0x40000000U

 /*-****************************************************************************
 *  Seekable Format
@@ -48,10 +48,19 @@ typedef struct ZSTD_seekTable_s ZSTD_seekTable;
 *
 *  Use ZSTD_seekable_initCStream() to initialize a ZSTD_seekable_CStream object
 *  for a new compression operation.
-*  `maxFrameSize` indicates the size at which to automatically start a new
-*  seekable frame.  `maxFrameSize == 0` implies the default maximum size.
-*  `checksumFlag` indicates whether or not the seek table should include frame
-*  checksums on the uncompressed data for verification.
+*  - `maxFrameSize` indicates the size at which to automatically start a new
+*            seekable frame.
+*            `maxFrameSize == 0` implies the default maximum size.
+*            Smaller frame sizes allow faster decompression of small segments,
+*            since retrieving a single byte requires decompression of
+*            the full frame where the byte belongs.
+*            In general, size the frames to roughly correspond to
+*            the access granularity (when it's known).
+*            But small sizes also reduce compression ratio.
+*            Avoid really tiny frame sizes (< 1 KB),
+*            that would hurt compression ratio considerably.
+*  - `checksumFlag` indicates whether or not the seek table should include frame
+*            checksums on the uncompressed data for verification.
 *  @return : a size hint for input to provide for compression, or an error code
 *            checkable with ZSTD_isError()
 *
@@ -230,6 +230,8 @@ size_t ZSTD_seekable_compressStream(ZSTD_seekable_CStream* zcs, ZSTD_outBuffer*
    const BYTE* const inBase = (const BYTE*) input->src + input->pos;
    size_t inLen = input->size - input->pos;

+    assert(zcs->maxFrameSize < INT_MAX);
+    ZSTD_CCtx_setParameter(zcs->cstream, ZSTD_c_srcSizeHint, (int)zcs->maxFrameSize);
    inLen = MIN(inLen, (size_t)(zcs->maxFrameSize - zcs->frameDSize));

    /* if we haven't finished flushing the last frame, don't start writing a new one */
@@ -493,7 +493,7 @@ size_t ZSTD_seekable_decompress(ZSTD_seekable* zs, void* dst, size_t len, unsign
    size_t srcBytesRead = 0;
    do {
        /* check if we can continue from a previous decompress job */
-        if (targetFrame != zs->curFrame || offset != zs->decompressedOffset) {
+        if (targetFrame != zs->curFrame || offset < zs->decompressedOffset) {
            zs->decompressedOffset = zs->seekTable.entries[targetFrame].dOffset;
            zs->curFrame = targetFrame;

@@ -517,9 +517,9 @@ size_t ZSTD_seekable_decompress(ZSTD_seekable* zs, void* dst, size_t len, unsign
            size_t forwardProgress;
            if (zs->decompressedOffset < offset) {
                /* dummy decompressions until we get to the target offset */
-                outTmp = (ZSTD_outBuffer){zs->outBuff, MIN(SEEKABLE_BUFF_SIZE, offset - zs->decompressedOffset), 0};
+                outTmp = (ZSTD_outBuffer){zs->outBuff, (size_t) (MIN(SEEKABLE_BUFF_SIZE, offset - zs->decompressedOffset)), 0};
            } else {
-                outTmp = (ZSTD_outBuffer){dst, len, zs->decompressedOffset - offset};
+                outTmp = (ZSTD_outBuffer){dst, len, (size_t) (zs->decompressedOffset - offset)};
            }

            prevOutPos = outTmp.pos;
@@ -6,12 +6,53 @@ Each entry will contain:
 1. The last affected decompressor versions.
 2. The decompressor components affected.
 2. Whether the compressed frame could ever be produced by the reference compressor.
-3. An example frame.
+3. An example frame (hexadecimal string when it can be short enough, link to golden file otherwise)
 4. A description of the bug.

 The document is in reverse chronological order, with the bugs that affect the most recent zstd decompressor versions listed first.


+No sequence using the 2-bytes format
+------------------------------------------------
+
+**Last affected version**: v1.5.5
+
+**Affected decompressor component(s)**: Library & CLI
+
+**Produced by the reference compressor**: No
+
+**Example Frame**: see zstd/tests/golden-decompression/zeroSeq_2B.zst
+
+The zstd decoder incorrectly expects FSE tables when there are 0 sequences present in the block
+if the value 0 is encoded using the 2-bytes format.
+Instead, it should immediately end the sequence section, and move on to next block.
+
+This situation was never generated by the reference compressor,
+because representing 0 sequences with the 2-bytes format is inefficient
+(the 1-byte format is always used in this case).
+
+
+Compressed block with a size of exactly 128 KB
+------------------------------------------------
+
+**Last affected version**: v1.5.2
+
+**Affected decompressor component(s)**: Library & CLI
+
+**Produced by the reference compressor**: No
+
+**Example Frame**: see zstd/tests/golden-decompression/block-128k.zst
+
+The zstd decoder incorrectly rejected blocks of type `Compressed_Block` when their size was exactly 128 KB.
+Note that `128 KB - 1` was accepted, and `128 KB + 1` is forbidden by the spec.
+
+This type of block was never generated by the reference compressor.
+
+These blocks used to be disallowed by the spec up until spec version 0.3.2 when the restriction was lifted by [PR#1689](https://github.com/facebook/zstd/pull/1689).
+
+> A Compressed_Block has the extra restriction that Block_Size is always strictly less than the decompressed size. If this condition cannot be respected, the block must be sent uncompressed instead (Raw_Block).
+
+
 Compressed block with 0 literals and 0 sequences
 ------------------------------------------------

@@ -31,6 +72,7 @@ Additionally, these blocks were disallowed by the spec up until spec version 0.3

 > A Compressed_Block has the extra restriction that Block_Size is always strictly less than the decompressed size. If this condition cannot be respected, the block must be sent uncompressed instead (Raw_Block).

+
 First block is RLE block
 ------------------------

@@ -52,6 +94,7 @@ block.

 https://github.com/facebook/zstd/blob/8814aa5bfa74f05a86e55e9d508da177a893ceeb/lib/compress/zstd_compress.c#L3527-L3535

+
 Tiny FSE Table & Block
 ----------------------

@@ -997,7 +997,8 @@ static void decompress_sequences(frame_context_t *const ctx,
                                 const size_t num_sequences);
 static sequence_command_t decode_sequence(sequence_states_t *const state,
                                          const u8 *const src,
-                                          i64 *const offset);
+                                          i64 *const offset,
+                                          int lastSequence);
 static void decode_seq_table(FSE_dtable *const table, istream_t *const in,
                               const seq_part_t type, const seq_mode_t mode);

@@ -1017,12 +1018,7 @@ static size_t decode_sequences(frame_context_t *const ctx, istream_t *in,
    // This is a variable size field using between 1 and 3 bytes. Let's call its
    // first byte byte0."
    u8 header = IO_read_bits(in, 8);
-    if (header == 0) {
-        // "There are no sequences. The sequence section stops there.
-        // Regenerated content is defined entirely by literals section."
-        *sequences = NULL;
-        return 0;
-    } else if (header < 128) {
+    if (header < 128) {
        // "Number_of_Sequences = byte0 . Uses 1 byte."
        num_sequences = header;
    } else if (header < 255) {
@@ -1033,6 +1029,12 @@ static size_t decode_sequences(frame_context_t *const ctx, istream_t *in,
        num_sequences = IO_read_bits(in, 16) + 0x7F00;
    }

+    if (num_sequences == 0) {
+        // "There are no sequences. The sequence section stops there."
+        *sequences = NULL;
+        return 0;
+    }
+
    *sequences = malloc(num_sequences * sizeof(sequence_command_t));
    if (!*sequences) {
        BAD_ALLOC();
@@ -1114,7 +1116,7 @@ static void decompress_sequences(frame_context_t *const ctx, istream_t *in,

    for (size_t i = 0; i < num_sequences; i++) {
        // Decode sequences one by one
-        sequences[i] = decode_sequence(&states, src, &bit_offset);
+        sequences[i] = decode_sequence(&states, src, &bit_offset, i==num_sequences-1);
    }

    if (bit_offset != 0) {
@@ -1125,7 +1127,8 @@ static void decompress_sequences(frame_context_t *const ctx, istream_t *in,
 // Decode a single sequence and update the state
 static sequence_command_t decode_sequence(sequence_states_t *const states,
                                          const u8 *const src,
-                                          i64 *const offset) {
+                                          i64 *const offset,
+                                          int lastSequence) {
    // "Each symbol is a code in its own context, which specifies Baseline and
    // Number_of_Bits to add. Codes are FSE compressed, and interleaved with raw
    // additional bits in the same bitstream."
@@ -1160,7 +1163,7 @@ static sequence_command_t decode_sequence(sequence_states_t *const states,
    // Literals_Length_State is updated, followed by Match_Length_State, and
    // then Offset_State."
    // If the stream is complete don't read bits to update state
-    if (*offset != 0) {
+    if (!lastSequence) {
        FSE_update_state(&states->ll_table, &states->ll_state, src, offset);
        FSE_update_state(&states->ml_table, &states->ml_state, src, offset);
        FSE_update_state(&states->of_table, &states->of_state, src, offset);
@@ -1399,7 +1402,7 @@ size_t ZSTD_get_decompressed_size(const void *src, const size_t src_len) {
 /******* END OUTPUT SIZE COUNTING *********************************************/

 /******* DICTIONARY PARSING ***************************************************/
-dictionary_t* create_dictionary() {
+dictionary_t* create_dictionary(void) {
    dictionary_t* const dict = calloc(1, sizeof(dictionary_t));
    if (!dict) {
        BAD_ALLOC();
@@ -16,7 +16,7 @@ Distribution of this document is unlimited.

 ### Version

-0.3.7 (2020-12-09)
+0.4.0 (2023-06-05)


 Introduction
@@ -470,6 +470,7 @@ This field uses 2 lowest bits of first byte, describing 4 different block types
        repeated `Regenerated_Size` times.
 - `Compressed_Literals_Block` - This is a standard Huffman-compressed block,
        starting with a Huffman tree description.
+        In this mode, there are at least 2 different literals represented in the Huffman tree description.
        See details below.
 - `Treeless_Literals_Block` - This is a Huffman-compressed block,
        using Huffman tree _from previous Huffman-compressed literals block_.
@@ -533,15 +534,20 @@ __`Size_Format` for `Compressed_Literals_Block` and `Treeless_Literals_Block`__
 Both `Compressed_Size` and `Regenerated_Size` fields follow __little-endian__ convention.
 Note: `Compressed_Size` __includes__ the size of the Huffman Tree description
 _when_ it is present.
+Note 2: `Compressed_Size` can never be `==0`.
+Even in single-stream scenario, assuming an empty content, it must be `>=1`,
+since it contains at least the final end bit flag.
+In 4-streams scenario, a valid `Compressed_Size` is necessarily `>= 10`
+(6 bytes for the jump table, + 4x1 bytes for the 4 streams).

-4 streams is superior to 1 stream in decompression speed,
+4 streams is faster than 1 stream in decompression speed,
 by exploiting instruction level parallelism.
 But it's also more expensive,
 costing on average ~7.3 bytes more than the 1 stream mode, mostly from the jump table.

 In general, use the 4 streams mode when there are more literals to decode,
 to favor higher decompression speeds.
-Beyond 1KB, the 4 streams mode is compulsory anyway.
+Note that beyond >1KB of literals, the 4 streams mode is compulsory.

 Note that a minimum of 6 bytes is required for the 4 streams mode.
 That's a technical minimum, but it's not recommended to employ the 4 streams mode
@@ -566,6 +572,7 @@ or from a dictionary.

 ### `Huffman_Tree_Description`
 This section is only present when `Literals_Block_Type` type is `Compressed_Literals_Block` (`2`).
+The tree describes the weights of all literals symbols that can be present in the literals block, at least 2 and up to 256.
 The format of the Huffman tree description can be found at [Huffman Tree description](#huffman-tree-description).
 The size of `Huffman_Tree_Description` is determined during decoding process,
 it must be used to determine where streams begin.
@@ -575,10 +582,10 @@ it must be used to determine where streams begin.
 ### Jump Table
 The Jump Table is only present when there are 4 Huffman-coded streams.

-Reminder : Huffman compressed data consists of either 1 or 4 Huffman-coded streams.
+Reminder : Huffman compressed data consists of either 1 or 4 streams.

 If only one stream is present, it is a single bitstream occupying the entire
-remaining portion of the literals block, encoded as described within
+remaining portion of the literals block, encoded as described in
 [Huffman-Coded Streams](#huffman-coded-streams).

 If there are four streams, `Literals_Section_Header` only provided
@@ -589,17 +596,18 @@ except for the last stream which may be up to 3 bytes smaller,
 to reach a total decompressed size as specified in `Regenerated_Size`.

 The compressed size of each stream is provided explicitly in the Jump Table.
-Jump Table is 6 bytes long, and consist of three 2-byte __little-endian__ fields,
+Jump Table is 6 bytes long, and consists of three 2-byte __little-endian__ fields,
 describing the compressed sizes of the first three streams.
-`Stream4_Size` is computed from total `Total_Streams_Size` minus sizes of other streams.
+`Stream4_Size` is computed from `Total_Streams_Size` minus sizes of other streams:

 `Stream4_Size = Total_Streams_Size - 6 - Stream1_Size - Stream2_Size - Stream3_Size`.

-Note: if `Stream1_Size + Stream2_Size + Stream3_Size > Total_Streams_Size`,
+`Stream4_Size` is necessarily `>= 1`. Therefore,
+if `Total_Streams_Size < Stream1_Size + Stream2_Size + Stream3_Size + 6 + 1`,
 data is considered corrupted.

 Each of these 4 bitstreams is then decoded independently as a Huffman-Coded stream,
-as described at [Huffman-Coded Streams](#huffman-coded-streams)
+as described in [Huffman-Coded Streams](#huffman-coded-streams)


 Sequences Section
@@ -642,13 +650,15 @@ __`Number_of_Sequences`__

 This is a variable size field using between 1 and 3 bytes.
 Let's call its first byte `byte0`.
- `if (byte0 == 0)` : there are no sequences.
-            The sequence section stops there.
-            Decompressed content is defined entirely as Literals Section content.
-            The FSE tables used in `Repeat_Mode` aren't updated.
 - `if (byte0 < 128)` : `Number_of_Sequences = byte0` . Uses 1 byte.
- `if (byte0 < 255)` : `Number_of_Sequences = ((byte0-128) << 8) + byte1` . Uses 2 bytes.
- `if (byte0 == 255)`: `Number_of_Sequences = byte1 + (byte2<<8) + 0x7F00` . Uses 3 bytes.
+- `if (byte0 < 255)` : `Number_of_Sequences = ((byte0 - 0x80) << 8) + byte1`. Uses 2 bytes.
+            Note that the 2 bytes format fully overlaps the 1 byte format.
+- `if (byte0 == 255)`: `Number_of_Sequences = byte1 + (byte2<<8) + 0x7F00`. Uses 3 bytes.
+
+`if (Number_of_Sequences == 0)` : there are no sequences.
+            The sequence section stops immediately,
+            FSE tables used in `Repeat_Mode` aren't updated.
+            Block's decompressed content is defined solely by the Literals Section content.

 __Symbol compression modes__

@@ -1197,7 +1207,7 @@ Huffman Coding
 --------------
 Zstandard Huffman-coded streams are read backwards,
 similar to the FSE bitstreams.
-Therefore, to find the start of the bitstream, it is therefore to
+Therefore, to find the start of the bitstream, it is required to
 know the offset of the last byte of the Huffman-coded stream.

 After writing the last bit containing information, the compressor
@@ -1239,9 +1249,15 @@ Transformation from `Weight` to `Number_of_Bits` follows this formula :
 ```
 Number_of_Bits = Weight ? (Max_Number_of_Bits + 1 - Weight) : 0
 ```
-The last symbol's `Weight` is deduced from previously decoded ones,
-by completing to the nearest power of 2.
-This power of 2 gives `Max_Number_of_Bits`, the depth of the current tree.
+When a literal value is not present, it receives a `Weight` of 0.
+The least frequent symbol receives a `Weight` of 1.
+Consequently, the `Weight` 1 is necessarily present.
+The most frequent symbol receives a `Weight` anywhere between 1 and 11 (max).
+The last symbol's `Weight` is deduced from previously retrieved Weights,
+by completing to the nearest power of 2. It's necessarily non 0.
+If it's not possible to reach a clean power of 2 with a single `Weight` value,
+the Huffman Tree Description is considered invalid.
+This final power of 2 gives `Max_Number_of_Bits`, the depth of the current tree.
 `Max_Number_of_Bits` must be <= 11,
 otherwise the representation is considered corrupted.

@@ -1254,7 +1270,7 @@ Let's presume the following Huffman tree must be described :

 The tree depth is 4, since its longest elements uses 4 bits
 (longest elements are the one with smallest frequency).
-Value `5` will not be listed, as it can be determined from values for 0-4,
+Literal value `5` will not be listed, as it can be determined from previous values 0-4,
 nor will values above `5` as they are all 0.
 Values from `0` to `4` will be listed using `Weight` instead of `Number_of_Bits`.
 Weight formula is :
@@ -1274,7 +1290,7 @@ The `Weight` of `5` can be determined by advancing to the next power of 2.
 The sum of `2^(Weight-1)` (excluding 0's) is :
 `8 + 4 + 2 + 0 + 1 = 15`.
 Nearest larger power of 2 value is 16.
-Therefore, `Max_Number_of_Bits = 4` and `Weight[5] = 16-15 = 1`.
+Therefore, `Max_Number_of_Bits = 4` and `Weight[5] = log_2(16 - 15) + 1 = 1`.

 #### Huffman Tree header

@@ -1683,6 +1699,9 @@ or at least provide a meaningful error code explaining for which reason it canno

 Version changes
 ---------------
+- 0.4.0 : fixed imprecise behavior for nbSeq==0, detected by Igor Pavlov
+- 0.3.9 : clarifications for Huffman-compressed literal sizes.
+- 0.3.8 : clarifications for Huffman Blocks and Huffman Tree descriptions.
 - 0.3.7 : clarifications for Repeat_Offsets, matching RFC8878
 - 0.3.6 : clarifications for Dictionary_ID
 - 0.3.5 : clarifications for Block_Maximum_Size
@@ -1,10 +1,10 @@
 <html>
 <head>
 <meta http-equiv="Content-Type" content="text/html; charset=ISO-8859-1">
-<title>zstd 1.5.3 Manual</title>
+<title>zstd 1.5.5 Manual</title>
 </head>
 <body>
-<h1>zstd 1.5.3 Manual</h1>
+<h1>zstd 1.5.5 Manual</h1>
 <hr>
 <a name="Contents"></a><h2>Contents</h2>
 <ol>
@@ -22,15 +22,15 @@
 <li><a href="#Chapter12">Dictionary helper functions</a></li>
 <li><a href="#Chapter13">Advanced dictionary and prefix API (Requires v1.4.0+)</a></li>
 <li><a href="#Chapter14">experimental API (static linking only)</a></li>
-<li><a href="#Chapter15">Frame size functions</a></li>
+<li><a href="#Chapter15">Frame header and size functions</a></li>
 <li><a href="#Chapter16">Memory management</a></li>
 <li><a href="#Chapter17">Advanced compression functions</a></li>
 <li><a href="#Chapter18">Advanced decompression functions</a></li>
 <li><a href="#Chapter19">Advanced streaming functions</a></li>
-<li><a href="#Chapter20">Buffer-less and synchronous inner streaming functions</a></li>
+<li><a href="#Chapter20">Buffer-less and synchronous inner streaming functions (DEPRECATED)</a></li>
 <li><a href="#Chapter21">Buffer-less streaming compression (synchronous mode)</a></li>
 <li><a href="#Chapter22">Buffer-less streaming decompression (synchronous mode)</a></li>
-<li><a href="#Chapter23">Block level API</a></li>
+<li><a href="#Chapter23">Block level API (DEPRECATED)</a></li>
 </ol>
 <hr>
 <a name="Chapter1"></a><h2>Introduction</h2><pre>
@@ -80,7 +80,8 @@
                const void* src, size_t srcSize,
                      int compressionLevel);
 </b><p>  Compresses `src` content as a single zstd compressed frame into already allocated `dst`.
-  Hint : compression runs faster if `dstCapacity` >=  `ZSTD_compressBound(srcSize)`.
+  NOTE: Providing `dstCapacity >= ZSTD_compressBound(srcSize)` guarantees that zstd will have
+        enough space to successfully compress the data.
  @return : compressed size written into `dst` (<= `dstCapacity),
            or an error code if it fails (which can be tested using ZSTD_isError()). 
 </p></pre><BR>
@@ -121,8 +122,8 @@ unsigned long long ZSTD_getFrameContentSize(const void *src, size_t srcSize);
   note 6 : This function replaces ZSTD_getDecompressedSize() 
 </p></pre><BR>

-<pre><b>ZSTDLIB_API
-ZSTD_DEPRECATED("Replaced by ZSTD_getFrameContentSize")
+<pre><b>ZSTD_DEPRECATED("Replaced by ZSTD_getFrameContentSize")
+ZSTDLIB_API
 unsigned long long ZSTD_getDecompressedSize(const void* src, size_t srcSize);
 </b><p>  NOTE: This function is now obsolete, in favor of ZSTD_getFrameContentSize().
  Both functions work the same way, but ZSTD_getDecompressedSize() blends
@@ -139,8 +140,30 @@ unsigned long long ZSTD_getDecompressedSize(const void* src, size_t srcSize);
        or an error code if input is invalid 
 </p></pre><BR>

-<h3>Helper functions</h3><pre></pre><b><pre>#define ZSTD_COMPRESSBOUND(srcSize)   ((srcSize) + ((srcSize)>>8) + (((srcSize) < (128<<10)) ? (((128<<10) - (srcSize)) >> 11) </b>/* margin, from 64 to 0 */ : 0))  /* this formula ensures that bound(A) + bound(B) <= bound(A+B) as long as A and B >= 128 KB */<b>
-size_t      ZSTD_compressBound(size_t srcSize); </b>/*!< maximum compressed size in worst case single-pass scenario */<b>
+<h3>Helper functions</h3><pre></pre><b><pre></b>/* ZSTD_compressBound() :<b>
+ * maximum compressed size in worst case single-pass scenario.
+ * When invoking `ZSTD_compress()` or any other one-pass compression function,
+ * it's recommended to provide @dstCapacity >= ZSTD_compressBound(srcSize)
+ * as it eliminates one potential failure scenario,
+ * aka not enough room in dst buffer to write the compressed frame.
+ * Note : ZSTD_compressBound() itself can fail, if @srcSize > ZSTD_MAX_INPUT_SIZE .
+ *        In which case, ZSTD_compressBound() will return an error code
+ *        which can be tested using ZSTD_isError().
+ *
+ * ZSTD_COMPRESSBOUND() :
+ * same as ZSTD_compressBound(), but as a macro.
+ * It can be used to produce constants, which can be useful for static allocation,
+ * for example to size a static array on stack.
+ * Will produce constant value 0 if srcSize too large.
+ */
+#define ZSTD_MAX_INPUT_SIZE ((sizeof(size_t)==8) ? 0xFF00FF00FF00FF00LLU : 0xFF00FF00U)
+#define ZSTD_COMPRESSBOUND(srcSize)   (((size_t)(srcSize) >= ZSTD_MAX_INPUT_SIZE) ? 0 : (srcSize) + ((srcSize)>>8) + (((srcSize) < (128<<10)) ? (((128<<10) - (srcSize)) >> 11) </b>/* margin, from 64 to 0 */ : 0))  /* this formula ensures that bound(A) + bound(B) <= bound(A+B) as long as A and B >= 128 KB */<b>
+size_t ZSTD_compressBound(size_t srcSize); </b>/*!< maximum compressed size in worst case single-pass scenario */<b>
+</b>/* ZSTD_isError() :<b>
+ * Most ZSTD_* functions returning a size_t value can be tested for error,
+ * using ZSTD_isError().
+ * @return 1 if error, 0 otherwise
+ */
 unsigned    ZSTD_isError(size_t code);          </b>/*!< tells if a `size_t` function result is an error code */<b>
 const char* ZSTD_getErrorName(size_t code);     </b>/*!< provides readable string from an error code */<b>
 int         ZSTD_minCLevel(void);               </b>/*!< minimum negative compression level allowed, requires v1.4.0+ */<b>
@@ -362,6 +385,8 @@ size_t     ZSTD_freeDCtx(ZSTD_DCtx* dctx);  </b>/* accept NULL pointer */<b>
     * ZSTD_c_useBlockSplitter
     * ZSTD_c_useRowMatchFinder
     * ZSTD_c_prefetchCDictTables
+     * ZSTD_c_enableSeqProducerFallback
+     * ZSTD_c_maxBlockSize
     * Because they are not stable, it's necessary to define ZSTD_STATIC_LINKING_ONLY to access them.
     * note : never ever use experimentalParam? names directly;
     *        also, the enums values themselves are unstable and can still change.
@@ -381,7 +406,10 @@ size_t     ZSTD_freeDCtx(ZSTD_DCtx* dctx);  </b>/* accept NULL pointer */<b>
     ZSTD_c_experimentalParam13=1010,
     ZSTD_c_experimentalParam14=1011,
     ZSTD_c_experimentalParam15=1012,
-     ZSTD_c_experimentalParam16=1013
+     ZSTD_c_experimentalParam16=1013,
+     ZSTD_c_experimentalParam17=1014,
+     ZSTD_c_experimentalParam18=1015,
+     ZSTD_c_experimentalParam19=1016
 } ZSTD_cParameter;
 </b></pre><BR>
 <pre><b>typedef struct {
@@ -444,7 +472,7 @@ size_t     ZSTD_freeDCtx(ZSTD_DCtx* dctx);  </b>/* accept NULL pointer */<b>
                  They will be used to compress next frame.
                  Resetting session never fails.
  - The parameters : changes all parameters back to "default".
-                  This removes any reference to any dictionary too.
+                  This also removes any reference to any dictionary or external sequence producer.
                  Parameters can only be changed between 2 sessions (i.e. no compression is currently ongoing)
                  otherwise the reset fails, and function returns an error value (which can be tested using ZSTD_isError())
  - Both : similar to resetting the session, followed by resetting parameters.
@@ -459,7 +487,8 @@ size_t     ZSTD_freeDCtx(ZSTD_DCtx* dctx);  </b>/* accept NULL pointer */<b>
  Should cctx hold data from a previously unfinished frame, everything about it is forgotten.
  - Compression parameters are pushed into CCtx before starting compression, using ZSTD_CCtx_set*()
  - The function is always blocking, returns when compression is completed.
-  Hint : compression runs faster if `dstCapacity` >=  `ZSTD_compressBound(srcSize)`.
+  NOTE: Providing `dstCapacity >= ZSTD_compressBound(srcSize)` guarantees that zstd will have
+        enough space to successfully compress the data, though it is possible it fails for other reasons.
 @return : compressed size written into `dst` (<= `dstCapacity),
           or an error code if it fails (which can be tested using ZSTD_isError()).
 
@@ -483,13 +512,15 @@ size_t     ZSTD_freeDCtx(ZSTD_DCtx* dctx);  </b>/* accept NULL pointer */<b>
     * ZSTD_d_stableOutBuffer
     * ZSTD_d_forceIgnoreChecksum
     * ZSTD_d_refMultipleDDicts
+     * ZSTD_d_disableHuffmanAssembly
     * Because they are not stable, it's necessary to define ZSTD_STATIC_LINKING_ONLY to access them.
     * note : never ever use experimentalParam? names directly
     */
     ZSTD_d_experimentalParam1=1000,
     ZSTD_d_experimentalParam2=1001,
     ZSTD_d_experimentalParam3=1002,
-     ZSTD_d_experimentalParam4=1003
+     ZSTD_d_experimentalParam4=1003,
+     ZSTD_d_experimentalParam5=1004

 } ZSTD_dParameter;
 </b></pre><BR>
@@ -658,6 +689,9 @@ size_t ZSTD_endStream(ZSTD_CStream* zcs, ZSTD_outBuffer* output);
     ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only);
     ZSTD_CCtx_refCDict(zcs, NULL); // clear the dictionary (if any)
     ZSTD_CCtx_setParameter(zcs, ZSTD_c_compressionLevel, compressionLevel);
+
+ Note that ZSTD_initCStream() clears any previously set dictionary. Use the new API
+ to compress with a dictionary.
 
 </p></pre><BR>

@@ -693,6 +727,33 @@ size_t ZSTD_endStream(ZSTD_CStream* zcs, ZSTD_outBuffer* output);
 size_t ZSTD_freeDStream(ZSTD_DStream* zds);  </b>/* accept NULL pointer */<b>
 </pre></b><BR>
 <h3>Streaming decompression functions</h3><pre></pre><b><pre></pre></b><BR>
+<pre><b>size_t ZSTD_initDStream(ZSTD_DStream* zds);
+</b><p> Initialize/reset DStream state for new decompression operation.
+ Call before new decompression operation using same DStream.
+
+ Note : This function is redundant with the advanced API and equivalent to:
+     ZSTD_DCtx_reset(zds, ZSTD_reset_session_only);
+     ZSTD_DCtx_refDDict(zds, NULL);
+ 
+</p></pre><BR>
+
+<pre><b>size_t ZSTD_decompressStream(ZSTD_DStream* zds, ZSTD_outBuffer* output, ZSTD_inBuffer* input);
+</b><p> Streaming decompression function.
+ Call repetitively to consume full input updating it as necessary.
+ Function will update both input and output `pos` fields exposing current state via these fields:
+ - `input.pos < input.size`, some input remaining and caller should provide remaining input
+   on the next call.
+ - `output.pos < output.size`, decoder finished and flushed all remaining buffers.
+ - `output.pos == output.size`, potentially uncflushed data present in the internal buffers,
+   call ZSTD_decompressStream() again to flush remaining data to output.
+ Note : with no additional input, amount of data flushed <= ZSTD_BLOCKSIZE_MAX.
+
+ @return : 0 when a frame is completely decoded and fully flushed,
+           or an error code, which can be tested using ZSTD_isError(),
+           or any other value > 0, which means there is some decoding or flushing to do to complete current frame.
+ 
+</p></pre><BR>
+
 <pre><b>size_t ZSTD_DStreamInSize(void);    </b>/*!< recommended size for input buffer */<b>
 </b></pre><BR>
 <pre><b>size_t ZSTD_DStreamOutSize(void);   </b>/*!< recommended size for output buffer. Guarantee to successfully flush at least one complete block in all circumstances. */<b>
@@ -807,9 +868,11 @@ size_t ZSTD_freeDStream(ZSTD_DStream* zds);  </b>/* accept NULL pointer */<b>

 <a name="Chapter13"></a><h2>Advanced dictionary and prefix API (Requires v1.4.0+)</h2><pre>
 This API allows dictionaries to be used with ZSTD_compress2(),
- ZSTD_compressStream2(), and ZSTD_decompressDCtx(). Dictionaries are sticky, and
- only reset with the context is reset with ZSTD_reset_parameters or
- ZSTD_reset_session_and_parameters. Prefixes are single-use.
+ ZSTD_compressStream2(), and ZSTD_decompressDCtx().
+ Dictionaries are sticky, they remain valid when same context is re-used,
+ they only reset when the context is reset
+ with ZSTD_reset_parameters or ZSTD_reset_session_and_parameters.
+ In contrast, Prefixes are single-use.
 <BR></pre>

 <pre><b>size_t ZSTD_CCtx_loadDictionary(ZSTD_CCtx* cctx, const void* dict, size_t dictSize);
@@ -818,8 +881,9 @@ size_t ZSTD_freeDStream(ZSTD_DStream* zds);  </b>/* accept NULL pointer */<b>
 @result : 0, or an error code (which can be tested with ZSTD_isError()).
  Special: Loading a NULL (or 0-size) dictionary invalidates previous dictionary,
           meaning "return to no-dictionary mode".
-  Note 1 : Dictionary is sticky, it will be used for all future compressed frames.
-           To return to "no-dictionary" situation, load a NULL dictionary (or reset parameters).
+  Note 1 : Dictionary is sticky, it will be used for all future compressed frames,
+           until parameters are reset, a new dictionary is loaded, or the dictionary
+           is explicitly invalidated by loading a NULL dictionary.
  Note 2 : Loading a dictionary involves building tables.
           It's also a CPU consuming operation, with non-negligible impact on latency.
           Tables are dependent on compression parameters, and for this reason,
@@ -828,11 +892,15 @@ size_t ZSTD_freeDStream(ZSTD_DStream* zds);  </b>/* accept NULL pointer */<b>
           Use experimental ZSTD_CCtx_loadDictionary_byReference() to reference content instead.
           In such a case, dictionary buffer must outlive its users.
  Note 4 : Use ZSTD_CCtx_loadDictionary_advanced()
-           to precisely select how dictionary content must be interpreted. 
+           to precisely select how dictionary content must be interpreted.
+  Note 5 : This method does not benefit from LDM (long distance mode).
+           If you want to employ LDM on some large dictionary content,
+           prefer employing ZSTD_CCtx_refPrefix() described below.
+ 
 </p></pre><BR>

 <pre><b>size_t ZSTD_CCtx_refCDict(ZSTD_CCtx* cctx, const ZSTD_CDict* cdict);
-</b><p>  Reference a prepared dictionary, to be used for all next compressed frames.
+</b><p>  Reference a prepared dictionary, to be used for all future compressed frames.
  Note that compression parameters are enforced from within CDict,
  and supersede any compression parameter previously set within CCtx.
  The parameters ignored are labelled as "superseded-by-cdict" in the ZSTD_cParameter enum docs.
@@ -852,6 +920,7 @@ size_t ZSTD_freeDStream(ZSTD_DStream* zds);  </b>/* accept NULL pointer */<b>
  Decompression will need same prefix to properly regenerate data.
  Compressing with a prefix is similar in outcome as performing a diff and compressing it,
  but performs much faster, especially during decompression (compression speed is tunable with compression level).
+  This method is compatible with LDM (long distance mode).
 @result : 0, or an error code (which can be tested with ZSTD_isError()).
  Special: Adding any prefix (including NULL) invalidates any previous prefix or dictionary
  Note 1 : Prefix buffer is referenced. It **must** outlive compression.
@@ -867,9 +936,9 @@ size_t ZSTD_freeDStream(ZSTD_DStream* zds);  </b>/* accept NULL pointer */<b>
 </p></pre><BR>

 <pre><b>size_t ZSTD_DCtx_loadDictionary(ZSTD_DCtx* dctx, const void* dict, size_t dictSize);
-</b><p>  Create an internal DDict from dict buffer,
-  to be used to decompress next frames.
-  The dictionary remains valid for all future frames, until explicitly invalidated.
+</b><p>  Create an internal DDict from dict buffer, to be used to decompress all future frames.
+  The dictionary remains valid for all future frames, until explicitly invalidated, or
+  a new dictionary is loaded.
 @result : 0, or an error code (which can be tested with ZSTD_isError()).
  Special : Adding a NULL (or 0-size) dictionary invalidates any previous dictionary,
            meaning "return to no-dictionary mode".
@@ -893,9 +962,10 @@ size_t ZSTD_freeDStream(ZSTD_DStream* zds);  </b>/* accept NULL pointer */<b>
  The memory for the table is allocated on the first call to refDDict, and can be
  freed with ZSTD_freeDCtx().

+  If called with ZSTD_d_refMultipleDDicts disabled (the default), only one dictionary
+  will be managed, and referencing a dictionary effectively "discards" any previous one.
+
 @result : 0, or an error code (which can be tested with ZSTD_isError()).
-  Note 1 : Currently, only one dictionary can be managed.
-           Referencing a new dictionary effectively "discards" any previous one.
  Special: referencing a NULL DDict means "return to no-dictionary mode".
  Note 2 : DDict is just referenced, its lifetime must outlive its usage from DCtx.
 
@@ -1085,7 +1155,7 @@ size_t ZSTD_sizeof_DDict(const ZSTD_DDict* ddict);
  ZSTD_ps_disable = 2       </b>/* Do not use the feature */<b>
 } ZSTD_paramSwitch_e;
 </b></pre><BR>
-<a name="Chapter15"></a><h2>Frame size functions</h2><pre></pre>
+<a name="Chapter15"></a><h2>Frame header and size functions</h2><pre></pre>

 <pre><b>ZSTDLIB_STATIC_API unsigned long long ZSTD_findDecompressedSize(const void* src, size_t srcSize);
 </b><p>  `src` should point to the start of a series of ZSTD encoded and/or skippable frames
@@ -1131,11 +1201,90 @@ size_t ZSTD_sizeof_DDict(const ZSTD_DDict* ddict);
           or an error code (if srcSize is too small) 
 </p></pre><BR>

+<pre><b>typedef enum { ZSTD_frame, ZSTD_skippableFrame } ZSTD_frameType_e;
+</b></pre><BR>
+<pre><b>typedef struct {
+    unsigned long long frameContentSize; </b>/* if == ZSTD_CONTENTSIZE_UNKNOWN, it means this field is not available. 0 means "empty" */<b>
+    unsigned long long windowSize;       </b>/* can be very large, up to <= frameContentSize */<b>
+    unsigned blockSizeMax;
+    ZSTD_frameType_e frameType;          </b>/* if == ZSTD_skippableFrame, frameContentSize is the size of skippable content */<b>
+    unsigned headerSize;
+    unsigned dictID;
+    unsigned checksumFlag;
+    unsigned _reserved1;
+    unsigned _reserved2;
+} ZSTD_frameHeader;
+</b></pre><BR>
+<pre><b>ZSTDLIB_STATIC_API size_t ZSTD_getFrameHeader(ZSTD_frameHeader* zfhPtr, const void* src, size_t srcSize);   </b>/**< doesn't consume input */<b>
+</b>/*! ZSTD_getFrameHeader_advanced() :<b>
+ *  same as ZSTD_getFrameHeader(),
+ *  with added capability to select a format (like ZSTD_f_zstd1_magicless) */
+ZSTDLIB_STATIC_API size_t ZSTD_getFrameHeader_advanced(ZSTD_frameHeader* zfhPtr, const void* src, size_t srcSize, ZSTD_format_e format);
+</b><p>  decode Frame Header, or requires larger `srcSize`.
+ @return : 0, `zfhPtr` is correctly filled,
+          >0, `srcSize` is too small, value is wanted `srcSize` amount,
+           or an error code, which can be tested using ZSTD_isError() 
+</p></pre><BR>
+
+<pre><b>ZSTDLIB_STATIC_API size_t ZSTD_decompressionMargin(const void* src, size_t srcSize);
+</b><p> Zstd supports in-place decompression, where the input and output buffers overlap.
+ In this case, the output buffer must be at least (Margin + Output_Size) bytes large,
+ and the input buffer must be at the end of the output buffer.
+
+  _______________________ Output Buffer ________________________
+ |                                                              |
+ |                                        ____ Input Buffer ____|
+ |                                       |                      |
+ v                                       v                      v
+ |---------------------------------------|-----------|----------|
+ ^                                                   ^          ^
+ |___________________ Output_Size ___________________|_ Margin _|
+
+ NOTE: See also ZSTD_DECOMPRESSION_MARGIN().
+ NOTE: This applies only to single-pass decompression through ZSTD_decompress() or
+ ZSTD_decompressDCtx().
+ NOTE: This function supports multi-frame input.
+
+ @param src The compressed frame(s)
+ @param srcSize The size of the compressed frame(s)
+ @returns The decompression margin or an error that can be checked with ZSTD_isError().
+ 
+</p></pre><BR>
+
+<pre><b>#define ZSTD_DECOMPRESSION_MARGIN(originalSize, blockSize) ((size_t)(                                              \
+        ZSTD_FRAMEHEADERSIZE_MAX                                                              </b>/* Frame header */ + \<b>
+        4                                                                                         </b>/* checksum */ + \<b>
+        ((originalSize) == 0 ? 0 : 3 * (((originalSize) + (blockSize) - 1) / blockSize)) </b>/* 3 bytes per block */ + \<b>
+        (blockSize)                                                                    </b>/* One block of margin */   \<b>
+    ))
+</b><p> Similar to ZSTD_decompressionMargin(), but instead of computing the margin from
+ the compressed frame, compute it from the original size and the blockSizeLog.
+ See ZSTD_decompressionMargin() for details.
+
+ WARNING: This macro does not support multi-frame input, the input must be a single
+ zstd frame. If you need that support use the function, or implement it yourself.
+
+ @param originalSize The original uncompressed size of the data.
+ @param blockSize    The block size == MIN(windowSize, ZSTD_BLOCKSIZE_MAX).
+                     Unless you explicitly set the windowLog smaller than
+                     ZSTD_BLOCKSIZELOG_MAX you can just use ZSTD_BLOCKSIZE_MAX.
+ 
+</p></pre><BR>
+
 <pre><b>typedef enum {
  ZSTD_sf_noBlockDelimiters = 0,         </b>/* Representation of ZSTD_Sequence has no block delimiters, sequences only */<b>
  ZSTD_sf_explicitBlockDelimiters = 1    </b>/* Representation of ZSTD_Sequence contains explicit block delimiters */<b>
 } ZSTD_sequenceFormat_e;
 </b></pre><BR>
+<pre><b>ZSTDLIB_STATIC_API size_t ZSTD_sequenceBound(size_t srcSize);
+</b><p> `srcSize` : size of the input buffer
+  @return : upper-bound for the number of sequences that can be generated
+            from a buffer of srcSize bytes
+
+  note : returns number of sequences - to get bytes, multiply by sizeof(ZSTD_Sequence).
+ 
+</p></pre><BR>
+
 <pre><b></b><p> Generate sequences using ZSTD_compress2(), given a source buffer.

