update HISTORY.md and version.h for 7.7.8

Summary:
Currently, `iterate_upper_bound` is not checked for range tombstone keys in MergingIterator. This may impact performance when there is a large number of range tombstones right after `iterate_upper_bound`. This PR fixes this issue by checking `iterate_upper_bound` in MergingIterator for range tombstone keys.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/10966

Test Plan:
- added unit test
- stress test: `python3 tools/db_crashtest.py whitebox --simple --verify_iterator_with_expected_state_one_in=5 --delrangepercent=5 --prefixpercent=18 --writepercent=48 --readpercen=15 --duration=36000 --range_deletion_width=100`
- ran different stress tests over sandcastle
- Falcon team ran some test traffic and saw reduced CPU usage on processing range tombstones.

.circleci/config.yml

@@ -1,7 +1,7 @@
 version: 2.1
 
 orbs:
-  win: circleci/windows@5.0.0
+  win: circleci/windows@2.4.0
 
 commands:
   install-cmake-on-macos:
@@ -45,38 +45,6 @@ commands:
             echo "export LZ4_DOWNLOAD_BASE=https://rocksdb-deps.s3.us-west-2.amazonaws.com/pkgs/lz4" >> $BASH_ENV
             echo "export ZSTD_DOWNLOAD_BASE=https://rocksdb-deps.s3.us-west-2.amazonaws.com/pkgs/zstd" >> $BASH_ENV
 
-  windows-build-steps:
-    steps:
-      - checkout
-      - run:
-          name: "Install thirdparty dependencies"
-          command: |
-            echo "Installing CMake..."
-            choco install cmake --installargs 'ADD_CMAKE_TO_PATH=System' -y
-            mkdir $Env:THIRDPARTY_HOME
-            cd $Env:THIRDPARTY_HOME
-            echo "Building Snappy dependency..."
-            curl https://github.com/google/snappy/archive/refs/tags/1.1.8.zip -O snappy-1.1.8.zip
-            unzip -q snappy-1.1.8.zip
-            cd snappy-1.1.8
-            mkdir build
-            cd build
-            & $Env:CMAKE_BIN -G "$Env:CMAKE_GENERATOR" ..
-            msbuild.exe Snappy.sln -maxCpuCount -property:Configuration=Debug -property:Platform=x64
-      - run:
-          name: "Build RocksDB"
-          command: |
-            mkdir build
-            cd build
-            & $Env:CMAKE_BIN -G "$Env:CMAKE_GENERATOR" -DCMAKE_BUILD_TYPE=Debug -DOPTDBG=1 -DPORTABLE=1 -DSNAPPY=1 -DJNI=1 ..
-            cd ..
-            echo "Building with VS version: $Env:CMAKE_GENERATOR"
-            msbuild.exe build/rocksdb.sln -maxCpuCount -property:Configuration=Debug -property:Platform=x64
-      - run:
-          name: "Test RocksDB"
-          shell: powershell.exe
-          command: |
-            build_tools\run_ci_db_test.ps1 -SuiteRun arena_test,db_basic_test,db_test,db_test2,db_merge_operand_test,bloom_test,c_test,coding_test,crc32c_test,dynamic_bloom_test,env_basic_test,env_test,hash_test,random_test -Concurrency 16
   pre-steps-macos:
       steps:
         - pre-steps
@@ -152,15 +120,15 @@ commands:
     steps:
       - run:
           name: "Test low-variance benchmarks"
-          command: ./tools/benchmark_ci.py --db_dir /tmp/rocksdb-benchmark-datadir --output_dir /tmp/benchmark-results --num_keys 20000000
+          command: ./tools/benchmark_ci.py --db_dir /tmp/rocksdb-benchmark-datadir --output_dir /tmp/benchmark-results --num_keys 50000
           environment:
             LD_LIBRARY_PATH: /usr/local/lib
             # How long to run parts of the test(s)
-            DURATION_RO: 300
-            DURATION_RW: 500
+            DURATION_RO: 400
+            DURATION_RW: 700
             # Keep threads within physical capacity of server (much lower than default)
             NUM_THREADS: 1
-            MAX_BACKGROUND_JOBS: 4
+            MAX_BACKGROUND_JOBS: 3
             # Don't run a couple of "optional" initial tests
             CI_TESTS_ONLY: "true"
             # Reduce configured size of levels to ensure more levels in the leveled compaction LSM tree
@@ -170,11 +138,7 @@ commands:
             # The benchmark host has 32GB memory
             # The following values are tailored to work with that
             # Note, tests may not exercise the targeted issues if the memory is increased on new test hosts.
-            COMPRESSION_TYPE: "none"
-            CACHE_INDEX_AND_FILTER_BLOCKS: 1
-            MIN_LEVEL_TO_COMPRESS: 3
-            CACHE_SIZE_MB: 10240
-            MB_WRITE_PER_SEC: 2
+
 
   post-benchmarks:
     steps:
@@ -190,6 +154,11 @@ commands:
             true
 
 executors:
+  windows-2xlarge:
+    machine:
+      image: 'windows-server-2019-vs2019:stable'
+      resource_class: windows.2xlarge
+      shell: bash.exe
   linux-docker:
     docker:
       # The image configuration is build_tools/ubuntu20_image/Dockerfile
@@ -270,15 +239,15 @@ jobs:
       - pre-steps
       - run: ENCRYPTED_ENV=1 ROCKSDB_DISABLE_SNAPPY=1 ROCKSDB_DISABLE_ZLIB=1 ROCKSDB_DISABLE_BZIP=1 ROCKSDB_DISABLE_LZ4=1 ROCKSDB_DISABLE_ZSTD=1 make V=1 J=32 -j32 check
       - run: |
-          ./sst_dump --help | grep -E -q 'Supported compression types: kNoCompression$' # Verify no compiled in compression
+          ./sst_dump --help | egrep -q 'Supported compression types: kNoCompression$' # Verify no compiled in compression
       - post-steps
 
-  build-linux-static_lib-alt_namespace-status_checked:
+  build-linux-shared_lib-alt_namespace-status_checked:
     executor: linux-docker
     resource_class: 2xlarge
     steps:
       - pre-steps
-      - run: ASSERT_STATUS_CHECKED=1 TEST_UINT128_COMPAT=1 ROCKSDB_MODIFY_NPHASH=1 LIB_MODE=static OPT="-DROCKSDB_NAMESPACE=alternative_rocksdb_ns" make V=1 -j24 check
+      - run: ASSERT_STATUS_CHECKED=1 TEST_UINT128_COMPAT=1 ROCKSDB_MODIFY_NPHASH=1 LIB_MODE=shared OPT="-DROCKSDB_NAMESPACE=alternative_rocksdb_ns" make V=1 -j32 check
       - post-steps
 
   build-linux-release:
@@ -286,21 +255,11 @@ jobs:
     resource_class: 2xlarge
     steps:
       - checkout # check out the code in the project directory
-      - run: make V=1 -j32 LIB_MODE=shared release
-      - run: ls librocksdb.so # ensure shared lib built
-      - run: ./db_stress --version # ensure with gflags
-      - run: make clean
       - run: make V=1 -j32 release
-      - run: ls librocksdb.a # ensure static lib built
       - run: ./db_stress --version # ensure with gflags
       - run: make clean
       - run: apt-get remove -y libgflags-dev
-      - run: make V=1 -j32 LIB_MODE=shared release
-      - run: ls librocksdb.so # ensure shared lib built
-      - run: if ./db_stress --version; then false; else true; fi # ensure without gflags
-      - run: make clean
       - run: make V=1 -j32 release
-      - run: ls librocksdb.a # ensure static lib built
       - run: if ./db_stress --version; then false; else true; fi # ensure without gflags
       - post-steps
 
@@ -316,6 +275,27 @@ jobs:
       - run: USE_RTTI=1 DEBUG_LEVEL=0 make V=1 -j16 static_lib tools db_bench
       - run: if ./db_stress --version; then false; else true; fi # ensure without gflags
 
+  build-linux-lite:
+    executor: linux-docker
+    resource_class: large
+    steps:
+      - pre-steps
+      - run: LITE=1 make V=1 J=8 -j8 check
+      - post-steps
+
+  build-linux-lite-release:
+    executor: linux-docker
+    resource_class: large
+    steps:
+      - checkout # check out the code in the project directory
+      - run: LITE=1 make V=1 -j8 release
+      - run: ./db_stress --version # ensure with gflags
+      - run: make clean
+      - run: apt-get remove -y libgflags-dev
+      - run: LITE=1 make V=1 -j8 release
+      - run: if ./db_stress --version; then false; else true; fi # ensure without gflags
+      - post-steps
+
   build-linux-clang-no_test_run:
     executor: linux-docker
     resource_class: xlarge
@@ -420,10 +400,7 @@ jobs:
     steps:
       - checkout # check out the code in the project directory
       - run: apt-get update -y && apt-get install -y libgflags-dev
-      - run:
-          name: "Unity build"
-          command: make V=1 -j8 unity_test
-          no_output_timeout: 20m
+      - run: make V=1 -j8 unity_test
       - run: make V=1 -j8 -k check-headers # could be moved to a different build
       - post-steps
 
@@ -434,7 +411,7 @@ jobs:
       - pre-steps
       - setup-folly
       - build-folly
-      - run: USE_FOLLY=1 LIB_MODE=static CC=gcc-7 CXX=g++-7 V=1 make -j32 check # TODO: LIB_MODE only to work around unresolved linker failures
+      - run: USE_FOLLY=1 CC=gcc-7 CXX=g++-7 V=1 make -j32 check
       - post-steps
 
   build-linux-gcc-7-with-folly-lite-no-test:
@@ -480,7 +457,7 @@ jobs:
     resource_class: 2xlarge
     steps:
       - pre-steps
-      - run: LIB_MODE=static CC=gcc-11 CXX=g++-11 V=1 make -j32 all microbench # TODO: LIB_MODE only to work around unresolved linker failures
+      - run: CC=gcc-11 CXX=g++-11 V=1 make -j32 all microbench
       - post-steps
 
   build-linux-clang-13-no_test_run:
@@ -499,7 +476,7 @@ jobs:
       - pre-steps
       - setup-folly
       - build-folly
-      - run: CC=clang-13 CXX=clang++-13 LIB_MODE=static USE_CLANG=1 USE_FOLLY=1 COMPILE_WITH_UBSAN=1 COMPILE_WITH_ASAN=1 make -j32 check # TODO: LIB_MODE only to work around unresolved linker failures
+      - run: CC=clang-13 CXX=clang++-13 USE_CLANG=1 USE_FOLLY=1 COMPILE_WITH_UBSAN=1 COMPILE_WITH_ASAN=1 make -j32 check
       - post-steps
 
   # This job is only to make sure the microbench tests are able to run, the benchmark result is not meaningful as the CI host is changing.
@@ -516,7 +493,7 @@ jobs:
     resource_class: large
     steps:
       - pre-steps
-      - run: ulimit -S -n `ulimit -H -n` && make V=1 -j8 CRASH_TEST_EXT_ARGS='--duration=960 --max_key=2500000 --use_io_uring=0' blackbox_crash_test_with_atomic_flush
+      - run: ulimit -S -n `ulimit -H -n` && make V=1 -j8 CRASH_TEST_EXT_ARGS='--duration=960 --max_key=2500000' blackbox_crash_test_with_atomic_flush
       - post-steps
 
   build-linux-crashtest-tiered-storage-bb:
@@ -526,7 +503,7 @@ jobs:
       - pre-steps
       - run:
           name: "run crashtest"
-          command: ulimit -S -n `ulimit -H -n` && make V=1 -j32 CRASH_TEST_EXT_ARGS='--duration=10800 --use_io_uring=0' blackbox_crash_test_with_tiered_storage
+          command: ulimit -S -n `ulimit -H -n` && make V=1 -j32 CRASH_TEST_EXT_ARGS=--duration=10800 blackbox_crash_test_with_tiered_storage
           no_output_timeout: 100m
       - post-steps
 
@@ -537,39 +514,77 @@ jobs:
       - pre-steps
       - run:
           name: "run crashtest"
-          command: ulimit -S -n `ulimit -H -n` && make V=1 -j32 CRASH_TEST_EXT_ARGS='--duration=10800 --use_io_uring=0' whitebox_crash_test_with_tiered_storage
+          command: ulimit -S -n `ulimit -H -n` && make V=1 -j32 CRASH_TEST_EXT_ARGS=--duration=10800 whitebox_crash_test_with_tiered_storage
           no_output_timeout: 100m
       - post-steps
 
-  build-windows-vs2022:
-    executor:
-      name: win/server-2022
-      size: 2xlarge
+  build-windows:
+    executor: windows-2xlarge
+    parameters:
+      extra_cmake_opt:
+        default: ""
+        type: string
+      vs_year:
+        default: "2019"
+        type: string
+      cmake_generator:
+        default: "Visual Studio 16 2019"
+        type: string
     environment:
       THIRDPARTY_HOME: C:/Users/circleci/thirdparty
-      CMAKE_HOME: C:/Program Files/CMake
-      CMAKE_BIN: C:/Program Files/CMake/bin/cmake.exe
-      SNAPPY_HOME: C:/Users/circleci/thirdparty/snappy-1.1.8
-      SNAPPY_INCLUDE: C:/Users/circleci/thirdparty/snappy-1.1.8;C:/Users/circleci/thirdparty/snappy-1.1.8/build
-      SNAPPY_LIB_DEBUG: C:/Users/circleci/thirdparty/snappy-1.1.8/build/Debug/snappy.lib
-      CMAKE_GENERATOR: Visual Studio 17 2022
+      CMAKE_HOME: C:/Users/circleci/thirdparty/cmake-3.16.4-win64-x64
+      CMAKE_BIN: C:/Users/circleci/thirdparty/cmake-3.16.4-win64-x64/bin/cmake.exe
+      SNAPPY_HOME: C:/Users/circleci/thirdparty/snappy-1.1.7
+      SNAPPY_INCLUDE: C:/Users/circleci/thirdparty/snappy-1.1.7;C:/Users/circleci/thirdparty/snappy-1.1.7/build
+      SNAPPY_LIB_DEBUG: C:/Users/circleci/thirdparty/snappy-1.1.7/build/Debug/snappy.lib
+      VS_YEAR: <<parameters.vs_year>>
+      CMAKE_GENERATOR: <<parameters.cmake_generator>>
     steps:
-      - windows-build-steps
-
-  build-windows-vs2019:
-    executor:
-      name: win/server-2019
-      size: 2xlarge
-    environment:
-      THIRDPARTY_HOME: C:/Users/circleci/thirdparty
-      CMAKE_HOME: C:/Program Files/CMake
-      CMAKE_BIN: C:/Program Files/CMake/bin/cmake.exe
-      SNAPPY_HOME: C:/Users/circleci/thirdparty/snappy-1.1.8
-      SNAPPY_INCLUDE: C:/Users/circleci/thirdparty/snappy-1.1.8;C:/Users/circleci/thirdparty/snappy-1.1.8/build
-      SNAPPY_LIB_DEBUG: C:/Users/circleci/thirdparty/snappy-1.1.8/build/Debug/snappy.lib
-      CMAKE_GENERATOR: Visual Studio 16 2019
-    steps:
-      - windows-build-steps
+      - checkout
+      - run:
+          name: "Setup VS"
+          command: |
+            if [[ "${VS_YEAR}" == "2019" ]]; then
+              echo "VS2019 already present."
+            elif [[ "${VS_YEAR}" == "2017" ]]; then
+              echo "Installing VS2017..."
+              powershell .circleci/vs2017_install.ps1
+            elif [[ "${VS_YEAR}" == "2015" ]]; then
+              echo "Installing VS2015..."
+              powershell .circleci/vs2015_install.ps1
+            fi
+      - store_artifacts:
+          path: \Users\circleci\AppData\Local\Temp\vslogs.zip
+      - run:
+          name: "Install thirdparty dependencies"
+          command: |
+            mkdir ${THIRDPARTY_HOME}
+            cd ${THIRDPARTY_HOME}
+            echo "Installing CMake..."
+            curl --fail --silent --show-error --output cmake-3.16.4-win64-x64.zip --location https://github.com/Kitware/CMake/releases/download/v3.16.4/cmake-3.16.4-win64-x64.zip
+            unzip -q cmake-3.16.4-win64-x64.zip
+            echo "Building Snappy dependency..."
+            curl --fail --silent --show-error --output snappy-1.1.7.zip --location https://github.com/google/snappy/archive/1.1.7.zip
+            unzip -q snappy-1.1.7.zip
+            cd snappy-1.1.7
+            mkdir build
+            cd build
+            ${CMAKE_BIN} -G "${CMAKE_GENERATOR}" ..
+            msbuild.exe Snappy.sln -maxCpuCount -property:Configuration=Debug -property:Platform=x64
+      - run:
+          name: "Build RocksDB"
+          command: |
+            mkdir build
+            cd build
+            ${CMAKE_BIN} -G "${CMAKE_GENERATOR}" -DCMAKE_BUILD_TYPE=Debug -DOPTDBG=1 -DPORTABLE=1 -DSNAPPY=1 -DJNI=1 << parameters.extra_cmake_opt >> ..
+            cd ..
+            echo "Building with VS version: ${CMAKE_GENERATOR}"
+            msbuild.exe build/rocksdb.sln -maxCpuCount -property:Configuration=Debug -property:Platform=x64
+      - run:
+          name: "Test RocksDB"
+          shell: powershell.exe
+          command: |
+            build_tools\run_ci_db_test.ps1 -SuiteRun db_basic_test,db_test,db_test2,db_merge_operand_test,bloom_test,c_test,coding_test,crc32c_test,dynamic_bloom_test,env_basic_test,env_test,hash_test,random_test -Concurrency 16
 
   build-linux-java:
     executor: linux-docker
@@ -817,16 +832,18 @@ workflows:
       - build-linux-cmake-with-folly-coroutines
       - build-linux-cmake-with-benchmark
       - build-linux-encrypted_env-no_compression
+      - build-linux-lite
   jobs-linux-run-tests-san:
     jobs:
       - build-linux-clang10-asan
       - build-linux-clang10-ubsan
       - build-linux-clang10-mini-tsan
-      - build-linux-static_lib-alt_namespace-status_checked
+      - build-linux-shared_lib-alt_namespace-status_checked
   jobs-linux-no-test-run:
     jobs:
       - build-linux-release
       - build-linux-release-rtti
+      - build-linux-lite-release
       - build-examples
       - build-fuzzers
       - build-linux-clang-no_test_run
@@ -842,8 +859,15 @@ workflows:
       - build-linux-mini-crashtest
   jobs-windows:
     jobs:
-      - build-windows-vs2022
-      - build-windows-vs2019
+      - build-windows:
+          name: "build-windows-vs2019"
+      - build-windows:
+          name: "build-windows-vs2019-cxx20"
+          extra_cmake_opt: -DCMAKE_CXX_STANDARD=20
+      - build-windows:
+          name: "build-windows-vs2017"
+          vs_year: "2017"
+          cmake_generator: "Visual Studio 15 Win64"
       - build-cmake-mingw
   jobs-java:
     jobs:

.circleci/vs2015_install.ps1

@@ -0,0 +1,24 @@
+$VS_DOWNLOAD_LINK = "https://go.microsoft.com/fwlink/?LinkId=691126"
+$COLLECT_DOWNLOAD_LINK = "https://aka.ms/vscollect.exe"
+curl.exe --retry 3 -kL $VS_DOWNLOAD_LINK --output vs_installer.exe
+if ($LASTEXITCODE -ne 0) {
+    echo "Download of the VS 2015 installer failed"
+    exit 1
+}
+$VS_INSTALL_ARGS = @("/Quiet", "/NoRestart")
+$process = Start-Process "${PWD}\vs_installer.exe" -ArgumentList $VS_INSTALL_ARGS -NoNewWindow -Wait -PassThru
+Remove-Item -Path vs_installer.exe -Force
+$exitCode = $process.ExitCode
+if (($exitCode -ne 0) -and ($exitCode -ne 3010)) {
+    echo "VS 2015 installer exited with code $exitCode, which should be one of [0, 3010]."
+    curl.exe --retry 3 -kL $COLLECT_DOWNLOAD_LINK --output Collect.exe
+    if ($LASTEXITCODE -ne 0) {
+        echo "Download of the VS Collect tool failed."
+        exit 1
+    }
+    Start-Process "${PWD}\Collect.exe" -NoNewWindow -Wait -PassThru
+    New-Item -Path "C:\w\build-results" -ItemType "directory" -Force
+    Copy-Item -Path "C:\Users\circleci\AppData\Local\Temp\vslogs.zip" -Destination "C:\w\build-results\"
+    exit 1
+}
+echo "VS 2015 installed."

.circleci/vs2017_install.ps1

@@ -0,0 +1,35 @@
+$VS_DOWNLOAD_LINK = "https://aka.ms/vs/15/release/vs_buildtools.exe"
+$COLLECT_DOWNLOAD_LINK = "https://aka.ms/vscollect.exe"
+$VS_INSTALL_ARGS = @("--nocache","--quiet","--wait", "--add Microsoft.VisualStudio.Workload.VCTools",
+                                                     "--add Microsoft.VisualStudio.Component.VC.Tools.14.13",
+                                                     "--add Microsoft.Component.MSBuild",
+                                                     "--add Microsoft.VisualStudio.Component.Roslyn.Compiler",
+                                                     "--add Microsoft.VisualStudio.Component.TextTemplating",
+                                                     "--add Microsoft.VisualStudio.Component.VC.CoreIde",
+                                                     "--add Microsoft.VisualStudio.Component.VC.Redist.14.Latest",
+                                                     "--add Microsoft.VisualStudio.ComponentGroup.NativeDesktop.Core",
+                                                     "--add Microsoft.VisualStudio.Component.VC.Tools.x86.x64",
+                                                     "--add Microsoft.VisualStudio.ComponentGroup.NativeDesktop.Win81")
+
+curl.exe --retry 3 -kL $VS_DOWNLOAD_LINK --output vs_installer.exe
+if ($LASTEXITCODE -ne 0) {
+    echo "Download of the VS 2017 installer failed"
+    exit 1
+}
+
+$process = Start-Process "${PWD}\vs_installer.exe" -ArgumentList $VS_INSTALL_ARGS -NoNewWindow -Wait -PassThru
+Remove-Item -Path vs_installer.exe -Force
+$exitCode = $process.ExitCode
+if (($exitCode -ne 0) -and ($exitCode -ne 3010)) {
+    echo "VS 2017 installer exited with code $exitCode, which should be one of [0, 3010]."
+    curl.exe --retry 3 -kL $COLLECT_DOWNLOAD_LINK --output Collect.exe
+    if ($LASTEXITCODE -ne 0) {
+        echo "Download of the VS Collect tool failed."
+        exit 1
+    }
+    Start-Process "${PWD}\Collect.exe" -NoNewWindow -Wait -PassThru
+    New-Item -Path "C:\w\build-results" -ItemType "directory" -Force
+    Copy-Item -Path "C:\Users\circleci\AppData\Local\Temp\vslogs.zip" -Destination "C:\w\build-results\"
+    exit 1
+}
+echo "VS 2017 installed."

.github/workflows/sanity_check.yml

@@ -1,8 +1,5 @@
 name: Check buck targets and code format
 on: [push, pull_request]
-permissions:
-  contents: read
-
 jobs:
   check:
     name: Check TARGETS file and code format

.gitignore

@@ -56,7 +56,6 @@ io_tracer_parser
 .DS_Store
 .vs
 .vscode
-.clangd
 
 java/out
 java/target

CMakeLists.txt

@@ -245,21 +245,11 @@ if(CMAKE_SYSTEM_PROCESSOR MATCHES "s390x")
   endif(HAS_S390X_MARCH_NATIVE)
 endif(CMAKE_SYSTEM_PROCESSOR MATCHES "s390x")
 
-if(CMAKE_SYSTEM_PROCESSOR MATCHES "loongarch64")
-  CHECK_C_COMPILER_FLAG("-march=loongarch64" HAS_LOONGARCH64)
-  if(HAS_LOONGARCH64)
-    set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -mcpu=loongarch64 -mtune=loongarch64")
-    set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -mcpu=loongarch64 -mtune=loongarch64")
-  endif(HAS_LOONGARCH64)
-endif(CMAKE_SYSTEM_PROCESSOR MATCHES "loongarch64")
-
 option(PORTABLE "build a portable binary" OFF)
 option(FORCE_SSE42 "force building with SSE4.2, even when PORTABLE=ON" OFF)
 option(FORCE_AVX "force building with AVX, even when PORTABLE=ON" OFF)
 option(FORCE_AVX2 "force building with AVX2, even when PORTABLE=ON" OFF)
 if(PORTABLE)
-  add_definitions(-DROCKSDB_PORTABLE)
-
   # MSVC does not need a separate compiler flag to enable SSE4.2; if nmmintrin.h
   # is available, it is available by default.
   if(FORCE_SSE42 AND NOT MSVC)
@@ -283,9 +273,6 @@ if(PORTABLE)
     if(CMAKE_SYSTEM_PROCESSOR MATCHES "^s390x")
       set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -march=z196")
     endif()
-    if(CMAKE_SYSTEM_PROCESSOR MATCHES "^loongarch64")
-      set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -march=loongarch64")
-    endif()
   endif()
 else()
   if(MSVC)
@@ -305,7 +292,8 @@ if(NOT MSVC)
   set(CMAKE_REQUIRED_FLAGS "-msse4.2 -mpclmul")
 endif()
 
-CHECK_CXX_SOURCE_COMPILES("
+if (NOT PORTABLE OR FORCE_SSE42)
+  CHECK_CXX_SOURCE_COMPILES("
 #include <cstdint>
 #include <nmmintrin.h>
 #include <wmmintrin.h>
@@ -317,11 +305,12 @@ int main() {
   auto d = _mm_cvtsi128_si64(c);
 }
 " HAVE_SSE42)
-if(HAVE_SSE42)
-  add_definitions(-DHAVE_SSE42)
-  add_definitions(-DHAVE_PCLMUL)
-elseif(FORCE_SSE42)
-  message(FATAL_ERROR "FORCE_SSE42=ON but unable to compile with SSE4.2 enabled")
+  if(HAVE_SSE42)
+    add_definitions(-DHAVE_SSE42)
+    add_definitions(-DHAVE_PCLMUL)
+  elseif(FORCE_SSE42)
+    message(FATAL_ERROR "FORCE_SSE42=ON but unable to compile with SSE4.2 enabled")
+  endif()
 endif()
 
 # Check if -latomic is required or not
@@ -496,6 +485,12 @@ if(CMAKE_COMPILER_IS_GNUCXX)
   set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -fno-builtin-memcmp")
 endif()
 
+option(ROCKSDB_LITE "Build RocksDBLite version" OFF)
+if(ROCKSDB_LITE)
+  add_definitions(-DROCKSDB_LITE)
+  set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -fno-exceptions -Os")
+endif()
+
 if(CMAKE_SYSTEM_NAME MATCHES "Cygwin")
   add_definitions(-fno-builtin-memcmp -DCYGWIN)
 elseif(CMAKE_SYSTEM_NAME MATCHES "Darwin")
@@ -654,14 +649,12 @@ set(SOURCES
         cache/cache.cc
         cache/cache_entry_roles.cc
         cache/cache_key.cc
-        cache/cache_helpers.cc
         cache/cache_reservation_manager.cc
         cache/charged_cache.cc
         cache/clock_cache.cc
         cache/compressed_secondary_cache.cc
+        cache/fast_lru_cache.cc
         cache/lru_cache.cc
-        cache/secondary_cache.cc
-        cache/secondary_cache_adapter.cc
         cache/sharded_cache.cc
         db/arena_wrapped_db_iter.cc
         db/blob/blob_contents.cc
@@ -748,7 +741,6 @@ set(SOURCES
         db/write_batch.cc
         db/write_batch_base.cc
         db/write_controller.cc
-        db/write_stall_stats.cc
         db/write_thread.cc
         env/composite_env.cc
         env/env.cc
@@ -804,7 +796,6 @@ set(SOURCES
         options/options.cc
         options/options_helper.cc
         options/options_parser.cc
-        port/mmap.cc
         port/stack_trace.cc
         table/adaptive/adaptive_table_factory.cc
         table/block_based/binary_search_index_reader.cc
@@ -814,7 +805,6 @@ set(SOURCES
         table/block_based/block_based_table_iterator.cc
         table/block_based/block_based_table_reader.cc
         table/block_based/block_builder.cc
-        table/block_based/block_cache.cc
         table/block_based/block_prefetcher.cc
         table/block_based/block_prefix_index.cc
         table/block_based/data_block_hash_index.cc
@@ -840,7 +830,6 @@ set(SOURCES
         table/get_context.cc
         table/iterator.cc
         table/merging_iterator.cc
-        table/compaction_merging_iterator.cc
         table/meta_blocks.cc
         table/persistent_cache_helper.cc
         table/plain/plain_table_bloom.cc
@@ -882,7 +871,6 @@ set(SOURCES
         util/compression_context_cache.cc
         util/concurrent_task_limiter_impl.cc
         util/crc32c.cc
-        util/data_structure.cc
         util/dynamic_bloom.cc
         util/hash.cc
         util/murmurhash.cc
@@ -990,12 +978,6 @@ if ( ROCKSDB_PLUGINS )
         plugin/${plugin}/${src}
         PROPERTIES COMPILE_FLAGS "${${plugin}_COMPILE_FLAGS}")
     endforeach()
-    foreach (test ${${plugin}_TESTS})
-      list(APPEND PLUGIN_TESTS plugin/${plugin}/${test})
-      set_source_files_properties(
-        plugin/${plugin}/${test}
-        PROPERTIES COMPILE_FLAGS "${${plugin}_COMPILE_FLAGS}")
-    endforeach()
     foreach (path ${${plugin}_INCLUDE_PATHS})
       include_directories(${path})
     endforeach()
@@ -1083,34 +1065,28 @@ set(ROCKSDB_PLUGIN_BUILTINS "")
 message(STATUS "ROCKSDB PLUGINS TO BUILD ${ROCKSDB_PLUGINS}")
 foreach(PLUGIN IN LISTS PLUGINS)
   set(PLUGIN_ROOT "${CMAKE_CURRENT_SOURCE_DIR}/plugin/${PLUGIN}/")
-  message(STATUS "PLUGIN ${PLUGIN} including rocksb plugin ${PLUGIN_ROOT}")
+  message("including rocksb plugin ${PLUGIN_ROOT}")
   set(PLUGINMKFILE "${PLUGIN_ROOT}${PLUGIN}.mk")
   if (NOT EXISTS ${PLUGINMKFILE})
-    message(FATAL_ERROR "PLUGIN ${PLUGIN} Missing plugin makefile: ${PLUGINMKFILE}")
+    message(FATAL_ERROR "Missing plugin makefile: ${PLUGINMKFILE}")
   endif()
   file(READ ${PLUGINMKFILE} PLUGINMK)
-
   string(REGEX MATCH "SOURCES = ([^\n]*)" FOO ${PLUGINMK})
   set(MK_SOURCES ${CMAKE_MATCH_1})
   separate_arguments(MK_SOURCES)
   foreach(MK_FILE IN LISTS MK_SOURCES)
     list(APPEND SOURCES "${PLUGIN_ROOT}${MK_FILE}")
-    message(STATUS "PLUGIN ${PLUGIN} Appending ${PLUGIN_ROOT}${MK_FILE} to SOURCES")
   endforeach()
-
   string(REGEX MATCH "_FUNC = ([^\n]*)" FOO ${PLUGINMK})
   if (NOT ${CMAKE_MATCH_1} STREQUAL "")
     string(APPEND ROCKSDB_PLUGIN_BUILTINS "{\"${PLUGIN}\", " ${CMAKE_MATCH_1} "},")
     string(APPEND ROCKSDB_PLUGIN_EXTERNS "int " ${CMAKE_MATCH_1} "(ROCKSDB_NAMESPACE::ObjectLibrary&, const std::string&); ")
   endif()
-
   string(REGEX MATCH "_LIBS = ([^\n]*)" FOO ${PLUGINMK})
-  separate_arguments(CMAKE_MATCH_1)
-  foreach(MK_LIB IN LISTS CMAKE_MATCH_1)
-    list(APPEND THIRDPARTY_LIBS "${MK_LIB}")
-  endforeach()
-  message(STATUS "PLUGIN ${PLUGIN} THIRDPARTY_LIBS=${THIRDPARTY_LIBS}")
-
+  if (NOT ${CMAKE_MATCH_1} STREQUAL "")
+    list(APPEND THIRDPARTY_LIBS "${CMAKE_MATCH_1}")
+  endif()
+  message("THIRDPARTY_LIBS=${THIRDPARTY_LIBS}")
   #TODO: We need to set any compile/link-time flags and add any link libraries
 endforeach()
 
@@ -1281,7 +1257,6 @@ if(WITH_TESTS OR WITH_BENCHMARK_TOOLS)
   add_subdirectory(third-party/gtest-1.8.1/fused-src/gtest)
   add_library(testharness STATIC
   test_util/mock_time_env.cc
-  test_util/secondary_cache_test_util.cc
   test_util/testharness.cc)
   target_link_libraries(testharness gtest)
 endif()
@@ -1488,7 +1463,6 @@ if(WITH_TESTS)
         utilities/ttl/ttl_test.cc
         utilities/util_merge_operators_test.cc
         utilities/write_batch_with_index/write_batch_with_index_test.cc
-	${PLUGIN_TESTS}
     )
   endif()
 
@@ -1592,9 +1566,9 @@ endif()
 
 option(WITH_TRACE_TOOLS "build with trace tools" ON)
 if(WITH_TRACE_TOOLS)
-  add_executable(block_cache_trace_analyzer${ARTIFACT_SUFFIX}
+  add_executable(block_cache_trace_analyzer_tool${ARTIFACT_SUFFIX}
     tools/block_cache_analyzer/block_cache_trace_analyzer_tool.cc)
-  target_link_libraries(block_cache_trace_analyzer${ARTIFACT_SUFFIX}
+  target_link_libraries(block_cache_trace_analyzer_tool${ARTIFACT_SUFFIX}
     ${ROCKSDB_LIB} ${GFLAGS_LIB} ${FOLLY_LIBS})
 
   add_executable(trace_analyzer${ARTIFACT_SUFFIX}

HISTORY.md

@@ -1,180 +1,42 @@
 # Rocksdb Change Log
-## 8.2.1 (05/04/2023)
-* Public API comments and a Meta-internal fix to the buck build
-
-## 8.2.0 (04/24/2023)
-### Public API Changes
-* `SstFileWriter::DeleteRange()` now returns `Status::InvalidArgument` if the range's end key comes before its start key according to the user comparator. Previously the behavior was undefined.
-* Add `multi_get_for_update` to C API.
-* Remove unnecessary constructor for CompressionOptions.
-
-### Behavior changes
-* Changed default block cache size from an 8MB to 32MB LRUCache, which increases the default number of cache shards from 16 to 64. This change is intended to minimize cache mutex contention under stress conditions. See https://github.com/facebook/rocksdb/wiki/Block-Cache for more information.
-* For level compaction with `level_compaction_dynamic_level_bytes=true`, RocksDB now trivially moves levels down to fill LSM starting from bottommost level during DB open. See more in comments for option `level_compaction_dynamic_level_bytes` (#11321).
-* User-provided `ReadOptions` take effect for more reads of non-`CacheEntryRole::kDataBlock` blocks.
-* For level compaction with `level_compaction_dynamic_level_bytes=true`, RocksDB now drains unnecessary levels through background compaction automatically (#11340). This together with #11321 makes it automatic to migrate other compaction settings to level compaction with `level_compaction_dynamic_level_bytes=true`. In addition, a live DB that becomes smaller will now have unnecessary levels drained which can help to reduce read and space amp.
-* If `CompactRange()` is called with `CompactRangeOptions::bottommost_level_compaction=kForce*` to compact from L0 to L1, RocksDB now will try to do trivial move from L0 to L1 and then do an intra L1 compaction, instead of a L0 to L1 compaction with trivial move disabled (#11375)).
-
-### Bug Fixes
-* In the DB::VerifyFileChecksums API, ensure that file system reads of SST files are equal to the readahead_size in ReadOptions, if specified. Previously, each read was 2x the readahead_size.
-* In block cache tracing, fixed some cases of bad hit/miss information (and more) with MultiGet.
-
-### New Features
-* Add experimental `PerfContext` counters `iter_{next|prev|seek}_count` for db iterator, each counting the times of corresponding API being called.
-* Allow runtime changes to whether `WriteBufferManager` allows stall or not by calling `SetAllowStall()`
-* Added statistics tickers BYTES_COMPRESSED_FROM, BYTES_COMPRESSED_TO, BYTES_COMPRESSION_BYPASSED, BYTES_COMPRESSION_REJECTED, NUMBER_BLOCK_COMPRESSION_BYPASSED, and NUMBER_BLOCK_COMPRESSION_REJECTED. Disabled/deprecated histograms BYTES_COMPRESSED and BYTES_DECOMPRESSED, and ticker NUMBER_BLOCK_NOT_COMPRESSED. The new tickers offer more inight into compression ratios, rejected vs. disabled compression, etc. (#11388)
-* New statistics `rocksdb.file.read.{flush|compaction}.micros` that measure read time of block-based SST tables or blob files during flush or compaction.
-
-## 8.1.0 (03/18/2023)
-### Behavior changes
-* Compaction output file cutting logic now considers range tombstone start keys. For example, SST partitioner now may receive ParitionRequest for range tombstone start keys.
-* If the async_io ReadOption is specified for MultiGet or NewIterator on a platform that doesn't support IO uring, the option is ignored and synchronous IO is used.
-
-### Bug Fixes
-* Fixed an issue for backward iteration when user defined timestamp is enabled in combination with BlobDB.
-* Fixed a couple of cases where a Merge operand encountered during iteration wasn't reflected in the `internal_merge_count` PerfContext counter.
-* Fixed a bug in CreateColumnFamilyWithImport()/ExportColumnFamily() which did not support range tombstones (#11252).
-* Fixed a bug where an excluded column family from an atomic flush contains unflushed data that should've been included in this atomic flush (i.e, data of seqno less than the max seqno of this atomic flush), leading to potential data loss in this excluded column family when `WriteOptions::disableWAL == true` (#11148).
-
-### New Features
-* Add statistics rocksdb.secondary.cache.filter.hits, rocksdb.secondary.cache.index.hits, and rocksdb.secondary.cache.filter.hits
-* Added a new PerfContext counter `internal_merge_point_lookup_count` which tracks the number of Merge operands applied while serving point lookup queries.
-* Add new statistics rocksdb.table.open.prefetch.tail.read.bytes, rocksdb.table.open.prefetch.tail.{miss|hit}
-* Add support for SecondaryCache with HyperClockCache (`HyperClockCacheOptions` inherits `secondary_cache` option from `ShardedCacheOptions`)
-* Add new db properties `rocksdb.cf-write-stall-stats`, `rocksdb.db-write-stall-stats`and APIs to examine them in a structured way. In particular, users of `GetMapProperty()` with property `kCFWriteStallStats`/`kDBWriteStallStats` can now use the functions in `WriteStallStatsMapKeys` to find stats in the map.
-
-### Public API Changes
-* Changed various functions and features in `Cache` that are mostly relevant to custom implementations or wrappers. Especially, asychronous lookup functionality is moved from `Lookup()` to a new `StartAsyncLookup()` function.
-
-## 8.0.0 (02/19/2023)
-### Behavior changes
-* `ReadOptions::verify_checksums=false` disables checksum verification for more reads of non-`CacheEntryRole::kDataBlock` blocks.
-* In case of scan with async_io enabled, if posix doesn't support IOUring, Status::NotSupported error will be returned to the users. Initially that error was swallowed and reads were switched to synchronous reads.
-
-### Bug Fixes
-* Fixed a data race on `ColumnFamilyData::flush_reason` caused by concurrent flushes.
-* Fixed an issue in `Get` and `MultiGet` when user-defined timestamps is enabled in combination with BlobDB.
-* Fixed some atypical behaviors for `LockWAL()` such as allowing concurrent/recursive use and not expecting `UnlockWAL()` after non-OK result. See API comments.
-* Fixed a feature interaction bug where for blobs `GetEntity` would expose the blob reference instead of the blob value.
-* Fixed `DisableManualCompaction()` and `CompactRangeOptions::canceled` to cancel compactions even when they are waiting on conflicting compactions to finish
-* Fixed a bug in which a successful `GetMergeOperands()` could transiently return `Status::MergeInProgress()`
-* Return the correct error (Status::NotSupported()) to MultiGet caller when ReadOptions::async_io flag is true and IO uring is not enabled. Previously, Status::Corruption() was being returned when the actual failure was lack of async IO support.
-* Fixed a bug in DB open/recovery from a compressed WAL that was caused due to incorrect handling of certain record fragments with the same offset within a WAL block.
-
-### Feature Removal
-* Remove RocksDB Lite.
-* The feature block_cache_compressed is removed. Statistics related to it are removed too.
-* Remove deprecated Env::LoadEnv(). Use Env::CreateFromString() instead.
-* Remove deprecated FileSystem::Load(). Use FileSystem::CreateFromString() instead.
-* Removed the deprecated version of these utility functions and the corresponding Java bindings: `LoadOptionsFromFile`, `LoadLatestOptions`, `CheckOptionsCompatibility`.
-* Remove the FactoryFunc from the LoadObject method from the Customizable helper methods.
-
-### Public API Changes
-* Moved rarely-needed Cache class definition to new advanced_cache.h, and added a CacheWrapper class to advanced_cache.h. Minor changes to SimCache API definitions.
-* Completely removed the following deprecated/obsolete statistics: the tickers `BLOCK_CACHE_INDEX_BYTES_EVICT`, `BLOCK_CACHE_FILTER_BYTES_EVICT`, `BLOOM_FILTER_MICROS`, `NO_FILE_CLOSES`, `STALL_L0_SLOWDOWN_MICROS`, `STALL_MEMTABLE_COMPACTION_MICROS`, `STALL_L0_NUM_FILES_MICROS`, `RATE_LIMIT_DELAY_MILLIS`, `NO_ITERATORS`, `NUMBER_FILTERED_DELETES`, `WRITE_TIMEDOUT`, `BLOB_DB_GC_NUM_KEYS_OVERWRITTEN`, `BLOB_DB_GC_NUM_KEYS_EXPIRED`, `BLOB_DB_GC_BYTES_OVERWRITTEN`, `BLOB_DB_GC_BYTES_EXPIRED`, `BLOCK_CACHE_COMPRESSION_DICT_BYTES_EVICT` as well as the histograms `STALL_L0_SLOWDOWN_COUNT`, `STALL_MEMTABLE_COMPACTION_COUNT`, `STALL_L0_NUM_FILES_COUNT`, `HARD_RATE_LIMIT_DELAY_COUNT`, `SOFT_RATE_LIMIT_DELAY_COUNT`, `BLOB_DB_GC_MICROS`, and `NUM_DATA_BLOCKS_READ_PER_LEVEL`. Note that as a result, the C++ enum values of the still supported statistics have changed. Developers are advised to not rely on the actual numeric values.
-* Deprecated IngestExternalFileOptions::write_global_seqno and change default to false. This option only needs to be set to true to generate a DB compatible with RocksDB versions before 5.16.0.
-* Remove deprecated APIs `GetColumnFamilyOptionsFrom{Map|String}(const ColumnFamilyOptions&, ..)`, `GetDBOptionsFrom{Map|String}(const DBOptions&, ..)`, `GetBlockBasedTableOptionsFrom{Map|String}(const BlockBasedTableOptions& table_options, ..)` and ` GetPlainTableOptionsFrom{Map|String}(const PlainTableOptions& table_options,..)`.
-* Added a subcode of `Status::Corruption`, `Status::SubCode::kMergeOperatorFailed`, for users to identify corruption failures originating in the merge operator, as opposed to RocksDB's internally identified data corruptions
-
-### Build Changes
-* The `make` build now builds a shared library by default instead of a static library. Use `LIB_MODE=static` to override.
-
-### New Features
-* Compaction filters are now supported for wide-column entities by means of the `FilterV3` API. See the comment of the API for more details.
-* Added `do_not_compress_roles` to `CompressedSecondaryCacheOptions` to disable compression on certain kinds of block. Filter blocks are now not compressed by CompressedSecondaryCache by default.
-* Added a new `MultiGetEntity` API that enables batched wide-column point lookups. See the API comments for more details.
-
-## 7.10.0 (01/23/2023)
-### Behavior changes
-* Make best-efforts recovery verify SST unique ID before Version construction (#10962)
-* Introduce `epoch_number` and sort L0 files by `epoch_number` instead of `largest_seqno`. `epoch_number` represents the order of a file being flushed or ingested/imported. Compaction output file will be assigned with the minimum `epoch_number` among input files'. For L0, larger `epoch_number` indicates newer L0 file.
-
+## 7.7.8 (11/27/2022)
 ### Bug Fixes
+* Fix failed memtable flush retry bug that could cause wrongly ordered updates, which would surface to writers as `Status::Corruption` in case of `force_consistency_checks=true` (default). It affects use cases that enable both parallel flush (`max_background_flushes > 1` or `max_background_jobs >= 8`) and non-default memtable count (`max_write_buffer_number > 2`).
+* Tiered Storage: fixed excessive keys written to penultimate level in non-debug builds.
 * Fixed a regression in iterator where range tombstones after `iterate_upper_bound` is processed.
-* Fixed a memory leak in MultiGet with async_io read option, caused by IO errors during table file open
-* Fixed a bug that multi-level FIFO compaction deletes one file in non-L0 even when `CompactionOptionsFIFO::max_table_files_size` is no exceeded since #10348 or 7.8.0.
-* Fixed a bug caused by `DB::SyncWAL()` affecting `track_and_verify_wals_in_manifest`. Without the fix, application may see "open error: Corruption: Missing WAL with log number" while trying to open the db. The corruption is a false alarm but prevents DB open (#10892).
-* Fixed a BackupEngine bug in which RestoreDBFromLatestBackup would fail if the latest backup was deleted and there is another valid backup available.
-* Fix L0 file misorder corruption caused by ingesting files of overlapping seqnos with memtable entries' through introducing `epoch_number`. Before the fix, `force_consistency_checks=true` may catch the corruption before it's exposed to readers, in which case writes returning `Status::Corruption` would be expected. Also replace the previous incomplete fix (#5958) to the same corruption with this new and more complete fix.
-* Fixed a bug in LockWAL() leading to re-locking mutex (#11020).
-* Fixed a heap use after free bug in async scan prefetching when the scan thread and another thread try to read and load the same seek block into cache.
-* Fixed a heap use after free in async scan prefetching if dictionary compression is enabled, in which case sync read of the compression dictionary gets mixed with async prefetching
-* Fixed a data race bug of `CompactRange()` under `change_level=true` acts on overlapping range with an ongoing file ingestion for level compaction. This will either result in overlapping file ranges corruption at a certain level caught by `force_consistency_checks=true` or protentially two same keys both with seqno 0 in two different levels (i.e, new data ends up in lower/older level). The latter will be caught by assertion in debug build but go silently and result in read returning wrong result in release build. This fix is general so it also replaced previous fixes to a similar problem for `CompactFiles()` (#4665), general `CompactRange()` and auto compaction (commit 5c64fb6 and 87dfc1d).
-* Fixed a bug in compaction output cutting where small output files were produced due to TTL file cutting states were not being updated (#11075).
 
-### New Features
-* When an SstPartitionerFactory is configured, CompactRange() now automatically selects for compaction any files overlapping a partition boundary that is in the compaction range, even if no actual entries are in the requested compaction range. With this feature, manual compaction can be used to (re-)establish SST partition points when SstPartitioner changes, without a full compaction.
-* Add BackupEngine feature to exclude files from backup that are known to be backed up elsewhere, using `CreateBackupOptions::exclude_files_callback`. To restore the DB, the excluded files must be provided in alternative backup directories using `RestoreOptions::alternate_dirs`.
-
-### Public API Changes
-* Substantial changes have been made to the Cache class to support internal development goals. Direct use of Cache class members is discouraged and further breaking modifications are expected in the future. SecondaryCache has some related changes and implementations will need to be updated. (Unlike Cache, SecondaryCache is still intended to support user implementations, and disruptive changes will be avoided.) (#10975)
-* Add `MergeOperationOutput::op_failure_scope` for merge operator users to control the blast radius of merge operator failures. Existing merge operator users do not need to make any change to preserve the old behavior
-
-### Performance Improvements
-* Updated xxHash source code, which should improve kXXH3 checksum speed, at least on ARM (#11098).
-* Improved CPU efficiency of DB reads, from block cache access improvements (#10975).
-
-## 7.9.0 (11/21/2022)
-### Performance Improvements
-* Fixed an iterator performance regression for delete range users when scanning through a consecutive sequence of range tombstones (#10877).
+## 7.7.7 (11/15/2022)
+### Bug Fixes
+* Fixed a regression in scan for async_io. During seek, valid buffers were getting cleared causing a regression.
 