 Each block will end with a dummy sequence
@@ -1260,8 +1409,11 @@ ZSTDLIB_STATIC_API size_t ZSTD_estimateDCtxSize(void);
  and ZSTD_estimateCCtxSize_usingCCtxParams(), which can be used in tandem with ZSTD_CCtxParams_setParameter().
  Both can be used to estimate memory using custom compression parameters and arbitrary srcSize limits.

-  Note 2 : only single-threaded compression is supported.
+  Note : only single-threaded compression is supported.
  ZSTD_estimateCCtxSize_usingCCtxParams() will return an error code if ZSTD_c_nbWorkers is >= 1.
+
+  Note 2 : ZSTD_estimateCCtxSize* functions are not compatible with the Block-Level Sequence Producer API at this time.
+  Size estimates assume that no external sequence producer is registered.
 
 </p></pre><BR>

@@ -1281,7 +1433,12 @@ ZSTDLIB_STATIC_API size_t ZSTD_estimateDStreamSize_fromFrame(const void* src, si
  or deducted from a valid frame Header, using ZSTD_estimateDStreamSize_fromFrame();
  Note : if streaming is init with function ZSTD_init?Stream_usingDict(),
         an internal ?Dict will be created, which additional size is not estimated here.
-         In this case, get total size by adding ZSTD_estimate?DictSize 
+         In this case, get total size by adding ZSTD_estimate?DictSize
+  Note 2 : only single-threaded compression is supported.
+  ZSTD_estimateCStreamSize_usingCCtxParams() will return an error code if ZSTD_c_nbWorkers is >= 1.
+  Note 3 : ZSTD_estimateCStreamSize* functions are not compatible with the Block-Level Sequence Producer API at this time.
+  Size estimates assume that no external sequence producer is registered.
+ 
 </p></pre><BR>

 <pre><b>ZSTDLIB_STATIC_API size_t ZSTD_estimateCDictSize(size_t dictSize, int compressionLevel);
@@ -1381,20 +1538,41 @@ ZSTDLIB_STATIC_API size_t ZSTD_CCtx_refThreadPool(ZSTD_CCtx* cctx, ZSTD_threadPo
  This function never fails (wide contract) 
 </p></pre><BR>

-<pre><b>ZSTDLIB_STATIC_API
-ZSTD_DEPRECATED("use ZSTD_compress2")
+<pre><b>ZSTDLIB_STATIC_API size_t ZSTD_CCtx_setCParams(ZSTD_CCtx* cctx, ZSTD_compressionParameters cparams);
+</b><p>  Set all parameters provided within @p cparams into the working @p cctx.
+  Note : if modifying parameters during compression (MT mode only),
+         note that changes to the .windowLog parameter will be ignored.
+ @return 0 on success, or an error code (can be checked with ZSTD_isError()).
+         On failure, no parameters are updated.
+ 
+</p></pre><BR>
+
+<pre><b>ZSTDLIB_STATIC_API size_t ZSTD_CCtx_setFParams(ZSTD_CCtx* cctx, ZSTD_frameParameters fparams);
+</b><p>  Set all parameters provided within @p fparams into the working @p cctx.
+ @return 0 on success, or an error code (can be checked with ZSTD_isError()).
+ 
+</p></pre><BR>
+
+<pre><b>ZSTDLIB_STATIC_API size_t ZSTD_CCtx_setParams(ZSTD_CCtx* cctx, ZSTD_parameters params);
+</b><p>  Set all parameters provided within @p params into the working @p cctx.
+ @return 0 on success, or an error code (can be checked with ZSTD_isError()).
+ 
+</p></pre><BR>
+
+<pre><b>ZSTD_DEPRECATED("use ZSTD_compress2")
+ZSTDLIB_STATIC_API
 size_t ZSTD_compress_advanced(ZSTD_CCtx* cctx,
-                              void* dst, size_t dstCapacity,
-                        const void* src, size_t srcSize,
-                        const void* dict,size_t dictSize,
-                              ZSTD_parameters params);
+                  void* dst, size_t dstCapacity,
+            const void* src, size_t srcSize,
+            const void* dict,size_t dictSize,
+                  ZSTD_parameters params);
 </b><p>  Note : this function is now DEPRECATED.
         It can be replaced by ZSTD_compress2(), in combination with ZSTD_CCtx_setParameter() and other parameter setters.
  This prototype will generate compilation warnings. 
 </p></pre><BR>

-<pre><b>ZSTDLIB_STATIC_API
-ZSTD_DEPRECATED("use ZSTD_compress2 with ZSTD_CCtx_loadDictionary")
+<pre><b>ZSTD_DEPRECATED("use ZSTD_compress2 with ZSTD_CCtx_loadDictionary")
+ZSTDLIB_STATIC_API
 size_t ZSTD_compress_usingCDict_advanced(ZSTD_CCtx* cctx,
                                  void* dst, size_t dstCapacity,
                            const void* src, size_t srcSize,
@@ -1555,8 +1733,8 @@ ZSTDLIB_STATIC_API size_t ZSTD_freeCCtxParams(ZSTD_CCtx_params* params);  </b>/*
 
 </p></pre><BR>

-<pre><b>ZSTDLIB_STATIC_API
-ZSTD_DEPRECATED("use ZSTD_DCtx_setParameter() instead")
+<pre><b>ZSTD_DEPRECATED("use ZSTD_DCtx_setParameter() instead")
+ZSTDLIB_STATIC_API
 size_t ZSTD_DCtx_setFormat(ZSTD_DCtx* dctx, ZSTD_format_e format);
 </b><p>  This function is REDUNDANT. Prefer ZSTD_DCtx_setParameter().
  Instruct the decoder context about what kind of data to decode next.
@@ -1582,8 +1760,8 @@ size_t ZSTD_DCtx_setFormat(ZSTD_DCtx* dctx, ZSTD_format_e format);
 <BR></pre>

 <h3>Advanced Streaming compression functions</h3><pre></pre><b><pre></pre></b><BR>
-<pre><b>ZSTDLIB_STATIC_API
-ZSTD_DEPRECATED("use ZSTD_CCtx_reset, see zstd.h for detailed instructions")
+<pre><b>ZSTD_DEPRECATED("use ZSTD_CCtx_reset, see zstd.h for detailed instructions")
+ZSTDLIB_STATIC_API
 size_t ZSTD_initCStream_srcSize(ZSTD_CStream* zcs,
             int compressionLevel,
             unsigned long long pledgedSrcSize);
@@ -1600,8 +1778,8 @@ size_t ZSTD_initCStream_srcSize(ZSTD_CStream* zcs,
 
 </p></pre><BR>

-<pre><b>ZSTDLIB_STATIC_API
-ZSTD_DEPRECATED("use ZSTD_CCtx_reset, see zstd.h for detailed instructions")
+<pre><b>ZSTD_DEPRECATED("use ZSTD_CCtx_reset, see zstd.h for detailed instructions")
+ZSTDLIB_STATIC_API
 size_t ZSTD_initCStream_usingDict(ZSTD_CStream* zcs,
         const void* dict, size_t dictSize,
               int compressionLevel);
@@ -1618,18 +1796,15 @@ size_t ZSTD_initCStream_usingDict(ZSTD_CStream* zcs,
 
 </p></pre><BR>

-<pre><b>ZSTDLIB_STATIC_API
-ZSTD_DEPRECATED("use ZSTD_CCtx_reset, see zstd.h for detailed instructions")
+<pre><b>ZSTD_DEPRECATED("use ZSTD_CCtx_reset, see zstd.h for detailed instructions")
+ZSTDLIB_STATIC_API
 size_t ZSTD_initCStream_advanced(ZSTD_CStream* zcs,
        const void* dict, size_t dictSize,
              ZSTD_parameters params,
              unsigned long long pledgedSrcSize);
-</b><p> This function is DEPRECATED, and is approximately equivalent to:
+</b><p> This function is DEPRECATED, and is equivalent to:
     ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only);
-     // Pseudocode: Set each zstd parameter and leave the rest as-is.
-     for ((param, value) : params) {
-         ZSTD_CCtx_setParameter(zcs, param, value);
-     }
+     ZSTD_CCtx_setParams(zcs, params);
     ZSTD_CCtx_setPledgedSrcSize(zcs, pledgedSrcSize);
     ZSTD_CCtx_loadDictionary(zcs, dict, dictSize);

@@ -1640,8 +1815,8 @@ size_t ZSTD_initCStream_advanced(ZSTD_CStream* zcs,
 
 </p></pre><BR>

-<pre><b>ZSTDLIB_STATIC_API
-ZSTD_DEPRECATED("use ZSTD_CCtx_reset and ZSTD_CCtx_refCDict, see zstd.h for detailed instructions")
+<pre><b>ZSTD_DEPRECATED("use ZSTD_CCtx_reset and ZSTD_CCtx_refCDict, see zstd.h for detailed instructions")
+ZSTDLIB_STATIC_API
 size_t ZSTD_initCStream_usingCDict(ZSTD_CStream* zcs, const ZSTD_CDict* cdict);
 </b><p> This function is DEPRECATED, and equivalent to:
     ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only);
@@ -1652,18 +1827,15 @@ size_t ZSTD_initCStream_usingCDict(ZSTD_CStream* zcs, const ZSTD_CDict* cdict);
 
 </p></pre><BR>

-<pre><b>ZSTDLIB_STATIC_API
-ZSTD_DEPRECATED("use ZSTD_CCtx_reset and ZSTD_CCtx_refCDict, see zstd.h for detailed instructions")
+<pre><b>ZSTD_DEPRECATED("use ZSTD_CCtx_reset and ZSTD_CCtx_refCDict, see zstd.h for detailed instructions")
+ZSTDLIB_STATIC_API
 size_t ZSTD_initCStream_usingCDict_advanced(ZSTD_CStream* zcs,
                   const ZSTD_CDict* cdict,
                         ZSTD_frameParameters fParams,
                         unsigned long long pledgedSrcSize);
-</b><p>   This function is DEPRECATED, and is approximately equivalent to:
+</b><p>   This function is DEPRECATED, and is equivalent to:
     ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only);
-     // Pseudocode: Set each zstd frame parameter and leave the rest as-is.
-     for ((fParam, value) : fParams) {
-         ZSTD_CCtx_setParameter(zcs, fParam, value);
-     }
+     ZSTD_CCtx_setFParams(zcs, fParams);
     ZSTD_CCtx_setPledgedSrcSize(zcs, pledgedSrcSize);
     ZSTD_CCtx_refCDict(zcs, cdict);

@@ -1674,8 +1846,8 @@ size_t ZSTD_initCStream_usingCDict_advanced(ZSTD_CStream* zcs,
 
 </p></pre><BR>

-<pre><b>ZSTDLIB_STATIC_API
-ZSTD_DEPRECATED("use ZSTD_CCtx_reset, see zstd.h for detailed instructions")
+<pre><b>ZSTD_DEPRECATED("use ZSTD_CCtx_reset, see zstd.h for detailed instructions")
+ZSTDLIB_STATIC_API
 size_t ZSTD_resetCStream(ZSTD_CStream* zcs, unsigned long long pledgedSrcSize);
 </b><p> This function is DEPRECATED, and is equivalent to:
     ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only);
@@ -1721,39 +1893,71 @@ size_t ZSTD_resetCStream(ZSTD_CStream* zcs, unsigned long long pledgedSrcSize);
 </p></pre><BR>

 <h3>Advanced Streaming decompression functions</h3><pre></pre><b><pre></pre></b><BR>
-<pre><b>ZSTDLIB_STATIC_API size_t ZSTD_initDStream_usingDict(ZSTD_DStream* zds, const void* dict, size_t dictSize);
+<pre><b>ZSTD_DEPRECATED("use ZSTD_DCtx_reset + ZSTD_DCtx_loadDictionary, see zstd.h for detailed instructions")
+ZSTDLIB_STATIC_API size_t ZSTD_initDStream_usingDict(ZSTD_DStream* zds, const void* dict, size_t dictSize);
 </b><p>
     ZSTD_DCtx_reset(zds, ZSTD_reset_session_only);
     ZSTD_DCtx_loadDictionary(zds, dict, dictSize);

 note: no dictionary will be used if dict == NULL or dictSize < 8
- Note : this prototype will be marked as deprecated and generate compilation warnings on reaching v1.5.x
 
 </p></pre><BR>

-<pre><b>ZSTDLIB_STATIC_API size_t ZSTD_initDStream_usingDDict(ZSTD_DStream* zds, const ZSTD_DDict* ddict);
+<pre><b>ZSTD_DEPRECATED("use ZSTD_DCtx_reset + ZSTD_DCtx_refDDict, see zstd.h for detailed instructions")
+ZSTDLIB_STATIC_API size_t ZSTD_initDStream_usingDDict(ZSTD_DStream* zds, const ZSTD_DDict* ddict);
 </b><p>
     ZSTD_DCtx_reset(zds, ZSTD_reset_session_only);
     ZSTD_DCtx_refDDict(zds, ddict);

 note : ddict is referenced, it must outlive decompression session
- Note : this prototype will be marked as deprecated and generate compilation warnings on reaching v1.5.x
 
 </p></pre><BR>

-<pre><b>ZSTDLIB_STATIC_API size_t ZSTD_resetDStream(ZSTD_DStream* zds);
+<pre><b>ZSTD_DEPRECATED("use ZSTD_DCtx_reset, see zstd.h for detailed instructions")
+ZSTDLIB_STATIC_API size_t ZSTD_resetDStream(ZSTD_DStream* zds);
 </b><p>
     ZSTD_DCtx_reset(zds, ZSTD_reset_session_only);

 re-use decompression parameters from previous init; saves dictionary loading
- Note : this prototype will be marked as deprecated and generate compilation warnings on reaching v1.5.x
 
 </p></pre><BR>

-<a name="Chapter20"></a><h2>Buffer-less and synchronous inner streaming functions</h2><pre>
-  This is an advanced API, giving full control over buffer management, for users which need direct control over memory.
-  But it's also a complex one, with several restrictions, documented below.
-  Prefer normal streaming API for an easier experience.
+<pre><b>ZSTDLIB_STATIC_API void
+ZSTD_registerSequenceProducer(
+  ZSTD_CCtx* cctx,
+  void* sequenceProducerState,
+  ZSTD_sequenceProducer_F* sequenceProducer
+);
+</b><p> Instruct zstd to use a block-level external sequence producer function.
+
+ The sequenceProducerState must be initialized by the caller, and the caller is
+ responsible for managing its lifetime. This parameter is sticky across
+ compressions. It will remain set until the user explicitly resets compression
+ parameters.
+
+ Sequence producer registration is considered to be an "advanced parameter",
+ part of the "advanced API". This means it will only have an effect on compression
+ APIs which respect advanced parameters, such as compress2() and compressStream2().
+ Older compression APIs such as compressCCtx(), which predate the introduction of
+ "advanced parameters", will ignore any external sequence producer setting.
+
+ The sequence producer can be "cleared" by registering a NULL function pointer. This
+ removes all limitations described above in the "LIMITATIONS" section of the API docs.
+
+ The user is strongly encouraged to read the full API documentation (above) before
+ calling this function. 
+</p></pre><BR>
+
+<a name="Chapter20"></a><h2>Buffer-less and synchronous inner streaming functions (DEPRECATED)</h2><pre>
+  This API is deprecated, and will be removed in a future version.
+  It allows streaming (de)compression with user allocated buffers.
+  However, it is hard to use, and not as well tested as the rest of
+  our API.
+
+  Please use the normal streaming API instead: ZSTD_compressStream2,
+  and ZSTD_decompressStream.
+  If there is functionality that you need, but it doesn't provide,
+  please open an issue on our GitHub.
 
 <BR></pre>

@@ -1764,7 +1968,6 @@ size_t ZSTD_resetCStream(ZSTD_CStream* zcs, unsigned long long pledgedSrcSize);

  Start by initializing a context.
  Use ZSTD_compressBegin(), or ZSTD_compressBegin_usingDict() for dictionary compression.
-  It's also possible to duplicate a reference context which has already been initialized, using ZSTD_copyCCtx()

  Then, consume your input using ZSTD_compressContinue().
  There are some important considerations to keep in mind when using this advanced function :
@@ -1785,11 +1988,15 @@ size_t ZSTD_resetCStream(ZSTD_CStream* zcs, unsigned long long pledgedSrcSize);
  `ZSTD_CCtx` object can be re-used (ZSTD_compressBegin()) to compress again.
 <BR></pre>

-<h3>Buffer-less streaming compression functions</h3><pre></pre><b><pre>ZSTDLIB_STATIC_API size_t ZSTD_compressBegin(ZSTD_CCtx* cctx, int compressionLevel);
+<h3>Buffer-less streaming compression functions</h3><pre></pre><b><pre>ZSTD_DEPRECATED("The buffer-less API is deprecated in favor of the normal streaming API. See docs.")
+ZSTDLIB_STATIC_API size_t ZSTD_compressBegin(ZSTD_CCtx* cctx, int compressionLevel);
+ZSTD_DEPRECATED("The buffer-less API is deprecated in favor of the normal streaming API. See docs.")
 ZSTDLIB_STATIC_API size_t ZSTD_compressBegin_usingDict(ZSTD_CCtx* cctx, const void* dict, size_t dictSize, int compressionLevel);
+ZSTD_DEPRECATED("The buffer-less API is deprecated in favor of the normal streaming API. See docs.")
 ZSTDLIB_STATIC_API size_t ZSTD_compressBegin_usingCDict(ZSTD_CCtx* cctx, const ZSTD_CDict* cdict); </b>/**< note: fails if cdict==NULL */<b>
-ZSTDLIB_STATIC_API size_t ZSTD_copyCCtx(ZSTD_CCtx* cctx, const ZSTD_CCtx* preparedCCtx, unsigned long long pledgedSrcSize); </b>/**<  note: if pledgedSrcSize is not known, use ZSTD_CONTENTSIZE_UNKNOWN */<b>
 </pre></b><BR>
+<pre><b>size_t ZSTD_copyCCtx(ZSTD_CCtx* cctx, const ZSTD_CCtx* preparedCCtx, unsigned long long pledgedSrcSize); </b>/**<  note: if pledgedSrcSize is not known, use ZSTD_CONTENTSIZE_UNKNOWN */<b>
+</b></pre><BR>
 <pre><b>size_t ZSTD_compressBegin_advanced(ZSTD_CCtx* cctx, const void* dict, size_t dictSize, ZSTD_parameters params, unsigned long long pledgedSrcSize); </b>/**< pledgedSrcSize : If srcSize is not known at init time, use ZSTD_CONTENTSIZE_UNKNOWN */<b>
 </b></pre><BR>
 <a name="Chapter22"></a><h2>Buffer-less streaming decompression (synchronous mode)</h2><pre>
@@ -1801,8 +2008,8 @@ ZSTDLIB_STATIC_API size_t ZSTD_copyCCtx(ZSTD_CCtx* cctx, const ZSTD_CCtx* prepar
  Frame header is extracted from the beginning of compressed frame, so providing only the frame's beginning is enough.
  Data fragment must be large enough to ensure successful decoding.
 `ZSTD_frameHeaderSize_max` bytes is guaranteed to always be large enough.
-  @result : 0 : successful decoding, the `ZSTD_frameHeader` structure is correctly filled.
-           >0 : `srcSize` is too small, please provide at least @result bytes on next attempt.
+  result  : 0 : successful decoding, the `ZSTD_frameHeader` structure is correctly filled.
+           >0 : `srcSize` is too small, please provide at least result bytes on next attempt.
           errorCode, which can be tested using ZSTD_isError().

  It fills a ZSTD_frameHeader structure with important information to correctly decode the frame,
@@ -1821,7 +2028,7 @@ ZSTDLIB_STATIC_API size_t ZSTD_copyCCtx(ZSTD_CCtx* cctx, const ZSTD_CCtx* prepar

  The most memory efficient way is to use a round buffer of sufficient size.
  Sufficient size is determined by invoking ZSTD_decodingBufferSize_min(),
-  which can @return an error code if required value is too large for current system (in 32-bits mode).
+  which can return an error code if required value is too large for current system (in 32-bits mode).
  In a round buffer methodology, ZSTD_decompressContinue() decompresses each block next to previous one,
  up to the moment there is not enough room left in the buffer to guarantee decoding another full block,
  which maximum size is provided in `ZSTD_frameHeader` structure, field `blockSizeMax`.
@@ -1841,7 +2048,7 @@ ZSTDLIB_STATIC_API size_t ZSTD_copyCCtx(ZSTD_CCtx* cctx, const ZSTD_CCtx* prepar
  ZSTD_nextSrcSizeToDecompress() tells how many bytes to provide as 'srcSize' to ZSTD_decompressContinue().
  ZSTD_decompressContinue() requires this _exact_ amount of bytes, or it will fail.

- @result of ZSTD_decompressContinue() is the number of bytes regenerated within 'dst' (necessarily <= dstCapacity).
+  result of ZSTD_decompressContinue() is the number of bytes regenerated within 'dst' (necessarily <= dstCapacity).
  It can be zero : it just means ZSTD_decompressContinue() has decoded some metadata item.
  It can also be an error code, which can be tested with ZSTD_isError().

@@ -1863,34 +2070,25 @@ ZSTDLIB_STATIC_API size_t ZSTD_copyCCtx(ZSTD_CCtx* cctx, const ZSTD_CCtx* prepar
  For skippable frames ZSTD_decompressContinue() always returns 0 : it only skips the content.
 <BR></pre>

-<h3>Buffer-less streaming decompression functions</h3><pre></pre><b><pre>typedef enum { ZSTD_frame, ZSTD_skippableFrame } ZSTD_frameType_e;
-typedef struct {
-    unsigned long long frameContentSize; </b>/* if == ZSTD_CONTENTSIZE_UNKNOWN, it means this field is not available. 0 means "empty" */<b>
-    unsigned long long windowSize;       </b>/* can be very large, up to <= frameContentSize */<b>
-    unsigned blockSizeMax;
-    ZSTD_frameType_e frameType;          </b>/* if == ZSTD_skippableFrame, frameContentSize is the size of skippable content */<b>
-    unsigned headerSize;
-    unsigned dictID;
-    unsigned checksumFlag;
-} ZSTD_frameHeader;
-</pre></b><BR>
-<pre><b>ZSTDLIB_STATIC_API size_t ZSTD_getFrameHeader(ZSTD_frameHeader* zfhPtr, const void* src, size_t srcSize);   </b>/**< doesn't consume input */<b>
-</b>/*! ZSTD_getFrameHeader_advanced() :<b>
- *  same as ZSTD_getFrameHeader(),
- *  with added capability to select a format (like ZSTD_f_zstd1_magicless) */
-ZSTDLIB_STATIC_API size_t ZSTD_getFrameHeader_advanced(ZSTD_frameHeader* zfhPtr, const void* src, size_t srcSize, ZSTD_format_e format);
-ZSTDLIB_STATIC_API size_t ZSTD_decodingBufferSize_min(unsigned long long windowSize, unsigned long long frameContentSize);  </b>/**< when frame content size is not known, pass in frameContentSize == ZSTD_CONTENTSIZE_UNKNOWN */<b>
-</b><p>  decode Frame Header, or requires larger `srcSize`.
- @return : 0, `zfhPtr` is correctly filled,
-          >0, `srcSize` is too small, value is wanted `srcSize` amount,
-           or an error code, which can be tested using ZSTD_isError() 
-</p></pre><BR>
-
+<h3>Buffer-less streaming decompression functions</h3><pre></pre><b><pre></pre></b><BR>
+<pre><b>ZSTDLIB_STATIC_API size_t ZSTD_decodingBufferSize_min(unsigned long long windowSize, unsigned long long frameContentSize);  </b>/**< when frame content size is not known, pass in frameContentSize == ZSTD_CONTENTSIZE_UNKNOWN */<b>
+</b></pre><BR>
 <pre><b>typedef enum { ZSTDnit_frameHeader, ZSTDnit_blockHeader, ZSTDnit_block, ZSTDnit_lastBlock, ZSTDnit_checksum, ZSTDnit_skippableFrame } ZSTD_nextInputType_e;
 </b></pre><BR>
-<a name="Chapter23"></a><h2>Block level API</h2><pre></pre>
+<a name="Chapter23"></a><h2>Block level API (DEPRECATED)</h2><pre></pre>

-<pre><b></b><p>    Frame metadata cost is typically ~12 bytes, which can be non-negligible for very small blocks (< 100 bytes).
+<pre><b></b><p>    You can get the frame header down to 2 bytes by setting:
+      - ZSTD_c_format = ZSTD_f_zstd1_magicless
+      - ZSTD_c_contentSizeFlag = 0
+      - ZSTD_c_checksumFlag = 0
+      - ZSTD_c_dictIDFlag = 0
+
+    This API is not as well tested as our normal API, so we recommend not using it.
+    We will be removing it in a future version. If the normal API doesn't provide
+    the functionality you need, please open a GitHub issue.
+
+    Block functions produce and decode raw zstd blocks, without frame metadata.
+    Frame metadata cost is typically ~12 bytes, which can be non-negligible for very small blocks (< 100 bytes).
    But users will have to take in charge needed metadata to regenerate data, such as compressed and content sizes.

    A few rules to respect :
@@ -1899,7 +2097,6 @@ ZSTDLIB_STATIC_API size_t ZSTD_decodingBufferSize_min(unsigned long long windowS
    - It is necessary to init context before starting
      + compression : any ZSTD_compressBegin*() variant, including with dictionary
      + decompression : any ZSTD_decompressBegin*() variant, including with dictionary
-      + copyCCtx() and copyDCtx() can be used too
    - Block size is limited, it must be <= ZSTD_getBlockSize() <= ZSTD_BLOCKSIZE_MAX == 128 KB
      + If input is larger than a block size, it's necessary to split input data into multiple blocks
      + For inputs larger than a single block, consider using regular ZSTD_compress() instead.
@@ -1915,9 +2112,13 @@ ZSTDLIB_STATIC_API size_t ZSTD_decodingBufferSize_min(unsigned long long windowS
        Use ZSTD_insertBlock() for such a case.
 </p></pre><BR>

-<h3>Raw zstd block functions</h3><pre></pre><b><pre>ZSTDLIB_STATIC_API size_t ZSTD_getBlockSize   (const ZSTD_CCtx* cctx);
+<h3>Raw zstd block functions</h3><pre></pre><b><pre>ZSTD_DEPRECATED("The block API is deprecated in favor of the normal compression API. See docs.")
+ZSTDLIB_STATIC_API size_t ZSTD_getBlockSize   (const ZSTD_CCtx* cctx);
+ZSTD_DEPRECATED("The block API is deprecated in favor of the normal compression API. See docs.")
 ZSTDLIB_STATIC_API size_t ZSTD_compressBlock  (ZSTD_CCtx* cctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize);
+ZSTD_DEPRECATED("The block API is deprecated in favor of the normal compression API. See docs.")
 ZSTDLIB_STATIC_API size_t ZSTD_decompressBlock(ZSTD_DCtx* dctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize);
+ZSTD_DEPRECATED("The block API is deprecated in favor of the normal compression API. See docs.")
 ZSTDLIB_STATIC_API size_t ZSTD_insertBlock    (ZSTD_DCtx* dctx, const void* blockStart, size_t blockSize);  </b>/**< insert uncompressed block into `dctx` history. Useful for multi-blocks decompression. */<b>
 </pre></b><BR>
 </html>
@@ -42,7 +42,13 @@ static void compressFile_orDie(const char* fname, const char* outName, int cLeve
     */
    CHECK_ZSTD( ZSTD_CCtx_setParameter(cctx, ZSTD_c_compressionLevel, cLevel) );
    CHECK_ZSTD( ZSTD_CCtx_setParameter(cctx, ZSTD_c_checksumFlag, 1) );
-    ZSTD_CCtx_setParameter(cctx, ZSTD_c_nbWorkers, nbThreads);
+    if (nbThreads > 1) {
+        size_t const r = ZSTD_CCtx_setParameter(cctx, ZSTD_c_nbWorkers, nbThreads);
+        if (ZSTD_isError(r)) {
+            fprintf (stderr, "Note: the linked libzstd library doesn't support multithreading. "
+                             "Reverting to single-thread mode. \n");
+        }
+    }

    /* This loop read from the input file, compresses that entire chunk,
     * and writes all output produced to the output file.
@@ -117,7 +123,7 @@ int main(int argc, const char** argv)
    }

    int cLevel = 1;
-    int nbThreads = 4;
+    int nbThreads = 1;

    if (argc >= 3) {
      cLevel = atoi (argv[2]);
@@ -8,6 +8,9 @@
 # You may select, at your option, one of the above-listed licenses.
 # ################################################################

+# default target (when runing `make` with no argument)
+lib-release:
+
 # Modules
 ZSTD_LIB_COMPRESSION ?= 1
 ZSTD_LIB_DECOMPRESSION ?= 1
@@ -54,12 +57,11 @@ VERSION := $(ZSTD_VERSION)
 # Note: by default, the static library is built single-threaded and dynamic library is built
 # multi-threaded. It is possible to force multi or single threaded builds by appending
 # -mt or -nomt to the build target (like lib-mt for multi-threaded, lib-nomt for single-threaded).
-.PHONY: default
-default: lib-release
+

 CPPFLAGS_DYNLIB  += -DZSTD_MULTITHREAD # dynamic library build defaults to multi-threaded
 LDFLAGS_DYNLIB   += -pthread
-CPPFLAGS_STATLIB +=                    # static library build defaults to single-threaded
+CPPFLAGS_STATICLIB +=                  # static library build defaults to single-threaded


 ifeq ($(findstring GCC,$(CCVER)),GCC)
@@ -91,7 +93,7 @@ all: lib


 .PHONY: libzstd.a  # must be run every time
-libzstd.a: CPPFLAGS += $(CPPFLAGS_STATLIB)
+libzstd.a: CPPFLAGS += $(CPPFLAGS_STATICLIB)

 SET_CACHE_DIRECTORY = \
   +$(MAKE) --no-print-directory $@ \
@@ -109,19 +111,19 @@ libzstd.a:
 else
 # BUILD_DIR is defined

-ZSTD_STATLIB_DIR := $(BUILD_DIR)/static
-ZSTD_STATLIB := $(ZSTD_STATLIB_DIR)/libzstd.a
-ZSTD_STATLIB_OBJ := $(addprefix $(ZSTD_STATLIB_DIR)/,$(ZSTD_LOCAL_OBJ))
-$(ZSTD_STATLIB): ARFLAGS = rcs
-$(ZSTD_STATLIB): | $(ZSTD_STATLIB_DIR)
-$(ZSTD_STATLIB): $(ZSTD_STATLIB_OBJ)
+ZSTD_STATICLIB_DIR := $(BUILD_DIR)/static
+ZSTD_STATICLIB := $(ZSTD_STATICLIB_DIR)/libzstd.a
+ZSTD_STATICLIB_OBJ := $(addprefix $(ZSTD_STATICLIB_DIR)/,$(ZSTD_LOCAL_OBJ))
+$(ZSTD_STATICLIB): ARFLAGS = rcs
+$(ZSTD_STATICLIB): | $(ZSTD_STATICLIB_DIR)
+$(ZSTD_STATICLIB): $(ZSTD_STATICLIB_OBJ)
  # Check for multithread flag at target execution time
 	$(if $(filter -DZSTD_MULTITHREAD,$(CPPFLAGS)),\
    @echo compiling multi-threaded static library $(LIBVER),\
    @echo compiling single-threaded static library $(LIBVER))
 	$(AR) $(ARFLAGS) $@ $^

-libzstd.a: $(ZSTD_STATLIB)
+libzstd.a: $(ZSTD_STATICLIB)
 	cp -f $< $@

 endif
@@ -182,14 +184,14 @@ lib : libzstd.a libzstd
 # make does not consider implicit pattern rule for .PHONY target

 %-mt : CPPFLAGS_DYNLIB  := -DZSTD_MULTITHREAD
-%-mt : CPPFLAGS_STATLIB := -DZSTD_MULTITHREAD
+%-mt : CPPFLAGS_STATICLIB := -DZSTD_MULTITHREAD
 %-mt : LDFLAGS_DYNLIB   := -pthread
 %-mt : %
 	@echo multi-threaded build completed

 %-nomt : CPPFLAGS_DYNLIB  :=
 %-nomt : LDFLAGS_DYNLIB   :=
-%-nomt : CPPFLAGS_STATLIB :=
+%-nomt : CPPFLAGS_STATICLIB :=
 %-nomt : %
 	@echo single-threaded build completed

@@ -200,42 +202,52 @@ lib : libzstd.a libzstd

 # Generate .h dependencies automatically

-DEPFLAGS = -MT $@ -MMD -MP -MF
+# -MMD: compiler generates dependency information as a side-effect of compilation, without system headers
+# -MP: adds phony target for each dependency other than main file.
+DEPFLAGS = -MMD -MP

-$(ZSTD_DYNLIB_DIR)/%.o : %.c $(ZSTD_DYNLIB_DIR)/%.d | $(ZSTD_DYNLIB_DIR)
+# ensure that ZSTD_DYNLIB_DIR exists prior to generating %.o
+$(ZSTD_DYNLIB_DIR)/%.o : %.c | $(ZSTD_DYNLIB_DIR)
 	@echo CC $@
-	$(COMPILE.c) $(DEPFLAGS) $(ZSTD_DYNLIB_DIR)/$*.d $(OUTPUT_OPTION) $<
+	$(COMPILE.c) $(DEPFLAGS) $(OUTPUT_OPTION) $<

-$(ZSTD_STATLIB_DIR)/%.o : %.c $(ZSTD_STATLIB_DIR)/%.d | $(ZSTD_STATLIB_DIR)
+$(ZSTD_STATICLIB_DIR)/%.o : %.c | $(ZSTD_STATICLIB_DIR)
 	@echo CC $@
-	$(COMPILE.c) $(DEPFLAGS) $(ZSTD_STATLIB_DIR)/$*.d $(OUTPUT_OPTION) $<
+	$(COMPILE.c) $(DEPFLAGS) $(OUTPUT_OPTION) $<

 $(ZSTD_DYNLIB_DIR)/%.o : %.S | $(ZSTD_DYNLIB_DIR)
 	@echo AS $@
 	$(COMPILE.S) $(OUTPUT_OPTION) $<

-$(ZSTD_STATLIB_DIR)/%.o : %.S | $(ZSTD_STATLIB_DIR)
+$(ZSTD_STATICLIB_DIR)/%.o : %.S | $(ZSTD_STATICLIB_DIR)
 	@echo AS $@
 	$(COMPILE.S) $(OUTPUT_OPTION) $<

-MKDIR ?= mkdir
-$(BUILD_DIR) $(ZSTD_DYNLIB_DIR) $(ZSTD_STATLIB_DIR):
-	$(MKDIR) -p $@
+MKDIR ?= mkdir -p
+$(BUILD_DIR) $(ZSTD_DYNLIB_DIR) $(ZSTD_STATICLIB_DIR):
+	$(MKDIR) $@

-DEPFILES := $(ZSTD_DYNLIB_OBJ:.o=.d) $(ZSTD_STATLIB_OBJ:.o=.d)
+DEPFILES := $(ZSTD_DYNLIB_OBJ:.o=.d) $(ZSTD_STATICLIB_OBJ:.o=.d)
 $(DEPFILES):

-include $(wildcard $(DEPFILES))
+# The leading '-' means: do not fail is include fails (ex: directory does not exist yet)
+-include $(wildcard $(DEPFILES))


-# Special case : building library in single-thread mode _and_ without zstdmt_compress.c
-ZSTDMT_FILES = compress/zstdmt_compress.c
-ZSTD_NOMT_FILES = $(filter-out $(ZSTDMT_FILES),$(ZSTD_FILES))
+# Special case : build library in single-thread mode _and_ without zstdmt_compress.c
+# Note : we still need threading.c and pool.c for the dictionary builder,
+# but they will correctly behave single-threaded.
+ZSTDMT_FILES = zstdmt_compress.c
+ZSTD_NOMT_FILES = $(filter-out $(ZSTDMT_FILES),$(notdir $(ZSTD_FILES)))
 libzstd-nomt: CFLAGS += -fPIC -fvisibility=hidden
 libzstd-nomt: LDFLAGS += -shared
 libzstd-nomt: $(ZSTD_NOMT_FILES)
 	@echo compiling single-thread dynamic library $(LIBVER)
 	@echo files : $(ZSTD_NOMT_FILES)
+	@if echo "$(ZSTD_NOMT_FILES)" | tr ' ' '\n' | $(GREP) -q zstdmt; then \
+        echo "Error: Found zstdmt in list."; \
+        exit 1; \
+    fi
 	$(CC) $(FLAGS) $^ $(LDFLAGS) $(SONAME_FLAGS) -o $@

 .PHONY: clean
@@ -249,7 +261,7 @@ clean:
 #-----------------------------------------------------------------------------
 # make install is validated only for below listed environments
 #-----------------------------------------------------------------------------
-ifneq (,$(filter $(UNAME),Linux Darwin GNU/kFreeBSD GNU OpenBSD FreeBSD NetBSD DragonFly SunOS Haiku AIX))
+ifneq (,$(filter $(UNAME),Linux Darwin GNU/kFreeBSD GNU OpenBSD FreeBSD NetBSD DragonFly SunOS Haiku AIX MSYS_NT CYGWIN_NT))

 lib: libzstd.pc

@@ -119,6 +119,15 @@ The file structure is designed to make this selection manually achievable for an
  binary is achieved by using `HUF_FORCE_DECOMPRESS_X1` and
  `ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT` (implied by `ZSTD_LIB_MINIFY`).