+## 7.7.6 (11/03/2022)
 ### Bug Fixes
 * Fix memory corruption error in scans if async_io is enabled. Memory corruption happened if there is IOError while reading the data leading to empty buffer and other buffer already in progress of async read goes again for reading.
-* Fix failed memtable flush retry bug that could cause wrongly ordered updates, which would surface to writers as `Status::Corruption` in case of `force_consistency_checks=true` (default). It affects use cases that enable both parallel flush (`max_background_flushes > 1` or `max_background_jobs >= 8`) and non-default memtable count (`max_write_buffer_number > 2`).
-* Fixed an issue where the `READ_NUM_MERGE_OPERANDS` ticker was not updated when the base key-value or tombstone was read from an SST file.
-* Fixed a memory safety bug when using a SecondaryCache with `block_cache_compressed`. `block_cache_compressed` no longer attempts to use SecondaryCache features.
-* Fixed a regression in scan for async_io. During seek, valid buffers were getting cleared causing a regression.
-* Tiered Storage: fixed excessive keys written to penultimate level in non-debug builds.
-
-### New Features
-* Add basic support for user-defined timestamp to Merge (#10819).
-* Add stats for ReadAsync time spent and async read errors.
-* Basic support for the wide-column data model is now available. Wide-column entities can be stored using the `PutEntity` API, and retrieved using `GetEntity` and the new `columns` API of iterator. For compatibility, the classic APIs `Get` and `MultiGet`, as well as iterator's `value` API return the value of the anonymous default column of wide-column entities; also, `GetEntity` and iterator's `columns` return any plain key-values in the form of an entity which only has the anonymous default column. `Merge` (and `GetMergeOperands`) currently also apply to the default column; any other columns of entities are unaffected by `Merge` operations. Note that some features like compaction filters, transactions, user-defined timestamps, and the SST file writer do not yet support wide-column entities; also, there is currently no `MultiGet`-like API to retrieve multiple entities at once. We plan to gradually close the above gaps and also implement new features like column-level operations (e.g. updating or querying only certain columns of an entity).
-* Marked HyperClockCache as a production-ready alternative to LRUCache for the block cache. HyperClockCache greatly improves hot-path CPU efficiency under high parallel load or high contention, with some documented caveats and limitations. As much as 4.5x higher ops/sec vs. LRUCache has been seen in db_bench under high parallel load.
-* Add periodic diagnostics to info_log (LOG file) for HyperClockCache block cache if performance is degraded by bad `estimated_entry_charge` option.
-
-### Public API Changes
-* Marked `block_cache_compressed` as a deprecated feature. Use SecondaryCache instead.
-* Added a `SecondaryCache::InsertSaved()` API, with default implementation depending on `Insert()`. Some implementations might need to add a custom implementation of `InsertSaved()`. (Details in API comments.)
-
-## 7.8.0 (10/22/2022)
-### New Features
-* `DeleteRange()` now supports user-defined timestamp.
-* Provide support for async_io with tailing iterators when ReadOptions.tailing is enabled during scans.
-* Tiered Storage: allow data moving up from the last level to the penultimate level if the input level is penultimate level or above.
-* Added `DB::Properties::kFastBlockCacheEntryStats`, which is similar to `DB::Properties::kBlockCacheEntryStats`, except returns cached (stale) values in more cases to reduce overhead.
-* FIFO compaction now supports migrating from a multi-level DB via DB::Open(). During the migration phase, FIFO compaction picker will:
-* picks the sst file with the smallest starting key in the bottom-most non-empty level.
-* Note that during the migration phase, the file purge order will only be an approximation of "FIFO" as files in lower-level might sometime contain newer keys than files in upper-level.
-* Added an option `ignore_max_compaction_bytes_for_input` to ignore max_compaction_bytes limit when adding files to be compacted from input level. This should help reduce write amplification. The option is enabled by default.
-* Tiered Storage: allow data moving up from the last level even if it's a last level only compaction, as long as the penultimate level is empty.
-* Add a new option IOOptions.do_not_recurse that can be used by underlying file systems to skip recursing through sub directories and list only files in GetChildren API.
-* Add option `preserve_internal_time_seconds` to preserve the time information for the latest data. Which can be used to determine the age of data when `preclude_last_level_data_seconds` is enabled. The time information is attached with SST in table property `rocksdb.seqno.time.map` which can be parsed by tool ldb or sst_dump.
 
+## 7.7.5 (10/28/2022)
+### Bug Fixes
+* Fixed an iterator performance regression for delete range users when scanning through a consecutive sequence of range tombstones (#10877).
+
+## 7.7.4 (10/28/2022)
+### Bug Fixes
+* Fixed a case of calling malloc_usable_size on result of operator new[].
+
+## 7.7.3 (10/11/2022)
+### Bug Fixes
+* Fixed a memory safety bug in experimental HyperClockCache (#10768)
+
+## 7.7.2 (10/05/2022)
 ### Bug Fixes
-* Fix a bug in io_uring_prep_cancel in AbortIO API for posix which expects sqe->addr to match with read request submitted and wrong paramter was being passed.
-* Fixed a regression in iterator performance when the entire DB is a single memtable introduced in #10449. The fix is in #10705 and #10716.
-* Fixed an optimistic transaction validation bug caused by DBImpl::GetLatestSequenceForKey() returning non-latest seq for merge (#10724).
-* Fixed a bug in iterator refresh which could segfault for DeleteRange users (#10739).
-* Fixed a bug causing manual flush with `flush_opts.wait=false` to stall when database has stopped all writes (#10001).
 * Fixed a bug in iterator refresh that was not freeing up SuperVersion, which could cause excessive resource pinniung (#10770).
 * Fixed a bug where RocksDB could be doing compaction endlessly when allow_ingest_behind is true and the bottommost level is not filled (#10767).
-* Fixed a memory safety bug in experimental HyperClockCache (#10768)
-* Fixed some cases where `ldb update_manifest` and `ldb unsafe_remove_sst_file` are not usable because they were requiring the DB files to match the existing manifest state (before updating the manifest to match a desired state).
-
-### Performance Improvements
-* Try to align the compaction output file boundaries to the next level ones, which can reduce more than 10% compaction load for the default level compaction. The feature is enabled by default, to disable, set `AdvancedColumnFamilyOptions.level_compaction_dynamic_file_size` to false. As a side effect, it can create SSTs larger than the target_file_size (capped at 2x target_file_size) or smaller files.
-* Improve RoundRobin TTL compaction, which is going to be the same as normal RoundRobin compaction to move the compaction cursor.
-* Fix a small CPU regression caused by a change that UserComparatorWrapper was made Customizable, because Customizable itself has small CPU overhead for initialization.
 
 ### Behavior Changes
 * Sanitize min_write_buffer_number_to_merge to 1 if atomic flush is enabled to prevent unexpected data loss when WAL is disabled in a multi-column-family setting (#10773).
-* With periodic stat dumper waits up every options.stats_dump_period_sec seconds, it won't dump stats for a CF if it has no change in the period, unless 7 periods have been skipped.
-* Only periodic stats dumper triggered by options.stats_dump_period_sec will update stats interval. Ones triggered by DB::GetProperty() will not update stats interval and will report based on an interval since the last time stats dump period.
 
-### Public API changes
-* Make kXXH3 checksum the new default, because it is faster on common hardware, especially with kCRC32c affected by a performance bug in some versions of clang (https://github.com/facebook/rocksdb/issues/9891). DBs written with this new setting can be read by RocksDB 6.27 and newer.
-* Refactor the classes, APIs and data structures for block cache tracing to allow a user provided trace writer to be used. Introduced an abstract BlockCacheTraceWriter class that takes a structured BlockCacheTraceRecord. The BlockCacheTraceWriter implementation can then format and log the record in whatever way it sees fit. The default BlockCacheTraceWriterImpl does file tracing using a user provided TraceWriter. More details in rocksdb/includb/block_cache_trace_writer.h.
+## 7.7.1 (09/26/2022)
+### Bug Fixes
+* Fixed an optimistic transaction validation bug caused by DBImpl::GetLatestSequenceForKey() returning non-latest seq for merge (#10724).
+* Fixed a bug in iterator refresh which could segfault for DeleteRange users (#10739).
 
 ## 7.7.0 (09/18/2022)
 ### Bug Fixes
@@ -187,6 +49,10 @@
 * Fix a bug in key range overlap checking with concurrent compactions when user-defined timestamp is enabled. User-defined timestamps should be EXCLUDED when checking if two ranges overlap.
 * Fixed a bug where the blob cache prepopulating logic did not consider the secondary cache (see #10603).
 * Fixed the rocksdb.num.sst.read.per.level, rocksdb.num.index.and.filter.blocks.read.per.level and rocksdb.num.level.read.per.multiget stats in the MultiGet coroutines
+* Fix a bug in io_uring_prep_cancel in AbortIO API for posix which expects sqe->addr to match with read request submitted and wrong paramter was being passed.
+* Fixed a regression in iterator performance when the entire DB is a single memtable introduced in #10449. The fix is in #10705 and #10716.
+* Fix a bug in io_uring_prep_cancel in AbortIO API for posix which expects sqe->addr to match with read request submitted and wrong paramter was being passed.
+* Fixed a regression in iterator performance when the entire DB is a single memtable introduced in #10449. The fix is in #10705 and #10716.
 
 ### Public API changes
 * Add `rocksdb_column_family_handle_get_id`, `rocksdb_column_family_handle_get_name` to get name, id of column family in C API

INSTALL.md

@@ -48,11 +48,6 @@ to build a portable binary, add `PORTABLE=1` before your make commands, like thi
 * If you wish to build the RocksJava static target, then cmake is required for building Snappy.
 
 * If you wish to run microbench (e.g, `make microbench`, `make ribbon_bench` or `cmake -DWITH_BENCHMARK=1`), Google benchmark >= 1.6.0 is needed.
-* You can do the following to install Google benchmark. These commands are copied from `./build_tools/ubuntu20_image/Dockerfile`:
-
-`$ git clone --depth 1 --branch v1.7.0 https://github.com/google/benchmark.git ~/benchmark`
-
-`$ cd ~/benchmark && mkdir build && cd build && cmake .. -GNinja -DCMAKE_BUILD_TYPE=Release -DBENCHMARK_ENABLE_GTEST_TESTS=0 && ninja && ninja install`
 
 ## Supported platforms
 
@@ -183,7 +178,7 @@ to build a portable binary, add `PORTABLE=1` before your make commands, like thi
         gmake rocksdbjava
 
 * **iOS**:
-  * Run: `TARGET_OS=IOS make static_lib`. When building the project which uses rocksdb iOS library, make sure to define an important pre-processing macros: `IOS_CROSS_COMPILE`.
+  * Run: `TARGET_OS=IOS make static_lib`. When building the project which uses rocksdb iOS library, make sure to define two important pre-processing macros: `ROCKSDB_LITE` and `IOS_CROSS_COMPILE`.
 
 * **Windows** (Visual Studio 2017 to up):
   * Read and follow the instructions at CMakeLists.txt

Makefile

@@ -44,6 +44,13 @@ quoted_perl_command = $(subst ','\'',$(perl_command))
 # Set the default DEBUG_LEVEL to 1
 DEBUG_LEVEL?=1
 
+# LIB_MODE says whether or not to use/build "shared" or "static" libraries.
+# Mode "static" means to link against static libraries (.a)
+# Mode "shared" means to link against shared libraries (.so, .sl, .dylib, etc)
+#
+# Set the default LIB_MODE to static
+LIB_MODE?=static
+
 # OBJ_DIR is where the object files reside.  Default to the current directory
 OBJ_DIR?=.
 
@@ -74,23 +81,29 @@ else ifneq ($(filter jtest rocksdbjava%, $(MAKECMDGOALS)),)
 	endif
 endif
 
-# LIB_MODE says whether or not to use/build "shared" or "static" libraries.
-# Mode "static" means to link against static libraries (.a)
-# Mode "shared" means to link against shared libraries (.so, .sl, .dylib, etc)
-#
-ifeq ($(DEBUG_LEVEL), 0)
-# For optimized, set the default LIB_MODE to static for code size/efficiency
-	LIB_MODE?=static
-else
-# For debug, set the default LIB_MODE to shared for efficient `make check` etc.
-	LIB_MODE?=shared
-endif
+$(info $$DEBUG_LEVEL is ${DEBUG_LEVEL})
 
-$(info $$DEBUG_LEVEL is $(DEBUG_LEVEL), $$LIB_MODE is $(LIB_MODE))
+# Lite build flag.
+LITE ?= 0
+ifeq ($(LITE), 0)
+ifneq ($(filter -DROCKSDB_LITE,$(OPT)),)
+  # Be backward compatible and support older format where OPT=-DROCKSDB_LITE is
+  # specified instead of LITE=1 on the command line.
+  LITE=1
+endif
+else ifeq ($(LITE), 1)
+ifeq ($(filter -DROCKSDB_LITE,$(OPT)),)
+	OPT += -DROCKSDB_LITE
+endif
+endif
 
 # Figure out optimize level.
 ifneq ($(DEBUG_LEVEL), 2)
+ifeq ($(LITE), 0)
 	OPTIMIZE_LEVEL ?= -O2
+else
+	OPTIMIZE_LEVEL ?= -Os
+endif
 endif
 # `OPTIMIZE_LEVEL` is empty when the user does not set it and `DEBUG_LEVEL=2`.
 # In that case, the compiler default (`-O0` for gcc and clang) will be used.
@@ -253,7 +266,6 @@ ROCKSDB_PLUGIN_EXTERNS = $(foreach p, $(ROCKSDB_PLUGIN_W_FUNCS), int $($(p)_FUNC
 ROCKSDB_PLUGIN_BUILTINS = $(foreach p, $(ROCKSDB_PLUGIN_W_FUNCS), {\"$(p)\"\, $($(p)_FUNC)}\,)
 ROCKSDB_PLUGIN_LDFLAGS = $(foreach plugin, $(ROCKSDB_PLUGINS), $($(plugin)_LDFLAGS))
 ROCKSDB_PLUGIN_PKGCONFIG_REQUIRES = $(foreach plugin, $(ROCKSDB_PLUGINS), $($(plugin)_PKGCONFIG_REQUIRES))
-ROCKSDB_PLUGIN_TESTS = $(foreach p, $(ROCKSDB_PLUGINS), $(foreach test, $($(p)_TESTS), plugin/$(p)/$(test)))
 
 CXXFLAGS += $(foreach plugin, $(ROCKSDB_PLUGINS), $($(plugin)_CXXFLAGS))
 PLATFORM_LDFLAGS += $(ROCKSDB_PLUGIN_LDFLAGS)
@@ -324,6 +336,13 @@ endif
 ifeq ($(PLATFORM), OS_SOLARIS)
 	PLATFORM_CXXFLAGS += -D _GLIBCXX_USE_C99
 endif
+ifneq ($(filter -DROCKSDB_LITE,$(OPT)),)
+	# found
+	CFLAGS += -fno-exceptions
+	CXXFLAGS += -fno-exceptions
+	# LUA is not supported under ROCKSDB_LITE
+	LUA_PATH =
+endif
 
 ifeq ($(LIB_MODE),shared)
 # So that binaries are executable from build location, in addition to install location
@@ -337,7 +356,7 @@ ifneq ($(MACHINE), arm64)
 # linking with jemalloc (as it won't be arm64-compatible) and remove some other options
 # set during platform detection
 DISABLE_JEMALLOC=1
-PLATFORM_CCFLAGS := $(filter-out -march=native -DHAVE_SSE42 -DHAVE_AVX2, $(PLATFORM_CCFLAGS))
+PLATFORM_CFLAGS := $(filter-out -march=native -DHAVE_SSE42 -DHAVE_AVX2, $(PLATFORM_CFLAGS))
 PLATFORM_CXXFLAGS := $(filter-out -march=native -DHAVE_SSE42 -DHAVE_AVX2, $(PLATFORM_CXXFLAGS))
 endif
 endif
@@ -490,7 +509,7 @@ ifneq ($(strip $(FOLLY_PATH)),)
 
 	# Add -ldl at the end as gcc resolves a symbol in a library by searching only in libraries specified later
 	# in the command line
-	PLATFORM_LDFLAGS += $(FOLLY_PATH)/lib/libfolly.a $(BOOST_PATH)/lib/libboost_context.a $(BOOST_PATH)/lib/libboost_filesystem.a $(BOOST_PATH)/lib/libboost_atomic.a $(BOOST_PATH)/lib/libboost_program_options.a $(BOOST_PATH)/lib/libboost_regex.a $(BOOST_PATH)/lib/libboost_system.a $(BOOST_PATH)/lib/libboost_thread.a $(DBL_CONV_PATH)/lib/libdouble-conversion.a $(FMT_LIB_PATH)/libfmt.a $(GLOG_LIB_PATH)/libglog.so $(GFLAGS_PATH)/lib/libgflags.so.2.2 $(LIBEVENT_PATH)/lib/libevent-2.1.so -ldl
+	PLATFORM_LDFLAGS += $(BOOST_PATH)/lib/libboost_context.a $(BOOST_PATH)/lib/libboost_filesystem.a $(BOOST_PATH)/lib/libboost_atomic.a $(BOOST_PATH)/lib/libboost_program_options.a $(BOOST_PATH)/lib/libboost_regex.a $(BOOST_PATH)/lib/libboost_system.a $(BOOST_PATH)/lib/libboost_thread.a $(DBL_CONV_PATH)/lib/libdouble-conversion.a $(FOLLY_PATH)/lib/libfolly.a $(FMT_LIB_PATH)/libfmt.a $(GLOG_LIB_PATH)/libglog.so $(GFLAGS_PATH)/lib/libgflags.so.2.2 -ldl
 	PLATFORM_LDFLAGS += -Wl,-rpath=$(GFLAGS_PATH)/lib -Wl,-rpath=$(GLOG_LIB_PATH) -Wl,-rpath=$(LIBEVENT_PATH)/lib -Wl,-rpath=$(LIBSODIUM_PATH)/lib -Wl,-rpath=$(LIBEVENT_PATH)/lib
 endif
 	PLATFORM_CCFLAGS += -DUSE_FOLLY -DFOLLY_NO_CONFIG
@@ -539,7 +558,7 @@ endif
 
 ifdef USE_CLANG
 	# Used by some teams in Facebook
-	WARNING_FLAGS += -Wshift-sign-overflow -Wambiguous-reversed-operator
+	WARNING_FLAGS += -Wshift-sign-overflow
 endif
 
 ifeq ($(PLATFORM), OS_OPENBSD)
@@ -628,12 +647,10 @@ STRESS_OBJECTS =  $(patsubst %.cc, $(OBJ_DIR)/%.o, $(STRESS_LIB_SOURCES))
 ALL_SOURCES  = $(filter-out util/build_version.cc, $(LIB_SOURCES)) $(TEST_LIB_SOURCES) $(MOCK_LIB_SOURCES) $(GTEST_DIR)/gtest/gtest-all.cc
 ALL_SOURCES += $(TOOL_LIB_SOURCES) $(BENCH_LIB_SOURCES) $(CACHE_BENCH_LIB_SOURCES) $(ANALYZER_LIB_SOURCES) $(STRESS_LIB_SOURCES)
 ALL_SOURCES += $(TEST_MAIN_SOURCES) $(TOOL_MAIN_SOURCES) $(BENCH_MAIN_SOURCES)
-ALL_SOURCES += $(ROCKSDB_PLUGIN_SOURCES) $(ROCKSDB_PLUGIN_TESTS)
+ALL_SOURCES += $(ROCKSDB_PLUGIN_SOURCES)
 
-PLUGIN_TESTS = $(patsubst %.cc, %, $(notdir $(ROCKSDB_PLUGIN_TESTS)))
 TESTS = $(patsubst %.cc, %, $(notdir $(TEST_MAIN_SOURCES)))
 TESTS += $(patsubst %.c, %, $(notdir $(TEST_MAIN_SOURCES_C)))
-TESTS += $(PLUGIN_TESTS)
 
 # `make check-headers` to very that each header file includes its own
 # dependencies
@@ -641,7 +658,7 @@ ifneq ($(filter check-headers, $(MAKECMDGOALS)),)
 # TODO: add/support JNI headers
 	DEV_HEADER_DIRS := $(sort include/ $(dir $(ALL_SOURCES)))
 # Some headers like in port/ are platform-specific
-	DEV_HEADERS := $(shell $(FIND) $(DEV_HEADER_DIRS) -type f -name '*.h' | grep -E -v 'port/|plugin/|lua/|range_tree/')
+	DEV_HEADERS := $(shell $(FIND) $(DEV_HEADER_DIRS) -type f -name '*.h' | egrep -v 'port/|plugin/|lua/|range_tree/')
 else
 	DEV_HEADERS :=
 endif
@@ -685,7 +702,6 @@ NON_PARALLEL_TEST = \
 	env_test \
 	deletefile_test \
 	db_bloom_filter_test \
-	$(PLUGIN_TESTS) \
 
 PARALLEL_TEST = $(filter-out $(NON_PARALLEL_TEST), $(TESTS))
 
@@ -779,9 +795,9 @@ TOOLS_LIBRARY=$(STATIC_TOOLS_LIBRARY)
 endif
 STRESS_LIBRARY=$(STATIC_STRESS_LIBRARY)
 
-ROCKSDB_MAJOR = $(shell grep -E "ROCKSDB_MAJOR.[0-9]" include/rocksdb/version.h | cut -d ' ' -f 3)
-ROCKSDB_MINOR = $(shell grep -E "ROCKSDB_MINOR.[0-9]" include/rocksdb/version.h | cut -d ' ' -f 3)
-ROCKSDB_PATCH = $(shell grep -E "ROCKSDB_PATCH.[0-9]" include/rocksdb/version.h | cut -d ' ' -f 3)
+ROCKSDB_MAJOR = $(shell egrep "ROCKSDB_MAJOR.[0-9]" include/rocksdb/version.h | cut -d ' ' -f 3)
+ROCKSDB_MINOR = $(shell egrep "ROCKSDB_MINOR.[0-9]" include/rocksdb/version.h | cut -d ' ' -f 3)
+ROCKSDB_PATCH = $(shell egrep "ROCKSDB_PATCH.[0-9]" include/rocksdb/version.h | cut -d ' ' -f 3)
 
 # If NO_UPDATE_BUILD_VERSION is set we don't update util/build_version.cc, but
 # the file needs to already exist or else the build will fail
@@ -1062,11 +1078,13 @@ check: all
 	rm -rf $(TEST_TMPDIR)
 ifneq ($(PLATFORM), OS_AIX)
 	$(PYTHON) tools/check_all_python.py
+ifeq ($(filter -DROCKSDB_LITE,$(OPT)),)
 ifndef ASSERT_STATUS_CHECKED # not yet working with these tests
 	$(PYTHON) tools/ldb_test.py
 	sh tools/rocksdb_dump_test.sh
 endif
 endif
+endif
 ifndef SKIP_FORMAT_BUCK_CHECKS
 	$(MAKE) check-format
 	$(MAKE) check-buck-targets
@@ -1222,9 +1240,9 @@ clean: clean-ext-libraries-all clean-rocks clean-rocksjava
 clean-not-downloaded: clean-ext-libraries-bin clean-rocks clean-not-downloaded-rocksjava
 
 clean-rocks:
-# Not practical to exactly match all versions/variants in naming (e.g. debug or not)
-	rm -f ${LIBNAME}*.so* ${LIBNAME}*.a
-	rm -f $(BENCHMARKS) $(TOOLS) $(TESTS) $(PARALLEL_TEST) $(MICROBENCHS)
+	echo shared=$(ALL_SHARED_LIBS)
+	echo static=$(ALL_STATIC_LIBS)
+	rm -f $(BENCHMARKS) $(TOOLS) $(TESTS) $(PARALLEL_TEST) $(ALL_STATIC_LIBS) $(ALL_SHARED_LIBS) $(MICROBENCHS)
 	rm -rf $(CLEAN_FILES) ios-x86 ios-arm scan_build_report
 	$(FIND) . -name "*.[oda]" -exec rm -f {} \;
 	$(FIND) . -type f \( -name "*.gcda" -o -name "*.gcno" \) -exec rm -f {} \;
@@ -1337,14 +1355,6 @@ db_sanity_test: $(OBJ_DIR)/tools/db_sanity_test.o $(LIBRARY)
 db_repl_stress: $(OBJ_DIR)/tools/db_repl_stress.o $(LIBRARY)
 	$(AM_LINK)
 
-define MakeTestRule
-$(notdir $(1:%.cc=%)): $(1:%.cc=$$(OBJ_DIR)/%.o) $$(TEST_LIBRARY) $$(LIBRARY)
-	$$(AM_LINK)
-endef
-
-# For each PLUGIN test, create a rule to generate the test executable
-$(foreach test, $(ROCKSDB_PLUGIN_TESTS), $(eval $(call MakeTestRule, $(test))))
-
 arena_test: $(OBJ_DIR)/memory/arena_test.o $(TEST_LIBRARY) $(LIBRARY)
 	$(AM_LINK)
 
@@ -1822,7 +1832,7 @@ memtable_list_test: $(OBJ_DIR)/db/memtable_list_test.o $(TEST_LIBRARY) $(LIBRARY
 write_callback_test: $(OBJ_DIR)/db/write_callback_test.o $(TEST_LIBRARY) $(LIBRARY)
 	$(AM_LINK)
 
-heap_test: $(OBJ_DIR)/util/heap_test.o $(TEST_LIBRARY) $(LIBRARY)
+heap_test: $(OBJ_DIR)/util/heap_test.o $(GTEST)
 	$(AM_LINK)
 
 point_lock_manager_test: utilities/transactions/lock/point/point_lock_manager_test.o $(TEST_LIBRARY) $(LIBRARY)
@@ -2048,7 +2058,7 @@ JAVA_INCLUDE = -I$(JAVA_HOME)/include/ -I$(JAVA_HOME)/include/linux
 ifeq ($(PLATFORM), OS_SOLARIS)
 	ARCH := $(shell isainfo -b)
 else ifeq ($(PLATFORM), OS_OPENBSD)
-	ifneq (,$(filter amd64 ppc64 ppc64le s390x arm64 aarch64 sparc64 loongarch64, $(MACHINE)))
+	ifneq (,$(filter amd64 ppc64 ppc64le s390x arm64 aarch64 sparc64, $(MACHINE)))
 		ARCH := 64
 	else
 		ARCH := 32
@@ -2069,7 +2079,7 @@ ifneq ($(origin JNI_LIBC), undefined)
 endif
 
 ifeq (,$(ROCKSDBJNILIB))
-ifneq (,$(filter ppc% s390x arm64 aarch64 sparc64 loongarch64, $(MACHINE)))
+ifneq (,$(filter ppc% s390x arm64 aarch64 sparc64, $(MACHINE)))
 	ROCKSDBJNILIB = librocksdbjni-linux-$(MACHINE)$(JNI_LIBC_POSTFIX).so
 else
 	ROCKSDBJNILIB = librocksdbjni-linux$(ARCH)$(JNI_LIBC_POSTFIX).so
@@ -2082,8 +2092,8 @@ ROCKSDB_JAVADOCS_JAR = rocksdbjni-$(ROCKSDB_JAVA_VERSION)-javadoc.jar
 ROCKSDB_SOURCES_JAR = rocksdbjni-$(ROCKSDB_JAVA_VERSION)-sources.jar
 SHA256_CMD = sha256sum
 
-ZLIB_VER ?= 1.2.13
-ZLIB_SHA256 ?= b3a24de97a8fdbc835b9833169501030b8977031bcb54b3b3ac13740f846ab30
+ZLIB_VER ?= 1.2.12
+ZLIB_SHA256 ?= 91844808532e5ce316b3c010929493c0244f3d37593afd6de04f71821d5136d9
 ZLIB_DOWNLOAD_BASE ?= http://zlib.net
 BZIP2_VER ?= 1.0.8
 BZIP2_SHA256 ?= ab5a03176ee106d3f0fa90e381da478ddae405918153cca248e682cd0c4a2269
@@ -2467,6 +2477,18 @@ build_size:
 	$(REPORT_BUILD_STATISTIC) rocksdb.build_size.shared_lib $$(stat --printf="%s" `readlink -f librocksdb.so`)
 	strip `readlink -f librocksdb.so`
 	$(REPORT_BUILD_STATISTIC) rocksdb.build_size.shared_lib_stripped $$(stat --printf="%s" `readlink -f librocksdb.so`)
+	# === lite build, static ===
+	$(MAKE) clean
+	$(MAKE) LITE=1 static_lib
+	$(REPORT_BUILD_STATISTIC) rocksdb.build_size.static_lib_lite $$(stat --printf="%s" librocksdb.a)
+	strip librocksdb.a
+	$(REPORT_BUILD_STATISTIC) rocksdb.build_size.static_lib_lite_stripped $$(stat --printf="%s" librocksdb.a)
+	# === lite build, shared ===
+	$(MAKE) clean
+	$(MAKE) LITE=1 shared_lib
+	$(REPORT_BUILD_STATISTIC) rocksdb.build_size.shared_lib_lite $$(stat --printf="%s" `readlink -f librocksdb.so`)
+	strip `readlink -f librocksdb.so`
+	$(REPORT_BUILD_STATISTIC) rocksdb.build_size.shared_lib_lite_stripped $$(stat --printf="%s" `readlink -f librocksdb.so`)
 
 # ---------------------------------------------------------------------------
 #  	Platform-specific compilation

README.md

@@ -1,6 +1,8 @@
 ## RocksDB: A Persistent Key-Value Store for Flash and RAM Storage
 
 [![CircleCI Status](https://circleci.com/gh/facebook/rocksdb.svg?style=svg)](https://circleci.com/gh/facebook/rocksdb)
+[![Appveyor Build status](https://ci.appveyor.com/api/projects/status/fbgfu0so3afcno78/branch/main?svg=true)](https://ci.appveyor.com/project/Facebook/rocksdb/branch/main)
+[![PPC64le Build Status](http://140-211-168-68-openstack.osuosl.org:8080/buildStatus/icon?job=rocksdb&style=plastic)](http://140-211-168-68-openstack.osuosl.org:8080/job/rocksdb)
 
 RocksDB is developed and maintained by Facebook Database Engineering Team.
 It is built on earlier work on [LevelDB](https://github.com/google/leveldb) by Sanjay Ghemawat (sanjay@google.com)

ROCKSDB_LITE.md

@@ -0,0 +1,21 @@
+# RocksDBLite
+
+RocksDBLite is a project focused on mobile use cases, which don't need a lot of fancy things we've built for server workloads and they are very sensitive to binary size. For that reason, we added a compile flag ROCKSDB_LITE that comments out a lot of the nonessential code and keeps the binary lean.
+
+Some examples of the features disabled by ROCKSDB_LITE:
+* compiled-in support for LDB tool
+* No backup engine
+* No support for replication (which we provide in form of TransactionalIterator)
+* No advanced monitoring tools
+* No special-purpose memtables that are highly optimized for specific use cases
+* No Transactions
+
+When adding a new big feature to RocksDB, please add ROCKSDB_LITE compile guard if:
+* Nobody from mobile really needs your feature,
+* Your feature is adding a lot of weight to the binary.
+
+Don't add ROCKSDB_LITE compile guard if:
+* It would introduce a lot of code complexity. Compile guards make code harder to read. It's a trade-off.
+* Your feature is not adding a lot of weight.
+
+If unsure, ask. :)

TARGETS

@@ -11,15 +11,13 @@ load("//rocks/buckifier:defs.bzl", "cpp_library_wrapper","rocks_cpp_library_wrap
 cpp_library_wrapper(name="rocksdb_lib", srcs=[
         "cache/cache.cc",
         "cache/cache_entry_roles.cc",
-        "cache/cache_helpers.cc",
         "cache/cache_key.cc",
         "cache/cache_reservation_manager.cc",
         "cache/charged_cache.cc",
         "cache/clock_cache.cc",
         "cache/compressed_secondary_cache.cc",
+        "cache/fast_lru_cache.cc",
         "cache/lru_cache.cc",
-        "cache/secondary_cache.cc",
-        "cache/secondary_cache_adapter.cc",
         "cache/sharded_cache.cc",
         "db/arena_wrapped_db_iter.cc",
         "db/blob/blob_contents.cc",
@@ -106,7 +104,6 @@ cpp_library_wrapper(name="rocksdb_lib", srcs=[
         "db/write_batch.cc",
         "db/write_batch_base.cc",
         "db/write_controller.cc",
-        "db/write_stall_stats.cc",
         "db/write_thread.cc",
         "env/composite_env.cc",
         "env/env.cc",
@@ -166,7 +163,6 @@ cpp_library_wrapper(name="rocksdb_lib", srcs=[
         "options/options.cc",
         "options/options_helper.cc",
         "options/options_parser.cc",
-        "port/mmap.cc",
         "port/port_posix.cc",
         "port/stack_trace.cc",
         "port/win/env_default.cc",
@@ -183,7 +179,6 @@ cpp_library_wrapper(name="rocksdb_lib", srcs=[
         "table/block_based/block_based_table_iterator.cc",
         "table/block_based/block_based_table_reader.cc",
         "table/block_based/block_builder.cc",
-        "table/block_based/block_cache.cc",
         "table/block_based/block_prefetcher.cc",
         "table/block_based/block_prefix_index.cc",
         "table/block_based/data_block_footer.cc",
@@ -202,7 +197,6 @@ cpp_library_wrapper(name="rocksdb_lib", srcs=[
         "table/block_based/reader_common.cc",
         "table/block_based/uncompression_dict_reader.cc",
         "table/block_fetcher.cc",
-        "table/compaction_merging_iterator.cc",
         "table/cuckoo/cuckoo_table_builder.cc",
         "table/cuckoo/cuckoo_table_factory.cc",
         "table/cuckoo/cuckoo_table_reader.cc",
@@ -250,7 +244,6 @@ cpp_library_wrapper(name="rocksdb_lib", srcs=[
         "util/concurrent_task_limiter_impl.cc",
         "util/crc32c.cc",
         "util/crc32c_arm64.cc",
-        "util/data_structure.cc",
         "util/dynamic_bloom.cc",
         "util/file_checksum_helper.cc",
         "util/hash.cc",
@@ -354,14 +347,350 @@ cpp_library_wrapper(name="rocksdb_lib", srcs=[
         "//folly/synchronization:distributed_mutex",
     ], headers=None, link_whole=False, extra_test_libs=False)
 
-cpp_library_wrapper(name="rocksdb_whole_archive_lib", srcs=[], deps=[":rocksdb_lib"], headers=None, link_whole=True, extra_test_libs=False)
+cpp_library_wrapper(name="rocksdb_whole_archive_lib", srcs=[
+        "cache/cache.cc",
+        "cache/cache_entry_roles.cc",
+        "cache/cache_key.cc",
+        "cache/cache_reservation_manager.cc",
+        "cache/charged_cache.cc",
+        "cache/clock_cache.cc",
+        "cache/compressed_secondary_cache.cc",
+        "cache/fast_lru_cache.cc",
+        "cache/lru_cache.cc",
+        "cache/sharded_cache.cc",
+        "db/arena_wrapped_db_iter.cc",
+        "db/blob/blob_contents.cc",
+        "db/blob/blob_fetcher.cc",
+        "db/blob/blob_file_addition.cc",
+        "db/blob/blob_file_builder.cc",
+        "db/blob/blob_file_cache.cc",
+        "db/blob/blob_file_garbage.cc",
+        "db/blob/blob_file_meta.cc",
+        "db/blob/blob_file_reader.cc",
+        "db/blob/blob_garbage_meter.cc",
+        "db/blob/blob_log_format.cc",
+        "db/blob/blob_log_sequential_reader.cc",
+        "db/blob/blob_log_writer.cc",
+        "db/blob/blob_source.cc",
+        "db/blob/prefetch_buffer_collection.cc",
+        "db/builder.cc",
+        "db/c.cc",
+        "db/column_family.cc",
+        "db/compaction/compaction.cc",
+        "db/compaction/compaction_iterator.cc",
+        "db/compaction/compaction_job.cc",
+        "db/compaction/compaction_outputs.cc",
+        "db/compaction/compaction_picker.cc",
+        "db/compaction/compaction_picker_fifo.cc",
+        "db/compaction/compaction_picker_level.cc",
+        "db/compaction/compaction_picker_universal.cc",
+        "db/compaction/compaction_service_job.cc",
+        "db/compaction/compaction_state.cc",
+        "db/compaction/sst_partitioner.cc",
+        "db/compaction/subcompaction_state.cc",
+        "db/convenience.cc",
+        "db/db_filesnapshot.cc",
+        "db/db_impl/compacted_db_impl.cc",
+        "db/db_impl/db_impl.cc",
+        "db/db_impl/db_impl_compaction_flush.cc",
+        "db/db_impl/db_impl_debug.cc",
+        "db/db_impl/db_impl_experimental.cc",
+        "db/db_impl/db_impl_files.cc",
+        "db/db_impl/db_impl_open.cc",
+        "db/db_impl/db_impl_readonly.cc",
+        "db/db_impl/db_impl_secondary.cc",
+        "db/db_impl/db_impl_write.cc",
+        "db/db_info_dumper.cc",
+        "db/db_iter.cc",
+        "db/dbformat.cc",
+        "db/error_handler.cc",
+        "db/event_helpers.cc",
+        "db/experimental.cc",
+        "db/external_sst_file_ingestion_job.cc",
+        "db/file_indexer.cc",
+        "db/flush_job.cc",
+        "db/flush_scheduler.cc",
+        "db/forward_iterator.cc",
+        "db/import_column_family_job.cc",
+        "db/internal_stats.cc",
+        "db/log_reader.cc",
+        "db/log_writer.cc",
+        "db/logs_with_prep_tracker.cc",
+        "db/malloc_stats.cc",
+        "db/memtable.cc",
+        "db/memtable_list.cc",
+        "db/merge_helper.cc",
+        "db/merge_operator.cc",
+        "db/output_validator.cc",
+        "db/periodic_task_scheduler.cc",
+        "db/range_del_aggregator.cc",
+        "db/range_tombstone_fragmenter.cc",
+        "db/repair.cc",
+        "db/seqno_to_time_mapping.cc",
+        "db/snapshot_impl.cc",
+        "db/table_cache.cc",
+        "db/table_properties_collector.cc",
+        "db/transaction_log_impl.cc",
+        "db/trim_history_scheduler.cc",
+        "db/version_builder.cc",
+        "db/version_edit.cc",
+        "db/version_edit_handler.cc",
+        "db/version_set.cc",
+        "db/wal_edit.cc",
+        "db/wal_manager.cc",
+        "db/wide/wide_column_serialization.cc",
+        "db/wide/wide_columns.cc",
+        "db/write_batch.cc",
+        "db/write_batch_base.cc",
+        "db/write_controller.cc",
+        "db/write_thread.cc",
+        "env/composite_env.cc",
+        "env/env.cc",
+        "env/env_chroot.cc",
+        "env/env_encryption.cc",
+        "env/env_posix.cc",
+        "env/file_system.cc",
+        "env/file_system_tracer.cc",
+        "env/fs_posix.cc",
+        "env/fs_remap.cc",
+        "env/io_posix.cc",
+        "env/mock_env.cc",
+        "env/unique_id_gen.cc",
+        "file/delete_scheduler.cc",
+        "file/file_prefetch_buffer.cc",
+        "file/file_util.cc",
+        "file/filename.cc",
+        "file/line_file_reader.cc",
+        "file/random_access_file_reader.cc",
+        "file/read_write_util.cc",
+        "file/readahead_raf.cc",
+        "file/sequence_file_reader.cc",
+        "file/sst_file_manager_impl.cc",
+        "file/writable_file_writer.cc",
+        "logging/auto_roll_logger.cc",
+        "logging/event_logger.cc",
+        "logging/log_buffer.cc",
+        "memory/arena.cc",
+        "memory/concurrent_arena.cc",
+        "memory/jemalloc_nodump_allocator.cc",
+        "memory/memkind_kmem_allocator.cc",
+        "memory/memory_allocator.cc",
+        "memtable/alloc_tracker.cc",
+        "memtable/hash_linklist_rep.cc",
+        "memtable/hash_skiplist_rep.cc",
+        "memtable/skiplistrep.cc",
+        "memtable/vectorrep.cc",
+        "memtable/write_buffer_manager.cc",
+        "monitoring/histogram.cc",
+        "monitoring/histogram_windowing.cc",
+        "monitoring/in_memory_stats_history.cc",
+        "monitoring/instrumented_mutex.cc",
+        "monitoring/iostats_context.cc",
+        "monitoring/perf_context.cc",
+        "monitoring/perf_level.cc",
+        "monitoring/persistent_stats_history.cc",
+        "monitoring/statistics.cc",
+        "monitoring/thread_status_impl.cc",
+        "monitoring/thread_status_updater.cc",
+        "monitoring/thread_status_updater_debug.cc",
+        "monitoring/thread_status_util.cc",
+        "monitoring/thread_status_util_debug.cc",
+        "options/cf_options.cc",
+        "options/configurable.cc",
+        "options/customizable.cc",
+        "options/db_options.cc",
+        "options/options.cc",
+        "options/options_helper.cc",
+        "options/options_parser.cc",
+        "port/port_posix.cc",
+        "port/stack_trace.cc",
+        "port/win/env_default.cc",
+        "port/win/env_win.cc",
+        "port/win/io_win.cc",
+        "port/win/port_win.cc",
+        "port/win/win_logger.cc",
+        "port/win/win_thread.cc",
+        "table/adaptive/adaptive_table_factory.cc",
+        "table/block_based/binary_search_index_reader.cc",
+        "table/block_based/block.cc",
+        "table/block_based/block_based_table_builder.cc",
+        "table/block_based/block_based_table_factory.cc",
+        "table/block_based/block_based_table_iterator.cc",
+        "table/block_based/block_based_table_reader.cc",
+        "table/block_based/block_builder.cc",
+        "table/block_based/block_prefetcher.cc",
+        "table/block_based/block_prefix_index.cc",
+        "table/block_based/data_block_footer.cc",
+        "table/block_based/data_block_hash_index.cc",
+        "table/block_based/filter_block_reader_common.cc",
+        "table/block_based/filter_policy.cc",
+        "table/block_based/flush_block_policy.cc",
+        "table/block_based/full_filter_block.cc",
+        "table/block_based/hash_index_reader.cc",
+        "table/block_based/index_builder.cc",
+        "table/block_based/index_reader_common.cc",
+        "table/block_based/parsed_full_filter_block.cc",
+        "table/block_based/partitioned_filter_block.cc",
+        "table/block_based/partitioned_index_iterator.cc",
+        "table/block_based/partitioned_index_reader.cc",
+        "table/block_based/reader_common.cc",
+        "table/block_based/uncompression_dict_reader.cc",
+        "table/block_fetcher.cc",
+        "table/cuckoo/cuckoo_table_builder.cc",
+        "table/cuckoo/cuckoo_table_factory.cc",
+        "table/cuckoo/cuckoo_table_reader.cc",
+        "table/format.cc",
+        "table/get_context.cc",
+        "table/iterator.cc",
+        "table/merging_iterator.cc",
+        "table/meta_blocks.cc",
+        "table/persistent_cache_helper.cc",
+        "table/plain/plain_table_bloom.cc",
+        "table/plain/plain_table_builder.cc",
+        "table/plain/plain_table_factory.cc",
+        "table/plain/plain_table_index.cc",
+        "table/plain/plain_table_key_coding.cc",
+        "table/plain/plain_table_reader.cc",
+        "table/sst_file_dumper.cc",
+        "table/sst_file_reader.cc",
+        "table/sst_file_writer.cc",
+        "table/table_factory.cc",
+        "table/table_properties.cc",
+        "table/two_level_iterator.cc",
+        "table/unique_id.cc",
+        "test_util/sync_point.cc",
+        "test_util/sync_point_impl.cc",
+        "test_util/transaction_test_util.cc",
+        "tools/dump/db_dump_tool.cc",
+        "tools/io_tracer_parser_tool.cc",
+        "tools/ldb_cmd.cc",
+        "tools/ldb_tool.cc",
+        "tools/sst_dump_tool.cc",
+        "trace_replay/block_cache_tracer.cc",
+        "trace_replay/io_tracer.cc",
+        "trace_replay/trace_record.cc",
+        "trace_replay/trace_record_handler.cc",
+        "trace_replay/trace_record_result.cc",
+        "trace_replay/trace_replay.cc",
+        "util/async_file_reader.cc",
+        "util/build_version.cc",
+        "util/cleanable.cc",
+        "util/coding.cc",
+        "util/compaction_job_stats_impl.cc",
+        "util/comparator.cc",
+        "util/compression.cc",
+        "util/compression_context_cache.cc",
+        "util/concurrent_task_limiter_impl.cc",
+        "util/crc32c.cc",
+        "util/crc32c_arm64.cc",
+        "util/dynamic_bloom.cc",
+        "util/file_checksum_helper.cc",
+        "util/hash.cc",
+        "util/murmurhash.cc",
+        "util/random.cc",
+        "util/rate_limiter.cc",
+        "util/ribbon_config.cc",
+        "util/slice.cc",
+        "util/status.cc",
+        "util/stderr_logger.cc",
+        "util/string_util.cc",
+        "util/thread_local.cc",
+        "util/threadpool_imp.cc",
+        "util/xxhash.cc",
+        "utilities/agg_merge/agg_merge.cc",
+        "utilities/backup/backup_engine.cc",
+        "utilities/blob_db/blob_compaction_filter.cc",
+        "utilities/blob_db/blob_db.cc",
+        "utilities/blob_db/blob_db_impl.cc",
+        "utilities/blob_db/blob_db_impl_filesnapshot.cc",
+        "utilities/blob_db/blob_dump_tool.cc",
+        "utilities/blob_db/blob_file.cc",
+        "utilities/cache_dump_load.cc",
+        "utilities/cache_dump_load_impl.cc",
+        "utilities/cassandra/cassandra_compaction_filter.cc",
+        "utilities/cassandra/format.cc",
+        "utilities/cassandra/merge_operator.cc",
+        "utilities/checkpoint/checkpoint_impl.cc",
+        "utilities/compaction_filters.cc",
+        "utilities/compaction_filters/remove_emptyvalue_compactionfilter.cc",
+        "utilities/convenience/info_log_finder.cc",
+        "utilities/counted_fs.cc",
+        "utilities/debug.cc",
+        "utilities/env_mirror.cc",
+        "utilities/env_timed.cc",
+        "utilities/fault_injection_env.cc",
+        "utilities/fault_injection_fs.cc",
+        "utilities/fault_injection_secondary_cache.cc",
+        "utilities/leveldb_options/leveldb_options.cc",
+        "utilities/memory/memory_util.cc",
+        "utilities/merge_operators.cc",
+        "utilities/merge_operators/bytesxor.cc",
+        "utilities/merge_operators/max.cc",
+        "utilities/merge_operators/put.cc",
+        "utilities/merge_operators/sortlist.cc",
+        "utilities/merge_operators/string_append/stringappend.cc",
+        "utilities/merge_operators/string_append/stringappend2.cc",
+        "utilities/merge_operators/uint64add.cc",
+        "utilities/object_registry.cc",
+        "utilities/option_change_migration/option_change_migration.cc",
+        "utilities/options/options_util.cc",
+        "utilities/persistent_cache/block_cache_tier.cc",
+        "utilities/persistent_cache/block_cache_tier_file.cc",
+        "utilities/persistent_cache/block_cache_tier_metadata.cc",
+        "utilities/persistent_cache/persistent_cache_tier.cc",
+        "utilities/persistent_cache/volatile_tier_impl.cc",
+        "utilities/simulator_cache/cache_simulator.cc",
+        "utilities/simulator_cache/sim_cache.cc",
+        "utilities/table_properties_collectors/compact_on_deletion_collector.cc",
+        "utilities/trace/file_trace_reader_writer.cc",
+        "utilities/trace/replayer_impl.cc",
+        "utilities/transactions/lock/lock_manager.cc",
+        "utilities/transactions/lock/point/point_lock_manager.cc",
+        "utilities/transactions/lock/point/point_lock_tracker.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/concurrent_tree.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/keyrange.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/lock_request.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/locktree.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/manager.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/range_buffer.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/treenode.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/txnid_set.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/wfg.cc",
+        "utilities/transactions/lock/range/range_tree/lib/standalone_port.cc",
+        "utilities/transactions/lock/range/range_tree/lib/util/dbt.cc",
+        "utilities/transactions/lock/range/range_tree/lib/util/memarena.cc",
+        "utilities/transactions/lock/range/range_tree/range_tree_lock_manager.cc",
+        "utilities/transactions/lock/range/range_tree/range_tree_lock_tracker.cc",
+        "utilities/transactions/optimistic_transaction.cc",
+        "utilities/transactions/optimistic_transaction_db_impl.cc",
+        "utilities/transactions/pessimistic_transaction.cc",
+        "utilities/transactions/pessimistic_transaction_db.cc",
+        "utilities/transactions/snapshot_checker.cc",
+        "utilities/transactions/transaction_base.cc",
+        "utilities/transactions/transaction_db_mutex_impl.cc",
+        "utilities/transactions/transaction_util.cc",
+        "utilities/transactions/write_prepared_txn.cc",
+        "utilities/transactions/write_prepared_txn_db.cc",
+        "utilities/transactions/write_unprepared_txn.cc",
+        "utilities/transactions/write_unprepared_txn_db.cc",
+        "utilities/ttl/db_ttl_impl.cc",
+        "utilities/wal_filter.cc",
+        "utilities/write_batch_with_index/write_batch_with_index.cc",
+        "utilities/write_batch_with_index/write_batch_with_index_internal.cc",
+    ], deps=[
+        "//folly/container:f14_hash",
+        "//folly/experimental/coro:blocking_wait",
+        "//folly/experimental/coro:collect",
+        "//folly/experimental/coro:coroutine",
+        "//folly/experimental/coro:task",
+        "//folly/synchronization:distributed_mutex",
+    ], headers=None, link_whole=True, extra_test_libs=False)
 
 cpp_library_wrapper(name="rocksdb_test_lib", srcs=[
         "db/db_test_util.cc",
         "db/db_with_timestamp_test_util.cc",
         "table/mock_table.cc",
         "test_util/mock_time_env.cc",
-        "test_util/secondary_cache_test_util.cc",
         "test_util/testharness.cc",
         "test_util/testutil.cc",
         "tools/block_cache_analyzer/block_cache_trace_analyzer.cc",

USERS.md

@@ -15,9 +15,6 @@ At Facebook, we use RocksDB as storage engines in multiple data management servi
 
 [2] https://code.facebook.com/posts/357056558062811/logdevice-a-distributed-data-store-for-logs/
 
-## Microsoft
-The Bing search engine from Microsoft uses RocksDB as the storage engine for its web data platform: https://blogs.bing.com/Engineering-Blog/october-2021/RocksDB-in-Microsoft-Bing
-
 ## LinkedIn
 Two different use cases at Linkedin are using RocksDB as a storage engine:
 
@@ -29,9 +26,6 @@ Learn more about those use cases in a Tech Talk by Ankit Gupta and Naveen Somasu
 ## Yahoo
 Yahoo is using RocksDB as a storage engine for their biggest distributed data store Sherpa. Learn more about it here: http://yahooeng.tumblr.com/post/120730204806/sherpa-scales-new-heights
 
-## Tencent
-[PaxosStore](https://github.com/Tencent/paxosstore) is a distributed database supporting WeChat. It uses RocksDB as its storage engine.
-
 ## Baidu
 [Apache Doris](http://doris.apache.org/master/en/) is a MPP analytical database engine released by Baidu. It [uses RocksDB](http://doris.apache.org/master/en/administrator-guide/operation/tablet-meta-tool.html) to manage its tablet's metadata.
 
@@ -85,9 +79,6 @@ quasardb uses a heavily tuned RocksDB as its persistence layer.
 ## TiKV
 [TiKV](https://github.com/pingcap/tikv) is a GEO-replicated, high-performance, distributed, transactional key-value database. TiKV is powered by Rust and Raft. TiKV uses RocksDB as its persistence layer.
 
-## Apache Spark
-[Spark Structured Streaming](https://docs.databricks.com/structured-streaming/rocksdb-state-store.html) uses RocksDB as the local state store.
-
 ## Apache Flink
 [Apache Flink](https://flink.apache.org/news/2016/03/08/release-1.0.0.html) uses RocksDB to store state locally on a machine.
 

buckifier/buckify_rocksdb.py

@@ -1,18 +1,19 @@
 #!/usr/bin/env python3
 # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved.
-from __future__ import absolute_import, division, print_function, unicode_literals
-
+from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
+from __future__ import unicode_literals
 try:
     from builtins import str
 except ImportError:
     from __builtin__ import str
-import fnmatch
+from targets_builder import TARGETSBuilder
 import json
 import os
+import fnmatch
 import sys
 
-from targets_builder import TARGETSBuilder
-
 from util import ColorString
 
 # This script generates TARGETS file for Buck.
@@ -26,7 +27,7 @@ from util import ColorString
 # $python3 buckifier/buckify_rocksdb.py \
 #        '{"fake": {
 #                      "extra_deps": [":test_dep", "//fakes/module:mock1"],
-#                      "extra_compiler_flags": ["-DFOO_BAR", "-Os"]
+#                      "extra_compiler_flags": ["-DROCKSDB_LITE", "-Os"]
 #                  }
 #         }'
 # (Generated TARGETS file has test_dep and mock1 as dependencies for RocksDB
@@ -43,13 +44,13 @@ def parse_src_mk(repo_path):
     src_files = {}
     for line in open(src_mk):
         line = line.strip()
-        if len(line) == 0 or line[0] == "#":
+        if len(line) == 0 or line[0] == '#':
             continue
-        if "=" in line:
-            current_src = line.split("=")[0].strip()
+        if '=' in line:
+            current_src = line.split('=')[0].strip()
             src_files[current_src] = []
-        elif ".c" in line:
-            src_path = line.split("\\")[0].strip()
+        elif '.c' in line:
+            src_path = line.split('\\')[0].strip()
             src_files[current_src].append(src_path)
     return src_files
 
@@ -57,16 +58,14 @@ def parse_src_mk(repo_path):
 # get all .cc / .c files
 def get_cc_files(repo_path):
     cc_files = []
-    for root, _dirnames, filenames in os.walk(
-        repo_path
-    ):  # noqa: B007 T25377293 Grandfathered in
-        root = root[(len(repo_path) + 1) :]
+    for root, dirnames, filenames in os.walk(repo_path):  # noqa: B007 T25377293 Grandfathered in
+        root = root[(len(repo_path) + 1):]
         if "java" in root:
             # Skip java
             continue
-        for filename in fnmatch.filter(filenames, "*.cc"):
+        for filename in fnmatch.filter(filenames, '*.cc'):
             cc_files.append(os.path.join(root, filename))
-        for filename in fnmatch.filter(filenames, "*.c"):
+        for filename in fnmatch.filter(filenames, '*.c'):
             cc_files.append(os.path.join(root, filename))
     return cc_files
 
@@ -94,10 +93,14 @@ def get_non_parallel_tests(repo_path):
 
     return s
 
-
 # Parse extra dependencies passed by user from command line
 def get_dependencies():
-    deps_map = {"": {"extra_deps": [], "extra_compiler_flags": []}}
+    deps_map = {
+        '': {
+            'extra_deps': [],
+            'extra_compiler_flags': []
+        }
+    }
     if len(sys.argv) < 2:
         return deps_map
 
@@ -108,7 +111,6 @@ def get_dependencies():
                 v = encode_dict(v)
             rv[k] = v
         return rv
-
     extra_deps = json.loads(sys.argv[1], object_hook=encode_dict)
     for target_alias, deps in extra_deps.items():
         deps_map[target_alias] = deps
@@ -141,66 +143,73 @@ def generate_targets(repo_path, deps_map):
         "rocksdb_lib",
         src_mk["LIB_SOURCES"] +
         # always add range_tree, it's only excluded on ppc64, which we don't use internally
-        src_mk["RANGE_TREE_SOURCES"] + src_mk["TOOL_LIB_SOURCES"],
+        src_mk["RANGE_TREE_SOURCES"] +
+        src_mk["TOOL_LIB_SOURCES"],
         deps=[
             "//folly/container:f14_hash",
             "//folly/experimental/coro:blocking_wait",
             "//folly/experimental/coro:collect",
             "//folly/experimental/coro:coroutine",
             "//folly/experimental/coro:task",
-            "//folly/synchronization:distributed_mutex",
-        ],
-    )
+            "//folly/synchronization:distributed_mutex"])
     # rocksdb_whole_archive_lib
     TARGETS.add_library(
         "rocksdb_whole_archive_lib",
-        [],
+        src_mk["LIB_SOURCES"] +
+        # always add range_tree, it's only excluded on ppc64, which we don't use internally
+        src_mk["RANGE_TREE_SOURCES"] +
+        src_mk["TOOL_LIB_SOURCES"],
         deps=[
-            ":rocksdb_lib",
-        ],
+            "//folly/container:f14_hash",
+            "//folly/experimental/coro:blocking_wait",
+            "//folly/experimental/coro:collect",
+            "//folly/experimental/coro:coroutine",
+            "//folly/experimental/coro:task",
+            "//folly/synchronization:distributed_mutex"],
         headers=None,
         extra_external_deps="",
-        link_whole=True,
-    )
+        link_whole=True)
     # rocksdb_test_lib
     TARGETS.add_library(
         "rocksdb_test_lib",
-        src_mk.get("MOCK_LIB_SOURCES", [])
-        + src_mk.get("TEST_LIB_SOURCES", [])
-        + src_mk.get("EXP_LIB_SOURCES", [])
-        + src_mk.get("ANALYZER_LIB_SOURCES", []),
+        src_mk.get("MOCK_LIB_SOURCES", []) +
+        src_mk.get("TEST_LIB_SOURCES", []) +
+        src_mk.get("EXP_LIB_SOURCES", []) +
+        src_mk.get("ANALYZER_LIB_SOURCES", []),
         [":rocksdb_lib"],
-        extra_test_libs=True,
-    )
+        extra_test_libs=True
+        )
     # rocksdb_tools_lib
     TARGETS.add_library(
         "rocksdb_tools_lib",
-        src_mk.get("BENCH_LIB_SOURCES", [])
-        + src_mk.get("ANALYZER_LIB_SOURCES", [])
-        + ["test_util/testutil.cc"],
-        [":rocksdb_lib"],
-    )
+        src_mk.get("BENCH_LIB_SOURCES", []) +
+        src_mk.get("ANALYZER_LIB_SOURCES", []) +
+        ["test_util/testutil.cc"],
+        [":rocksdb_lib"])
     # rocksdb_cache_bench_tools_lib
     TARGETS.add_library(
         "rocksdb_cache_bench_tools_lib",
         src_mk.get("CACHE_BENCH_LIB_SOURCES", []),
-        [":rocksdb_lib"],
-    )
+        [":rocksdb_lib"])
     # rocksdb_stress_lib
     TARGETS.add_rocksdb_library(
         "rocksdb_stress_lib",
         src_mk.get("ANALYZER_LIB_SOURCES", [])
-        + src_mk.get("STRESS_LIB_SOURCES", [])
-        + ["test_util/testutil.cc"],
-    )
+        + src_mk.get('STRESS_LIB_SOURCES', [])
+        + ["test_util/testutil.cc"])
     # db_stress binary
-    TARGETS.add_binary(
-        "db_stress", ["db_stress_tool/db_stress.cc"], [":rocksdb_stress_lib"]
-    )
+    TARGETS.add_binary("db_stress",
+                       ["db_stress_tool/db_stress.cc"],
+                       [":rocksdb_stress_lib"])
     # bench binaries
     for src in src_mk.get("MICROBENCH_SOURCES", []):
-        name = src.rsplit("/", 1)[1].split(".")[0] if "/" in src else src.split(".")[0]
-        TARGETS.add_binary(name, [src], [], extra_bench_libs=True)
+        name =  src.rsplit('/',1)[1].split('.')[0] if '/' in src else src.split('.')[0]
+        TARGETS.add_binary(
+            name,
+            [src],
+            [],
+            extra_bench_libs=True
+            )
     print("Extra dependencies:\n{0}".format(json.dumps(deps_map)))
 
     # Dictionary test executable name -> relative source file path
@@ -210,7 +219,7 @@ def generate_targets(repo_path, deps_map):
     # are more than one .c test file, we need to extend
     # TARGETS.add_c_test() to include other C tests too.
     for test_src in src_mk.get("TEST_MAIN_SOURCES_C", []):
-        if test_src != "db/c_test.c":
+        if test_src != 'db/c_test.c':
             print("Don't know how to deal with " + test_src)
             return False
     TARGETS.add_c_test()
@@ -220,7 +229,7 @@ def generate_targets(repo_path, deps_map):
             fast_fancy_bench_config_list = json.load(json_file)
             for config_dict in fast_fancy_bench_config_list:
                 clean_benchmarks = {}
-                benchmarks = config_dict["benchmarks"]
+                benchmarks = config_dict['benchmarks']
                 for binary, benchmark_dict in benchmarks.items():
                     clean_benchmarks[binary] = {}
                     for benchmark, overloaded_metric_list in benchmark_dict.items():
@@ -228,20 +237,13 @@ def generate_targets(repo_path, deps_map):
                         for metric in overloaded_metric_list:
                             if not isinstance(metric, dict):
                                 clean_benchmarks[binary][benchmark].append(metric)
-                TARGETS.add_fancy_bench_config(
-                    config_dict["name"],
-                    clean_benchmarks,
-                    False,
-                    config_dict["expected_runtime_one_iter"],
-                    config_dict["sl_iterations"],
-                    config_dict["regression_threshold"],
-                )
+                TARGETS.add_fancy_bench_config(config_dict['name'], clean_benchmarks, False, config_dict['expected_runtime_one_iter'], config_dict['sl_iterations'], config_dict['regression_threshold'])
 
         with open(f"{repo_path}/buckifier/bench-slow.json") as json_file:
             slow_fancy_bench_config_list = json.load(json_file)
             for config_dict in slow_fancy_bench_config_list:
                 clean_benchmarks = {}
-                benchmarks = config_dict["benchmarks"]
+                benchmarks = config_dict['benchmarks']
                 for binary, benchmark_dict in benchmarks.items():
                     clean_benchmarks[binary] = {}
                     for benchmark, overloaded_metric_list in benchmark_dict.items():
@@ -250,14 +252,7 @@ def generate_targets(repo_path, deps_map):
                             if not isinstance(metric, dict):
                                 clean_benchmarks[binary][benchmark].append(metric)
             for config_dict in slow_fancy_bench_config_list:
-                TARGETS.add_fancy_bench_config(
-                    config_dict["name"] + "_slow",
-                    clean_benchmarks,
-                    True,
-                    config_dict["expected_runtime_one_iter"],
-                    config_dict["sl_iterations"],
-                    config_dict["regression_threshold"],
-                )
+                TARGETS.add_fancy_bench_config(config_dict['name']+"_slow", clean_benchmarks, True, config_dict['expected_runtime_one_iter'], config_dict['sl_iterations'], config_dict['regression_threshold'])
     # it is better servicelab experiments break
     # than rocksdb github ci
     except Exception:
@@ -266,7 +261,7 @@ def generate_targets(repo_path, deps_map):
     TARGETS.add_test_header()
 
     for test_src in src_mk.get("TEST_MAIN_SOURCES", []):
-        test = test_src.split(".c")[0].strip().split("/")[-1].strip()
+        test = test_src.split('.c')[0].strip().split('/')[-1].strip()
         test_source_map[test] = test_src
         print("" + test + " " + test_src)
 
@@ -276,29 +271,23 @@ def generate_targets(repo_path, deps_map):
                 print(ColorString.warning("Failed to get test name for %s" % test_src))
                 continue
 
-            test_target_name = test if not target_alias else test + "_" + target_alias
+            test_target_name = \
+                test if not target_alias else test + "_" + target_alias
 
             if test in _EXPORTED_TEST_LIBS:
                 test_library = "%s_lib" % test_target_name
-                TARGETS.add_library(
-                    test_library,
-                    [test_src],
-                    deps=[":rocksdb_test_lib"],
-                    extra_test_libs=True,
-                )
+                TARGETS.add_library(test_library, [test_src], deps=[":rocksdb_test_lib"], extra_test_libs=True)
                 TARGETS.register_test(
                     test_target_name,
                     test_src,
-                    deps=json.dumps(deps["extra_deps"] + [":" + test_library]),
-                    extra_compiler_flags=json.dumps(deps["extra_compiler_flags"]),
-                )
+                    deps = json.dumps(deps['extra_deps'] + [':'+test_library]),
+                    extra_compiler_flags = json.dumps(deps['extra_compiler_flags']))
             else:
                 TARGETS.register_test(
                     test_target_name,
                     test_src,
-                    deps=json.dumps(deps["extra_deps"] + [":rocksdb_test_lib"]),
-                    extra_compiler_flags=json.dumps(deps["extra_compiler_flags"]),
-                )
+                    deps = json.dumps(deps['extra_deps'] + [":rocksdb_test_lib"] ),
+                    extra_compiler_flags = json.dumps(deps['extra_compiler_flags']))
 
     print(ColorString.info("Generated TARGETS Summary:"))
     print(ColorString.info("- %d libs" % TARGETS.total_lib))
@@ -311,7 +300,8 @@ def get_rocksdb_path():
     # rocksdb = {script_dir}/..
     script_dir = os.path.dirname(sys.argv[0])
     script_dir = os.path.abspath(script_dir)
-    rocksdb_path = os.path.abspath(os.path.join(script_dir, "../"))
+    rocksdb_path = os.path.abspath(
+        os.path.join(script_dir, "../"))
 
     return rocksdb_path
 
@@ -328,6 +318,5 @@ def main():
     if not ok:
         exit_with_error("Failed to generate TARGETS files")
 
-
 if __name__ == "__main__":
     main()

buckifier/targets_builder.py

@@ -1,150 +1,113 @@
 # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved.
-from __future__ import absolute_import, division, print_function, unicode_literals
-
+from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
+from __future__ import unicode_literals
 try:
-    from builtins import object, str
+    from builtins import object
+    from builtins import str
 except ImportError:
-    from __builtin__ import object, str
-import pprint
-
+    from __builtin__ import object
+    from __builtin__ import str
 import targets_cfg
-
+import pprint
 
 def pretty_list(lst, indent=8):
     if lst is None or len(lst) == 0:
         return ""
 
     if len(lst) == 1:
-        return '"%s"' % lst[0]
+        return "\"%s\"" % lst[0]
 
-    separator = '",\n%s"' % (" " * indent)
+    separator = "\",\n%s\"" % (" " * indent)
     res = separator.join(sorted(lst))
-    res = "\n" + (" " * indent) + '"' + res + '",\n' + (" " * (indent - 4))
+    res = "\n" + (" " * indent) + "\"" + res + "\",\n" + (" " * (indent - 4))
     return res
 
 
 class TARGETSBuilder(object):
     def __init__(self, path, extra_argv):
         self.path = path
+        self.targets_file = open(path, 'wb')
         header = targets_cfg.rocksdb_target_header_template.format(
-            extra_argv=extra_argv
-        )
-        with open(path, "wb") as targets_file:
-            targets_file.write(header.encode("utf-8"))
+            extra_argv=extra_argv)
+        self.targets_file.write(header.encode("utf-8"))
         self.total_lib = 0
         self.total_bin = 0
         self.total_test = 0
         self.tests_cfg = ""
 
-    def add_library(
-        self,
-        name,
-        srcs,
-        deps=None,
-        headers=None,
-        extra_external_deps="",
-        link_whole=False,
-        external_dependencies=None,
-        extra_test_libs=False,
-    ):
+    def __del__(self):
+        self.targets_file.close()
+
+    def add_library(self, name, srcs, deps=None, headers=None,
+                    extra_external_deps="", link_whole=False,
+                     external_dependencies=None, extra_test_libs=False):
         if headers is not None:
             headers = "[" + pretty_list(headers) + "]"
-        with open(self.path, "ab") as targets_file:
-            targets_file.write(
-                targets_cfg.library_template.format(
-                    name=name,
-                    srcs=pretty_list(srcs),
-                    headers=headers,
-                    deps=pretty_list(deps),
-                    extra_external_deps=extra_external_deps,
-                    link_whole=link_whole,
-                    external_dependencies=pretty_list(external_dependencies),
-                    extra_test_libs=extra_test_libs,
-                ).encode("utf-8")
+        self.targets_file.write(targets_cfg.library_template.format(
+            name=name,
+            srcs=pretty_list(srcs),
+            headers=headers,
+            deps=pretty_list(deps),
+            extra_external_deps=extra_external_deps,
+            link_whole=link_whole,
+            external_dependencies=pretty_list(external_dependencies),
+            extra_test_libs=extra_test_libs
+            ).encode("utf-8"))
+        self.total_lib = self.total_lib + 1
+
+    def add_rocksdb_library(self, name, srcs, headers=None,
+                            external_dependencies=None):
+        if headers is not None:
+            headers = "[" + pretty_list(headers) + "]"
+        self.targets_file.write(targets_cfg.rocksdb_library_template.format(
+            name=name,
+            srcs=pretty_list(srcs),
+            headers=headers,
+            external_dependencies=pretty_list(external_dependencies)
+            ).encode("utf-8")
             )
         self.total_lib = self.total_lib + 1
 
-    def add_rocksdb_library(self, name, srcs, headers=None, external_dependencies=None):
-        if headers is not None:
-            headers = "[" + pretty_list(headers) + "]"
-        with open(self.path, "ab") as targets_file:
-            targets_file.write(
-                targets_cfg.rocksdb_library_template.format(
-                    name=name,
-                    srcs=pretty_list(srcs),
-                    headers=headers,
-                    external_dependencies=pretty_list(external_dependencies),
-                ).encode("utf-8")
-            )
-        self.total_lib = self.total_lib + 1
-
-    def add_binary(
-        self,
-        name,
-        srcs,
-        deps=None,
-        extra_preprocessor_flags=None,
-        extra_bench_libs=False,
-    ):
-        with open(self.path, "ab") as targets_file:
-            targets_file.write(
-                targets_cfg.binary_template.format(
-                    name=name,
-                    srcs=pretty_list(srcs),
-                    deps=pretty_list(deps),
-                    extra_preprocessor_flags=pretty_list(extra_preprocessor_flags),
-                    extra_bench_libs=extra_bench_libs,
-                ).encode("utf-8")
-            )
+    def add_binary(self, name, srcs, deps=None, extra_preprocessor_flags=None,extra_bench_libs=False):
+        self.targets_file.write(targets_cfg.binary_template.format(
+            name=name,
+            srcs=pretty_list(srcs),
+            deps=pretty_list(deps),
+            extra_preprocessor_flags=pretty_list(extra_preprocessor_flags),
+            extra_bench_libs=extra_bench_libs,
+            ).encode("utf-8"))
         self.total_bin = self.total_bin + 1
 
     def add_c_test(self):
-        with open(self.path, "ab") as targets_file:
-            targets_file.write(
-                b"""
+        self.targets_file.write(b"""
 add_c_test_wrapper()
-"""
-            )
+""")
 
     def add_test_header(self):
-        with open(self.path, "ab") as targets_file:
-            targets_file.write(
-                b"""
+        self.targets_file.write(b"""
         # Generate a test rule for each entry in ROCKS_TESTS
         # Do not build the tests in opt mode, since SyncPoint and other test code
         # will not be included.
-"""
-            )
+""")
 
-    def add_fancy_bench_config(
-        self,
-        name,
-        bench_config,
-        slow,
-        expected_runtime,
-        sl_iterations,
-        regression_threshold,
-    ):
-        with open(self.path, "ab") as targets_file:
-            targets_file.write(
-                targets_cfg.fancy_bench_template.format(
+    def add_fancy_bench_config(self, name, bench_config, slow, expected_runtime, sl_iterations, regression_threshold):
+        self.targets_file.write(targets_cfg.fancy_bench_template.format(
                     name=name,
                     bench_config=pprint.pformat(bench_config),
                     slow=slow,
                     expected_runtime=expected_runtime,
                     sl_iterations=sl_iterations,
-                    regression_threshold=regression_threshold,
-                ).encode("utf-8")
-            )
+                    regression_threshold=regression_threshold
+                    ).encode("utf-8"))
 
-    def register_test(self, test_name, src, deps, extra_compiler_flags):
-        with open(self.path, "ab") as targets_file:
-            targets_file.write(
-                targets_cfg.unittests_template.format(
-                    test_name=test_name,
-                    test_cc=str(src),
-                    deps=deps,
-                    extra_compiler_flags=extra_compiler_flags,
-                ).encode("utf-8")
-            )
+    def register_test(self,
+                      test_name,
+                      src,
+                      deps,
+                      extra_compiler_flags):
+
+        self.targets_file.write(targets_cfg.unittests_template.format(test_name=test_name,test_cc=str(src),deps=deps,
+            extra_compiler_flags=extra_compiler_flags).encode("utf-8"))
         self.total_test = self.total_test + 1

buckifier/targets_cfg.py

@@ -1,7 +1,11 @@
 # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved.
-from __future__ import absolute_import, division, print_function, unicode_literals
+from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
+from __future__ import unicode_literals
 
-rocksdb_target_header_template = """# This file \100generated by:
+rocksdb_target_header_template = \
+    """# This file \100generated by:
 #$ python3 buckifier/buckify_rocksdb.py{extra_argv}
 # --> DO NOT EDIT MANUALLY <--
 # This file is a Facebook-specific integration for buck builds, so can
@@ -23,6 +27,7 @@ rocks_cpp_library_wrapper(name="{name}", srcs=[{srcs}], headers={headers})
 """
 
 
+
 binary_template = """
 cpp_binary_wrapper(name="{name}", srcs=[{srcs}], deps=[{deps}], extra_preprocessor_flags=[{extra_preprocessor_flags}], extra_bench_libs={extra_bench_libs})
 """

buckifier/util.py

@@ -2,35 +2,37 @@
 """
 This module keeps commonly used components.
 """
-from __future__ import absolute_import, division, print_function, unicode_literals
-
+from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
+from __future__ import unicode_literals
 try:
     from builtins import object
 except ImportError:
     from __builtin__ import object
-import os
 import subprocess
 import sys
+import os
 import time
 
-
 class ColorString(object):
-    """Generate colorful strings on terminal"""
-
-    HEADER = "\033[95m"
-    BLUE = "\033[94m"
-    GREEN = "\033[92m"
-    WARNING = "\033[93m"
-    FAIL = "\033[91m"
-    ENDC = "\033[0m"
+    """ Generate colorful strings on terminal """
+    HEADER = '\033[95m'
+    BLUE = '\033[94m'
+    GREEN = '\033[92m'
+    WARNING = '\033[93m'
+    FAIL = '\033[91m'
+    ENDC = '\033[0m'
 
     @staticmethod
     def _make_color_str(text, color):
         # In Python2, default encoding for unicode string is ASCII
         if sys.version_info.major <= 2:
-            return "".join([color, text.encode("utf-8"), ColorString.ENDC])
+            return "".join(
+                [color, text.encode('utf-8'), ColorString.ENDC])
         # From Python3, default encoding for unicode string is UTF-8
-        return "".join([color, text, ColorString.ENDC])
+        return "".join(
+            [color, text, ColorString.ENDC])
 
     @staticmethod
     def ok(text):
@@ -66,38 +68,37 @@ class ColorString(object):
 
 
 def run_shell_command(shell_cmd, cmd_dir=None):
-    """Run a single shell command.
-    @returns a tuple of shell command return code, stdout, stderr"""
+    """ Run a single shell command.
+        @returns a tuple of shell command return code, stdout, stderr """
 
     if cmd_dir is not None and not os.path.exists(cmd_dir):
         run_shell_command("mkdir -p %s" % cmd_dir)
 
     start = time.time()
     print("\t>>> Running: " + shell_cmd)
-    p = subprocess.Popen(  # noqa
-        shell_cmd,
-        shell=True,
-        stdout=subprocess.PIPE,
-        stderr=subprocess.PIPE,
-        cwd=cmd_dir,
-    )
+    p = subprocess.Popen(shell_cmd,
+                         shell=True,
+                         stdout=subprocess.PIPE,
+                         stderr=subprocess.PIPE,
+                         cwd=cmd_dir)
     stdout, stderr = p.communicate()
     end = time.time()
 
     # Report time if we spent more than 5 minutes executing a command
     execution_time = end - start
     if execution_time > (60 * 5):
-        mins = execution_time / 60
-        secs = execution_time % 60
+        mins = (execution_time / 60)
+        secs = (execution_time % 60)
         print("\t>time spent: %d minutes %d seconds" % (mins, secs))
 
+
     return p.returncode, stdout, stderr
 
 
 def run_shell_commands(shell_cmds, cmd_dir=None, verbose=False):
-    """Execute a sequence of shell commands, which is equivalent to
-    running `cmd1 && cmd2 && cmd3`
-    @returns boolean indication if all commands succeeds.
+    """ Execute a sequence of shell commands, which is equivalent to
+        running `cmd1 && cmd2 && cmd3`
+        @returns boolean indication if all commands succeeds.
     """
 
     if cmd_dir:

build_tools/amalgamate.py

@@ -28,15 +28,14 @@
 from __future__ import print_function
 
 import argparse
+from os import path
 import re
 import sys
-from os import path
 
 include_re = re.compile('^[ \t]*#include[ \t]+"(.*)"[ \t]*$')
 included = set()
 excluded = set()
 
-
 def find_header(name, abs_path, include_paths):
     samedir = path.join(path.dirname(abs_path), name)
     if path.exists(samedir):
@@ -47,31 +46,17 @@ def find_header(name, abs_path, include_paths):
             return include_path
     return None
 
-
-def expand_include(
-    include_path,
-    f,
-    abs_path,
-    source_out,
-    header_out,
-    include_paths,
-    public_include_paths,
-):
+def expand_include(include_path, f, abs_path, source_out, header_out, include_paths, public_include_paths):
     if include_path in included:
         return False
 
     included.add(include_path)
     with open(include_path) as f:
         print('#line 1 "{}"'.format(include_path), file=source_out)
-        process_file(
-            f, include_path, source_out, header_out, include_paths, public_include_paths
-        )
+        process_file(f, include_path, source_out, header_out, include_paths, public_include_paths)
     return True
 
-
-def process_file(
-    f, abs_path, source_out, header_out, include_paths, public_include_paths
-):
+def process_file(f, abs_path, source_out, header_out, include_paths, public_include_paths):
     for (line, text) in enumerate(f):
         m = include_re.match(text)
         if m:
@@ -83,15 +68,7 @@ def process_file(
                     source_out.write(text)
                     expanded = False
                 else:
-                    expanded = expand_include(
-                        include_path,
-                        f,
-                        abs_path,
-                        source_out,
-                        header_out,
-                        include_paths,
-                        public_include_paths,
-                    )
+                    expanded = expand_include(include_path, f, abs_path, source_out, header_out, include_paths, public_include_paths)
             else:
                 # now try public headers
                 include_path = find_header(filename, abs_path, public_include_paths)
@@ -101,52 +78,23 @@ def process_file(
                     if include_path in excluded:
                         source_out.write(text)
                     else:
-                        expand_include(
-                            include_path,
-                            f,
-                            abs_path,
-                            header_out,
-                            None,
-                            public_include_paths,
-                            [],
-                        )
+                        expand_include(include_path, f, abs_path, header_out, None, public_include_paths, [])
                 else:
-                    sys.exit(
-                        "unable to find {}, included in {} on line {}".format(
-                            filename, abs_path, line
-                        )
-                    )
+                    sys.exit("unable to find {}, included in {} on line {}".format(filename, abs_path, line))
 
             if expanded:
-                print('#line {} "{}"'.format(line + 1, abs_path), file=source_out)
+                print('#line {} "{}"'.format(line+1, abs_path), file=source_out)
         elif text != "#pragma once\n":
             source_out.write(text)
 
-
 def main():
-    parser = argparse.ArgumentParser(
-        description="Transform a unity build into an amalgamation"
-    )
+    parser = argparse.ArgumentParser(description="Transform a unity build into an amalgamation")
     parser.add_argument("source", help="source file")
-    parser.add_argument(
-        "-I",
-        action="append",
-        dest="include_paths",
-        help="include paths for private headers",
-    )
-    parser.add_argument(
-        "-i",
-        action="append",
-        dest="public_include_paths",
-        help="include paths for public headers",
-    )
-    parser.add_argument(
-        "-x", action="append", dest="excluded", help="excluded header files"
-    )
+    parser.add_argument("-I", action="append", dest="include_paths", help="include paths for private headers")
+    parser.add_argument("-i", action="append", dest="public_include_paths", help="include paths for public headers")
+    parser.add_argument("-x", action="append", dest="excluded", help="excluded header files")
     parser.add_argument("-o", dest="source_out", help="output C++ file", required=True)
-    parser.add_argument(
-        "-H", dest="header_out", help="output C++ header file", required=True
-    )
+    parser.add_argument("-H", dest="header_out", help="output C++ header file", required=True)
     args = parser.parse_args()
 
     include_paths = list(map(path.abspath, args.include_paths or []))
@@ -154,15 +102,10 @@ def main():
     excluded.update(map(path.abspath, args.excluded or []))
     filename = args.source
     abs_path = path.abspath(filename)
-    with open(filename) as f, open(args.source_out, "w") as source_out, open(
-        args.header_out, "w"
-    ) as header_out:
+    with open(filename) as f, open(args.source_out, 'w') as source_out, open(args.header_out, 'w') as header_out:
         print('#line 1 "{}"'.format(filename), file=source_out)
         print('#include "{}"'.format(header_out.name), file=source_out)
-        process_file(
-            f, abs_path, source_out, header_out, include_paths, public_include_paths
-        )
-
+        process_file(f, abs_path, source_out, header_out, include_paths, public_include_paths)
 
 if __name__ == "__main__":
     main()

build_tools/benchmark_log_tool.py

@@ -4,27 +4,23 @@
 #  COPYING file in the root directory) and Apache 2.0 License
 #  (found in the LICENSE.Apache file in the root directory).
 
-"""Access the results of benchmark runs
+'''Access the results of benchmark runs
 Send these results on to OpenSearch graphing service
-"""
+'''
 
 import argparse
 import itertools
-import logging
 import os
 import re
 import sys
-
 import requests
 from dateutil import parser
+import logging
 
 logging.basicConfig(level=logging.DEBUG)
-
-
 class Configuration:
-    opensearch_user = os.environ["ES_USER"]
-    opensearch_pass = os.environ["ES_PASS"]
-
+    opensearch_user = os.environ['ES_USER']
+    opensearch_pass = os.environ['ES_PASS']
 
 class BenchmarkResultException(Exception):
     def __init__(self, message, content):
@@ -34,71 +30,45 @@ class BenchmarkResultException(Exception):
 
 class BenchmarkUtils:
 
-    expected_keys = [
-        "ops_sec",
-        "mb_sec",
-        "lsm_sz",
-        "blob_sz",
-        "c_wgb",
-        "w_amp",
-        "c_mbps",
-        "c_wsecs",
-        "c_csecs",
-        "b_rgb",
-        "b_wgb",
-        "usec_op",
-        "p50",
-        "p99",
-        "p99.9",
-        "p99.99",
-        "pmax",
-        "uptime",
-        "stall%",
-        "Nstall",
-        "u_cpu",
-        "s_cpu",
-        "rss",
-        "test",
-        "date",
-        "version",
-        "job_id",
-    ]
+    expected_keys = ['ops_sec', 'mb_sec',  'lsm_sz',  'blob_sz', 'c_wgb',   'w_amp',
+    'c_mbps',  'c_wsecs', 'c_csecs', 'b_rgb',   'b_wgb',   'usec_op',
+    'p50',     'p99',     'p99.9',   'p99.99', 'pmax',
+    'uptime',  'stall%', 'Nstall',  'u_cpu',   's_cpu',   'rss', 'test', 'date', 'version', 'job_id']
 
     def sanity_check(row):
-        if "test" not in row:
+        if not 'test' in row:
             logging.debug(f"not 'test' in row: {row}")
             return False
-        if row["test"] == "":
+        if row['test'] == '':
             logging.debug(f"row['test'] == '': {row}")
             return False
-        if "date" not in row:
+        if not 'date' in row:
             logging.debug(f"not 'date' in row: {row}")
             return False
-        if "ops_sec" not in row:
+        if not 'ops_sec' in row:
             logging.debug(f"not 'ops_sec' in row: {row}")
             return False
         try:
-            _ = int(row["ops_sec"])
+            v = int(row['ops_sec'])
         except (ValueError, TypeError):
             logging.debug(f"int(row['ops_sec']): {row}")
             return False
         try:
-            (_, _) = parser.parse(row["date"], fuzzy_with_tokens=True)
+            (_, _) = parser.parse(row['date'], fuzzy_with_tokens=True)
         except (parser.ParserError):
-            logging.error(
-                f"parser.parse((row['date']): not a valid format for date in row: {row}"
-            )
+            logging.error(f"parser.parse((row['date']): not a valid format for date in row: {row}")
             return False
         return True
 
     def conform_opensearch(row):
-        (dt, _) = parser.parse(row["date"], fuzzy_with_tokens=True)
+        (dt, _) = parser.parse(row['date'], fuzzy_with_tokens=True)
         # create a test_date field, which was previously what was expected
         # repair the date field, which has what can be a WRONG ISO FORMAT, (no leading 0 on single-digit day-of-month)
         # e.g. 2022-07-1T00:14:55 should be 2022-07-01T00:14:55
-        row["test_date"] = dt.isoformat()
-        row["date"] = dt.isoformat()
-        return {key.replace(".", "_"): value for key, value in row.items()}
+        row['test_date'] = dt.isoformat()
+        row['date'] = dt.isoformat()
+        return dict((key.replace('.', '_'), value)
+                    for (key, value) in row.items())
 
 
 class ResultParser:
@@ -110,24 +80,24 @@ class ResultParser:
     def ignore(self, l_in: str):
         if len(l_in) == 0:
             return True
-        if l_in[0:1] == "#":
+        if l_in[0:1] == '#':
             return True
         return False
 
-    def line(self, line_in: str):
-        """Parse a line into items
+    def line(self, l_in: str):
+        '''Parse a line into items
         Being clever about separators
-        """
-        line = line_in
+        '''
+        l = l_in
         row = []
-        while line != "":
-            match_item = self.field.match(line)
+        while l != '':
+            match_item = self.field.match(l)
             if match_item:
                 item = match_item.group(0)
                 row.append(item)
-                line = line[len(item) :]
+                l = l[len(item):]
             else:
-                match_intra = self.intra.match(line)
+                match_intra = self.intra.match(l)
                 if match_intra:
                     intra = match_intra.group(0)
                     # Count the separators
@@ -137,29 +107,26 @@ class ResultParser:
                     sep_count = len(tabbed) - 1
                     if sep_count == 0:
                         sep_count = 1
-                    for _ in range(sep_count - 1):
-                        row.append("")
-                    line = line[len(intra) :]
+                    for i in range(sep_count-1):
+                        row.append('')
+                    l = l[len(intra):]
                 else:
                     raise BenchmarkResultException(
-                        "Invalid TSV line", f"{line_in} at {line}"
-                    )
+                        'Invalid TSV line', f"{l_in} at {l}")
         return row
 
     def parse(self, lines):
-        """Parse something that iterates lines"""
+        '''Parse something that iterates lines'''
         rows = [self.line(line) for line in lines if not self.ignore(line)]
         header = rows[0]
         width = len(header)
-        records = [
-            {k: v for (k, v) in itertools.zip_longest(header, row[:width])}
-            for row in rows[1:]
-        ]
+        records = [{k: v for (k, v) in itertools.zip_longest(
+            header, row[:width])} for row in rows[1:]]
         return records
 
 
 def load_report_from_tsv(filename: str):
-    file = open(filename, "r")
+    file = open(filename, 'r')
     contents = file.readlines()
     file.close()
     parser = ResultParser()
@@ -169,70 +136,52 @@ def load_report_from_tsv(filename: str):
 
 
 def push_report_to_opensearch(report, esdocument):
-    sanitized = [
-        BenchmarkUtils.conform_opensearch(row)
-        for row in report
-        if BenchmarkUtils.sanity_check(row)
-    ]
-    logging.debug(
-        f"upload {len(sanitized)} sane of {len(report)} benchmarks to opensearch"
-    )
+    sanitized = [BenchmarkUtils.conform_opensearch(row)
+                 for row in report if BenchmarkUtils.sanity_check(row)]
+    logging.debug(f"upload {len(sanitized)} sane of {len(report)} benchmarks to opensearch")
     for single_benchmark in sanitized:
         logging.debug(f"upload benchmark: {single_benchmark}")
         response = requests.post(
             esdocument,
-            json=single_benchmark,
-            auth=(os.environ["ES_USER"], os.environ["ES_PASS"]),
-        )
+            json=single_benchmark, auth=(os.environ['ES_USER'], os.environ['ES_PASS']))
         logging.debug(
-            f"Sent to OpenSearch, status: {response.status_code}, result: {response.text}"
-        )
+            f"Sent to OpenSearch, status: {response.status_code}, result: {response.text}")
         response.raise_for_status()
 
-
 def push_report_to_null(report):
-
+    
     for row in report:
         if BenchmarkUtils.sanity_check(row):
             logging.debug(f"row {row}")
             conformed = BenchmarkUtils.conform_opensearch(row)
             logging.debug(f"conformed row {conformed}")
 
-
 def main():
-    """Tool for fetching, parsing and uploading benchmark results to OpenSearch / ElasticSearch
+    '''Tool for fetching, parsing and uploading benchmark results to OpenSearch / ElasticSearch
     This tool will
 
     (1) Open a local tsv benchmark report file
     (2) Upload to OpenSearch document, via https/JSON
-    """
+    '''
 
-    parser = argparse.ArgumentParser(description="CircleCI benchmark scraper.")
+    parser = argparse.ArgumentParser(
+        description='CircleCI benchmark scraper.')
 
     # --tsvfile is the name of the file to read results from
     # --esdocument is the ElasticSearch document to push these results into
     #
-    parser.add_argument(
-        "--tsvfile",
-        default="build_tools/circle_api_scraper_input.txt",
-        help="File from which to read tsv report",
-    )
-    parser.add_argument(
-        "--esdocument",
-        help="ElasticSearch/OpenSearch document URL to upload report into",
-    )
-    parser.add_argument(
-        "--upload", choices=["opensearch", "none"], default="opensearch"
-    )
+    parser.add_argument('--tsvfile', default='build_tools/circle_api_scraper_input.txt',
+                        help='File from which to read tsv report')
+    parser.add_argument('--esdocument', help='ElasticSearch/OpenSearch document URL to upload report into')
+    parser.add_argument('--upload', choices=['opensearch', 'none'], default='opensearch')
 
     args = parser.parse_args()
     logging.debug(f"Arguments: {args}")
     reports = load_report_from_tsv(args.tsvfile)
-    if args.upload == "opensearch":
+    if (args.upload == 'opensearch'):
         push_report_to_opensearch(reports, args.esdocument)
     else:
         push_report_to_null(reports)
 
-
-if __name__ == "__main__":
+if __name__ == '__main__':
     sys.exit(main())

build_tools/build_detect_platform

@@ -63,7 +63,13 @@ if [ -z "$ROCKSDB_NO_FBCODE" -a -d /mnt/gvfs/third-party ]; then
     if [ "$LIB_MODE" == "shared" ]; then
       PIC_BUILD=1
     fi
-    source "$PWD/build_tools/fbcode_config_platform010.sh"
+    if [ -n "$ROCKSDB_FBCODE_BUILD_WITH_PLATFORM010" ]; then
+      source "$PWD/build_tools/fbcode_config_platform010.sh"
+    elif [ -n "$ROCKSDB_FBCODE_BUILD_WITH_PLATFORM009" ]; then
+      source "$PWD/build_tools/fbcode_config_platform009.sh"
+    else
+      source "$PWD/build_tools/fbcode_config_platform009.sh"
+    fi
 fi
 
 # Delete existing output, if it exists
@@ -148,7 +154,7 @@ case "$TARGET_OS" in
         ;;
     IOS)
         PLATFORM=IOS
-        COMMON_FLAGS="$COMMON_FLAGS -DOS_MACOSX -DIOS_CROSS_COMPILE "
+        COMMON_FLAGS="$COMMON_FLAGS -DOS_MACOSX -DIOS_CROSS_COMPILE -DROCKSDB_LITE"
         PLATFORM_SHARED_EXT=dylib
         PLATFORM_SHARED_LDFLAGS="-dynamiclib -install_name "
         CROSS_COMPILE=true

build_tools/dependencies_platform009.sh

@@ -0,0 +1,22 @@
+# Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved.
+GCC_BASE=/mnt/gvfs/third-party2/gcc/1795efe5f06778c15a92c8f9a2aba5dc496d9d4d/9.x/centos7-native/3bed279
+CLANG_BASE=/mnt/gvfs/third-party2/llvm-fb/7318eaac22659b6ff2fe43918e4b69fd0772a8a7/9.0.0/platform009/651ee30
+LIBGCC_BASE=/mnt/gvfs/third-party2/libgcc/4959b39cfbe5965a37c861c4c327fa7c5c759b87/9.x/platform009/9202ce7
+GLIBC_BASE=/mnt/gvfs/third-party2/glibc/45ce3375cdc77ecb2520bbf8f0ecddd3f98efd7a/2.30/platform009/f259413
+SNAPPY_BASE=/mnt/gvfs/third-party2/snappy/be4de3205e029101b18aa8103daa696c2bef3b19/1.1.3/platform009/7f3b187
+ZLIB_BASE=/mnt/gvfs/third-party2/zlib/3c160ac5c67e257501e24c6c1d00ad5e01d73db6/1.2.8/platform009/7f3b187
+BZIP2_BASE=/mnt/gvfs/third-party2/bzip2/73a237ac5bc0a5f5d67b39b8d253cfebaab88684/1.0.6/platform009/7f3b187
+LZ4_BASE=/mnt/gvfs/third-party2/lz4/6ca38d3c390be2774d61a300f151464bbd632d62/1.9.1/platform009/7f3b187
+ZSTD_BASE=/mnt/gvfs/third-party2/zstd/64c58a207d2495e83abc57a500a956df09b79a7c/1.4.x/platform009/ba86d1f
+GFLAGS_BASE=/mnt/gvfs/third-party2/gflags/824d0a8a5abb5b121afd1b35fc3896407ea50092/2.2.0/platform009/7f3b187
+JEMALLOC_BASE=/mnt/gvfs/third-party2/jemalloc/b62912d333ef33f9760efa6219dbe3fe6abb3b0e/master/platform009/c305944
+NUMA_BASE=/mnt/gvfs/third-party2/numa/0af65f71e23a67bf65dc91b11f95caa39325c432/2.0.11/platform009/7f3b187
+LIBUNWIND_BASE=/mnt/gvfs/third-party2/libunwind/02486dac347645d31dce116f44e1de3177315be2/1.4/platform009/5191652
+TBB_BASE=/mnt/gvfs/third-party2/tbb/2e0ec671e550bfca347300bf3f789d9c0fff24ad/2018_U5/platform009/7f3b187
+LIBURING_BASE=/mnt/gvfs/third-party2/liburing/70dbd9cfee63a25611417d09433a86d7711b3990/20200729/platform009/7f3b187
+KERNEL_HEADERS_BASE=/mnt/gvfs/third-party2/kernel-headers/32b8a2407b634df3f8f948ba373fc4acc6a18296/fb/platform009/da39a3e
+BINUTILS_BASE=/mnt/gvfs/third-party2/binutils/08634589372fa5f237bfd374e8c644a8364e78c1/2.32/platform009/ba86d1f/
+VALGRIND_BASE=/mnt/gvfs/third-party2/valgrind/6ae525939ad02e5e676855082fbbc7828dbafeac/3.15.0/platform009/7f3b187
+LUA_BASE=/mnt/gvfs/third-party2/lua/162efd9561a3d21f6869f4814011e9cf1b3ff4dc/5.3.4/platform009/a6271c4
+BENCHMARK_BASE=/mnt/gvfs/third-party2/benchmark/30bf49ad6414325e17f3425b0edcb64239427ae3/1.6.1/platform009/7f3b187
+GLOG_BASE=/mnt/gvfs/third-party2/glog/32d751bd5673375b438158717ab6a57c1cc57e3d/0.3.2_fb/platform009/10a364d

build_tools/error_filter.py

@@ -3,13 +3,16 @@
 #  COPYING file in the root directory) and Apache 2.0 License
 #  (found in the LICENSE.Apache file in the root directory).
 