+  On the compressor side, Zstd's compression levels map to several internal
+  strategies. In environments where the higher compression levels aren't used,
+  it is possible to exclude all but the fastest strategy with
+  `ZSTD_LIB_EXCLUDE_COMPRESSORS_DFAST_AND_UP=1`. (Note that this will change
+  the behavior of the default compression level.) Or if you want to retain the
+  default compressor as well, you can set
+  `ZSTD_LIB_EXCLUDE_COMPRESSORS_GREEDY_AND_UP=1`, at the cost of an additional
+  ~20KB or so.
+
  For squeezing the last ounce of size out, you can also define
  `ZSTD_NO_INLINE`, which disables inlining, and `ZSTD_STRIP_ERROR_STRINGS`,
  which removes the error messages that are otherwise returned by
@@ -0,0 +1,55 @@
+/*
+ * Copyright (c) Meta Platforms, Inc. and affiliates.
+ * All rights reserved.
+ *
+ * This source code is licensed under both the BSD-style license (found in the
+ * LICENSE file in the root directory of this source tree) and the GPLv2 (found
+ * in the COPYING file in the root directory of this source tree).
+ * You may select, at your option, one of the above-listed licenses.
+ */
+
+/* This file provides custom allocation primitives
+ */
+
+#define ZSTD_DEPS_NEED_MALLOC
+#include "zstd_deps.h"   /* ZSTD_malloc, ZSTD_calloc, ZSTD_free, ZSTD_memset */
+
+#include "compiler.h" /* MEM_STATIC */
+#define ZSTD_STATIC_LINKING_ONLY
+#include "../zstd.h" /* ZSTD_customMem */
+
+#ifndef ZSTD_ALLOCATIONS_H
+#define ZSTD_ALLOCATIONS_H
+
+/* custom memory allocation functions */
+
+MEM_STATIC void* ZSTD_customMalloc(size_t size, ZSTD_customMem customMem)
+{
+    if (customMem.customAlloc)
+        return customMem.customAlloc(customMem.opaque, size);
+    return ZSTD_malloc(size);
+}
+
+MEM_STATIC void* ZSTD_customCalloc(size_t size, ZSTD_customMem customMem)
+{
+    if (customMem.customAlloc) {
+        /* calloc implemented as malloc+memset;
+         * not as efficient as calloc, but next best guess for custom malloc */
+        void* const ptr = customMem.customAlloc(customMem.opaque, size);
+        ZSTD_memset(ptr, 0, size);
+        return ptr;
+    }
+    return ZSTD_calloc(1, size);
+}
+
+MEM_STATIC void ZSTD_customFree(void* ptr, ZSTD_customMem customMem)
+{
+    if (ptr!=NULL) {
+        if (customMem.customFree)
+            customMem.customFree(customMem.opaque, ptr);
+        else
+            ZSTD_free(ptr);
+    }
+}
+
+#endif /* ZSTD_ALLOCATIONS_H */
@@ -17,7 +17,7 @@ MEM_STATIC unsigned ZSTD_countTrailingZeros32_fallback(U32 val)
 {
    assert(val != 0);
    {
-        static const int DeBruijnBytePos[32] = {0, 1, 28, 2, 29, 14, 24, 3,
+        static const U32 DeBruijnBytePos[32] = {0, 1, 28, 2, 29, 14, 24, 3,
                                                30, 22, 20, 15, 25, 17, 4, 8,
                                                31, 27, 13, 23, 21, 19, 16, 7,
                                                26, 12, 18, 6, 11, 5, 10, 9};
@@ -30,7 +30,7 @@ MEM_STATIC unsigned ZSTD_countTrailingZeros32(U32 val)
    assert(val != 0);
 #   if defined(_MSC_VER)
 #       if STATIC_BMI2 == 1
-            return _tzcnt_u32(val);
+            return (unsigned)_tzcnt_u32(val);
 #       else
            if (val != 0) {
                unsigned long r;
@@ -69,7 +69,7 @@ MEM_STATIC unsigned ZSTD_countLeadingZeros32(U32 val)
    assert(val != 0);
 #   if defined(_MSC_VER)
 #       if STATIC_BMI2 == 1
-            return _lzcnt_u32(val);
+            return (unsigned)_lzcnt_u32(val);
 #       else
            if (val != 0) {
                unsigned long r;
@@ -92,7 +92,7 @@ MEM_STATIC unsigned ZSTD_countTrailingZeros64(U64 val)
    assert(val != 0);
 #   if defined(_MSC_VER) && defined(_WIN64)
 #       if STATIC_BMI2 == 1
-            return _tzcnt_u64(val);
+            return (unsigned)_tzcnt_u64(val);
 #       else
            if (val != 0) {
                unsigned long r;
@@ -123,7 +123,7 @@ MEM_STATIC unsigned ZSTD_countLeadingZeros64(U64 val)
    assert(val != 0);
 #   if defined(_MSC_VER) && defined(_WIN64)
 #       if STATIC_BMI2 == 1
-            return _lzcnt_u64(val);
+            return (unsigned)_lzcnt_u64(val);
 #       else
            if (val != 0) {
                unsigned long r;
@@ -172,4 +172,29 @@ MEM_STATIC unsigned ZSTD_highbit32(U32 val)   /* compress, dictBuilder, decodeCo
    return 31 - ZSTD_countLeadingZeros32(val);
 }

+/* ZSTD_rotateRight_*():
+ * Rotates a bitfield to the right by "count" bits.
+ * https://en.wikipedia.org/w/index.php?title=Circular_shift&oldid=991635599#Implementing_circular_shifts
+ */
+MEM_STATIC
+U64 ZSTD_rotateRight_U64(U64 const value, U32 count) {
+    assert(count < 64);
+    count &= 0x3F; /* for fickle pattern recognition */
+    return (value >> count) | (U64)(value << ((0U - count) & 0x3F));
+}
+
+MEM_STATIC
+U32 ZSTD_rotateRight_U32(U32 const value, U32 count) {
+    assert(count < 32);
+    count &= 0x1F; /* for fickle pattern recognition */
+    return (value >> count) | (U32)(value << ((0U - count) & 0x1F));
+}
+
+MEM_STATIC
+U16 ZSTD_rotateRight_U16(U16 const value, U32 count) {
+    assert(count < 16);
+    count &= 0x0F; /* for fickle pattern recognition */
+    return (value >> count) | (U16)(value << ((0U - count) & 0x0F));
+}
+
 #endif /* ZSTD_BITS_H */
@@ -90,19 +90,20 @@ MEM_STATIC size_t BIT_closeCStream(BIT_CStream_t* bitC);
 /*-********************************************
 *  bitStream decoding API (read backward)
 **********************************************/
+typedef size_t BitContainerType;
 typedef struct {
-    size_t   bitContainer;
+    BitContainerType bitContainer;
    unsigned bitsConsumed;
    const char* ptr;
    const char* start;
    const char* limitPtr;
 } BIT_DStream_t;

-typedef enum { BIT_DStream_unfinished = 0,
-               BIT_DStream_endOfBuffer = 1,
-               BIT_DStream_completed = 2,
-               BIT_DStream_overflow = 3 } BIT_DStream_status;  /* result of BIT_reloadDStream() */
-               /* 1,2,4,8 would be better for bitmap combinations, but slows down performance a bit ... :( */
+typedef enum { BIT_DStream_unfinished = 0,  /* fully refilled */
+               BIT_DStream_endOfBuffer = 1, /* still some bits left in bitstream */
+               BIT_DStream_completed = 2,   /* bitstream entirely consumed, bit-exact */
+               BIT_DStream_overflow = 3     /* user requested more bits than present in bitstream */
+    } BIT_DStream_status;  /* result of BIT_reloadDStream() */

 MEM_STATIC size_t   BIT_initDStream(BIT_DStream_t* bitD, const void* srcBuffer, size_t srcSize);
 MEM_STATIC size_t   BIT_readBits(BIT_DStream_t* bitD, unsigned nbBits);
@@ -112,7 +113,7 @@ MEM_STATIC unsigned BIT_endOfDStream(const BIT_DStream_t* bitD);

 /* Start by invoking BIT_initDStream().
 *  A chunk of the bitStream is then stored into a local register.
-*  Local register size is 64-bits on 64-bits systems, 32-bits on 32-bits systems (size_t).
+*  Local register size is 64-bits on 64-bits systems, 32-bits on 32-bits systems (BitContainerType).
 *  You can then retrieve bitFields stored into the local register, **in reverse order**.
 *  Local register is explicitly reloaded from memory by the BIT_reloadDStream() method.
 *  A reload guarantee a minimum of ((8*sizeof(bitD->bitContainer))-7) bits when its result is BIT_DStream_unfinished.
@@ -162,7 +163,7 @@ MEM_STATIC size_t BIT_initCStream(BIT_CStream_t* bitC,
    return 0;
 }

-MEM_STATIC FORCE_INLINE_ATTR size_t BIT_getLowerBits(size_t bitContainer, U32 const nbBits)
+FORCE_INLINE_TEMPLATE size_t BIT_getLowerBits(size_t bitContainer, U32 const nbBits)
 {
 #if defined(STATIC_BMI2) && STATIC_BMI2 == 1 && !defined(ZSTD_NO_INTRINSICS)
    return  _bzhi_u64(bitContainer, nbBits);
@@ -267,22 +268,22 @@ MEM_STATIC size_t BIT_initDStream(BIT_DStream_t* bitD, const void* srcBuffer, si
        bitD->bitContainer = *(const BYTE*)(bitD->start);
        switch(srcSize)
        {
-        case 7: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[6]) << (sizeof(bitD->bitContainer)*8 - 16);
+        case 7: bitD->bitContainer += (BitContainerType)(((const BYTE*)(srcBuffer))[6]) << (sizeof(bitD->bitContainer)*8 - 16);
                ZSTD_FALLTHROUGH;

-        case 6: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[5]) << (sizeof(bitD->bitContainer)*8 - 24);
+        case 6: bitD->bitContainer += (BitContainerType)(((const BYTE*)(srcBuffer))[5]) << (sizeof(bitD->bitContainer)*8 - 24);
                ZSTD_FALLTHROUGH;

-        case 5: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[4]) << (sizeof(bitD->bitContainer)*8 - 32);
+        case 5: bitD->bitContainer += (BitContainerType)(((const BYTE*)(srcBuffer))[4]) << (sizeof(bitD->bitContainer)*8 - 32);
                ZSTD_FALLTHROUGH;

-        case 4: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[3]) << 24;
+        case 4: bitD->bitContainer += (BitContainerType)(((const BYTE*)(srcBuffer))[3]) << 24;
                ZSTD_FALLTHROUGH;

-        case 3: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[2]) << 16;
+        case 3: bitD->bitContainer += (BitContainerType)(((const BYTE*)(srcBuffer))[2]) << 16;
                ZSTD_FALLTHROUGH;

-        case 2: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[1]) <<  8;
+        case 2: bitD->bitContainer += (BitContainerType)(((const BYTE*)(srcBuffer))[1]) <<  8;
                ZSTD_FALLTHROUGH;

        default: break;
@@ -297,12 +298,12 @@ MEM_STATIC size_t BIT_initDStream(BIT_DStream_t* bitD, const void* srcBuffer, si
    return srcSize;
 }

-MEM_STATIC FORCE_INLINE_ATTR size_t BIT_getUpperBits(size_t bitContainer, U32 const start)
+FORCE_INLINE_TEMPLATE size_t BIT_getUpperBits(BitContainerType bitContainer, U32 const start)
 {
    return bitContainer >> start;
 }

-MEM_STATIC FORCE_INLINE_ATTR size_t BIT_getMiddleBits(size_t bitContainer, U32 const start, U32 const nbBits)
+FORCE_INLINE_TEMPLATE size_t BIT_getMiddleBits(BitContainerType bitContainer, U32 const start, U32 const nbBits)
 {
    U32 const regMask = sizeof(bitContainer)*8 - 1;
    /* if start > regMask, bitstream is corrupted, and result is undefined */
@@ -325,7 +326,7 @@ MEM_STATIC FORCE_INLINE_ATTR size_t BIT_getMiddleBits(size_t bitContainer, U32 c
 *  On 32-bits, maxNbBits==24.
 *  On 64-bits, maxNbBits==56.
 * @return : value extracted */
-MEM_STATIC  FORCE_INLINE_ATTR size_t BIT_lookBits(const BIT_DStream_t*  bitD, U32 nbBits)
+FORCE_INLINE_TEMPLATE size_t BIT_lookBits(const BIT_DStream_t*  bitD, U32 nbBits)
 {
    /* arbitrate between double-shift and shift+mask */
 #if 1
@@ -348,7 +349,7 @@ MEM_STATIC size_t BIT_lookBitsFast(const BIT_DStream_t* bitD, U32 nbBits)
    return (bitD->bitContainer << (bitD->bitsConsumed & regMask)) >> (((regMask+1)-nbBits) & regMask);
 }

-MEM_STATIC FORCE_INLINE_ATTR void BIT_skipBits(BIT_DStream_t* bitD, U32 nbBits)
+FORCE_INLINE_TEMPLATE void BIT_skipBits(BIT_DStream_t* bitD, U32 nbBits)
 {
    bitD->bitsConsumed += nbBits;
 }
@@ -357,7 +358,7 @@ MEM_STATIC FORCE_INLINE_ATTR void BIT_skipBits(BIT_DStream_t* bitD, U32 nbBits)
 *  Read (consume) next n bits from local register and update.
 *  Pay attention to not read more than nbBits contained into local register.
 * @return : extracted value. */
-MEM_STATIC FORCE_INLINE_ATTR size_t BIT_readBits(BIT_DStream_t* bitD, unsigned nbBits)
+FORCE_INLINE_TEMPLATE size_t BIT_readBits(BIT_DStream_t* bitD, unsigned nbBits)
 {
    size_t const value = BIT_lookBits(bitD, nbBits);
    BIT_skipBits(bitD, nbBits);
@@ -374,6 +375,21 @@ MEM_STATIC size_t BIT_readBitsFast(BIT_DStream_t* bitD, unsigned nbBits)
    return value;
 }

+/*! BIT_reloadDStream_internal() :
+ *  Simple variant of BIT_reloadDStream(), with two conditions:
+ *  1. bitstream is valid : bitsConsumed <= sizeof(bitD->bitContainer)*8
+ *  2. look window is valid after shifted down : bitD->ptr >= bitD->start
+ */
+MEM_STATIC BIT_DStream_status BIT_reloadDStream_internal(BIT_DStream_t* bitD)
+{
+    assert(bitD->bitsConsumed <= sizeof(bitD->bitContainer)*8);
+    bitD->ptr -= bitD->bitsConsumed >> 3;
+    assert(bitD->ptr >= bitD->start);
+    bitD->bitsConsumed &= 7;
+    bitD->bitContainer = MEM_readLEST(bitD->ptr);
+    return BIT_DStream_unfinished;
+}
+
 /*! BIT_reloadDStreamFast() :
 *  Similar to BIT_reloadDStream(), but with two differences:
 *  1. bitsConsumed <= sizeof(bitD->bitContainer)*8 must hold!
@@ -384,31 +400,35 @@ MEM_STATIC BIT_DStream_status BIT_reloadDStreamFast(BIT_DStream_t* bitD)
 {
    if (UNLIKELY(bitD->ptr < bitD->limitPtr))
        return BIT_DStream_overflow;
-    assert(bitD->bitsConsumed <= sizeof(bitD->bitContainer)*8);
-    bitD->ptr -= bitD->bitsConsumed >> 3;
-    bitD->bitsConsumed &= 7;
-    bitD->bitContainer = MEM_readLEST(bitD->ptr);
-    return BIT_DStream_unfinished;
+    return BIT_reloadDStream_internal(bitD);
 }

 /*! BIT_reloadDStream() :
 *  Refill `bitD` from buffer previously set in BIT_initDStream() .
- *  This function is safe, it guarantees it will not read beyond src buffer.
+ *  This function is safe, it guarantees it will not never beyond src buffer.
 * @return : status of `BIT_DStream_t` internal register.
 *           when status == BIT_DStream_unfinished, internal register is filled with at least 25 or 57 bits */
-MEM_STATIC BIT_DStream_status BIT_reloadDStream(BIT_DStream_t* bitD)
+FORCE_INLINE_TEMPLATE BIT_DStream_status BIT_reloadDStream(BIT_DStream_t* bitD)
 {
-    if (bitD->bitsConsumed > (sizeof(bitD->bitContainer)*8))  /* overflow detected, like end of stream */
+    /* note : once in overflow mode, a bitstream remains in this mode until it's reset */
+    if (UNLIKELY(bitD->bitsConsumed > (sizeof(bitD->bitContainer)*8))) {
+        static const BitContainerType zeroFilled = 0;
+        bitD->ptr = (const char*)&zeroFilled; /* aliasing is allowed for char */
+        /* overflow detected, erroneous scenario or end of stream: no update */
        return BIT_DStream_overflow;
+    }
+
+    assert(bitD->ptr >= bitD->start);

    if (bitD->ptr >= bitD->limitPtr) {
-        return BIT_reloadDStreamFast(bitD);
+        return BIT_reloadDStream_internal(bitD);
    }
    if (bitD->ptr == bitD->start) {
+        /* reached end of bitStream => no update */
        if (bitD->bitsConsumed < sizeof(bitD->bitContainer)*8) return BIT_DStream_endOfBuffer;
        return BIT_DStream_completed;
    }
-    /* start < ptr < limitPtr */
+    /* start < ptr < limitPtr => cautious update */
    {   U32 nbBytes = bitD->bitsConsumed >> 3;
        BIT_DStream_status result = BIT_DStream_unfinished;
        if (bitD->ptr - nbBytes < bitD->start) {
@@ -51,12 +51,19 @@
 #  define WIN_CDECL
 #endif

+/* UNUSED_ATTR tells the compiler it is okay if the function is unused. */
+#if defined(__GNUC__)
+#  define UNUSED_ATTR __attribute__((unused))
+#else
+#  define UNUSED_ATTR
+#endif
+
 /**
 * FORCE_INLINE_TEMPLATE is used to define C "templates", which take constant
 * parameters. They must be inlined for the compiler to eliminate the constant
 * branches.
 */
-#define FORCE_INLINE_TEMPLATE static INLINE_KEYWORD FORCE_INLINE_ATTR
+#define FORCE_INLINE_TEMPLATE static INLINE_KEYWORD FORCE_INLINE_ATTR UNUSED_ATTR
 /**
 * HINT_INLINE is used to help the compiler generate better code. It is *not*
 * used for "templates", so it can be tweaked based on the compilers
@@ -71,14 +78,28 @@
 #if !defined(__clang__) && defined(__GNUC__) && __GNUC__ >= 4 && __GNUC_MINOR__ >= 8 && __GNUC__ < 5
 #  define HINT_INLINE static INLINE_KEYWORD
 #else
-#  define HINT_INLINE static INLINE_KEYWORD FORCE_INLINE_ATTR
+#  define HINT_INLINE FORCE_INLINE_TEMPLATE
 #endif

-/* UNUSED_ATTR tells the compiler it is okay if the function is unused. */
+/* "soft" inline :
+ * The compiler is free to select if it's a good idea to inline or not.
+ * The main objective is to silence compiler warnings
+ * when a defined function in included but not used.
+ *
+ * Note : this macro is prefixed `MEM_` because it used to be provided by `mem.h` unit.
+ * Updating the prefix is probably preferable, but requires a fairly large codemod,
+ * since this name is used everywhere.
+ */
+#ifndef MEM_STATIC  /* already defined in Linux Kernel mem.h */
 #if defined(__GNUC__)
-#  define UNUSED_ATTR __attribute__((unused))
+#  define MEM_STATIC static __inline UNUSED_ATTR
+#elif defined (__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */)
+#  define MEM_STATIC static inline
+#elif defined(_MSC_VER)
+#  define MEM_STATIC static __inline
 #else
-#  define UNUSED_ATTR
+#  define MEM_STATIC static  /* this version may generate warnings for unused static functions; disable the relevant warning */
+#endif
 #endif

 /* force no inlining */
@@ -112,7 +133,7 @@
 #  define PREFETCH_L1(ptr)  (void)(ptr)  /* disabled */
 #  define PREFETCH_L2(ptr)  (void)(ptr)  /* disabled */
 #else
-#  if defined(_MSC_VER) && (defined(_M_X64) || defined(_M_I86))  /* _mm_prefetch() is not defined outside of x86/x64 */
+#  if defined(_MSC_VER) && (defined(_M_X64) || defined(_M_I86)) && !defined(_M_ARM64EC)  /* _mm_prefetch() is not defined outside of x86/x64 */
 #    include <mmintrin.h>   /* https://msdn.microsoft.com/fr-fr/library/84szxsww(v=vs.90).aspx */
 #    define PREFETCH_L1(ptr)  _mm_prefetch((const char*)(ptr), _MM_HINT_T0)
 #    define PREFETCH_L2(ptr)  _mm_prefetch((const char*)(ptr), _MM_HINT_T1)
@@ -281,7 +302,18 @@
 *  Sanitizer
 *****************************************************************/

-#if ZSTD_MEMORY_SANITIZER
+/* Issue #3240 reports an ASAN failure on an llvm-mingw build. Out of an
+ * abundance of caution, disable our custom poisoning on mingw. */
+#ifdef __MINGW32__
+#ifndef ZSTD_ASAN_DONT_POISON_WORKSPACE
+#define ZSTD_ASAN_DONT_POISON_WORKSPACE 1
+#endif
+#ifndef ZSTD_MSAN_DONT_POISON_WORKSPACE
+#define ZSTD_MSAN_DONT_POISON_WORKSPACE 1
+#endif
+#endif
+
+#if ZSTD_MEMORY_SANITIZER && !defined(ZSTD_MSAN_DONT_POISON_WORKSPACE)
 /* Not all platforms that support msan provide sanitizers/msan_interface.h.
 * We therefore declare the functions we need ourselves, rather than trying to
 * include the header file... */
@@ -300,9 +332,13 @@ void __msan_poison(const volatile void *a, size_t size);
 /* Returns the offset of the first (at least partially) poisoned byte in the
   memory range, or -1 if the whole range is good. */
 intptr_t __msan_test_shadow(const volatile void *x, size_t size);
+
+/* Print shadow and origin for the memory range to stderr in a human-readable
+   format. */
+void __msan_print_shadow(const volatile void *x, size_t size);
 #endif

-#if ZSTD_ADDRESS_SANITIZER
+#if ZSTD_ADDRESS_SANITIZER && !defined(ZSTD_ASAN_DONT_POISON_WORKSPACE)
 /* Not all platforms that support asan provide sanitizers/asan_interface.h.
 * We therefore declare the functions we need ourselves, rather than trying to
 * include the header file... */
@@ -19,7 +19,6 @@
 #include "error_private.h"       /* ERR_*, ERROR */
 #define FSE_STATIC_LINKING_ONLY  /* FSE_MIN_TABLELOG */
 #include "fse.h"
-#define HUF_STATIC_LINKING_ONLY  /* HUF_TABLELOG_ABSOLUTEMAX */
 #include "huf.h"
 #include "bits.h"                /* ZSDT_highbit32, ZSTD_countTrailingZeros32 */

@@ -237,7 +236,7 @@ size_t HUF_readStats(BYTE* huffWeight, size_t hwSize, U32* rankStats,
                     const void* src, size_t srcSize)
 {
    U32 wksp[HUF_READ_STATS_WORKSPACE_SIZE_U32];
-    return HUF_readStats_wksp(huffWeight, hwSize, rankStats, nbSymbolsPtr, tableLogPtr, src, srcSize, wksp, sizeof(wksp), /* bmi2 */ 0);
+    return HUF_readStats_wksp(huffWeight, hwSize, rankStats, nbSymbolsPtr, tableLogPtr, src, srcSize, wksp, sizeof(wksp), /* flags */ 0);
 }

 FORCE_INLINE_TEMPLATE size_t
@@ -329,13 +328,13 @@ size_t HUF_readStats_wksp(BYTE* huffWeight, size_t hwSize, U32* rankStats,
                     U32* nbSymbolsPtr, U32* tableLogPtr,
                     const void* src, size_t srcSize,
                     void* workSpace, size_t wkspSize,
-                     int bmi2)
+                     int flags)
 {
 #if DYNAMIC_BMI2
-    if (bmi2) {
+    if (flags & HUF_flags_bmi2) {
        return HUF_readStats_body_bmi2(huffWeight, hwSize, rankStats, nbSymbolsPtr, tableLogPtr, src, srcSize, workSpace, wkspSize);
    }
 #endif
-    (void)bmi2;
+    (void)flags;
    return HUF_readStats_body_default(huffWeight, hwSize, rankStats, nbSymbolsPtr, tableLogPtr, src, srcSize, workSpace, wkspSize);
 }
@@ -47,12 +47,15 @@ const char* ERR_getErrorString(ERR_enum code)
    case PREFIX(dstSize_tooSmall): return "Destination buffer is too small";
    case PREFIX(srcSize_wrong): return "Src size is incorrect";
    case PREFIX(dstBuffer_null): return "Operation on NULL destination buffer";
+    case PREFIX(noForwardProgress_destFull): return "Operation made no progress over multiple calls, due to output buffer being full";
+    case PREFIX(noForwardProgress_inputEmpty): return "Operation made no progress over multiple calls, due to input being empty";
        /* following error codes are not stable and may be removed or changed in a future version */
    case PREFIX(frameIndex_tooLarge): return "Frame index is too large";
    case PREFIX(seekableIO): return "An I/O error occurred when reading/seeking";
    case PREFIX(dstBuffer_wrong): return "Destination buffer is wrong";
    case PREFIX(srcBuffer_wrong): return "Source buffer is wrong";
-    case PREFIX(externalMatchFinder_failed): return "External matchfinder returned an error code";
+    case PREFIX(sequenceProducer_failed): return "Block-level external sequence producer returned an error code";
+    case PREFIX(externalSequences_invalid): return "External sequences are not valid";
    case PREFIX(maxCode):
    default: return notErrorCode;
    }
@@ -53,34 +53,6 @@ extern "C" {
 FSE_PUBLIC_API unsigned FSE_versionNumber(void);   /**< library version number; to be used when checking dll version */


-/*-****************************************
-*  FSE simple functions
-******************************************/
-/*! FSE_compress() :
-    Compress content of buffer 'src', of size 'srcSize', into destination buffer 'dst'.
-    'dst' buffer must be already allocated. Compression runs faster is dstCapacity >= FSE_compressBound(srcSize).
-    @return : size of compressed data (<= dstCapacity).
-    Special values : if return == 0, srcData is not compressible => Nothing is stored within dst !!!
-                     if return == 1, srcData is a single byte symbol * srcSize times. Use RLE compression instead.
-                     if FSE_isError(return), compression failed (more details using FSE_getErrorName())
-*/
-FSE_PUBLIC_API size_t FSE_compress(void* dst, size_t dstCapacity,
-                             const void* src, size_t srcSize);
-
-/*! FSE_decompress():
-    Decompress FSE data from buffer 'cSrc', of size 'cSrcSize',
-    into already allocated destination buffer 'dst', of size 'dstCapacity'.
-    @return : size of regenerated data (<= maxDstSize),
-              or an error code, which can be tested using FSE_isError() .
-
-    ** Important ** : FSE_decompress() does not decompress non-compressible nor RLE data !!!
-    Why ? : making this distinction requires a header.
-    Header management is intentionally delegated to the user layer, which can better manage special cases.
-*/
-FSE_PUBLIC_API size_t FSE_decompress(void* dst,  size_t dstCapacity,
-                               const void* cSrc, size_t cSrcSize);
-
-
 /*-*****************************************
 *  Tool functions
 ******************************************/
@@ -91,20 +63,6 @@ FSE_PUBLIC_API unsigned    FSE_isError(size_t code);        /* tells if a return
 FSE_PUBLIC_API const char* FSE_getErrorName(size_t code);   /* provides error code string (useful for debugging) */


-/*-*****************************************
-*  FSE advanced functions
-******************************************/
-/*! FSE_compress2() :
-    Same as FSE_compress(), but allows the selection of 'maxSymbolValue' and 'tableLog'
-    Both parameters can be defined as '0' to mean : use default value
-    @return : size of compressed data
-    Special values : if return == 0, srcData is not compressible => Nothing is stored within cSrc !!!
-                     if return == 1, srcData is a single byte symbol * srcSize times. Use RLE compression.
-                     if FSE_isError(return), it's an error code.
-*/
-FSE_PUBLIC_API size_t FSE_compress2 (void* dst, size_t dstSize, const void* src, size_t srcSize, unsigned maxSymbolValue, unsigned tableLog);
-
-
 /*-*****************************************
 *  FSE detailed API
 ******************************************/
@@ -164,8 +122,6 @@ FSE_PUBLIC_API size_t FSE_writeNCount (void* buffer, size_t bufferSize,
 /*! Constructor and Destructor of FSE_CTable.
    Note that FSE_CTable size depends on 'tableLog' and 'maxSymbolValue' */
 typedef unsigned FSE_CTable;   /* don't allocate that. It's only meant to be more restrictive than void* */
-FSE_PUBLIC_API FSE_CTable* FSE_createCTable (unsigned maxSymbolValue, unsigned tableLog);
-FSE_PUBLIC_API void        FSE_freeCTable (FSE_CTable* ct);

 /*! FSE_buildCTable():
    Builds `ct`, which must be already allocated, using FSE_createCTable().
@@ -241,23 +197,7 @@ FSE_PUBLIC_API size_t FSE_readNCount_bmi2(short* normalizedCounter,
                           unsigned* maxSymbolValuePtr, unsigned* tableLogPtr,
                           const void* rBuffer, size_t rBuffSize, int bmi2);

-/*! Constructor and Destructor of FSE_DTable.
-    Note that its size depends on 'tableLog' */
 typedef unsigned FSE_DTable;   /* don't allocate that. It's just a way to be more restrictive than void* */
-FSE_PUBLIC_API FSE_DTable* FSE_createDTable(unsigned tableLog);
-FSE_PUBLIC_API void        FSE_freeDTable(FSE_DTable* dt);
-
-/*! FSE_buildDTable():
-    Builds 'dt', which must be already allocated, using FSE_createDTable().
-    return : 0, or an errorCode, which can be tested using FSE_isError() */
-FSE_PUBLIC_API size_t FSE_buildDTable (FSE_DTable* dt, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog);
-
-/*! FSE_decompress_usingDTable():
-    Decompress compressed source `cSrc` of size `cSrcSize` using `dt`
-    into `dst` which must be already allocated.
-    @return : size of regenerated data (necessarily <= `dstCapacity`),
-              or an errorCode, which can be tested using FSE_isError() */
-FSE_PUBLIC_API size_t FSE_decompress_usingDTable(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, const FSE_DTable* dt);

 /*!
 Tutorial :
@@ -320,16 +260,6 @@ If there is an error, the function will return an error code, which can be teste
 unsigned FSE_optimalTableLog_internal(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue, unsigned minus);
 /**< same as FSE_optimalTableLog(), which used `minus==2` */

-/* FSE_compress_wksp() :
- * Same as FSE_compress2(), but using an externally allocated scratch buffer (`workSpace`).
- * FSE_COMPRESS_WKSP_SIZE_U32() provides the minimum size required for `workSpace` as a table of FSE_CTable.
- */
-#define FSE_COMPRESS_WKSP_SIZE_U32(maxTableLog, maxSymbolValue)   ( FSE_CTABLE_SIZE_U32(maxTableLog, maxSymbolValue) + ((maxTableLog > 12) ? (1 << (maxTableLog - 2)) : 1024) )
-size_t FSE_compress_wksp (void* dst, size_t dstSize, const void* src, size_t srcSize, unsigned maxSymbolValue, unsigned tableLog, void* workSpace, size_t wkspSize);
-
-size_t FSE_buildCTable_raw (FSE_CTable* ct, unsigned nbBits);
-/**< build a fake FSE_CTable, designed for a flat distribution, where each symbol uses nbBits */
-
 size_t FSE_buildCTable_rle (FSE_CTable* ct, unsigned char symbolValue);
 /**< build a fake FSE_CTable, designed to compress always the same symbolValue */

@@ -347,19 +277,11 @@ size_t FSE_buildCTable_wksp(FSE_CTable* ct, const short* normalizedCounter, unsi
 FSE_PUBLIC_API size_t FSE_buildDTable_wksp(FSE_DTable* dt, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog, void* workSpace, size_t wkspSize);
 /**< Same as FSE_buildDTable(), using an externally allocated `workspace` produced with `FSE_BUILD_DTABLE_WKSP_SIZE_U32(maxSymbolValue)` */

-size_t FSE_buildDTable_raw (FSE_DTable* dt, unsigned nbBits);
-/**< build a fake FSE_DTable, designed to read a flat distribution where each symbol uses nbBits */
-
-size_t FSE_buildDTable_rle (FSE_DTable* dt, unsigned char symbolValue);
-/**< build a fake FSE_DTable, designed to always generate the same symbolValue */
-
 #define FSE_DECOMPRESS_WKSP_SIZE_U32(maxTableLog, maxSymbolValue) (FSE_DTABLE_SIZE_U32(maxTableLog) + 1 + FSE_BUILD_DTABLE_WKSP_SIZE_U32(maxTableLog, maxSymbolValue) + (FSE_MAX_SYMBOL_VALUE + 1) / 2 + 1)
 #define FSE_DECOMPRESS_WKSP_SIZE(maxTableLog, maxSymbolValue) (FSE_DECOMPRESS_WKSP_SIZE_U32(maxTableLog, maxSymbolValue) * sizeof(unsigned))
-size_t FSE_decompress_wksp(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, unsigned maxLog, void* workSpace, size_t wkspSize);
-/**< same as FSE_decompress(), using an externally allocated `workSpace` produced with `FSE_DECOMPRESS_WKSP_SIZE_U32(maxLog, maxSymbolValue)` */
-
 size_t FSE_decompress_wksp_bmi2(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, unsigned maxLog, void* workSpace, size_t wkspSize, int bmi2);
-/**< Same as FSE_decompress_wksp() but with dynamic BMI2 support. Pass 1 if your CPU supports BMI2 or 0 if it doesn't. */
+/**< same as FSE_decompress(), using an externally allocated `workSpace` produced with `FSE_DECOMPRESS_WKSP_SIZE_U32(maxLog, maxSymbolValue)`.
+ * Set bmi2 to 1 if your CPU supports BMI2 or 0 if it doesn't */

 typedef enum {
   FSE_repeat_none,  /**< Cannot use the previous table */
@@ -56,19 +56,6 @@
 #define FSE_FUNCTION_NAME(X,Y) FSE_CAT(X,Y)
 #define FSE_TYPE_NAME(X,Y) FSE_CAT(X,Y)

-
-/* Function templates */
-FSE_DTable* FSE_createDTable (unsigned tableLog)
-{
-    if (tableLog > FSE_TABLELOG_ABSOLUTE_MAX) tableLog = FSE_TABLELOG_ABSOLUTE_MAX;
-    return (FSE_DTable*)ZSTD_malloc( FSE_DTABLE_SIZE_U32(tableLog) * sizeof (U32) );
-}
-
-void FSE_freeDTable (FSE_DTable* dt)
-{
-    ZSTD_free(dt);
-}
-
 static size_t FSE_buildDTable_internal(FSE_DTable* dt, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog, void* workSpace, size_t wkspSize)
 {
    void* const tdPtr = dt+1;   /* because *dt is unsigned, 32-bits aligned on 32-bits */
@@ -185,49 +172,6 @@ size_t FSE_buildDTable_wksp(FSE_DTable* dt, const short* normalizedCounter, unsi
 /*-*******************************************************
 *  Decompression (Byte symbols)
 *********************************************************/
-size_t FSE_buildDTable_rle (FSE_DTable* dt, BYTE symbolValue)
-{
-    void* ptr = dt;
-    FSE_DTableHeader* const DTableH = (FSE_DTableHeader*)ptr;
-    void* dPtr = dt + 1;
-    FSE_decode_t* const cell = (FSE_decode_t*)dPtr;
-
-    DTableH->tableLog = 0;
-    DTableH->fastMode = 0;
-
-    cell->newState = 0;
-    cell->symbol = symbolValue;
-    cell->nbBits = 0;
-
-    return 0;
-}
-
-
-size_t FSE_buildDTable_raw (FSE_DTable* dt, unsigned nbBits)
-{
-    void* ptr = dt;
-    FSE_DTableHeader* const DTableH = (FSE_DTableHeader*)ptr;
-    void* dPtr = dt + 1;
-    FSE_decode_t* const dinfo = (FSE_decode_t*)dPtr;
-    const unsigned tableSize = 1 << nbBits;
-    const unsigned tableMask = tableSize - 1;
-    const unsigned maxSV1 = tableMask+1;
-    unsigned s;
-
-    /* Sanity checks */
-    if (nbBits < 1) return ERROR(GENERIC);         /* min size */
-
-    /* Build Decoding Table */
-    DTableH->tableLog = (U16)nbBits;
-    DTableH->fastMode = 1;
-    for (s=0; s<maxSV1; s++) {
-        dinfo[s].newState = 0;
-        dinfo[s].symbol = (BYTE)s;
-        dinfo[s].nbBits = (BYTE)nbBits;
-    }
-
-    return 0;
-}

 FORCE_INLINE_TEMPLATE size_t FSE_decompress_usingDTable_generic(
          void* dst, size_t maxDstSize,
@@ -291,26 +235,6 @@ FORCE_INLINE_TEMPLATE size_t FSE_decompress_usingDTable_generic(
    return op-ostart;
 }

-
-size_t FSE_decompress_usingDTable(void* dst, size_t originalSize,
-                            const void* cSrc, size_t cSrcSize,
-                            const FSE_DTable* dt)
-{
-    const void* ptr = dt;
-    const FSE_DTableHeader* DTableH = (const FSE_DTableHeader*)ptr;
-    const U32 fastMode = DTableH->fastMode;
-
-    /* select fast mode (static) */
-    if (fastMode) return FSE_decompress_usingDTable_generic(dst, originalSize, cSrc, cSrcSize, dt, 1);
-    return FSE_decompress_usingDTable_generic(dst, originalSize, cSrc, cSrcSize, dt, 0);
-}
-
-
-size_t FSE_decompress_wksp(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, unsigned maxLog, void* workSpace, size_t wkspSize)
-{
-    return FSE_decompress_wksp_bmi2(dst, dstCapacity, cSrc, cSrcSize, maxLog, workSpace, wkspSize, /* bmi2 */ 0);
-}
-
 typedef struct {
    short ncount[FSE_MAX_SYMBOL_VALUE + 1];
    FSE_DTable dtable[1]; /* Dynamically sized */
@@ -384,22 +308,4 @@ size_t FSE_decompress_wksp_bmi2(void* dst, size_t dstCapacity, const void* cSrc,
    return FSE_decompress_wksp_body_default(dst, dstCapacity, cSrc, cSrcSize, maxLog, workSpace, wkspSize);
 }

-
-typedef FSE_DTable DTable_max_t[FSE_DTABLE_SIZE_U32(FSE_MAX_TABLELOG)];
-
-#ifndef ZSTD_NO_UNUSED_FUNCTIONS
-size_t FSE_buildDTable(FSE_DTable* dt, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog) {
-    U32 wksp[FSE_BUILD_DTABLE_WKSP_SIZE_U32(FSE_TABLELOG_ABSOLUTE_MAX, FSE_MAX_SYMBOL_VALUE)];
-    return FSE_buildDTable_wksp(dt, normalizedCounter, maxSymbolValue, tableLog, wksp, sizeof(wksp));
-}
-
-size_t FSE_decompress(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize)
-{
-    /* Static analyzer seems unable to understand this table will be properly initialized later */
-    U32 wksp[FSE_DECOMPRESS_WKSP_SIZE_U32(FSE_MAX_TABLELOG, FSE_MAX_SYMBOL_VALUE)];
-    return FSE_decompress_wksp(dst, dstCapacity, cSrc, cSrcSize, FSE_MAX_TABLELOG, wksp, sizeof(wksp));
-}
-#endif
-
-
 #endif   /* FSE_COMMONDEFS_ONLY */
@@ -21,102 +21,23 @@ extern "C" {

 /* *** Dependencies *** */
 #include "zstd_deps.h"    /* size_t */
-
-
-/* *** library symbols visibility *** */
-/* Note : when linking with -fvisibility=hidden on gcc, or by default on Visual,
- *        HUF symbols remain "private" (internal symbols for library only).
- *        Set macro FSE_DLL_EXPORT to 1 if you want HUF symbols visible on DLL interface */
-#if defined(FSE_DLL_EXPORT) && (FSE_DLL_EXPORT==1) && defined(__GNUC__) && (__GNUC__ >= 4)
-#  define HUF_PUBLIC_API __attribute__ ((visibility ("default")))
-#elif defined(FSE_DLL_EXPORT) && (FSE_DLL_EXPORT==1)   /* Visual expected */
-#  define HUF_PUBLIC_API __declspec(dllexport)
-#elif defined(FSE_DLL_IMPORT) && (FSE_DLL_IMPORT==1)
-#  define HUF_PUBLIC_API __declspec(dllimport)  /* not required, just to generate faster code (saves a function pointer load from IAT and an indirect jump) */
-#else
-#  define HUF_PUBLIC_API
-#endif
-
-
-/* ========================== */
-/* ***  simple functions  *** */
-/* ========================== */
-
-/** HUF_compress() :
- *  Compress content from buffer 'src', of size 'srcSize', into buffer 'dst'.
- * 'dst' buffer must be already allocated.
- *  Compression runs faster if `dstCapacity` >= HUF_compressBound(srcSize).
- * `srcSize` must be <= `HUF_BLOCKSIZE_MAX` == 128 KB.
- * @return : size of compressed data (<= `dstCapacity`).
- *  Special values : if return == 0, srcData is not compressible => Nothing is stored within dst !!!
- *                   if HUF_isError(return), compression failed (more details using HUF_getErrorName())
- */
-HUF_PUBLIC_API size_t HUF_compress(void* dst, size_t dstCapacity,
-                             const void* src, size_t srcSize);
-
-/** HUF_decompress() :
- *  Decompress HUF data from buffer 'cSrc', of size 'cSrcSize',
- *  into already allocated buffer 'dst', of minimum size 'dstSize'.
- * `originalSize` : **must** be the ***exact*** size of original (uncompressed) data.
- *  Note : in contrast with FSE, HUF_decompress can regenerate
- *         RLE (cSrcSize==1) and uncompressed (cSrcSize==dstSize) data,
- *         because it knows size to regenerate (originalSize).
- * @return : size of regenerated data (== originalSize),
- *           or an error code, which can be tested using HUF_isError()
- */
-HUF_PUBLIC_API size_t HUF_decompress(void* dst,  size_t originalSize,
-                               const void* cSrc, size_t cSrcSize);
-
-
-/* ***   Tool functions *** */
-#define HUF_BLOCKSIZE_MAX (128 * 1024)                  /**< maximum input size for a single block compressed with HUF_compress */
-HUF_PUBLIC_API size_t HUF_compressBound(size_t size);   /**< maximum compressed size (worst case) */
-
-/* Error Management */
-HUF_PUBLIC_API unsigned    HUF_isError(size_t code);       /**< tells if a return value is an error code */
-HUF_PUBLIC_API const char* HUF_getErrorName(size_t code);  /**< provides error code string (useful for debugging) */
-
-
-/* ***   Advanced function   *** */
-
-/** HUF_compress2() :
- *  Same as HUF_compress(), but offers control over `maxSymbolValue` and `tableLog`.
- * `maxSymbolValue` must be <= HUF_SYMBOLVALUE_MAX .
- * `tableLog` must be `<= HUF_TABLELOG_MAX` . */
-HUF_PUBLIC_API size_t HUF_compress2 (void* dst, size_t dstCapacity,
-                               const void* src, size_t srcSize,
-                               unsigned maxSymbolValue, unsigned tableLog);
-
-/** HUF_compress4X_wksp() :
- *  Same as HUF_compress2(), but uses externally allocated @workSpace.
- * @workSpace's size, aka @wkspSize, must be >= HUF_WORKSPACE_SIZE
- * @srcSize must be >= 6
- */
-#define HUF_WORKSPACE_SIZE ((8 << 10) + 512 /* sorting scratch space */)
-#define HUF_WORKSPACE_SIZE_U64 (HUF_WORKSPACE_SIZE / sizeof(U64))
-HUF_PUBLIC_API size_t HUF_compress4X_wksp (void* dst, size_t dstCapacity,
-                                     const void* src, size_t srcSize,
-                                     unsigned maxSymbolValue, unsigned tableLog,
-                                     void* workSpace, size_t wkspSize);
-
-#endif   /* HUF_H_298734234 */
-
-/* ******************************************************************
- *  WARNING !!
- *  The following section contains advanced and experimental definitions
- *  which shall never be used in the context of a dynamic library,
- *  because they are not guaranteed to remain stable in the future.
- *  Only consider them in association with static linking.
- * *****************************************************************/
-#if defined(HUF_STATIC_LINKING_ONLY) && !defined(HUF_H_HUF_STATIC_LINKING_ONLY)
-#define HUF_H_HUF_STATIC_LINKING_ONLY
-
-/* *** Dependencies *** */
-#include "mem.h"   /* U32 */
+#include "mem.h"          /* U32 */
 #define FSE_STATIC_LINKING_ONLY
 #include "fse.h"