-"""Filter for error messages in test output:
+'''Filter for error messages in test output:
     - Receives merged stdout/stderr from test on stdin
     - Finds patterns of known error messages for test name (first argument)
     - Prints those error messages to stdout
-"""
+'''
 
-from __future__ import absolute_import, division, print_function, unicode_literals
+from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
+from __future__ import unicode_literals
 
 import re
 import sys
@@ -17,24 +20,23 @@ import sys
 
 class ErrorParserBase(object):
     def parse_error(self, line):
-        """Parses a line of test output. If it contains an error, returns a
+        '''Parses a line of test output. If it contains an error, returns a
         formatted message describing the error; otherwise, returns None.
         Subclasses must override this method.
-        """
+        '''
         raise NotImplementedError
 
 
 class GTestErrorParser(ErrorParserBase):
-    """A parser that remembers the last test that began running so it can print
+    '''A parser that remembers the last test that began running so it can print
     that test's name upon detecting failure.
-    """
-
-    _GTEST_NAME_PATTERN = re.compile(r"\[ RUN      \] (\S+)$")
+    '''
+    _GTEST_NAME_PATTERN = re.compile(r'\[ RUN      \] (\S+)$')
     # format: '<filename or "unknown file">:<line #>: Failure'
-    _GTEST_FAIL_PATTERN = re.compile(r"(unknown file|\S+:\d+): Failure$")
+    _GTEST_FAIL_PATTERN = re.compile(r'(unknown file|\S+:\d+): Failure$')
 
     def __init__(self):
-        self._last_gtest_name = "Unknown test"
+        self._last_gtest_name = 'Unknown test'
 
     def parse_error(self, line):
         gtest_name_match = self._GTEST_NAME_PATTERN.match(line)
@@ -43,13 +45,14 @@ class GTestErrorParser(ErrorParserBase):
             return None
         gtest_fail_match = self._GTEST_FAIL_PATTERN.match(line)
         if gtest_fail_match:
-            return "%s failed: %s" % (self._last_gtest_name, gtest_fail_match.group(1))
+            return '%s failed: %s' % (
+                    self._last_gtest_name, gtest_fail_match.group(1))
         return None
 
 
 class MatchErrorParser(ErrorParserBase):
-    """A simple parser that returns the whole line if it matches the pattern."""
-
+    '''A simple parser that returns the whole line if it matches the pattern.
+    '''
     def __init__(self, pattern):
         self._pattern = re.compile(pattern)
 
@@ -66,104 +69,97 @@ class CompilerErrorParser(MatchErrorParser):
         # format (link error):
         #   '<filename>:<line #>: error: <error msg>'
         # The below regex catches both
-        super(CompilerErrorParser, self).__init__(r"\S+:\d+: error:")
+        super(CompilerErrorParser, self).__init__(r'\S+:\d+: error:')
 
 
 class ScanBuildErrorParser(MatchErrorParser):
     def __init__(self):
-        super(ScanBuildErrorParser, self).__init__(r"scan-build: \d+ bugs found.$")
+        super(ScanBuildErrorParser, self).__init__(
+                r'scan-build: \d+ bugs found.$')
 
 
 class DbCrashErrorParser(MatchErrorParser):
     def __init__(self):
-        super(DbCrashErrorParser, self).__init__(r"\*\*\*.*\^$|TEST FAILED.")
+        super(DbCrashErrorParser, self).__init__(r'\*\*\*.*\^$|TEST FAILED.')
 
 
 class WriteStressErrorParser(MatchErrorParser):
     def __init__(self):
         super(WriteStressErrorParser, self).__init__(
-            r"ERROR: write_stress died with exitcode=\d+"
-        )
+                r'ERROR: write_stress died with exitcode=\d+')
 
 
 class AsanErrorParser(MatchErrorParser):
     def __init__(self):
-        super(AsanErrorParser, self).__init__(r"==\d+==ERROR: AddressSanitizer:")
+        super(AsanErrorParser, self).__init__(
+                r'==\d+==ERROR: AddressSanitizer:')
 
 
 class UbsanErrorParser(MatchErrorParser):
     def __init__(self):
         # format: '<filename>:<line #>:<column #>: runtime error: <error msg>'
-        super(UbsanErrorParser, self).__init__(r"\S+:\d+:\d+: runtime error:")
+        super(UbsanErrorParser, self).__init__(r'\S+:\d+:\d+: runtime error:')
 
 
 class ValgrindErrorParser(MatchErrorParser):
     def __init__(self):
         # just grab the summary, valgrind doesn't clearly distinguish errors
         # from other log messages.
-        super(ValgrindErrorParser, self).__init__(r"==\d+== ERROR SUMMARY:")
+        super(ValgrindErrorParser, self).__init__(r'==\d+== ERROR SUMMARY:')
 
 
 class CompatErrorParser(MatchErrorParser):
     def __init__(self):
-        super(CompatErrorParser, self).__init__(r"==== .*[Ee]rror.* ====$")
+        super(CompatErrorParser, self).__init__(r'==== .*[Ee]rror.* ====$')
 
 
 class TsanErrorParser(MatchErrorParser):
     def __init__(self):
-        super(TsanErrorParser, self).__init__(r"WARNING: ThreadSanitizer:")
+        super(TsanErrorParser, self).__init__(r'WARNING: ThreadSanitizer:')
 
 
 _TEST_NAME_TO_PARSERS = {
-    "punit": [CompilerErrorParser, GTestErrorParser],
-    "unit": [CompilerErrorParser, GTestErrorParser],
-    "release": [CompilerErrorParser, GTestErrorParser],
-    "unit_481": [CompilerErrorParser, GTestErrorParser],
-    "release_481": [CompilerErrorParser, GTestErrorParser],
-    "clang_unit": [CompilerErrorParser, GTestErrorParser],
-    "clang_release": [CompilerErrorParser, GTestErrorParser],
-    "clang_analyze": [CompilerErrorParser, ScanBuildErrorParser],
-    "code_cov": [CompilerErrorParser, GTestErrorParser],
-    "unity": [CompilerErrorParser, GTestErrorParser],
-    "lite": [CompilerErrorParser],
-    "lite_test": [CompilerErrorParser, GTestErrorParser],
-    "stress_crash": [CompilerErrorParser, DbCrashErrorParser],
-    "stress_crash_with_atomic_flush": [CompilerErrorParser, DbCrashErrorParser],
-    "stress_crash_with_txn": [CompilerErrorParser, DbCrashErrorParser],
-    "write_stress": [CompilerErrorParser, WriteStressErrorParser],
-    "asan": [CompilerErrorParser, GTestErrorParser, AsanErrorParser],
-    "asan_crash": [CompilerErrorParser, AsanErrorParser, DbCrashErrorParser],
-    "asan_crash_with_atomic_flush": [
-        CompilerErrorParser,
-        AsanErrorParser,
-        DbCrashErrorParser,
-    ],
-    "asan_crash_with_txn": [CompilerErrorParser, AsanErrorParser, DbCrashErrorParser],
-    "ubsan": [CompilerErrorParser, GTestErrorParser, UbsanErrorParser],
-    "ubsan_crash": [CompilerErrorParser, UbsanErrorParser, DbCrashErrorParser],
-    "ubsan_crash_with_atomic_flush": [
-        CompilerErrorParser,
-        UbsanErrorParser,
-        DbCrashErrorParser,
-    ],
-    "ubsan_crash_with_txn": [CompilerErrorParser, UbsanErrorParser, DbCrashErrorParser],
-    "valgrind": [CompilerErrorParser, GTestErrorParser, ValgrindErrorParser],
-    "tsan": [CompilerErrorParser, GTestErrorParser, TsanErrorParser],
-    "format_compatible": [CompilerErrorParser, CompatErrorParser],
-    "run_format_compatible": [CompilerErrorParser, CompatErrorParser],
-    "no_compression": [CompilerErrorParser, GTestErrorParser],
-    "run_no_compression": [CompilerErrorParser, GTestErrorParser],
-    "regression": [CompilerErrorParser],
-    "run_regression": [CompilerErrorParser],
+    'punit': [CompilerErrorParser, GTestErrorParser],
+    'unit': [CompilerErrorParser, GTestErrorParser],
+    'release': [CompilerErrorParser, GTestErrorParser],
+    'unit_481': [CompilerErrorParser, GTestErrorParser],
+    'release_481': [CompilerErrorParser, GTestErrorParser],
+    'clang_unit': [CompilerErrorParser, GTestErrorParser],
+    'clang_release': [CompilerErrorParser, GTestErrorParser],
+    'clang_analyze': [CompilerErrorParser, ScanBuildErrorParser],
+    'code_cov': [CompilerErrorParser, GTestErrorParser],
+    'unity': [CompilerErrorParser, GTestErrorParser],
+    'lite': [CompilerErrorParser],
+    'lite_test': [CompilerErrorParser, GTestErrorParser],
+    'stress_crash': [CompilerErrorParser, DbCrashErrorParser],
+    'stress_crash_with_atomic_flush': [CompilerErrorParser, DbCrashErrorParser],
+    'stress_crash_with_txn': [CompilerErrorParser, DbCrashErrorParser],
+    'write_stress': [CompilerErrorParser, WriteStressErrorParser],
+    'asan': [CompilerErrorParser, GTestErrorParser, AsanErrorParser],
+    'asan_crash': [CompilerErrorParser, AsanErrorParser, DbCrashErrorParser],
+    'asan_crash_with_atomic_flush': [CompilerErrorParser, AsanErrorParser, DbCrashErrorParser],
+    'asan_crash_with_txn': [CompilerErrorParser, AsanErrorParser, DbCrashErrorParser],
+    'ubsan': [CompilerErrorParser, GTestErrorParser, UbsanErrorParser],
+    'ubsan_crash': [CompilerErrorParser, UbsanErrorParser, DbCrashErrorParser],
+    'ubsan_crash_with_atomic_flush': [CompilerErrorParser, UbsanErrorParser, DbCrashErrorParser],
+    'ubsan_crash_with_txn': [CompilerErrorParser, UbsanErrorParser, DbCrashErrorParser],
+    'valgrind': [CompilerErrorParser, GTestErrorParser, ValgrindErrorParser],
+    'tsan': [CompilerErrorParser, GTestErrorParser, TsanErrorParser],
+    'format_compatible': [CompilerErrorParser, CompatErrorParser],
+    'run_format_compatible': [CompilerErrorParser, CompatErrorParser],
+    'no_compression': [CompilerErrorParser, GTestErrorParser],
+    'run_no_compression': [CompilerErrorParser, GTestErrorParser],
+    'regression': [CompilerErrorParser],
+    'run_regression': [CompilerErrorParser],
 }
 
 
 def main():
     if len(sys.argv) != 2:
-        return "Usage: %s <test name>" % sys.argv[0]
+        return 'Usage: %s <test name>' % sys.argv[0]
     test_name = sys.argv[1]
     if test_name not in _TEST_NAME_TO_PARSERS:
-        return "Unknown test name: %s" % test_name
+        return 'Unknown test name: %s' % test_name
 
     error_parsers = []
     for parser_cls in _TEST_NAME_TO_PARSERS[test_name]:
@@ -177,5 +173,5 @@ def main():
                 print(error_msg)
 
 
-if __name__ == "__main__":
+if __name__ == '__main__':
     sys.exit(main())

build_tools/fbcode_config_platform009.sh

@@ -0,0 +1,170 @@
+#!/bin/sh
+# Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved.
+#
+# Set environment variables so that we can compile rocksdb using
+# fbcode settings.  It uses the latest g++ and clang compilers and also
+# uses jemalloc
+# Environment variables that change the behavior of this script:
+# PIC_BUILD -- if true, it will only take pic versions of libraries from fbcode. libraries that don't have pic variant will not be included
+
+
+BASEDIR=`dirname $BASH_SOURCE`
+source "$BASEDIR/dependencies_platform009.sh"
+
+CFLAGS=""
+
+# libgcc
+LIBGCC_INCLUDE="$LIBGCC_BASE/include/c++/9.3.0 -I $LIBGCC_BASE/include/c++/9.3.0/backward"
+LIBGCC_LIBS=" -L $LIBGCC_BASE/lib"
+
+# glibc
+GLIBC_INCLUDE="$GLIBC_BASE/include"
+GLIBC_LIBS=" -L $GLIBC_BASE/lib"
+
+if test -z $PIC_BUILD; then
+  MAYBE_PIC=
+else
+  MAYBE_PIC=_pic
+fi
+
+if ! test $ROCKSDB_DISABLE_SNAPPY; then
+  # snappy
+  SNAPPY_INCLUDE=" -I $SNAPPY_BASE/include/"
+  SNAPPY_LIBS=" $SNAPPY_BASE/lib/libsnappy${MAYBE_PIC}.a"
+  CFLAGS+=" -DSNAPPY"
+fi
+
+if ! test $ROCKSDB_DISABLE_ZLIB; then
+  # location of zlib headers and libraries
+  ZLIB_INCLUDE=" -I $ZLIB_BASE/include/"
+  ZLIB_LIBS=" $ZLIB_BASE/lib/libz${MAYBE_PIC}.a"
+  CFLAGS+=" -DZLIB"
+fi
+
+if ! test $ROCKSDB_DISABLE_BZIP; then
+  # location of bzip headers and libraries
+  BZIP_INCLUDE=" -I $BZIP2_BASE/include/"
+  BZIP_LIBS=" $BZIP2_BASE/lib/libbz2${MAYBE_PIC}.a"
+  CFLAGS+=" -DBZIP2"
+fi
+
+if ! test $ROCKSDB_DISABLE_LZ4; then
+  LZ4_INCLUDE=" -I $LZ4_BASE/include/"
+  LZ4_LIBS=" $LZ4_BASE/lib/liblz4${MAYBE_PIC}.a"
+  CFLAGS+=" -DLZ4"
+fi
+
+if ! test $ROCKSDB_DISABLE_ZSTD; then
+  ZSTD_INCLUDE=" -I $ZSTD_BASE/include/"
+  ZSTD_LIBS=" $ZSTD_BASE/lib/libzstd${MAYBE_PIC}.a"
+  CFLAGS+=" -DZSTD"
+fi
+
+# location of gflags headers and libraries
+GFLAGS_INCLUDE=" -I $GFLAGS_BASE/include/"
+GFLAGS_LIBS=" $GFLAGS_BASE/lib/libgflags${MAYBE_PIC}.a"
+CFLAGS+=" -DGFLAGS=gflags"
+
+BENCHMARK_INCLUDE=" -I $BENCHMARK_BASE/include/"
+BENCHMARK_LIBS=" $BENCHMARK_BASE/lib/libbenchmark${MAYBE_PIC}.a"
+
+GLOG_INCLUDE=" -I $GLOG_BASE/include/"
+GLOG_LIBS=" $GLOG_BASE/lib/libglog${MAYBE_PIC}.a"
+
+# location of jemalloc
+JEMALLOC_INCLUDE=" -I $JEMALLOC_BASE/include/"
+JEMALLOC_LIB=" $JEMALLOC_BASE/lib/libjemalloc${MAYBE_PIC}.a"
+
+# location of numa
+NUMA_INCLUDE=" -I $NUMA_BASE/include/"
+NUMA_LIB=" $NUMA_BASE/lib/libnuma${MAYBE_PIC}.a"
+CFLAGS+=" -DNUMA"
+
+# location of libunwind
+LIBUNWIND="$LIBUNWIND_BASE/lib/libunwind${MAYBE_PIC}.a"
+
+# location of TBB
+TBB_INCLUDE=" -isystem $TBB_BASE/include/"
+TBB_LIBS="$TBB_BASE/lib/libtbb${MAYBE_PIC}.a"
+CFLAGS+=" -DTBB"
+
+# location of LIBURING
+LIBURING_INCLUDE=" -isystem $LIBURING_BASE/include/"
+LIBURING_LIBS="$LIBURING_BASE/lib/liburing${MAYBE_PIC}.a"
+CFLAGS+=" -DLIBURING"
+
+test "$USE_SSE" || USE_SSE=1
+export USE_SSE
+test "$PORTABLE" || PORTABLE=1
+export PORTABLE
+
+BINUTILS="$BINUTILS_BASE/bin"
+AR="$BINUTILS/ar"
+AS="$BINUTILS/as"
+
+DEPS_INCLUDE="$SNAPPY_INCLUDE $ZLIB_INCLUDE $BZIP_INCLUDE $LZ4_INCLUDE $ZSTD_INCLUDE $GFLAGS_INCLUDE $NUMA_INCLUDE $TBB_INCLUDE $LIBURING_INCLUDE $BENCHMARK_INCLUDE $GLOG_INCLUDE"
+
+STDLIBS="-L $GCC_BASE/lib64"
+
+CLANG_BIN="$CLANG_BASE/bin"
+CLANG_LIB="$CLANG_BASE/lib"
+CLANG_SRC="$CLANG_BASE/../../src"
+
+CLANG_ANALYZER="$CLANG_BIN/clang++"
+CLANG_SCAN_BUILD="$CLANG_SRC/llvm/clang/tools/scan-build/bin/scan-build"
+
+if [ -z "$USE_CLANG" ]; then
+  # gcc
+  CC="$GCC_BASE/bin/gcc"
+  CXX="$GCC_BASE/bin/g++"
+  AR="$GCC_BASE/bin/gcc-ar"
+
+  CFLAGS+=" -B$BINUTILS"
+  CFLAGS+=" -isystem $LIBGCC_INCLUDE"
+  CFLAGS+=" -isystem $GLIBC_INCLUDE"
+  JEMALLOC=1
+else
+  # clang
+  CLANG_INCLUDE="$CLANG_LIB/clang/stable/include"
+  CC="$CLANG_BIN/clang"
+  CXX="$CLANG_BIN/clang++"
+  AR="$CLANG_BIN/llvm-ar"
+
+  KERNEL_HEADERS_INCLUDE="$KERNEL_HEADERS_BASE/include"
+
+  CFLAGS+=" -B$BINUTILS -nostdinc -nostdlib"
+  CFLAGS+=" -isystem $LIBGCC_BASE/include/c++/9.x "
+  CFLAGS+=" -isystem $LIBGCC_BASE/include/c++/9.x/x86_64-facebook-linux "
+  CFLAGS+=" -isystem $GLIBC_INCLUDE"
+  CFLAGS+=" -isystem $LIBGCC_INCLUDE"
+  CFLAGS+=" -isystem $CLANG_INCLUDE"
+  CFLAGS+=" -isystem $KERNEL_HEADERS_INCLUDE/linux "
+  CFLAGS+=" -isystem $KERNEL_HEADERS_INCLUDE "
+  CFLAGS+=" -Wno-expansion-to-defined "
+  CXXFLAGS="-nostdinc++"
+fi
+
+CFLAGS+=" $DEPS_INCLUDE"
+CFLAGS+=" -DROCKSDB_PLATFORM_POSIX -DROCKSDB_LIB_IO_POSIX -DROCKSDB_FALLOCATE_PRESENT -DROCKSDB_MALLOC_USABLE_SIZE -DROCKSDB_RANGESYNC_PRESENT -DROCKSDB_SCHED_GETCPU_PRESENT -DHAVE_SSE42 -DROCKSDB_IOURING_PRESENT"
+CXXFLAGS+=" $CFLAGS"
+
+EXEC_LDFLAGS=" $SNAPPY_LIBS $ZLIB_LIBS $BZIP_LIBS $LZ4_LIBS $ZSTD_LIBS $GFLAGS_LIBS $NUMA_LIB $TBB_LIBS $LIBURING_LIBS $BENCHMARK_LIBS"
+EXEC_LDFLAGS+=" -Wl,--dynamic-linker,/usr/local/fbcode/platform009/lib/ld.so"
+EXEC_LDFLAGS+=" $LIBUNWIND"
+EXEC_LDFLAGS+=" -Wl,-rpath=/usr/local/fbcode/platform009/lib"
+EXEC_LDFLAGS+=" -Wl,-rpath=$GCC_BASE/lib64"
+# required by libtbb
+EXEC_LDFLAGS+=" -ldl"
+
+PLATFORM_LDFLAGS="$LIBGCC_LIBS $GLIBC_LIBS $STDLIBS -lgcc -lstdc++"
+PLATFORM_LDFLAGS+=" -B$BINUTILS"
+
+EXEC_LDFLAGS_SHARED="$SNAPPY_LIBS $ZLIB_LIBS $BZIP_LIBS $LZ4_LIBS $ZSTD_LIBS $GFLAGS_LIBS $TBB_LIBS $LIBURING_LIBS $BENCHMARK_LIBS"
+
+VALGRIND_VER="$VALGRIND_BASE/bin/"
+
+# lua not supported because it's on track for deprecation, I think
+LUA_PATH=
+LUA_LIB=
+
+export CC CXX AR AS CFLAGS CXXFLAGS EXEC_LDFLAGS EXEC_LDFLAGS_SHARED VALGRIND_VER JEMALLOC_LIB JEMALLOC_INCLUDE CLANG_ANALYZER CLANG_SCAN_BUILD LUA_PATH LUA_LIB

build_tools/gnu_parallel

@@ -1170,7 +1170,7 @@ sub parse_env_var {
     # Check if any variables contain \n
     if(my @v = map { s/BASH_FUNC_(.*)\(\)/$1/; $_ } grep { $ENV{$_}=~/\n/ } @vars) {
 	# \n is bad for csh and will cause it to fail.
-	$Global::envwarn = ::shell_quote_scalar(q{echo $SHELL | grep -E "/t?csh" > /dev/null && echo CSH/TCSH DO NOT SUPPORT newlines IN VARIABLES/FUNCTIONS. Unset }."@v".q{ && exec false;}."\n\n") . $Global::envwarn;
+	$Global::envwarn = ::shell_quote_scalar(q{echo $SHELL | egrep "/t?csh" > /dev/null && echo CSH/TCSH DO NOT SUPPORT newlines IN VARIABLES/FUNCTIONS. Unset }."@v".q{ && exec false;}."\n\n") . $Global::envwarn;
     }
 
     if(not @qcsh) { push @qcsh, "true"; }
@@ -1916,8 +1916,7 @@ sub drain_job_queue {
                 } elsif (not $ps_reported and (time() - $last_progress_time) >= 60) {
                     # No progress in at least 60 seconds: run ps
                     print $Global::original_stderr "\n";
-                    my $script_dir = ::dirname($0);
-                    system("$script_dir/ps_with_stack || ps -wwf");
+                    system("ps", "-wf");
                     $ps_reported = 1;
                 }
                 $last_left = $Global::left;

build_tools/ps_with_stack

@@ -1,38 +0,0 @@
-#!/usr/bin/env perl
-
-use strict;
-
-open(my $ps, "-|", "ps -wwf");
-my $cols_known = 0;
-my $cmd_col = 0;
-my $pid_col = 0;
-while (<$ps>) {
-  print;
-  my @cols = split(/\s+/);
-
-  if (!$cols_known && /CMD/) {
-    # Parse relevant ps column headers
-    for (my $i = 0; $i <= $#cols; $i++) {
-      if ($cols[$i] eq "CMD") {
-        $cmd_col = $i;
-      }
-      if ($cols[$i] eq "PID") {
-        $pid_col = $i;
-      }
-    }
-    $cols_known = 1;
-  } else {
-    my $pid = $cols[$pid_col];
-    my $cmd = $cols[$cmd_col];
-    # Match numeric PID and relative path command
-    # -> The intention is only to dump stack traces for hangs in code under
-    # test, which means we probably just built it and are executing by
-    # relative path (e.g. ./my_test or foo/bar_test) rather then by absolute
-    # path (e.g. /usr/bin/time) or PATH search (e.g. grep).
-    if ($pid =~ /^[0-9]+$/ && $cmd =~ /^[^\/ ]+[\/]/) {
-      print "Dumping stacks for $pid...\n";
-      system("pstack $pid || gdb -batch -p $pid -ex 'thread apply all bt'");
-    }
-  }
-}
-close $ps;

build_tools/regression_build_test.sh

@@ -360,7 +360,7 @@ function send_to_ods {
     echo >&2 "ERROR: Key $key doesn't have a value."
     return
   fi
-  curl --silent "https://www.facebook.com/intern/agent/ods_set.php?entity=rocksdb_build&key=$key&value=$value" \
+  curl --silent "https://www.intern.facebook.com/intern/agent/ods_set.php?entity=rocksdb_build&key=$key&value=$value" \
     --connect-timeout 60
 }
 

build_tools/ubuntu20_image/Dockerfile

@@ -16,7 +16,6 @@ RUN apt-get install -y libgflags-dev libtbb-dev
 RUN apt-get install -y libsnappy-dev zlib1g-dev libbz2-dev liblz4-dev libzstd-dev
 # install cmake
 RUN apt-get install -y cmake
-RUN apt-get install -y libssl-dev
 # install clang-13
 WORKDIR /root
 RUN wget https://apt.llvm.org/llvm.sh

build_tools/update_dependencies.sh

@@ -104,3 +104,46 @@ get_lib_base valgrind   LATEST platform010
 get_lib_base lua        5.3.4  platform010
 
 git diff $OUTPUT
+
+
+###########################################################
+#                platform009 dependencies                 #
+###########################################################
+
+OUTPUT="$BASEDIR/dependencies_platform009.sh"
+
+rm -f "$OUTPUT"
+touch "$OUTPUT"
+
+echo "Writing dependencies to $OUTPUT"
+
+# Compilers locations
+GCC_BASE=`readlink -f $TP2_LATEST/gcc/9.x/centos7-native/*/`
+CLANG_BASE=`readlink -f $TP2_LATEST/llvm-fb/9.0.0/platform009/*/`
+
+log_header
+log_variable GCC_BASE
+log_variable CLANG_BASE
+
+# Libraries locations
+get_lib_base libgcc     9.x     platform009
+get_lib_base glibc      2.30    platform009
+get_lib_base snappy     LATEST  platform009
+get_lib_base zlib       LATEST  platform009
+get_lib_base bzip2      LATEST  platform009
+get_lib_base lz4        LATEST  platform009
+get_lib_base zstd       LATEST  platform009
+get_lib_base gflags     LATEST  platform009
+get_lib_base jemalloc   LATEST  platform009
+get_lib_base numa       LATEST  platform009
+get_lib_base libunwind  LATEST  platform009
+get_lib_base tbb        2018_U5 platform009
+get_lib_base liburing   LATEST  platform009
+get_lib_base benchmark  LATEST  platform009
+
+get_lib_base kernel-headers fb platform009
+get_lib_base binutils   LATEST centos7-native
+get_lib_base valgrind   LATEST platform009
+get_lib_base lua        5.3.4  platform009
+
+git diff $OUTPUT

cache/cache.cc

@@ -16,8 +16,7 @@
 #include "util/string_util.h"
 
 namespace ROCKSDB_NAMESPACE {
-const Cache::CacheItemHelper kNoopCacheItemHelper{};
-
+#ifndef ROCKSDB_LITE
 static std::unordered_map<std::string, OptionTypeInfo>
     lru_cache_options_type_info = {
         {"capacity",
@@ -65,6 +64,7 @@ static std::unordered_map<std::string, OptionTypeInfo>
           OptionType::kBoolean, OptionVerificationType::kNormal,
           OptionTypeFlags::kMutable}},
 };
+#endif  // ROCKSDB_LITE
 
 Status SecondaryCache::CreateFromString(
     const ConfigOptions& config_options, const std::string& value,
@@ -75,6 +75,7 @@ Status SecondaryCache::CreateFromString(
     Status status;
     std::shared_ptr<SecondaryCache> sec_cache;
 
+#ifndef ROCKSDB_LITE
     CompressedSecondaryCacheOptions sec_cache_opts;
     status = OptionTypeInfo::ParseStruct(config_options, "",
                                          &comp_sec_cache_options_type_info, "",
@@ -83,13 +84,19 @@ Status SecondaryCache::CreateFromString(
       sec_cache = NewCompressedSecondaryCache(sec_cache_opts);
     }
 
+#else
+    (void)config_options;
+    status = Status::NotSupported(
+        "Cannot load compressed secondary cache in LITE mode ", args);
+#endif  //! ROCKSDB_LITE
 
     if (status.ok()) {
       result->swap(sec_cache);
     }
     return status;
   } else {
-    return LoadSharedObject<SecondaryCache>(config_options, value, result);
+    return LoadSharedObject<SecondaryCache>(config_options, value, nullptr,
+                                            result);
   }
 }
 
@@ -101,6 +108,7 @@ Status Cache::CreateFromString(const ConfigOptions& config_options,
   if (value.find('=') == std::string::npos) {
     cache = NewLRUCache(ParseSizeT(value));
   } else {
+#ifndef ROCKSDB_LITE
     LRUCacheOptions cache_opts;
     status = OptionTypeInfo::ParseStruct(config_options, "",
                                          &lru_cache_options_type_info, "",
@@ -108,51 +116,14 @@ Status Cache::CreateFromString(const ConfigOptions& config_options,
     if (status.ok()) {
       cache = NewLRUCache(cache_opts);
     }
+#else
+    (void)config_options;
+    status = Status::NotSupported("Cannot load cache in LITE mode ", value);
+#endif  //! ROCKSDB_LITE
   }
   if (status.ok()) {
     result->swap(cache);
   }
   return status;
 }
-
-bool Cache::AsyncLookupHandle::IsReady() {
-  return pending_handle == nullptr || pending_handle->IsReady();
-}
-
-bool Cache::AsyncLookupHandle::IsPending() { return pending_handle != nullptr; }
-
-Cache::Handle* Cache::AsyncLookupHandle::Result() {
-  assert(!IsPending());
-  return result_handle;
-}
-
-void Cache::StartAsyncLookup(AsyncLookupHandle& async_handle) {
-  async_handle.found_dummy_entry = false;  // in case re-used
-  assert(!async_handle.IsPending());
-  async_handle.result_handle =
-      Lookup(async_handle.key, async_handle.helper, async_handle.create_context,
-             async_handle.priority, async_handle.stats);
-}
-
-Cache::Handle* Cache::Wait(AsyncLookupHandle& async_handle) {
-  WaitAll(&async_handle, 1);
-  return async_handle.Result();
-}
-
-void Cache::WaitAll(AsyncLookupHandle* async_handles, size_t count) {
-  for (size_t i = 0; i < count; ++i) {
-    if (async_handles[i].IsPending()) {
-      // If a pending handle gets here, it should be marked at "to be handled
-      // by a caller" by that caller erasing the pending_cache on it.
-      assert(async_handles[i].pending_cache == nullptr);
-    }
-  }
-}
-
-void Cache::SetEvictionCallback(EvictionCallback&& fn) {
-  // Overwriting non-empty with non-empty could indicate a bug
-  assert(!eviction_callback_ || !fn);
-  eviction_callback_ = std::move(fn);
-}
-
 }  // namespace ROCKSDB_NAMESPACE

cache/cache_bench_tool.cc

@@ -13,10 +13,11 @@
 #include <set>
 #include <sstream>
 
+#include "cache/fast_lru_cache.h"
 #include "db/db_impl/db_impl.h"
 #include "monitoring/histogram.h"
 #include "port/port.h"
-#include "rocksdb/advanced_cache.h"
+#include "rocksdb/cache.h"
 #include "rocksdb/convenience.h"
 #include "rocksdb/db.h"
 #include "rocksdb/env.h"
@@ -77,9 +78,11 @@ DEFINE_bool(lean, false,
             "If true, no additional computation is performed besides cache "
             "operations.");
 
+#ifndef ROCKSDB_LITE
 DEFINE_string(secondary_cache_uri, "",
               "Full URI for creating a custom secondary cache object");
 static class std::shared_ptr<ROCKSDB_NAMESPACE::SecondaryCache> secondary_cache;
+#endif  // ROCKSDB_LITE
 
 DEFINE_string(cache_type, "lru_cache", "Type of block cache.");
 
@@ -224,7 +227,7 @@ struct KeyGen {
   }
 };
 
-Cache::ObjectPtr createValue(Random64& rnd) {
+char* createValue(Random64& rnd) {
   char* rv = new char[FLAGS_value_bytes];
   // Fill with some filler data, and take some CPU time
   for (uint32_t i = 0; i < FLAGS_value_bytes; i += 8) {
@@ -234,36 +237,28 @@ Cache::ObjectPtr createValue(Random64& rnd) {
 }
 
 // Callbacks for secondary cache
-size_t SizeFn(Cache::ObjectPtr /*obj*/) { return FLAGS_value_bytes; }
+size_t SizeFn(void* /*obj*/) { return FLAGS_value_bytes; }
 
-Status SaveToFn(Cache::ObjectPtr from_obj, size_t /*from_offset*/,
-                size_t length, char* out) {
-  memcpy(out, from_obj, length);
+Status SaveToFn(void* obj, size_t /*offset*/, size_t size, void* out) {
+  memcpy(out, obj, size);
   return Status::OK();
 }
 
-Status CreateFn(const Slice& data, Cache::CreateContext* /*context*/,
-                MemoryAllocator* /*allocator*/, Cache::ObjectPtr* out_obj,
-                size_t* out_charge) {
-  *out_obj = new char[data.size()];
-  memcpy(*out_obj, data.data(), data.size());
-  *out_charge = data.size();
-  return Status::OK();
-};
-
-void DeleteFn(Cache::ObjectPtr value, MemoryAllocator* /*alloc*/) {
+// Different deleters to simulate using deleter to gather
+// stats on the code origin and kind of cache entries.
+void deleter1(const Slice& /*key*/, void* value) {
+  delete[] static_cast<char*>(value);
+}
+void deleter2(const Slice& /*key*/, void* value) {
+  delete[] static_cast<char*>(value);
+}
+void deleter3(const Slice& /*key*/, void* value) {
   delete[] static_cast<char*>(value);
 }
 
-Cache::CacheItemHelper helper1_wos(CacheEntryRole::kDataBlock, DeleteFn);
-Cache::CacheItemHelper helper1(CacheEntryRole::kDataBlock, DeleteFn, SizeFn,
-                               SaveToFn, CreateFn, &helper1_wos);
-Cache::CacheItemHelper helper2_wos(CacheEntryRole::kIndexBlock, DeleteFn);
-Cache::CacheItemHelper helper2(CacheEntryRole::kIndexBlock, DeleteFn, SizeFn,
-                               SaveToFn, CreateFn, &helper2_wos);
-Cache::CacheItemHelper helper3_wos(CacheEntryRole::kFilterBlock, DeleteFn);
-Cache::CacheItemHelper helper3(CacheEntryRole::kFilterBlock, DeleteFn, SizeFn,
-                               SaveToFn, CreateFn, &helper3_wos);
+Cache::CacheItemHelper helper1(SizeFn, SaveToFn, deleter1);
+Cache::CacheItemHelper helper2(SizeFn, SaveToFn, deleter2);
+Cache::CacheItemHelper helper3(SizeFn, SaveToFn, deleter3);
 }  // namespace
 
 class CacheBench {
@@ -302,10 +297,15 @@ class CacheBench {
       cache_ = HyperClockCacheOptions(FLAGS_cache_size, FLAGS_value_bytes,
                                       FLAGS_num_shard_bits)
                    .MakeSharedCache();
+    } else if (FLAGS_cache_type == "fast_lru_cache") {
+      cache_ = NewFastLRUCache(
+          FLAGS_cache_size, FLAGS_value_bytes, FLAGS_num_shard_bits,
+          false /*strict_capacity_limit*/, kDefaultCacheMetadataChargePolicy);
     } else if (FLAGS_cache_type == "lru_cache") {
       LRUCacheOptions opts(FLAGS_cache_size, FLAGS_num_shard_bits,
                            false /* strict_capacity_limit */,
                            0.5 /* high_pri_pool_ratio */);
+#ifndef ROCKSDB_LITE
       if (!FLAGS_secondary_cache_uri.empty()) {
         Status s = SecondaryCache::CreateFromString(
             ConfigOptions(), FLAGS_secondary_cache_uri, &secondary_cache);
@@ -318,6 +318,7 @@ class CacheBench {
         }
         opts.secondary_cache = secondary_cache;
       }
+#endif  // ROCKSDB_LITE
 
       cache_ = NewLRUCache(opts);
     } else {
@@ -440,7 +441,7 @@ class CacheBench {
     uint64_t total_entry_count = 0;
     uint64_t table_occupancy = 0;
     uint64_t table_size = 0;
-    std::set<const Cache::CacheItemHelper*> helpers;
+    std::set<Cache::DeleterFn> deleters;
     StopWatchNano timer(clock);
 
     for (;;) {
@@ -465,7 +466,7 @@ class CacheBench {
                  << BytesToHumanString(static_cast<uint64_t>(
                         1.0 * total_charge / total_entry_count))
                  << "\n"
-                 << "Unique helpers: " << helpers.size() << "\n";
+                 << "Unique deleters: " << deleters.size() << "\n";
             *stats_report = ostr.str();
             return;
           }
@@ -481,14 +482,14 @@ class CacheBench {
       total_key_size = 0;
       total_charge = 0;
       total_entry_count = 0;
-      helpers.clear();
-      auto fn = [&](const Slice& key, Cache::ObjectPtr /*value*/, size_t charge,
-                    const Cache::CacheItemHelper* helper) {
+      deleters.clear();
+      auto fn = [&](const Slice& key, void* /*value*/, size_t charge,
+                    Cache::DeleterFn deleter) {
         total_key_size += key.size();
         total_charge += charge;
         ++total_entry_count;
-        // Something slightly more expensive as in stats by category
-        helpers.insert(helper);
+        // Something slightly more expensive as in (future) stats by category
+        deleters.insert(deleter);
       };
       timer.Start();
       Cache::ApplyToAllEntriesOptions opts;
@@ -537,6 +538,14 @@ class CacheBench {
     for (uint64_t i = 0; i < FLAGS_ops_per_thread; i++) {
       Slice key = gen.GetRand(thread->rnd, max_key_, max_log_);
       uint64_t random_op = thread->rnd.Next();
+      Cache::CreateCallback create_cb = [](const void* buf, size_t size,
+                                           void** out_obj,
+                                           size_t* charge) -> Status {
+        *out_obj = reinterpret_cast<void*>(new char[size]);
+        memcpy(*out_obj, buf, size);
+        *charge = size;
+        return Status::OK();
+      };
 
       timer.Start();
 
@@ -546,8 +555,8 @@ class CacheBench {
           handle = nullptr;
         }
         // do lookup
-        handle = cache_->Lookup(key, &helper2, /*context*/ nullptr,
-                                Cache::Priority::LOW);
+        handle = cache_->Lookup(key, &helper2, create_cb, Cache::Priority::LOW,
+                                true);
         if (handle) {
           if (!FLAGS_lean) {
             // do something with the data
@@ -575,8 +584,8 @@ class CacheBench {
           handle = nullptr;
         }
         // do lookup
-        handle = cache_->Lookup(key, &helper2, /*context*/ nullptr,
-                                Cache::Priority::LOW);
+        handle = cache_->Lookup(key, &helper2, create_cb, Cache::Priority::LOW,
+                                true);
         if (handle) {
           if (!FLAGS_lean) {
             // do something with the data

cache/cache_entry_roles.cc

@@ -101,4 +101,34 @@ std::string BlockCacheEntryStatsMapKeys::UsedPercent(CacheEntryRole role) {
   return GetPrefixedCacheEntryRoleName(kPrefix, role);
 }
 
+namespace {
+
+struct Registry {
+  std::mutex mutex;
+  UnorderedMap<Cache::DeleterFn, CacheEntryRole> role_map;
+  void Register(Cache::DeleterFn fn, CacheEntryRole role) {
+    std::lock_guard<std::mutex> lock(mutex);
+    role_map[fn] = role;
+  }
+  UnorderedMap<Cache::DeleterFn, CacheEntryRole> Copy() {
+    std::lock_guard<std::mutex> lock(mutex);
+    return role_map;
+  }
+};
+
+Registry& GetRegistry() {
+  STATIC_AVOID_DESTRUCTION(Registry, registry);
+  return registry;
+}
+
+}  // namespace
+
+void RegisterCacheDeleterRole(Cache::DeleterFn fn, CacheEntryRole role) {
+  GetRegistry().Register(fn, role);
+}
+
+UnorderedMap<Cache::DeleterFn, CacheEntryRole> CopyCacheDeleterRoleMap() {
+  return GetRegistry().Copy();
+}
+
 }  // namespace ROCKSDB_NAMESPACE

cache/cache_entry_roles.h

@@ -7,8 +7,11 @@
 
 #include <array>
 #include <cstdint>
+#include <memory>
+#include <type_traits>
 
 #include "rocksdb/cache.h"
+#include "util/hash_containers.h"
 
 namespace ROCKSDB_NAMESPACE {
 
@@ -17,4 +20,84 @@ extern std::array<std::string, kNumCacheEntryRoles>
 extern std::array<std::string, kNumCacheEntryRoles>
     kCacheEntryRoleToHyphenString;
 
+// To associate cache entries with their role, we use a hack on the
+// existing Cache interface. Because the deleter of an entry can authenticate
+// the code origin of an entry, we can elaborate the choice of deleter to
+// also encode role information, without inferring false role information
+// from entries not choosing to encode a role.
+//
+// The rest of this file is for handling mappings between deleters and
+// roles.
+
+// To infer a role from a deleter, the deleter must be registered. This
+// can be done "manually" with this function. This function is thread-safe,
+// and the registration mappings go into private but static storage. (Note
+// that DeleterFn is a function pointer, not std::function. Registrations
+// should not be too many.)
+void RegisterCacheDeleterRole(Cache::DeleterFn fn, CacheEntryRole role);
+
+// Gets a copy of the registered deleter -> role mappings. This is the only
+// function for reading the mappings made with RegisterCacheDeleterRole.
+// Why only this interface for reading?
+// * This function has to be thread safe, which could incur substantial
+// overhead. We should not pay this overhead for every deleter look-up.
+// * This is suitable for preparing for batch operations, like with
+// CacheEntryStatsCollector.
+// * The number of mappings should be sufficiently small (dozens).
+UnorderedMap<Cache::DeleterFn, CacheEntryRole> CopyCacheDeleterRoleMap();
+
+// ************************************************************** //
+// An automatic registration infrastructure. This enables code
+// to simply ask for a deleter associated with a particular type
+// and role, and registration is automatic. In a sense, this is
+// a small dependency injection infrastructure, because linking
+// in new deleter instantiations is essentially sufficient for
+// making stats collection (using CopyCacheDeleterRoleMap) aware
+// of them.
+
+namespace cache_entry_roles_detail {
+
+template <typename T, CacheEntryRole R>
+struct RegisteredDeleter {
+  RegisteredDeleter() { RegisterCacheDeleterRole(Delete, R); }
+
+  // These have global linkage to help ensure compiler optimizations do not
+  // break uniqueness for each <T,R>
+  static void Delete(const Slice& /* key */, void* value) {
+    // Supports T == Something[], unlike delete operator
+    std::default_delete<T>()(
+        static_cast<typename std::remove_extent<T>::type*>(value));
+  }
+};
+
+template <CacheEntryRole R>
+struct RegisteredNoopDeleter {
+  RegisteredNoopDeleter() { RegisterCacheDeleterRole(Delete, R); }
+
+  static void Delete(const Slice& /* key */, void* /* value */) {
+    // Here was `assert(value == nullptr);` but we can also put pointers
+    // to static data in Cache, for testing at least.
+  }
+};
+
+}  // namespace cache_entry_roles_detail
+
+// Get an automatically registered deleter for value type T and role R.
+// Based on C++ semantics, registration is invoked exactly once in a
+// thread-safe way on first call to this function, for each <T, R>.
+template <typename T, CacheEntryRole R>
+Cache::DeleterFn GetCacheEntryDeleterForRole() {
+  static cache_entry_roles_detail::RegisteredDeleter<T, R> reg;
+  return reg.Delete;
+}
+
+// Get an automatically registered no-op deleter (value should be nullptr)
+// and associated with role R. This is used for Cache "reservation" entries
+// such as for WriteBufferManager.
+template <CacheEntryRole R>
+Cache::DeleterFn GetNoopDeleterForRole() {
+  static cache_entry_roles_detail::RegisteredNoopDeleter<R> reg;
+  return reg.Delete;
+}
+
 }  // namespace ROCKSDB_NAMESPACE

cache/cache_entry_stats.h

@@ -10,8 +10,8 @@
 #include <memory>
 #include <mutex>
 
+#include "cache/cache_helpers.h"
 #include "cache/cache_key.h"
-#include "cache/typed_cache.h"
 #include "port/lang.h"
 #include "rocksdb/cache.h"
 #include "rocksdb/status.h"
@@ -111,14 +111,11 @@ class CacheEntryStatsCollector {
   // Gets or creates a shared instance of CacheEntryStatsCollector in the
   // cache itself, and saves into `ptr`. This shared_ptr will hold the
   // entry in cache until all refs are destroyed.
-  static Status GetShared(Cache *raw_cache, SystemClock *clock,
+  static Status GetShared(Cache *cache, SystemClock *clock,
                           std::shared_ptr<CacheEntryStatsCollector> *ptr) {
-    assert(raw_cache);
-    BasicTypedCacheInterface<CacheEntryStatsCollector, CacheEntryRole::kMisc>
-        cache{raw_cache};
-
     const Slice &cache_key = GetCacheKey();
-    auto h = cache.Lookup(cache_key);
+
+    Cache::Handle *h = cache->Lookup(cache_key);
     if (h == nullptr) {
       // Not yet in cache, but Cache doesn't provide a built-in way to
       // avoid racing insert. So we double-check under a shared mutex,
@@ -126,15 +123,15 @@ class CacheEntryStatsCollector {
       STATIC_AVOID_DESTRUCTION(std::mutex, static_mutex);
       std::lock_guard<std::mutex> lock(static_mutex);
 
-      h = cache.Lookup(cache_key);
+      h = cache->Lookup(cache_key);
       if (h == nullptr) {
-        auto new_ptr = new CacheEntryStatsCollector(cache.get(), clock);
+        auto new_ptr = new CacheEntryStatsCollector(cache, clock);
         // TODO: non-zero charge causes some tests that count block cache
         // usage to go flaky. Fix the problem somehow so we can use an
         // accurate charge.
         size_t charge = 0;
-        Status s =
-            cache.Insert(cache_key, new_ptr, charge, &h, Cache::Priority::HIGH);
+        Status s = cache->Insert(cache_key, new_ptr, charge, Deleter, &h,
+                                 Cache::Priority::HIGH);
         if (!s.ok()) {
           assert(h == nullptr);
           delete new_ptr;
@@ -143,11 +140,11 @@ class CacheEntryStatsCollector {
       }
     }
     // If we reach here, shared entry is in cache with handle `h`.
-    assert(cache.get()->GetCacheItemHelper(h) == cache.GetBasicHelper());
+    assert(cache->GetDeleter(h) == Deleter);
 
     // Build an aliasing shared_ptr that keeps `ptr` in cache while there
     // are references.
-    *ptr = cache.SharedGuard(h);
+    *ptr = MakeSharedCacheHandleGuard<CacheEntryStatsCollector>(cache, h);
     return Status::OK();
   }
 
@@ -160,6 +157,10 @@ class CacheEntryStatsCollector {
         cache_(cache),
         clock_(clock) {}
 
+  static void Deleter(const Slice &, void *value) {
+    delete static_cast<CacheEntryStatsCollector *>(value);
+  }
+
   static const Slice &GetCacheKey() {
     // For each template instantiation
     static CacheKey ckey = CacheKey::CreateUniqueForProcessLifetime();

cache/cache_helpers.cc

@@ -1,40 +0,0 @@
-//  Copyright (c) Meta Platforms, Inc. and affiliates.
-//  This source code is licensed under both the GPLv2 (found in the
-//  COPYING file in the root directory) and Apache 2.0 License
-//  (found in the LICENSE.Apache file in the root directory).
-
-#include "cache/cache_helpers.h"
-
-namespace ROCKSDB_NAMESPACE {
-
-void ReleaseCacheHandleCleanup(void* arg1, void* arg2) {
-  Cache* const cache = static_cast<Cache*>(arg1);
-  assert(cache);
-
-  Cache::Handle* const cache_handle = static_cast<Cache::Handle*>(arg2);
-  assert(cache_handle);
-
-  cache->Release(cache_handle);
-}
-
-Status WarmInCache(Cache* cache, const Slice& key, const Slice& saved,
-                   Cache::CreateContext* create_context,
-                   const Cache::CacheItemHelper* helper,
-                   Cache::Priority priority, size_t* out_charge) {
-  assert(helper);
-  assert(helper->create_cb);
-  Cache::ObjectPtr value;
-  size_t charge;
-  Status st = helper->create_cb(saved, create_context,
-                                cache->memory_allocator(), &value, &charge);
-  if (st.ok()) {
-    st =
-        cache->Insert(key, value, helper, charge, /*handle*/ nullptr, priority);
-    if (out_charge) {
-      *out_charge = charge;
-    }
-  }
-  return st;
-}
-
-}  // namespace ROCKSDB_NAMESPACE

cache/cache_helpers.h

@@ -7,7 +7,7 @@
 
 #include <cassert>
 
-#include "rocksdb/advanced_cache.h"
+#include "rocksdb/cache.h"
 #include "rocksdb/rocksdb_namespace.h"
 
 namespace ROCKSDB_NAMESPACE {
@@ -17,17 +17,22 @@ template <typename T>
 T* GetFromCacheHandle(Cache* cache, Cache::Handle* handle) {
   assert(cache);
   assert(handle);
+
   return static_cast<T*>(cache->Value(handle));
 }
 
+// Simple generic deleter for Cache (to be used with Cache::Insert).
+template <typename T>
+void DeleteCacheEntry(const Slice& /* key */, void* value) {
+  delete static_cast<T*>(value);
+}
+
 // Turns a T* into a Slice so it can be used as a key with Cache.
 template <typename T>
-Slice GetSliceForKey(const T* t) {
+Slice GetSlice(const T* t) {
   return Slice(reinterpret_cast<const char*>(t), sizeof(T));
 }
 
-void ReleaseCacheHandleCleanup(void* arg1, void* arg2);
-
 // Generic resource management object for cache handles that releases the handle
 // when destroyed. Has unique ownership of the handle, so copying it is not
 // allowed, while moving it transfers ownership.
@@ -83,7 +88,7 @@ class CacheHandleGuard {
     if (cleanable) {
       if (handle_ != nullptr) {
         assert(cache_);
-        cleanable->RegisterCleanup(&ReleaseCacheHandleCleanup, cache_, handle_);
+        cleanable->RegisterCleanup(&ReleaseCacheHandle, cache_, handle_);
       }
     }
     ResetFields();
@@ -110,6 +115,16 @@ class CacheHandleGuard {
     value_ = nullptr;
   }
 
+  static void ReleaseCacheHandle(void* arg1, void* arg2) {
+    Cache* const cache = static_cast<Cache*>(arg1);
+    assert(cache);
+
+    Cache::Handle* const cache_handle = static_cast<Cache::Handle*>(arg2);
+    assert(cache_handle);
+
+    cache->Release(cache_handle);
+  }
+
  private:
   Cache* cache_ = nullptr;
   Cache::Handle* handle_ = nullptr;
@@ -124,16 +139,7 @@ template <typename T>
 std::shared_ptr<T> MakeSharedCacheHandleGuard(Cache* cache,
                                               Cache::Handle* handle) {
   auto wrapper = std::make_shared<CacheHandleGuard<T>>(cache, handle);
-  return std::shared_ptr<T>(wrapper, GetFromCacheHandle<T>(cache, handle));
+  return std::shared_ptr<T>(wrapper, static_cast<T*>(cache->Value(handle)));
 }
 
-// Given the persistable data (saved) for a block cache entry, parse that
-// into a cache entry object and insert it into the given cache. The charge
-// of the new entry can be returned to the caller through `out_charge`.
-Status WarmInCache(Cache* cache, const Slice& key, const Slice& saved,
-                   Cache::CreateContext* create_context,
-                   const Cache::CacheItemHelper* helper,
-                   Cache::Priority priority = Cache::Priority::LOW,
-                   size_t* out_charge = nullptr);
-
 }  // namespace ROCKSDB_NAMESPACE

cache/cache_key.cc

@@ -8,7 +8,7 @@
 #include <algorithm>
 #include <atomic>
 
-#include "rocksdb/advanced_cache.h"
+#include "rocksdb/cache.h"
 #include "table/unique_id_impl.h"
 #include "util/hash.h"
 #include "util/math.h"

cache/cache_reservation_manager.cc

@@ -13,6 +13,7 @@
 #include <cstring>
 #include <memory>
 
+#include "cache/cache_entry_roles.h"
 #include "rocksdb/cache.h"
 #include "rocksdb/slice.h"
 #include "rocksdb/status.h"
@@ -40,17 +41,17 @@ CacheReservationManagerImpl<
 template <CacheEntryRole R>
 CacheReservationManagerImpl<R>::CacheReservationManagerImpl(
     std::shared_ptr<Cache> cache, bool delayed_decrease)
-    : cache_(cache),
-      delayed_decrease_(delayed_decrease),
+    : delayed_decrease_(delayed_decrease),
       cache_allocated_size_(0),
       memory_used_(0) {
   assert(cache != nullptr);
+  cache_ = cache;
 }
 
 template <CacheEntryRole R>
 CacheReservationManagerImpl<R>::~CacheReservationManagerImpl() {
   for (auto* handle : dummy_handles_) {
-    cache_.ReleaseAndEraseIfLastRef(handle);
+    cache_->Release(handle, true);
   }
 }
 
@@ -114,7 +115,8 @@ Status CacheReservationManagerImpl<R>::IncreaseCacheReservation(
   Status return_status = Status::OK();
   while (new_mem_used > cache_allocated_size_.load(std::memory_order_relaxed)) {
     Cache::Handle* handle = nullptr;
-    return_status = cache_.Insert(GetNextCacheKey(), kSizeDummyEntry, &handle);
+    return_status = cache_->Insert(GetNextCacheKey(), nullptr, kSizeDummyEntry,
+                                   GetNoopDeleterForRole<R>(), &handle);
 
     if (return_status != Status::OK()) {
       return return_status;
@@ -139,7 +141,7 @@ Status CacheReservationManagerImpl<R>::DecreaseCacheReservation(
          cache_allocated_size_.load(std::memory_order_relaxed)) {
     assert(!dummy_handles_.empty());
     auto* handle = dummy_handles_.back();
-    cache_.ReleaseAndEraseIfLastRef(handle);
+    cache_->Release(handle, true);
     dummy_handles_.pop_back();
     cache_allocated_size_ -= kSizeDummyEntry;
   }
@@ -167,9 +169,8 @@ Slice CacheReservationManagerImpl<R>::GetNextCacheKey() {
 }
 
 template <CacheEntryRole R>
-const Cache::CacheItemHelper*
-CacheReservationManagerImpl<R>::TEST_GetCacheItemHelperForRole() {
-  return CacheInterface::GetHelper();
+Cache::DeleterFn CacheReservationManagerImpl<R>::TEST_GetNoopDeleterForRole() {
+  return GetNoopDeleterForRole<R>();
 }
 
 template class CacheReservationManagerImpl<

cache/cache_reservation_manager.h

@@ -18,7 +18,7 @@
 
 #include "cache/cache_entry_roles.h"
 #include "cache/cache_key.h"
-#include "cache/typed_cache.h"
+#include "rocksdb/cache.h"
 #include "rocksdb/slice.h"
 #include "rocksdb/status.h"
 #include "util/coding.h"
@@ -197,10 +197,10 @@ class CacheReservationManagerImpl
 
   static constexpr std::size_t GetDummyEntrySize() { return kSizeDummyEntry; }
 
-  // For testing only - it is to help ensure the CacheItemHelperForRole<R>
+  // For testing only - it is to help ensure the NoopDeleterForRole<R>
   // accessed from CacheReservationManagerImpl and the one accessed from the
   // test are from the same translation units
-  static const Cache::CacheItemHelper *TEST_GetCacheItemHelperForRole();
+  static Cache::DeleterFn TEST_GetNoopDeleterForRole();
 
  private:
   static constexpr std::size_t kSizeDummyEntry = 256 * 1024;
@@ -211,8 +211,7 @@ class CacheReservationManagerImpl
   Status IncreaseCacheReservation(std::size_t new_mem_used);
   Status DecreaseCacheReservation(std::size_t new_mem_used);
 
-  using CacheInterface = PlaceholderSharedCacheInterface<R>;
-  CacheInterface cache_;
+  std::shared_ptr<Cache> cache_;
   bool delayed_decrease_;
   std::atomic<std::size_t> cache_allocated_size_;
   std::size_t memory_used_;

cache/cache_reservation_manager_test.cc

@@ -463,7 +463,6 @@ TEST(CacheReservationHandleTest, HandleTest) {
 }  // namespace ROCKSDB_NAMESPACE
 
 int main(int argc, char** argv) {
-  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
   ::testing::InitGoogleTest(&argc, argv);
   return RUN_ALL_TESTS();
 }

cache/cache_test.cc

@@ -15,16 +15,15 @@
 #include <string>
 #include <vector>
 
+#include "cache/fast_lru_cache.h"
 #include "cache/lru_cache.h"
-#include "cache/typed_cache.h"
 #include "port/stack_trace.h"
-#include "test_util/secondary_cache_test_util.h"
 #include "test_util/testharness.h"
 #include "util/coding.h"
 #include "util/string_util.h"
 
-// HyperClockCache only supports 16-byte keys, so some of the tests
-// originally written for LRUCache do not work on the other caches.
+// FastLRUCache and HyperClockCache only support 16-byte keys, so some of
+// the tests originally wrote for LRUCache do not work on the other caches.
 // Those tests were adapted to use 16-byte keys. We kept the original ones.
 // TODO: Remove the original tests if they ever become unused.
 