+/* ***   Tool functions *** */
+#define HUF_BLOCKSIZE_MAX (128 * 1024)   /**< maximum input size for a single block compressed with HUF_compress */
+size_t HUF_compressBound(size_t size);   /**< maximum compressed size (worst case) */
+
+/* Error Management */
+unsigned    HUF_isError(size_t code);       /**< tells if a return value is an error code */
+const char* HUF_getErrorName(size_t code);  /**< provides error code string (useful for debugging) */
+
+
+#define HUF_WORKSPACE_SIZE ((8 << 10) + 512 /* sorting scratch space */)
+#define HUF_WORKSPACE_SIZE_U64 (HUF_WORKSPACE_SIZE / sizeof(U64))
+
 /* *** Constants *** */
 #define HUF_TABLELOG_MAX      12      /* max runtime value of tableLog (due to static allocation); can be modified up to HUF_TABLELOG_ABSOLUTEMAX */
 #define HUF_TABLELOG_DEFAULT  11      /* default tableLog value when none specified */
@@ -156,30 +77,49 @@ typedef U32 HUF_DTable;
 /* ****************************************
 *  Advanced decompression functions
 ******************************************/
-size_t HUF_decompress4X1 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize);   /**< single-symbol decoder */
-#ifndef HUF_FORCE_DECOMPRESS_X1
-size_t HUF_decompress4X2 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize);   /**< double-symbols decoder */
-#endif

-size_t HUF_decompress4X_DCtx (HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize);   /**< decodes RLE and uncompressed */
-size_t HUF_decompress4X_hufOnly(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< considers RLE and uncompressed as errors */
-size_t HUF_decompress4X_hufOnly_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize); /**< considers RLE and uncompressed as errors */
-size_t HUF_decompress4X1_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize);   /**< single-symbol decoder */
-size_t HUF_decompress4X1_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize);   /**< single-symbol decoder */
-#ifndef HUF_FORCE_DECOMPRESS_X1
-size_t HUF_decompress4X2_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize);   /**< double-symbols decoder */
-size_t HUF_decompress4X2_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize);   /**< double-symbols decoder */
-#endif
+/**
+ * Huffman flags bitset.
+ * For all flags, 0 is the default value.
+ */
+typedef enum {
+    /**
+     * If compiled with DYNAMIC_BMI2: Set flag only if the CPU supports BMI2 at runtime.
+     * Otherwise: Ignored.
+     */
+    HUF_flags_bmi2 = (1 << 0),
+    /**
+     * If set: Test possible table depths to find the one that produces the smallest header + encoded size.
+     * If unset: Use heuristic to find the table depth.
+     */
+    HUF_flags_optimalDepth = (1 << 1),
+    /**
+     * If set: If the previous table can encode the input, always reuse the previous table.
+     * If unset: If the previous table can encode the input, reuse the previous table if it results in a smaller output.
+     */
+    HUF_flags_preferRepeat = (1 << 2),
+    /**
+     * If set: Sample the input and check if the sample is uncompressible, if it is then don't attempt to compress.
+     * If unset: Always histogram the entire input.
+     */
+    HUF_flags_suspectUncompressible = (1 << 3),
+    /**
+     * If set: Don't use assembly implementations
+     * If unset: Allow using assembly implementations
+     */
+    HUF_flags_disableAsm = (1 << 4),
+    /**
+     * If set: Don't use the fast decoding loop, always use the fallback decoding loop.
+     * If unset: Use the fast decoding loop when possible.
+     */
+    HUF_flags_disableFast = (1 << 5)
+} HUF_flags_e;


 /* ****************************************
 *  HUF detailed API
 * ****************************************/
 #define HUF_OPTIMAL_DEPTH_THRESHOLD ZSTD_btultra
-typedef enum {
-   HUF_depth_fast, /** Use heuristic to find the table depth**/
-   HUF_depth_optimal /** Test possible table depths to find the one that produces the smallest header + encoded size**/
- } HUF_depth_mode;

 /*! HUF_compress() does the following:
 *  1. count symbol occurrence from source[] into table count[] using FSE_count() (exposed within "fse.h")
@@ -195,12 +135,9 @@ typedef enum {
 unsigned HUF_minTableLog(unsigned symbolCardinality);
 unsigned HUF_cardinality(const unsigned* count, unsigned maxSymbolValue);
 unsigned HUF_optimalTableLog(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue, void* workSpace,
- size_t wkspSize, HUF_CElt* table, const unsigned* count, HUF_depth_mode depthMode); /* table is used as scratch space for building and testing tables, not a return value */
-size_t HUF_buildCTable (HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue, unsigned maxNbBits);   /* @return : maxNbBits; CTable and count can overlap. In which case, CTable will overwrite count content */
-size_t HUF_writeCTable (void* dst, size_t maxDstSize, const HUF_CElt* CTable, unsigned maxSymbolValue, unsigned huffLog);
+ size_t wkspSize, HUF_CElt* table, const unsigned* count, int flags); /* table is used as scratch space for building and testing tables, not a return value */
 size_t HUF_writeCTable_wksp(void* dst, size_t maxDstSize, const HUF_CElt* CTable, unsigned maxSymbolValue, unsigned huffLog, void* workspace, size_t workspaceSize);
-size_t HUF_compress4X_usingCTable(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable);
-size_t HUF_compress4X_usingCTable_bmi2(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable, int bmi2);
+size_t HUF_compress4X_usingCTable(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable, int flags);
 size_t HUF_estimateCompressedSize(const HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue);
 int HUF_validateCTable(const HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue);

@@ -220,14 +157,13 @@ size_t HUF_compress4X_repeat(void* dst, size_t dstSize,
                       const void* src, size_t srcSize,
                       unsigned maxSymbolValue, unsigned tableLog,
                       void* workSpace, size_t wkspSize,    /**< `workSpace` must be aligned on 4-bytes boundaries, `wkspSize` must be >= HUF_WORKSPACE_SIZE */
-                       HUF_CElt* hufTable, HUF_repeat* repeat, int preferRepeat, int bmi2,
-                       unsigned suspectUncompressible, HUF_depth_mode depthMode);
+                       HUF_CElt* hufTable, HUF_repeat* repeat, int flags);

 /** HUF_buildCTable_wksp() :
 *  Same as HUF_buildCTable(), but using externally allocated scratch buffer.
 * `workSpace` must be aligned on 4-bytes boundaries, and its size must be >= HUF_CTABLE_WORKSPACE_SIZE.
 */
-#define HUF_CTABLE_WORKSPACE_SIZE_U32 (2*HUF_SYMBOLVALUE_MAX +1 +1)
+#define HUF_CTABLE_WORKSPACE_SIZE_U32 ((4 * (HUF_SYMBOLVALUE_MAX + 1)) + 192)
 #define HUF_CTABLE_WORKSPACE_SIZE (HUF_CTABLE_WORKSPACE_SIZE_U32 * sizeof(unsigned))
 size_t HUF_buildCTable_wksp (HUF_CElt* tree,
                       const unsigned* count, U32 maxSymbolValue, U32 maxNbBits,
@@ -253,7 +189,7 @@ size_t HUF_readStats_wksp(BYTE* huffWeight, size_t hwSize,
                          U32* rankStats, U32* nbSymbolsPtr, U32* tableLogPtr,
                          const void* src, size_t srcSize,
                          void* workspace, size_t wkspSize,
-                          int bmi2);
+                          int flags);

 /** HUF_readCTable() :
 *  Loading a CTable saved with HUF_writeCTable() */
@@ -261,9 +197,22 @@ size_t HUF_readCTable (HUF_CElt* CTable, unsigned* maxSymbolValuePtr, const void

 /** HUF_getNbBitsFromCTable() :
 *  Read nbBits from CTable symbolTable, for symbol `symbolValue` presumed <= HUF_SYMBOLVALUE_MAX
- *  Note 1 : is not inlined, as HUF_CElt definition is private */
+ *  Note 1 : If symbolValue > HUF_readCTableHeader(symbolTable).maxSymbolValue, returns 0
+ *  Note 2 : is not inlined, as HUF_CElt definition is private
+ */
 U32 HUF_getNbBitsFromCTable(const HUF_CElt* symbolTable, U32 symbolValue);

+typedef struct {
+    BYTE tableLog;
+    BYTE maxSymbolValue;
+    BYTE unused[sizeof(size_t) - 2];
+} HUF_CTableHeader;
+
+/** HUF_readCTableHeader() :
+ *  @returns The header from the CTable specifying the tableLog and the maxSymbolValue.
+ */
+HUF_CTableHeader HUF_readCTableHeader(HUF_CElt const* ctable);
+
 /*
 * HUF_decompress() does the following:
 * 1. select the decompression algorithm (X1, X2) based on pre-computed heuristics
@@ -291,32 +240,12 @@ U32 HUF_selectDecoder (size_t dstSize, size_t cSrcSize);
 #define HUF_DECOMPRESS_WORKSPACE_SIZE ((2 << 10) + (1 << 9))
 #define HUF_DECOMPRESS_WORKSPACE_SIZE_U32 (HUF_DECOMPRESS_WORKSPACE_SIZE / sizeof(U32))

-#ifndef HUF_FORCE_DECOMPRESS_X2
-size_t HUF_readDTableX1 (HUF_DTable* DTable, const void* src, size_t srcSize);
-size_t HUF_readDTableX1_wksp (HUF_DTable* DTable, const void* src, size_t srcSize, void* workSpace, size_t wkspSize);
-#endif
-#ifndef HUF_FORCE_DECOMPRESS_X1
-size_t HUF_readDTableX2 (HUF_DTable* DTable, const void* src, size_t srcSize);
-size_t HUF_readDTableX2_wksp (HUF_DTable* DTable, const void* src, size_t srcSize, void* workSpace, size_t wkspSize);
-#endif
-
-size_t HUF_decompress4X_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable);
-#ifndef HUF_FORCE_DECOMPRESS_X2
-size_t HUF_decompress4X1_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable);
-#endif
-#ifndef HUF_FORCE_DECOMPRESS_X1
-size_t HUF_decompress4X2_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable);
-#endif
-

 /* ====================== */
 /* single stream variants */
 /* ====================== */

-size_t HUF_compress1X (void* dst, size_t dstSize, const void* src, size_t srcSize, unsigned maxSymbolValue, unsigned tableLog);
-size_t HUF_compress1X_wksp (void* dst, size_t dstSize, const void* src, size_t srcSize, unsigned maxSymbolValue, unsigned tableLog, void* workSpace, size_t wkspSize);  /**< `workSpace` must be a table of at least HUF_WORKSPACE_SIZE_U64 U64 */
-size_t HUF_compress1X_usingCTable(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable);
-size_t HUF_compress1X_usingCTable_bmi2(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable, int bmi2);
+size_t HUF_compress1X_usingCTable(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable, int flags);
 /** HUF_compress1X_repeat() :
 *  Same as HUF_compress1X_wksp(), but considers using hufTable if *repeat != HUF_repeat_none.
 *  If it uses hufTable it does not modify hufTable or repeat.
@@ -327,50 +256,30 @@ size_t HUF_compress1X_repeat(void* dst, size_t dstSize,
                       const void* src, size_t srcSize,
                       unsigned maxSymbolValue, unsigned tableLog,
                       void* workSpace, size_t wkspSize,   /**< `workSpace` must be aligned on 4-bytes boundaries, `wkspSize` must be >= HUF_WORKSPACE_SIZE */
-                       HUF_CElt* hufTable, HUF_repeat* repeat, int preferRepeat, int bmi2,
-                       unsigned suspectUncompressible, HUF_depth_mode depthMode);
+                       HUF_CElt* hufTable, HUF_repeat* repeat, int flags);

-size_t HUF_decompress1X1 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize);   /* single-symbol decoder */
+size_t HUF_decompress1X_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int flags);
 #ifndef HUF_FORCE_DECOMPRESS_X1
-size_t HUF_decompress1X2 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize);   /* double-symbol decoder */
-#endif
-
-size_t HUF_decompress1X_DCtx (HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize);
-size_t HUF_decompress1X_DCtx_wksp (HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize);
-#ifndef HUF_FORCE_DECOMPRESS_X2
-size_t HUF_decompress1X1_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize);   /**< single-symbol decoder */
-size_t HUF_decompress1X1_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize);   /**< single-symbol decoder */
-#endif
-#ifndef HUF_FORCE_DECOMPRESS_X1
-size_t HUF_decompress1X2_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize);   /**< double-symbols decoder */
-size_t HUF_decompress1X2_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize);   /**< double-symbols decoder */
-#endif
-
-size_t HUF_decompress1X_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable);   /**< automatic selection of sing or double symbol decoder, based on DTable */
-#ifndef HUF_FORCE_DECOMPRESS_X2
-size_t HUF_decompress1X1_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable);
-#endif
-#ifndef HUF_FORCE_DECOMPRESS_X1
-size_t HUF_decompress1X2_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable);
+size_t HUF_decompress1X2_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int flags);   /**< double-symbols decoder */
 #endif

 /* BMI2 variants.
 * If the CPU has BMI2 support, pass bmi2=1, otherwise pass bmi2=0.
 */
-size_t HUF_decompress1X_usingDTable_bmi2(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable, int bmi2);
+size_t HUF_decompress1X_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable, int flags);
 #ifndef HUF_FORCE_DECOMPRESS_X2
-size_t HUF_decompress1X1_DCtx_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int bmi2);
+size_t HUF_decompress1X1_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int flags);
 #endif
-size_t HUF_decompress4X_usingDTable_bmi2(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable, int bmi2);
-size_t HUF_decompress4X_hufOnly_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int bmi2);
+size_t HUF_decompress4X_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable, int flags);
+size_t HUF_decompress4X_hufOnly_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int flags);
 #ifndef HUF_FORCE_DECOMPRESS_X2
-size_t HUF_readDTableX1_wksp_bmi2(HUF_DTable* DTable, const void* src, size_t srcSize, void* workSpace, size_t wkspSize, int bmi2);
+size_t HUF_readDTableX1_wksp(HUF_DTable* DTable, const void* src, size_t srcSize, void* workSpace, size_t wkspSize, int flags);
 #endif
 #ifndef HUF_FORCE_DECOMPRESS_X1
-size_t HUF_readDTableX2_wksp_bmi2(HUF_DTable* DTable, const void* src, size_t srcSize, void* workSpace, size_t wkspSize, int bmi2);
+size_t HUF_readDTableX2_wksp(HUF_DTable* DTable, const void* src, size_t srcSize, void* workSpace, size_t wkspSize, int flags);
 #endif

-#endif /* HUF_STATIC_LINKING_ONLY */
+#endif   /* HUF_H_298734234 */

 #if defined (__cplusplus)
 }
@@ -31,15 +31,6 @@ extern "C" {
 #   include <stdlib.h>  /* _byteswap_ulong */
 #   include <intrin.h>  /* _byteswap_* */
 #endif
-#if defined(__GNUC__)
-#  define MEM_STATIC static __inline __attribute__((unused))
-#elif defined (__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */)
-#  define MEM_STATIC static inline
-#elif defined(_MSC_VER)
-#  define MEM_STATIC static __inline
-#else
-#  define MEM_STATIC static  /* this version may generate warnings for unused static functions; disable the relevant warning */
-#endif

 /*-**************************************************************
 *  Basic Types
@@ -10,9 +10,9 @@


 /* ======   Dependencies   ======= */
+#include "../common/allocations.h"  /* ZSTD_customCalloc, ZSTD_customFree */
 #include "zstd_deps.h" /* size_t */
 #include "debug.h"     /* assert */
-#include "zstd_internal.h"  /* ZSTD_customCalloc, ZSTD_customFree */
 #include "pool.h"

 /* ======   Compiler specifics   ====== */
@@ -271,7 +271,9 @@ static int isQueueFull(POOL_ctx const* ctx) {
 static void
 POOL_add_internal(POOL_ctx* ctx, POOL_function function, void *opaque)
 {
-    POOL_job const job = {function, opaque};
+    POOL_job job;
+    job.function = function;
+    job.opaque = opaque;
    assert(ctx != NULL);
    if (ctx->shutdown) return;

@@ -68,6 +68,8 @@
 /* Mark the internal assembly functions as hidden  */
 #ifdef __ELF__
 # define ZSTD_HIDE_ASM_FUNCTION(func) .hidden func
+#elif defined(__APPLE__)
+# define ZSTD_HIDE_ASM_FUNCTION(func) .private_extern func
 #else
 # define ZSTD_HIDE_ASM_FUNCTION(func)
 #endif
@@ -137,12 +139,20 @@
 /*
 * For x86 ELF targets, add .note.gnu.property section for Intel CET in
 * assembly sources when CET is enabled.
+ *
+ * Additionally, any function that may be called indirectly must begin
+ * with ZSTD_CET_ENDBRANCH.
 */
 #if defined(__ELF__) && (defined(__x86_64__) || defined(__i386__)) \
    && defined(__has_include)
 # if __has_include(<cet.h>)
 #  include <cet.h>
+#  define ZSTD_CET_ENDBRANCH _CET_ENDBR
 # endif
 #endif

+#ifndef ZSTD_CET_ENDBRANCH
+# define ZSTD_CET_ENDBRANCH
+#endif
+
 #endif /* ZSTD_PORTABILITY_MACROS_H */
@@ -47,7 +47,7 @@ static unsigned __stdcall worker(void *arg)
    void* (*start_routine)(void*);
    void* thread_arg;

-    /* Inialized thread_arg and start_routine and signal main thread that we don't need it
+    /* Initialized thread_arg and start_routine and signal main thread that we don't need it
     * to wait any longer.
     */
    {
@@ -73,10 +73,12 @@ int ZSTD_pthread_create(ZSTD_pthread_t* thread, const void* unused,
    ZSTD_thread_params_t thread_param;
    (void)unused;

+    if (thread==NULL) return -1;
+    *thread = NULL;
+
    thread_param.start_routine = start_routine;
    thread_param.arg = arg;
    thread_param.initialized = 0;
-    *thread = NULL;

    /* Setup thread initialization synchronization */
    if(ZSTD_pthread_cond_init(&thread_param.initialized_cond, NULL)) {
@@ -91,7 +93,7 @@ int ZSTD_pthread_create(ZSTD_pthread_t* thread, const void* unused,

    /* Spawn thread */
    *thread = (HANDLE)_beginthreadex(NULL, 0, worker, &thread_param, 0, NULL);
-    if (!thread) {
+    if (*thread==NULL) {
        ZSTD_pthread_mutex_destroy(&thread_param.initialized_mutex);
        ZSTD_pthread_cond_destroy(&thread_param.initialized_cond);
        return errno;
@@ -137,6 +139,7 @@ int ZSTD_pthread_join(ZSTD_pthread_t thread)

 int ZSTD_pthread_mutex_init(ZSTD_pthread_mutex_t* mutex, pthread_mutexattr_t const* attr)
 {
+    assert(mutex != NULL);
    *mutex = (pthread_mutex_t*)ZSTD_malloc(sizeof(pthread_mutex_t));
    if (!*mutex)
        return 1;
@@ -145,6 +148,7 @@ int ZSTD_pthread_mutex_init(ZSTD_pthread_mutex_t* mutex, pthread_mutexattr_t con

 int ZSTD_pthread_mutex_destroy(ZSTD_pthread_mutex_t* mutex)
 {
+    assert(mutex != NULL);
    if (!*mutex)
        return 0;
    {
@@ -156,6 +160,7 @@ int ZSTD_pthread_mutex_destroy(ZSTD_pthread_mutex_t* mutex)

 int ZSTD_pthread_cond_init(ZSTD_pthread_cond_t* cond, pthread_condattr_t const* attr)
 {
+    assert(cond != NULL);
    *cond = (pthread_cond_t*)ZSTD_malloc(sizeof(pthread_cond_t));
    if (!*cond)
        return 1;
@@ -164,6 +169,7 @@ int ZSTD_pthread_cond_init(ZSTD_pthread_cond_t* cond, pthread_condattr_t const*

 int ZSTD_pthread_cond_destroy(ZSTD_pthread_cond_t* cond)
 {
+    assert(cond != NULL);
    if (!*cond)
        return 0;
    {
@@ -3166,7 +3166,7 @@ XXH_FORCE_INLINE xxh_u64x2 XXH_vec_mule(xxh_u32x4 a, xxh_u32x4 b)
 #if defined(XXH_NO_PREFETCH)
 #  define XXH_PREFETCH(ptr)  (void)(ptr)  /* disabled */
 #else
-#  if defined(_MSC_VER) && (defined(_M_X64) || defined(_M_IX86))  /* _mm_prefetch() not defined outside of x86/x64 */
+#  if defined(_MSC_VER) && (defined(_M_X64) || defined(_M_IX86)) && !defined(_M_ARM64EC)  /* _mm_prefetch() not defined outside of x86/x64 */
 #    include <mmintrin.h>   /* https://msdn.microsoft.com/fr-fr/library/84szxsww(v=vs.90).aspx */
 #    define XXH_PREFETCH(ptr)  _mm_prefetch((const char*)(ptr), _MM_HINT_T0)
 #  elif defined(__GNUC__) && ( (__GNUC__ >= 4) || ( (__GNUC__ == 3) && (__GNUC_MINOR__ >= 1) ) )
@@ -14,7 +14,6 @@
 *  Dependencies
 ***************************************/
 #define ZSTD_DEPS_NEED_MALLOC
-#include "zstd_deps.h"   /* ZSTD_malloc, ZSTD_calloc, ZSTD_free, ZSTD_memset */
 #include "error_private.h"
 #include "zstd_internal.h"

@@ -47,37 +46,3 @@ ZSTD_ErrorCode ZSTD_getErrorCode(size_t code) { return ERR_getErrorCode(code); }
 /*! ZSTD_getErrorString() :
 *  provides error code string from enum */
 const char* ZSTD_getErrorString(ZSTD_ErrorCode code) { return ERR_getErrorString(code); }
-
-
-
-/*=**************************************************************
-*  Custom allocator
-****************************************************************/
-void* ZSTD_customMalloc(size_t size, ZSTD_customMem customMem)
-{
-    if (customMem.customAlloc)
-        return customMem.customAlloc(customMem.opaque, size);
-    return ZSTD_malloc(size);
-}
-
-void* ZSTD_customCalloc(size_t size, ZSTD_customMem customMem)
-{
-    if (customMem.customAlloc) {
-        /* calloc implemented as malloc+memset;
-         * not as efficient as calloc, but next best guess for custom malloc */
-        void* const ptr = customMem.customAlloc(customMem.opaque, size);
-        ZSTD_memset(ptr, 0, size);
-        return ptr;
-    }
-    return ZSTD_calloc(1, size);
-}
-
-void ZSTD_customFree(void* ptr, ZSTD_customMem customMem)
-{
-    if (ptr!=NULL) {
-        if (customMem.customFree)
-            customMem.customFree(customMem.opaque, ptr);
-        else
-            ZSTD_free(ptr);
-    }
-}
@@ -28,7 +28,6 @@
 #include "../zstd.h"
 #define FSE_STATIC_LINKING_ONLY
 #include "fse.h"
-#define HUF_STATIC_LINKING_ONLY
 #include "huf.h"
 #ifndef XXH_STATIC_LINKING_ONLY
 #  define XXH_STATIC_LINKING_ONLY  /* XXH64_state_t */
@@ -114,6 +113,8 @@ typedef enum { set_basic, set_rle, set_compressed, set_repeat } symbolEncodingTy
 #define LLFSELog    9
 #define OffFSELog   8
 #define MaxFSELog  MAX(MAX(MLFSELog, LLFSELog), OffFSELog)
+#define MaxMLBits 16
+#define MaxLLBits 16

 #define ZSTD_MAX_HUF_HEADER_SIZE 128 /* header + <= 127 byte tree description */
 /* Each table cannot take more than #symbols * FSELog bits */
@@ -341,17 +342,13 @@ MEM_STATIC ZSTD_sequenceLength ZSTD_getSequenceLength(seqStore_t const* seqStore
 *          `decompressedBound != ZSTD_CONTENTSIZE_ERROR`
 */
 typedef struct {
+    size_t nbBlocks;
    size_t compressedSize;
    unsigned long long decompressedBound;
 } ZSTD_frameSizeInfo;   /* decompress & legacy */

 const seqStore_t* ZSTD_getSeqStore(const ZSTD_CCtx* ctx);   /* compress & dictBuilder */
-void ZSTD_seqToCodes(const seqStore_t* seqStorePtr);   /* compress, dictBuilder, decodeCorpus (shouldn't get its definition from here) */
-
-/* custom memory allocation functions */
-void* ZSTD_customMalloc(size_t size, ZSTD_customMem customMem);
-void* ZSTD_customCalloc(size_t size, ZSTD_customMem customMem);
-void ZSTD_customFree(void* ptr, ZSTD_customMem customMem);
+int ZSTD_seqToCodes(const seqStore_t* seqStorePtr);   /* compress, dictBuilder, decodeCorpus (shouldn't get its definition from here) */


 /* ZSTD_invalidateRepCodes() :
@@ -369,13 +366,13 @@ typedef struct {

 /*! ZSTD_getcBlockSize() :
 *  Provides the size of compressed block from block header `src` */
-/* Used by: decompress, fullbench (does not get its definition from here) */
+/*  Used by: decompress, fullbench */
 size_t ZSTD_getcBlockSize(const void* src, size_t srcSize,
                          blockProperties_t* bpPtr);

 /*! ZSTD_decodeSeqHeaders() :
 *  decode sequence header from src */
-/* Used by: decompress, fullbench (does not get its definition from here) */
+/*  Used by: zstd_decompress_block, fullbench */
 size_t ZSTD_decodeSeqHeaders(ZSTD_DCtx* dctx, int* nbSeqPtr,
                       const void* src, size_t srcSize);

@@ -343,16 +343,6 @@ size_t FSE_writeNCount (void* buffer, size_t bufferSize,
 *  FSE Compression Code
 ****************************************************************/

-FSE_CTable* FSE_createCTable (unsigned maxSymbolValue, unsigned tableLog)
-{
-    size_t size;
-    if (tableLog > FSE_TABLELOG_ABSOLUTE_MAX) tableLog = FSE_TABLELOG_ABSOLUTE_MAX;
-    size = FSE_CTABLE_SIZE_U32 (tableLog, maxSymbolValue) * sizeof(U32);
-    return (FSE_CTable*)ZSTD_malloc(size);
-}
-
-void FSE_freeCTable (FSE_CTable* ct) { ZSTD_free(ct); }
-
 /* provides the minimum logSize to safely represent a distribution */
 static unsigned FSE_minTableLog(size_t srcSize, unsigned maxSymbolValue)
 {
@@ -533,40 +523,6 @@ size_t FSE_normalizeCount (short* normalizedCounter, unsigned tableLog,
    return tableLog;
 }

-
-/* fake FSE_CTable, for raw (uncompressed) input */
-size_t FSE_buildCTable_raw (FSE_CTable* ct, unsigned nbBits)
-{
-    const unsigned tableSize = 1 << nbBits;
-    const unsigned tableMask = tableSize - 1;
-    const unsigned maxSymbolValue = tableMask;
-    void* const ptr = ct;
-    U16* const tableU16 = ( (U16*) ptr) + 2;
-    void* const FSCT = ((U32*)ptr) + 1 /* header */ + (tableSize>>1);   /* assumption : tableLog >= 1 */
-    FSE_symbolCompressionTransform* const symbolTT = (FSE_symbolCompressionTransform*) (FSCT);
-    unsigned s;
-
-    /* Sanity checks */
-    if (nbBits < 1) return ERROR(GENERIC);             /* min size */
-
-    /* header */
-    tableU16[-2] = (U16) nbBits;
-    tableU16[-1] = (U16) maxSymbolValue;
-
-    /* Build table */
-    for (s=0; s<tableSize; s++)
-        tableU16[s] = (U16)(tableSize + s);
-
-    /* Build Symbol Transformation Table */
-    {   const U32 deltaNbBits = (nbBits << 16) - (1 << nbBits);
-        for (s=0; s<=maxSymbolValue; s++) {
-            symbolTT[s].deltaNbBits = deltaNbBits;
-            symbolTT[s].deltaFindState = s-1;
-    }   }
-
-    return 0;
-}
-
 /* fake FSE_CTable, for rle input (always same symbol) */
 size_t FSE_buildCTable_rle (FSE_CTable* ct, BYTE symbolValue)
 {
@@ -665,78 +621,4 @@ size_t FSE_compress_usingCTable (void* dst, size_t dstSize,

 size_t FSE_compressBound(size_t size) { return FSE_COMPRESSBOUND(size); }

-#ifndef ZSTD_NO_UNUSED_FUNCTIONS
-/* FSE_compress_wksp() :
- * Same as FSE_compress2(), but using an externally allocated scratch buffer (`workSpace`).
- * `wkspSize` size must be `(1<<tableLog)`.
- */
-size_t FSE_compress_wksp (void* dst, size_t dstSize, const void* src, size_t srcSize, unsigned maxSymbolValue, unsigned tableLog, void* workSpace, size_t wkspSize)
-{
-    BYTE* const ostart = (BYTE*) dst;
-    BYTE* op = ostart;
-    BYTE* const oend = ostart + dstSize;
-
-    unsigned count[FSE_MAX_SYMBOL_VALUE+1];
-    S16   norm[FSE_MAX_SYMBOL_VALUE+1];
-    FSE_CTable* CTable = (FSE_CTable*)workSpace;
-    size_t const CTableSize = FSE_CTABLE_SIZE_U32(tableLog, maxSymbolValue);
-    void* scratchBuffer = (void*)(CTable + CTableSize);
-    size_t const scratchBufferSize = wkspSize - (CTableSize * sizeof(FSE_CTable));
-
-    /* init conditions */
-    if (wkspSize < FSE_COMPRESS_WKSP_SIZE_U32(tableLog, maxSymbolValue)) return ERROR(tableLog_tooLarge);
-    if (srcSize <= 1) return 0;  /* Not compressible */
-    if (!maxSymbolValue) maxSymbolValue = FSE_MAX_SYMBOL_VALUE;
-    if (!tableLog) tableLog = FSE_DEFAULT_TABLELOG;
-
-    /* Scan input and build symbol stats */
-    {   CHECK_V_F(maxCount, HIST_count_wksp(count, &maxSymbolValue, src, srcSize, scratchBuffer, scratchBufferSize) );
-        if (maxCount == srcSize) return 1;   /* only a single symbol in src : rle */
-        if (maxCount == 1) return 0;         /* each symbol present maximum once => not compressible */
-        if (maxCount < (srcSize >> 7)) return 0;   /* Heuristic : not compressible enough */
-    }
-
-    tableLog = FSE_optimalTableLog(tableLog, srcSize, maxSymbolValue);
-    CHECK_F( FSE_normalizeCount(norm, tableLog, count, srcSize, maxSymbolValue, /* useLowProbCount */ srcSize >= 2048) );
-
-    /* Write table description header */
-    {   CHECK_V_F(nc_err, FSE_writeNCount(op, oend-op, norm, maxSymbolValue, tableLog) );
-        op += nc_err;
-    }
-
-    /* Compress */
-    CHECK_F( FSE_buildCTable_wksp(CTable, norm, maxSymbolValue, tableLog, scratchBuffer, scratchBufferSize) );
-    {   CHECK_V_F(cSize, FSE_compress_usingCTable(op, oend - op, src, srcSize, CTable) );
-        if (cSize == 0) return 0;   /* not enough space for compressed data */
-        op += cSize;
-    }
-
-    /* check compressibility */
-    if ( (size_t)(op-ostart) >= srcSize-1 ) return 0;
-
-    return op-ostart;
-}
-
-typedef struct {
-    FSE_CTable CTable_max[FSE_CTABLE_SIZE_U32(FSE_MAX_TABLELOG, FSE_MAX_SYMBOL_VALUE)];
-    union {
-      U32 hist_wksp[HIST_WKSP_SIZE_U32];
-      BYTE scratchBuffer[1 << FSE_MAX_TABLELOG];
-    } workspace;
-} fseWkspMax_t;
-
-size_t FSE_compress2 (void* dst, size_t dstCapacity, const void* src, size_t srcSize, unsigned maxSymbolValue, unsigned tableLog)
-{
-    fseWkspMax_t scratchBuffer;
-    DEBUG_STATIC_ASSERT(sizeof(scratchBuffer) >= FSE_COMPRESS_WKSP_SIZE_U32(FSE_MAX_TABLELOG, FSE_MAX_SYMBOL_VALUE));   /* compilation failures here means scratchBuffer is not large enough */
-    if (tableLog > FSE_MAX_TABLELOG) return ERROR(tableLog_tooLarge);
-    return FSE_compress_wksp(dst, dstCapacity, src, srcSize, maxSymbolValue, tableLog, &scratchBuffer, sizeof(scratchBuffer));
-}
-
-size_t FSE_compress (void* dst, size_t dstCapacity, const void* src, size_t srcSize)
-{
-    return FSE_compress2(dst, dstCapacity, src, srcSize, FSE_MAX_SYMBOL_VALUE, FSE_DEFAULT_TABLELOG);
-}
-#endif
-
 #endif   /* FSE_COMMONDEFS_ONLY */
@@ -29,7 +29,6 @@
 #include "hist.h"
 #define FSE_STATIC_LINKING_ONLY   /* FSE_optimalTableLog_internal */
 #include "../common/fse.h"        /* header compression */
-#define HUF_STATIC_LINKING_ONLY
 #include "../common/huf.h"
 #include "../common/error_private.h"
 #include "../common/bits.h"       /* ZSTD_highbit32 */
@@ -221,6 +220,25 @@ static void HUF_setValue(HUF_CElt* elt, size_t value)
    }
 }

+HUF_CTableHeader HUF_readCTableHeader(HUF_CElt const* ctable)
+{
+    HUF_CTableHeader header;
+    ZSTD_memcpy(&header, ctable, sizeof(header));
+    return header;
+}
+
+static void HUF_writeCTableHeader(HUF_CElt* ctable, U32 tableLog, U32 maxSymbolValue)
+{
+    HUF_CTableHeader header;
+    HUF_STATIC_ASSERT(sizeof(ctable[0]) == sizeof(header));
+    ZSTD_memset(&header, 0, sizeof(header));
+    assert(tableLog < 256);
+    header.tableLog = (BYTE)tableLog;
+    assert(maxSymbolValue < 256);
+    header.maxSymbolValue = (BYTE)maxSymbolValue;
+    ZSTD_memcpy(ctable, &header, sizeof(header));
+}
+
 typedef struct {
    HUF_CompressWeightsWksp wksp;
    BYTE bitsToWeight[HUF_TABLELOG_MAX + 1];   /* precomputed conversion table */
@@ -236,6 +254,11 @@ size_t HUF_writeCTable_wksp(void* dst, size_t maxDstSize,
    U32 n;
    HUF_WriteCTableWksp* wksp = (HUF_WriteCTableWksp*)HUF_alignUpWorkspace(workspace, &workspaceSize, ZSTD_ALIGNOF(U32));

+    HUF_STATIC_ASSERT(HUF_CTABLE_WORKSPACE_SIZE >= sizeof(HUF_WriteCTableWksp));
+
+    assert(HUF_readCTableHeader(CTable).maxSymbolValue == maxSymbolValue);
+    assert(HUF_readCTableHeader(CTable).tableLog == huffLog);
+
    /* check conditions */
    if (workspaceSize < sizeof(HUF_WriteCTableWksp)) return ERROR(GENERIC);
    if (maxSymbolValue > HUF_SYMBOLVALUE_MAX) return ERROR(maxSymbolValue_tooLarge);
@@ -265,16 +288,6 @@ size_t HUF_writeCTable_wksp(void* dst, size_t maxDstSize,
    return ((maxSymbolValue+1)/2) + 1;
 }

-/*! HUF_writeCTable() :
-    `CTable` : Huffman tree to save, using huf representation.
-    @return : size of saved CTable */
-size_t HUF_writeCTable (void* dst, size_t maxDstSize,
-                        const HUF_CElt* CTable, unsigned maxSymbolValue, unsigned huffLog)
-{
-    HUF_WriteCTableWksp wksp;
-    return HUF_writeCTable_wksp(dst, maxDstSize, CTable, maxSymbolValue, huffLog, &wksp, sizeof(wksp));
-}
-

 size_t HUF_readCTable (HUF_CElt* CTable, unsigned* maxSymbolValuePtr, const void* src, size_t srcSize, unsigned* hasZeroWeights)
 {
@@ -292,7 +305,9 @@ size_t HUF_readCTable (HUF_CElt* CTable, unsigned* maxSymbolValuePtr, const void
    if (tableLog > HUF_TABLELOG_MAX) return ERROR(tableLog_tooLarge);
    if (nbSymbols > *maxSymbolValuePtr+1) return ERROR(maxSymbolValue_tooSmall);

-    CTable[0] = tableLog;
+    *maxSymbolValuePtr = nbSymbols - 1;
+
+    HUF_writeCTableHeader(CTable, tableLog, *maxSymbolValuePtr);

    /* Prepare base value per rank */
    {   U32 n, nextRankStart = 0;
@@ -324,7 +339,6 @@ size_t HUF_readCTable (HUF_CElt* CTable, unsigned* maxSymbolValuePtr, const void
        { U32 n; for (n=0; n<nbSymbols; n++) HUF_setValue(ct + n, valPerRank[HUF_getNbBits(ct[n])]++); }
    }

-    *maxSymbolValuePtr = nbSymbols - 1;
    return readSize;
 }

@@ -332,6 +346,8 @@ U32 HUF_getNbBitsFromCTable(HUF_CElt const* CTable, U32 symbolValue)
 {
    const HUF_CElt* const ct = CTable + 1;
    assert(symbolValue <= HUF_SYMBOLVALUE_MAX);
+    if (symbolValue > HUF_readCTableHeader(CTable).maxSymbolValue)
+        return 0;
    return (U32)HUF_getNbBits(ct[symbolValue]);
 }

@@ -486,7 +502,7 @@ typedef struct {
    U16 curr;
 } rankPos;

-typedef nodeElt huffNodeTable[HUF_CTABLE_WORKSPACE_SIZE_U32];
+typedef nodeElt huffNodeTable[2 * (HUF_SYMBOLVALUE_MAX + 1)];

 /* Number of buckets available for HUF_sort() */
 #define RANK_POSITION_TABLE_SIZE 192
@@ -732,7 +748,8 @@ static void HUF_buildCTableFromTree(HUF_CElt* CTable, nodeElt const* huffNode, i
        HUF_setNbBits(ct + huffNode[n].byte, huffNode[n].nbBits);   /* push nbBits per symbol, symbol order */
    for (n=0; n<alphabetSize; n++)
        HUF_setValue(ct + n, valPerRank[HUF_getNbBits(ct[n])]++);   /* assign value within rank, symbol order */
-    CTable[0] = maxNbBits;
+
+    HUF_writeCTableHeader(CTable, maxNbBits, maxSymbolValue);
 }

 size_t
@@ -745,6 +762,8 @@ HUF_buildCTable_wksp(HUF_CElt* CTable, const unsigned* count, U32 maxSymbolValue
    nodeElt* const huffNode = huffNode0+1;
    int nonNullRank;

+    HUF_STATIC_ASSERT(HUF_CTABLE_WORKSPACE_SIZE == sizeof(HUF_buildCTable_wksp_tables));
+
    DEBUGLOG(5, "HUF_buildCTable_wksp (alphabet size = %u)", maxSymbolValue+1);

    /* safety checks */
@@ -783,13 +802,20 @@ size_t HUF_estimateCompressedSize(const HUF_CElt* CTable, const unsigned* count,
 }

 int HUF_validateCTable(const HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue) {
-  HUF_CElt const* ct = CTable + 1;
-  int bad = 0;
-  int s;
-  for (s = 0; s <= (int)maxSymbolValue; ++s) {
-    bad |= (count[s] != 0) & (HUF_getNbBits(ct[s]) == 0);
-  }
-  return !bad;
+    HUF_CTableHeader header = HUF_readCTableHeader(CTable);
+    HUF_CElt const* ct = CTable + 1;
+    int bad = 0;
+    int s;
+
+    assert(header.tableLog <= HUF_TABLELOG_ABSOLUTEMAX);
+
+    if (header.maxSymbolValue < maxSymbolValue)
+        return 0;
+
+    for (s = 0; s <= (int)maxSymbolValue; ++s) {
+        bad |= (count[s] != 0) & (HUF_getNbBits(ct[s]) == 0);
+    }
+    return !bad;
 }

 size_t HUF_compressBound(size_t size) { return HUF_COMPRESSBOUND(size); }
@@ -1031,7 +1057,7 @@ HUF_compress1X_usingCTable_internal_body(void* dst, size_t dstSize,
                                   const void* src, size_t srcSize,
                                   const HUF_CElt* CTable)
 {
-    U32 const tableLog = (U32)CTable[0];
+    U32 const tableLog = HUF_readCTableHeader(CTable).tableLog;
    HUF_CElt const* ct = CTable + 1;
    const BYTE* ip = (const BYTE*) src;
    BYTE* const ostart = (BYTE*)dst;
@@ -1112,9 +1138,9 @@ HUF_compress1X_usingCTable_internal_default(void* dst, size_t dstSize,
 static size_t
 HUF_compress1X_usingCTable_internal(void* dst, size_t dstSize,
                              const void* src, size_t srcSize,
-                              const HUF_CElt* CTable, const int bmi2)
+                              const HUF_CElt* CTable, const int flags)
 {
-    if (bmi2) {
+    if (flags & HUF_flags_bmi2) {
        return HUF_compress1X_usingCTable_internal_bmi2(dst, dstSize, src, srcSize, CTable);
    }
    return HUF_compress1X_usingCTable_internal_default(dst, dstSize, src, srcSize, CTable);
@@ -1125,28 +1151,23 @@ HUF_compress1X_usingCTable_internal(void* dst, size_t dstSize,
 static size_t
 HUF_compress1X_usingCTable_internal(void* dst, size_t dstSize,
                              const void* src, size_t srcSize,
-                              const HUF_CElt* CTable, const int bmi2)
+                              const HUF_CElt* CTable, const int flags)
 {
-    (void)bmi2;
+    (void)flags;
    return HUF_compress1X_usingCTable_internal_body(dst, dstSize, src, srcSize, CTable);
 }