@@ -57,43 +56,43 @@ int DecodeKey32Bits(const Slice& k) {
   return DecodeFixed32(k.data());
 }
 
-Cache::ObjectPtr EncodeValue(uintptr_t v) {
-  return reinterpret_cast<Cache::ObjectPtr>(v);
-}
+void* EncodeValue(uintptr_t v) { return reinterpret_cast<void*>(v); }
 
 int DecodeValue(void* v) {
   return static_cast<int>(reinterpret_cast<uintptr_t>(v));
 }
 
-const Cache::CacheItemHelper kDumbHelper{
-    CacheEntryRole::kMisc,
-    [](Cache::ObjectPtr /*value*/, MemoryAllocator* /*alloc*/) {}};
+void DumbDeleter(const Slice& /*key*/, void* /*value*/) {}
 
-const Cache::CacheItemHelper kEraseOnDeleteHelper1{
-    CacheEntryRole::kMisc,
-    [](Cache::ObjectPtr value, MemoryAllocator* /*alloc*/) {
-      Cache* cache = static_cast<Cache*>(value);
-      cache->Erase("foo");
-    }};
+void EraseDeleter1(const Slice& /*key*/, void* value) {
+  Cache* cache = reinterpret_cast<Cache*>(value);
+  cache->Erase("foo");
+}
+
+void EraseDeleter2(const Slice& /*key*/, void* value) {
+  Cache* cache = reinterpret_cast<Cache*>(value);
+  cache->Erase(EncodeKey16Bytes(1234));
+}
+
+const std::string kLRU = "lru";
+const std::string kHyperClock = "hyper_clock";
+const std::string kFast = "fast";
 
-const Cache::CacheItemHelper kEraseOnDeleteHelper2{
-    CacheEntryRole::kMisc,
-    [](Cache::ObjectPtr value, MemoryAllocator* /*alloc*/) {
-      Cache* cache = static_cast<Cache*>(value);
-      cache->Erase(EncodeKey16Bytes(1234));
-    }};
 }  // anonymous namespace
 
-class CacheTest : public testing::Test,
-                  public secondary_cache_test_util::WithCacheTypeParam {
+class CacheTest : public testing::TestWithParam<std::string> {
  public:
   static CacheTest* current_;
   static std::string type_;
 
-  static void Deleter(Cache::ObjectPtr v, MemoryAllocator*) {
+  static void Deleter(const Slice& key, void* v) {
+    if (type_ == kFast || type_ == kHyperClock) {
+      current_->deleted_keys_.push_back(DecodeKey16Bytes(key));
+    } else {
+      current_->deleted_keys_.push_back(DecodeKey32Bits(key));
+    }
     current_->deleted_values_.push_back(DecodeValue(v));
   }
-  static const Cache::CacheItemHelper kHelper;
 
   static const int kCacheSize = 1000;
   static const int kNumShardBits = 4;
@@ -101,10 +100,13 @@ class CacheTest : public testing::Test,
   static const int kCacheSize2 = 100;
   static const int kNumShardBits2 = 2;
 
+  std::vector<int> deleted_keys_;
   std::vector<int> deleted_values_;
   std::shared_ptr<Cache> cache_;
   std::shared_ptr<Cache> cache2_;
 
+  size_t estimated_value_size_ = 1;
+
   CacheTest()
       : cache_(NewCache(kCacheSize, kNumShardBits, false)),
         cache2_(NewCache(kCacheSize2, kNumShardBits2, false)) {
@@ -114,13 +116,58 @@ class CacheTest : public testing::Test,
 
   ~CacheTest() override {}
 
+  std::shared_ptr<Cache> NewCache(size_t capacity) {
+    auto type = GetParam();
+    if (type == kLRU) {
+      return NewLRUCache(capacity);
+    }
+    if (type == kHyperClock) {
+      return HyperClockCacheOptions(
+                 capacity, estimated_value_size_ /*estimated_value_size*/)
+          .MakeSharedCache();
+    }
+    if (type == kFast) {
+      return NewFastLRUCache(
+          capacity, estimated_value_size_, -1 /*num_shard_bits*/,
+          false /*strict_capacity_limit*/, kDefaultCacheMetadataChargePolicy);
+    }
+    return nullptr;
+  }
+
+  std::shared_ptr<Cache> NewCache(
+      size_t capacity, int num_shard_bits, bool strict_capacity_limit,
+      CacheMetadataChargePolicy charge_policy = kDontChargeCacheMetadata) {
+    auto type = GetParam();
+    if (type == kLRU) {
+      LRUCacheOptions co;
+      co.capacity = capacity;
+      co.num_shard_bits = num_shard_bits;
+      co.strict_capacity_limit = strict_capacity_limit;
+      co.high_pri_pool_ratio = 0;
+      co.metadata_charge_policy = charge_policy;
+      return NewLRUCache(co);
+    }
+    if (type == kHyperClock) {
+      return HyperClockCacheOptions(capacity, 1 /*estimated_value_size*/,
+                                    num_shard_bits, strict_capacity_limit,
+                                    nullptr /*allocator*/, charge_policy)
+          .MakeSharedCache();
+    }
+    if (type == kFast) {
+      return NewFastLRUCache(capacity, 1 /*estimated_value_size*/,
+                             num_shard_bits, strict_capacity_limit,
+                             charge_policy);
+    }
+    return nullptr;
+  }
+
   // These functions encode/decode keys in tests cases that use
   // int keys.
   // Currently, HyperClockCache requires keys to be 16B long, whereas
   // LRUCache doesn't, so the encoding depends on the cache type.
   std::string EncodeKey(int k) {
     auto type = GetParam();
-    if (type == kHyperClock) {
+    if (type == kFast || type == kHyperClock) {
       return EncodeKey16Bytes(k);
     } else {
       return EncodeKey32Bits(k);
@@ -129,7 +176,7 @@ class CacheTest : public testing::Test,
 
   int DecodeKey(const Slice& k) {
     auto type = GetParam();
-    if (type == kHyperClock) {
+    if (type == kFast || type == kHyperClock) {
       return DecodeKey16Bytes(k);
     } else {
       return DecodeKey32Bits(k);
@@ -147,34 +194,39 @@ class CacheTest : public testing::Test,
 
   void Insert(std::shared_ptr<Cache> cache, int key, int value,
               int charge = 1) {
-    EXPECT_OK(cache->Insert(EncodeKey(key), EncodeValue(value), &kHelper,
-                            charge, /*handle*/ nullptr, Cache::Priority::HIGH));
+    EXPECT_OK(cache->Insert(EncodeKey(key), EncodeValue(value), charge,
+                            &CacheTest::Deleter));
   }
 
   void Erase(std::shared_ptr<Cache> cache, int key) {
     cache->Erase(EncodeKey(key));
   }
 
-  int Lookup(int key) { return Lookup(cache_, key); }
+  int Lookup(int key) {
+    return Lookup(cache_, key);
+  }
 
   void Insert(int key, int value, int charge = 1) {
     Insert(cache_, key, value, charge);
   }
 
-  void Erase(int key) { Erase(cache_, key); }
+  void Erase(int key) {
+    Erase(cache_, key);
+  }
 
-  int Lookup2(int key) { return Lookup(cache2_, key); }
+  int Lookup2(int key) {
+    return Lookup(cache2_, key);
+  }
 
   void Insert2(int key, int value, int charge = 1) {
     Insert(cache2_, key, value, charge);
   }
 
-  void Erase2(int key) { Erase(cache2_, key); }
+  void Erase2(int key) {
+    Erase(cache2_, key);
+  }
 };
 
-const Cache::CacheItemHelper CacheTest::kHelper{CacheEntryRole::kMisc,
-                                                &CacheTest::Deleter};
-
 CacheTest* CacheTest::current_;
 std::string CacheTest::type_;
 
@@ -204,8 +256,10 @@ TEST_P(CacheTest, UsageTest) {
       key = EncodeKey(i);
     }
     auto kv_size = key.size() + 5;
-    ASSERT_OK(cache->Insert(key, value, &kDumbHelper, kv_size));
-    ASSERT_OK(precise_cache->Insert(key, value, &kDumbHelper, kv_size));
+    ASSERT_OK(cache->Insert(key, reinterpret_cast<void*>(value), kv_size,
+                            DumbDeleter));
+    ASSERT_OK(precise_cache->Insert(key, reinterpret_cast<void*>(value),
+                                    kv_size, DumbDeleter));
     usage += kv_size;
     ASSERT_EQ(usage, cache->GetUsage());
     if (type == kHyperClock) {
@@ -228,8 +282,10 @@ TEST_P(CacheTest, UsageTest) {
     } else {
       key = EncodeKey(static_cast<int>(1000 + i));
     }
-    ASSERT_OK(cache->Insert(key, value, &kDumbHelper, key.size() + 5));
-    ASSERT_OK(precise_cache->Insert(key, value, &kDumbHelper, key.size() + 5));
+    ASSERT_OK(cache->Insert(key, reinterpret_cast<void*>(value), key.size() + 5,
+                            DumbDeleter));
+    ASSERT_OK(precise_cache->Insert(key, reinterpret_cast<void*>(value),
+                                    key.size() + 5, DumbDeleter));
   }
 
   // the usage should be close to the capacity
@@ -284,9 +340,11 @@ TEST_P(CacheTest, PinnedUsageTest) {
     auto kv_size = key.size() + 5;
     Cache::Handle* handle;
     Cache::Handle* handle_in_precise_cache;
-    ASSERT_OK(cache->Insert(key, value, &kDumbHelper, kv_size, &handle));
+    ASSERT_OK(cache->Insert(key, reinterpret_cast<void*>(value), kv_size,
+                            DumbDeleter, &handle));
     assert(handle);
-    ASSERT_OK(precise_cache->Insert(key, value, &kDumbHelper, kv_size,
+    ASSERT_OK(precise_cache->Insert(key, reinterpret_cast<void*>(value),
+                                    kv_size, DumbDeleter,
                                     &handle_in_precise_cache));
     assert(handle_in_precise_cache);
     pinned_usage += kv_size;
@@ -327,8 +385,10 @@ TEST_P(CacheTest, PinnedUsageTest) {
     } else {
       key = EncodeKey(static_cast<int>(1000 + i));
     }
-    ASSERT_OK(cache->Insert(key, value, &kDumbHelper, key.size() + 5));
-    ASSERT_OK(precise_cache->Insert(key, value, &kDumbHelper, key.size() + 5));
+    ASSERT_OK(cache->Insert(key, reinterpret_cast<void*>(value), key.size() + 5,
+                            DumbDeleter));
+    ASSERT_OK(precise_cache->Insert(key, reinterpret_cast<void*>(value),
+                                    key.size() + 5, DumbDeleter));
   }
   ASSERT_EQ(pinned_usage, cache->GetPinnedUsage());
   ASSERT_EQ(precise_cache_pinned_usage, precise_cache->GetPinnedUsage());
@@ -358,13 +418,13 @@ TEST_P(CacheTest, HitAndMiss) {
 
   Insert(100, 101);
   ASSERT_EQ(101, Lookup(100));
-  ASSERT_EQ(-1, Lookup(200));
-  ASSERT_EQ(-1, Lookup(300));
+  ASSERT_EQ(-1,  Lookup(200));
+  ASSERT_EQ(-1,  Lookup(300));
 
   Insert(200, 201);
   ASSERT_EQ(101, Lookup(100));
   ASSERT_EQ(201, Lookup(200));
-  ASSERT_EQ(-1, Lookup(300));
+  ASSERT_EQ(-1,  Lookup(300));
 
   Insert(100, 102);
   if (GetParam() == kHyperClock) {
@@ -374,9 +434,10 @@ TEST_P(CacheTest, HitAndMiss) {
     ASSERT_EQ(102, Lookup(100));
   }
   ASSERT_EQ(201, Lookup(200));
-  ASSERT_EQ(-1, Lookup(300));
+  ASSERT_EQ(-1,  Lookup(300));
 
-  ASSERT_EQ(1U, deleted_values_.size());
+  ASSERT_EQ(1U, deleted_keys_.size());
+  ASSERT_EQ(100, deleted_keys_[0]);
   if (GetParam() == kHyperClock) {
     ASSERT_EQ(102, deleted_values_[0]);
   } else {
@@ -397,20 +458,21 @@ TEST_P(CacheTest, InsertSameKey) {
 
 TEST_P(CacheTest, Erase) {
   Erase(200);
-  ASSERT_EQ(0U, deleted_values_.size());
+  ASSERT_EQ(0U, deleted_keys_.size());
 
   Insert(100, 101);
   Insert(200, 201);
   Erase(100);
-  ASSERT_EQ(-1, Lookup(100));
+  ASSERT_EQ(-1,  Lookup(100));
   ASSERT_EQ(201, Lookup(200));
-  ASSERT_EQ(1U, deleted_values_.size());
+  ASSERT_EQ(1U, deleted_keys_.size());
+  ASSERT_EQ(100, deleted_keys_[0]);
   ASSERT_EQ(101, deleted_values_[0]);
 
   Erase(100);
-  ASSERT_EQ(-1, Lookup(100));
+  ASSERT_EQ(-1,  Lookup(100));
   ASSERT_EQ(201, Lookup(200));
-  ASSERT_EQ(1U, deleted_values_.size());
+  ASSERT_EQ(1U, deleted_keys_.size());
 }
 
 TEST_P(CacheTest, EntriesArePinned) {
@@ -427,21 +489,23 @@ TEST_P(CacheTest, EntriesArePinned) {
   Insert(100, 102);
   Cache::Handle* h2 = cache_->Lookup(EncodeKey(100));
   ASSERT_EQ(102, DecodeValue(cache_->Value(h2)));
-  ASSERT_EQ(0U, deleted_values_.size());
+  ASSERT_EQ(0U, deleted_keys_.size());
   ASSERT_EQ(2U, cache_->GetUsage());
 
   cache_->Release(h1);
-  ASSERT_EQ(1U, deleted_values_.size());
+  ASSERT_EQ(1U, deleted_keys_.size());
+  ASSERT_EQ(100, deleted_keys_[0]);
   ASSERT_EQ(101, deleted_values_[0]);
   ASSERT_EQ(1U, cache_->GetUsage());
 
   Erase(100);
   ASSERT_EQ(-1, Lookup(100));
-  ASSERT_EQ(1U, deleted_values_.size());
+  ASSERT_EQ(1U, deleted_keys_.size());
   ASSERT_EQ(1U, cache_->GetUsage());
 
   cache_->Release(h2);
-  ASSERT_EQ(2U, deleted_values_.size());
+  ASSERT_EQ(2U, deleted_keys_.size());
+  ASSERT_EQ(100, deleted_keys_[1]);
   ASSERT_EQ(102, deleted_values_[1]);
   ASSERT_EQ(0U, cache_->GetUsage());
 }
@@ -451,7 +515,7 @@ TEST_P(CacheTest, EvictionPolicy) {
   Insert(200, 201);
   // Frequently used entry must be kept around
   for (int i = 0; i < 2 * kCacheSize; i++) {
-    Insert(1000 + i, 2000 + i);
+    Insert(1000+i, 2000+i);
     ASSERT_EQ(101, Lookup(100));
   }
   ASSERT_EQ(101, Lookup(100));
@@ -544,9 +608,9 @@ TEST_P(CacheTest, EvictEmptyCache) {
   // Insert item large than capacity to trigger eviction on empty cache.
   auto cache = NewCache(1, 0, false);
   if (type == kLRU) {
-    ASSERT_OK(cache->Insert("foo", nullptr, &kDumbHelper, 10));
+    ASSERT_OK(cache->Insert("foo", nullptr, 10, DumbDeleter));
   } else {
-    ASSERT_OK(cache->Insert(EncodeKey(1000), nullptr, &kDumbHelper, 10));
+    ASSERT_OK(cache->Insert(EncodeKey(1000), nullptr, 10, DumbDeleter));
   }
 }
 
@@ -557,19 +621,19 @@ TEST_P(CacheTest, EraseFromDeleter) {
   // the cache at that point.
   std::shared_ptr<Cache> cache = NewCache(10, 0, false);
   std::string foo, bar;
-  const Cache::CacheItemHelper* erase_helper;
+  Cache::DeleterFn erase_deleter;
   if (type == kLRU) {
     foo = "foo";
     bar = "bar";
-    erase_helper = &kEraseOnDeleteHelper1;
+    erase_deleter = EraseDeleter1;
   } else {
     foo = EncodeKey(1234);
     bar = EncodeKey(5678);
-    erase_helper = &kEraseOnDeleteHelper2;
+    erase_deleter = EraseDeleter2;
   }
 
-  ASSERT_OK(cache->Insert(foo, nullptr, &kDumbHelper, 1));
-  ASSERT_OK(cache->Insert(bar, cache.get(), erase_helper, 1));
+  ASSERT_OK(cache->Insert(foo, nullptr, 1, DumbDeleter));
+  ASSERT_OK(cache->Insert(bar, cache.get(), 1, erase_deleter));
 
   cache->Erase(bar);
   ASSERT_EQ(nullptr, cache->Lookup(foo));
@@ -622,7 +686,7 @@ TEST_P(CacheTest, HeavyEntries) {
       ASSERT_EQ(1000 + i, r);
     }
   }
-  ASSERT_LE(cached_weight, kCacheSize + kCacheSize / 10);
+  ASSERT_LE(cached_weight, kCacheSize + kCacheSize/10);
 }
 
 TEST_P(CacheTest, NewId) {
@@ -631,54 +695,53 @@ TEST_P(CacheTest, NewId) {
   ASSERT_NE(a, b);
 }
 
-TEST_P(CacheTest, ReleaseAndErase) {
-  std::shared_ptr<Cache> cache = NewCache(5, 0, false);
-  Cache::Handle* handle;
-  Status s =
-      cache->Insert(EncodeKey(100), EncodeValue(100), &kHelper, 1, &handle);
-  ASSERT_TRUE(s.ok());
-  ASSERT_EQ(5U, cache->GetCapacity());
-  ASSERT_EQ(1U, cache->GetUsage());
-  ASSERT_EQ(0U, deleted_values_.size());
-  auto erased = cache->Release(handle, true);
-  ASSERT_TRUE(erased);
-  // This tests that deleter has been called
-  ASSERT_EQ(1U, deleted_values_.size());
-}
-
-TEST_P(CacheTest, ReleaseWithoutErase) {
-  std::shared_ptr<Cache> cache = NewCache(5, 0, false);
-  Cache::Handle* handle;
-  Status s =
-      cache->Insert(EncodeKey(100), EncodeValue(100), &kHelper, 1, &handle);
-  ASSERT_TRUE(s.ok());
-  ASSERT_EQ(5U, cache->GetCapacity());
-  ASSERT_EQ(1U, cache->GetUsage());
-  ASSERT_EQ(0U, deleted_values_.size());
-  auto erased = cache->Release(handle);
-  ASSERT_FALSE(erased);
-  // This tests that deleter is not called. When cache has free capacity it is
-  // not expected to immediately erase the released items.
-  ASSERT_EQ(0U, deleted_values_.size());
-}
-
-namespace {
 class Value {
  public:
   explicit Value(int v) : v_(v) {}
 
   int v_;
-
-  static constexpr auto kCacheEntryRole = CacheEntryRole::kMisc;
 };
 
-using SharedCache = BasicTypedSharedCacheInterface<Value>;
-using TypedHandle = SharedCache::TypedHandle;
+namespace {
+void deleter(const Slice& /*key*/, void* value) {
+  delete static_cast<Value *>(value);
+}
 }  // namespace
 
+TEST_P(CacheTest, ReleaseAndErase) {
+  std::shared_ptr<Cache> cache = NewCache(5, 0, false);
+  Cache::Handle* handle;
+  Status s = cache->Insert(EncodeKey(100), EncodeValue(100), 1,
+                           &CacheTest::Deleter, &handle);
+  ASSERT_TRUE(s.ok());
+  ASSERT_EQ(5U, cache->GetCapacity());
+  ASSERT_EQ(1U, cache->GetUsage());
+  ASSERT_EQ(0U, deleted_keys_.size());
+  auto erased = cache->Release(handle, true);
+  ASSERT_TRUE(erased);
+  // This tests that deleter has been called
+  ASSERT_EQ(1U, deleted_keys_.size());
+}
+
+TEST_P(CacheTest, ReleaseWithoutErase) {
+  std::shared_ptr<Cache> cache = NewCache(5, 0, false);
+  Cache::Handle* handle;
+  Status s = cache->Insert(EncodeKey(100), EncodeValue(100), 1,
+                           &CacheTest::Deleter, &handle);
+  ASSERT_TRUE(s.ok());
+  ASSERT_EQ(5U, cache->GetCapacity());
+  ASSERT_EQ(1U, cache->GetUsage());
+  ASSERT_EQ(0U, deleted_keys_.size());
+  auto erased = cache->Release(handle);
+  ASSERT_FALSE(erased);
+  // This tests that deleter is not called. When cache has free capacity it is
+  // not expected to immediately erase the released items.
+  ASSERT_EQ(0U, deleted_keys_.size());
+}
+
 TEST_P(CacheTest, SetCapacity) {
   auto type = GetParam();
-  if (type == kHyperClock) {
+  if (type == kFast || type == kHyperClock) {
     ROCKSDB_GTEST_BYPASS(
         "FastLRUCache and HyperClockCache don't support arbitrary capacity "
         "adjustments.");
@@ -688,19 +751,19 @@ TEST_P(CacheTest, SetCapacity) {
   // lets create a cache with capacity 5,
   // then, insert 5 elements, then increase capacity
   // to 10, returned capacity should be 10, usage=5
-  SharedCache cache{NewCache(5, 0, false)};
-  std::vector<TypedHandle*> handles(10);
+  std::shared_ptr<Cache> cache = NewCache(5, 0, false);
+  std::vector<Cache::Handle*> handles(10);
   // Insert 5 entries, but not releasing.
   for (int i = 0; i < 5; i++) {
     std::string key = EncodeKey(i + 1);
-    Status s = cache.Insert(key, new Value(i + 1), 1, &handles[i]);
+    Status s = cache->Insert(key, new Value(i + 1), 1, &deleter, &handles[i]);
     ASSERT_TRUE(s.ok());
   }
-  ASSERT_EQ(5U, cache.get()->GetCapacity());
-  ASSERT_EQ(5U, cache.get()->GetUsage());
-  cache.get()->SetCapacity(10);
-  ASSERT_EQ(10U, cache.get()->GetCapacity());
-  ASSERT_EQ(5U, cache.get()->GetUsage());
+  ASSERT_EQ(5U, cache->GetCapacity());
+  ASSERT_EQ(5U, cache->GetUsage());
+  cache->SetCapacity(10);
+  ASSERT_EQ(10U, cache->GetCapacity());
+  ASSERT_EQ(5U, cache->GetUsage());
 
   // test2: decrease capacity
   // insert 5 more elements to cache, then release 5,
@@ -708,77 +771,85 @@ TEST_P(CacheTest, SetCapacity) {
   // and usage should be 7
   for (int i = 5; i < 10; i++) {
     std::string key = EncodeKey(i + 1);
-    Status s = cache.Insert(key, new Value(i + 1), 1, &handles[i]);
+    Status s = cache->Insert(key, new Value(i + 1), 1, &deleter, &handles[i]);
     ASSERT_TRUE(s.ok());
   }
-  ASSERT_EQ(10U, cache.get()->GetCapacity());
-  ASSERT_EQ(10U, cache.get()->GetUsage());
+  ASSERT_EQ(10U, cache->GetCapacity());
+  ASSERT_EQ(10U, cache->GetUsage());
   for (int i = 0; i < 5; i++) {
-    cache.Release(handles[i]);
+    cache->Release(handles[i]);
   }
-  ASSERT_EQ(10U, cache.get()->GetCapacity());
-  ASSERT_EQ(10U, cache.get()->GetUsage());
-  cache.get()->SetCapacity(7);
-  ASSERT_EQ(7, cache.get()->GetCapacity());
-  ASSERT_EQ(7, cache.get()->GetUsage());
+  ASSERT_EQ(10U, cache->GetCapacity());
+  ASSERT_EQ(10U, cache->GetUsage());
+  cache->SetCapacity(7);
+  ASSERT_EQ(7, cache->GetCapacity());
+  ASSERT_EQ(7, cache->GetUsage());
 
   // release remaining 5 to keep valgrind happy
   for (int i = 5; i < 10; i++) {
-    cache.Release(handles[i]);
+    cache->Release(handles[i]);
   }
 
   // Make sure this doesn't crash or upset ASAN/valgrind
-  cache.get()->DisownData();
+  cache->DisownData();
 }
 
 TEST_P(LRUCacheTest, SetStrictCapacityLimit) {
+  auto type = GetParam();
+  if (type == kFast) {
+    ROCKSDB_GTEST_BYPASS(
+        "FastLRUCache only supports a limited number of "
+        "inserts beyond "
+        "capacity.");
+    return;
+  }
   // test1: set the flag to false. Insert more keys than capacity. See if they
   // all go through.
-  SharedCache cache{NewCache(5, 0, false)};
-  std::vector<TypedHandle*> handles(10);
+  std::shared_ptr<Cache> cache = NewCache(5, 0, false);
+  std::vector<Cache::Handle*> handles(10);
   Status s;
   for (int i = 0; i < 10; i++) {
     std::string key = EncodeKey(i + 1);
-    s = cache.Insert(key, new Value(i + 1), 1, &handles[i]);
+    s = cache->Insert(key, new Value(i + 1), 1, &deleter, &handles[i]);
     ASSERT_OK(s);
     ASSERT_NE(nullptr, handles[i]);
   }
-  ASSERT_EQ(10, cache.get()->GetUsage());
+  ASSERT_EQ(10, cache->GetUsage());
 
   // test2: set the flag to true. Insert and check if it fails.
   std::string extra_key = EncodeKey(100);
   Value* extra_value = new Value(0);
-  cache.get()->SetStrictCapacityLimit(true);
-  TypedHandle* handle;
-  s = cache.Insert(extra_key, extra_value, 1, &handle);
+  cache->SetStrictCapacityLimit(true);
+  Cache::Handle* handle;
+  s = cache->Insert(extra_key, extra_value, 1, &deleter, &handle);
   ASSERT_TRUE(s.IsMemoryLimit());
   ASSERT_EQ(nullptr, handle);
-  ASSERT_EQ(10, cache.get()->GetUsage());
+  ASSERT_EQ(10, cache->GetUsage());
 
   for (int i = 0; i < 10; i++) {
-    cache.Release(handles[i]);
+    cache->Release(handles[i]);
   }
 
   // test3: init with flag being true.
-  SharedCache cache2{NewCache(5, 0, true)};
+  std::shared_ptr<Cache> cache2 = NewCache(5, 0, true);
   for (int i = 0; i < 5; i++) {
     std::string key = EncodeKey(i + 1);
-    s = cache2.Insert(key, new Value(i + 1), 1, &handles[i]);
+    s = cache2->Insert(key, new Value(i + 1), 1, &deleter, &handles[i]);
     ASSERT_OK(s);
     ASSERT_NE(nullptr, handles[i]);
   }
-  s = cache2.Insert(extra_key, extra_value, 1, &handle);
+  s = cache2->Insert(extra_key, extra_value, 1, &deleter, &handle);
   ASSERT_TRUE(s.IsMemoryLimit());
   ASSERT_EQ(nullptr, handle);
   // test insert without handle
-  s = cache2.Insert(extra_key, extra_value, 1);
+  s = cache2->Insert(extra_key, extra_value, 1, &deleter);
   // AS if the key have been inserted into cache but get evicted immediately.
   ASSERT_OK(s);
-  ASSERT_EQ(5, cache2.get()->GetUsage());
-  ASSERT_EQ(nullptr, cache2.Lookup(extra_key));
+  ASSERT_EQ(5, cache2->GetUsage());
+  ASSERT_EQ(nullptr, cache2->Lookup(extra_key));
 
   for (int i = 0; i < 5; i++) {
-    cache2.Release(handles[i]);
+    cache2->Release(handles[i]);
   }
 }
 
@@ -786,54 +857,55 @@ TEST_P(CacheTest, OverCapacity) {
   size_t n = 10;
 
   // a LRUCache with n entries and one shard only
-  SharedCache cache{NewCache(n, 0, false)};
-  std::vector<TypedHandle*> handles(n + 1);
+  std::shared_ptr<Cache> cache = NewCache(n, 0, false);
+
+  std::vector<Cache::Handle*> handles(n+1);
 
   // Insert n+1 entries, but not releasing.
   for (int i = 0; i < static_cast<int>(n + 1); i++) {
     std::string key = EncodeKey(i + 1);
-    Status s = cache.Insert(key, new Value(i + 1), 1, &handles[i]);
+    Status s = cache->Insert(key, new Value(i + 1), 1, &deleter, &handles[i]);
     ASSERT_TRUE(s.ok());
   }
 
   // Guess what's in the cache now?
   for (int i = 0; i < static_cast<int>(n + 1); i++) {
     std::string key = EncodeKey(i + 1);
-    auto h = cache.Lookup(key);
+    auto h = cache->Lookup(key);
     ASSERT_TRUE(h != nullptr);
-    if (h) cache.Release(h);
+    if (h) cache->Release(h);
   }
 
   // the cache is over capacity since nothing could be evicted
-  ASSERT_EQ(n + 1U, cache.get()->GetUsage());
+  ASSERT_EQ(n + 1U, cache->GetUsage());
   for (int i = 0; i < static_cast<int>(n + 1); i++) {
-    cache.Release(handles[i]);
+    cache->Release(handles[i]);
   }
 
   if (GetParam() == kHyperClock) {
     // Make sure eviction is triggered.
-    ASSERT_OK(cache.Insert(EncodeKey(-1), nullptr, 1, &handles[0]));
+    ASSERT_OK(cache->Insert(EncodeKey(-1), nullptr, 1, &deleter, &handles[0]));
 
     // cache is under capacity now since elements were released
-    ASSERT_GE(n, cache.get()->GetUsage());
+    ASSERT_GE(n, cache->GetUsage());
 
     // clean up
-    cache.Release(handles[0]);
+    cache->Release(handles[0]);
   } else {
     // LRUCache checks for over-capacity in Release.
 
     // cache is exactly at capacity now with minimal eviction
-    ASSERT_EQ(n, cache.get()->GetUsage());
+    ASSERT_EQ(n, cache->GetUsage());
 
     // element 0 is evicted and the rest is there
     // This is consistent with the LRU policy since the element 0
     // was released first
     for (int i = 0; i < static_cast<int>(n + 1); i++) {
       std::string key = EncodeKey(i + 1);
-      auto h = cache.Lookup(key);
+      auto h = cache->Lookup(key);
       if (h) {
         ASSERT_NE(static_cast<size_t>(i), 0U);
-        cache.Release(h);
+        cache->Release(h);
       } else {
         ASSERT_EQ(static_cast<size_t>(i), 0U);
       }
@@ -841,15 +913,40 @@ TEST_P(CacheTest, OverCapacity) {
   }
 }
 
+namespace {
+std::vector<std::pair<int, int>> legacy_callback_state;
+void legacy_callback(void* value, size_t charge) {
+  legacy_callback_state.push_back(
+      {DecodeValue(value), static_cast<int>(charge)});
+}
+};
+
+TEST_P(CacheTest, ApplyToAllCacheEntriesTest) {
+  std::vector<std::pair<int, int>> inserted;
+  legacy_callback_state.clear();
+
+  for (int i = 0; i < 10; ++i) {
+    Insert(i, i * 2, i + 1);
+    inserted.push_back({i * 2, i + 1});
+  }
+  cache_->ApplyToAllCacheEntries(legacy_callback, true);
+
+  std::sort(inserted.begin(), inserted.end());
+  std::sort(legacy_callback_state.begin(), legacy_callback_state.end());
+  ASSERT_EQ(inserted.size(), legacy_callback_state.size());
+  for (int i = 0; i < static_cast<int>(inserted.size()); ++i) {
+    EXPECT_EQ(inserted[i], legacy_callback_state[i]);
+  }
+}
+
 TEST_P(CacheTest, ApplyToAllEntriesTest) {
   std::vector<std::string> callback_state;
-  const auto callback = [&](const Slice& key, Cache::ObjectPtr value,
-                            size_t charge,
-                            const Cache::CacheItemHelper* helper) {
+  const auto callback = [&](const Slice& key, void* value, size_t charge,
+                            Cache::DeleterFn deleter) {
     callback_state.push_back(std::to_string(DecodeKey(key)) + "," +
                              std::to_string(DecodeValue(value)) + "," +
                              std::to_string(charge));
-    assert(helper == &CacheTest::kHelper);
+    assert(deleter == &CacheTest::Deleter);
   };
 
   std::vector<std::string> inserted;
@@ -888,8 +985,8 @@ TEST_P(CacheTest, ApplyToAllEntriesDuringResize) {
 
   // For callback
   int special_count = 0;
-  const auto callback = [&](const Slice&, Cache::ObjectPtr, size_t charge,
-                            const Cache::CacheItemHelper*) {
+  const auto callback = [&](const Slice&, void*, size_t charge,
+                            Cache::DeleterFn) {
     if (charge == static_cast<size_t>(kSpecialCharge)) {
       ++special_count;
     }
@@ -926,21 +1023,21 @@ TEST_P(CacheTest, DefaultShardBits) {
       (GetParam() == kHyperClock ? 32U * 1024U : 512U) * 1024U;
 
   std::shared_ptr<Cache> cache = NewCache(32U * min_shard_size);
-  ShardedCacheBase* sc = dynamic_cast<ShardedCacheBase*>(cache.get());
+  ShardedCache* sc = dynamic_cast<ShardedCache*>(cache.get());
   ASSERT_EQ(5, sc->GetNumShardBits());
 
   cache = NewCache(min_shard_size / 1000U * 999U);
-  sc = dynamic_cast<ShardedCacheBase*>(cache.get());
+  sc = dynamic_cast<ShardedCache*>(cache.get());
   ASSERT_EQ(0, sc->GetNumShardBits());
 
   cache = NewCache(3U * 1024U * 1024U * 1024U);
-  sc = dynamic_cast<ShardedCacheBase*>(cache.get());
+  sc = dynamic_cast<ShardedCache*>(cache.get());
   // current maximum of 6
   ASSERT_EQ(6, sc->GetNumShardBits());
 
   if constexpr (sizeof(size_t) > 4) {
     cache = NewCache(128U * min_shard_size);
-    sc = dynamic_cast<ShardedCacheBase*>(cache.get());
+    sc = dynamic_cast<ShardedCache*>(cache.get());
     // current maximum of 6
     ASSERT_EQ(6, sc->GetNumShardBits());
   }
@@ -951,14 +1048,14 @@ TEST_P(CacheTest, GetChargeAndDeleter) {
   Cache::Handle* h1 = cache_->Lookup(EncodeKey(1));
   ASSERT_EQ(2, DecodeValue(cache_->Value(h1)));
   ASSERT_EQ(1, cache_->GetCharge(h1));
-  ASSERT_EQ(&CacheTest::kHelper, cache_->GetCacheItemHelper(h1));
+  ASSERT_EQ(&CacheTest::Deleter, cache_->GetDeleter(h1));
   cache_->Release(h1);
 }
 
 INSTANTIATE_TEST_CASE_P(CacheTestInstance, CacheTest,
-                        secondary_cache_test_util::GetTestingCacheTypes());
+                        testing::Values(kLRU, kHyperClock, kFast));
 INSTANTIATE_TEST_CASE_P(CacheTestInstance, LRUCacheTest,
-                        testing::Values(secondary_cache_test_util::kLRU));
+                        testing::Values(kLRU, kFast));
 
 }  // namespace ROCKSDB_NAMESPACE
 

cache/charged_cache.cc

@@ -11,57 +11,65 @@ namespace ROCKSDB_NAMESPACE {
 
 ChargedCache::ChargedCache(std::shared_ptr<Cache> cache,
                            std::shared_ptr<Cache> block_cache)
-    : CacheWrapper(cache),
+    : cache_(cache),
       cache_res_mgr_(std::make_shared<ConcurrentCacheReservationManager>(
           std::make_shared<
               CacheReservationManagerImpl<CacheEntryRole::kBlobCache>>(
               block_cache))) {}
 
-Status ChargedCache::Insert(const Slice& key, ObjectPtr obj,
-                            const CacheItemHelper* helper, size_t charge,
-                            Handle** handle, Priority priority) {
-  Status s = target_->Insert(key, obj, helper, charge, handle, priority);
+Status ChargedCache::Insert(const Slice& key, void* value, size_t charge,
+                            DeleterFn deleter, Handle** handle,
+                            Priority priority) {
+  Status s = cache_->Insert(key, value, charge, deleter, handle, priority);
   if (s.ok()) {
     // Insert may cause the cache entry eviction if the cache is full. So we
     // directly call the reservation manager to update the total memory used
     // in the cache.
     assert(cache_res_mgr_);
-    cache_res_mgr_->UpdateCacheReservation(target_->GetUsage())
+    cache_res_mgr_->UpdateCacheReservation(cache_->GetUsage())
         .PermitUncheckedError();
   }
   return s;
 }
 
+Status ChargedCache::Insert(const Slice& key, void* value,
+                            const CacheItemHelper* helper, size_t charge,
+                            Handle** handle, Priority priority) {
+  Status s = cache_->Insert(key, value, helper, charge, handle, priority);
+  if (s.ok()) {
+    // Insert may cause the cache entry eviction if the cache is full. So we
+    // directly call the reservation manager to update the total memory used
+    // in the cache.
+    assert(cache_res_mgr_);
+    cache_res_mgr_->UpdateCacheReservation(cache_->GetUsage())
+        .PermitUncheckedError();
+  }
+  return s;
+}
+
+Cache::Handle* ChargedCache::Lookup(const Slice& key, Statistics* stats) {
+  return cache_->Lookup(key, stats);
+}
+
 Cache::Handle* ChargedCache::Lookup(const Slice& key,
                                     const CacheItemHelper* helper,
-                                    CreateContext* create_context,
-                                    Priority priority, Statistics* stats) {
-  auto handle = target_->Lookup(key, helper, create_context, priority, stats);
+                                    const CreateCallback& create_cb,
+                                    Priority priority, bool wait,
+                                    Statistics* stats) {
+  auto handle = cache_->Lookup(key, helper, create_cb, priority, wait, stats);
   // Lookup may promote the KV pair from the secondary cache to the primary
   // cache. So we directly call the reservation manager to update the total
   // memory used in the cache.
-  if (helper && helper->create_cb) {
-    assert(cache_res_mgr_);
-    cache_res_mgr_->UpdateCacheReservation(target_->GetUsage())
-        .PermitUncheckedError();
-  }
-  return handle;
-}
-
-void ChargedCache::WaitAll(AsyncLookupHandle* async_handles, size_t count) {
-  target_->WaitAll(async_handles, count);
-  // In case of any promotions. Although some could finish by return of
-  // StartAsyncLookup, Wait/WaitAll will generally be used, so simpler to
-  // update here.
   assert(cache_res_mgr_);
-  cache_res_mgr_->UpdateCacheReservation(target_->GetUsage())
+  cache_res_mgr_->UpdateCacheReservation(cache_->GetUsage())
       .PermitUncheckedError();
+  return handle;
 }
 
 bool ChargedCache::Release(Cache::Handle* handle, bool useful,
                            bool erase_if_last_ref) {
-  size_t memory_used_delta = target_->GetUsage(handle);
-  bool erased = target_->Release(handle, useful, erase_if_last_ref);
+  size_t memory_used_delta = cache_->GetUsage(handle);
+  bool erased = cache_->Release(handle, useful, erase_if_last_ref);
   if (erased) {
     assert(cache_res_mgr_);
     cache_res_mgr_
@@ -72,8 +80,8 @@ bool ChargedCache::Release(Cache::Handle* handle, bool useful,
 }
 
 bool ChargedCache::Release(Cache::Handle* handle, bool erase_if_last_ref) {
-  size_t memory_used_delta = target_->GetUsage(handle);
-  bool erased = target_->Release(handle, erase_if_last_ref);
+  size_t memory_used_delta = cache_->GetUsage(handle);
+  bool erased = cache_->Release(handle, erase_if_last_ref);
   if (erased) {
     assert(cache_res_mgr_);
     cache_res_mgr_
@@ -84,25 +92,25 @@ bool ChargedCache::Release(Cache::Handle* handle, bool erase_if_last_ref) {
 }
 
 void ChargedCache::Erase(const Slice& key) {
-  target_->Erase(key);
+  cache_->Erase(key);
   assert(cache_res_mgr_);
-  cache_res_mgr_->UpdateCacheReservation(target_->GetUsage())
+  cache_res_mgr_->UpdateCacheReservation(cache_->GetUsage())
       .PermitUncheckedError();
 }
 
 void ChargedCache::EraseUnRefEntries() {
-  target_->EraseUnRefEntries();
+  cache_->EraseUnRefEntries();
   assert(cache_res_mgr_);
-  cache_res_mgr_->UpdateCacheReservation(target_->GetUsage())
+  cache_res_mgr_->UpdateCacheReservation(cache_->GetUsage())
       .PermitUncheckedError();
 }
 
 void ChargedCache::SetCapacity(size_t capacity) {
-  target_->SetCapacity(capacity);
+  cache_->SetCapacity(capacity);
   // SetCapacity can result in evictions when the cache capacity is decreased,
   // so we would want to update the cache reservation here as well.
   assert(cache_res_mgr_);
-  cache_res_mgr_->UpdateCacheReservation(target_->GetUsage())
+  cache_res_mgr_->UpdateCacheReservation(cache_->GetUsage())
       .PermitUncheckedError();
 }
 

cache/charged_cache.h

@@ -8,7 +8,7 @@
 #include <string>
 
 #include "port/port.h"
-#include "rocksdb/advanced_cache.h"
+#include "rocksdb/cache.h"
 
 namespace ROCKSDB_NAMESPACE {
 
@@ -17,21 +17,22 @@ class ConcurrentCacheReservationManager;
 // A cache interface which wraps around another cache and takes care of
 // reserving space in block cache towards a single global memory limit, and
 // forwards all the calls to the underlying cache.
-class ChargedCache : public CacheWrapper {
+class ChargedCache : public Cache {
  public:
   ChargedCache(std::shared_ptr<Cache> cache,
                std::shared_ptr<Cache> block_cache);
+  ~ChargedCache() override = default;
 
-  Status Insert(const Slice& key, ObjectPtr obj, const CacheItemHelper* helper,
+  Status Insert(const Slice& key, void* value, size_t charge, DeleterFn deleter,
+                Handle** handle, Priority priority) override;
+  Status Insert(const Slice& key, void* value, const CacheItemHelper* helper,
                 size_t charge, Handle** handle = nullptr,
                 Priority priority = Priority::LOW) override;
 
+  Cache::Handle* Lookup(const Slice& key, Statistics* stats) override;
   Cache::Handle* Lookup(const Slice& key, const CacheItemHelper* helper,
-                        CreateContext* create_context,
-                        Priority priority = Priority::LOW,
-                        Statistics* stats = nullptr) override;
-
-  void WaitAll(AsyncLookupHandle* async_handles, size_t count) override;
+                        const CreateCallback& create_cb, Priority priority,
+                        bool wait, Statistics* stats = nullptr) override;
 
   bool Release(Cache::Handle* handle, bool useful,
                bool erase_if_last_ref = false) override;
@@ -43,9 +44,69 @@ class ChargedCache : public CacheWrapper {
   static const char* kClassName() { return "ChargedCache"; }
   const char* Name() const override { return kClassName(); }
 
+  uint64_t NewId() override { return cache_->NewId(); }
+
   void SetCapacity(size_t capacity) override;
 
-  inline Cache* GetCache() const { return target_.get(); }
+  void SetStrictCapacityLimit(bool strict_capacity_limit) override {
+    cache_->SetStrictCapacityLimit(strict_capacity_limit);
+  }
+
+  bool HasStrictCapacityLimit() const override {
+    return cache_->HasStrictCapacityLimit();
+  }
+
+  void* Value(Cache::Handle* handle) override { return cache_->Value(handle); }
+
+  bool IsReady(Cache::Handle* handle) override {
+    return cache_->IsReady(handle);
+  }
+
+  void Wait(Cache::Handle* handle) override { cache_->Wait(handle); }
+
+  void WaitAll(std::vector<Handle*>& handles) override {
+    cache_->WaitAll(handles);
+  }
+
+  bool Ref(Cache::Handle* handle) override { return cache_->Ref(handle); }
+
+  size_t GetCapacity() const override { return cache_->GetCapacity(); }
+
+  size_t GetUsage() const override { return cache_->GetUsage(); }
+
+  size_t GetUsage(Cache::Handle* handle) const override {
+    return cache_->GetUsage(handle);
+  }
+
+  size_t GetPinnedUsage() const override { return cache_->GetPinnedUsage(); }
+
+  size_t GetCharge(Cache::Handle* handle) const override {
+    return cache_->GetCharge(handle);
+  }
+
+  Cache::DeleterFn GetDeleter(Cache::Handle* handle) const override {
+    return cache_->GetDeleter(handle);
+  }
+
+  void ApplyToAllEntries(
+      const std::function<void(const Slice& key, void* value, size_t charge,
+                               Cache::DeleterFn deleter)>& callback,
+      const Cache::ApplyToAllEntriesOptions& opts) override {
+    cache_->ApplyToAllEntries(callback, opts);
+  }
+
+  void ApplyToAllCacheEntries(void (*callback)(void* value, size_t charge),
+                              bool thread_safe) override {
+    cache_->ApplyToAllCacheEntries(callback, thread_safe);
+  }
+
+  std::string GetPrintableOptions() const override {
+    return cache_->GetPrintableOptions();
+  }
+
+  void DisownData() override { return cache_->DisownData(); }
+
+  inline Cache* GetCache() const { return cache_.get(); }
 
   inline ConcurrentCacheReservationManager* TEST_GetCacheReservationManager()
       const {
@@ -53,6 +114,7 @@ class ChargedCache : public CacheWrapper {
   }
 
  private:
+  std::shared_ptr<Cache> cache_;
   std::shared_ptr<ConcurrentCacheReservationManager> cache_res_mgr_;
 };
 

cache/clock_cache.h

@@ -27,13 +27,12 @@
 
 namespace ROCKSDB_NAMESPACE {
 
-namespace clock_cache {
+namespace hyper_clock_cache {
 
 // Forward declaration of friend class.
 class ClockCacheTest;
 
-// HyperClockCache is an alternative to LRUCache specifically tailored for
-// use as BlockBasedTableOptions::block_cache
+// HyperClockCache is an experimental alternative to LRUCache.
 //
 // Benefits
 // --------
@@ -145,7 +144,7 @@ class ClockCacheTest;
 //     (erased by user) but can be read by existing references, and ref count
 //     changed by Ref and Release.
 //
-// A special case is "standalone" entries, which are heap-allocated handles
+// A special case is "detached" entries, which are heap-allocated handles
 // not in the table. They are always Invisible and freed on zero refs.
 //
 // State transitions:
@@ -200,8 +199,8 @@ class ClockCacheTest;
 // table occupancy limit has been reached. If strict_capacity_limit=false,
 // we must never fail Insert, and if a Handle* is provided, we have to return
 // a usable Cache handle on success. The solution to this (typically rare)
-// problem is "standalone" handles, which are usable by the caller but not
-// actually available for Lookup in the Cache. Standalone handles are allocated
+// problem is "detached" handles, which are usable by the caller but not
+// actually available for Lookup in the Cache. Detached handles are allocated
 // independently on the heap and specially marked so that they are freed on
 // the heap when their last reference is released.
 //
@@ -304,24 +303,30 @@ constexpr double kLoadFactor = 0.7;
 // strict upper bound on the load factor.
 constexpr double kStrictLoadFactor = 0.84;
 
+using CacheKeyBytes = std::array<char, kCacheKeySize>;
+
 struct ClockHandleBasicData {
-  Cache::ObjectPtr value = nullptr;
-  const Cache::CacheItemHelper* helper = nullptr;
-  // A lossless, reversible hash of the fixed-size (16 byte) cache key. This
-  // eliminates the need to store a hash separately.
-  UniqueId64x2 hashed_key = kNullUniqueId64x2;
+  void* value = nullptr;
+  Cache::DeleterFn deleter = nullptr;
+  CacheKeyBytes key = {};
   size_t total_charge = 0;
 
-  inline size_t GetTotalCharge() const { return total_charge; }
+  Slice KeySlice() const { return Slice(key.data(), kCacheKeySize); }
 
-  // Calls deleter (if non-null) on cache key and value
-  void FreeData(MemoryAllocator* allocator) const;
-
-  // Required by concept HandleImpl
-  const UniqueId64x2& GetHash() const { return hashed_key; }
+  void FreeData() const {
+    if (deleter) {
+      (*deleter)(KeySlice(), value);
+    }
+  }
 };
 
-struct ClockHandle : public ClockHandleBasicData {
+struct ClockHandleMoreData : public ClockHandleBasicData {
+  uint32_t hash = 0;
+};
+
+// Target size to be exactly a common cache line size (see static_assert in
+// clock_cache.cc)
+struct ALIGN_AS(64U) ClockHandle : public ClockHandleMoreData {
   // Constants for handling the atomic `meta` word, which tracks most of the
   // state of the handle. The meta word looks like this:
   // low bits                                                     high bits
@@ -373,96 +378,66 @@ struct ClockHandle : public ClockHandleBasicData {
 
   // See above
   std::atomic<uint64_t> meta{};
+  // The number of elements that hash to this slot or a lower one, but wind
+  // up in this slot or a higher one.
+  std::atomic<uint32_t> displacements{};
 
-  // Anticipating use for SecondaryCache support
-  void* reserved_for_future_use = nullptr;
+  // True iff the handle is allocated separately from hash table.
+  bool detached = false;
 };  // struct ClockHandle
 
-class HyperClockTable {
+class ClockHandleTable {
  public:
-  // Target size to be exactly a common cache line size (see static_assert in
-  // clock_cache.cc)
-  struct ALIGN_AS(64U) HandleImpl : public ClockHandle {
-    // The number of elements that hash to this slot or a lower one, but wind
-    // up in this slot or a higher one.
-    std::atomic<uint32_t> displacements{};
+  explicit ClockHandleTable(int hash_bits, bool initial_charge_metadata);
+  ~ClockHandleTable();
 
-    // Whether this is a "deteched" handle that is independently allocated
-    // with `new` (so must be deleted with `delete`).
-    // TODO: ideally this would be packed into some other data field, such
-    // as upper bits of total_charge, but that incurs a measurable performance
-    // regression.
-    bool standalone = false;
-
-    inline bool IsStandalone() const { return standalone; }
-
-    inline void SetStandalone() { standalone = true; }
-  };  // struct HandleImpl
-
-  struct Opts {
-    size_t estimated_value_size;
-  };
-
-  HyperClockTable(size_t capacity, bool strict_capacity_limit,
-                  CacheMetadataChargePolicy metadata_charge_policy,
-                  MemoryAllocator* allocator,
-                  const Cache::EvictionCallback* eviction_callback,
-                  const Opts& opts);
-  ~HyperClockTable();
-
-  Status Insert(const ClockHandleBasicData& proto, HandleImpl** handle,
+  Status Insert(const ClockHandleMoreData& proto, ClockHandle** handle,
                 Cache::Priority priority, size_t capacity,
                 bool strict_capacity_limit);
 
-  HandleImpl* CreateStandalone(ClockHandleBasicData& proto, size_t capacity,
-                               bool strict_capacity_limit,
-                               bool allow_uncharged);
+  ClockHandle* Lookup(const CacheKeyBytes& key, uint32_t hash);
 
-  HandleImpl* Lookup(const UniqueId64x2& hashed_key);
+  bool Release(ClockHandle* handle, bool useful, bool erase_if_last_ref);
 
-  bool Release(HandleImpl* handle, bool useful, bool erase_if_last_ref);
+  void Ref(ClockHandle& handle);
 
-  void Ref(HandleImpl& handle);
+  void Erase(const CacheKeyBytes& key, uint32_t hash);
 
-  void Erase(const UniqueId64x2& hashed_key);
-
-  void ConstApplyToEntriesRange(std::function<void(const HandleImpl&)> func,
-                                size_t index_begin, size_t index_end,
+  void ConstApplyToEntriesRange(std::function<void(const ClockHandle&)> func,
+                                uint32_t index_begin, uint32_t index_end,
                                 bool apply_if_will_be_deleted) const;
 
   void EraseUnRefEntries();
 
-  size_t GetTableSize() const { return size_t{1} << length_bits_; }
+  uint32_t GetTableSize() const { return uint32_t{1} << length_bits_; }
 
   int GetLengthBits() const { return length_bits_; }
 
-  size_t GetOccupancy() const {
+  uint32_t GetOccupancyLimit() const { return occupancy_limit_; }
+
+  uint32_t GetOccupancy() const {
     return occupancy_.load(std::memory_order_relaxed);
   }
 
-  size_t GetOccupancyLimit() const { return occupancy_limit_; }
-
   size_t GetUsage() const { return usage_.load(std::memory_order_relaxed); }
 
-  size_t GetStandaloneUsage() const {
-    return standalone_usage_.load(std::memory_order_relaxed);
+  size_t GetDetachedUsage() const {
+    return detached_usage_.load(std::memory_order_relaxed);
   }
 
   // Acquire/release N references
-  void TEST_RefN(HandleImpl& handle, size_t n);
-  void TEST_ReleaseN(HandleImpl* handle, size_t n);
+  void TEST_RefN(ClockHandle& handle, size_t n);
+  void TEST_ReleaseN(ClockHandle* handle, size_t n);
 
  private:  // functions
   // Returns x mod 2^{length_bits_}.
-  inline size_t ModTableSize(uint64_t x) {
-    return static_cast<size_t>(x) & length_bits_mask_;
-  }
+  uint32_t ModTableSize(uint32_t x) { return x & length_bits_mask_; }
 
   // Runs the clock eviction algorithm trying to reclaim at least
   // requested_charge. Returns how much is evicted, which could be less
   // if it appears impossible to evict the requested amount without blocking.
-  inline void Evict(size_t requested_charge, size_t* freed_charge,
-                    size_t* freed_count);
+  void Evict(size_t requested_charge, size_t* freed_charge,
+             uint32_t* freed_count);
 
   // Returns the first slot in the probe sequence, starting from the given
   // probe number, with a handle e such that match(e) is true. At every
@@ -475,56 +450,15 @@ class HyperClockTable {
   // value of probe is one more than the last non-aborting probe during the
   // call. This is so that that the variable can be used to keep track of
   // progress across consecutive calls to FindSlot.
-  inline HandleImpl* FindSlot(const UniqueId64x2& hashed_key,
-                              std::function<bool(HandleImpl*)> match,
-                              std::function<bool(HandleImpl*)> stop,
-                              std::function<void(HandleImpl*)> update,
-                              size_t& probe);
+  inline ClockHandle* FindSlot(uint32_t hash,
+                               std::function<bool(ClockHandle*)> match,
+                               std::function<bool(ClockHandle*)> stop,
+                               std::function<void(ClockHandle*)> update,
+                               uint32_t& probe);
 
   // Re-decrement all displacements in probe path starting from beginning
   // until (not including) the given handle
-  inline void Rollback(const UniqueId64x2& hashed_key, const HandleImpl* h);
-
-  // Subtracts `total_charge` from `usage_` and 1 from `occupancy_`.
-  // Ideally this comes after releasing the entry itself so that we
-  // actually have the available occupancy/usage that is claimed.
-  // However, that means total_charge has to be saved from the handle
-  // before releasing it so that it can be provided to this function.
-  inline void ReclaimEntryUsage(size_t total_charge);
-
-  // Helper for updating `usage_` for new entry with given `total_charge`
-  // and evicting if needed under strict_capacity_limit=true rules. This
-  // means the operation might fail with Status::MemoryLimit. If
-  // `need_evict_for_occupancy`, then eviction of at least one entry is
-  // required, and the operation should fail if not possible.
-  // NOTE: Otherwise, occupancy_ is not managed in this function
-  inline Status ChargeUsageMaybeEvictStrict(size_t total_charge,
-                                            size_t capacity,
-                                            bool need_evict_for_occupancy);
-
-  // Helper for updating `usage_` for new entry with given `total_charge`
-  // and evicting if needed under strict_capacity_limit=false rules. This
-  // means that updating `usage_` always succeeds even if forced to exceed
-  // capacity. If `need_evict_for_occupancy`, then eviction of at least one
-  // entry is required, and the operation should return false if such eviction
-  // is not possible. `usage_` is not updated in that case. Otherwise, returns
-  // true, indicating success.
-  // NOTE: occupancy_ is not managed in this function
-  inline bool ChargeUsageMaybeEvictNonStrict(size_t total_charge,
-                                             size_t capacity,
-                                             bool need_evict_for_occupancy);
-
-  // Creates a "standalone" handle for returning from an Insert operation that
-  // cannot be completed by actually inserting into the table.
-  // Updates `standalone_usage_` but not `usage_` nor `occupancy_`.
-  inline HandleImpl* StandaloneInsert(const ClockHandleBasicData& proto);
-
-  MemoryAllocator* GetAllocator() const { return allocator_; }
-
-  // Returns the number of bits used to hash an element in the hash
-  // table.
-  static int CalcHashBits(size_t capacity, size_t estimated_value_size,
-                          CacheMetadataChargePolicy metadata_charge_policy);
+  void Rollback(uint32_t hash, const ClockHandle* h);
 
  private:  // data
   // Number of hash bits used for table index.
@@ -532,19 +466,13 @@ class HyperClockTable {
   const int length_bits_;
 
   // For faster computation of ModTableSize.
-  const size_t length_bits_mask_;
+  const uint32_t length_bits_mask_;
 
   // Maximum number of elements the user can store in the table.
-  const size_t occupancy_limit_;
+  const uint32_t occupancy_limit_;
 
   // Array of slots comprising the hash table.
-  const std::unique_ptr<HandleImpl[]> array_;
-
-  // From Cache, for deleter
-  MemoryAllocator* const allocator_;
-
-  // A reference to Cache::eviction_callback_
-  const Cache::EvictionCallback& eviction_callback_;
+  const std::unique_ptr<ClockHandle[]> array_;
 
   // We partition the following members into different cache lines
   // to avoid false sharing among Lookup, Release, Erase and Insert
@@ -556,116 +484,96 @@ class HyperClockTable {
 
   ALIGN_AS(CACHE_LINE_SIZE)
   // Number of elements in the table.
-  std::atomic<size_t> occupancy_{};
+  std::atomic<uint32_t> occupancy_{};
 
-  // Memory usage by entries tracked by the cache (including standalone)
+  // Memory usage by entries tracked by the cache (including detached)
   std::atomic<size_t> usage_{};
 
-  // Part of usage by standalone entries (not in table)
-  std::atomic<size_t> standalone_usage_{};
-};  // class HyperClockTable
+  // Part of usage by detached entries (not in table)
+  std::atomic<size_t> detached_usage_{};
+};  // class ClockHandleTable
 
 // A single shard of sharded cache.
-template <class Table>
-class ALIGN_AS(CACHE_LINE_SIZE) ClockCacheShard final : public CacheShardBase {
+class ALIGN_AS(CACHE_LINE_SIZE) ClockCacheShard final : public CacheShard {
  public:
-  ClockCacheShard(size_t capacity, bool strict_capacity_limit,
-                  CacheMetadataChargePolicy metadata_charge_policy,
-                  MemoryAllocator* allocator,
-                  const Cache::EvictionCallback* eviction_callback,
-                  const typename Table::Opts& opts);
+  ClockCacheShard(size_t capacity, size_t estimated_value_size,
+                  bool strict_capacity_limit,
+                  CacheMetadataChargePolicy metadata_charge_policy);
+  ~ClockCacheShard() override = default;
 
-  // For CacheShard concept
-  using HandleImpl = typename Table::HandleImpl;
-  // Hash is lossless hash of 128-bit key
-  using HashVal = UniqueId64x2;
-  using HashCref = const HashVal&;
-  static inline uint32_t HashPieceForSharding(HashCref hash) {
-    return Upper32of64(hash[0]);
-  }
-  static inline HashVal ComputeHash(const Slice& key) {
-    assert(key.size() == kCacheKeySize);
-    HashVal in;
-    HashVal out;
-    // NOTE: endian dependence
-    // TODO: use GetUnaligned?
-    std::memcpy(&in, key.data(), kCacheKeySize);
-    BijectiveHash2x64(in[1], in[0], &out[1], &out[0]);
-    return out;
-  }
+  // TODO: document limitations
+  void SetCapacity(size_t capacity) override;
 
-  // For reconstructing key from hashed_key. Requires the caller to provide
-  // backing storage for the Slice in `unhashed`
-  static inline Slice ReverseHash(const UniqueId64x2& hashed,
-                                  UniqueId64x2* unhashed) {
-    BijectiveUnhash2x64(hashed[1], hashed[0], &(*unhashed)[1], &(*unhashed)[0]);
-    // NOTE: endian dependence
-    return Slice(reinterpret_cast<const char*>(unhashed), kCacheKeySize);
-  }
+  void SetStrictCapacityLimit(bool strict_capacity_limit) override;
 
-  // Although capacity is dynamically changeable, the number of table slots is
-  // not, so growing capacity substantially could lead to hitting occupancy
-  // limit.
-  void SetCapacity(size_t capacity);
+  Status Insert(const Slice& key, uint32_t hash, void* value, size_t charge,
+                Cache::DeleterFn deleter, Cache::Handle** handle,
+                Cache::Priority priority) override;
 
-  void SetStrictCapacityLimit(bool strict_capacity_limit);
+  Cache::Handle* Lookup(const Slice& key, uint32_t hash) override;
 
-  Status Insert(const Slice& key, const UniqueId64x2& hashed_key,
-                Cache::ObjectPtr value, const Cache::CacheItemHelper* helper,
-                size_t charge, HandleImpl** handle, Cache::Priority priority);
+  bool Release(Cache::Handle* handle, bool useful,
+               bool erase_if_last_ref) override;
 
-  HandleImpl* CreateStandalone(const Slice& key, const UniqueId64x2& hashed_key,
-                               Cache::ObjectPtr obj,
-                               const Cache::CacheItemHelper* helper,
-                               size_t charge, bool allow_uncharged);
+  bool Release(Cache::Handle* handle, bool erase_if_last_ref = false) override;
 
-  HandleImpl* Lookup(const Slice& key, const UniqueId64x2& hashed_key);
+  bool Ref(Cache::Handle* handle) override;
 
-  bool Release(HandleImpl* handle, bool useful, bool erase_if_last_ref);
+  void Erase(const Slice& key, uint32_t hash) override;
 
-  bool Release(HandleImpl* handle, bool erase_if_last_ref = false);
+  size_t GetUsage() const override;
 
-  bool Ref(HandleImpl* handle);
+  size_t GetPinnedUsage() const override;
 
-  void Erase(const Slice& key, const UniqueId64x2& hashed_key);
+  size_t GetOccupancyCount() const override;
 
-  size_t GetCapacity() const;
-
-  size_t GetUsage() const;
-
-  size_t GetStandaloneUsage() const;
-
-  size_t GetPinnedUsage() const;
-
-  size_t GetOccupancyCount() const;
-
-  size_t GetOccupancyLimit() const;
-
-  size_t GetTableAddressCount() const;
+  size_t GetTableAddressCount() const override;
 
   void ApplyToSomeEntries(
-      const std::function<void(const Slice& key, Cache::ObjectPtr obj,
-                               size_t charge,
-                               const Cache::CacheItemHelper* helper)>& callback,
-      size_t average_entries_per_lock, size_t* state);
+      const std::function<void(const Slice& key, void* value, size_t charge,
+                               DeleterFn deleter)>& callback,
+      uint32_t average_entries_per_lock, uint32_t* state) override;
 
-  void EraseUnRefEntries();
+  void EraseUnRefEntries() override;
 
-  std::string GetPrintableOptions() const { return std::string{}; }
+  std::string GetPrintableOptions() const override { return std::string{}; }
 
-  HandleImpl* Lookup(const Slice& key, const UniqueId64x2& hashed_key,
-                     const Cache::CacheItemHelper* /*helper*/,
-                     Cache::CreateContext* /*create_context*/,
-                     Cache::Priority /*priority*/, Statistics* /*stats*/) {
-    return Lookup(key, hashed_key);
+  // SecondaryCache not yet supported
+  Status Insert(const Slice& key, uint32_t hash, void* value,
+                const Cache::CacheItemHelper* helper, size_t charge,
+                Cache::Handle** handle, Cache::Priority priority) override {
+    return Insert(key, hash, value, charge, helper->del_cb, handle, priority);
   }
 
+  Cache::Handle* Lookup(const Slice& key, uint32_t hash,
+                        const Cache::CacheItemHelper* /*helper*/,
+                        const Cache::CreateCallback& /*create_cb*/,
+                        Cache::Priority /*priority*/, bool /*wait*/,
+                        Statistics* /*stats*/) override {
+    return Lookup(key, hash);
+  }
+
+  bool IsReady(Cache::Handle* /*handle*/) override { return true; }
+
+  void Wait(Cache::Handle* /*handle*/) override {}
+
   // Acquire/release N references
-  void TEST_RefN(HandleImpl* handle, size_t n);
-  void TEST_ReleaseN(HandleImpl* handle, size_t n);
+  void TEST_RefN(Cache::Handle* handle, size_t n);
+  void TEST_ReleaseN(Cache::Handle* handle, size_t n);
+
+ private:  // functions
+  friend class ClockCache;
+  friend class ClockCacheTest;
+
+  ClockHandle* DetachedInsert(const ClockHandleMoreData& h);
+
+  // Returns the number of bits used to hash an element in the hash
+  // table.
+  static int CalcHashBits(size_t capacity, size_t estimated_value_size,
+                          CacheMetadataChargePolicy metadata_charge_policy);
 
  private:  // data
-  Table table_;
+  ClockHandleTable table_;
 
   // Maximum total charge of all elements stored in the table.
   std::atomic<size_t> capacity_;
@@ -678,27 +586,37 @@ class HyperClockCache
 #ifdef NDEBUG
     final
 #endif
-    : public ShardedCache<ClockCacheShard<HyperClockTable>> {
+    : public ShardedCache {
  public:
-  using Shard = ClockCacheShard<HyperClockTable>;
-
   HyperClockCache(size_t capacity, size_t estimated_value_size,
                   int num_shard_bits, bool strict_capacity_limit,
-                  CacheMetadataChargePolicy metadata_charge_policy,
-                  std::shared_ptr<MemoryAllocator> memory_allocator);
+                  CacheMetadataChargePolicy metadata_charge_policy =
+                      kDontChargeCacheMetadata);
+
+  ~HyperClockCache() override;
 
   const char* Name() const override { return "HyperClockCache"; }
 
-  Cache::ObjectPtr Value(Handle* handle) override;
+  CacheShard* GetShard(uint32_t shard) override;
+
+  const CacheShard* GetShard(uint32_t shard) const override;
+
+  void* Value(Handle* handle) override;
 
   size_t GetCharge(Handle* handle) const override;
 
-  const CacheItemHelper* GetCacheItemHelper(Handle* handle) const override;
+  uint32_t GetHash(Handle* handle) const override;
 
-  void ReportProblems(
-      const std::shared_ptr<Logger>& /*info_log*/) const override;
+  DeleterFn GetDeleter(Handle* handle) const override;
+
+  void DisownData() override;
+
+ private:
+  ClockCacheShard* shards_ = nullptr;
+
+  int num_shards_;
 };  // class HyperClockCache
 
-}  // namespace clock_cache
+}  // namespace hyper_clock_cache
 
 }  // namespace ROCKSDB_NAMESPACE

cache/compressed_secondary_cache.cc

@@ -22,13 +22,12 @@ CompressedSecondaryCache::CompressedSecondaryCache(
     std::shared_ptr<MemoryAllocator> memory_allocator, bool use_adaptive_mutex,
     CacheMetadataChargePolicy metadata_charge_policy,
     CompressionType compression_type, uint32_t compress_format_version,
-    bool enable_custom_split_merge,
-    const CacheEntryRoleSet& do_not_compress_roles)
+    bool enable_custom_split_merge)
     : cache_options_(capacity, num_shard_bits, strict_capacity_limit,
                      high_pri_pool_ratio, low_pri_pool_ratio, memory_allocator,
                      use_adaptive_mutex, metadata_charge_policy,
                      compression_type, compress_format_version,
-                     enable_custom_split_merge, do_not_compress_roles) {
+                     enable_custom_split_merge) {
   cache_ =
       NewLRUCache(capacity, num_shard_bits, strict_capacity_limit,
                   high_pri_pool_ratio, memory_allocator, use_adaptive_mutex,
@@ -38,12 +37,10 @@ CompressedSecondaryCache::CompressedSecondaryCache(
 CompressedSecondaryCache::~CompressedSecondaryCache() { cache_.reset(); }
 
 std::unique_ptr<SecondaryCacheResultHandle> CompressedSecondaryCache::Lookup(
-    const Slice& key, const Cache::CacheItemHelper* helper,
-    Cache::CreateContext* create_context, bool /*wait*/, bool advise_erase,
-    bool& kept_in_sec_cache) {
-  assert(helper);
+    const Slice& key, const Cache::CreateCallback& create_cb, bool /*wait*/,
+    bool advise_erase, bool& is_in_sec_cache) {
   std::unique_ptr<SecondaryCacheResultHandle> handle;
-  kept_in_sec_cache = false;
+  is_in_sec_cache = false;
   Cache::Handle* lru_handle = cache_->Lookup(key);
   if (lru_handle == nullptr) {
     return nullptr;
@@ -67,15 +64,12 @@ std::unique_ptr<SecondaryCacheResultHandle> CompressedSecondaryCache::Lookup(
     ptr = reinterpret_cast<CacheAllocationPtr*>(handle_value);
     handle_value_charge = cache_->GetCharge(lru_handle);
   }
-  MemoryAllocator* allocator = cache_options_.memory_allocator.get();
 
   Status s;
-  Cache::ObjectPtr value{nullptr};
+  void* value{nullptr};
   size_t charge{0};
-  if (cache_options_.compression_type == kNoCompression ||
-      cache_options_.do_not_compress_roles.Contains(helper->role)) {
-    s = helper->create_cb(Slice(ptr->get(), handle_value_charge),
-                          create_context, allocator, &value, &charge);
+  if (cache_options_.compression_type == kNoCompression) {
+    s = create_cb(ptr->get(), handle_value_charge, &value, &charge);
   } else {
     UncompressionContext uncompression_context(cache_options_.compression_type);
     UncompressionInfo uncompression_info(uncompression_context,
@@ -85,14 +79,14 @@ std::unique_ptr<SecondaryCacheResultHandle> CompressedSecondaryCache::Lookup(
     size_t uncompressed_size{0};
     CacheAllocationPtr uncompressed = UncompressData(
         uncompression_info, (char*)ptr->get(), handle_value_charge,
-        &uncompressed_size, cache_options_.compress_format_version, allocator);
+        &uncompressed_size, cache_options_.compress_format_version,
+        cache_options_.memory_allocator.get());
 
     if (!uncompressed) {
       cache_->Release(lru_handle, /*erase_if_last_ref=*/true);
       return nullptr;
     }
-    s = helper->create_cb(Slice(uncompressed.get(), uncompressed_size),
-                          create_context, allocator, &value, &charge);
+    s = create_cb(uncompressed.get(), uncompressed_size, &value, &charge);
   }
 
   if (!s.ok()) {
@@ -104,32 +98,30 @@ std::unique_ptr<SecondaryCacheResultHandle> CompressedSecondaryCache::Lookup(
     cache_->Release(lru_handle, /*erase_if_last_ref=*/true);
     // Insert a dummy handle.
     cache_
-        ->Insert(key, /*obj=*/nullptr,
-                 GetHelper(cache_options_.enable_custom_split_merge),
-                 /*charge=*/0)
+        ->Insert(key, /*value=*/nullptr, /*charge=*/0,
+                 GetDeletionCallback(cache_options_.enable_custom_split_merge))
         .PermitUncheckedError();
   } else {
-    kept_in_sec_cache = true;
+    is_in_sec_cache = true;
     cache_->Release(lru_handle, /*erase_if_last_ref=*/false);
   }
   handle.reset(new CompressedSecondaryCacheResultHandle(value, charge));
   return handle;
 }
 
-Status CompressedSecondaryCache::Insert(const Slice& key,
-                                        Cache::ObjectPtr value,
+Status CompressedSecondaryCache::Insert(const Slice& key, void* value,
                                         const Cache::CacheItemHelper* helper) {
   if (value == nullptr) {
     return Status::InvalidArgument();
   }
 
   Cache::Handle* lru_handle = cache_->Lookup(key);
-  auto internal_helper = GetHelper(cache_options_.enable_custom_split_merge);
+  Cache::DeleterFn del_cb =
+      GetDeletionCallback(cache_options_.enable_custom_split_merge);
   if (lru_handle == nullptr) {
     PERF_COUNTER_ADD(compressed_sec_cache_insert_dummy_count, 1);
     // Insert a dummy handle if the handle is evicted for the first time.
-    return cache_->Insert(key, /*obj=*/nullptr, internal_helper,
-                          /*charge=*/0);
+    return cache_->Insert(key, /*value=*/nullptr, /*charge=*/0, del_cb);
   } else {
     cache_->Release(lru_handle, /*erase_if_last_ref=*/false);
   }
@@ -145,8 +137,7 @@ Status CompressedSecondaryCache::Insert(const Slice& key,
   Slice val(ptr.get(), size);
 
   std::string compressed_val;
-  if (cache_options_.compression_type != kNoCompression &&
-      !cache_options_.do_not_compress_roles.Contains(helper->role)) {
+  if (cache_options_.compression_type != kNoCompression) {
     PERF_COUNTER_ADD(compressed_sec_cache_uncompressed_bytes, size);
     CompressionOptions compression_opts;
     CompressionContext compression_context(cache_options_.compression_type);
@@ -178,28 +169,15 @@ Status CompressedSecondaryCache::Insert(const Slice& key,
     size_t charge{0};
     CacheValueChunk* value_chunks_head =
         SplitValueIntoChunks(val, cache_options_.compression_type, charge);
-    return cache_->Insert(key, value_chunks_head, internal_helper, charge);
+    return cache_->Insert(key, value_chunks_head, charge, del_cb);
   } else {
     CacheAllocationPtr* buf = new CacheAllocationPtr(std::move(ptr));
-    return cache_->Insert(key, buf, internal_helper, size);
+    return cache_->Insert(key, buf, size, del_cb);
   }
 }
 
 void CompressedSecondaryCache::Erase(const Slice& key) { cache_->Erase(key); }
 
-Status CompressedSecondaryCache::SetCapacity(size_t capacity) {
-  MutexLock l(&capacity_mutex_);
-  cache_options_.capacity = capacity;
-  cache_->SetCapacity(capacity);
-  return Status::OK();
-}
-
-Status CompressedSecondaryCache::GetCapacity(size_t& capacity) {
-  MutexLock l(&capacity_mutex_);
-  capacity = cache_options_.capacity;
-  return Status::OK();
-}
-
 std::string CompressedSecondaryCache::GetPrintableOptions() const {
   std::string ret;
   ret.reserve(20000);
@@ -216,9 +194,9 @@ std::string CompressedSecondaryCache::GetPrintableOptions() const {
 }
 
 CompressedSecondaryCache::CacheValueChunk*
-CompressedSecondaryCache::SplitValueIntoChunks(const Slice& value,
-                                               CompressionType compression_type,
-                                               size_t& charge) {
+CompressedSecondaryCache::SplitValueIntoChunks(
+    const Slice& value, const CompressionType compression_type,
+    size_t& charge) {
   assert(!value.empty());
   const char* src_ptr = value.data();
   size_t src_size{value.size()};
@@ -285,29 +263,23 @@ CacheAllocationPtr CompressedSecondaryCache::MergeChunksIntoValue(
   return ptr;
 }
 
-const Cache::CacheItemHelper* CompressedSecondaryCache::GetHelper(
-    bool enable_custom_split_merge) const {
+Cache::DeleterFn CompressedSecondaryCache::GetDeletionCallback(
+    bool enable_custom_split_merge) {
   if (enable_custom_split_merge) {
-    static const Cache::CacheItemHelper kHelper{
-        CacheEntryRole::kMisc,
-        [](Cache::ObjectPtr obj, MemoryAllocator* /*alloc*/) {
-          CacheValueChunk* chunks_head = static_cast<CacheValueChunk*>(obj);
-          while (chunks_head != nullptr) {
-            CacheValueChunk* tmp_chunk = chunks_head;
-            chunks_head = chunks_head->next;
-            tmp_chunk->Free();
-            obj = nullptr;
-          };
-        }};
-    return &kHelper;
+    return [](const Slice& /*key*/, void* obj) {
+      CacheValueChunk* chunks_head = reinterpret_cast<CacheValueChunk*>(obj);
+      while (chunks_head != nullptr) {
+        CacheValueChunk* tmp_chunk = chunks_head;
+        chunks_head = chunks_head->next;
+        tmp_chunk->Free();
+        obj = nullptr;
+      };
+    };
   } else {
-    static const Cache::CacheItemHelper kHelper{
-        CacheEntryRole::kMisc,
-        [](Cache::ObjectPtr obj, MemoryAllocator* /*alloc*/) {
-          delete static_cast<CacheAllocationPtr*>(obj);
-          obj = nullptr;
-        }};
-    return &kHelper;
+    return [](const Slice& /*key*/, void* obj) {
+      delete reinterpret_cast<CacheAllocationPtr*>(obj);
+      obj = nullptr;
+    };
   }
 }
 
@@ -317,13 +289,12 @@ std::shared_ptr<SecondaryCache> NewCompressedSecondaryCache(
     std::shared_ptr<MemoryAllocator> memory_allocator, bool use_adaptive_mutex,
     CacheMetadataChargePolicy metadata_charge_policy,
     CompressionType compression_type, uint32_t compress_format_version,
-    bool enable_custom_split_merge,
-    const CacheEntryRoleSet& do_not_compress_roles) {
+    bool enable_custom_split_merge) {
   return std::make_shared<CompressedSecondaryCache>(
       capacity, num_shard_bits, strict_capacity_limit, high_pri_pool_ratio,
       low_pri_pool_ratio, memory_allocator, use_adaptive_mutex,
       metadata_charge_policy, compression_type, compress_format_version,
-      enable_custom_split_merge, do_not_compress_roles);
+      enable_custom_split_merge);
 }
 
 std::shared_ptr<SecondaryCache> NewCompressedSecondaryCache(
@@ -335,7 +306,7 @@ std::shared_ptr<SecondaryCache> NewCompressedSecondaryCache(
       opts.high_pri_pool_ratio, opts.low_pri_pool_ratio, opts.memory_allocator,
       opts.use_adaptive_mutex, opts.metadata_charge_policy,
       opts.compression_type, opts.compress_format_version,
-      opts.enable_custom_split_merge, opts.do_not_compress_roles);
+      opts.enable_custom_split_merge);
 }
 
 }  // namespace ROCKSDB_NAMESPACE

cache/compressed_secondary_cache.h

@@ -15,15 +15,14 @@
 #include "rocksdb/slice.h"
 #include "rocksdb/status.h"
 #include "util/compression.h"
-#include "util/mutexlock.h"
 
 namespace ROCKSDB_NAMESPACE {
 
 class CompressedSecondaryCacheResultHandle : public SecondaryCacheResultHandle {
  public:
-  CompressedSecondaryCacheResultHandle(Cache::ObjectPtr value, size_t size)
+  CompressedSecondaryCacheResultHandle(void* value, size_t size)
       : value_(value), size_(size) {}
-  ~CompressedSecondaryCacheResultHandle() override = default;
+  virtual ~CompressedSecondaryCacheResultHandle() override = default;
 
   CompressedSecondaryCacheResultHandle(
       const CompressedSecondaryCacheResultHandle&) = delete;
@@ -34,12 +33,12 @@ class CompressedSecondaryCacheResultHandle : public SecondaryCacheResultHandle {
 
   void Wait() override {}
 
-  Cache::ObjectPtr Value() override { return value_; }
+  void* Value() override { return value_; }
 
   size_t Size() override { return size_; }
 
  private:
-  Cache::ObjectPtr value_;
+  void* value_;
   size_t size_;
 };
 
@@ -78,20 +77,17 @@ class CompressedSecondaryCache : public SecondaryCache {
           kDefaultCacheMetadataChargePolicy,
       CompressionType compression_type = CompressionType::kLZ4Compression,
       uint32_t compress_format_version = 2,
-      bool enable_custom_split_merge = false,
-      const CacheEntryRoleSet& do_not_compress_roles = {
-          CacheEntryRole::kFilterBlock});
-  ~CompressedSecondaryCache() override;
+      bool enable_custom_split_merge = false);
+  virtual ~CompressedSecondaryCache() override;
 
   const char* Name() const override { return "CompressedSecondaryCache"; }
 
-  Status Insert(const Slice& key, Cache::ObjectPtr value,
+  Status Insert(const Slice& key, void* value,
                 const Cache::CacheItemHelper* helper) override;
 
   std::unique_ptr<SecondaryCacheResultHandle> Lookup(
-      const Slice& key, const Cache::CacheItemHelper* helper,
-      Cache::CreateContext* create_context, bool /*wait*/, bool advise_erase,
-      bool& kept_in_sec_cache) override;
+      const Slice& key, const Cache::CreateCallback& create_cb, bool /*wait*/,
+      bool advise_erase, bool& is_in_sec_cache) override;
 
   bool SupportForceErase() const override { return true; }
 
@@ -99,14 +95,10 @@ class CompressedSecondaryCache : public SecondaryCache {
 
   void WaitAll(std::vector<SecondaryCacheResultHandle*> /*handles*/) override {}
 
-  Status SetCapacity(size_t capacity) override;
-
-  Status GetCapacity(size_t& capacity) override;
-
   std::string GetPrintableOptions() const override;
 
  private:
-  friend class CompressedSecondaryCacheTestBase;
+  friend class CompressedSecondaryCacheTest;
   static constexpr std::array<uint16_t, 8> malloc_bin_sizes_{
       128, 256, 512, 1024, 2048, 4096, 8192, 16384};
 
@@ -124,7 +116,7 @@ class CompressedSecondaryCache : public SecondaryCache {
   // are stored in CacheValueChunk and extra charge is needed for each chunk,
   // so the cache charge is recalculated here.
   CacheValueChunk* SplitValueIntoChunks(const Slice& value,
-                                        CompressionType compression_type,
+                                        const CompressionType compression_type,
                                         size_t& charge);
 
   // After merging chunks, the extra charge for each chunk is removed, so
@@ -132,11 +124,10 @@ class CompressedSecondaryCache : public SecondaryCache {
   CacheAllocationPtr MergeChunksIntoValue(const void* chunks_head,
                                           size_t& charge);
 
-  // TODO: clean up to use cleaner interfaces in typed_cache.h
-  const Cache::CacheItemHelper* GetHelper(bool enable_custom_split_merge) const;
+  // An implementation of Cache::DeleterFn.
+  static Cache::DeleterFn GetDeletionCallback(bool enable_custom_split_merge);
   std::shared_ptr<Cache> cache_;
   CompressedSecondaryCacheOptions cache_options_;
-  mutable port::Mutex capacity_mutex_;
 };
 
 }  // namespace ROCKSDB_NAMESPACE

cache/compressed_secondary_cache_test.cc

@@ -5,43 +5,93 @@
 
 #include "cache/compressed_secondary_cache.h"
 
-#include <array>
+#include <algorithm>
+#include <cstdint>
 #include <iterator>
-#include <memory>
 #include <tuple>
 
+#include "cache/lru_cache.h"
 #include "memory/jemalloc_nodump_allocator.h"
+#include "memory/memory_allocator.h"
+#include "rocksdb/compression_type.h"
 #include "rocksdb/convenience.h"
-#include "test_util/secondary_cache_test_util.h"
+#include "rocksdb/secondary_cache.h"
 #include "test_util/testharness.h"
 #include "test_util/testutil.h"
+#include "util/compression.h"
+#include "util/random.h"
 
 namespace ROCKSDB_NAMESPACE {
 
-using secondary_cache_test_util::GetTestingCacheTypes;
-using secondary_cache_test_util::WithCacheType;
-
-// 16 bytes for HCC compatibility
-const std::string key0 = "____    ____key0";
-const std::string key1 = "____    ____key1";
-const std::string key2 = "____    ____key2";
-const std::string key3 = "____    ____key3";
-
-class CompressedSecondaryCacheTestBase : public testing::Test,
-                                         public WithCacheType {
+class CompressedSecondaryCacheTest : public testing::Test {
  public:
-  CompressedSecondaryCacheTestBase() {}
-  ~CompressedSecondaryCacheTestBase() override = default;
+  CompressedSecondaryCacheTest() : fail_create_(false) {}
+  ~CompressedSecondaryCacheTest() {}
 
  protected:
+  class TestItem {
+   public:
+    TestItem(const char* buf, size_t size) : buf_(new char[size]), size_(size) {
+      memcpy(buf_.get(), buf, size);
+    }
+    ~TestItem() {}
+
+    char* Buf() { return buf_.get(); }
+    size_t Size() { return size_; }
+
+   private:
+    std::unique_ptr<char[]> buf_;
+    size_t size_;
+  };
+
+  static size_t SizeCallback(void* obj) {
+    return reinterpret_cast<TestItem*>(obj)->Size();
+  }
+
+  static Status SaveToCallback(void* from_obj, size_t from_offset,
+                               size_t length, void* out) {
+    TestItem* item = reinterpret_cast<TestItem*>(from_obj);
+    const char* buf = item->Buf();
+    EXPECT_EQ(length, item->Size());
+    EXPECT_EQ(from_offset, 0);
+    memcpy(out, buf, length);
+    return Status::OK();
+  }
+
+  static void DeletionCallback(const Slice& /*key*/, void* obj) {
+    delete reinterpret_cast<TestItem*>(obj);
+    obj = nullptr;
+  }
+
+  static Cache::CacheItemHelper helper_;
+
+  static Status SaveToCallbackFail(void* /*obj*/, size_t /*offset*/,
+                                   size_t /*size*/, void* /*out*/) {
+    return Status::NotSupported();
+  }
+
+  static Cache::CacheItemHelper helper_fail_;
+
+  Cache::CreateCallback test_item_creator = [&](const void* buf, size_t size,
+                                                void** out_obj,
+                                                size_t* charge) -> Status {
+    if (fail_create_) {
+      return Status::NotSupported();
+    }
+    *out_obj = reinterpret_cast<void*>(new TestItem((char*)buf, size));
+    *charge = size;
+    return Status::OK();
+  };
+
+  void SetFailCreate(bool fail) { fail_create_ = fail; }
+
   void BasicTestHelper(std::shared_ptr<SecondaryCache> sec_cache,
                        bool sec_cache_is_compressed) {
     get_perf_context()->Reset();
-    bool kept_in_sec_cache{true};
+    bool is_in_sec_cache{true};
     // Lookup an non-existent key.
-    std::unique_ptr<SecondaryCacheResultHandle> handle0 =
-        sec_cache->Lookup(key0, GetHelper(), this, true, /*advise_erase=*/true,
-                          kept_in_sec_cache);
+    std::unique_ptr<SecondaryCacheResultHandle> handle0 = sec_cache->Lookup(
+        "k0", test_item_creator, true, /*advise_erase=*/true, is_in_sec_cache);
     ASSERT_EQ(handle0, nullptr);
 
     Random rnd(301);
@@ -49,25 +99,25 @@ class CompressedSecondaryCacheTestBase : public testing::Test,
     std::string str1(rnd.RandomString(1000));
     TestItem item1(str1.data(), str1.length());
     // A dummy handle is inserted if the item is inserted for the first time.
-    ASSERT_OK(sec_cache->Insert(key1, &item1, GetHelper()));
+    ASSERT_OK(sec_cache->Insert("k1", &item1,
+                                &CompressedSecondaryCacheTest::helper_));
     ASSERT_EQ(get_perf_context()->compressed_sec_cache_insert_dummy_count, 1);
     ASSERT_EQ(get_perf_context()->compressed_sec_cache_uncompressed_bytes, 0);
     ASSERT_EQ(get_perf_context()->compressed_sec_cache_compressed_bytes, 0);
 
-    std::unique_ptr<SecondaryCacheResultHandle> handle1_1 =
-        sec_cache->Lookup(key1, GetHelper(), this, true, /*advise_erase=*/false,
-                          kept_in_sec_cache);
+    std::unique_ptr<SecondaryCacheResultHandle> handle1_1 = sec_cache->Lookup(
+        "k1", test_item_creator, true, /*advise_erase=*/false, is_in_sec_cache);
     ASSERT_EQ(handle1_1, nullptr);
 
     // Insert and Lookup the item k1 for the second time and advise erasing it.
-    ASSERT_OK(sec_cache->Insert(key1, &item1, GetHelper()));
+    ASSERT_OK(sec_cache->Insert("k1", &item1,
+                                &CompressedSecondaryCacheTest::helper_));
     ASSERT_EQ(get_perf_context()->compressed_sec_cache_insert_real_count, 1);
 
-    std::unique_ptr<SecondaryCacheResultHandle> handle1_2 =
-        sec_cache->Lookup(key1, GetHelper(), this, true, /*advise_erase=*/true,
-                          kept_in_sec_cache);
+    std::unique_ptr<SecondaryCacheResultHandle> handle1_2 = sec_cache->Lookup(
+        "k1", test_item_creator, true, /*advise_erase=*/true, is_in_sec_cache);
     ASSERT_NE(handle1_2, nullptr);
-    ASSERT_FALSE(kept_in_sec_cache);
+    ASSERT_FALSE(is_in_sec_cache);
     if (sec_cache_is_compressed) {
       ASSERT_EQ(get_perf_context()->compressed_sec_cache_uncompressed_bytes,
                 1000);
@@ -84,22 +134,22 @@ class CompressedSecondaryCacheTestBase : public testing::Test,
     ASSERT_EQ(memcmp(val1->Buf(), item1.Buf(), item1.Size()), 0);
 
     // Lookup the item k1 again.
-    std::unique_ptr<SecondaryCacheResultHandle> handle1_3 =
-        sec_cache->Lookup(key1, GetHelper(), this, true, /*advise_erase=*/true,
-                          kept_in_sec_cache);
+    std::unique_ptr<SecondaryCacheResultHandle> handle1_3 = sec_cache->Lookup(
+        "k1", test_item_creator, true, /*advise_erase=*/true, is_in_sec_cache);
     ASSERT_EQ(handle1_3, nullptr);
 
     // Insert and Lookup the item k2.
     std::string str2(rnd.RandomString(1000));
     TestItem item2(str2.data(), str2.length());
-    ASSERT_OK(sec_cache->Insert(key2, &item2, GetHelper()));
+    ASSERT_OK(sec_cache->Insert("k2", &item2,
+                                &CompressedSecondaryCacheTest::helper_));
     ASSERT_EQ(get_perf_context()->compressed_sec_cache_insert_dummy_count, 2);
-    std::unique_ptr<SecondaryCacheResultHandle> handle2_1 =
-        sec_cache->Lookup(key2, GetHelper(), this, true, /*advise_erase=*/false,
-                          kept_in_sec_cache);
+    std::unique_ptr<SecondaryCacheResultHandle> handle2_1 = sec_cache->Lookup(
+        "k2", test_item_creator, true, /*advise_erase=*/false, is_in_sec_cache);
     ASSERT_EQ(handle2_1, nullptr);
 
-    ASSERT_OK(sec_cache->Insert(key2, &item2, GetHelper()));
+    ASSERT_OK(sec_cache->Insert("k2", &item2,
+                                &CompressedSecondaryCacheTest::helper_));
     ASSERT_EQ(get_perf_context()->compressed_sec_cache_insert_real_count, 2);
     if (sec_cache_is_compressed) {
       ASSERT_EQ(get_perf_context()->compressed_sec_cache_uncompressed_bytes,
@@ -110,9 +160,8 @@ class CompressedSecondaryCacheTestBase : public testing::Test,
       ASSERT_EQ(get_perf_context()->compressed_sec_cache_uncompressed_bytes, 0);
       ASSERT_EQ(get_perf_context()->compressed_sec_cache_compressed_bytes, 0);
     }
-    std::unique_ptr<SecondaryCacheResultHandle> handle2_2 =
-        sec_cache->Lookup(key2, GetHelper(), this, true, /*advise_erase=*/false,
-                          kept_in_sec_cache);
+    std::unique_ptr<SecondaryCacheResultHandle> handle2_2 = sec_cache->Lookup(
+        "k2", test_item_creator, true, /*advise_erase=*/false, is_in_sec_cache);
     ASSERT_NE(handle2_2, nullptr);
     std::unique_ptr<TestItem> val2 =
         std::unique_ptr<TestItem>(static_cast<TestItem*>(handle2_2->Value()));
@@ -181,26 +230,28 @@ class CompressedSecondaryCacheTestBase : public testing::Test,
     std::string str1(rnd.RandomString(1000));
     TestItem item1(str1.data(), str1.length());
     // Insert a dummy handle.
-    ASSERT_OK(sec_cache->Insert(key1, &item1, GetHelper()));
+    ASSERT_OK(sec_cache->Insert("k1", &item1,
+                                &CompressedSecondaryCacheTest::helper_));
     // Insert k1.
-    ASSERT_OK(sec_cache->Insert(key1, &item1, GetHelper()));
+    ASSERT_OK(sec_cache->Insert("k1", &item1,
+                                &CompressedSecondaryCacheTest::helper_));
 
     // Insert and Lookup the second item.
     std::string str2(rnd.RandomString(200));
     TestItem item2(str2.data(), str2.length());
     // Insert a dummy handle, k1 is not evicted.
-    ASSERT_OK(sec_cache->Insert(key2, &item2, GetHelper()));
-    bool kept_in_sec_cache{false};
-    std::unique_ptr<SecondaryCacheResultHandle> handle1 =
-        sec_cache->Lookup(key1, GetHelper(), this, true, /*advise_erase=*/false,
-                          kept_in_sec_cache);
+    ASSERT_OK(sec_cache->Insert("k2", &item2,
+                                &CompressedSecondaryCacheTest::helper_));
+    bool is_in_sec_cache{false};
+    std::unique_ptr<SecondaryCacheResultHandle> handle1 = sec_cache->Lookup(
+        "k1", test_item_creator, true, /*advise_erase=*/false, is_in_sec_cache);
     ASSERT_EQ(handle1, nullptr);
 