 #endif

-size_t HUF_compress1X_usingCTable(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable)
+size_t HUF_compress1X_usingCTable(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable, int flags)
 {
-    return HUF_compress1X_usingCTable_bmi2(dst, dstSize, src, srcSize, CTable, /* bmi2 */ 0);
-}
-
-size_t HUF_compress1X_usingCTable_bmi2(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable, int bmi2)
-{
-    return HUF_compress1X_usingCTable_internal(dst, dstSize, src, srcSize, CTable, bmi2);
+    return HUF_compress1X_usingCTable_internal(dst, dstSize, src, srcSize, CTable, flags);
 }

 static size_t
 HUF_compress4X_usingCTable_internal(void* dst, size_t dstSize,
                              const void* src, size_t srcSize,
-                              const HUF_CElt* CTable, int bmi2)
+                              const HUF_CElt* CTable, int flags)
 {
    size_t const segmentSize = (srcSize+3)/4;   /* first 3 segments */
    const BYTE* ip = (const BYTE*) src;
@@ -1160,7 +1181,7 @@ HUF_compress4X_usingCTable_internal(void* dst, size_t dstSize,
    op += 6;   /* jumpTable */

    assert(op <= oend);
-    {   CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, (size_t)(oend-op), ip, segmentSize, CTable, bmi2) );
+    {   CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, (size_t)(oend-op), ip, segmentSize, CTable, flags) );
        if (cSize == 0 || cSize > 65535) return 0;
        MEM_writeLE16(ostart, (U16)cSize);
        op += cSize;
@@ -1168,7 +1189,7 @@ HUF_compress4X_usingCTable_internal(void* dst, size_t dstSize,

    ip += segmentSize;
    assert(op <= oend);
-    {   CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, (size_t)(oend-op), ip, segmentSize, CTable, bmi2) );
+    {   CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, (size_t)(oend-op), ip, segmentSize, CTable, flags) );
        if (cSize == 0 || cSize > 65535) return 0;
        MEM_writeLE16(ostart+2, (U16)cSize);
        op += cSize;
@@ -1176,7 +1197,7 @@ HUF_compress4X_usingCTable_internal(void* dst, size_t dstSize,

    ip += segmentSize;
    assert(op <= oend);
-    {   CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, (size_t)(oend-op), ip, segmentSize, CTable, bmi2) );
+    {   CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, (size_t)(oend-op), ip, segmentSize, CTable, flags) );
        if (cSize == 0 || cSize > 65535) return 0;
        MEM_writeLE16(ostart+4, (U16)cSize);
        op += cSize;
@@ -1185,7 +1206,7 @@ HUF_compress4X_usingCTable_internal(void* dst, size_t dstSize,
    ip += segmentSize;
    assert(op <= oend);
    assert(ip <= iend);
-    {   CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, (size_t)(oend-op), ip, (size_t)(iend-ip), CTable, bmi2) );
+    {   CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, (size_t)(oend-op), ip, (size_t)(iend-ip), CTable, flags) );
        if (cSize == 0 || cSize > 65535) return 0;
        op += cSize;
    }
@@ -1193,14 +1214,9 @@ HUF_compress4X_usingCTable_internal(void* dst, size_t dstSize,
    return (size_t)(op-ostart);
 }

-size_t HUF_compress4X_usingCTable(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable)
+size_t HUF_compress4X_usingCTable(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable, int flags)
 {
-    return HUF_compress4X_usingCTable_bmi2(dst, dstSize, src, srcSize, CTable, /* bmi2 */ 0);
-}
-
-size_t HUF_compress4X_usingCTable_bmi2(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable, int bmi2)
-{
-    return HUF_compress4X_usingCTable_internal(dst, dstSize, src, srcSize, CTable, bmi2);
+    return HUF_compress4X_usingCTable_internal(dst, dstSize, src, srcSize, CTable, flags);
 }

 typedef enum { HUF_singleStream, HUF_fourStreams } HUF_nbStreams_e;
@@ -1208,11 +1224,11 @@ typedef enum { HUF_singleStream, HUF_fourStreams } HUF_nbStreams_e;
 static size_t HUF_compressCTable_internal(
                BYTE* const ostart, BYTE* op, BYTE* const oend,
                const void* src, size_t srcSize,
-                HUF_nbStreams_e nbStreams, const HUF_CElt* CTable, const int bmi2)
+                HUF_nbStreams_e nbStreams, const HUF_CElt* CTable, const int flags)
 {
    size_t const cSize = (nbStreams==HUF_singleStream) ?
-                         HUF_compress1X_usingCTable_internal(op, (size_t)(oend - op), src, srcSize, CTable, bmi2) :
-                         HUF_compress4X_usingCTable_internal(op, (size_t)(oend - op), src, srcSize, CTable, bmi2);
+                         HUF_compress1X_usingCTable_internal(op, (size_t)(oend - op), src, srcSize, CTable, flags) :
+                         HUF_compress4X_usingCTable_internal(op, (size_t)(oend - op), src, srcSize, CTable, flags);
    if (HUF_isError(cSize)) { return cSize; }
    if (cSize==0) { return 0; }   /* uncompressible */
    op += cSize;
@@ -1260,12 +1276,12 @@ unsigned HUF_optimalTableLog(
            void* workSpace, size_t wkspSize,
            HUF_CElt* table,
      const unsigned* count,
-            HUF_depth_mode depthMode)
+            int flags)
 {
    assert(srcSize > 1); /* Not supported, RLE should be used instead */
    assert(wkspSize >= sizeof(HUF_buildCTable_wksp_tables));

-    if (depthMode != HUF_depth_optimal) {
+    if (!(flags & HUF_flags_optimalDepth)) {
        /* cheap evaluation, based on FSE */
        return FSE_optimalTableLog_internal(maxTableLog, srcSize, maxSymbolValue, 1);
    }
@@ -1317,8 +1333,7 @@ HUF_compress_internal (void* dst, size_t dstSize,
                       unsigned maxSymbolValue, unsigned huffLog,
                       HUF_nbStreams_e nbStreams,
                       void* workSpace, size_t wkspSize,
-                       HUF_CElt* oldHufTable, HUF_repeat* repeat, int preferRepeat,
-                 const int bmi2, unsigned suspectUncompressible, HUF_depth_mode depthMode)
+                       HUF_CElt* oldHufTable, HUF_repeat* repeat, int flags)
 {
    HUF_compress_tables_t* const table = (HUF_compress_tables_t*)HUF_alignUpWorkspace(workSpace, &wkspSize, ZSTD_ALIGNOF(size_t));
    BYTE* const ostart = (BYTE*)dst;
@@ -1339,15 +1354,15 @@ HUF_compress_internal (void* dst, size_t dstSize,
    if (!huffLog) huffLog = HUF_TABLELOG_DEFAULT;

    /* Heuristic : If old table is valid, use it for small inputs */
-    if (preferRepeat && repeat && *repeat == HUF_repeat_valid) {
+    if ((flags & HUF_flags_preferRepeat) && repeat && *repeat == HUF_repeat_valid) {
        return HUF_compressCTable_internal(ostart, op, oend,
                                           src, srcSize,
-                                           nbStreams, oldHufTable, bmi2);
+                                           nbStreams, oldHufTable, flags);
    }

    /* If uncompressible data is suspected, do a smaller sampling first */
    DEBUG_STATIC_ASSERT(SUSPECT_INCOMPRESSIBLE_SAMPLE_RATIO >= 2);
-    if (suspectUncompressible && srcSize >= (SUSPECT_INCOMPRESSIBLE_SAMPLE_SIZE * SUSPECT_INCOMPRESSIBLE_SAMPLE_RATIO)) {
+    if ((flags & HUF_flags_suspectUncompressible) && srcSize >= (SUSPECT_INCOMPRESSIBLE_SAMPLE_SIZE * SUSPECT_INCOMPRESSIBLE_SAMPLE_RATIO)) {
        size_t largestTotal = 0;
        DEBUGLOG(5, "input suspected incompressible : sampling to check");
        {   unsigned maxSymbolValueBegin = maxSymbolValue;
@@ -1375,14 +1390,14 @@ HUF_compress_internal (void* dst, size_t dstSize,
        *repeat = HUF_repeat_none;
    }
    /* Heuristic : use existing table for small inputs */
-    if (preferRepeat && repeat && *repeat != HUF_repeat_none) {
+    if ((flags & HUF_flags_preferRepeat) && repeat && *repeat != HUF_repeat_none) {
        return HUF_compressCTable_internal(ostart, op, oend,
                                           src, srcSize,
-                                           nbStreams, oldHufTable, bmi2);
+                                           nbStreams, oldHufTable, flags);
    }

    /* Build Huffman Tree */
-    huffLog = HUF_optimalTableLog(huffLog, srcSize, maxSymbolValue, &table->wksps, sizeof(table->wksps), table->CTable, table->count, depthMode);
+    huffLog = HUF_optimalTableLog(huffLog, srcSize, maxSymbolValue, &table->wksps, sizeof(table->wksps), table->CTable, table->count, flags);
    {   size_t const maxBits = HUF_buildCTable_wksp(table->CTable, table->count,
                                            maxSymbolValue, huffLog,
                                            &table->wksps.buildCTable_wksp, sizeof(table->wksps.buildCTable_wksp));
@@ -1390,12 +1405,6 @@ HUF_compress_internal (void* dst, size_t dstSize,
        huffLog = (U32)maxBits;
        DEBUGLOG(6, "bit distribution completed (%zu symbols)", showCTableBits(table->CTable + 1, maxSymbolValue+1));
    }
-    /* Zero unused symbols in CTable, so we can check it for validity */
-    {
-        size_t const ctableSize = HUF_CTABLE_SIZE_ST(maxSymbolValue);
-        size_t const unusedSize = sizeof(table->CTable) - ctableSize * sizeof(HUF_CElt);
-        ZSTD_memset(table->CTable + ctableSize, 0, unusedSize);
-    }

    /* Write table description header */
    {   CHECK_V_F(hSize, HUF_writeCTable_wksp(op, dstSize, table->CTable, maxSymbolValue, huffLog,
@@ -1407,7 +1416,7 @@ HUF_compress_internal (void* dst, size_t dstSize,
            if (oldSize <= hSize + newSize || hSize + 12 >= srcSize) {
                return HUF_compressCTable_internal(ostart, op, oend,
                                                   src, srcSize,
-                                                   nbStreams, oldHufTable, bmi2);
+                                                   nbStreams, oldHufTable, flags);
        }   }

        /* Use the new huffman table */
@@ -1419,48 +1428,20 @@ HUF_compress_internal (void* dst, size_t dstSize,
    }
    return HUF_compressCTable_internal(ostart, op, oend,
                                       src, srcSize,
-                                       nbStreams, table->CTable, bmi2);
-}
-
-
-size_t HUF_compress1X_wksp (void* dst, size_t dstSize,
-                      const void* src, size_t srcSize,
-                      unsigned maxSymbolValue, unsigned huffLog,
-                      void* workSpace, size_t wkspSize)
-{
-    return HUF_compress_internal(dst, dstSize, src, srcSize,
-                                 maxSymbolValue, huffLog, HUF_singleStream,
-                                 workSpace, wkspSize,
-                                 NULL, NULL, 0, 0 /*bmi2*/, 0, HUF_depth_fast);
+                                       nbStreams, table->CTable, flags);
 }

 size_t HUF_compress1X_repeat (void* dst, size_t dstSize,
                      const void* src, size_t srcSize,
                      unsigned maxSymbolValue, unsigned huffLog,
                      void* workSpace, size_t wkspSize,
-                      HUF_CElt* hufTable, HUF_repeat* repeat, int preferRepeat,
-                      int bmi2, unsigned suspectUncompressible, HUF_depth_mode depthMode)
+                      HUF_CElt* hufTable, HUF_repeat* repeat, int flags)
 {
    DEBUGLOG(5, "HUF_compress1X_repeat (srcSize = %zu)", srcSize);
    return HUF_compress_internal(dst, dstSize, src, srcSize,
                                 maxSymbolValue, huffLog, HUF_singleStream,
                                 workSpace, wkspSize, hufTable,
-                                 repeat, preferRepeat, bmi2, suspectUncompressible, depthMode);
-}
-
-/* HUF_compress4X_repeat():
- * compress input using 4 streams.
- * provide workspace to generate compression tables */
-size_t HUF_compress4X_wksp (void* dst, size_t dstSize,
-                      const void* src, size_t srcSize,
-                      unsigned maxSymbolValue, unsigned huffLog,
-                      void* workSpace, size_t wkspSize)
-{
-    DEBUGLOG(5, "HUF_compress4X_wksp (srcSize = %zu)", srcSize);
-    return HUF_compress_internal(dst, dstSize, src, srcSize,
-                                 maxSymbolValue, huffLog, HUF_fourStreams,
-                                 workSpace, wkspSize,
-                                 NULL, NULL, 0, 0 /*bmi2*/, 0, HUF_depth_fast);
+                                 repeat, flags);
 }

 /* HUF_compress4X_repeat():
@@ -1471,45 +1452,11 @@ size_t HUF_compress4X_repeat (void* dst, size_t dstSize,
                      const void* src, size_t srcSize,
                      unsigned maxSymbolValue, unsigned huffLog,
                      void* workSpace, size_t wkspSize,
-                      HUF_CElt* hufTable, HUF_repeat* repeat, int preferRepeat, int bmi2,
-                      unsigned suspectUncompressible, HUF_depth_mode depthMode)
+                      HUF_CElt* hufTable, HUF_repeat* repeat, int flags)
 {
    DEBUGLOG(5, "HUF_compress4X_repeat (srcSize = %zu)", srcSize);
    return HUF_compress_internal(dst, dstSize, src, srcSize,
                                 maxSymbolValue, huffLog, HUF_fourStreams,
                                 workSpace, wkspSize,
-                                 hufTable, repeat, preferRepeat, bmi2, suspectUncompressible, depthMode);
+                                 hufTable, repeat, flags);
 }
-
-#ifndef ZSTD_NO_UNUSED_FUNCTIONS
-/** HUF_buildCTable() :
- * @return : maxNbBits
- *  Note : count is used before tree is written, so they can safely overlap
- */
-size_t HUF_buildCTable (HUF_CElt* tree, const unsigned* count, unsigned maxSymbolValue, unsigned maxNbBits)
-{
-    HUF_buildCTable_wksp_tables workspace;
-    return HUF_buildCTable_wksp(tree, count, maxSymbolValue, maxNbBits, &workspace, sizeof(workspace));
-}
-
-size_t HUF_compress1X (void* dst, size_t dstSize,
-                 const void* src, size_t srcSize,
-                 unsigned maxSymbolValue, unsigned huffLog)
-{
-    U64 workSpace[HUF_WORKSPACE_SIZE_U64];
-    return HUF_compress1X_wksp(dst, dstSize, src, srcSize, maxSymbolValue, huffLog, workSpace, sizeof(workSpace));
-}
-
-size_t HUF_compress2 (void* dst, size_t dstSize,
-                const void* src, size_t srcSize,
-                unsigned maxSymbolValue, unsigned huffLog)
-{
-    U64 workSpace[HUF_WORKSPACE_SIZE_U64];
-    return HUF_compress4X_wksp(dst, dstSize, src, srcSize, maxSymbolValue, huffLog, workSpace, sizeof(workSpace));
-}
-
-size_t HUF_compress (void* dst, size_t maxDstSize, const void* src, size_t srcSize)
-{
-    return HUF_compress2(dst, maxDstSize, src, srcSize, 255, HUF_TABLELOG_DEFAULT);
-}
-#endif
@@ -226,8 +226,10 @@ struct ZSTD_matchState_t {
    U32 hashLog3;           /* dispatch table for matches of len==3 : larger == faster, more memory */

    U32 rowHashLog;                          /* For row-based matchfinder: Hashlog based on nb of rows in the hashTable.*/
-    U16* tagTable;                           /* For row-based matchFinder: A row-based table containing the hashes and head index. */
+    BYTE* tagTable;                          /* For row-based matchFinder: A row-based table containing the hashes and head index. */
    U32 hashCache[ZSTD_ROW_HASH_CACHE_SIZE]; /* For row-based matchFinder: a cache of hashes to improve speed */
+    U64 hashSalt;                            /* For row-based matchFinder: salts the hash for re-use of tag table */
+    U32 hashSaltEntropy;                     /* For row-based matchFinder: collects entropy for salt generation */

    U32* hashTable;
    U32* hashTable3;
@@ -247,6 +249,13 @@ struct ZSTD_matchState_t {
     * This behavior is controlled from the cctx ms.
     * This parameter has no effect in the cdict ms. */
    int prefetchCDictTables;
+
+    /* When == 0, lazy match finders insert every position.
+     * When != 0, lazy match finders only insert positions they search.
+     * This allows them to skip much faster over incompressible data,
+     * at a small cost to compression ratio.
+     */
+    int lazySkipping;
 };

 typedef struct {
@@ -353,8 +362,14 @@ struct ZSTD_CCtx_params_s {

    /* Indicates whether an external matchfinder has been referenced.
     * Users can't set this externally.
-     * It is set internally in ZSTD_registerExternalMatchFinder(). */
-    int useExternalMatchFinder;
+     * It is set internally in ZSTD_registerSequenceProducer(). */
+    int useSequenceProducer;
+
+    /* Adjust the max block size*/
+    size_t maxBlockSize;
+
+    /* Controls repcode search in external sequence parsing */
+    ZSTD_paramSwitch_e searchForExternalRepcodes;
 };  /* typedef'd to ZSTD_CCtx_params within "zstd.h" */

 #define COMPRESS_SEQUENCES_WORKSPACE_SIZE (sizeof(unsigned) * (MaxSeq + 2))
@@ -389,7 +404,7 @@ typedef struct {
 /* Context for block-level external matchfinder API */
 typedef struct {
  void* mState;
-  ZSTD_externalMatchFinder_F* mFinder;
+  ZSTD_sequenceProducer_F* mFinder;
  ZSTD_Sequence* seqBuffer;
  size_t seqBufferCapacity;
 } ZSTD_externalMatchCtx;
@@ -491,7 +506,7 @@ typedef enum {
                                 * In this mode we take both the source size and the dictionary size
                                 * into account when selecting and adjusting the parameters.
                                 */
-    ZSTD_cpm_unknown = 3,       /* ZSTD_getCParams, ZSTD_getParams, ZSTD_adjustParams.
+    ZSTD_cpm_unknown = 3        /* ZSTD_getCParams, ZSTD_getParams, ZSTD_adjustParams.
                                 * We don't know what these parameters are for. We default to the legacy
                                 * behavior of taking both the source size and the dict size into account
                                 * when selecting and adjusting parameters.
@@ -781,28 +796,35 @@ ZSTD_count_2segments(const BYTE* ip, const BYTE* match,
 *  Hashes
 ***************************************/
 static const U32 prime3bytes = 506832829U;
-static U32    ZSTD_hash3(U32 u, U32 h) { assert(h <= 32); return ((u << (32-24)) * prime3bytes)  >> (32-h) ; }
-MEM_STATIC size_t ZSTD_hash3Ptr(const void* ptr, U32 h) { return ZSTD_hash3(MEM_readLE32(ptr), h); } /* only in zstd_opt.h */
+static U32    ZSTD_hash3(U32 u, U32 h, U32 s) { assert(h <= 32); return (((u << (32-24)) * prime3bytes) ^ s)  >> (32-h) ; }
+MEM_STATIC size_t ZSTD_hash3Ptr(const void* ptr, U32 h) { return ZSTD_hash3(MEM_readLE32(ptr), h, 0); } /* only in zstd_opt.h */
+MEM_STATIC size_t ZSTD_hash3PtrS(const void* ptr, U32 h, U32 s) { return ZSTD_hash3(MEM_readLE32(ptr), h, s); }

 static const U32 prime4bytes = 2654435761U;
-static U32    ZSTD_hash4(U32 u, U32 h) { assert(h <= 32); return (u * prime4bytes) >> (32-h) ; }
-static size_t ZSTD_hash4Ptr(const void* ptr, U32 h) { return ZSTD_hash4(MEM_readLE32(ptr), h); }
+static U32    ZSTD_hash4(U32 u, U32 h, U32 s) { assert(h <= 32); return ((u * prime4bytes) ^ s) >> (32-h) ; }
+static size_t ZSTD_hash4Ptr(const void* ptr, U32 h) { return ZSTD_hash4(MEM_readLE32(ptr), h, 0); }
+static size_t ZSTD_hash4PtrS(const void* ptr, U32 h, U32 s) { return ZSTD_hash4(MEM_readLE32(ptr), h, s); }

 static const U64 prime5bytes = 889523592379ULL;
-static size_t ZSTD_hash5(U64 u, U32 h) { assert(h <= 64); return (size_t)(((u  << (64-40)) * prime5bytes) >> (64-h)) ; }
-static size_t ZSTD_hash5Ptr(const void* p, U32 h) { return ZSTD_hash5(MEM_readLE64(p), h); }
+static size_t ZSTD_hash5(U64 u, U32 h, U64 s) { assert(h <= 64); return (size_t)((((u  << (64-40)) * prime5bytes) ^ s) >> (64-h)) ; }
+static size_t ZSTD_hash5Ptr(const void* p, U32 h) { return ZSTD_hash5(MEM_readLE64(p), h, 0); }
+static size_t ZSTD_hash5PtrS(const void* p, U32 h, U64 s) { return ZSTD_hash5(MEM_readLE64(p), h, s); }

 static const U64 prime6bytes = 227718039650203ULL;
-static size_t ZSTD_hash6(U64 u, U32 h) { assert(h <= 64); return (size_t)(((u  << (64-48)) * prime6bytes) >> (64-h)) ; }
-static size_t ZSTD_hash6Ptr(const void* p, U32 h) { return ZSTD_hash6(MEM_readLE64(p), h); }
+static size_t ZSTD_hash6(U64 u, U32 h, U64 s) { assert(h <= 64); return (size_t)((((u  << (64-48)) * prime6bytes) ^ s) >> (64-h)) ; }
+static size_t ZSTD_hash6Ptr(const void* p, U32 h) { return ZSTD_hash6(MEM_readLE64(p), h, 0); }
+static size_t ZSTD_hash6PtrS(const void* p, U32 h, U64 s) { return ZSTD_hash6(MEM_readLE64(p), h, s); }

 static const U64 prime7bytes = 58295818150454627ULL;
-static size_t ZSTD_hash7(U64 u, U32 h) { assert(h <= 64); return (size_t)(((u  << (64-56)) * prime7bytes) >> (64-h)) ; }
-static size_t ZSTD_hash7Ptr(const void* p, U32 h) { return ZSTD_hash7(MEM_readLE64(p), h); }
+static size_t ZSTD_hash7(U64 u, U32 h, U64 s) { assert(h <= 64); return (size_t)((((u  << (64-56)) * prime7bytes) ^ s) >> (64-h)) ; }
+static size_t ZSTD_hash7Ptr(const void* p, U32 h) { return ZSTD_hash7(MEM_readLE64(p), h, 0); }
+static size_t ZSTD_hash7PtrS(const void* p, U32 h, U64 s) { return ZSTD_hash7(MEM_readLE64(p), h, s); }

 static const U64 prime8bytes = 0xCF1BBCDCB7A56463ULL;
-static size_t ZSTD_hash8(U64 u, U32 h) { assert(h <= 64); return (size_t)(((u) * prime8bytes) >> (64-h)) ; }
-static size_t ZSTD_hash8Ptr(const void* p, U32 h) { return ZSTD_hash8(MEM_readLE64(p), h); }
+static size_t ZSTD_hash8(U64 u, U32 h, U64 s) { assert(h <= 64); return (size_t)((((u) * prime8bytes)  ^ s) >> (64-h)) ; }
+static size_t ZSTD_hash8Ptr(const void* p, U32 h) { return ZSTD_hash8(MEM_readLE64(p), h, 0); }
+static size_t ZSTD_hash8PtrS(const void* p, U32 h, U64 s) { return ZSTD_hash8(MEM_readLE64(p), h, s); }
+

 MEM_STATIC FORCE_INLINE_ATTR
 size_t ZSTD_hashPtr(const void* p, U32 hBits, U32 mls)
@@ -822,6 +844,24 @@ size_t ZSTD_hashPtr(const void* p, U32 hBits, U32 mls)
    }
 }

+MEM_STATIC FORCE_INLINE_ATTR
+size_t ZSTD_hashPtrSalted(const void* p, U32 hBits, U32 mls, const U64 hashSalt) {
+    /* Although some of these hashes do support hBits up to 64, some do not.
+     * To be on the safe side, always avoid hBits > 32. */
+    assert(hBits <= 32);
+
+    switch(mls)
+    {
+        default:
+        case 4: return ZSTD_hash4PtrS(p, hBits, (U32)hashSalt);
+        case 5: return ZSTD_hash5PtrS(p, hBits, hashSalt);
+        case 6: return ZSTD_hash6PtrS(p, hBits, hashSalt);
+        case 7: return ZSTD_hash7PtrS(p, hBits, hashSalt);
+        case 8: return ZSTD_hash8PtrS(p, hBits, hashSalt);
+    }
+}
+
+
 /** ZSTD_ipow() :
 * Return base^exponent.
 */
@@ -1169,10 +1209,15 @@ ZSTD_checkDictValidity(const ZSTD_window_t* window,
                    (unsigned)blockEndIdx, (unsigned)maxDist, (unsigned)loadedDictEnd);
        assert(blockEndIdx >= loadedDictEnd);

-        if (blockEndIdx > loadedDictEnd + maxDist) {
+        if (blockEndIdx > loadedDictEnd + maxDist || loadedDictEnd != window->dictLimit) {
            /* On reaching window size, dictionaries are invalidated.
             * For simplification, if window size is reached anywhere within next block,
             * the dictionary is invalidated for the full block.
+             *
+             * We also have to invalidate the dictionary if ZSTD_window_update() has detected
+             * non-contiguous segments, which means that loadedDictEnd != window->dictLimit.
+             * loadedDictEnd may be 0, if forceWindow is true, but in that case we never use
+             * dictMatchState, so setting it to NULL is not a problem.
             */
            DEBUGLOG(6, "invalidating dictionary for current block (distance > windowSize)");
            *loadedDictEndPtr = 0;
@@ -1422,11 +1467,10 @@ size_t ZSTD_writeLastEmptyBlock(void* dst, size_t dstCapacity);
 * This cannot be used when long range matching is enabled.
 * Zstd will use these sequences, and pass the literals to a secondary block
 * compressor.
- * @return : An error code on failure.
 * NOTE: seqs are not verified! Invalid sequences can cause out-of-bounds memory
 * access and data corruption.
 */
-size_t ZSTD_referenceExternalSequences(ZSTD_CCtx* cctx, rawSeq* seq, size_t nbSeq);
+void ZSTD_referenceExternalSequences(ZSTD_CCtx* cctx, rawSeq* seq, size_t nbSeq);

 /** ZSTD_cycleLog() :
 *  condition for correct operation : hashLog > 1 */
@@ -1445,7 +1489,7 @@ size_t
 ZSTD_copySequencesToSeqStoreExplicitBlockDelim(ZSTD_CCtx* cctx,
                                              ZSTD_sequencePosition* seqPos,
                                        const ZSTD_Sequence* const inSeqs, size_t inSeqsSize,
-                                        const void* src, size_t blockSize);
+                                        const void* src, size_t blockSize, ZSTD_paramSwitch_e externalRepSearch);

 /* Returns the number of bytes to move the current read position back by.
 * Only non-zero if we ended up splitting a sequence.
@@ -1462,6 +1506,26 @@ ZSTD_copySequencesToSeqStoreExplicitBlockDelim(ZSTD_CCtx* cctx,
 size_t
 ZSTD_copySequencesToSeqStoreNoBlockDelim(ZSTD_CCtx* cctx, ZSTD_sequencePosition* seqPos,
                                   const ZSTD_Sequence* const inSeqs, size_t inSeqsSize,
-                                   const void* src, size_t blockSize);
+                                   const void* src, size_t blockSize, ZSTD_paramSwitch_e externalRepSearch);
+
+
+/* ===============================================================
+ * Deprecated definitions that are still used internally to avoid
+ * deprecation warnings. These functions are exactly equivalent to
+ * their public variants, but avoid the deprecation warnings.
+ * =============================================================== */
+
+size_t ZSTD_compressBegin_usingCDict_deprecated(ZSTD_CCtx* cctx, const ZSTD_CDict* cdict);
+
+size_t ZSTD_compressContinue_public(ZSTD_CCtx* cctx,
+                                    void* dst, size_t dstCapacity,
+                              const void* src, size_t srcSize);
+
+size_t ZSTD_compressEnd_public(ZSTD_CCtx* cctx,
+                               void* dst, size_t dstCapacity,
+                         const void* src, size_t srcSize);
+
+size_t ZSTD_compressBlock_deprecated(ZSTD_CCtx* cctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize);
+

 #endif /* ZSTD_COMPRESS_H */
@@ -24,9 +24,9 @@ static size_t showHexa(const void* src, size_t srcSize)
    const BYTE* const ip = (const BYTE*)src;
    size_t u;
    for (u=0; u<srcSize; u++) {
-        RAWLOG(6, " %02X", ip[u]); (void)ip;
+        RAWLOG(5, " %02X", ip[u]); (void)ip;
    }
-    RAWLOG(6, " \n");
+    RAWLOG(5, " \n");
    return srcSize;
 }

@@ -65,12 +65,26 @@ size_t ZSTD_noCompressLiterals (void* dst, size_t dstCapacity, const void* src,
    return srcSize + flSize;
 }

+static int allBytesIdentical(const void* src, size_t srcSize)
+{
+    assert(srcSize >= 1);
+    assert(src != NULL);
+    {   const BYTE b = ((const BYTE*)src)[0];
+        size_t p;
+        for (p=1; p<srcSize; p++) {
+            if (((const BYTE*)src)[p] != b) return 0;
+        }
+        return 1;
+    }
+}
+
 size_t ZSTD_compressRleLiteralsBlock (void* dst, size_t dstCapacity, const void* src, size_t srcSize)
 {
    BYTE* const ostart = (BYTE*)dst;
    U32   const flSize = 1 + (srcSize>31) + (srcSize>4095);

-    (void)dstCapacity;  /* dstCapacity already guaranteed to be >=4, hence large enough */
+    assert(dstCapacity >= 4); (void)dstCapacity;
+    assert(allBytesIdentical(src, srcSize));

    switch(flSize)
    {
@@ -88,7 +102,7 @@ size_t ZSTD_compressRleLiteralsBlock (void* dst, size_t dstCapacity, const void*
    }

    ostart[flSize] = *(const BYTE*)src;
-    DEBUGLOG(5, "RLE literals: %u -> %u", (U32)srcSize, (U32)flSize + 1);
+    DEBUGLOG(5, "RLE : Repeated Literal (%02X: %u times) -> %u bytes encoded", ((const BYTE*)src)[0], (U32)srcSize, (U32)flSize + 1);
    return flSize+1;
 }

@@ -105,8 +119,8 @@ ZSTD_minLiteralsToCompress(ZSTD_strategy strategy, HUF_repeat huf_repeat)
    /* btultra2 : min 8 bytes;
     * then 2x larger for each successive compression strategy
     * max threshold 64 bytes */
-    {   int const shift = MIN(9-strategy, 3);
-        size_t const mintc = (huf_repeat == HUF_repeat_valid) ? 6 : 8 << shift;
+    {   int const shift = MIN(9-(int)strategy, 3);
+        size_t const mintc = (huf_repeat == HUF_repeat_valid) ? 6 : (size_t)8 << shift;
        DEBUGLOG(7, "minLiteralsToCompress = %zu", mintc);
        return mintc;
    }
@@ -146,9 +160,13 @@ size_t ZSTD_compressLiterals (

    RETURN_ERROR_IF(dstCapacity < lhSize+1, dstSize_tooSmall, "not enough space for compression");
    {   HUF_repeat repeat = prevHuf->repeatMode;
-        int const preferRepeat = (strategy < ZSTD_lazy) ? srcSize <= 1024 : 0;
-        HUF_depth_mode const depthMode = (strategy >= HUF_OPTIMAL_DEPTH_THRESHOLD) ? HUF_depth_optimal : HUF_depth_fast;
-        typedef size_t (*huf_compress_f)(void*, size_t, const void*, size_t, unsigned, unsigned, void*, size_t, HUF_CElt*, HUF_repeat*, int, int, unsigned, HUF_depth_mode);
+        int const flags = 0
+            | (bmi2 ? HUF_flags_bmi2 : 0)
+            | (strategy < ZSTD_lazy && srcSize <= 1024 ? HUF_flags_preferRepeat : 0)
+            | (strategy >= HUF_OPTIMAL_DEPTH_THRESHOLD ? HUF_flags_optimalDepth : 0)
+            | (suspectUncompressible ? HUF_flags_suspectUncompressible : 0);
+
+        typedef size_t (*huf_compress_f)(void*, size_t, const void*, size_t, unsigned, unsigned, void*, size_t, HUF_CElt*, HUF_repeat*, int);
        huf_compress_f huf_compress;
        if (repeat == HUF_repeat_valid && lhSize == 3) singleStream = 1;
        huf_compress = singleStream ? HUF_compress1X_repeat : HUF_compress4X_repeat;
@@ -157,11 +175,11 @@ size_t ZSTD_compressLiterals (
                                HUF_SYMBOLVALUE_MAX, LitHufLog,
                                entropyWorkspace, entropyWorkspaceSize,
                                (HUF_CElt*)nextHuf->CTable,
-                                &repeat, preferRepeat,
-                                bmi2, suspectUncompressible, depthMode);
+                                &repeat, flags);
+        DEBUGLOG(5, "%zu literals compressed into %zu bytes (before header)", srcSize, cLitSize);
        if (repeat != HUF_repeat_none) {
            /* reused the existing table */
-            DEBUGLOG(5, "Reusing previous huffman table");
+            DEBUGLOG(5, "reusing statistics from previous huffman block");
            hType = set_repeat;
        }
    }
@@ -172,9 +190,15 @@ size_t ZSTD_compressLiterals (
            return ZSTD_noCompressLiterals(dst, dstCapacity, src, srcSize);
    }   }
    if (cLitSize==1) {
-        ZSTD_memcpy(nextHuf, prevHuf, sizeof(*prevHuf));
-        return ZSTD_compressRleLiteralsBlock(dst, dstCapacity, src, srcSize);
-    }
+        /* A return value of 1 signals that the alphabet consists of a single symbol.
+         * However, in some rare circumstances, it could be the compressed size (a single byte).
+         * For that outcome to have a chance to happen, it's necessary that `srcSize < 8`.
+         * (it's also necessary to not generate statistics).
+         * Therefore, in such a case, actively check that all bytes are identical. */
+        if ((srcSize >= 8) || allBytesIdentical(src, srcSize)) {
+            ZSTD_memcpy(nextHuf, prevHuf, sizeof(*prevHuf));
+            return ZSTD_compressRleLiteralsBlock(dst, dstCapacity, src, srcSize);
+    }   }

    if (hType == set_compressed) {
        /* using a newly constructed table */
@@ -16,6 +16,10 @@

 size_t ZSTD_noCompressLiterals (void* dst, size_t dstCapacity, const void* src, size_t srcSize);

+/* ZSTD_compressRleLiteralsBlock() :
+ * Conditions :
+ * - All bytes in @src are identical
+ * - dstCapacity >= 4 */
 size_t ZSTD_compressRleLiteralsBlock (void* dst, size_t dstCapacity, const void* src, size_t srcSize);

 /* ZSTD_compressLiterals():
@@ -54,8 +54,6 @@ ZSTD_compressSubBlock_literal(const HUF_CElt* hufTable,
    symbolEncodingType_e hType = writeEntropy ? hufMetadata->hType : set_repeat;
    size_t cLitSize = 0;

-    (void)bmi2; /* TODO bmi2... */
-
    DEBUGLOG(5, "ZSTD_compressSubBlock_literal (litSize=%zu, lhSize=%zu, writeEntropy=%d)", litSize, lhSize, writeEntropy);

    *entropyWritten = 0;
@@ -77,9 +75,9 @@ ZSTD_compressSubBlock_literal(const HUF_CElt* hufTable,
        DEBUGLOG(5, "ZSTD_compressSubBlock_literal (hSize=%zu)", hufMetadata->hufDesSize);
    }

-    /* TODO bmi2 */
-    {   const size_t cSize = singleStream ? HUF_compress1X_usingCTable(op, oend-op, literals, litSize, hufTable)
-                                          : HUF_compress4X_usingCTable(op, oend-op, literals, litSize, hufTable);
+    {   int const flags = bmi2 ? HUF_flags_bmi2 : 0;
+        const size_t cSize = singleStream ? HUF_compress1X_usingCTable(op, oend-op, literals, litSize, hufTable, flags)
+                                          : HUF_compress4X_usingCTable(op, oend-op, literals, litSize, hufTable, flags);
        op += cSize;
        cLitSize += cSize;
        if (cSize == 0 || ERR_isError(cSize)) {
@@ -182,7 +180,7 @@ ZSTD_compressSubBlock_sequences(const ZSTD_fseCTables_t* fseTables,
    /* Sequences Header */
    RETURN_ERROR_IF((oend-op) < 3 /*max nbSeq Size*/ + 1 /*seqHead*/,
                    dstSize_tooSmall, "");
-    if (nbSeq < 0x7F)
+    if (nbSeq < 128)
        *op++ = (BYTE)nbSeq;
    else if (nbSeq < LONGNBSEQ)
        op[0] = (BYTE)((nbSeq>>8) + 0x80), op[1] = (BYTE)nbSeq, op+=2;
@@ -14,7 +14,9 @@
 /*-*************************************
 *  Dependencies
 ***************************************/
+#include "../common/allocations.h"  /* ZSTD_customMalloc, ZSTD_customFree */
 #include "../common/zstd_internal.h"
+#include "../common/portability_macros.h"

 #if defined (__cplusplus)
 extern "C" {
@@ -44,8 +46,9 @@ extern "C" {
 ***************************************/
 typedef enum {
    ZSTD_cwksp_alloc_objects,
-    ZSTD_cwksp_alloc_buffers,
-    ZSTD_cwksp_alloc_aligned
+    ZSTD_cwksp_alloc_aligned_init_once,
+    ZSTD_cwksp_alloc_aligned,
+    ZSTD_cwksp_alloc_buffers
 } ZSTD_cwksp_alloc_phase_e;

 /**
@@ -98,8 +101,8 @@ typedef enum {
 *
 * Workspace Layout:
 *
- * [                        ... workspace ...                         ]
- * [objects][tables ... ->] free space [<- ... aligned][<- ... buffers]
+ * [                        ... workspace ...                           ]
+ * [objects][tables ->] free space [<- buffers][<- aligned][<- init once]
 *
 * The various objects that live in the workspace are divided into the
 * following categories, and are allocated separately:
@@ -123,9 +126,18 @@ typedef enum {
 *   uint32_t arrays, all of whose values are between 0 and (nextSrc - base).
 *   Their sizes depend on the cparams. These tables are 64-byte aligned.
 *
- * - Aligned: these buffers are used for various purposes that require 4 byte
- *   alignment, but don't require any initialization before they're used. These
- *   buffers are each aligned to 64 bytes.
+ * - Init once: these buffers require to be initialized at least once before
+ *   use. They should be used when we want to skip memory initialization
+ *   while not triggering memory checkers (like Valgrind) when reading from
+ *   from this memory without writing to it first.
+ *   These buffers should be used carefully as they might contain data
+ *   from previous compressions.
+ *   Buffers are aligned to 64 bytes.
+ *
+ * - Aligned: these buffers don't require any initialization before they're
+ *   used. The user of the buffer should make sure they write into a buffer
+ *   location before reading from it.
+ *   Buffers are aligned to 64 bytes.
 *
 * - Buffers: these buffers are used for various purposes that don't require
 *   any alignment or initialization before they're used. This means they can
@@ -137,8 +149,9 @@ typedef enum {
 * correctly packed into the workspace buffer. That order is:
 *
 * 1. Objects
- * 2. Buffers
- * 3. Aligned/Tables
+ * 2. Init once / Tables
+ * 3. Aligned / Tables
+ * 4. Buffers / Tables
 *
 * Attempts to reserve objects of different types out of order will fail.
 */
@@ -150,6 +163,7 @@ typedef struct {
    void* tableEnd;
    void* tableValidEnd;
    void* allocStart;
+    void* initOnceStart;

    BYTE allocFailed;
    int workspaceOversizedDuration;
@@ -162,6 +176,7 @@ typedef struct {
 ***************************************/

 MEM_STATIC size_t ZSTD_cwksp_available_space(ZSTD_cwksp* ws);
+MEM_STATIC void*  ZSTD_cwksp_initialAllocStart(ZSTD_cwksp* ws);

 MEM_STATIC void ZSTD_cwksp_assert_internal_consistency(ZSTD_cwksp* ws) {
    (void)ws;
@@ -171,6 +186,21 @@ MEM_STATIC void ZSTD_cwksp_assert_internal_consistency(ZSTD_cwksp* ws) {
    assert(ws->tableEnd <= ws->allocStart);
    assert(ws->tableValidEnd <= ws->allocStart);
    assert(ws->allocStart <= ws->workspaceEnd);
+    assert(ws->initOnceStart <= ZSTD_cwksp_initialAllocStart(ws));
+    assert(ws->workspace <= ws->initOnceStart);
+#if ZSTD_MEMORY_SANITIZER
+    {
+        intptr_t const offset = __msan_test_shadow(ws->initOnceStart,
+            (U8*)ZSTD_cwksp_initialAllocStart(ws) - (U8*)ws->initOnceStart);
+        (void)offset;
+#if defined(ZSTD_MSAN_PRINT)
+        if(offset!=-1) {
+            __msan_print_shadow((U8*)ws->initOnceStart + offset - 8, 32);
+        }
+#endif
+        assert(offset==-1);
+    };
+#endif
 }

 /**
@@ -217,14 +247,10 @@ MEM_STATIC size_t ZSTD_cwksp_aligned_alloc_size(size_t size) {
 * for internal purposes (currently only alignment).
 */
 MEM_STATIC size_t ZSTD_cwksp_slack_space_required(void) {
-    /* For alignment, the wksp will always allocate an additional n_1=[1, 64] bytes
-     * to align the beginning of tables section, as well as another n_2=[0, 63] bytes
-     * to align the beginning of the aligned section.
-     *
-     * n_1 + n_2 == 64 bytes if the cwksp is freshly allocated, due to tables and
-     * aligneds being sized in multiples of 64 bytes.
+    /* For alignment, the wksp will always allocate an additional 2*ZSTD_CWKSP_ALIGNMENT_BYTES
+     * bytes to align the beginning of tables section and end of buffers;
     */
-    size_t const slackSpace = ZSTD_CWKSP_ALIGNMENT_BYTES;
+    size_t const slackSpace = ZSTD_CWKSP_ALIGNMENT_BYTES * 2;
    return slackSpace;
 }

@@ -237,10 +263,18 @@ MEM_STATIC size_t ZSTD_cwksp_bytes_to_align_ptr(void* ptr, const size_t alignByt
    size_t const alignBytesMask = alignBytes - 1;
    size_t const bytes = (alignBytes - ((size_t)ptr & (alignBytesMask))) & alignBytesMask;
    assert((alignBytes & alignBytesMask) == 0);
-    assert(bytes != ZSTD_CWKSP_ALIGNMENT_BYTES);
+    assert(bytes < alignBytes);
    return bytes;
 }

+/**
+ * Returns the initial value for allocStart which is used to determine the position from
+ * which we can allocate from the end of the workspace.
+ */
+MEM_STATIC void*  ZSTD_cwksp_initialAllocStart(ZSTD_cwksp* ws) {
+    return (void*)((size_t)ws->workspaceEnd & ~(ZSTD_CWKSP_ALIGNMENT_BYTES-1));
+}
+
 /**
 * Internal function. Do not use directly.
 * Reserves the given number of bytes within the aligned/buffer segment of the wksp,
@@ -281,27 +315,16 @@ ZSTD_cwksp_internal_advance_phase(ZSTD_cwksp* ws, ZSTD_cwksp_alloc_phase_e phase
 {
    assert(phase >= ws->phase);
    if (phase > ws->phase) {
-        /* Going from allocating objects to allocating buffers */
-        if (ws->phase < ZSTD_cwksp_alloc_buffers &&
-                phase >= ZSTD_cwksp_alloc_buffers) {
+        /* Going from allocating objects to allocating initOnce / tables */
+        if (ws->phase < ZSTD_cwksp_alloc_aligned_init_once &&
+            phase >= ZSTD_cwksp_alloc_aligned_init_once) {
            ws->tableValidEnd = ws->objectEnd;
-        }
+            ws->initOnceStart = ZSTD_cwksp_initialAllocStart(ws);

-        /* Going from allocating buffers to allocating aligneds/tables */
-        if (ws->phase < ZSTD_cwksp_alloc_aligned &&
-                phase >= ZSTD_cwksp_alloc_aligned) {
-            {   /* Align the start of the "aligned" to 64 bytes. Use [1, 64] bytes. */
-                size_t const bytesToAlign =
-                    ZSTD_CWKSP_ALIGNMENT_BYTES - ZSTD_cwksp_bytes_to_align_ptr(ws->allocStart, ZSTD_CWKSP_ALIGNMENT_BYTES);
-                DEBUGLOG(5, "reserving aligned alignment addtl space: %zu", bytesToAlign);
-                ZSTD_STATIC_ASSERT((ZSTD_CWKSP_ALIGNMENT_BYTES & (ZSTD_CWKSP_ALIGNMENT_BYTES - 1)) == 0); /* power of 2 */
-                RETURN_ERROR_IF(!ZSTD_cwksp_reserve_internal_buffer_space(ws, bytesToAlign),
-                                memory_allocation, "aligned phase - alignment initial allocation failed!");
-            }
            {   /* Align the start of the tables to 64 bytes. Use [0, 63] bytes */
-                void* const alloc = ws->objectEnd;
+                void *const alloc = ws->objectEnd;
                size_t const bytesToAlign = ZSTD_cwksp_bytes_to_align_ptr(alloc, ZSTD_CWKSP_ALIGNMENT_BYTES);
-                void* const objectEnd = (BYTE*)alloc + bytesToAlign;
+                void *const objectEnd = (BYTE *) alloc + bytesToAlign;
                DEBUGLOG(5, "reserving table alignment addtl space: %zu", bytesToAlign);
                RETURN_ERROR_IF(objectEnd > ws->workspaceEnd, memory_allocation,
                                "table phase - alignment initial allocation failed!");
@@ -309,7 +332,9 @@ ZSTD_cwksp_internal_advance_phase(ZSTD_cwksp* ws, ZSTD_cwksp_alloc_phase_e phase
                ws->tableEnd = objectEnd;  /* table area starts being empty */
                if (ws->tableValidEnd < ws->tableEnd) {
                    ws->tableValidEnd = ws->tableEnd;
-        }   }   }
+                }
+            }
+        }
        ws->phase = phase;
        ZSTD_cwksp_assert_internal_consistency(ws);
    }
@@ -321,7 +346,7 @@ ZSTD_cwksp_internal_advance_phase(ZSTD_cwksp* ws, ZSTD_cwksp_alloc_phase_e phase
 */
 MEM_STATIC int ZSTD_cwksp_owns_buffer(const ZSTD_cwksp* ws, const void* ptr)
 {
-    return (ptr != NULL) && (ws->workspace <= ptr) && (ptr <= ws->workspaceEnd);
+    return (ptr != NULL) && (ws->workspace <= ptr) && (ptr < ws->workspaceEnd);
 }

 /**
@@ -348,7 +373,9 @@ ZSTD_cwksp_reserve_internal(ZSTD_cwksp* ws, size_t bytes, ZSTD_cwksp_alloc_phase
    if (alloc) {
        alloc = (BYTE *)alloc + ZSTD_CWKSP_ASAN_REDZONE_SIZE;
        if (ws->isStatic == ZSTD_cwksp_dynamic_alloc) {
-            __asan_unpoison_memory_region(alloc, bytes);
+            /* We need to keep the redzone poisoned while unpoisoning the bytes that
+             * are actually allocated. */
+            __asan_unpoison_memory_region(alloc, bytes - 2 * ZSTD_CWKSP_ASAN_REDZONE_SIZE);
        }
    }
 #endif
@@ -364,6 +391,36 @@ MEM_STATIC BYTE* ZSTD_cwksp_reserve_buffer(ZSTD_cwksp* ws, size_t bytes)
    return (BYTE*)ZSTD_cwksp_reserve_internal(ws, bytes, ZSTD_cwksp_alloc_buffers);
 }

+/**
+ * Reserves and returns memory sized on and aligned on ZSTD_CWKSP_ALIGNMENT_BYTES (64 bytes).
+ * This memory has been initialized at least once in the past.
+ * This doesn't mean it has been initialized this time, and it might contain data from previous
+ * operations.
+ * The main usage is for algorithms that might need read access into uninitialized memory.
+ * The algorithm must maintain safety under these conditions and must make sure it doesn't
+ * leak any of the past data (directly or in side channels).
+ */
+MEM_STATIC void* ZSTD_cwksp_reserve_aligned_init_once(ZSTD_cwksp* ws, size_t bytes)
+{
+    size_t const alignedBytes = ZSTD_cwksp_align(bytes, ZSTD_CWKSP_ALIGNMENT_BYTES);
+    void* ptr = ZSTD_cwksp_reserve_internal(ws, alignedBytes, ZSTD_cwksp_alloc_aligned_init_once);
+    assert(((size_t)ptr & (ZSTD_CWKSP_ALIGNMENT_BYTES-1))== 0);
+    if(ptr && ptr < ws->initOnceStart) {
+        /* We assume the memory following the current allocation is either:
+         * 1. Not usable as initOnce memory (end of workspace)
+         * 2. Another initOnce buffer that has been allocated before (and so was previously memset)
+         * 3. An ASAN redzone, in which case we don't want to write on it
+         * For these reasons it should be fine to not explicitly zero every byte up to ws->initOnceStart.
+         * Note that we assume here that MSAN and ASAN cannot run in the same time. */
+        ZSTD_memset(ptr, 0, MIN((size_t)((U8*)ws->initOnceStart - (U8*)ptr), alignedBytes));
+        ws->initOnceStart = ptr;
+    }
+#if ZSTD_MEMORY_SANITIZER
+    assert(__msan_test_shadow(ptr, bytes) == -1);
+#endif
+    return ptr;
+}
+
 /**
 * Reserves and returns memory sized on and aligned on ZSTD_CWKSP_ALIGNMENT_BYTES (64 bytes).
 */
@@ -382,13 +439,17 @@ MEM_STATIC void* ZSTD_cwksp_reserve_aligned(ZSTD_cwksp* ws, size_t bytes)
 */
 MEM_STATIC void* ZSTD_cwksp_reserve_table(ZSTD_cwksp* ws, size_t bytes)
 {
-    const ZSTD_cwksp_alloc_phase_e phase = ZSTD_cwksp_alloc_aligned;
+    const ZSTD_cwksp_alloc_phase_e phase = ZSTD_cwksp_alloc_aligned_init_once;
    void* alloc;
    void* end;
    void* top;

-    if (ZSTD_isError(ZSTD_cwksp_internal_advance_phase(ws, phase))) {
-        return NULL;
+    /* We can only start allocating tables after we are done reserving space for objects at the
+     * start of the workspace */
+    if(ws->phase < phase) {
+        if (ZSTD_isError(ZSTD_cwksp_internal_advance_phase(ws, phase))) {
+            return NULL;
+        }
    }
    alloc = ws->tableEnd;
    end = (BYTE *)alloc + bytes;
@@ -467,11 +528,19 @@ MEM_STATIC void ZSTD_cwksp_mark_tables_dirty(ZSTD_cwksp* ws)
 #if ZSTD_MEMORY_SANITIZER && !defined (ZSTD_MSAN_DONT_POISON_WORKSPACE)
    /* To validate that the table re-use logic is sound, and that we don't
     * access table space that we haven't cleaned, we re-"poison" the table
-     * space every time we mark it dirty. */
+     * space every time we mark it dirty.
+     * Since tableValidEnd space and initOnce space may overlap we don't poison
+     * the initOnce portion as it break its promise. This means that this poisoning
+     * check isn't always applied fully. */
    {
        size_t size = (BYTE*)ws->tableValidEnd - (BYTE*)ws->objectEnd;
        assert(__msan_test_shadow(ws->objectEnd, size) == -1);
-        __msan_poison(ws->objectEnd, size);
+        if((BYTE*)ws->tableValidEnd < (BYTE*)ws->initOnceStart) {
+            __msan_poison(ws->objectEnd, size);
+        } else {
+            assert(ws->initOnceStart >= ws->objectEnd);
+            __msan_poison(ws->objectEnd, (BYTE*)ws->initOnceStart - (BYTE*)ws->objectEnd);
+        }
    }
 #endif

@@ -536,11 +605,14 @@ MEM_STATIC void ZSTD_cwksp_clear(ZSTD_cwksp* ws) {
 #if ZSTD_MEMORY_SANITIZER && !defined (ZSTD_MSAN_DONT_POISON_WORKSPACE)
    /* To validate that the context re-use logic is sound, and that we don't
     * access stuff that this compression hasn't initialized, we re-"poison"
-     * the workspace (or at least the non-static, non-table parts of it)
-     * every time we start a new compression. */
+     * the workspace except for the areas in which we expect memory re-use
+     * without initialization (objects, valid tables area and init once
+     * memory). */
    {
-        size_t size = (BYTE*)ws->workspaceEnd - (BYTE*)ws->tableValidEnd;
-        __msan_poison(ws->tableValidEnd, size);
+        if((BYTE*)ws->tableValidEnd < (BYTE*)ws->initOnceStart) {
+            size_t size = (BYTE*)ws->initOnceStart - (BYTE*)ws->tableValidEnd;
+            __msan_poison(ws->tableValidEnd, size);
+        }
    }
 #endif

@@ -556,14 +628,23 @@ MEM_STATIC void ZSTD_cwksp_clear(ZSTD_cwksp* ws) {
 #endif

    ws->tableEnd = ws->objectEnd;
-    ws->allocStart = ws->workspaceEnd;
+    ws->allocStart = ZSTD_cwksp_initialAllocStart(ws);
    ws->allocFailed = 0;
-    if (ws->phase > ZSTD_cwksp_alloc_buffers) {
-        ws->phase = ZSTD_cwksp_alloc_buffers;
+    if (ws->phase > ZSTD_cwksp_alloc_aligned_init_once) {
+        ws->phase = ZSTD_cwksp_alloc_aligned_init_once;
    }
    ZSTD_cwksp_assert_internal_consistency(ws);
 }

+MEM_STATIC size_t ZSTD_cwksp_sizeof(const ZSTD_cwksp* ws) {
+    return (size_t)((BYTE*)ws->workspaceEnd - (BYTE*)ws->workspace);
+}
+
+MEM_STATIC size_t ZSTD_cwksp_used(const ZSTD_cwksp* ws) {
+    return (size_t)((BYTE*)ws->tableEnd - (BYTE*)ws->workspace)
+         + (size_t)((BYTE*)ws->workspaceEnd - (BYTE*)ws->allocStart);
+}
+
 /**
 * The provided workspace takes ownership of the buffer [start, start+size).
 * Any existing values in the workspace are ignored (the previously managed
@@ -576,6 +657,7 @@ MEM_STATIC void ZSTD_cwksp_init(ZSTD_cwksp* ws, void* start, size_t size, ZSTD_c
    ws->workspaceEnd = (BYTE*)start + size;
    ws->objectEnd = ws->workspace;
    ws->tableValidEnd = ws->objectEnd;
+    ws->initOnceStart = ZSTD_cwksp_initialAllocStart(ws);
    ws->phase = ZSTD_cwksp_alloc_objects;
    ws->isStatic = isStatic;
    ZSTD_cwksp_clear(ws);
@@ -594,6 +676,11 @@ MEM_STATIC size_t ZSTD_cwksp_create(ZSTD_cwksp* ws, size_t size, ZSTD_customMem
 MEM_STATIC void ZSTD_cwksp_free(ZSTD_cwksp* ws, ZSTD_customMem customMem) {
    void *ptr = ws->workspace;
    DEBUGLOG(4, "cwksp: freeing workspace");
+#if ZSTD_MEMORY_SANITIZER && !defined(ZSTD_MSAN_DONT_POISON_WORKSPACE)
+    if (ptr != NULL && customMem.customFree != NULL) {
+        __msan_unpoison(ptr, ZSTD_cwksp_sizeof(ws));
+    }
+#endif
    ZSTD_memset(ws, 0, sizeof(ZSTD_cwksp));
    ZSTD_customFree(ptr, customMem);
 }
@@ -607,15 +694,6 @@ MEM_STATIC void ZSTD_cwksp_move(ZSTD_cwksp* dst, ZSTD_cwksp* src) {
    ZSTD_memset(src, 0, sizeof(ZSTD_cwksp));
 }

-MEM_STATIC size_t ZSTD_cwksp_sizeof(const ZSTD_cwksp* ws) {
-    return (size_t)((BYTE*)ws->workspaceEnd - (BYTE*)ws->workspace);
-}
-
-MEM_STATIC size_t ZSTD_cwksp_used(const ZSTD_cwksp* ws) {
-    return (size_t)((BYTE*)ws->tableEnd - (BYTE*)ws->workspace)
-         + (size_t)((BYTE*)ws->workspaceEnd - (BYTE*)ws->allocStart);
-}
-
 MEM_STATIC int ZSTD_cwksp_reserve_failed(const ZSTD_cwksp* ws) {
    return ws->allocFailed;
 }
@@ -628,17 +706,11 @@ MEM_STATIC int ZSTD_cwksp_reserve_failed(const ZSTD_cwksp* ws) {
 * Returns if the estimated space needed for a wksp is within an acceptable limit of the
 * actual amount of space used.
 */
-MEM_STATIC int ZSTD_cwksp_estimated_space_within_bounds(const ZSTD_cwksp* const ws,
-                                                        size_t const estimatedSpace, int resizedWorkspace) {
-    if (resizedWorkspace) {
-        /* Resized/newly allocated wksp should have exact bounds */
-        return ZSTD_cwksp_used(ws) == estimatedSpace;
-    } else {
-        /* Due to alignment, when reusing a workspace, we can actually consume 63 fewer or more bytes
-         * than estimatedSpace. See the comments in zstd_cwksp.h for details.
-         */
-        return (ZSTD_cwksp_used(ws) >= estimatedSpace - 63) && (ZSTD_cwksp_used(ws) <= estimatedSpace + 63);
-    }
+MEM_STATIC int ZSTD_cwksp_estimated_space_within_bounds(const ZSTD_cwksp *const ws, size_t const estimatedSpace) {
+    /* We have an alignment space between objects and tables between tables and buffers, so we can have up to twice
+     * the alignment bytes difference between estimation and actual usage */
+    return (estimatedSpace - ZSTD_cwksp_slack_space_required()) <= ZSTD_cwksp_used(ws) &&
+           ZSTD_cwksp_used(ws) <= estimatedSpace;
 }


@@ -11,6 +11,8 @@
 #include "zstd_compress_internal.h"
 #include "zstd_double_fast.h"

+#ifndef ZSTD_EXCLUDE_DFAST_BLOCK_COMPRESSOR
+
 static void ZSTD_fillDoubleHashTableForCDict(ZSTD_matchState_t* ms,
                              void const* end, ZSTD_dictTableLoadMethod_e dtlm)
 {
@@ -756,3 +758,5 @@ size_t ZSTD_compressBlock_doubleFast_extDict(
        return ZSTD_compressBlock_doubleFast_extDict_7(ms, seqStore, rep, src, srcSize);
    }
 }
+
+#endif /* ZSTD_EXCLUDE_DFAST_BLOCK_COMPRESSOR */
@@ -18,9 +18,12 @@ extern "C" {
 #include "../common/mem.h"      /* U32 */
 #include "zstd_compress_internal.h"     /* ZSTD_CCtx, size_t */

+#ifndef ZSTD_EXCLUDE_DFAST_BLOCK_COMPRESSOR
+
 void ZSTD_fillDoubleHashTable(ZSTD_matchState_t* ms,
                              void const* end, ZSTD_dictTableLoadMethod_e dtlm,
                              ZSTD_tableFillPurpose_e tfp);
+
 size_t ZSTD_compressBlock_doubleFast(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize);
@@ -31,6 +34,14 @@ size_t ZSTD_compressBlock_doubleFast_extDict(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize);

+#define ZSTD_COMPRESSBLOCK_DOUBLEFAST ZSTD_compressBlock_doubleFast
+#define ZSTD_COMPRESSBLOCK_DOUBLEFAST_DICTMATCHSTATE ZSTD_compressBlock_doubleFast_dictMatchState
+#define ZSTD_COMPRESSBLOCK_DOUBLEFAST_EXTDICT ZSTD_compressBlock_doubleFast_extDict
+#else
+#define ZSTD_COMPRESSBLOCK_DOUBLEFAST NULL
+#define ZSTD_COMPRESSBLOCK_DOUBLEFAST_DICTMATCHSTATE NULL
+#define ZSTD_COMPRESSBLOCK_DOUBLEFAST_EXTDICT NULL
+#endif /* ZSTD_EXCLUDE_DFAST_BLOCK_COMPRESSOR */

 #if defined (__cplusplus)
 }
@@ -12,6 +12,13 @@
 #include "zstd_lazy.h"
 #include "../common/bits.h" /* ZSTD_countTrailingZeros64 */

+#if !defined(ZSTD_EXCLUDE_GREEDY_BLOCK_COMPRESSOR) \
+ || !defined(ZSTD_EXCLUDE_LAZY_BLOCK_COMPRESSOR) \
+ || !defined(ZSTD_EXCLUDE_LAZY2_BLOCK_COMPRESSOR) \
+ || !defined(ZSTD_EXCLUDE_BTLAZY2_BLOCK_COMPRESSOR)
+
+#define kLazySkippingStep 8
+

 /*-*************************************
 *  Binary Tree search
@@ -618,7 +625,7 @@ size_t ZSTD_dedicatedDictSearch_lazy_search(size_t* offsetPtr, size_t ml, U32 nb
 FORCE_INLINE_TEMPLATE U32 ZSTD_insertAndFindFirstIndex_internal(
                        ZSTD_matchState_t* ms,
                        const ZSTD_compressionParameters* const cParams,
-                        const BYTE* ip, U32 const mls)
+                        const BYTE* ip, U32 const mls, U32 const lazySkipping)
 {
    U32* const hashTable  = ms->hashTable;
    const U32 hashLog = cParams->hashLog;
@@ -633,6 +640,9 @@ FORCE_INLINE_TEMPLATE U32 ZSTD_insertAndFindFirstIndex_internal(
        NEXT_IN_CHAIN(idx, chainMask) = hashTable[h];
        hashTable[h] = idx;
        idx++;
+        /* Stop inserting every position when in the lazy skipping mode. */
+        if (lazySkipping)
+            break;
    }

    ms->nextToUpdate = target;
@@ -641,7 +651,7 @@ FORCE_INLINE_TEMPLATE U32 ZSTD_insertAndFindFirstIndex_internal(

 U32 ZSTD_insertAndFindFirstIndex(ZSTD_matchState_t* ms, const BYTE* ip) {
    const ZSTD_compressionParameters* const cParams = &ms->cParams;
-    return ZSTD_insertAndFindFirstIndex_internal(ms, cParams, ip, ms->cParams.minMatch);
+    return ZSTD_insertAndFindFirstIndex_internal(ms, cParams, ip, ms->cParams.minMatch, /* lazySkipping*/ 0);
 }

 /* inlining is important to hardwire a hot branch (template emulation) */
@@ -685,7 +695,7 @@ size_t ZSTD_HcFindBestMatch(
    }

    /* HC4 match finder */
-    matchIndex = ZSTD_insertAndFindFirstIndex_internal(ms, cParams, ip, mls);
+    matchIndex = ZSTD_insertAndFindFirstIndex_internal(ms, cParams, ip, mls, ms->lazySkipping);

    for ( ; (matchIndex>=lowLimit) & (nbAttempts>0) ; nbAttempts--) {
        size_t currentMl=0;
@@ -758,8 +768,6 @@ size_t ZSTD_HcFindBestMatch(
 * (SIMD) Row-based matchfinder
 ***********************************/
 /* Constants for row-based hash */
-#define ZSTD_ROW_HASH_TAG_OFFSET 16     /* byte offset of hashes in the match state's tagTable from the beginning of a row */
-#define ZSTD_ROW_HASH_TAG_BITS 8        /* nb bits to use for the tag */
 #define ZSTD_ROW_HASH_TAG_MASK ((1u << ZSTD_ROW_HASH_TAG_BITS) - 1)
 #define ZSTD_ROW_HASH_MAX_ENTRIES 64    /* absolute maximum number of entries per row, for all configurations */

@@ -775,39 +783,15 @@ MEM_STATIC U32 ZSTD_VecMask_next(ZSTD_VecMask val) {
    return ZSTD_countTrailingZeros64(val);
 }

-/* ZSTD_rotateRight_*():
- * Rotates a bitfield to the right by "count" bits.
- * https://en.wikipedia.org/w/index.php?title=Circular_shift&oldid=991635599#Implementing_circular_shifts
- */
-FORCE_INLINE_TEMPLATE
-U64 ZSTD_rotateRight_U64(U64 const value, U32 count) {
-    assert(count < 64);
-    count &= 0x3F; /* for fickle pattern recognition */
-    return (value >> count) | (U64)(value << ((0U - count) & 0x3F));
-}
-
-FORCE_INLINE_TEMPLATE
-U32 ZSTD_rotateRight_U32(U32 const value, U32 count) {
-    assert(count < 32);
-    count &= 0x1F; /* for fickle pattern recognition */
-    return (value >> count) | (U32)(value << ((0U - count) & 0x1F));
-}
-
-FORCE_INLINE_TEMPLATE
-U16 ZSTD_rotateRight_U16(U16 const value, U32 count) {
-    assert(count < 16);
-    count &= 0x0F; /* for fickle pattern recognition */
-    return (value >> count) | (U16)(value << ((0U - count) & 0x0F));
-}
-
 /* ZSTD_row_nextIndex():
 * Returns the next index to insert at within a tagTable row, and updates the "head"
- * value to reflect the update. Essentially cycles backwards from [0, {entries per row})
+ * value to reflect the update. Essentially cycles backwards from [1, {entries per row})
 */
 FORCE_INLINE_TEMPLATE U32 ZSTD_row_nextIndex(BYTE* const tagRow, U32 const rowMask) {
-  U32 const next = (*tagRow - 1) & rowMask;
-  *tagRow = (BYTE)next;
-  return next;
+    U32 next = (*tagRow-1) & rowMask;
+    next += (next == 0) ? rowMask : 0; /* skip first position */
+    *tagRow = (BYTE)next;
+    return next;
 }

 /* ZSTD_isAligned():
@@ -821,7 +805,7 @@ MEM_STATIC int ZSTD_isAligned(void const* ptr, size_t align) {
 /* ZSTD_row_prefetch():
 * Performs prefetching for the hashTable and tagTable at a given row.
 */
-FORCE_INLINE_TEMPLATE void ZSTD_row_prefetch(U32 const* hashTable, U16 const* tagTable, U32 const relRow, U32 const rowLog) {
+FORCE_INLINE_TEMPLATE void ZSTD_row_prefetch(U32 const* hashTable, BYTE const* tagTable, U32 const relRow, U32 const rowLog) {
    PREFETCH_L1(hashTable + relRow);
    if (rowLog >= 5) {
        PREFETCH_L1(hashTable + relRow + 16);
@@ -845,13 +829,13 @@ FORCE_INLINE_TEMPLATE void ZSTD_row_fillHashCache(ZSTD_matchState_t* ms, const B
                                   U32 idx, const BYTE* const iLimit)
 {
    U32 const* const hashTable = ms->hashTable;
-    U16 const* const tagTable = ms->tagTable;
+    BYTE const* const tagTable = ms->tagTable;
    U32 const hashLog = ms->rowHashLog;
    U32 const maxElemsToPrefetch = (base + idx) > iLimit ? 0 : (U32)(iLimit - (base + idx) + 1);
    U32 const lim = idx + MIN(ZSTD_ROW_HASH_CACHE_SIZE, maxElemsToPrefetch);

    for (; idx < lim; ++idx) {
-        U32 const hash = (U32)ZSTD_hashPtr(base + idx, hashLog + ZSTD_ROW_HASH_TAG_BITS, mls);
+        U32 const hash = (U32)ZSTD_hashPtrSalted(base + idx, hashLog + ZSTD_ROW_HASH_TAG_BITS, mls, ms->hashSalt);
        U32 const row = (hash >> ZSTD_ROW_HASH_TAG_BITS) << rowLog;
        ZSTD_row_prefetch(hashTable, tagTable, row, rowLog);
        ms->hashCache[idx & ZSTD_ROW_HASH_CACHE_MASK] = hash;
@@ -867,11 +851,12 @@ FORCE_INLINE_TEMPLATE void ZSTD_row_fillHashCache(ZSTD_matchState_t* ms, const B
 * base + idx + ZSTD_ROW_HASH_CACHE_SIZE. Also prefetches the appropriate rows from hashTable and tagTable.
 */
 FORCE_INLINE_TEMPLATE U32 ZSTD_row_nextCachedHash(U32* cache, U32 const* hashTable,
-                                                  U16 const* tagTable, BYTE const* base,
+                                                  BYTE const* tagTable, BYTE const* base,
                                                  U32 idx, U32 const hashLog,
-                                                  U32 const rowLog, U32 const mls)
+                                                  U32 const rowLog, U32 const mls,
+                                                  U64 const hashSalt)
 {
-    U32 const newHash = (U32)ZSTD_hashPtr(base+idx+ZSTD_ROW_HASH_CACHE_SIZE, hashLog + ZSTD_ROW_HASH_TAG_BITS, mls);
+    U32 const newHash = (U32)ZSTD_hashPtrSalted(base+idx+ZSTD_ROW_HASH_CACHE_SIZE, hashLog + ZSTD_ROW_HASH_TAG_BITS, mls, hashSalt);
    U32 const row = (newHash >> ZSTD_ROW_HASH_TAG_BITS) << rowLog;
    ZSTD_row_prefetch(hashTable, tagTable, row, rowLog);
    {   U32 const hash = cache[idx & ZSTD_ROW_HASH_CACHE_MASK];
@@ -889,22 +874,21 @@ FORCE_INLINE_TEMPLATE void ZSTD_row_update_internalImpl(ZSTD_matchState_t* ms,
                                                        U32 const rowMask, U32 const useCache)
 {
    U32* const hashTable = ms->hashTable;
-    U16* const tagTable = ms->tagTable;
+    BYTE* const tagTable = ms->tagTable;
    U32 const hashLog = ms->rowHashLog;
    const BYTE* const base = ms->window.base;

    DEBUGLOG(6, "ZSTD_row_update_internalImpl(): updateStartIdx=%u, updateEndIdx=%u", updateStartIdx, updateEndIdx);
    for (; updateStartIdx < updateEndIdx; ++updateStartIdx) {
-        U32 const hash = useCache ? ZSTD_row_nextCachedHash(ms->hashCache, hashTable, tagTable, base, updateStartIdx, hashLog, rowLog, mls)
-                                  : (U32)ZSTD_hashPtr(base + updateStartIdx, hashLog + ZSTD_ROW_HASH_TAG_BITS, mls);
+        U32 const hash = useCache ? ZSTD_row_nextCachedHash(ms->hashCache, hashTable, tagTable, base, updateStartIdx, hashLog, rowLog, mls, ms->hashSalt)
+                                  : (U32)ZSTD_hashPtrSalted(base + updateStartIdx, hashLog + ZSTD_ROW_HASH_TAG_BITS, mls, ms->hashSalt);
        U32 const relRow = (hash >> ZSTD_ROW_HASH_TAG_BITS) << rowLog;
        U32* const row = hashTable + relRow;
-        BYTE* tagRow = (BYTE*)(tagTable + relRow);  /* Though tagTable is laid out as a table of U16, each tag is only 1 byte.
-                                                       Explicit cast allows us to get exact desired position within each row */
+        BYTE* tagRow = tagTable + relRow;
        U32 const pos = ZSTD_row_nextIndex(tagRow, rowMask);

-        assert(hash == ZSTD_hashPtr(base + updateStartIdx, hashLog + ZSTD_ROW_HASH_TAG_BITS, mls));
-        ((BYTE*)tagRow)[pos + ZSTD_ROW_HASH_TAG_OFFSET] = hash & ZSTD_ROW_HASH_TAG_MASK;
+        assert(hash == ZSTD_hashPtrSalted(base + updateStartIdx, hashLog + ZSTD_ROW_HASH_TAG_BITS, mls, ms->hashSalt));
+        tagRow[pos] = hash & ZSTD_ROW_HASH_TAG_MASK;
        row[pos] = updateStartIdx;
    }
 }
@@ -1060,7 +1044,7 @@ ZSTD_row_getNEONMask(const U32 rowEntries, const BYTE* const src, const BYTE tag
 FORCE_INLINE_TEMPLATE ZSTD_VecMask
 ZSTD_row_getMatchMask(const BYTE* const tagRow, const BYTE tag, const U32 headGrouped, const U32 rowEntries)
 {
-    const BYTE* const src = tagRow + ZSTD_ROW_HASH_TAG_OFFSET;
+    const BYTE* const src = tagRow;
    assert((rowEntries == 16) || (rowEntries == 32) || rowEntries == 64);
    assert(rowEntries <= ZSTD_ROW_HASH_MAX_ENTRIES);
    assert(ZSTD_row_matchMaskGroupWidth(rowEntries) * rowEntries <= sizeof(ZSTD_VecMask) * 8);
@@ -1145,7 +1129,7 @@ size_t ZSTD_RowFindBestMatch(
                        const U32 rowLog)
 {
    U32* const hashTable = ms->hashTable;
-    U16* const tagTable = ms->tagTable;
+    BYTE* const tagTable = ms->tagTable;
    U32* const hashCache = ms->hashCache;
    const U32 hashLog = ms->rowHashLog;
    const ZSTD_compressionParameters* const cParams = &ms->cParams;
@@ -1164,8 +1148,10 @@ size_t ZSTD_RowFindBestMatch(
    const U32 rowMask = rowEntries - 1;
    const U32 cappedSearchLog = MIN(cParams->searchLog, rowLog); /* nb of searches is capped at nb entries per row */
    const U32 groupWidth = ZSTD_row_matchMaskGroupWidth(rowEntries);
+    const U64 hashSalt = ms->hashSalt;
    U32 nbAttempts = 1U << cappedSearchLog;
    size_t ml=4-1;
+    U32 hash;

    /* DMS/DDS variables that may be referenced laster */
    const ZSTD_matchState_t* const dms = ms->dictMatchState;
@@ -1189,7 +1175,7 @@ size_t ZSTD_RowFindBestMatch(
    if (dictMode == ZSTD_dictMatchState) {
        /* Prefetch DMS rows */
        U32* const dmsHashTable = dms->hashTable;
-        U16* const dmsTagTable = dms->tagTable;
+        BYTE* const dmsTagTable = dms->tagTable;
        U32 const dmsHash = (U32)ZSTD_hashPtr(ip, dms->rowHashLog + ZSTD_ROW_HASH_TAG_BITS, mls);
        U32 const dmsRelRow = (dmsHash >> ZSTD_ROW_HASH_TAG_BITS) << rowLog;
        dmsTag = dmsHash & ZSTD_ROW_HASH_TAG_MASK;
@@ -1199,9 +1185,19 @@ size_t ZSTD_RowFindBestMatch(
    }

    /* Update the hashTable and tagTable up to (but not including) ip */
-    ZSTD_row_update_internal(ms, ip, mls, rowLog, rowMask, 1 /* useCache */);
+    if (!ms->lazySkipping) {
+        ZSTD_row_update_internal(ms, ip, mls, rowLog, rowMask, 1 /* useCache */);
+        hash = ZSTD_row_nextCachedHash(hashCache, hashTable, tagTable, base, curr, hashLog, rowLog, mls, hashSalt);
+    } else {
+        /* Stop inserting every position when in the lazy skipping mode.
+         * The hash cache is also not kept up to date in this mode.
+         */
+        hash = (U32)ZSTD_hashPtrSalted(ip, hashLog + ZSTD_ROW_HASH_TAG_BITS, mls, hashSalt);
+        ms->nextToUpdate = curr;
+    }
+    ms->hashSaltEntropy += hash; /* collect salt entropy */
+
    {   /* Get the hash for ip, compute the appropriate row */
-        U32 const hash = ZSTD_row_nextCachedHash(hashCache, hashTable, tagTable, base, curr, hashLog, rowLog, mls);
        U32 const relRow = (hash >> ZSTD_ROW_HASH_TAG_BITS) << rowLog;
        U32 const tag = hash & ZSTD_ROW_HASH_TAG_MASK;
        U32* const row = hashTable + relRow;
@@ -1213,9 +1209,10 @@ size_t ZSTD_RowFindBestMatch(
        ZSTD_VecMask matches = ZSTD_row_getMatchMask(tagRow, (BYTE)tag, headGrouped, rowEntries);

        /* Cycle through the matches and prefetch */
-        for (; (matches > 0) && (nbAttempts > 0); --nbAttempts, matches &= (matches - 1)) {
+        for (; (matches > 0) && (nbAttempts > 0); matches &= (matches - 1)) {
            U32 const matchPos = ((headGrouped + ZSTD_VecMask_next(matches)) / groupWidth) & rowMask;
            U32 const matchIndex = row[matchPos];
+            if(matchPos == 0) continue;
            assert(numMatches < rowEntries);
            if (matchIndex < lowLimit)
                break;
@@ -1225,13 +1222,14 @@ size_t ZSTD_RowFindBestMatch(
                PREFETCH_L1(dictBase + matchIndex);
            }
            matchBuffer[numMatches++] = matchIndex;
+            --nbAttempts;
        }

        /* Speed opt: insert current byte into hashtable too. This allows us to avoid one iteration of the loop
           in ZSTD_row_update_internal() at the next search. */
        {
            U32 const pos = ZSTD_row_nextIndex(tagRow, rowMask);
-            tagRow[pos + ZSTD_ROW_HASH_TAG_OFFSET] = (BYTE)tag;
+            tagRow[pos] = (BYTE)tag;
            row[pos] = ms->nextToUpdate++;
        }

@@ -1282,13 +1280,15 @@ size_t ZSTD_RowFindBestMatch(
            size_t currMatch = 0;
            ZSTD_VecMask matches = ZSTD_row_getMatchMask(dmsTagRow, (BYTE)dmsTag, headGrouped, rowEntries);

-            for (; (matches > 0) && (nbAttempts > 0); --nbAttempts, matches &= (matches - 1)) {
+            for (; (matches > 0) && (nbAttempts > 0); matches &= (matches - 1)) {
                U32 const matchPos = ((headGrouped + ZSTD_VecMask_next(matches)) / groupWidth) & rowMask;
                U32 const matchIndex = dmsRow[matchPos];
+                if(matchPos == 0) continue;
                if (matchIndex < dmsLowestIndex)
                    break;
                PREFETCH_L1(dmsBase + matchIndex);
                matchBuffer[numMatches++] = matchIndex;
+                --nbAttempts;
            }

            /* Return the longest match */
@@ -1545,10 +1545,11 @@ ZSTD_compressBlock_lazy_generic(
        assert(offset_2 <= dictAndPrefixLength);
    }

+    /* Reset the lazy skipping state */
+    ms->lazySkipping = 0;
+
    if (searchMethod == search_rowHash) {
-        ZSTD_row_fillHashCache(ms, base, rowLog,
-                            MIN(ms->cParams.minMatch, 6 /* mls caps out at 6 */),
-                            ms->nextToUpdate, ilimit);
+        ZSTD_row_fillHashCache(ms, base, rowLog, mls, ms->nextToUpdate, ilimit);
    }

    /* Match Loop */
@@ -1592,7 +1593,16 @@ ZSTD_compressBlock_lazy_generic(
        }

        if (matchLength < 4) {
-            ip += ((ip-anchor) >> kSearchStrength) + 1;   /* jump faster over incompressible sections */
+            size_t const step = ((size_t)(ip-anchor) >> kSearchStrength) + 1;   /* jump faster over incompressible sections */;
+            ip += step;
+            /* Enter the lazy skipping mode once we are skipping more than 8 bytes at a time.
+             * In this mode we stop inserting every position into our tables, and only insert
+             * positions that we search, which is one in step positions.
+             * The exact cutoff is flexible, I've just chosen a number that is reasonably high,
+             * so we minimize the compression ratio loss in "normal" scenarios. This mode gets
+             * triggered once we've gone 2KB without finding any matches.
+             */
+            ms->lazySkipping = step > kLazySkippingStep;
            continue;
        }

@@ -1696,6 +1706,13 @@ _storeSequence:
            ZSTD_storeSeq(seqStore, litLength, anchor, iend, (U32)offBase, matchLength);
            anchor = ip = start + matchLength;
        }
+        if (ms->lazySkipping) {
+            /* We've found a match, disable lazy skipping mode, and refill the hash cache. */
+            if (searchMethod == search_rowHash) {
+                ZSTD_row_fillHashCache(ms, base, rowLog, mls, ms->nextToUpdate, ilimit);
+            }
+            ms->lazySkipping = 0;
+        }

        /* check immediate repcode */
        if (isDxS) {
@@ -1742,29 +1759,10 @@ _storeSequence:
    /* Return the last literals size */
    return (size_t)(iend - anchor);
 }
+#endif /* build exclusions */