     // Insert k2 and k1 is evicted.
-    ASSERT_OK(sec_cache->Insert(key2, &item2, GetHelper()));
-    std::unique_ptr<SecondaryCacheResultHandle> handle2 =
-        sec_cache->Lookup(key2, GetHelper(), this, true, /*advise_erase=*/false,
-                          kept_in_sec_cache);
+    ASSERT_OK(sec_cache->Insert("k2", &item2,
+                                &CompressedSecondaryCacheTest::helper_));
+    std::unique_ptr<SecondaryCacheResultHandle> handle2 = sec_cache->Lookup(
+        "k2", test_item_creator, true, /*advise_erase=*/false, is_in_sec_cache);
     ASSERT_NE(handle2, nullptr);
     std::unique_ptr<TestItem> val2 =
         std::unique_ptr<TestItem>(static_cast<TestItem*>(handle2->Value()));
@@ -208,26 +259,27 @@ class CompressedSecondaryCacheTestBase : public testing::Test,
     ASSERT_EQ(memcmp(val2->Buf(), item2.Buf(), item2.Size()), 0);
 
     // Insert k1 again and a dummy handle is inserted.
-    ASSERT_OK(sec_cache->Insert(key1, &item1, GetHelper()));
+    ASSERT_OK(sec_cache->Insert("k1", &item1,
+                                &CompressedSecondaryCacheTest::helper_));
 
-    std::unique_ptr<SecondaryCacheResultHandle> handle1_1 =
-        sec_cache->Lookup(key1, GetHelper(), this, true, /*advise_erase=*/false,
-                          kept_in_sec_cache);
+    std::unique_ptr<SecondaryCacheResultHandle> handle1_1 = sec_cache->Lookup(
+        "k1", test_item_creator, true, /*advise_erase=*/false, is_in_sec_cache);
     ASSERT_EQ(handle1_1, nullptr);
 
     // Create Fails.
     SetFailCreate(true);
-    std::unique_ptr<SecondaryCacheResultHandle> handle2_1 =
-        sec_cache->Lookup(key2, GetHelper(), this, true, /*advise_erase=*/true,
-                          kept_in_sec_cache);
+    std::unique_ptr<SecondaryCacheResultHandle> handle2_1 = sec_cache->Lookup(
+        "k2", test_item_creator, true, /*advise_erase=*/true, is_in_sec_cache);
     ASSERT_EQ(handle2_1, nullptr);
 
     // Save Fails.
     std::string str3 = rnd.RandomString(10);
     TestItem item3(str3.data(), str3.length());
     // The Status is OK because a dummy handle is inserted.
-    ASSERT_OK(sec_cache->Insert(key3, &item3, GetHelperFail()));
-    ASSERT_NOK(sec_cache->Insert(key3, &item3, GetHelperFail()));
+    ASSERT_OK(sec_cache->Insert("k3", &item3,
+                                &CompressedSecondaryCacheTest::helper_fail_));
+    ASSERT_NOK(sec_cache->Insert("k3", &item3,
+                                 &CompressedSecondaryCacheTest::helper_fail_));
 
     sec_cache.reset();
   }
@@ -251,43 +303,52 @@ class CompressedSecondaryCacheTestBase : public testing::Test,
     secondary_cache_opts.enable_custom_split_merge = enable_custom_split_merge;
     std::shared_ptr<SecondaryCache> secondary_cache =
         NewCompressedSecondaryCache(secondary_cache_opts);
-    std::shared_ptr<Cache> cache = NewCache(
+    LRUCacheOptions lru_cache_opts(
         /*_capacity =*/1300, /*_num_shard_bits =*/0,
-        /*_strict_capacity_limit =*/true, secondary_cache);
+        /*_strict_capacity_limit =*/false, /*_high_pri_pool_ratio =*/0.5,
+        /*_memory_allocator =*/nullptr, kDefaultToAdaptiveMutex,
+        kDefaultCacheMetadataChargePolicy, /*_low_pri_pool_ratio =*/0.0);
+    lru_cache_opts.secondary_cache = secondary_cache;
+    std::shared_ptr<Cache> cache = NewLRUCache(lru_cache_opts);
     std::shared_ptr<Statistics> stats = CreateDBStatistics();
 
     get_perf_context()->Reset();
     Random rnd(301);
     std::string str1 = rnd.RandomString(1001);
-    auto item1_1 = new TestItem(str1.data(), str1.length());
-    ASSERT_OK(cache->Insert(key1, item1_1, GetHelper(), str1.length()));
+    TestItem* item1_1 = new TestItem(str1.data(), str1.length());
+    ASSERT_OK(cache->Insert(
+        "k1", item1_1, &CompressedSecondaryCacheTest::helper_, str1.length()));
 
     std::string str2 = rnd.RandomString(1012);
-    auto item2_1 = new TestItem(str2.data(), str2.length());
+    TestItem* item2_1 = new TestItem(str2.data(), str2.length());
     // After this Insert, primary cache contains k2 and secondary cache contains
     // k1's dummy item.
-    ASSERT_OK(cache->Insert(key2, item2_1, GetHelper(), str2.length()));
+    ASSERT_OK(cache->Insert(
+        "k2", item2_1, &CompressedSecondaryCacheTest::helper_, str2.length()));
     ASSERT_EQ(get_perf_context()->compressed_sec_cache_insert_dummy_count, 1);
     ASSERT_EQ(get_perf_context()->compressed_sec_cache_uncompressed_bytes, 0);
     ASSERT_EQ(get_perf_context()->compressed_sec_cache_compressed_bytes, 0);
 
     std::string str3 = rnd.RandomString(1024);
-    auto item3_1 = new TestItem(str3.data(), str3.length());
+    TestItem* item3_1 = new TestItem(str3.data(), str3.length());
     // After this Insert, primary cache contains k3 and secondary cache contains
     // k1's dummy item and k2's dummy item.
-    ASSERT_OK(cache->Insert(key3, item3_1, GetHelper(), str3.length()));
+    ASSERT_OK(cache->Insert(
+        "k3", item3_1, &CompressedSecondaryCacheTest::helper_, str3.length()));
     ASSERT_EQ(get_perf_context()->compressed_sec_cache_insert_dummy_count, 2);
 
     // After this Insert, primary cache contains k1 and secondary cache contains
     // k1's dummy item, k2's dummy item, and k3's dummy item.
-    auto item1_2 = new TestItem(str1.data(), str1.length());
-    ASSERT_OK(cache->Insert(key1, item1_2, GetHelper(), str1.length()));
+    TestItem* item1_2 = new TestItem(str1.data(), str1.length());
+    ASSERT_OK(cache->Insert(
+        "k1", item1_2, &CompressedSecondaryCacheTest::helper_, str1.length()));
     ASSERT_EQ(get_perf_context()->compressed_sec_cache_insert_dummy_count, 3);
 
     // After this Insert, primary cache contains k2 and secondary cache contains
     // k1's item, k2's dummy item, and k3's dummy item.
-    auto item2_2 = new TestItem(str2.data(), str2.length());
-    ASSERT_OK(cache->Insert(key2, item2_2, GetHelper(), str2.length()));
+    TestItem* item2_2 = new TestItem(str2.data(), str2.length());
+    ASSERT_OK(cache->Insert(
+        "k2", item2_2, &CompressedSecondaryCacheTest::helper_, str2.length()));
     ASSERT_EQ(get_perf_context()->compressed_sec_cache_insert_real_count, 1);
     if (sec_cache_is_compressed) {
       ASSERT_EQ(get_perf_context()->compressed_sec_cache_uncompressed_bytes,
@@ -301,8 +362,9 @@ class CompressedSecondaryCacheTestBase : public testing::Test,
 
     // After this Insert, primary cache contains k3 and secondary cache contains
     // k1's item and k2's item.
-    auto item3_2 = new TestItem(str3.data(), str3.length());
-    ASSERT_OK(cache->Insert(key3, item3_2, GetHelper(), str3.length()));
+    TestItem* item3_2 = new TestItem(str3.data(), str3.length());
+    ASSERT_OK(cache->Insert(
+        "k3", item3_2, &CompressedSecondaryCacheTest::helper_, str3.length()));
     ASSERT_EQ(get_perf_context()->compressed_sec_cache_insert_real_count, 2);
     if (sec_cache_is_compressed) {
       ASSERT_EQ(get_perf_context()->compressed_sec_cache_uncompressed_bytes,
@@ -315,89 +377,50 @@ class CompressedSecondaryCacheTestBase : public testing::Test,
     }
 
     Cache::Handle* handle;
-    handle = cache->Lookup(key3, GetHelper(), this, Cache::Priority::LOW,
+    handle = cache->Lookup("k3", &CompressedSecondaryCacheTest::helper_,
+                           test_item_creator, Cache::Priority::LOW, true,
                            stats.get());
     ASSERT_NE(handle, nullptr);
-    auto val3 = static_cast<TestItem*>(cache->Value(handle));
+    TestItem* val3 = static_cast<TestItem*>(cache->Value(handle));
     ASSERT_NE(val3, nullptr);
     ASSERT_EQ(memcmp(val3->Buf(), item3_2->Buf(), item3_2->Size()), 0);
     cache->Release(handle);
 
     // Lookup an non-existent key.
-    handle = cache->Lookup(key0, GetHelper(), this, Cache::Priority::LOW,
+    handle = cache->Lookup("k0", &CompressedSecondaryCacheTest::helper_,
+                           test_item_creator, Cache::Priority::LOW, true,
                            stats.get());
     ASSERT_EQ(handle, nullptr);
 
     // This Lookup should just insert a dummy handle in the primary cache
     // and the k1 is still in the secondary cache.
-    handle = cache->Lookup(key1, GetHelper(), this, Cache::Priority::LOW,
+    handle = cache->Lookup("k1", &CompressedSecondaryCacheTest::helper_,
+                           test_item_creator, Cache::Priority::LOW, true,
                            stats.get());
     ASSERT_NE(handle, nullptr);
     ASSERT_EQ(get_perf_context()->block_cache_standalone_handle_count, 1);
-    auto val1_1 = static_cast<TestItem*>(cache->Value(handle));
+    TestItem* val1_1 = static_cast<TestItem*>(cache->Value(handle));
     ASSERT_NE(val1_1, nullptr);
     ASSERT_EQ(memcmp(val1_1->Buf(), str1.data(), str1.size()), 0);
     cache->Release(handle);
 
     // This Lookup should erase k1 from the secondary cache and insert
     // it into primary cache; then k3 is demoted.
-    // k2 and k3 are in secondary cache.
-    handle = cache->Lookup(key1, GetHelper(), this, Cache::Priority::LOW,
+    handle = cache->Lookup("k1", &CompressedSecondaryCacheTest::helper_,
+                           test_item_creator, Cache::Priority::LOW, true,
                            stats.get());
     ASSERT_NE(handle, nullptr);
     ASSERT_EQ(get_perf_context()->block_cache_standalone_handle_count, 1);
-    ASSERT_EQ(get_perf_context()->compressed_sec_cache_insert_real_count, 3);
     cache->Release(handle);
 
     // k2 is still in secondary cache.
-    handle = cache->Lookup(key2, GetHelper(), this, Cache::Priority::LOW,
+    handle = cache->Lookup("k2", &CompressedSecondaryCacheTest::helper_,
+                           test_item_creator, Cache::Priority::LOW, true,
                            stats.get());
     ASSERT_NE(handle, nullptr);
     ASSERT_EQ(get_perf_context()->block_cache_standalone_handle_count, 2);
     cache->Release(handle);
 
-    // Testing SetCapacity().
-    ASSERT_OK(secondary_cache->SetCapacity(0));
-    handle = cache->Lookup(key3, GetHelper(), this, Cache::Priority::LOW,
-                           stats.get());
-    ASSERT_EQ(handle, nullptr);
-
-    ASSERT_OK(secondary_cache->SetCapacity(7000));
-    size_t capacity;
-    ASSERT_OK(secondary_cache->GetCapacity(capacity));
-    ASSERT_EQ(capacity, 7000);
-    auto item1_3 = new TestItem(str1.data(), str1.length());
-    // After this Insert, primary cache contains k1.
-    ASSERT_OK(cache->Insert(key1, item1_3, GetHelper(), str2.length()));
-    ASSERT_EQ(get_perf_context()->compressed_sec_cache_insert_dummy_count, 3);
-    ASSERT_EQ(get_perf_context()->compressed_sec_cache_insert_real_count, 4);
-
-    auto item2_3 = new TestItem(str2.data(), str2.length());
-    // After this Insert, primary cache contains k2 and secondary cache contains
-    // k1's dummy item.
-    ASSERT_OK(cache->Insert(key2, item2_3, GetHelper(), str1.length()));
-    ASSERT_EQ(get_perf_context()->compressed_sec_cache_insert_dummy_count, 4);
-
-    auto item1_4 = new TestItem(str1.data(), str1.length());
-    // After this Insert, primary cache contains k1 and secondary cache contains
-    // k1's dummy item and k2's dummy item.
-    ASSERT_OK(cache->Insert(key1, item1_4, GetHelper(), str2.length()));
-    ASSERT_EQ(get_perf_context()->compressed_sec_cache_insert_dummy_count, 5);
-
-    auto item2_4 = new TestItem(str2.data(), str2.length());
-    // After this Insert, primary cache contains k2 and secondary cache contains
-    // k1's real item and k2's dummy item.
-    ASSERT_OK(cache->Insert(key2, item2_4, GetHelper(), str2.length()));
-    ASSERT_EQ(get_perf_context()->compressed_sec_cache_insert_real_count, 5);
-    // This Lookup should just insert a dummy handle in the primary cache
-    // and the k1 is still in the secondary cache.
-    handle = cache->Lookup(key1, GetHelper(), this, Cache::Priority::LOW,
-                           stats.get());
-
-    ASSERT_NE(handle, nullptr);
-    cache->Release(handle);
-    ASSERT_EQ(get_perf_context()->block_cache_standalone_handle_count, 3);
-
     cache.reset();
     secondary_cache.reset();
   }
@@ -419,31 +442,32 @@ class CompressedSecondaryCacheTestBase : public testing::Test,
     std::shared_ptr<SecondaryCache> secondary_cache =
         NewCompressedSecondaryCache(secondary_cache_opts);
 
-    std::shared_ptr<Cache> cache = NewCache(
+    LRUCacheOptions opts(
         /*_capacity=*/1300, /*_num_shard_bits=*/0,
-        /*_strict_capacity_limit=*/false, secondary_cache);
+        /*_strict_capacity_limit=*/false, /*_high_pri_pool_ratio=*/0.5,
+        /*_memory_allocator=*/nullptr, kDefaultToAdaptiveMutex,
+        kDefaultCacheMetadataChargePolicy, /*_low_pri_pool_ratio=*/0.0);
+    opts.secondary_cache = secondary_cache;
+    std::shared_ptr<Cache> cache = NewLRUCache(opts);
 
     Random rnd(301);
     std::string str1 = rnd.RandomString(1001);
-    auto item1 = std::make_unique<TestItem>(str1.data(), str1.length());
-    ASSERT_OK(cache->Insert(key1, item1.get(), GetHelper(), str1.length()));
+    auto item1 =
+        std::unique_ptr<TestItem>(new TestItem(str1.data(), str1.length()));
+    ASSERT_NOK(cache->Insert("k1", item1.get(), nullptr, str1.length()));
+    ASSERT_OK(cache->Insert("k1", item1.get(),
+                            &CompressedSecondaryCacheTest::helper_,
+                            str1.length()));
     item1.release();  // Appease clang-analyze "potential memory leak"
 
     Cache::Handle* handle;
-    handle = cache->Lookup(key2, nullptr, this, Cache::Priority::LOW);
+    handle = cache->Lookup("k2", nullptr, test_item_creator,
+                           Cache::Priority::LOW, true);
     ASSERT_EQ(handle, nullptr);
-    handle = cache->Lookup(key2, GetHelper(), this, Cache::Priority::LOW);
+    handle = cache->Lookup("k2", &CompressedSecondaryCacheTest::helper_,
+                           test_item_creator, Cache::Priority::LOW, false);
     ASSERT_EQ(handle, nullptr);
 
-    Cache::AsyncLookupHandle ah;
-    ah.key = key2;
-    ah.helper = GetHelper();
-    ah.create_context = this;
-    ah.priority = Cache::Priority::LOW;
-    cache->StartAsyncLookup(ah);
-    cache->Wait(ah);
-    ASSERT_EQ(ah.Result(), nullptr);
-
     cache.reset();
     secondary_cache.reset();
   }
@@ -466,29 +490,40 @@ class CompressedSecondaryCacheTestBase : public testing::Test,
     std::shared_ptr<SecondaryCache> secondary_cache =
         NewCompressedSecondaryCache(secondary_cache_opts);
 
-    std::shared_ptr<Cache> cache = NewCache(
+    LRUCacheOptions opts(
         /*_capacity=*/1300, /*_num_shard_bits=*/0,
-        /*_strict_capacity_limit=*/true, secondary_cache);
+        /*_strict_capacity_limit=*/false, /*_high_pri_pool_ratio=*/0.5,
+        /*_memory_allocator=*/nullptr, kDefaultToAdaptiveMutex,
+        kDefaultCacheMetadataChargePolicy, /*_low_pri_pool_ratio=*/0.0);
+    opts.secondary_cache = secondary_cache;
+    std::shared_ptr<Cache> cache = NewLRUCache(opts);
 
     Random rnd(301);
     std::string str1 = rnd.RandomString(1001);
-    auto item1 = new TestItem(str1.data(), str1.length());
-    ASSERT_OK(cache->Insert(key1, item1, GetHelperFail(), str1.length()));
+    TestItem* item1 = new TestItem(str1.data(), str1.length());
+    ASSERT_OK(cache->Insert("k1", item1,
+                            &CompressedSecondaryCacheTest::helper_fail_,
+                            str1.length()));
 
     std::string str2 = rnd.RandomString(1002);
-    auto item2 = new TestItem(str2.data(), str2.length());
+    TestItem* item2 = new TestItem(str2.data(), str2.length());
     // k1 should be demoted to the secondary cache.
-    ASSERT_OK(cache->Insert(key2, item2, GetHelperFail(), str2.length()));
+    ASSERT_OK(cache->Insert("k2", item2,
+                            &CompressedSecondaryCacheTest::helper_fail_,
+                            str2.length()));
 
     Cache::Handle* handle;
-    handle = cache->Lookup(key2, GetHelperFail(), this, Cache::Priority::LOW);
+    handle = cache->Lookup("k2", &CompressedSecondaryCacheTest::helper_fail_,
+                           test_item_creator, Cache::Priority::LOW, true);
     ASSERT_NE(handle, nullptr);
     cache->Release(handle);
     // This lookup should fail, since k1 demotion would have failed.
-    handle = cache->Lookup(key1, GetHelperFail(), this, Cache::Priority::LOW);
+    handle = cache->Lookup("k1", &CompressedSecondaryCacheTest::helper_fail_,
+                           test_item_creator, Cache::Priority::LOW, true);
     ASSERT_EQ(handle, nullptr);
     // Since k1 was not promoted, k2 should still be in cache.
-    handle = cache->Lookup(key2, GetHelperFail(), this, Cache::Priority::LOW);
+    handle = cache->Lookup("k2", &CompressedSecondaryCacheTest::helper_fail_,
+                           test_item_creator, Cache::Priority::LOW, true);
     ASSERT_NE(handle, nullptr);
     cache->Release(handle);
 
@@ -514,30 +549,39 @@ class CompressedSecondaryCacheTestBase : public testing::Test,
     std::shared_ptr<SecondaryCache> secondary_cache =
         NewCompressedSecondaryCache(secondary_cache_opts);
 
-    std::shared_ptr<Cache> cache = NewCache(
+    LRUCacheOptions opts(
         /*_capacity=*/1300, /*_num_shard_bits=*/0,
-        /*_strict_capacity_limit=*/true, secondary_cache);
+        /*_strict_capacity_limit=*/false, /*_high_pri_pool_ratio=*/0.5,
+        /*_memory_allocator=*/nullptr, kDefaultToAdaptiveMutex,
+        kDefaultCacheMetadataChargePolicy, /*_low_pri_pool_ratio=*/0.0);
+    opts.secondary_cache = secondary_cache;
+    std::shared_ptr<Cache> cache = NewLRUCache(opts);
 
     Random rnd(301);
     std::string str1 = rnd.RandomString(1001);
-    auto item1 = new TestItem(str1.data(), str1.length());
-    ASSERT_OK(cache->Insert(key1, item1, GetHelper(), str1.length()));
+    TestItem* item1 = new TestItem(str1.data(), str1.length());
+    ASSERT_OK(cache->Insert("k1", item1, &CompressedSecondaryCacheTest::helper_,
+                            str1.length()));
 
     std::string str2 = rnd.RandomString(1002);
-    auto item2 = new TestItem(str2.data(), str2.length());
+    TestItem* item2 = new TestItem(str2.data(), str2.length());
     // k1 should be demoted to the secondary cache.
-    ASSERT_OK(cache->Insert(key2, item2, GetHelper(), str2.length()));
+    ASSERT_OK(cache->Insert("k2", item2, &CompressedSecondaryCacheTest::helper_,
+                            str2.length()));
 
     Cache::Handle* handle;
     SetFailCreate(true);
-    handle = cache->Lookup(key2, GetHelper(), this, Cache::Priority::LOW);
+    handle = cache->Lookup("k2", &CompressedSecondaryCacheTest::helper_,
+                           test_item_creator, Cache::Priority::LOW, true);
     ASSERT_NE(handle, nullptr);
     cache->Release(handle);
     // This lookup should fail, since k1 creation would have failed
-    handle = cache->Lookup(key1, GetHelper(), this, Cache::Priority::LOW);
+    handle = cache->Lookup("k1", &CompressedSecondaryCacheTest::helper_,
+                           test_item_creator, Cache::Priority::LOW, true);
     ASSERT_EQ(handle, nullptr);
     // Since k1 didn't get promoted, k2 should still be in cache
-    handle = cache->Lookup(key2, GetHelper(), this, Cache::Priority::LOW);
+    handle = cache->Lookup("k2", &CompressedSecondaryCacheTest::helper_,
+                           test_item_creator, Cache::Priority::LOW, true);
     ASSERT_NE(handle, nullptr);
     cache->Release(handle);
 
@@ -563,34 +607,43 @@ class CompressedSecondaryCacheTestBase : public testing::Test,
     std::shared_ptr<SecondaryCache> secondary_cache =
         NewCompressedSecondaryCache(secondary_cache_opts);
 
-    std::shared_ptr<Cache> cache = NewCache(
+    LRUCacheOptions opts(
         /*_capacity=*/1300, /*_num_shard_bits=*/0,
-        /*_strict_capacity_limit=*/false, secondary_cache);
+        /*_strict_capacity_limit=*/false, /*_high_pri_pool_ratio=*/0.5,
+        /*_memory_allocator=*/nullptr, kDefaultToAdaptiveMutex,
+        kDefaultCacheMetadataChargePolicy, /*_low_pri_pool_ratio=*/0.0);
+    opts.secondary_cache = secondary_cache;
+    std::shared_ptr<Cache> cache = NewLRUCache(opts);
 
     Random rnd(301);
     std::string str1 = rnd.RandomString(1001);
-    auto item1_1 = new TestItem(str1.data(), str1.length());
-    ASSERT_OK(cache->Insert(key1, item1_1, GetHelper(), str1.length()));
+    TestItem* item1_1 = new TestItem(str1.data(), str1.length());
+    ASSERT_OK(cache->Insert(
+        "k1", item1_1, &CompressedSecondaryCacheTest::helper_, str1.length()));
 
     std::string str2 = rnd.RandomString(1002);
     std::string str2_clone{str2};
-    auto item2 = new TestItem(str2.data(), str2.length());
+    TestItem* item2 = new TestItem(str2.data(), str2.length());
     // After this Insert, primary cache contains k2 and secondary cache contains
     // k1's dummy item.
-    ASSERT_OK(cache->Insert(key2, item2, GetHelper(), str2.length()));
+    ASSERT_OK(cache->Insert("k2", item2, &CompressedSecondaryCacheTest::helper_,
+                            str2.length()));
 
     // After this Insert, primary cache contains k1 and secondary cache contains
     // k1's dummy item and k2's dummy item.
-    auto item1_2 = new TestItem(str1.data(), str1.length());
-    ASSERT_OK(cache->Insert(key1, item1_2, GetHelper(), str1.length()));
+    TestItem* item1_2 = new TestItem(str1.data(), str1.length());
+    ASSERT_OK(cache->Insert(
+        "k1", item1_2, &CompressedSecondaryCacheTest::helper_, str1.length()));
 
-    auto item2_2 = new TestItem(str2.data(), str2.length());
+    TestItem* item2_2 = new TestItem(str2.data(), str2.length());
     // After this Insert, primary cache contains k2 and secondary cache contains
     // k1's item and k2's dummy item.
-    ASSERT_OK(cache->Insert(key2, item2_2, GetHelper(), str2.length()));
+    ASSERT_OK(cache->Insert(
+        "k2", item2_2, &CompressedSecondaryCacheTest::helper_, str2.length()));
 
     Cache::Handle* handle2;
-    handle2 = cache->Lookup(key2, GetHelper(), this, Cache::Priority::LOW);
+    handle2 = cache->Lookup("k2", &CompressedSecondaryCacheTest::helper_,
+                            test_item_creator, Cache::Priority::LOW, true);
     ASSERT_NE(handle2, nullptr);
     cache->Release(handle2);
 
@@ -598,12 +651,14 @@ class CompressedSecondaryCacheTestBase : public testing::Test,
     // strict_capacity_limit is true, but the lookup should still succeed.
     // A k1's dummy item is inserted into primary cache.
     Cache::Handle* handle1;
-    handle1 = cache->Lookup(key1, GetHelper(), this, Cache::Priority::LOW);
+    handle1 = cache->Lookup("k1", &CompressedSecondaryCacheTest::helper_,
+                            test_item_creator, Cache::Priority::LOW, true);
     ASSERT_NE(handle1, nullptr);
     cache->Release(handle1);
 
     // Since k1 didn't get inserted, k2 should still be in cache
-    handle2 = cache->Lookup(key2, GetHelper(), this, Cache::Priority::LOW);
+    handle2 = cache->Lookup("k2", &CompressedSecondaryCacheTest::helper_,
+                            test_item_creator, Cache::Priority::LOW, true);
     ASSERT_NE(handle2, nullptr);
     cache->Release(handle2);
 
@@ -644,7 +699,7 @@ class CompressedSecondaryCacheTestBase : public testing::Test,
     current_chunk = current_chunk->next;
     ASSERT_EQ(current_chunk->size, 98);
 
-    sec_cache->GetHelper(true)->del_cb(chunks_head, /*alloc*/ nullptr);
+    sec_cache->GetDeletionCallback(true)("dummy", chunks_head);
   }
 
   void MergeChunksIntoValueTest() {
@@ -725,29 +780,31 @@ class CompressedSecondaryCacheTestBase : public testing::Test,
     std::string value_str{value.get(), charge};
     ASSERT_EQ(strcmp(value_str.data(), str.data()), 0);
 
-    sec_cache->GetHelper(true)->del_cb(chunks_head, /*alloc*/ nullptr);
+    sec_cache->GetDeletionCallback(true)("dummy", chunks_head);
   }
+
+ private:
+  bool fail_create_;
 };
 
-class CompressedSecondaryCacheTest
-    : public CompressedSecondaryCacheTestBase,
-      public testing::WithParamInterface<std::string> {
-  const std::string& Type() override { return GetParam(); }
-};
+Cache::CacheItemHelper CompressedSecondaryCacheTest::helper_(
+    CompressedSecondaryCacheTest::SizeCallback,
+    CompressedSecondaryCacheTest::SaveToCallback,
+    CompressedSecondaryCacheTest::DeletionCallback);
 
-INSTANTIATE_TEST_CASE_P(CompressedSecondaryCacheTest,
-                        CompressedSecondaryCacheTest, GetTestingCacheTypes());
+Cache::CacheItemHelper CompressedSecondaryCacheTest::helper_fail_(
+    CompressedSecondaryCacheTest::SizeCallback,
+    CompressedSecondaryCacheTest::SaveToCallbackFail,
+    CompressedSecondaryCacheTest::DeletionCallback);
 
 class CompressedSecCacheTestWithCompressAndAllocatorParam
-    : public CompressedSecondaryCacheTestBase,
-      public ::testing::WithParamInterface<
-          std::tuple<bool, bool, std::string>> {
+    : public CompressedSecondaryCacheTest,
+      public ::testing::WithParamInterface<std::tuple<bool, bool>> {
  public:
   CompressedSecCacheTestWithCompressAndAllocatorParam() {
     sec_cache_is_compressed_ = std::get<0>(GetParam());
     use_jemalloc_ = std::get<1>(GetParam());
   }
-  const std::string& Type() override { return std::get<2>(GetParam()); }
   bool sec_cache_is_compressed_;
   bool use_jemalloc_;
 };
@@ -758,20 +815,20 @@ TEST_P(CompressedSecCacheTestWithCompressAndAllocatorParam, BasicTes) {
 
 INSTANTIATE_TEST_CASE_P(CompressedSecCacheTests,
                         CompressedSecCacheTestWithCompressAndAllocatorParam,
-                        ::testing::Combine(testing::Bool(), testing::Bool(),
-                                           GetTestingCacheTypes()));
+                        ::testing::Combine(testing::Bool(), testing::Bool()));
 
 class CompressedSecondaryCacheTestWithCompressionParam
-    : public CompressedSecondaryCacheTestBase,
-      public ::testing::WithParamInterface<std::tuple<bool, std::string>> {
+    : public CompressedSecondaryCacheTest,
+      public ::testing::WithParamInterface<bool> {
  public:
   CompressedSecondaryCacheTestWithCompressionParam() {
-    sec_cache_is_compressed_ = std::get<0>(GetParam());
+    sec_cache_is_compressed_ = GetParam();
   }
-  const std::string& Type() override { return std::get<1>(GetParam()); }
   bool sec_cache_is_compressed_;
 };
 
+#ifndef ROCKSDB_LITE
+
 TEST_P(CompressedSecondaryCacheTestWithCompressionParam, BasicTestFromString) {
   std::shared_ptr<SecondaryCache> sec_cache{nullptr};
   std::string sec_cache_uri;
@@ -835,6 +892,7 @@ TEST_P(CompressedSecondaryCacheTestWithCompressionParam,
   BasicTestHelper(sec_cache, sec_cache_is_compressed_);
 }
 
+#endif  // ROCKSDB_LITE
 
 TEST_P(CompressedSecondaryCacheTestWithCompressionParam, FailsTest) {
   FailsTest(sec_cache_is_compressed_);
@@ -860,92 +918,18 @@ TEST_P(CompressedSecondaryCacheTestWithCompressionParam,
   IntegrationFullCapacityTest(sec_cache_is_compressed_);
 }
 
-TEST_P(CompressedSecondaryCacheTestWithCompressionParam, EntryRoles) {
-  CompressedSecondaryCacheOptions opts;
-  opts.capacity = 2048;
-  opts.num_shard_bits = 0;
-
-  if (sec_cache_is_compressed_) {
-    if (!LZ4_Supported()) {
-      ROCKSDB_GTEST_SKIP("This test requires LZ4 support.");
-      return;
-    }
-  } else {
-    opts.compression_type = CompressionType::kNoCompression;
-  }
-
-  // Select a random subset to include, for fast test
-  Random& r = *Random::GetTLSInstance();
-  CacheEntryRoleSet do_not_compress;
-  for (uint32_t i = 0; i < kNumCacheEntryRoles; ++i) {
-    // A few included on average, but decent chance of zero
-    if (r.OneIn(5)) {
-      do_not_compress.Add(static_cast<CacheEntryRole>(i));
-    }
-  }
-  opts.do_not_compress_roles = do_not_compress;
-
-  std::shared_ptr<SecondaryCache> sec_cache = NewCompressedSecondaryCache(opts);
-
-  // Fixed seed to ensure consistent compressibility (doesn't compress)
-  std::string junk(Random(301).RandomString(1000));
-
-  for (uint32_t i = 0; i < kNumCacheEntryRoles; ++i) {
-    CacheEntryRole role = static_cast<CacheEntryRole>(i);
-
-    // Uniquify `junk`
-    junk[0] = static_cast<char>(i);
-    TestItem item{junk.data(), junk.length()};
-    Slice ith_key = Slice(junk.data(), 16);
-
-    get_perf_context()->Reset();
-    ASSERT_OK(sec_cache->Insert(ith_key, &item, GetHelper(role)));
-    ASSERT_EQ(get_perf_context()->compressed_sec_cache_insert_dummy_count, 1U);
-
-    ASSERT_OK(sec_cache->Insert(ith_key, &item, GetHelper(role)));
-    ASSERT_EQ(get_perf_context()->compressed_sec_cache_insert_real_count, 1U);
-
-    bool kept_in_sec_cache{true};
-    std::unique_ptr<SecondaryCacheResultHandle> handle =
-        sec_cache->Lookup(ith_key, GetHelper(role), this, true,
-                          /*advise_erase=*/true, kept_in_sec_cache);
-    ASSERT_NE(handle, nullptr);
-
-    // Lookup returns the right data
-    std::unique_ptr<TestItem> val =
-        std::unique_ptr<TestItem>(static_cast<TestItem*>(handle->Value()));
-    ASSERT_NE(val, nullptr);
-    ASSERT_EQ(memcmp(val->Buf(), item.Buf(), item.Size()), 0);
-
-    bool compressed =
-        sec_cache_is_compressed_ && !do_not_compress.Contains(role);
-    if (compressed) {
-      ASSERT_EQ(get_perf_context()->compressed_sec_cache_uncompressed_bytes,
-                1000);
-      ASSERT_EQ(get_perf_context()->compressed_sec_cache_compressed_bytes,
-                1007);
-    } else {
-      ASSERT_EQ(get_perf_context()->compressed_sec_cache_uncompressed_bytes, 0);
-      ASSERT_EQ(get_perf_context()->compressed_sec_cache_compressed_bytes, 0);
-    }
-  }
-}
-
 INSTANTIATE_TEST_CASE_P(CompressedSecCacheTests,
                         CompressedSecondaryCacheTestWithCompressionParam,
-                        testing::Combine(testing::Bool(),
-                                         GetTestingCacheTypes()));
+                        testing::Bool());
 
 class CompressedSecCacheTestWithCompressAndSplitParam
-    : public CompressedSecondaryCacheTestBase,
-      public ::testing::WithParamInterface<
-          std::tuple<bool, bool, std::string>> {
+    : public CompressedSecondaryCacheTest,
+      public ::testing::WithParamInterface<std::tuple<bool, bool>> {
  public:
   CompressedSecCacheTestWithCompressAndSplitParam() {
     sec_cache_is_compressed_ = std::get<0>(GetParam());
     enable_custom_split_merge_ = std::get<1>(GetParam());
   }
-  const std::string& Type() override { return std::get<2>(GetParam()); }
   bool sec_cache_is_compressed_;
   bool enable_custom_split_merge_;
 };
@@ -956,25 +940,23 @@ TEST_P(CompressedSecCacheTestWithCompressAndSplitParam, BasicIntegrationTest) {
 
 INSTANTIATE_TEST_CASE_P(CompressedSecCacheTests,
                         CompressedSecCacheTestWithCompressAndSplitParam,
-                        ::testing::Combine(testing::Bool(), testing::Bool(),
-                                           GetTestingCacheTypes()));
+                        ::testing::Combine(testing::Bool(), testing::Bool()));
 
-TEST_P(CompressedSecondaryCacheTest, SplitValueIntoChunksTest) {
+TEST_F(CompressedSecondaryCacheTest, SplitValueIntoChunksTest) {
   SplitValueIntoChunksTest();
 }
 
-TEST_P(CompressedSecondaryCacheTest, MergeChunksIntoValueTest) {
+TEST_F(CompressedSecondaryCacheTest, MergeChunksIntoValueTest) {
   MergeChunksIntoValueTest();
 }
 
-TEST_P(CompressedSecondaryCacheTest, SplictValueAndMergeChunksTest) {
+TEST_F(CompressedSecondaryCacheTest, SplictValueAndMergeChunksTest) {
   SplictValueAndMergeChunksTest();
 }
 
 }  // namespace ROCKSDB_NAMESPACE
 
 int main(int argc, char** argv) {
-  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
   ::testing::InitGoogleTest(&argc, argv);
   return RUN_ALL_TESTS();
 }

cache/fast_lru_cache.cc

@@ -0,0 +1,622 @@
+//  Copyright (c) 2011-present, Facebook, Inc.  All rights reserved.
+//  This source code is licensed under both the GPLv2 (found in the
+//  COPYING file in the root directory) and Apache 2.0 License
+//  (found in the LICENSE.Apache file in the root directory).
+//
+// Copyright (c) 2011 The LevelDB Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file. See the AUTHORS file for names of contributors.
+
+#include "cache/fast_lru_cache.h"
+
+#include <cassert>
+#include <cstdint>
+#include <cstdio>
+#include <functional>
+
+#include "monitoring/perf_context_imp.h"
+#include "monitoring/statistics.h"
+#include "port/lang.h"
+#include "util/distributed_mutex.h"
+#include "util/hash.h"
+#include "util/math.h"
+#include "util/random.h"
+
+namespace ROCKSDB_NAMESPACE {
+
+namespace fast_lru_cache {
+
+LRUHandleTable::LRUHandleTable(int hash_bits)
+    : length_bits_(hash_bits),
+      length_bits_mask_((uint32_t{1} << length_bits_) - 1),
+      occupancy_(0),
+      occupancy_limit_(static_cast<uint32_t>((uint32_t{1} << length_bits_) *
+                                             kStrictLoadFactor)),
+      array_(new LRUHandle[size_t{1} << length_bits_]) {
+  assert(hash_bits <= 32);
+}
+
+LRUHandleTable::~LRUHandleTable() {
+  ApplyToEntriesRange([](LRUHandle* h) { h->FreeData(); }, 0, GetTableSize());
+}
+
+LRUHandle* LRUHandleTable::Lookup(const Slice& key, uint32_t hash) {
+  int probe = 0;
+  int slot = FindVisibleElement(key, hash, probe, 0);
+  return (slot == -1) ? nullptr : &array_[slot];
+}
+
+LRUHandle* LRUHandleTable::Insert(LRUHandle* h, LRUHandle** old) {
+  int probe = 0;
+  int slot = FindVisibleElementOrAvailableSlot(h->key(), h->hash, probe,
+                                               1 /*displacement*/);
+  *old = nullptr;
+  if (slot == -1) {
+    // TODO(Guido) Don't we need to roll back displacements here?
+    return nullptr;
+  }
+
+  if (array_[slot].IsEmpty() || array_[slot].IsTombstone()) {
+    bool empty = array_[slot].IsEmpty();
+    Assign(slot, h);
+    LRUHandle* new_entry = &array_[slot];
+    if (empty) {
+      // This used to be an empty slot.
+      return new_entry;
+    }
+    // It used to be a tombstone, so there may already be a copy of the
+    // key in the table.
+    slot = FindVisibleElement(h->key(), h->hash, probe, 0 /*displacement*/);
+    if (slot == -1) {
+      // No existing copy of the key.
+      return new_entry;
+    }
+    *old = &array_[slot];
+    return new_entry;
+  } else {
+    // There is an existing copy of the key.
+    *old = &array_[slot];
+    // Find an available slot for the new element.
+    array_[slot].displacements++;
+    slot = FindAvailableSlot(h->key(), probe, 1 /*displacement*/);
+    if (slot == -1) {
+      // No available slots. Roll back displacements.
+      probe = 0;
+      slot = FindVisibleElement(h->key(), h->hash, probe, -1);
+      array_[slot].displacements--;
+      FindAvailableSlot(h->key(), probe, -1);
+      return nullptr;
+    }
+    Assign(slot, h);
+    return &array_[slot];
+  }
+}
+
+void LRUHandleTable::Remove(LRUHandle* h) {
+  assert(h->next == nullptr &&
+         h->prev == nullptr);  // Already off the LRU list.
+  int probe = 0;
+  FindSlot(
+      h->key(), [&h](LRUHandle* e) { return e == h; }, probe,
+      -1 /*displacement*/);
+  h->SetIsVisible(false);
+  h->SetIsElement(false);
+  occupancy_--;
+}
+
+void LRUHandleTable::Assign(int slot, LRUHandle* h) {
+  LRUHandle* dst = &array_[slot];
+  uint32_t disp = dst->displacements;
+  *dst = *h;
+  dst->displacements = disp;
+  dst->SetIsVisible(true);
+  dst->SetIsElement(true);
+  occupancy_++;
+}
+
+void LRUHandleTable::Exclude(LRUHandle* h) { h->SetIsVisible(false); }
+
+int LRUHandleTable::FindVisibleElement(const Slice& key, uint32_t hash,
+                                       int& probe, int displacement) {
+  return FindSlot(
+      key,
+      [&](LRUHandle* h) { return h->Matches(key, hash) && h->IsVisible(); },
+      probe, displacement);
+}
+
+int LRUHandleTable::FindAvailableSlot(const Slice& key, int& probe,
+                                      int displacement) {
+  return FindSlot(
+      key, [](LRUHandle* h) { return h->IsEmpty() || h->IsTombstone(); }, probe,
+      displacement);
+}
+
+int LRUHandleTable::FindVisibleElementOrAvailableSlot(const Slice& key,
+                                                      uint32_t hash, int& probe,
+                                                      int displacement) {
+  return FindSlot(
+      key,
+      [&](LRUHandle* h) {
+        return h->IsEmpty() || h->IsTombstone() ||
+               (h->Matches(key, hash) && h->IsVisible());
+      },
+      probe, displacement);
+}
+
+inline int LRUHandleTable::FindSlot(const Slice& key,
+                                    std::function<bool(LRUHandle*)> cond,
+                                    int& probe, int displacement) {
+  uint32_t base = ModTableSize(Hash(key.data(), key.size(), kProbingSeed1));
+  uint32_t increment =
+      ModTableSize((Hash(key.data(), key.size(), kProbingSeed2) << 1) | 1);
+  uint32_t current = ModTableSize(base + probe * increment);
+  while (true) {
+    LRUHandle* h = &array_[current];
+    probe++;
+    if (current == base && probe > 1) {
+      // We looped back.
+      return -1;
+    }
+    if (cond(h)) {
+      return current;
+    }
+    if (h->IsEmpty()) {
+      // We check emptyness after the condition, because
+      // the condition may be emptyness.
+      return -1;
+    }
+    h->displacements += displacement;
+    current = ModTableSize(current + increment);
+  }
+}
+
+LRUCacheShard::LRUCacheShard(size_t capacity, size_t estimated_value_size,
+                             bool strict_capacity_limit,
+                             CacheMetadataChargePolicy metadata_charge_policy)
+    : CacheShard(metadata_charge_policy),
+      capacity_(capacity),
+      strict_capacity_limit_(strict_capacity_limit),
+      table_(
+          CalcHashBits(capacity, estimated_value_size, metadata_charge_policy)),
+      usage_(0),
+      lru_usage_(0) {
+  // Make empty circular linked list.
+  lru_.next = &lru_;
+  lru_.prev = &lru_;
+  lru_low_pri_ = &lru_;
+}
+
+void LRUCacheShard::EraseUnRefEntries() {
+  autovector<LRUHandle> last_reference_list;
+  {
+    DMutexLock l(mutex_);
+    while (lru_.next != &lru_) {
+      LRUHandle* old = lru_.next;
+      // LRU list contains only elements which can be evicted.
+      assert(old->IsVisible() && !old->HasRefs());
+      LRU_Remove(old);
+      table_.Remove(old);
+      assert(usage_ >= old->total_charge);
+      usage_ -= old->total_charge;
+      last_reference_list.push_back(*old);
+    }
+  }
+
+  // Free the entries here outside of mutex for performance reasons.
+  for (auto& h : last_reference_list) {
+    h.FreeData();
+  }
+}
+
+void LRUCacheShard::ApplyToSomeEntries(
+    const std::function<void(const Slice& key, void* value, size_t charge,
+                             DeleterFn deleter)>& callback,
+    uint32_t average_entries_per_lock, uint32_t* state) {
+  // The state is essentially going to be the starting hash, which works
+  // nicely even if we resize between calls because we use upper-most
+  // hash bits for table indexes.
+  DMutexLock l(mutex_);
+  uint32_t length_bits = table_.GetLengthBits();
+  uint32_t length = table_.GetTableSize();
+
+  assert(average_entries_per_lock > 0);
+  // Assuming we are called with same average_entries_per_lock repeatedly,
+  // this simplifies some logic (index_end will not overflow).
+  assert(average_entries_per_lock < length || *state == 0);
+
+  uint32_t index_begin = *state >> (32 - length_bits);
+  uint32_t index_end = index_begin + average_entries_per_lock;
+  if (index_end >= length) {
+    // Going to end
+    index_end = length;
+    *state = UINT32_MAX;
+  } else {
+    *state = index_end << (32 - length_bits);
+  }
+
+  table_.ApplyToEntriesRange(
+      [callback,
+       metadata_charge_policy = metadata_charge_policy_](LRUHandle* h) {
+        callback(h->key(), h->value, h->GetCharge(metadata_charge_policy),
+                 h->deleter);
+      },
+      index_begin, index_end);
+}
+
+void LRUCacheShard::LRU_Remove(LRUHandle* h) {
+  assert(h->next != nullptr);
+  assert(h->prev != nullptr);
+  h->next->prev = h->prev;
+  h->prev->next = h->next;
+  h->prev = h->next = nullptr;
+  assert(lru_usage_ >= h->total_charge);
+  lru_usage_ -= h->total_charge;
+}
+
+void LRUCacheShard::LRU_Insert(LRUHandle* h) {
+  assert(h->next == nullptr);
+  assert(h->prev == nullptr);
+  // Insert h to head of LRU list.
+  h->next = &lru_;
+  h->prev = lru_.prev;
+  h->prev->next = h;
+  h->next->prev = h;
+  lru_usage_ += h->total_charge;
+}
+
+void LRUCacheShard::EvictFromLRU(size_t charge,
+                                 autovector<LRUHandle>* deleted) {
+  while ((usage_ + charge) > capacity_ && lru_.next != &lru_) {
+    LRUHandle* old = lru_.next;
+    // LRU list contains only elements which can be evicted.
+    assert(old->IsVisible() && !old->HasRefs());
+    LRU_Remove(old);
+    table_.Remove(old);
+    assert(usage_ >= old->total_charge);
+    usage_ -= old->total_charge;
+    deleted->push_back(*old);
+  }
+}
+
+size_t LRUCacheShard::CalcEstimatedHandleCharge(
+    size_t estimated_value_size,
+    CacheMetadataChargePolicy metadata_charge_policy) {
+  LRUHandle h;
+  h.CalcTotalCharge(estimated_value_size, metadata_charge_policy);
+  return h.total_charge;
+}
+
+int LRUCacheShard::CalcHashBits(
+    size_t capacity, size_t estimated_value_size,
+    CacheMetadataChargePolicy metadata_charge_policy) {
+  size_t handle_charge =
+      CalcEstimatedHandleCharge(estimated_value_size, metadata_charge_policy);
+  assert(handle_charge > 0);
+  uint32_t num_entries =
+      static_cast<uint32_t>(capacity / (kLoadFactor * handle_charge)) + 1;
+  assert(num_entries <= uint32_t{1} << 31);
+  return FloorLog2((num_entries << 1) - 1);
+}
+
+void LRUCacheShard::SetCapacity(size_t capacity) {
+  autovector<LRUHandle> last_reference_list;
+  {
+    DMutexLock l(mutex_);
+    if (capacity > capacity_) {
+      assert(false);  // Not supported.
+    }
+    capacity_ = capacity;
+    EvictFromLRU(0, &last_reference_list);
+  }
+
+  // Free the entries here outside of mutex for performance reasons.
+  for (auto& h : last_reference_list) {
+    h.FreeData();
+  }
+}
+
+void LRUCacheShard::SetStrictCapacityLimit(bool strict_capacity_limit) {
+  DMutexLock l(mutex_);
+  strict_capacity_limit_ = strict_capacity_limit;
+}
+
+Status LRUCacheShard::Insert(const Slice& key, uint32_t hash, void* value,
+                             size_t charge, Cache::DeleterFn deleter,
+                             Cache::Handle** handle,
+                             Cache::Priority /*priority*/) {
+  if (key.size() != kCacheKeySize) {
+    return Status::NotSupported("FastLRUCache only supports key size " +
+                                std::to_string(kCacheKeySize) + "B");
+  }
+
+  LRUHandle tmp;
+  tmp.value = value;
+  tmp.deleter = deleter;
+  tmp.hash = hash;
+  tmp.CalcTotalCharge(charge, metadata_charge_policy_);
+  for (int i = 0; i < kCacheKeySize; i++) {
+    tmp.key_data[i] = key.data()[i];
+  }
+
+  Status s = Status::OK();
+  autovector<LRUHandle> last_reference_list;
+  {
+    DMutexLock l(mutex_);
+    assert(table_.GetOccupancy() <= table_.GetOccupancyLimit());
+
+    // Free the space following strict LRU policy until enough space
+    // is freed or the lru list is empty.
+    EvictFromLRU(tmp.total_charge, &last_reference_list);
+    if ((usage_ + tmp.total_charge > capacity_ &&
+         (strict_capacity_limit_ || handle == nullptr)) ||
+        table_.GetOccupancy() == table_.GetOccupancyLimit()) {
+      // There are two measures of capacity:
+      // - Space (or charge) capacity: The maximum possible sum of the charges
+      //    of the elements.
+      // - Table capacity: The number of slots in the hash table.
+      // These are incomparable, in the sense that one doesn't imply the other.
+      // Typically we will reach space capacity before table capacity---
+      // if the user always inserts values with size equal to
+      // estimated_value_size, then at most a kLoadFactor fraction of slots
+      // will ever be occupied. But in some cases we may reach table capacity
+      // before space capacity---if the user initially claims a very large
+      // estimated_value_size but then inserts tiny values, more elements than
+      // initially estimated will be inserted.
+
+      // TODO(Guido) Some tests (at least two from cache_test, as well as the
+      // stress tests) currently assume the table capacity is unbounded.
+      if (handle == nullptr) {
+        // Don't insert the entry but still return ok, as if the entry inserted
+        // into cache and get evicted immediately.
+        last_reference_list.push_back(tmp);
+      } else {
+        if (table_.GetOccupancy() == table_.GetOccupancyLimit()) {
+          // TODO: Consider using a distinct status for this case, but usually
+          // it will be handled the same way as reaching charge capacity limit
+          s = Status::MemoryLimit(
+              "Insert failed because all slots in the hash table are full.");
+        } else {
+          s = Status::MemoryLimit(
+              "Insert failed because the total charge has exceeded the "
+              "capacity.");
+        }
+      }
+    } else {
+      // Insert into the cache. Note that the cache might get larger than its
+      // capacity if not enough space was freed up.
+      LRUHandle* old;
+      LRUHandle* h = table_.Insert(&tmp, &old);
+      assert(h != nullptr);  // We're below occupancy, so this insertion should
+                             // never fail.
+      usage_ += h->total_charge;
+      if (old != nullptr) {
+        s = Status::OkOverwritten();
+        assert(old->IsVisible());
+        table_.Exclude(old);
+        if (!old->HasRefs()) {
+          // old is on LRU because it's in cache and its reference count is 0.
+          LRU_Remove(old);
+          table_.Remove(old);
+          assert(usage_ >= old->total_charge);
+          usage_ -= old->total_charge;
+          last_reference_list.push_back(*old);
+        }
+      }
+      if (handle == nullptr) {
+        LRU_Insert(h);
+      } else {
+        // If caller already holds a ref, no need to take one here.
+        if (!h->HasRefs()) {
+          h->Ref();
+        }
+        *handle = reinterpret_cast<Cache::Handle*>(h);
+      }
+    }
+  }
+
+  // Free the entries here outside of mutex for performance reasons.
+  for (auto& h : last_reference_list) {
+    h.FreeData();
+  }
+
+  return s;
+}
+
+Cache::Handle* LRUCacheShard::Lookup(const Slice& key, uint32_t hash) {
+  LRUHandle* h = nullptr;
+  {
+    DMutexLock l(mutex_);
+    h = table_.Lookup(key, hash);
+    if (h != nullptr) {
+      assert(h->IsVisible());
+      if (!h->HasRefs()) {
+        // The entry is in LRU since it's in hash and has no external
+        // references.
+        LRU_Remove(h);
+      }
+      h->Ref();
+    }
+  }
+  return reinterpret_cast<Cache::Handle*>(h);
+}
+
+bool LRUCacheShard::Ref(Cache::Handle* h) {
+  LRUHandle* e = reinterpret_cast<LRUHandle*>(h);
+  DMutexLock l(mutex_);
+  // To create another reference - entry must be already externally referenced.
+  assert(e->HasRefs());
+  e->Ref();
+  return true;
+}
+
+bool LRUCacheShard::Release(Cache::Handle* handle, bool erase_if_last_ref) {
+  if (handle == nullptr) {
+    return false;
+  }
+  LRUHandle* h = reinterpret_cast<LRUHandle*>(handle);
+  LRUHandle copy;
+  bool last_reference = false;
+  {
+    DMutexLock l(mutex_);
+    last_reference = h->Unref();
+    if (last_reference && h->IsVisible()) {
+      // The item is still in cache, and nobody else holds a reference to it.
+      if (usage_ > capacity_ || erase_if_last_ref) {
+        // The LRU list must be empty since the cache is full.
+        assert(lru_.next == &lru_ || erase_if_last_ref);
+        // Take this opportunity and remove the item.
+        table_.Remove(h);
+      } else {
+        // Put the item back on the LRU list, and don't free it.
+        LRU_Insert(h);
+        last_reference = false;
+      }
+    }
+    // If it was the last reference, then decrement the cache usage.
+    if (last_reference) {
+      assert(usage_ >= h->total_charge);
+      usage_ -= h->total_charge;
+      copy = *h;
+    }
+  }
+
+  // Free the entry here outside of mutex for performance reasons.
+  if (last_reference) {
+    copy.FreeData();
+  }
+  return last_reference;
+}
+
+void LRUCacheShard::Erase(const Slice& key, uint32_t hash) {
+  LRUHandle copy;
+  bool last_reference = false;
+  {
+    DMutexLock l(mutex_);
+    LRUHandle* h = table_.Lookup(key, hash);
+    if (h != nullptr) {
+      table_.Exclude(h);
+      if (!h->HasRefs()) {
+        // The entry is in LRU since it's in cache and has no external
+        // references.
+        LRU_Remove(h);
+        table_.Remove(h);
+        assert(usage_ >= h->total_charge);
+        usage_ -= h->total_charge;
+        last_reference = true;
+        copy = *h;
+      }
+    }
+  }
+  // Free the entry here outside of mutex for performance reasons.
+  // last_reference will only be true if e != nullptr.
+  if (last_reference) {
+    copy.FreeData();
+  }
+}
+
+size_t LRUCacheShard::GetUsage() const {
+  DMutexLock l(mutex_);
+  return usage_;
+}
+
+size_t LRUCacheShard::GetPinnedUsage() const {
+  DMutexLock l(mutex_);
+  assert(usage_ >= lru_usage_);
+  return usage_ - lru_usage_;
+}
+
+size_t LRUCacheShard::GetOccupancyCount() const {
+  DMutexLock l(mutex_);
+  return table_.GetOccupancy();
+}
+
+size_t LRUCacheShard::GetTableAddressCount() const {
+  DMutexLock l(mutex_);
+  return table_.GetTableSize();
+}
+
+std::string LRUCacheShard::GetPrintableOptions() const { return std::string{}; }
+
+LRUCache::LRUCache(size_t capacity, size_t estimated_value_size,
+                   int num_shard_bits, bool strict_capacity_limit,
+                   CacheMetadataChargePolicy metadata_charge_policy)
+    : ShardedCache(capacity, num_shard_bits, strict_capacity_limit) {
+  assert(estimated_value_size > 0 ||
+         metadata_charge_policy != kDontChargeCacheMetadata);
+  num_shards_ = 1 << num_shard_bits;
+  shards_ = reinterpret_cast<LRUCacheShard*>(
+      port::cacheline_aligned_alloc(sizeof(LRUCacheShard) * num_shards_));
+  size_t per_shard = (capacity + (num_shards_ - 1)) / num_shards_;
+  for (int i = 0; i < num_shards_; i++) {
+    new (&shards_[i])
+        LRUCacheShard(per_shard, estimated_value_size, strict_capacity_limit,
+                      metadata_charge_policy);
+  }
+}
+
+LRUCache::~LRUCache() {
+  if (shards_ != nullptr) {
+    assert(num_shards_ > 0);
+    for (int i = 0; i < num_shards_; i++) {
+      shards_[i].~LRUCacheShard();
+    }
+    port::cacheline_aligned_free(shards_);
+  }
+}
+
+CacheShard* LRUCache::GetShard(uint32_t shard) {
+  return reinterpret_cast<CacheShard*>(&shards_[shard]);
+}
+
+const CacheShard* LRUCache::GetShard(uint32_t shard) const {
+  return reinterpret_cast<CacheShard*>(&shards_[shard]);
+}
+
+void* LRUCache::Value(Handle* handle) {
+  return reinterpret_cast<const LRUHandle*>(handle)->value;
+}
+
+size_t LRUCache::GetCharge(Handle* handle) const {
+  CacheMetadataChargePolicy metadata_charge_policy = kDontChargeCacheMetadata;
+  if (num_shards_ > 0) {
+    metadata_charge_policy = shards_[0].metadata_charge_policy_;
+  }
+  return reinterpret_cast<const LRUHandle*>(handle)->GetCharge(
+      metadata_charge_policy);
+}
+
+Cache::DeleterFn LRUCache::GetDeleter(Handle* handle) const {
+  auto h = reinterpret_cast<const LRUHandle*>(handle);
+  return h->deleter;
+}
+
+uint32_t LRUCache::GetHash(Handle* handle) const {
+  return reinterpret_cast<const LRUHandle*>(handle)->hash;
+}
+
+void LRUCache::DisownData() {
+  // Leak data only if that won't generate an ASAN/valgrind warning.
+  if (!kMustFreeHeapAllocations) {
+    shards_ = nullptr;
+    num_shards_ = 0;
+  }
+}
+
+}  // namespace fast_lru_cache
+
+std::shared_ptr<Cache> NewFastLRUCache(
+    size_t capacity, size_t estimated_value_size, int num_shard_bits,
+    bool strict_capacity_limit,
+    CacheMetadataChargePolicy metadata_charge_policy) {
+  if (num_shard_bits >= 20) {
+    return nullptr;  // The cache cannot be sharded into too many fine pieces.
+  }
+  if (num_shard_bits < 0) {
+    num_shard_bits = GetDefaultCacheShardBits(capacity);
+  }
+  return std::make_shared<fast_lru_cache::LRUCache>(
+      capacity, estimated_value_size, num_shard_bits, strict_capacity_limit,
+      metadata_charge_policy);
+}
+
+}  // namespace ROCKSDB_NAMESPACE