-size_t ZSTD_compressBlock_btlazy2(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize)
-{
-    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_binaryTree, 2, ZSTD_noDict);
-}
-
-size_t ZSTD_compressBlock_lazy2(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize)
-{
-    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 2, ZSTD_noDict);
-}
-
-size_t ZSTD_compressBlock_lazy(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize)
-{
-    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 1, ZSTD_noDict);
-}
-
+#ifndef ZSTD_EXCLUDE_GREEDY_BLOCK_COMPRESSOR
 size_t ZSTD_compressBlock_greedy(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize)
@@ -1772,27 +1770,6 @@ size_t ZSTD_compressBlock_greedy(
    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 0, ZSTD_noDict);
 }

-size_t ZSTD_compressBlock_btlazy2_dictMatchState(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize)
-{
-    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_binaryTree, 2, ZSTD_dictMatchState);
-}
-
-size_t ZSTD_compressBlock_lazy2_dictMatchState(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize)
-{
-    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 2, ZSTD_dictMatchState);
-}
-
-size_t ZSTD_compressBlock_lazy_dictMatchState(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize)
-{
-    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 1, ZSTD_dictMatchState);
-}
-
 size_t ZSTD_compressBlock_greedy_dictMatchState(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize)
@@ -1800,21 +1777,6 @@ size_t ZSTD_compressBlock_greedy_dictMatchState(
    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 0, ZSTD_dictMatchState);
 }

-
-size_t ZSTD_compressBlock_lazy2_dedicatedDictSearch(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize)
-{
-    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 2, ZSTD_dedicatedDictSearch);
-}
-
-size_t ZSTD_compressBlock_lazy_dedicatedDictSearch(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize)
-{
-    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 1, ZSTD_dedicatedDictSearch);
-}
-
 size_t ZSTD_compressBlock_greedy_dedicatedDictSearch(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize)
@@ -1822,21 +1784,6 @@ size_t ZSTD_compressBlock_greedy_dedicatedDictSearch(
    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 0, ZSTD_dedicatedDictSearch);
 }

-/* Row-based matchfinder */
-size_t ZSTD_compressBlock_lazy2_row(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize)
-{
-    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_rowHash, 2, ZSTD_noDict);
-}
-
-size_t ZSTD_compressBlock_lazy_row(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize)
-{
-    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_rowHash, 1, ZSTD_noDict);
-}
-
 size_t ZSTD_compressBlock_greedy_row(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize)
@@ -1844,11 +1791,48 @@ size_t ZSTD_compressBlock_greedy_row(
    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_rowHash, 0, ZSTD_noDict);
 }

-size_t ZSTD_compressBlock_lazy2_dictMatchState_row(
+size_t ZSTD_compressBlock_greedy_dictMatchState_row(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize)
 {
-    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_rowHash, 2, ZSTD_dictMatchState);
+    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_rowHash, 0, ZSTD_dictMatchState);
+}
+
+size_t ZSTD_compressBlock_greedy_dedicatedDictSearch_row(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize)
+{
+    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_rowHash, 0, ZSTD_dedicatedDictSearch);
+}
+#endif
+
+#ifndef ZSTD_EXCLUDE_LAZY_BLOCK_COMPRESSOR
+size_t ZSTD_compressBlock_lazy(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize)
+{
+    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 1, ZSTD_noDict);
+}
+
+size_t ZSTD_compressBlock_lazy_dictMatchState(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize)
+{
+    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 1, ZSTD_dictMatchState);
+}
+
+size_t ZSTD_compressBlock_lazy_dedicatedDictSearch(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize)
+{
+    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 1, ZSTD_dedicatedDictSearch);
+}
+
+size_t ZSTD_compressBlock_lazy_row(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize)
+{
+    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_rowHash, 1, ZSTD_noDict);
 }

 size_t ZSTD_compressBlock_lazy_dictMatchState_row(
@@ -1858,13 +1842,49 @@ size_t ZSTD_compressBlock_lazy_dictMatchState_row(
    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_rowHash, 1, ZSTD_dictMatchState);
 }

-size_t ZSTD_compressBlock_greedy_dictMatchState_row(
+size_t ZSTD_compressBlock_lazy_dedicatedDictSearch_row(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize)
 {
-    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_rowHash, 0, ZSTD_dictMatchState);
+    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_rowHash, 1, ZSTD_dedicatedDictSearch);
+}
+#endif
+
+#ifndef ZSTD_EXCLUDE_LAZY2_BLOCK_COMPRESSOR
+size_t ZSTD_compressBlock_lazy2(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize)
+{
+    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 2, ZSTD_noDict);
 }

+size_t ZSTD_compressBlock_lazy2_dictMatchState(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize)
+{
+    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 2, ZSTD_dictMatchState);
+}
+
+size_t ZSTD_compressBlock_lazy2_dedicatedDictSearch(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize)
+{
+    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 2, ZSTD_dedicatedDictSearch);
+}
+
+size_t ZSTD_compressBlock_lazy2_row(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize)
+{
+    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_rowHash, 2, ZSTD_noDict);
+}
+
+size_t ZSTD_compressBlock_lazy2_dictMatchState_row(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize)
+{
+    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_rowHash, 2, ZSTD_dictMatchState);
+}

 size_t ZSTD_compressBlock_lazy2_dedicatedDictSearch_row(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
@@ -1872,21 +1892,28 @@ size_t ZSTD_compressBlock_lazy2_dedicatedDictSearch_row(
 {
    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_rowHash, 2, ZSTD_dedicatedDictSearch);
 }
+#endif

-size_t ZSTD_compressBlock_lazy_dedicatedDictSearch_row(
+#ifndef ZSTD_EXCLUDE_BTLAZY2_BLOCK_COMPRESSOR
+size_t ZSTD_compressBlock_btlazy2(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize)
 {
-    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_rowHash, 1, ZSTD_dedicatedDictSearch);
+    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_binaryTree, 2, ZSTD_noDict);
 }

-size_t ZSTD_compressBlock_greedy_dedicatedDictSearch_row(
+size_t ZSTD_compressBlock_btlazy2_dictMatchState(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize)
 {
-    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_rowHash, 0, ZSTD_dedicatedDictSearch);
+    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_binaryTree, 2, ZSTD_dictMatchState);
 }
+#endif

+#if !defined(ZSTD_EXCLUDE_GREEDY_BLOCK_COMPRESSOR) \
+ || !defined(ZSTD_EXCLUDE_LAZY_BLOCK_COMPRESSOR) \
+ || !defined(ZSTD_EXCLUDE_LAZY2_BLOCK_COMPRESSOR) \
+ || !defined(ZSTD_EXCLUDE_BTLAZY2_BLOCK_COMPRESSOR)
 FORCE_INLINE_TEMPLATE
 size_t ZSTD_compressBlock_lazy_extDict_generic(
                        ZSTD_matchState_t* ms, seqStore_t* seqStore,
@@ -1913,12 +1940,13 @@ size_t ZSTD_compressBlock_lazy_extDict_generic(

    DEBUGLOG(5, "ZSTD_compressBlock_lazy_extDict_generic (searchFunc=%u)", (U32)searchMethod);

+    /* Reset the lazy skipping state */
+    ms->lazySkipping = 0;
+
    /* init */
    ip += (ip == prefixStart);
    if (searchMethod == search_rowHash) {
-        ZSTD_row_fillHashCache(ms, base, rowLog,
-                               MIN(ms->cParams.minMatch, 6 /* mls caps out at 6 */),
-                               ms->nextToUpdate, ilimit);
+        ZSTD_row_fillHashCache(ms, base, rowLog, mls, ms->nextToUpdate, ilimit);
    }

    /* Match Loop */
@@ -1956,7 +1984,16 @@ size_t ZSTD_compressBlock_lazy_extDict_generic(
        }

        if (matchLength < 4) {
-            ip += ((ip-anchor) >> kSearchStrength) + 1;   /* jump faster over incompressible sections */
+            size_t const step = ((size_t)(ip-anchor) >> kSearchStrength);
+            ip += step + 1;   /* jump faster over incompressible sections */
+            /* Enter the lazy skipping mode once we are skipping more than 8 bytes at a time.
+             * In this mode we stop inserting every position into our tables, and only insert
+             * positions that we search, which is one in step positions.
+             * The exact cutoff is flexible, I've just chosen a number that is reasonably high,
+             * so we minimize the compression ratio loss in "normal" scenarios. This mode gets
+             * triggered once we've gone 2KB without finding any matches.
+             */
+            ms->lazySkipping = step > kLazySkippingStep;
            continue;
        }

@@ -2042,6 +2079,13 @@ _storeSequence:
            ZSTD_storeSeq(seqStore, litLength, anchor, iend, (U32)offBase, matchLength);
            anchor = ip = start + matchLength;
        }
+        if (ms->lazySkipping) {
+            /* We've found a match, disable lazy skipping mode, and refill the hash cache. */
+            if (searchMethod == search_rowHash) {
+                ZSTD_row_fillHashCache(ms, base, rowLog, mls, ms->nextToUpdate, ilimit);
+            }
+            ms->lazySkipping = 0;
+        }

        /* check immediate repcode */
        while (ip <= ilimit) {
@@ -2072,8 +2116,9 @@ _storeSequence:
    /* Return the last literals size */
    return (size_t)(iend - anchor);
 }
+#endif /* build exclusions */

-
+#ifndef ZSTD_EXCLUDE_GREEDY_BLOCK_COMPRESSOR
 size_t ZSTD_compressBlock_greedy_extDict(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize)
@@ -2081,6 +2126,15 @@ size_t ZSTD_compressBlock_greedy_extDict(
    return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 0);
 }

+size_t ZSTD_compressBlock_greedy_extDict_row(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize)
+{
+    return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, search_rowHash, 0);
+}
+#endif
+
+#ifndef ZSTD_EXCLUDE_LAZY_BLOCK_COMPRESSOR
 size_t ZSTD_compressBlock_lazy_extDict(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize)
@@ -2089,29 +2143,6 @@ size_t ZSTD_compressBlock_lazy_extDict(
    return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 1);
 }

-size_t ZSTD_compressBlock_lazy2_extDict(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize)
-
-{
-    return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 2);
-}
-
-size_t ZSTD_compressBlock_btlazy2_extDict(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize)
-
-{
-    return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, search_binaryTree, 2);
-}
-
-size_t ZSTD_compressBlock_greedy_extDict_row(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize)
-{
-    return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, search_rowHash, 0);
-}
-
 size_t ZSTD_compressBlock_lazy_extDict_row(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize)
@@ -2119,6 +2150,16 @@ size_t ZSTD_compressBlock_lazy_extDict_row(
 {
    return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, search_rowHash, 1);
 }
+#endif
+
+#ifndef ZSTD_EXCLUDE_LAZY2_BLOCK_COMPRESSOR
+size_t ZSTD_compressBlock_lazy2_extDict(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize)
+
+{
+    return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 2);
+}

 size_t ZSTD_compressBlock_lazy2_extDict_row(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
@@ -2126,3 +2167,14 @@ size_t ZSTD_compressBlock_lazy2_extDict_row(
 {
    return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, search_rowHash, 2);
 }
+#endif
+
+#ifndef ZSTD_EXCLUDE_BTLAZY2_BLOCK_COMPRESSOR
+size_t ZSTD_compressBlock_btlazy2_extDict(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize)
+
+{
+    return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, search_binaryTree, 2);
+}
+#endif
@@ -25,98 +25,175 @@ extern "C" {
 */
 #define ZSTD_LAZY_DDSS_BUCKET_LOG 2

+#define ZSTD_ROW_HASH_TAG_BITS 8        /* nb bits to use for the tag */
+
+#if !defined(ZSTD_EXCLUDE_GREEDY_BLOCK_COMPRESSOR) \
+ || !defined(ZSTD_EXCLUDE_LAZY_BLOCK_COMPRESSOR) \
+ || !defined(ZSTD_EXCLUDE_LAZY2_BLOCK_COMPRESSOR) \
+ || !defined(ZSTD_EXCLUDE_BTLAZY2_BLOCK_COMPRESSOR)
 U32 ZSTD_insertAndFindFirstIndex(ZSTD_matchState_t* ms, const BYTE* ip);
 void ZSTD_row_update(ZSTD_matchState_t* const ms, const BYTE* ip);

 void ZSTD_dedicatedDictSearch_lazy_loadDictionary(ZSTD_matchState_t* ms, const BYTE* const ip);

 void ZSTD_preserveUnsortedMark (U32* const table, U32 const size, U32 const reducerValue);  /*! used in ZSTD_reduceIndex(). preemptively increase value of ZSTD_DUBT_UNSORTED_MARK */
+#endif

-size_t ZSTD_compressBlock_btlazy2(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize);
-size_t ZSTD_compressBlock_lazy2(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize);
-size_t ZSTD_compressBlock_lazy(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize);
+#ifndef ZSTD_EXCLUDE_GREEDY_BLOCK_COMPRESSOR
 size_t ZSTD_compressBlock_greedy(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize);
-size_t ZSTD_compressBlock_lazy2_row(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize);
-size_t ZSTD_compressBlock_lazy_row(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize);
 size_t ZSTD_compressBlock_greedy_row(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize);
-
-size_t ZSTD_compressBlock_btlazy2_dictMatchState(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize);
-size_t ZSTD_compressBlock_lazy2_dictMatchState(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize);
-size_t ZSTD_compressBlock_lazy_dictMatchState(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize);
 size_t ZSTD_compressBlock_greedy_dictMatchState(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize);
-size_t ZSTD_compressBlock_lazy2_dictMatchState_row(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize);
-size_t ZSTD_compressBlock_lazy_dictMatchState_row(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize);
 size_t ZSTD_compressBlock_greedy_dictMatchState_row(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize);
-
-size_t ZSTD_compressBlock_lazy2_dedicatedDictSearch(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize);
-size_t ZSTD_compressBlock_lazy_dedicatedDictSearch(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize);
 size_t ZSTD_compressBlock_greedy_dedicatedDictSearch(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize);
-size_t ZSTD_compressBlock_lazy2_dedicatedDictSearch_row(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize);
-size_t ZSTD_compressBlock_lazy_dedicatedDictSearch_row(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize);
 size_t ZSTD_compressBlock_greedy_dedicatedDictSearch_row(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize);
-
 size_t ZSTD_compressBlock_greedy_extDict(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize);
-size_t ZSTD_compressBlock_lazy_extDict(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize);
-size_t ZSTD_compressBlock_lazy2_extDict(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize);
 size_t ZSTD_compressBlock_greedy_extDict_row(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize);
+
+#define ZSTD_COMPRESSBLOCK_GREEDY ZSTD_compressBlock_greedy
+#define ZSTD_COMPRESSBLOCK_GREEDY_ROW ZSTD_compressBlock_greedy_row
+#define ZSTD_COMPRESSBLOCK_GREEDY_DICTMATCHSTATE ZSTD_compressBlock_greedy_dictMatchState
+#define ZSTD_COMPRESSBLOCK_GREEDY_DICTMATCHSTATE_ROW ZSTD_compressBlock_greedy_dictMatchState_row
+#define ZSTD_COMPRESSBLOCK_GREEDY_DEDICATEDDICTSEARCH ZSTD_compressBlock_greedy_dedicatedDictSearch
+#define ZSTD_COMPRESSBLOCK_GREEDY_DEDICATEDDICTSEARCH_ROW ZSTD_compressBlock_greedy_dedicatedDictSearch_row
+#define ZSTD_COMPRESSBLOCK_GREEDY_EXTDICT ZSTD_compressBlock_greedy_extDict
+#define ZSTD_COMPRESSBLOCK_GREEDY_EXTDICT_ROW ZSTD_compressBlock_greedy_extDict_row
+#else
+#define ZSTD_COMPRESSBLOCK_GREEDY NULL
+#define ZSTD_COMPRESSBLOCK_GREEDY_ROW NULL
+#define ZSTD_COMPRESSBLOCK_GREEDY_DICTMATCHSTATE NULL
+#define ZSTD_COMPRESSBLOCK_GREEDY_DICTMATCHSTATE_ROW NULL
+#define ZSTD_COMPRESSBLOCK_GREEDY_DEDICATEDDICTSEARCH NULL
+#define ZSTD_COMPRESSBLOCK_GREEDY_DEDICATEDDICTSEARCH_ROW NULL
+#define ZSTD_COMPRESSBLOCK_GREEDY_EXTDICT NULL
+#define ZSTD_COMPRESSBLOCK_GREEDY_EXTDICT_ROW NULL
+#endif
+
+#ifndef ZSTD_EXCLUDE_LAZY_BLOCK_COMPRESSOR
+size_t ZSTD_compressBlock_lazy(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize);
+size_t ZSTD_compressBlock_lazy_row(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize);
+size_t ZSTD_compressBlock_lazy_dictMatchState(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize);
+size_t ZSTD_compressBlock_lazy_dictMatchState_row(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize);
+size_t ZSTD_compressBlock_lazy_dedicatedDictSearch(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize);
+size_t ZSTD_compressBlock_lazy_dedicatedDictSearch_row(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize);
+size_t ZSTD_compressBlock_lazy_extDict(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize);
 size_t ZSTD_compressBlock_lazy_extDict_row(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize);
+
+#define ZSTD_COMPRESSBLOCK_LAZY ZSTD_compressBlock_lazy
+#define ZSTD_COMPRESSBLOCK_LAZY_ROW ZSTD_compressBlock_lazy_row
+#define ZSTD_COMPRESSBLOCK_LAZY_DICTMATCHSTATE ZSTD_compressBlock_lazy_dictMatchState
+#define ZSTD_COMPRESSBLOCK_LAZY_DICTMATCHSTATE_ROW ZSTD_compressBlock_lazy_dictMatchState_row
+#define ZSTD_COMPRESSBLOCK_LAZY_DEDICATEDDICTSEARCH ZSTD_compressBlock_lazy_dedicatedDictSearch
+#define ZSTD_COMPRESSBLOCK_LAZY_DEDICATEDDICTSEARCH_ROW ZSTD_compressBlock_lazy_dedicatedDictSearch_row
+#define ZSTD_COMPRESSBLOCK_LAZY_EXTDICT ZSTD_compressBlock_lazy_extDict
+#define ZSTD_COMPRESSBLOCK_LAZY_EXTDICT_ROW ZSTD_compressBlock_lazy_extDict_row
+#else
+#define ZSTD_COMPRESSBLOCK_LAZY NULL
+#define ZSTD_COMPRESSBLOCK_LAZY_ROW NULL
+#define ZSTD_COMPRESSBLOCK_LAZY_DICTMATCHSTATE NULL
+#define ZSTD_COMPRESSBLOCK_LAZY_DICTMATCHSTATE_ROW NULL
+#define ZSTD_COMPRESSBLOCK_LAZY_DEDICATEDDICTSEARCH NULL
+#define ZSTD_COMPRESSBLOCK_LAZY_DEDICATEDDICTSEARCH_ROW NULL
+#define ZSTD_COMPRESSBLOCK_LAZY_EXTDICT NULL
+#define ZSTD_COMPRESSBLOCK_LAZY_EXTDICT_ROW NULL
+#endif
+
+#ifndef ZSTD_EXCLUDE_LAZY2_BLOCK_COMPRESSOR
+size_t ZSTD_compressBlock_lazy2(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize);
+size_t ZSTD_compressBlock_lazy2_row(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize);
+size_t ZSTD_compressBlock_lazy2_dictMatchState(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize);
+size_t ZSTD_compressBlock_lazy2_dictMatchState_row(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize);
+size_t ZSTD_compressBlock_lazy2_dedicatedDictSearch(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize);
+size_t ZSTD_compressBlock_lazy2_dedicatedDictSearch_row(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize);
+size_t ZSTD_compressBlock_lazy2_extDict(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize);
 size_t ZSTD_compressBlock_lazy2_extDict_row(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize);
+
+#define ZSTD_COMPRESSBLOCK_LAZY2 ZSTD_compressBlock_lazy2
+#define ZSTD_COMPRESSBLOCK_LAZY2_ROW ZSTD_compressBlock_lazy2_row
+#define ZSTD_COMPRESSBLOCK_LAZY2_DICTMATCHSTATE ZSTD_compressBlock_lazy2_dictMatchState
+#define ZSTD_COMPRESSBLOCK_LAZY2_DICTMATCHSTATE_ROW ZSTD_compressBlock_lazy2_dictMatchState_row
+#define ZSTD_COMPRESSBLOCK_LAZY2_DEDICATEDDICTSEARCH ZSTD_compressBlock_lazy2_dedicatedDictSearch
+#define ZSTD_COMPRESSBLOCK_LAZY2_DEDICATEDDICTSEARCH_ROW ZSTD_compressBlock_lazy2_dedicatedDictSearch_row
+#define ZSTD_COMPRESSBLOCK_LAZY2_EXTDICT ZSTD_compressBlock_lazy2_extDict
+#define ZSTD_COMPRESSBLOCK_LAZY2_EXTDICT_ROW ZSTD_compressBlock_lazy2_extDict_row
+#else
+#define ZSTD_COMPRESSBLOCK_LAZY2 NULL
+#define ZSTD_COMPRESSBLOCK_LAZY2_ROW NULL
+#define ZSTD_COMPRESSBLOCK_LAZY2_DICTMATCHSTATE NULL
+#define ZSTD_COMPRESSBLOCK_LAZY2_DICTMATCHSTATE_ROW NULL
+#define ZSTD_COMPRESSBLOCK_LAZY2_DEDICATEDDICTSEARCH NULL
+#define ZSTD_COMPRESSBLOCK_LAZY2_DEDICATEDDICTSEARCH_ROW NULL
+#define ZSTD_COMPRESSBLOCK_LAZY2_EXTDICT NULL
+#define ZSTD_COMPRESSBLOCK_LAZY2_EXTDICT_ROW NULL
+#endif
+
+#ifndef ZSTD_EXCLUDE_BTLAZY2_BLOCK_COMPRESSOR
+size_t ZSTD_compressBlock_btlazy2(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize);
+size_t ZSTD_compressBlock_btlazy2_dictMatchState(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize);
 size_t ZSTD_compressBlock_btlazy2_extDict(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize);
-        
+
+#define ZSTD_COMPRESSBLOCK_BTLAZY2 ZSTD_compressBlock_btlazy2
+#define ZSTD_COMPRESSBLOCK_BTLAZY2_DICTMATCHSTATE ZSTD_compressBlock_btlazy2_dictMatchState
+#define ZSTD_COMPRESSBLOCK_BTLAZY2_EXTDICT ZSTD_compressBlock_btlazy2_extDict
+#else
+#define ZSTD_COMPRESSBLOCK_BTLAZY2 NULL
+#define ZSTD_COMPRESSBLOCK_BTLAZY2_DICTMATCHSTATE NULL
+#define ZSTD_COMPRESSBLOCK_BTLAZY2_EXTDICT NULL
+#endif
+

 #if defined (__cplusplus)
 }
@@ -246,7 +246,11 @@ static size_t ZSTD_ldm_fillFastTables(ZSTD_matchState_t* ms,
        break;

    case ZSTD_dfast:
+#ifndef ZSTD_EXCLUDE_DFAST_BLOCK_COMPRESSOR
        ZSTD_fillDoubleHashTable(ms, iend, ZSTD_dtlm_fast, ZSTD_tfp_forCCtx);
+#else
+        assert(0); /* shouldn't be called: cparams should've been adjusted. */
+#endif
        break;

    case ZSTD_greedy:
@@ -12,6 +12,9 @@
 #include "hist.h"
 #include "zstd_opt.h"

+#if !defined(ZSTD_EXCLUDE_BTLAZY2_BLOCK_COMPRESSOR) \
+ || !defined(ZSTD_EXCLUDE_BTOPT_BLOCK_COMPRESSOR) \
+ || !defined(ZSTD_EXCLUDE_BTULTRA_BLOCK_COMPRESSOR)

 #define ZSTD_LITFREQ_ADD    2   /* scaling factor for litFreq, so that frequencies adapt faster to new stats */
 #define ZSTD_MAX_PRICE     (1<<30)
@@ -1086,6 +1089,8 @@ ZSTD_compressBlock_opt_generic(ZSTD_matchState_t* ms,
    ZSTD_optimal_t lastSequence;
    ZSTD_optLdm_t optLdm;

+    ZSTD_memset(&lastSequence, 0, sizeof(ZSTD_optimal_t));
+
    optLdm.seqStore = ms->ldmSeqStore ? *ms->ldmSeqStore : kNullRawSeqStore;
    optLdm.endPosInBlock = optLdm.startPosInBlock = optLdm.offset = 0;
    ZSTD_opt_getNextMatchAndUpdateSeqStore(&optLdm, (U32)(ip-istart), (U32)(iend-ip));
@@ -1344,21 +1349,27 @@ _shortestPath:   /* cur, last_pos, best_mlen, best_off have to be set */
    /* Return the last literals size */
    return (size_t)(iend - anchor);
 }
+#endif /* build exclusions */

+#ifndef ZSTD_EXCLUDE_BTOPT_BLOCK_COMPRESSOR
 static size_t ZSTD_compressBlock_opt0(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        const void* src, size_t srcSize, const ZSTD_dictMode_e dictMode)
 {
    return ZSTD_compressBlock_opt_generic(ms, seqStore, rep, src, srcSize, 0 /* optLevel */, dictMode);
 }
+#endif

+#ifndef ZSTD_EXCLUDE_BTULTRA_BLOCK_COMPRESSOR
 static size_t ZSTD_compressBlock_opt2(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        const void* src, size_t srcSize, const ZSTD_dictMode_e dictMode)
 {
    return ZSTD_compressBlock_opt_generic(ms, seqStore, rep, src, srcSize, 2 /* optLevel */, dictMode);
 }
+#endif

+#ifndef ZSTD_EXCLUDE_BTOPT_BLOCK_COMPRESSOR
 size_t ZSTD_compressBlock_btopt(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        const void* src, size_t srcSize)
@@ -1366,10 +1377,12 @@ size_t ZSTD_compressBlock_btopt(
    DEBUGLOG(5, "ZSTD_compressBlock_btopt");
    return ZSTD_compressBlock_opt0(ms, seqStore, rep, src, srcSize, ZSTD_noDict);
 }
+#endif




+#ifndef ZSTD_EXCLUDE_BTULTRA_BLOCK_COMPRESSOR
 /* ZSTD_initStats_ultra():
 * make a first compression pass, just to seed stats with more accurate starting values.
 * only works on first block, with no dictionary and no ldm.
@@ -1436,7 +1449,9 @@ size_t ZSTD_compressBlock_btultra2(

    return ZSTD_compressBlock_opt2(ms, seqStore, rep, src, srcSize, ZSTD_noDict);
 }
+#endif

+#ifndef ZSTD_EXCLUDE_BTOPT_BLOCK_COMPRESSOR
 size_t ZSTD_compressBlock_btopt_dictMatchState(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        const void* src, size_t srcSize)
@@ -1444,19 +1459,21 @@ size_t ZSTD_compressBlock_btopt_dictMatchState(
    return ZSTD_compressBlock_opt0(ms, seqStore, rep, src, srcSize, ZSTD_dictMatchState);
 }

-size_t ZSTD_compressBlock_btultra_dictMatchState(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        const void* src, size_t srcSize)
-{
-    return ZSTD_compressBlock_opt2(ms, seqStore, rep, src, srcSize, ZSTD_dictMatchState);
-}
-
 size_t ZSTD_compressBlock_btopt_extDict(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        const void* src, size_t srcSize)
 {
    return ZSTD_compressBlock_opt0(ms, seqStore, rep, src, srcSize, ZSTD_extDict);
 }
+#endif
+
+#ifndef ZSTD_EXCLUDE_BTULTRA_BLOCK_COMPRESSOR
+size_t ZSTD_compressBlock_btultra_dictMatchState(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        const void* src, size_t srcSize)
+{
+    return ZSTD_compressBlock_opt2(ms, seqStore, rep, src, srcSize, ZSTD_dictMatchState);
+}

 size_t ZSTD_compressBlock_btultra_extDict(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
@@ -1464,6 +1481,7 @@ size_t ZSTD_compressBlock_btultra_extDict(
 {
    return ZSTD_compressBlock_opt2(ms, seqStore, rep, src, srcSize, ZSTD_extDict);
 }
+#endif

 /* note : no btultra2 variant for extDict nor dictMatchState,
 * because btultra2 is not meant to work with dictionaries
@@ -17,28 +17,38 @@ extern "C" {

 #include "zstd_compress_internal.h"

+#if !defined(ZSTD_EXCLUDE_BTLAZY2_BLOCK_COMPRESSOR) \
+ || !defined(ZSTD_EXCLUDE_BTOPT_BLOCK_COMPRESSOR) \
+ || !defined(ZSTD_EXCLUDE_BTULTRA_BLOCK_COMPRESSOR)
 /* used in ZSTD_loadDictionaryContent() */
 void ZSTD_updateTree(ZSTD_matchState_t* ms, const BYTE* ip, const BYTE* iend);
+#endif

+#ifndef ZSTD_EXCLUDE_BTOPT_BLOCK_COMPRESSOR
 size_t ZSTD_compressBlock_btopt(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize);
-size_t ZSTD_compressBlock_btultra(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize);
-size_t ZSTD_compressBlock_btultra2(
-        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
-        void const* src, size_t srcSize);
-
-
 size_t ZSTD_compressBlock_btopt_dictMatchState(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize);
-size_t ZSTD_compressBlock_btultra_dictMatchState(
+size_t ZSTD_compressBlock_btopt_extDict(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize);

-size_t ZSTD_compressBlock_btopt_extDict(
+#define ZSTD_COMPRESSBLOCK_BTOPT ZSTD_compressBlock_btopt
+#define ZSTD_COMPRESSBLOCK_BTOPT_DICTMATCHSTATE ZSTD_compressBlock_btopt_dictMatchState
+#define ZSTD_COMPRESSBLOCK_BTOPT_EXTDICT ZSTD_compressBlock_btopt_extDict
+#else
+#define ZSTD_COMPRESSBLOCK_BTOPT NULL
+#define ZSTD_COMPRESSBLOCK_BTOPT_DICTMATCHSTATE NULL
+#define ZSTD_COMPRESSBLOCK_BTOPT_EXTDICT NULL
+#endif
+
+#ifndef ZSTD_EXCLUDE_BTULTRA_BLOCK_COMPRESSOR
+size_t ZSTD_compressBlock_btultra(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize);
+size_t ZSTD_compressBlock_btultra_dictMatchState(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize);
 size_t ZSTD_compressBlock_btultra_extDict(
@@ -48,6 +58,20 @@ size_t ZSTD_compressBlock_btultra_extDict(
        /* note : no btultra2 variant for extDict nor dictMatchState,
         * because btultra2 is not meant to work with dictionaries
         * and is only specific for the first block (no prefix) */
+size_t ZSTD_compressBlock_btultra2(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize);
+
+#define ZSTD_COMPRESSBLOCK_BTULTRA ZSTD_compressBlock_btultra
+#define ZSTD_COMPRESSBLOCK_BTULTRA_DICTMATCHSTATE ZSTD_compressBlock_btultra_dictMatchState
+#define ZSTD_COMPRESSBLOCK_BTULTRA_EXTDICT ZSTD_compressBlock_btultra_extDict
+#define ZSTD_COMPRESSBLOCK_BTULTRA2 ZSTD_compressBlock_btultra2
+#else
+#define ZSTD_COMPRESSBLOCK_BTULTRA NULL
+#define ZSTD_COMPRESSBLOCK_BTULTRA_DICTMATCHSTATE NULL
+#define ZSTD_COMPRESSBLOCK_BTULTRA_EXTDICT NULL
+#define ZSTD_COMPRESSBLOCK_BTULTRA2 NULL
+#endif

 #if defined (__cplusplus)
 }
@@ -20,6 +20,7 @@


 /* ======   Dependencies   ====== */
+#include "../common/allocations.h"  /* ZSTD_customMalloc, ZSTD_customCalloc, ZSTD_customFree */
 #include "../common/zstd_deps.h"   /* ZSTD_memcpy, ZSTD_memset, INT_MAX, UINT_MAX */
 #include "../common/mem.h"         /* MEM_STATIC */
 #include "../common/pool.h"        /* threadpool */
@@ -600,11 +601,8 @@ static void ZSTDMT_serialState_update(serialState_t* serialState,
    ZSTD_pthread_mutex_unlock(&serialState->mutex);

    if (seqStore.size > 0) {
-        size_t const err = ZSTD_referenceExternalSequences(
-            jobCCtx, seqStore.seq, seqStore.size);
+        ZSTD_referenceExternalSequences(jobCCtx, seqStore.seq, seqStore.size);
        assert(serialState->params.ldmParams.enableLdm == ZSTD_ps_enable);
-        assert(!ZSTD_isError(err));
-        (void)err;
    }
 }

@@ -719,7 +717,7 @@ static void ZSTDMT_compressionJob(void* jobDescription)
    ZSTDMT_serialState_update(job->serial, cctx, rawSeqStore, job->src, job->jobID);

    if (!job->firstJob) {  /* flush and overwrite frame header when it's not first job */
-        size_t const hSize = ZSTD_compressContinue(cctx, dstBuff.start, dstBuff.capacity, job->src.start, 0);
+        size_t const hSize = ZSTD_compressContinue_public(cctx, dstBuff.start, dstBuff.capacity, job->src.start, 0);
        if (ZSTD_isError(hSize)) JOB_ERROR(hSize);
        DEBUGLOG(5, "ZSTDMT_compressionJob: flush and overwrite %u bytes of frame header (not first job)", (U32)hSize);
        ZSTD_invalidateRepCodes(cctx);
@@ -737,7 +735,7 @@ static void ZSTDMT_compressionJob(void* jobDescription)
        DEBUGLOG(5, "ZSTDMT_compressionJob: compress %u bytes in %i blocks", (U32)job->src.size, nbChunks);
        assert(job->cSize == 0);
        for (chunkNb = 1; chunkNb < nbChunks; chunkNb++) {
-            size_t const cSize = ZSTD_compressContinue(cctx, op, oend-op, ip, chunkSize);
+            size_t const cSize = ZSTD_compressContinue_public(cctx, op, oend-op, ip, chunkSize);
            if (ZSTD_isError(cSize)) JOB_ERROR(cSize);
            ip += chunkSize;
            op += cSize; assert(op < oend);
@@ -757,8 +755,8 @@ static void ZSTDMT_compressionJob(void* jobDescription)
            size_t const lastBlockSize1 = job->src.size & (chunkSize-1);
            size_t const lastBlockSize = ((lastBlockSize1==0) & (job->src.size>=chunkSize)) ? chunkSize : lastBlockSize1;
            size_t const cSize = (job->lastJob) ?
-                 ZSTD_compressEnd     (cctx, op, oend-op, ip, lastBlockSize) :
-                 ZSTD_compressContinue(cctx, op, oend-op, ip, lastBlockSize);
+                 ZSTD_compressEnd_public(cctx, op, oend-op, ip, lastBlockSize) :
+                 ZSTD_compressContinue_public(cctx, op, oend-op, ip, lastBlockSize);
            if (ZSTD_isError(cSize)) JOB_ERROR(cSize);
            lastCBlockSize = cSize;
    }   }
@@ -30,14 +30,14 @@
 * TODO: Support Windows calling convention.
 */

-ZSTD_HIDE_ASM_FUNCTION(HUF_decompress4X1_usingDTable_internal_bmi2_asm_loop)
-ZSTD_HIDE_ASM_FUNCTION(HUF_decompress4X2_usingDTable_internal_bmi2_asm_loop)
-ZSTD_HIDE_ASM_FUNCTION(_HUF_decompress4X2_usingDTable_internal_bmi2_asm_loop)
-ZSTD_HIDE_ASM_FUNCTION(_HUF_decompress4X1_usingDTable_internal_bmi2_asm_loop)
-.global HUF_decompress4X1_usingDTable_internal_bmi2_asm_loop
-.global HUF_decompress4X2_usingDTable_internal_bmi2_asm_loop
-.global _HUF_decompress4X1_usingDTable_internal_bmi2_asm_loop
-.global _HUF_decompress4X2_usingDTable_internal_bmi2_asm_loop
+ZSTD_HIDE_ASM_FUNCTION(HUF_decompress4X1_usingDTable_internal_fast_asm_loop)
+ZSTD_HIDE_ASM_FUNCTION(HUF_decompress4X2_usingDTable_internal_fast_asm_loop)
+ZSTD_HIDE_ASM_FUNCTION(_HUF_decompress4X2_usingDTable_internal_fast_asm_loop)
+ZSTD_HIDE_ASM_FUNCTION(_HUF_decompress4X1_usingDTable_internal_fast_asm_loop)
+.global HUF_decompress4X1_usingDTable_internal_fast_asm_loop
+.global HUF_decompress4X2_usingDTable_internal_fast_asm_loop
+.global _HUF_decompress4X1_usingDTable_internal_fast_asm_loop
+.global _HUF_decompress4X2_usingDTable_internal_fast_asm_loop
 .text

 /* Sets up register mappings for clarity.
@@ -95,8 +95,9 @@ ZSTD_HIDE_ASM_FUNCTION(_HUF_decompress4X1_usingDTable_internal_bmi2_asm_loop)
 /* Define both _HUF_* & HUF_* symbols because MacOS
 * C symbols are prefixed with '_' & Linux symbols aren't.
 */
-_HUF_decompress4X1_usingDTable_internal_bmi2_asm_loop:
-HUF_decompress4X1_usingDTable_internal_bmi2_asm_loop:
+_HUF_decompress4X1_usingDTable_internal_fast_asm_loop:
+HUF_decompress4X1_usingDTable_internal_fast_asm_loop:
+    ZSTD_CET_ENDBRANCH
    /* Save all registers - even if they are callee saved for simplicity. */
    push %rax
    push %rbx
@@ -350,8 +351,9 @@ HUF_decompress4X1_usingDTable_internal_bmi2_asm_loop:
    pop %rax
    ret

-_HUF_decompress4X2_usingDTable_internal_bmi2_asm_loop:
-HUF_decompress4X2_usingDTable_internal_bmi2_asm_loop:
+_HUF_decompress4X2_usingDTable_internal_fast_asm_loop:
+HUF_decompress4X2_usingDTable_internal_fast_asm_loop:
+    ZSTD_CET_ENDBRANCH
    /* Save all registers - even if they are callee saved for simplicity. */
    push %rax
    push %rbx
@@ -14,12 +14,12 @@
 /*-*******************************************************
 *  Dependencies
 *********************************************************/
+#include "../common/allocations.h"  /* ZSTD_customMalloc, ZSTD_customFree */
 #include "../common/zstd_deps.h"   /* ZSTD_memcpy, ZSTD_memmove, ZSTD_memset */
 #include "../common/cpu.h"         /* bmi2 */
 #include "../common/mem.h"         /* low level memory routines */
 #define FSE_STATIC_LINKING_ONLY
 #include "../common/fse.h"
-#define HUF_STATIC_LINKING_ONLY
 #include "../common/huf.h"
 #include "zstd_decompress_internal.h"
 #include "zstd_ddict.h"
@@ -55,11 +55,11 @@
 /*-*******************************************************
 *  Dependencies
 *********************************************************/
+#include "../common/allocations.h"  /* ZSTD_customMalloc, ZSTD_customCalloc, ZSTD_customFree */
 #include "../common/zstd_deps.h"   /* ZSTD_memcpy, ZSTD_memmove, ZSTD_memset */
 #include "../common/mem.h"         /* low level memory routines */
 #define FSE_STATIC_LINKING_ONLY
 #include "../common/fse.h"
-#define HUF_STATIC_LINKING_ONLY
 #include "../common/huf.h"
 #include "../common/xxhash.h" /* XXH64_reset, XXH64_update, XXH64_digest, XXH64 */
 #include "../common/zstd_internal.h"  /* blockProperties_t */
@@ -244,6 +244,8 @@ static void ZSTD_DCtx_resetParameters(ZSTD_DCtx* dctx)
    dctx->outBufferMode = ZSTD_bm_buffered;
    dctx->forceIgnoreChecksum = ZSTD_d_validateChecksum;
    dctx->refMultipleDDicts = ZSTD_rmd_refSingleDDict;
+    dctx->disableHufAsm = 0;
+    dctx->maxBlockSizeParam = 0;
 }

 static void ZSTD_initDCtx_internal(ZSTD_DCtx* dctx)
@@ -264,6 +266,7 @@ static void ZSTD_initDCtx_internal(ZSTD_DCtx* dctx)
 #endif
    dctx->noForwardProgress = 0;
    dctx->oversizedDuration = 0;
+    dctx->isFrameDecompression = 1;
 #if DYNAMIC_BMI2
    dctx->bmi2 = ZSTD_cpuSupportsBmi2();
 #endif
@@ -588,49 +591,52 @@ static size_t readSkippableFrameSize(void const* src, size_t srcSize)
    sizeU32 = MEM_readLE32((BYTE const*)src + ZSTD_FRAMEIDSIZE);
    RETURN_ERROR_IF((U32)(sizeU32 + ZSTD_SKIPPABLEHEADERSIZE) < sizeU32,
                    frameParameter_unsupported, "");
-    {
-        size_t const skippableSize = skippableHeaderSize + sizeU32;
+    {   size_t const skippableSize = skippableHeaderSize + sizeU32;
        RETURN_ERROR_IF(skippableSize > srcSize, srcSize_wrong, "");
        return skippableSize;
    }
 }

 /*! ZSTD_readSkippableFrame() :
- * Retrieves a zstd skippable frame containing data given by src, and writes it to dst buffer.
+ * Retrieves content of a skippable frame, and writes it to dst buffer.
 *
 * The parameter magicVariant will receive the magicVariant that was supplied when the frame was written,
 * i.e. magicNumber - ZSTD_MAGIC_SKIPPABLE_START.  This can be NULL if the caller is not interested
 * in the magicVariant.
 *
- * Returns an error if destination buffer is not large enough, or if the frame is not skippable.
+ * Returns an error if destination buffer is not large enough, or if this is not a valid skippable frame.
 *
 * @return : number of bytes written or a ZSTD error.
 */
-ZSTDLIB_API size_t ZSTD_readSkippableFrame(void* dst, size_t dstCapacity, unsigned* magicVariant,
-                                            const void* src, size_t srcSize)
+size_t ZSTD_readSkippableFrame(void* dst, size_t dstCapacity,
+                               unsigned* magicVariant,  /* optional, can be NULL */
+                         const void* src, size_t srcSize)
 {
-    U32 const magicNumber = MEM_readLE32(src);
-    size_t skippableFrameSize = readSkippableFrameSize(src, srcSize);
-    size_t skippableContentSize = skippableFrameSize - ZSTD_SKIPPABLEHEADERSIZE;
+    RETURN_ERROR_IF(srcSize < ZSTD_SKIPPABLEHEADERSIZE, srcSize_wrong, "");

-    /* check input validity */
-    RETURN_ERROR_IF(!ZSTD_isSkippableFrame(src, srcSize), frameParameter_unsupported, "");
-    RETURN_ERROR_IF(skippableFrameSize < ZSTD_SKIPPABLEHEADERSIZE || skippableFrameSize > srcSize, srcSize_wrong, "");
-    RETURN_ERROR_IF(skippableContentSize > dstCapacity, dstSize_tooSmall, "");
+    {   U32 const magicNumber = MEM_readLE32(src);
+        size_t skippableFrameSize = readSkippableFrameSize(src, srcSize);
+        size_t skippableContentSize = skippableFrameSize - ZSTD_SKIPPABLEHEADERSIZE;

-    /* deliver payload */
-    if (skippableContentSize > 0  && dst != NULL)
-        ZSTD_memcpy(dst, (const BYTE *)src + ZSTD_SKIPPABLEHEADERSIZE, skippableContentSize);
-    if (magicVariant != NULL)
-        *magicVariant = magicNumber - ZSTD_MAGIC_SKIPPABLE_START;
-    return skippableContentSize;
+        /* check input validity */
+        RETURN_ERROR_IF(!ZSTD_isSkippableFrame(src, srcSize), frameParameter_unsupported, "");
+        RETURN_ERROR_IF(skippableFrameSize < ZSTD_SKIPPABLEHEADERSIZE || skippableFrameSize > srcSize, srcSize_wrong, "");
+        RETURN_ERROR_IF(skippableContentSize > dstCapacity, dstSize_tooSmall, "");
+
+        /* deliver payload */
+        if (skippableContentSize > 0  && dst != NULL)
+            ZSTD_memcpy(dst, (const BYTE *)src + ZSTD_SKIPPABLEHEADERSIZE, skippableContentSize);
+        if (magicVariant != NULL)
+            *magicVariant = magicNumber - ZSTD_MAGIC_SKIPPABLE_START;
+        return skippableContentSize;
+    }
 }

 /** ZSTD_findDecompressedSize() :
- *  compatible with legacy mode
 *  `srcSize` must be the exact length of some number of ZSTD compressed and/or
 *      skippable frames
- *  @return : decompressed size of the frames contained */
+ *  note: compatible with legacy mode
+ * @return : decompressed size of the frames contained */
 unsigned long long ZSTD_findDecompressedSize(const void* src, size_t srcSize)
 {
    unsigned long long totalDstSize = 0;
@@ -640,9 +646,7 @@ unsigned long long ZSTD_findDecompressedSize(const void* src, size_t srcSize)

        if ((magicNumber & ZSTD_MAGIC_SKIPPABLE_MASK) == ZSTD_MAGIC_SKIPPABLE_START) {
            size_t const skippableSize = readSkippableFrameSize(src, srcSize);
-            if (ZSTD_isError(skippableSize)) {
-                return ZSTD_CONTENTSIZE_ERROR;
-            }
+            if (ZSTD_isError(skippableSize)) return ZSTD_CONTENTSIZE_ERROR;
            assert(skippableSize <= srcSize);

            src = (const BYTE *)src + skippableSize;
@@ -650,17 +654,17 @@ unsigned long long ZSTD_findDecompressedSize(const void* src, size_t srcSize)
            continue;
        }

-        {   unsigned long long const ret = ZSTD_getFrameContentSize(src, srcSize);
-            if (ret >= ZSTD_CONTENTSIZE_ERROR) return ret;
+        {   unsigned long long const fcs = ZSTD_getFrameContentSize(src, srcSize);
+            if (fcs >= ZSTD_CONTENTSIZE_ERROR) return fcs;

-            /* check for overflow */
-            if (totalDstSize + ret < totalDstSize) return ZSTD_CONTENTSIZE_ERROR;
-            totalDstSize += ret;
+            if (totalDstSize + fcs < totalDstSize)
+                return ZSTD_CONTENTSIZE_ERROR; /* check for overflow */
+            totalDstSize += fcs;
        }
+        /* skip to next frame */
        {   size_t const frameSrcSize = ZSTD_findFrameCompressedSize(src, srcSize);
-            if (ZSTD_isError(frameSrcSize)) {
-                return ZSTD_CONTENTSIZE_ERROR;
-            }
+            if (ZSTD_isError(frameSrcSize)) return ZSTD_CONTENTSIZE_ERROR;
+            assert(frameSrcSize <= srcSize);

            src = (const BYTE *)src + frameSrcSize;
            srcSize -= frameSrcSize;
@@ -782,6 +786,7 @@ static ZSTD_frameSizeInfo ZSTD_findFrameSizeInfo(const void* src, size_t srcSize
            ip += 4;
        }

+        frameSizeInfo.nbBlocks = nbBlocks;
        frameSizeInfo.compressedSize = (size_t)(ip - ipstart);
        frameSizeInfo.decompressedBound = (zfh.frameContentSize != ZSTD_CONTENTSIZE_UNKNOWN)
                                        ? zfh.frameContentSize
@@ -825,6 +830,48 @@ unsigned long long ZSTD_decompressBound(const void* src, size_t srcSize)
    return bound;
 }

+size_t ZSTD_decompressionMargin(void const* src, size_t srcSize)
+{
+    size_t margin = 0;
+    unsigned maxBlockSize = 0;
+
+    /* Iterate over each frame */
+    while (srcSize > 0) {
+        ZSTD_frameSizeInfo const frameSizeInfo = ZSTD_findFrameSizeInfo(src, srcSize);
+        size_t const compressedSize = frameSizeInfo.compressedSize;
+        unsigned long long const decompressedBound = frameSizeInfo.decompressedBound;
+        ZSTD_frameHeader zfh;
+
+        FORWARD_IF_ERROR(ZSTD_getFrameHeader(&zfh, src, srcSize), "");
+        if (ZSTD_isError(compressedSize) || decompressedBound == ZSTD_CONTENTSIZE_ERROR)
+            return ERROR(corruption_detected);
+
+        if (zfh.frameType == ZSTD_frame) {
+            /* Add the frame header to our margin */
+            margin += zfh.headerSize;
+            /* Add the checksum to our margin */
+            margin += zfh.checksumFlag ? 4 : 0;
+            /* Add 3 bytes per block */
+            margin += 3 * frameSizeInfo.nbBlocks;
+
+            /* Compute the max block size */
+            maxBlockSize = MAX(maxBlockSize, zfh.blockSizeMax);
+        } else {
+            assert(zfh.frameType == ZSTD_skippableFrame);
+            /* Add the entire skippable frame size to our margin. */
+            margin += compressedSize;
+        }
+
+        assert(srcSize >= compressedSize);
+        src = (const BYTE*)src + compressedSize;
+        srcSize -= compressedSize;
+    }
+
+    /* Add the max block size back to the margin. */
+    margin += maxBlockSize;
+
+    return margin;
+}

 /*-*************************************************************
 *   Frame decoding
@@ -850,7 +897,7 @@ static size_t ZSTD_copyRawBlock(void* dst, size_t dstCapacity,
        if (srcSize == 0) return 0;
        RETURN_ERROR(dstBuffer_null, "");
    }
-    ZSTD_memcpy(dst, src, srcSize);
+    ZSTD_memmove(dst, src, srcSize);
    return srcSize;
 }

@@ -926,8 +973,13 @@ static size_t ZSTD_decompressFrame(ZSTD_DCtx* dctx,
        ip += frameHeaderSize; remainingSrcSize -= frameHeaderSize;
    }

+    /* Shrink the blockSizeMax if enabled */
+    if (dctx->maxBlockSizeParam != 0)
+        dctx->fParams.blockSizeMax = MIN(dctx->fParams.blockSizeMax, (unsigned)dctx->maxBlockSizeParam);
+
    /* Loop on each block */
    while (1) {
+        BYTE* oBlockEnd = oend;
        size_t decodedSize;
        blockProperties_t blockProperties;
        size_t const cBlockSize = ZSTD_getcBlockSize(ip, remainingSrcSize, &blockProperties);
@@ -937,16 +989,35 @@ static size_t ZSTD_decompressFrame(ZSTD_DCtx* dctx,
        remainingSrcSize -= ZSTD_blockHeaderSize;
        RETURN_ERROR_IF(cBlockSize > remainingSrcSize, srcSize_wrong, "");

+        if (ip >= op && ip < oBlockEnd) {
+            /* We are decompressing in-place. Limit the output pointer so that we
+             * don't overwrite the block that we are currently reading. This will
+             * fail decompression if the input & output pointers aren't spaced
+             * far enough apart.
+             *
+             * This is important to set, even when the pointers are far enough
+             * apart, because ZSTD_decompressBlock_internal() can decide to store
+             * literals in the output buffer, after the block it is decompressing.
+             * Since we don't want anything to overwrite our input, we have to tell
+             * ZSTD_decompressBlock_internal to never write past ip.
+             *
+             * See ZSTD_allocateLiteralsBuffer() for reference.
+             */
+            oBlockEnd = op + (ip - op);
+        }
+
        switch(blockProperties.blockType)
        {
        case bt_compressed:
-            decodedSize = ZSTD_decompressBlock_internal(dctx, op, (size_t)(oend-op), ip, cBlockSize, /* frame */ 1, not_streaming);
+            assert(dctx->isFrameDecompression == 1);
+            decodedSize = ZSTD_decompressBlock_internal(dctx, op, (size_t)(oBlockEnd-op), ip, cBlockSize, not_streaming);
            break;
        case bt_raw :
+            /* Use oend instead of oBlockEnd because this function is safe to overlap. It uses memmove. */
            decodedSize = ZSTD_copyRawBlock(op, (size_t)(oend-op), ip, cBlockSize);
            break;
        case bt_rle :
-            decodedSize = ZSTD_setRleBlock(op, (size_t)(oend-op), *ip, blockProperties.origSize);
+            decodedSize = ZSTD_setRleBlock(op, (size_t)(oBlockEnd-op), *ip, blockProperties.origSize);
            break;
        case bt_reserved :
        default:
@@ -981,6 +1052,7 @@ static size_t ZSTD_decompressFrame(ZSTD_DCtx* dctx,
    }
    ZSTD_DCtx_trace_end(dctx, (U64)(op-ostart), (U64)(ip-istart), /* streaming */ 0);
    /* Allow caller to get size read */
+    DEBUGLOG(4, "ZSTD_decompressFrame: decompressed frame of size %zi, consuming %zi bytes of input", op-ostart, ip - (const BYTE*)*srcPtr);
    *srcPtr = ip;
    *srcSizePtr = remainingSrcSize;
    return (size_t)(op-ostart);
@@ -1027,17 +1099,18 @@ static size_t ZSTD_decompressMultiFrame(ZSTD_DCtx* dctx,
        }
 #endif

-        {   U32 const magicNumber = MEM_readLE32(src);
-            DEBUGLOG(4, "reading magic number %08X (expecting %08X)",
-                        (unsigned)magicNumber, ZSTD_MAGICNUMBER);
+        if (srcSize >= 4) {
+            U32 const magicNumber = MEM_readLE32(src);
+            DEBUGLOG(5, "reading magic number %08X", (unsigned)magicNumber);
            if ((magicNumber & ZSTD_MAGIC_SKIPPABLE_MASK) == ZSTD_MAGIC_SKIPPABLE_START) {
+                /* skippable frame detected : skip it */
                size_t const skippableSize = readSkippableFrameSize(src, srcSize);
-                FORWARD_IF_ERROR(skippableSize, "readSkippableFrameSize failed");
+                FORWARD_IF_ERROR(skippableSize, "invalid skippable frame");
                assert(skippableSize <= srcSize);

                src = (const BYTE *)src + skippableSize;
                srcSize -= skippableSize;
-                continue;
+                continue; /* check next frame */
        }   }

        if (ddict) {
@@ -1253,7 +1326,8 @@ size_t ZSTD_decompressContinue(ZSTD_DCtx* dctx, void* dst, size_t dstCapacity, c
            {
            case bt_compressed:
                DEBUGLOG(5, "ZSTD_decompressContinue: case bt_compressed");
-                rSize = ZSTD_decompressBlock_internal(dctx, dst, dstCapacity, src, srcSize, /* frame */ 1, is_streaming);
+                assert(dctx->isFrameDecompression == 1);
+                rSize = ZSTD_decompressBlock_internal(dctx, dst, dstCapacity, src, srcSize, is_streaming);
                dctx->expected = 0;  /* Streaming not supported */
                break;
            case bt_raw :
@@ -1375,11 +1449,11 @@ ZSTD_loadDEntropy(ZSTD_entropyDTables_t* entropy,
        /* in minimal huffman, we always use X1 variants */
        size_t const hSize = HUF_readDTableX1_wksp(entropy->hufTable,
                                                dictPtr, dictEnd - dictPtr,
-                                                workspace, workspaceSize);
+                                                workspace, workspaceSize, /* flags */ 0);
 #else
        size_t const hSize = HUF_readDTableX2_wksp(entropy->hufTable,
                                                dictPtr, (size_t)(dictEnd - dictPtr),
-                                                workspace, workspaceSize);
+                                                workspace, workspaceSize, /* flags */ 0);
 #endif
        RETURN_ERROR_IF(HUF_isError(hSize), dictionary_corrupted, "");
        dictPtr += hSize;
@@ -1474,6 +1548,12 @@ size_t ZSTD_decompressBegin(ZSTD_DCtx* dctx)
    dctx->stage = ZSTDds_getFrameHeaderSize;
    dctx->processedCSize = 0;
    dctx->decodedSize = 0;
+    /* Set to non-null because ZSTD_prefetchMatch() may end up doing addition
+     * with this value for corrupted frames. However, it then just passes the
+     * pointer to PREFETCH_L1(), which doesn't require valid pointers. But,
+     * if it is NULL we get nullptr-with-nonzero-offset UBSAN warnings.
+     */
+    dctx->previousDstEnd = "";
    dctx->previousDstEnd = NULL;
    dctx->prefixStart = NULL;
    dctx->virtualStart = NULL;
@@ -1482,6 +1562,7 @@ size_t ZSTD_decompressBegin(ZSTD_DCtx* dctx)
    dctx->litEntropy = dctx->fseEntropy = 0;
    dctx->dictID = 0;
    dctx->bType = bt_reserved;
+    dctx->isFrameDecompression = 1;
    ZSTD_STATIC_ASSERT(sizeof(dctx->entropy.rep) == sizeof(repStartValue));
    ZSTD_memcpy(dctx->entropy.rep, repStartValue, sizeof(repStartValue));  /* initial repcodes */
    dctx->LLTptr = dctx->entropy.LLTable;
@@ -1656,7 +1737,9 @@ size_t ZSTD_initDStream_usingDict(ZSTD_DStream* zds, const void* dict, size_t di
 size_t ZSTD_initDStream(ZSTD_DStream* zds)
 {
    DEBUGLOG(4, "ZSTD_initDStream");
-    return ZSTD_initDStream_usingDDict(zds, NULL);
+    FORWARD_IF_ERROR(ZSTD_DCtx_reset(zds, ZSTD_reset_session_only), "");
+    FORWARD_IF_ERROR(ZSTD_DCtx_refDDict(zds, NULL), "");
+    return ZSTD_startingInputLength(zds->format);
 }

 /* ZSTD_initDStream_usingDDict() :
@@ -1664,6 +1747,7 @@ size_t ZSTD_initDStream(ZSTD_DStream* zds)
 * this function cannot fail */
 size_t ZSTD_initDStream_usingDDict(ZSTD_DStream* dctx, const ZSTD_DDict* ddict)
 {
+    DEBUGLOG(4, "ZSTD_initDStream_usingDDict");
    FORWARD_IF_ERROR( ZSTD_DCtx_reset(dctx, ZSTD_reset_session_only) , "");
    FORWARD_IF_ERROR( ZSTD_DCtx_refDDict(dctx, ddict) , "");
    return ZSTD_startingInputLength(dctx->format);
@@ -1674,6 +1758,7 @@ size_t ZSTD_initDStream_usingDDict(ZSTD_DStream* dctx, const ZSTD_DDict* ddict)
 * this function cannot fail */
 size_t ZSTD_resetDStream(ZSTD_DStream* dctx)
 {
+    DEBUGLOG(4, "ZSTD_resetDStream");
    FORWARD_IF_ERROR(ZSTD_DCtx_reset(dctx, ZSTD_reset_session_only), "");
    return ZSTD_startingInputLength(dctx->format);
 }
@@ -1745,6 +1830,15 @@ ZSTD_bounds ZSTD_dParam_getBounds(ZSTD_dParameter dParam)
            bounds.lowerBound = (int)ZSTD_rmd_refSingleDDict;
            bounds.upperBound = (int)ZSTD_rmd_refMultipleDDicts;
            return bounds;
+        case ZSTD_d_disableHuffmanAssembly:
+            bounds.lowerBound = 0;
+            bounds.upperBound = 1;
+            return bounds;
+        case ZSTD_d_maxBlockSize:
+            bounds.lowerBound = ZSTD_BLOCKSIZE_MAX_MIN;
+            bounds.upperBound = ZSTD_BLOCKSIZE_MAX;
+            return bounds;
+
        default:;
    }
    bounds.error = ERROR(parameter_unsupported);
@@ -1785,6 +1879,12 @@ size_t ZSTD_DCtx_getParameter(ZSTD_DCtx* dctx, ZSTD_dParameter param, int* value
        case ZSTD_d_refMultipleDDicts:
            *value = (int)dctx->refMultipleDDicts;
            return 0;
+        case ZSTD_d_disableHuffmanAssembly:
+            *value = (int)dctx->disableHufAsm;
+            return 0;
+        case ZSTD_d_maxBlockSize:
+            *value = dctx->maxBlockSizeParam;
+            return 0;
        default:;
    }
    RETURN_ERROR(parameter_unsupported, "");
@@ -1818,6 +1918,14 @@ size_t ZSTD_DCtx_setParameter(ZSTD_DCtx* dctx, ZSTD_dParameter dParam, int value
            }
            dctx->refMultipleDDicts = (ZSTD_refMultipleDDicts_e)value;
            return 0;
+        case ZSTD_d_disableHuffmanAssembly:
+            CHECK_DBOUNDS(ZSTD_d_disableHuffmanAssembly, value);
+            dctx->disableHufAsm = value != 0;
+            return 0;
+        case ZSTD_d_maxBlockSize:
+            if (value != 0) CHECK_DBOUNDS(ZSTD_d_maxBlockSize, value);
+            dctx->maxBlockSizeParam = value;
+            return 0;
        default:;
    }
    RETURN_ERROR(parameter_unsupported, "");
@@ -1829,6 +1937,7 @@ size_t ZSTD_DCtx_reset(ZSTD_DCtx* dctx, ZSTD_ResetDirective reset)
      || (reset == ZSTD_reset_session_and_parameters) ) {
        dctx->streamStage = zdss_init;
        dctx->noForwardProgress = 0;
+        dctx->isFrameDecompression = 1;
    }
    if ( (reset == ZSTD_reset_parameters)
      || (reset == ZSTD_reset_session_and_parameters) ) {
@@ -1845,11 +1954,17 @@ size_t ZSTD_sizeof_DStream(const ZSTD_DStream* dctx)
    return ZSTD_sizeof_DCtx(dctx);
 }

-size_t ZSTD_decodingBufferSize_min(unsigned long long windowSize, unsigned long long frameContentSize)
+static size_t ZSTD_decodingBufferSize_internal(unsigned long long windowSize, unsigned long long frameContentSize, size_t blockSizeMax)
 {
-    size_t const blockSize = (size_t) MIN(windowSize, ZSTD_BLOCKSIZE_MAX);
-    /* space is needed to store the litbuffer after the output of a given block without stomping the extDict of a previous run, as well as to cover both windows against wildcopy*/
-    unsigned long long const neededRBSize = windowSize + blockSize + ZSTD_BLOCKSIZE_MAX + (WILDCOPY_OVERLENGTH * 2);
+    size_t const blockSize = MIN((size_t)MIN(windowSize, ZSTD_BLOCKSIZE_MAX), blockSizeMax);
+    /* We need blockSize + WILDCOPY_OVERLENGTH worth of buffer so that if a block
+     * ends at windowSize + WILDCOPY_OVERLENGTH + 1 bytes, we can start writing
+     * the block at the beginning of the output buffer, and maintain a full window.
+     *
+     * We need another blockSize worth of buffer so that we can store split
+     * literals at the end of the block without overwriting the extDict window.
+     */
+    unsigned long long const neededRBSize = windowSize + (blockSize * 2) + (WILDCOPY_OVERLENGTH * 2);
    unsigned long long const neededSize = MIN(frameContentSize, neededRBSize);
    size_t const minRBSize = (size_t) neededSize;
    RETURN_ERROR_IF((unsigned long long)minRBSize != neededSize,
@@ -1857,6 +1972,11 @@ size_t ZSTD_decodingBufferSize_min(unsigned long long windowSize, unsigned long
    return minRBSize;
 }

+size_t ZSTD_decodingBufferSize_min(unsigned long long windowSize, unsigned long long frameContentSize)
+{
+    return ZSTD_decodingBufferSize_internal(windowSize, frameContentSize, ZSTD_BLOCKSIZE_MAX);
+}
+
 size_t ZSTD_estimateDStreamSize(size_t windowSize)
 {
    size_t const blockSize = MIN(windowSize, ZSTD_BLOCKSIZE_MAX);
@@ -2095,11 +2215,13 @@ size_t ZSTD_decompressStream(ZSTD_DStream* zds, ZSTD_outBuffer* output, ZSTD_inB
            zds->fParams.windowSize = MAX(zds->fParams.windowSize, 1U << ZSTD_WINDOWLOG_ABSOLUTEMIN);
            RETURN_ERROR_IF(zds->fParams.windowSize > zds->maxWindowSize,
                            frameParameter_windowTooLarge, "");
+            if (zds->maxBlockSizeParam != 0)
+                zds->fParams.blockSizeMax = MIN(zds->fParams.blockSizeMax, (unsigned)zds->maxBlockSizeParam);

            /* Adapt buffer sizes to frame header instructions */
            {   size_t const neededInBuffSize = MAX(zds->fParams.blockSizeMax, 4 /* frame checksum */);
                size_t const neededOutBuffSize = zds->outBufferMode == ZSTD_bm_buffered
-                        ? ZSTD_decodingBufferSize_min(zds->fParams.windowSize, zds->fParams.frameContentSize)
+                        ? ZSTD_decodingBufferSize_internal(zds->fParams.windowSize, zds->fParams.frameContentSize, zds->fParams.blockSizeMax)
                        : 0;

                ZSTD_DCtx_updateOversizedDuration(zds, neededInBuffSize, neededOutBuffSize);
@@ -2159,7 +2281,7 @@ size_t ZSTD_decompressStream(ZSTD_DStream* zds, ZSTD_outBuffer* output, ZSTD_inB
                int const isSkipFrame = ZSTD_isSkipFrame(zds);
                size_t loadedSize;
                /* At this point we shouldn't be decompressing a block that we can stream. */
-                assert(neededInSize == ZSTD_nextSrcSizeToDecompressWithInputSize(zds, iend - ip));
+                assert(neededInSize == ZSTD_nextSrcSizeToDecompressWithInputSize(zds, (size_t)(iend - ip)));
                if (isSkipFrame) {
                    loadedSize = MIN(toLoad, (size_t)(iend-ip));
                } else {
@@ -2219,8 +2341,8 @@ size_t ZSTD_decompressStream(ZSTD_DStream* zds, ZSTD_outBuffer* output, ZSTD_inB
    if ((ip==istart) && (op==ostart)) {  /* no forward progress */
        zds->noForwardProgress ++;
        if (zds->noForwardProgress >= ZSTD_NO_FORWARD_PROGRESS_MAX) {
-            RETURN_ERROR_IF(op==oend, dstSize_tooSmall, "");
-            RETURN_ERROR_IF(ip==iend, srcSize_wrong, "");
+            RETURN_ERROR_IF(op==oend, noForwardProgress_destFull, "");
+            RETURN_ERROR_IF(ip==iend, noForwardProgress_inputEmpty, "");
            assert(0);
        }
    } else {
@@ -2257,11 +2379,17 @@ size_t ZSTD_decompressStream_simpleArgs (
                            void* dst, size_t dstCapacity, size_t* dstPos,
                      const void* src, size_t srcSize, size_t* srcPos)
 {
-    ZSTD_outBuffer output = { dst, dstCapacity, *dstPos };
-    ZSTD_inBuffer  input  = { src, srcSize, *srcPos };
-    /* ZSTD_compress_generic() will check validity of dstPos and srcPos */
-    size_t const cErr = ZSTD_decompressStream(dctx, &output, &input);
-    *dstPos = output.pos;
-    *srcPos = input.pos;
-    return cErr;
+    ZSTD_outBuffer output;
+    ZSTD_inBuffer  input;
+    output.dst = dst;
+    output.size = dstCapacity;
+    output.pos = *dstPos;
+    input.src = src;
+    input.size = srcSize;
+    input.pos = *srcPos;
+    {   size_t const cErr = ZSTD_decompressStream(dctx, &output, &input);
+        *dstPos = output.pos;
+        *srcPos = input.pos;
+        return cErr;
+    }
 }
@@ -47,7 +47,7 @@ typedef enum {
 */
 size_t ZSTD_decompressBlock_internal(ZSTD_DCtx* dctx,
                               void* dst, size_t dstCapacity,
-                         const void* src, size_t srcSize, const int frame, const streaming_operation streaming);
+                         const void* src, size_t srcSize, const streaming_operation streaming);

 /* ZSTD_buildFSETable() :
 * generate FSE decoding table for one symbol (ll, ml or off)
@@ -64,5 +64,10 @@ void ZSTD_buildFSETable(ZSTD_seqSymbol* dt,
                   unsigned tableLog, void* wksp, size_t wkspSize,
                   int bmi2);

+/* Internal definition of ZSTD_decompressBlock() to avoid deprecation warnings. */
+size_t ZSTD_decompressBlock_deprecated(ZSTD_DCtx* dctx,
+                            void* dst, size_t dstCapacity,
+                      const void* src, size_t srcSize);
+

 #endif /* ZSTD_DEC_BLOCK_H */
@@ -153,6 +153,7 @@ struct ZSTD_DCtx_s
    size_t litSize;
    size_t rleSize;
    size_t staticSize;
+    int isFrameDecompression;
 #if DYNAMIC_BMI2 != 0
    int bmi2;                     /* == 1 if the CPU supports BMI2 and 0 otherwise. CPU support is determined dynamically once per context lifetime. */
 #endif
@@ -165,6 +166,8 @@ struct ZSTD_DCtx_s
    ZSTD_dictUses_e dictUses;
    ZSTD_DDictHashSet* ddictSet;                    /* Hash set for multiple ddicts */
    ZSTD_refMultipleDDicts_e refMultipleDDicts;     /* User specified: if == 1, will allow references to multiple DDicts. Default == 0 (disabled) */
+    int disableHufAsm;
+    int maxBlockSizeParam;

    /* streaming */
    ZSTD_dStreamStage streamStage;
@@ -13,6 +13,8 @@
 /* *************************************
 *  Dependencies
 ***************************************/
+#define ZSTD_DISABLE_DEPRECATE_WARNINGS  /* suppress warning on ZSTD_initDStream_usingDict */
+#include "../zstd.h"        /* ZSTD_CStream, ZSTD_DStream, ZSTDLIB_API */
 #define ZBUFF_STATIC_LINKING_ONLY
 #include "zbuff.h"

@@ -733,7 +733,7 @@ static size_t COVER_buildDictionary(const COVER_ctx_t *ctx, U32 *freqs,
  return tail;
 }

-ZDICTLIB_API size_t ZDICT_trainFromBuffer_cover(
+ZDICTLIB_STATIC_API size_t ZDICT_trainFromBuffer_cover(
    void *dictBuffer, size_t dictBufferCapacity,
    const void *samplesBuffer, const size_t *samplesSizes, unsigned nbSamples,
    ZDICT_cover_params_t parameters)
@@ -951,9 +951,17 @@ void COVER_best_finish(COVER_best_t *best, ZDICT_cover_params_t parameters,
  }
 }

+static COVER_dictSelection_t setDictSelection(BYTE* buf, size_t s, size_t csz)
+{
+    COVER_dictSelection_t ds;
+    ds.dictContent = buf;
+    ds.dictSize = s;
+    ds.totalCompressedSize = csz;
+    return ds;
+}
+
 COVER_dictSelection_t COVER_dictSelectionError(size_t error) {
-    COVER_dictSelection_t selection = { NULL, 0, error };
-    return selection;
+    return setDictSelection(NULL, 0, error);
 }

 unsigned COVER_dictSelectionIsError(COVER_dictSelection_t selection) {
@@ -1006,9 +1014,8 @@ COVER_dictSelection_t COVER_selectDict(BYTE* customDictContent, size_t dictBuffe
  }

  if (params.shrinkDict == 0) {
-    COVER_dictSelection_t selection = { largestDictbuffer, dictContentSize, totalCompressedSize };
    free(candidateDictBuffer);
-    return selection;
+    return setDictSelection(largestDictbuffer, dictContentSize, totalCompressedSize);
  }

  largestDict = dictContentSize;
@@ -1041,19 +1048,15 @@ COVER_dictSelection_t COVER_selectDict(BYTE* customDictContent, size_t dictBuffe
    }

    if ((double)totalCompressedSize <= (double)largestCompressed * regressionTolerance) {
-      COVER_dictSelection_t selection = { candidateDictBuffer, dictContentSize, totalCompressedSize };
      free(largestDictbuffer);
-      return selection;
+      return setDictSelection( candidateDictBuffer, dictContentSize, totalCompressedSize );
    }
    dictContentSize *= 2;
  }
  dictContentSize = largestDict;
  totalCompressedSize = largestCompressed;
-  {
-    COVER_dictSelection_t selection = { largestDictbuffer, dictContentSize, totalCompressedSize };
-    free(candidateDictBuffer);
-    return selection;
-  }
+  free(candidateDictBuffer);
+  return setDictSelection( largestDictbuffer, dictContentSize, totalCompressedSize );
 }

 /**
@@ -1116,7 +1119,7 @@ _cleanup:
  free(freqs);
 }

-ZDICTLIB_API size_t ZDICT_optimizeTrainFromBuffer_cover(
+ZDICTLIB_STATIC_API size_t ZDICT_optimizeTrainFromBuffer_cover(
    void* dictBuffer, size_t dictBufferCapacity, const void* samplesBuffer,
    const size_t* samplesSizes, unsigned nbSamples,
    ZDICT_cover_params_t* parameters)
@@ -545,7 +545,7 @@ FASTCOVER_convertToFastCoverParams(ZDICT_cover_params_t coverParams,
 }


-ZDICTLIB_API size_t
+ZDICTLIB_STATIC_API size_t
 ZDICT_trainFromBuffer_fastCover(void* dictBuffer, size_t dictBufferCapacity,
                                const void* samplesBuffer,
                                const size_t* samplesSizes, unsigned nbSamples,
@@ -614,7 +614,7 @@ ZDICT_trainFromBuffer_fastCover(void* dictBuffer, size_t dictBufferCapacity,
 }


-ZDICTLIB_API size_t
+ZDICTLIB_STATIC_API size_t
 ZDICT_optimizeTrainFromBuffer_fastCover(
                    void* dictBuffer, size_t dictBufferCapacity,
                    const void* samplesBuffer,
@@ -44,7 +44,6 @@
 #ifndef ZDICT_STATIC_LINKING_ONLY
 #  define ZDICT_STATIC_LINKING_ONLY
 #endif
-#define HUF_STATIC_LINKING_ONLY

 #include "../common/mem.h"           /* read */
 #include "../common/fse.h"           /* FSE_normalizeCount, FSE_writeNCount */
@@ -567,11 +566,11 @@ static void ZDICT_countEStats(EStats_ress_t esr, const ZSTD_parameters* params,
    size_t cSize;

    if (srcSize > blockSizeMax) srcSize = blockSizeMax;   /* protection vs large samples */
-    {   size_t const errorCode = ZSTD_compressBegin_usingCDict(esr.zc, esr.dict);
+    {   size_t const errorCode = ZSTD_compressBegin_usingCDict_deprecated(esr.zc, esr.dict);
        if (ZSTD_isError(errorCode)) { DISPLAYLEVEL(1, "warning : ZSTD_compressBegin_usingCDict failed \n"); return; }

    }
-    cSize = ZSTD_compressBlock(esr.zc, esr.workPlace, ZSTD_BLOCKSIZE_MAX, src, srcSize);
+    cSize = ZSTD_compressBlock_deprecated(esr.zc, esr.workPlace, ZSTD_BLOCKSIZE_MAX, src, srcSize);
    if (ZSTD_isError(cSize)) { DISPLAYLEVEL(3, "warning : could not compress sample size %u \n", (unsigned)srcSize); return; }

    if (cSize) {  /* if == 0; block is not compressible */
@@ -676,6 +675,7 @@ static size_t ZDICT_analyzeEntropy(void*  dstBuffer, size_t maxDstSize,
    size_t const totalSrcSize = ZDICT_totalSampleSize(fileSizes, nbFiles);
    size_t const averageSampleSize = totalSrcSize / (nbFiles + !nbFiles);
    BYTE* dstPtr = (BYTE*)dstBuffer;
+    U32 wksp[HUF_CTABLE_WORKSPACE_SIZE_U32];

    /* init */
    DEBUGLOG(4, "ZDICT_analyzeEntropy");
@@ -716,7 +716,7 @@ static size_t ZDICT_analyzeEntropy(void*  dstBuffer, size_t maxDstSize,
    }   }

    /* analyze, build stats, starting with literals */
-    {   size_t maxNbBits = HUF_buildCTable (hufTable, countLit, 255, huffLog);
+    {   size_t maxNbBits = HUF_buildCTable_wksp(hufTable, countLit, 255, huffLog, wksp, sizeof(wksp));
        if (HUF_isError(maxNbBits)) {
            eSize = maxNbBits;
            DISPLAYLEVEL(1, " HUF_buildCTable error \n");
@@ -725,7 +725,7 @@ static size_t ZDICT_analyzeEntropy(void*  dstBuffer, size_t maxDstSize,
        if (maxNbBits==8) {  /* not compressible : will fail on HUF_writeCTable() */
            DISPLAYLEVEL(2, "warning : pathological dataset : literals are not compressible : samples are noisy or too regular \n");
            ZDICT_flatLit(countLit);  /* replace distribution by a fake "mostly flat but still compressible" distribution, that HUF_writeCTable() can encode */
-            maxNbBits = HUF_buildCTable (hufTable, countLit, 255, huffLog);
+            maxNbBits = HUF_buildCTable_wksp(hufTable, countLit, 255, huffLog, wksp, sizeof(wksp));
            assert(maxNbBits==9);
        }
        huffLog = (U32)maxNbBits;
@@ -766,7 +766,7 @@ static size_t ZDICT_analyzeEntropy(void*  dstBuffer, size_t maxDstSize,
    llLog = (U32)errorCode;

    /* write result to buffer */
-    {   size_t const hhSize = HUF_writeCTable(dstPtr, maxDstSize, hufTable, 255, huffLog);
+    {   size_t const hhSize = HUF_writeCTable_wksp(dstPtr, maxDstSize, hufTable, 255, huffLog, wksp, sizeof(wksp));
        if (HUF_isError(hhSize)) {
            eSize = hhSize;
            DISPLAYLEVEL(1, "HUF_writeCTable error \n");
@@ -242,6 +242,13 @@ MEM_STATIC ZSTD_frameSizeInfo ZSTD_findFrameSizeInfoLegacy(const void *src, size
        frameSizeInfo.compressedSize = ERROR(srcSize_wrong);
        frameSizeInfo.decompressedBound = ZSTD_CONTENTSIZE_ERROR;
    }
+    /* In all cases, decompressedBound == nbBlocks * ZSTD_BLOCKSIZE_MAX.
+     * So we can compute nbBlocks without having to change every function.
+     */
+    if (frameSizeInfo.decompressedBound != ZSTD_CONTENTSIZE_ERROR) {
+        assert((frameSizeInfo.decompressedBound & (ZSTD_BLOCKSIZE_MAX - 1)) == 0);
+        frameSizeInfo.nbBlocks = (size_t)(frameSizeInfo.decompressedBound / ZSTD_BLOCKSIZE_MAX);
+    }
    return frameSizeInfo;
 }

@@ -1720,20 +1720,26 @@ static size_t ZSTD_execSequence(BYTE* op,
    static const int dec32table[] = {0, 1, 2, 1, 4, 4, 4, 4};   /* added */
    static const int dec64table[] = {8, 8, 8, 7, 8, 9,10,11};   /* subtracted */
    const BYTE* const ostart = op;
+    BYTE* const oLitEnd = op + sequence.litLength;
    const size_t litLength = sequence.litLength;
    BYTE* const endMatch = op + litLength + sequence.matchLength;    /* risk : address space overflow (32-bits) */
    const BYTE* const litEnd = *litPtr + litLength;

-    /* check */
+    /* checks */
+    size_t const seqLength = sequence.litLength + sequence.matchLength;
+
+    if (seqLength > (size_t)(oend - op)) return ERROR(dstSize_tooSmall);
+    if (sequence.litLength > (size_t)(litLimit - *litPtr)) return ERROR(corruption_detected);
+    /* Now we know there are no overflow in literal nor match lengths, can use pointer checks */
+    if (sequence.offset > (U32)(oLitEnd - base)) return ERROR(corruption_detected);
+
    if (endMatch > oend) return ERROR(dstSize_tooSmall);   /* overwrite beyond dst buffer */
-    if (litEnd > litLimit) return ERROR(corruption_detected);
-    if (sequence.matchLength > (size_t)(*litPtr-op))  return ERROR(dstSize_tooSmall);    /* overwrite literal segment */
+    if (litEnd > litLimit) return ERROR(corruption_detected);   /* overRead beyond lit buffer */
+    if (sequence.matchLength > (size_t)(*litPtr-op)) return ERROR(dstSize_tooSmall);  /* overwrite literal segment */

    /* copy Literals */
-    if (((size_t)(*litPtr - op) < 8) || ((size_t)(oend-litEnd) < 8) || (op+litLength > oend-8))
-        memmove(op, *litPtr, litLength);   /* overwrite risk */
-    else
-        ZSTD_wildcopy(op, *litPtr, litLength);
+    ZSTD_memmove(op, *litPtr, sequence.litLength);   /* note : v0.1 seems to allow scenarios where output or input are close to end of buffer */
+
    op += litLength;
    *litPtr = litEnd;   /* update for next sequence */

@@ -3066,12 +3066,19 @@ static size_t ZSTD_execSequence(BYTE* op,
    const BYTE* const litEnd = *litPtr + sequence.litLength;

    /* checks */
-    if (oLitEnd > oend_8) return ERROR(dstSize_tooSmall);   /* last match must start at a minimum distance of 8 from oend */
+    size_t const seqLength = sequence.litLength + sequence.matchLength;
+
+    if (seqLength > (size_t)(oend - op)) return ERROR(dstSize_tooSmall);
+    if (sequence.litLength > (size_t)(litLimit - *litPtr)) return ERROR(corruption_detected);
+    /* Now we know there are no overflow in literal nor match lengths, can use the pointer check */
+    if (oLitEnd > oend_8) return ERROR(dstSize_tooSmall);
+    if (sequence.offset > (U32)(oLitEnd - base)) return ERROR(corruption_detected);
+
    if (oMatchEnd > oend) return ERROR(dstSize_tooSmall);   /* overwrite beyond dst buffer */
    if (litEnd > litLimit) return ERROR(corruption_detected);   /* overRead beyond lit buffer */

    /* copy Literals */
-    ZSTD_wildcopy(op, *litPtr, sequence.litLength);   /* note : oLitEnd <= oend-8 : no risk of overwrite beyond oend */
+    ZSTD_wildcopy(op, *litPtr, (ptrdiff_t)sequence.litLength);   /* note : oLitEnd <= oend-8 : no risk of overwrite beyond oend */
    op = oLitEnd;
    *litPtr = litEnd;   /* update for next sequence */

--- a/Show More
+++ b/Show More