cache/fast_lru_cache.h

@@ -0,0 +1,476 @@
+//  Copyright (c) 2011-present, Facebook, Inc.  All rights reserved
+//  This source code is licensed under both the GPLv2 (found in the
+//  COPYING file in the root directory) and Apache 2.0 License
+//  (found in the LICENSE.Apache file in the root directory).
+//
+// Copyright (c) 2011 The LevelDB Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file. See the AUTHORS file for names of contributors.
+#pragma once
+
+#include <array>
+#include <memory>
+#include <string>
+
+#include "cache/cache_key.h"
+#include "cache/sharded_cache.h"
+#include "port/lang.h"
+#include "port/malloc.h"
+#include "port/port.h"
+#include "rocksdb/secondary_cache.h"
+#include "util/autovector.h"
+#include "util/distributed_mutex.h"
+
+namespace ROCKSDB_NAMESPACE {
+
+namespace fast_lru_cache {
+
+// Forward declaration of friend class.
+class FastLRUCacheTest;
+
+// LRU cache implementation using an open-address hash table.
+//
+// Every slot in the hash table is an LRUHandle. Because handles can be
+// referenced externally, we can't discard them immediately once they are
+// deleted (via a delete or an LRU eviction) or replaced by a new version
+// (via an insert of the same key). The state of an element is defined by
+// the following two properties:
+// (R) Referenced: An element can be referenced externally (refs > 0), or not.
+//    Importantly, an element can be evicted if and only if it's not
+//    referenced. In particular, when an element becomes referenced, it's
+//    temporarily taken out of the LRU list until all references to it
+//    are dropped.
+// (V) Visible: An element can visible for lookups (IS_VISIBLE set), or not.
+//    Initially, every element is visible. An element that is not visible is
+//    called a ghost.
+// These properties induce 4 different states, with transitions defined as
+// follows:
+// - V --> not V: When a visible element is deleted or replaced by a new
+//    version.
+// - Not V --> V: This cannot happen. A ghost remains in that state until it's
+//    not referenced any more, at which point it's ready to be removed from the
+//    hash table. (A ghost simply waits to transition to the afterlife---it will
+//    never be visible again.)
+// - R --> not R: When all references to an element are dropped.
+// - Not R --> R: When an unreferenced element becomes referenced. This can only
+//    happen if the element is V, since references to an element can only be
+//    created when it's visible.
+//
+// Internally, the cache uses an open-addressed hash table to index the handles.
+// We use tombstone counters to keep track of displacements.
+// Because of the tombstones and the two possible visibility states of an
+// element, the table slots can be in 4 different states:
+// 1. Visible element (IS_ELEMENT set and IS_VISIBLE set): The slot contains a
+//    key-value element.
+// 2. Ghost element (IS_ELEMENT set and IS_VISIBLE unset): The slot contains an
+//    element that has been removed, but it's still referenced. It's invisible
+//    to lookups.
+// 3. Tombstone (IS_ELEMENT unset and displacements > 0): The slot contains a
+//    tombstone.
+// 4. Empty (IS_ELEMENT unset and displacements == 0): The slot is unused.
+//    A slot that is an element can further have IS_VISIBLE set or not.
+// When a ghost is removed from the table, it can either transition to being a
+// tombstone or an empty slot, depending on the number of displacements of the
+// slot. In any case, the slot becomes available. When a handle is inserted
+// into that slot, it becomes a visible element again.
+
+// The load factor p is a real number in (0, 1) such that at all
+// times at most a fraction p of all slots, without counting tombstones,
+// are occupied by elements. This means that the probability that a
+// random probe hits an empty slot is at most p, and thus at most 1/p probes
+// are required on average. For example, p = 70% implies that between 1 and 2
+// probes are needed on average (bear in mind that this reasoning doesn't
+// consider the effects of clustering over time).
+// Because the size of the hash table is always rounded up to the next
+// power of 2, p is really an upper bound on the actual load factor---the
+// actual load factor is anywhere between p/2 and p. This is a bit wasteful,
+// but bear in mind that slots only hold metadata, not actual values.
+// Since space cost is dominated by the values (the LSM blocks),
+// overprovisioning the table with metadata only increases the total cache space
+// usage by a tiny fraction.
+constexpr double kLoadFactor = 0.35;
+
+// The user can exceed kLoadFactor if the sizes of the inserted values don't
+// match estimated_value_size, or if strict_capacity_limit == false. To
+// avoid performance to plunge, we set a strict upper bound on the load factor.
+constexpr double kStrictLoadFactor = 0.7;
+
+// Arbitrary seeds.
+constexpr uint32_t kProbingSeed1 = 0xbc9f1d34;
+constexpr uint32_t kProbingSeed2 = 0x7a2bb9d5;
+
+// An experimental (under development!) alternative to LRUCache
+
+struct LRUHandle {
+  void* value;
+  Cache::DeleterFn deleter;
+  LRUHandle* next;
+  LRUHandle* prev;
+  size_t total_charge;  // TODO(opt): Only allow uint32_t?
+  // The hash of key(). Used for fast sharding and comparisons.
+  uint32_t hash;
+  // The number of external refs to this entry.
+  uint32_t refs;
+
+  enum Flags : uint8_t {
+    // Whether the handle is visible to Lookups.
+    IS_VISIBLE = (1 << 0),
+    // Whether the slot is in use by an element.
+    IS_ELEMENT = (1 << 1),
+  };
+  uint8_t flags;
+
+  // The number of elements that hash to this slot or a lower one,
+  // but wind up in a higher slot.
+  uint32_t displacements;
+
+  std::array<char, kCacheKeySize> key_data;
+
+  LRUHandle() {
+    value = nullptr;
+    deleter = nullptr;
+    next = nullptr;
+    prev = nullptr;
+    total_charge = 0;
+    hash = 0;
+    refs = 0;
+    flags = 0;
+    displacements = 0;
+    key_data.fill(0);
+  }
+
+  Slice key() const { return Slice(key_data.data(), kCacheKeySize); }
+
+  // Increase the reference count by 1.
+  void Ref() { refs++; }
+
+  // Just reduce the reference count by 1. Return true if it was last reference.
+  bool Unref() {
+    assert(refs > 0);
+    refs--;
+    return refs == 0;
+  }
+
+  // Return true if there are external refs, false otherwise.
+  bool HasRefs() const { return refs > 0; }
+
+  bool IsVisible() const { return flags & IS_VISIBLE; }
+
+  void SetIsVisible(bool is_visible) {
+    if (is_visible) {
+      flags |= IS_VISIBLE;
+    } else {
+      flags &= ~IS_VISIBLE;
+    }
+  }
+
+  bool IsElement() const { return flags & IS_ELEMENT; }
+
+  void SetIsElement(bool is_element) {
+    if (is_element) {
+      flags |= IS_ELEMENT;
+    } else {
+      flags &= ~IS_ELEMENT;
+    }
+  }
+
+  void FreeData() {
+    assert(refs == 0);
+    if (deleter) {
+      (*deleter)(key(), value);
+    }
+  }
+
+  // Calculate the memory usage by metadata.
+  inline size_t CalcMetaCharge(
+      CacheMetadataChargePolicy metadata_charge_policy) const {
+    if (metadata_charge_policy != kFullChargeCacheMetadata) {
+      return 0;
+    } else {
+      // #ifdef ROCKSDB_MALLOC_USABLE_SIZE
+      //       return malloc_usable_size(
+      //           const_cast<void*>(static_cast<const void*>(this)));
+      // #else
+      // TODO(Guido) malloc_usable_size only works when we call it on
+      // a pointer allocated with malloc. Because our handles are all
+      // allocated in a single shot as an array, the user can't call
+      // CalcMetaCharge (or CalcTotalCharge or GetCharge) on a handle
+      // pointer returned by the cache. Moreover, malloc_usable_size
+      // expects a heap-allocated handle, but sometimes in our code we
+      // wish to pass a stack-allocated handle (this is only a performance
+      // concern).
+      // What is the right way to compute metadata charges with pre-allocated
+      // handles?
+      return sizeof(LRUHandle);
+      // #endif
+    }
+  }
+
+  inline void CalcTotalCharge(
+      size_t charge, CacheMetadataChargePolicy metadata_charge_policy) {
+    total_charge = charge + CalcMetaCharge(metadata_charge_policy);
+  }
+
+  inline size_t GetCharge(
+      CacheMetadataChargePolicy metadata_charge_policy) const {
+    size_t meta_charge = CalcMetaCharge(metadata_charge_policy);
+    assert(total_charge >= meta_charge);
+    return total_charge - meta_charge;
+  }
+
+  inline bool IsEmpty() {
+    return !this->IsElement() && this->displacements == 0;
+  }
+
+  inline bool IsTombstone() {
+    return !this->IsElement() && this->displacements > 0;
+  }
+
+  inline bool Matches(const Slice& some_key, uint32_t some_hash) {
+    return this->IsElement() && this->hash == some_hash &&
+           this->key() == some_key;
+  }
+};
+
+
+class LRUHandleTable {
+ public:
+  explicit LRUHandleTable(int hash_bits);
+  ~LRUHandleTable();
+
+  // Returns a pointer to a visible element matching the key/hash, or
+  // nullptr if not present.
+  LRUHandle* Lookup(const Slice& key, uint32_t hash);
+
+  // Inserts a copy of h into the hash table.
+  // Returns a pointer to the inserted handle, or nullptr if no slot
+  // available was found. If an existing visible element matching the
+  // key/hash is already present in the hash table, the argument old
+  // is set to pointe to it; otherwise, it's set to nullptr.
+  LRUHandle* Insert(LRUHandle* h, LRUHandle** old);
+
+  // Removes h from the hash table. The handle must already be off
+  // the LRU list.
+  void Remove(LRUHandle* h);
+
+  // Turns a visible element h into a ghost (i.e., not visible).
+  void Exclude(LRUHandle* h);
+
+  // Assigns a copy of h to the given slot.
+  void Assign(int slot, LRUHandle* h);
+
+  template <typename T>
+  void ApplyToEntriesRange(T func, uint32_t index_begin, uint32_t index_end) {
+    for (uint32_t i = index_begin; i < index_end; i++) {
+      LRUHandle* h = &array_[i];
+      if (h->IsVisible()) {
+        func(h);
+      }
+    }
+  }
+
+  uint32_t GetTableSize() const { return uint32_t{1} << length_bits_; }
+
+  int GetLengthBits() const { return length_bits_; }
+
+  uint32_t GetOccupancyLimit() const { return occupancy_limit_; }
+
+  uint32_t GetOccupancy() const { return occupancy_; }
+
+  // Returns x mod 2^{length_bits_}.
+  uint32_t ModTableSize(uint32_t x) { return x & length_bits_mask_; }
+
+ private:
+  int FindVisibleElement(const Slice& key, uint32_t hash, int& probe,
+                         int displacement);
+
+  int FindAvailableSlot(const Slice& key, int& probe, int displacement);
+
+  int FindVisibleElementOrAvailableSlot(const Slice& key, uint32_t hash,
+                                        int& probe, int displacement);
+
+  // Returns the index of the first slot probed (hashing with
+  // the given key) with a handle e such that cond(e) is true.
+  // Otherwise, if no match is found, returns -1.
+  // For every handle e probed except the final slot, updates
+  // e->displacements += displacement.
+  // The argument probe is modified such that consecutive calls
+  // to FindSlot continue probing right after where the previous
+  // call left.
+  int FindSlot(const Slice& key, std::function<bool(LRUHandle*)> cond,
+               int& probe, int displacement);
+
+  // Number of hash bits used for table index.
+  // The size of the table is 1 << length_bits_.
+  int length_bits_;
+
+  const uint32_t length_bits_mask_;
+
+  // Number of elements in the table.
+  uint32_t occupancy_;
+
+  // Maximum number of elements the user can store in the table.
+  uint32_t occupancy_limit_;
+
+  std::unique_ptr<LRUHandle[]> array_;
+};
+
+// A single shard of sharded cache.
+class ALIGN_AS(CACHE_LINE_SIZE) LRUCacheShard final : public CacheShard {
+ public:
+  LRUCacheShard(size_t capacity, size_t estimated_value_size,
+                bool strict_capacity_limit,
+                CacheMetadataChargePolicy metadata_charge_policy);
+  ~LRUCacheShard() override = default;
+
+  // Separate from constructor so caller can easily make an array of LRUCache
+  // if current usage is more than new capacity, the function will attempt to
+  // free the needed space.
+  void SetCapacity(size_t capacity) override;
+
+  // Set the flag to reject insertion if cache if full.
+  void SetStrictCapacityLimit(bool strict_capacity_limit) override;
+
+  // Like Cache methods, but with an extra "hash" parameter.
+  // Insert an item into the hash table and, if handle is null, insert into
+  // the LRU list. Older items are evicted as necessary. If the cache is full
+  // and free_handle_on_fail is true, the item is deleted and handle is set to
+  // nullptr.
+  Status Insert(const Slice& key, uint32_t hash, void* value, size_t charge,
+                Cache::DeleterFn deleter, Cache::Handle** handle,
+                Cache::Priority priority) override;
+
+  Status Insert(const Slice& key, uint32_t hash, void* value,
+                const Cache::CacheItemHelper* helper, size_t charge,
+                Cache::Handle** handle, Cache::Priority priority) override {
+    return Insert(key, hash, value, charge, helper->del_cb, handle, priority);
+  }
+
+  Cache::Handle* Lookup(const Slice& key, uint32_t hash,
+                        const Cache::CacheItemHelper* /*helper*/,
+                        const Cache::CreateCallback& /*create_cb*/,
+                        Cache::Priority /*priority*/, bool /*wait*/,
+                        Statistics* /*stats*/) override {
+    return Lookup(key, hash);
+  }
+  Cache::Handle* Lookup(const Slice& key, uint32_t hash) override;
+
+  bool Release(Cache::Handle* handle, bool /*useful*/,
+               bool erase_if_last_ref) override {
+    return Release(handle, erase_if_last_ref);
+  }
+  bool IsReady(Cache::Handle* /*handle*/) override { return true; }
+  void Wait(Cache::Handle* /*handle*/) override {}
+
+  bool Ref(Cache::Handle* handle) override;
+  bool Release(Cache::Handle* handle, bool erase_if_last_ref = false) override;
+  void Erase(const Slice& key, uint32_t hash) override;
+
+  size_t GetUsage() const override;
+  size_t GetPinnedUsage() const override;
+  size_t GetOccupancyCount() const override;
+  size_t GetTableAddressCount() const override;
+
+  void ApplyToSomeEntries(
+      const std::function<void(const Slice& key, void* value, size_t charge,
+                               DeleterFn deleter)>& callback,
+      uint32_t average_entries_per_lock, uint32_t* state) override;
+
+  void EraseUnRefEntries() override;
+
+  std::string GetPrintableOptions() const override;
+
+ private:
+  friend class LRUCache;
+  friend class FastLRUCacheTest;
+
+  void LRU_Remove(LRUHandle* e);
+  void LRU_Insert(LRUHandle* e);
+
+  // Free some space following strict LRU policy until enough space
+  // to hold (usage_ + charge) is freed or the LRU list is empty
+  // This function is not thread safe - it needs to be executed while
+  // holding the mutex_.
+  void EvictFromLRU(size_t charge, autovector<LRUHandle>* deleted);
+
+  // Returns the charge of a single handle.
+  static size_t CalcEstimatedHandleCharge(
+      size_t estimated_value_size,
+      CacheMetadataChargePolicy metadata_charge_policy);
+
+  // Returns the number of bits used to hash an element in the hash
+  // table.
+  static int CalcHashBits(size_t capacity, size_t estimated_value_size,
+                          CacheMetadataChargePolicy metadata_charge_policy);
+
+  // Initialized before use.
+  size_t capacity_;
+
+  // Whether to reject insertion if cache reaches its full capacity.
+  bool strict_capacity_limit_;
+
+  // Dummy head of LRU list.
+  // lru.prev is newest entry, lru.next is oldest entry.
+  // LRU contains items which can be evicted, ie reference only by cache
+  LRUHandle lru_;
+
+  // Pointer to head of low-pri pool in LRU list.
+  LRUHandle* lru_low_pri_;
+
+  // ------------^^^^^^^^^^^^^-----------
+  // Not frequently modified data members
+  // ------------------------------------
+  //
+  // We separate data members that are updated frequently from the ones that
+  // are not frequently updated so that they don't share the same cache line
+  // which will lead into false cache sharing
+  //
+  // ------------------------------------
+  // Frequently modified data members
+  // ------------vvvvvvvvvvvvv-----------
+  LRUHandleTable table_;
+
+  // Memory size for entries residing in the cache.
+  size_t usage_;
+
+  // Memory size for entries residing only in the LRU list.
+  size_t lru_usage_;
+
+  // mutex_ protects the following state.
+  // We don't count mutex_ as the cache's internal state so semantically we
+  // don't mind mutex_ invoking the non-const actions.
+  mutable DMutex mutex_;
+};
+
+class LRUCache
+#ifdef NDEBUG
+    final
+#endif
+    : public ShardedCache {
+ public:
+  LRUCache(size_t capacity, size_t estimated_value_size, int num_shard_bits,
+           bool strict_capacity_limit,
+           CacheMetadataChargePolicy metadata_charge_policy =
+               kDontChargeCacheMetadata);
+  ~LRUCache() override;
+  const char* Name() const override { return "LRUCache"; }
+  CacheShard* GetShard(uint32_t shard) override;
+  const CacheShard* GetShard(uint32_t shard) const override;
+  void* Value(Handle* handle) override;
+  size_t GetCharge(Handle* handle) const override;
+  uint32_t GetHash(Handle* handle) const override;
+  DeleterFn GetDeleter(Handle* handle) const override;
+  void DisownData() override;
+
+ private:
+  LRUCacheShard* shards_ = nullptr;
+  int num_shards_ = 0;
+};
+}  // namespace fast_lru_cache
+
+std::shared_ptr<Cache> NewFastLRUCache(
+    size_t capacity, size_t estimated_value_size, int num_shard_bits,
+    bool strict_capacity_limit,
+    CacheMetadataChargePolicy metadata_charge_policy);
+
+}  // namespace ROCKSDB_NAMESPACE

cache/lru_cache.cc

@@ -14,7 +14,6 @@
 #include <cstdio>
 #include <cstdlib>
 
-#include "cache/secondary_cache_adapter.h"
 #include "monitoring/perf_context_imp.h"
 #include "monitoring/statistics.h"
 #include "port/lang.h"
@@ -23,23 +22,20 @@
 namespace ROCKSDB_NAMESPACE {
 namespace lru_cache {
 
-LRUHandleTable::LRUHandleTable(int max_upper_hash_bits,
-                               MemoryAllocator* allocator)
+LRUHandleTable::LRUHandleTable(int max_upper_hash_bits)
     : length_bits_(/* historical starting size*/ 4),
       list_(new LRUHandle* [size_t{1} << length_bits_] {}),
       elems_(0),
-      max_length_bits_(max_upper_hash_bits),
-      allocator_(allocator) {}
+      max_length_bits_(max_upper_hash_bits) {}
 
 LRUHandleTable::~LRUHandleTable() {
-  auto alloc = allocator_;
   ApplyToEntriesRange(
-      [alloc](LRUHandle* h) {
+      [](LRUHandle* h) {
         if (!h->HasRefs()) {
-          h->Free(alloc);
+          h->Free();
         }
       },
-      0, size_t{1} << length_bits_);
+      0, uint32_t{1} << length_bits_);
 }
 
 LRUHandle* LRUHandleTable::Lookup(const Slice& key, uint32_t hash) {
@@ -96,7 +92,7 @@ void LRUHandleTable::Resize() {
   std::unique_ptr<LRUHandle* []> new_list {
     new LRUHandle* [size_t{1} << new_length_bits] {}
   };
-  [[maybe_unused]] uint32_t count = 0;
+  uint32_t count = 0;
   for (uint32_t i = 0; i < old_length; i++) {
     LRUHandle* h = list_[i];
     while (h != nullptr) {
@@ -114,14 +110,12 @@ void LRUHandleTable::Resize() {
   length_bits_ = new_length_bits;
 }
 
-LRUCacheShard::LRUCacheShard(size_t capacity, bool strict_capacity_limit,
-                             double high_pri_pool_ratio,
-                             double low_pri_pool_ratio, bool use_adaptive_mutex,
-                             CacheMetadataChargePolicy metadata_charge_policy,
-                             int max_upper_hash_bits,
-                             MemoryAllocator* allocator,
-                             const Cache::EvictionCallback* eviction_callback)
-    : CacheShardBase(metadata_charge_policy),
+LRUCacheShard::LRUCacheShard(
+    size_t capacity, bool strict_capacity_limit, double high_pri_pool_ratio,
+    double low_pri_pool_ratio, bool use_adaptive_mutex,
+    CacheMetadataChargePolicy metadata_charge_policy, int max_upper_hash_bits,
+    const std::shared_ptr<SecondaryCache>& secondary_cache)
+    : CacheShard(metadata_charge_policy),
       capacity_(0),
       high_pri_pool_usage_(0),
       low_pri_pool_usage_(0),
@@ -130,11 +124,11 @@ LRUCacheShard::LRUCacheShard(size_t capacity, bool strict_capacity_limit,
       high_pri_pool_capacity_(0),
       low_pri_pool_ratio_(low_pri_pool_ratio),
       low_pri_pool_capacity_(0),
-      table_(max_upper_hash_bits, allocator),
+      table_(max_upper_hash_bits),
       usage_(0),
       lru_usage_(0),
       mutex_(use_adaptive_mutex),
-      eviction_callback_(*eviction_callback) {
+      secondary_cache_(secondary_cache) {
   // Make empty circular linked list.
   lru_.next = &lru_;
   lru_.prev = &lru_;
@@ -161,42 +155,44 @@ void LRUCacheShard::EraseUnRefEntries() {
   }
 
   for (auto entry : last_reference_list) {
-    entry->Free(table_.GetAllocator());
+    entry->Free();
   }
 }
 
 void LRUCacheShard::ApplyToSomeEntries(
-    const std::function<void(const Slice& key, Cache::ObjectPtr value,
-                             size_t charge,
-                             const Cache::CacheItemHelper* helper)>& callback,
-    size_t average_entries_per_lock, size_t* state) {
+    const std::function<void(const Slice& key, void* value, size_t charge,
+                             DeleterFn deleter)>& callback,
+    uint32_t average_entries_per_lock, uint32_t* state) {
   // The state is essentially going to be the starting hash, which works
   // nicely even if we resize between calls because we use upper-most
   // hash bits for table indexes.
   DMutexLock l(mutex_);
-  int length_bits = table_.GetLengthBits();
-  size_t length = size_t{1} << length_bits;
+  uint32_t length_bits = table_.GetLengthBits();
+  uint32_t length = uint32_t{1} << length_bits;
 
   assert(average_entries_per_lock > 0);
   // Assuming we are called with same average_entries_per_lock repeatedly,
   // this simplifies some logic (index_end will not overflow).
   assert(average_entries_per_lock < length || *state == 0);
 
-  size_t index_begin = *state >> (sizeof(size_t) * 8u - length_bits);
-  size_t index_end = index_begin + average_entries_per_lock;
+  uint32_t index_begin = *state >> (32 - length_bits);
+  uint32_t index_end = index_begin + average_entries_per_lock;
   if (index_end >= length) {
     // Going to end
     index_end = length;
-    *state = SIZE_MAX;
+    *state = UINT32_MAX;
   } else {
-    *state = index_end << (sizeof(size_t) * 8u - length_bits);
+    *state = index_end << (32 - length_bits);
   }
 
   table_.ApplyToEntriesRange(
       [callback,
        metadata_charge_policy = metadata_charge_policy_](LRUHandle* h) {
+        DeleterFn deleter = h->IsSecondaryCacheCompatible()
+                                ? h->info_.helper->del_cb
+                                : h->info_.deleter;
         callback(h->key(), h->value, h->GetCharge(metadata_charge_policy),
-                 h->helper);
+                 deleter);
       },
       index_begin, index_end);
 }
@@ -334,19 +330,16 @@ void LRUCacheShard::EvictFromLRU(size_t charge,
   }
 }
 
-void LRUCacheShard::NotifyEvicted(
-    const autovector<LRUHandle*>& evicted_handles) {
-  MemoryAllocator* alloc = table_.GetAllocator();
-  for (LRUHandle* entry : evicted_handles) {
-    if (eviction_callback_ &&
-        eviction_callback_(entry->key(),
-                           reinterpret_cast<Cache::Handle*>(entry))) {
-      // Callback took ownership of obj; just free handle
-      free(entry);
-    } else {
-      // Free the entries here outside of mutex for performance reasons.
-      entry->Free(alloc);
+void LRUCacheShard::TryInsertIntoSecondaryCache(
+    autovector<LRUHandle*> evicted_handles) {
+  for (auto entry : evicted_handles) {
+    if (secondary_cache_ && entry->IsSecondaryCacheCompatible() &&
+        !entry->IsInSecondaryCache()) {
+      secondary_cache_->Insert(entry->key(), entry->value, entry->info_.helper)
+          .PermitUncheckedError();
     }
+    // Free the entries here outside of mutex for performance reasons.
+    entry->Free();
   }
 }
 
@@ -360,7 +353,7 @@ void LRUCacheShard::SetCapacity(size_t capacity) {
     EvictFromLRU(0, &last_reference_list);
   }
 
-  NotifyEvicted(last_reference_list);
+  TryInsertIntoSecondaryCache(last_reference_list);
 }
 
 void LRUCacheShard::SetStrictCapacityLimit(bool strict_capacity_limit) {
@@ -368,7 +361,8 @@ void LRUCacheShard::SetStrictCapacityLimit(bool strict_capacity_limit) {
   strict_capacity_limit_ = strict_capacity_limit;
 }
 
-Status LRUCacheShard::InsertItem(LRUHandle* e, LRUHandle** handle) {
+Status LRUCacheShard::InsertItem(LRUHandle* e, Cache::Handle** handle,
+                                 bool free_handle_on_fail) {
   Status s = Status::OK();
   autovector<LRUHandle*> last_reference_list;
 
@@ -387,9 +381,10 @@ Status LRUCacheShard::InsertItem(LRUHandle* e, LRUHandle** handle) {
         // into cache and get evicted immediately.
         last_reference_list.push_back(e);
       } else {
-        free(e);
-        e = nullptr;
-        *handle = nullptr;
+        if (free_handle_on_fail) {
+          free(e);
+          *handle = nullptr;
+        }
         s = Status::MemoryLimit("Insert failed due to LRU cache being full.");
       }
     } else {
@@ -416,37 +411,208 @@ Status LRUCacheShard::InsertItem(LRUHandle* e, LRUHandle** handle) {
         if (!e->HasRefs()) {
           e->Ref();
         }
-        *handle = e;
+        *handle = reinterpret_cast<Cache::Handle*>(e);
       }
     }
   }
 
-  NotifyEvicted(last_reference_list);
+  TryInsertIntoSecondaryCache(last_reference_list);
 
   return s;
 }
 
-LRUHandle* LRUCacheShard::Lookup(const Slice& key, uint32_t hash,
-                                 const Cache::CacheItemHelper* /*helper*/,
-                                 Cache::CreateContext* /*create_context*/,
-                                 Cache::Priority /*priority*/,
-                                 Statistics* /*stats*/) {
-  DMutexLock l(mutex_);
-  LRUHandle* e = table_.Lookup(key, hash);
-  if (e != nullptr) {
-    assert(e->InCache());
-    if (!e->HasRefs()) {
-      // The entry is in LRU since it's in hash and has no external
-      // references.
-      LRU_Remove(e);
+void LRUCacheShard::Promote(LRUHandle* e) {
+  SecondaryCacheResultHandle* secondary_handle = e->sec_handle;
+
+  assert(secondary_handle->IsReady());
+  e->SetIncomplete(false);
+  e->SetInCache(false);
+  e->value = secondary_handle->Value();
+  e->CalcTotalCharge(secondary_handle->Size(), metadata_charge_policy_);
+  delete secondary_handle;
+
+  if (e->value) {
+    Status s;
+    if (secondary_cache_ && secondary_cache_->SupportForceErase() &&
+        e->IsStandalone()) {
+      // Insert a dummy handle and return a standalone handle to caller.
+      // Charge the standalone handle.
+      autovector<LRUHandle*> last_reference_list;
+      bool free_standalone_handle{false};
+      {
+        DMutexLock l(mutex_);
+
+        // Free the space following strict LRU policy until enough space
+        // is freed or the lru list is empty.
+        EvictFromLRU(e->total_charge, &last_reference_list);
+
+        if ((usage_ + e->total_charge) > capacity_ && strict_capacity_limit_) {
+          free_standalone_handle = true;
+        } else {
+          usage_ += e->total_charge;
+        }
+      }
+
+      TryInsertIntoSecondaryCache(last_reference_list);
+      if (free_standalone_handle) {
+        e->Unref();
+        e->Free();
+        e = nullptr;
+      } else {
+        PERF_COUNTER_ADD(block_cache_standalone_handle_count, 1);
+      }
+
+      // Insert a dummy handle into the primary cache. This dummy handle is
+      // not IsSecondaryCacheCompatible().
+      Cache::Priority priority =
+          e->IsHighPri() ? Cache::Priority::HIGH : Cache::Priority::LOW;
+      s = Insert(e->key(), e->hash, /*value=*/nullptr, 0,
+                 /*deleter=*/nullptr, /*helper=*/nullptr, /*handle=*/nullptr,
+                 priority);
+    } else {
+      e->SetInCache(true);
+      e->SetIsStandalone(false);
+      Cache::Handle* handle = reinterpret_cast<Cache::Handle*>(e);
+      // This InsertItem() could fail if the cache is over capacity and
+      // strict_capacity_limit_ is true. In such a case, we don't want
+      // InsertItem() to free the handle, since the item is already in memory
+      // and the caller will most likely just read it from disk if we erase it
+      // here.
+      s = InsertItem(e, &handle, /*free_handle_on_fail=*/false);
+      if (s.ok()) {
+        PERF_COUNTER_ADD(block_cache_real_handle_count, 1);
+      }
     }
-    e->Ref();
-    e->SetHit();
+
+    if (!s.ok()) {
+      // Item is in memory, but not accounted against the cache capacity.
+      // When the handle is released, the item should get deleted.
+      assert(!e->InCache());
+    }
+
+  } else {
+    // Since the secondary cache lookup failed, mark the item as not in cache
+    // Don't charge the cache as its only metadata that'll shortly be released
+    DMutexLock l(mutex_);
+    // TODO
+    e->CalcTotalCharge(0, metadata_charge_policy_);
+    e->SetInCache(false);
+    e->SetIsStandalone(false);
   }
-  return e;
 }
 
-bool LRUCacheShard::Ref(LRUHandle* e) {
+Cache::Handle* LRUCacheShard::Lookup(
+    const Slice& key, uint32_t hash,
+    const ShardedCache::CacheItemHelper* helper,
+    const ShardedCache::CreateCallback& create_cb, Cache::Priority priority,
+    bool wait, Statistics* stats) {
+  LRUHandle* e = nullptr;
+  bool found_dummy_entry{false};
+  {
+    DMutexLock l(mutex_);
+    e = table_.Lookup(key, hash);
+    if (e != nullptr) {
+      assert(e->InCache());
+      if (!e->HasRefs()) {
+        // The entry is in LRU since it's in hash and has no external
+        // references.
+        LRU_Remove(e);
+      }
+      e->Ref();
+      e->SetHit();
+
+      // For a dummy handle, if it was retrieved from secondary cache,
+      // it may still exist in secondary cache.
+      // If the handle exists in secondary cache, the value should be
+      // erased from sec cache and be inserted into primary cache.
+      if (!e->value && secondary_cache_ &&
+          secondary_cache_->SupportForceErase()) {
+        found_dummy_entry = true;
+      }
+    }
+  }
+
+  // If handle table lookup failed or the handle is a dummy one, allocate
+  // a handle outside the mutex if we re going to lookup in the secondary cache.
+  //
+  // When a block is firstly Lookup from CompressedSecondaryCache, we just
+  // insert a dummy block into the primary cache (charging the actual size of
+  // the block) and don't erase the block from CompressedSecondaryCache. A
+  // standalone handle is returned to the caller. Only if the block is hit
+  // again, we erase it from CompressedSecondaryCache and add it into the
+  // primary cache.
+  //
+  // Only support synchronous for now.
+  // TODO: Support asynchronous lookup in secondary cache
+  if ((!e || found_dummy_entry) && secondary_cache_ && helper &&
+      helper->saveto_cb) {
+    // For objects from the secondary cache, we expect the caller to provide
+    // a way to create/delete the primary cache object. The only case where
+    // a deleter would not be required is for dummy entries inserted for
+    // accounting purposes, which we won't demote to the secondary cache
+    // anyway.
+    assert(create_cb && helper->del_cb);
+    // Release the dummy handle.
+    if (e) {
+      Release(reinterpret_cast<Cache::Handle*>(e), true /*erase_if_last_ref*/);
+    }
+    bool is_in_sec_cache{false};
+    std::unique_ptr<SecondaryCacheResultHandle> secondary_handle =
+        secondary_cache_->Lookup(key, create_cb, wait, found_dummy_entry,
+                                 is_in_sec_cache);
+    if (secondary_handle != nullptr) {
+      e = static_cast<LRUHandle*>(malloc(sizeof(LRUHandle) - 1 + key.size()));
+
+      e->flags = 0;
+      e->SetSecondaryCacheCompatible(true);
+      e->info_.helper = helper;
+      e->key_length = key.size();
+      e->hash = hash;
+      e->refs = 0;
+      e->next = e->prev = nullptr;
+      e->SetPriority(priority);
+      memcpy(e->key_data, key.data(), key.size());
+      e->value = nullptr;
+      e->sec_handle = secondary_handle.release();
+      e->total_charge = 0;
+      e->Ref();
+      e->SetIsInSecondaryCache(is_in_sec_cache);
+
+      if (secondary_cache_->SupportForceErase() && !found_dummy_entry) {
+        e->SetIsStandalone(true);
+      }
+
+      if (wait) {
+        Promote(e);
+        if (e) {
+          if (!e->value) {
+            // The secondary cache returned a handle, but the lookup failed.
+            e->Unref();
+            e->Free();
+            e = nullptr;
+          } else {
+            PERF_COUNTER_ADD(secondary_cache_hit_count, 1);
+            RecordTick(stats, SECONDARY_CACHE_HITS);
+          }
+        }
+      } else {
+        // If wait is false, we always return a handle and let the caller
+        // release the handle after checking for success or failure.
+        e->SetIncomplete(true);
+        // This may be slightly inaccurate, if the lookup eventually fails.
+        // But the probability is very low.
+        PERF_COUNTER_ADD(secondary_cache_hit_count, 1);
+        RecordTick(stats, SECONDARY_CACHE_HITS);
+      }
+    } else {
+      e = nullptr;
+    }
+  }
+  return reinterpret_cast<Cache::Handle*>(e);
+}
+
+bool LRUCacheShard::Ref(Cache::Handle* h) {
+  LRUHandle* e = reinterpret_cast<LRUHandle*>(h);
   DMutexLock l(mutex_);
   // To create another reference - entry must be already externally referenced.
   assert(e->HasRefs());
@@ -468,18 +634,16 @@ void LRUCacheShard::SetLowPriorityPoolRatio(double low_pri_pool_ratio) {
   MaintainPoolSize();
 }
 
-bool LRUCacheShard::Release(LRUHandle* e, bool /*useful*/,
-                            bool erase_if_last_ref) {
-  if (e == nullptr) {
+bool LRUCacheShard::Release(Cache::Handle* handle, bool erase_if_last_ref) {
+  if (handle == nullptr) {
     return false;
   }
-  bool must_free;
-  bool was_in_cache;
+  LRUHandle* e = reinterpret_cast<LRUHandle*>(handle);
+  bool last_reference = false;
   {
     DMutexLock l(mutex_);
-    must_free = e->Unref();
-    was_in_cache = e->InCache();
-    if (must_free && was_in_cache) {
+    last_reference = e->Unref();
+    if (last_reference && e->InCache()) {
       // The item is still in cache, and nobody else holds a reference to it.
       if (usage_ > capacity_ || erase_if_last_ref) {
         // The LRU list must be empty since the cache is full.
@@ -490,39 +654,32 @@ bool LRUCacheShard::Release(LRUHandle* e, bool /*useful*/,
       } else {
         // Put the item back on the LRU list, and don't free it.
         LRU_Insert(e);
-        must_free = false;
+        last_reference = false;
       }
     }
-    // If about to be freed, then decrement the cache usage.
-    if (must_free) {
+    // If it was the last reference, and the entry is either not secondary
+    // cache compatible (i.e a dummy entry for accounting), or is secondary
+    // cache compatible and has a non-null value, then decrement the cache
+    // usage. If value is null in the latter case, that means the lookup
+    // failed and we didn't charge the cache.
+    if (last_reference && (!e->IsSecondaryCacheCompatible() || e->value)) {
       assert(usage_ >= e->total_charge);
       usage_ -= e->total_charge;
     }
   }
 
   // Free the entry here outside of mutex for performance reasons.
-  if (must_free) {
-    // Only call eviction callback if we're sure no one requested erasure
-    // FIXME: disabled because of test churn
-    if (false && was_in_cache && !erase_if_last_ref && eviction_callback_ &&
-        eviction_callback_(e->key(), reinterpret_cast<Cache::Handle*>(e))) {
-      // Callback took ownership of obj; just free handle
-      free(e);
-    } else {
-      e->Free(table_.GetAllocator());
-    }
+  if (last_reference) {
+    e->Free();
   }
-  return must_free;
+  return last_reference;
 }
 
-LRUHandle* LRUCacheShard::CreateHandle(const Slice& key, uint32_t hash,
-                                       Cache::ObjectPtr value,
-                                       const Cache::CacheItemHelper* helper,
-                                       size_t charge) {
-  assert(helper);
-  // value == nullptr is reserved for indicating failure in SecondaryCache
-  assert(!(helper->IsSecondaryCacheCompatible() && value == nullptr));
-
+Status LRUCacheShard::Insert(const Slice& key, uint32_t hash, void* value,
+                             size_t charge,
+                             void (*deleter)(const Slice& key, void* value),
+                             const Cache::CacheItemHelper* helper,
+                             Cache::Handle** handle, Cache::Priority priority) {
   // Allocate the memory here outside of the mutex.
   // If the cache is full, we'll have to release it.
   // It shouldn't happen very often though.
@@ -530,60 +687,26 @@ LRUHandle* LRUCacheShard::CreateHandle(const Slice& key, uint32_t hash,
       static_cast<LRUHandle*>(malloc(sizeof(LRUHandle) - 1 + key.size()));
 
   e->value = value;
-  e->m_flags = 0;
-  e->im_flags = 0;
-  e->helper = helper;
+  e->flags = 0;
+  if (helper) {
+    e->SetSecondaryCacheCompatible(true);
+    e->info_.helper = helper;
+  } else {
+#ifdef __SANITIZE_THREAD__
+    e->is_secondary_cache_compatible_for_tsan = false;
+#endif  // __SANITIZE_THREAD__
+    e->info_.deleter = deleter;
+  }
   e->key_length = key.size();
   e->hash = hash;
   e->refs = 0;
   e->next = e->prev = nullptr;
+  e->SetInCache(true);
+  e->SetPriority(priority);
   memcpy(e->key_data, key.data(), key.size());
   e->CalcTotalCharge(charge, metadata_charge_policy_);
 
-  return e;
-}
-
-Status LRUCacheShard::Insert(const Slice& key, uint32_t hash,
-                             Cache::ObjectPtr value,
-                             const Cache::CacheItemHelper* helper,
-                             size_t charge, LRUHandle** handle,
-                             Cache::Priority priority) {
-  LRUHandle* e = CreateHandle(key, hash, value, helper, charge);
-  e->SetPriority(priority);
-  e->SetInCache(true);
-  return InsertItem(e, handle);
-}
-
-LRUHandle* LRUCacheShard::CreateStandalone(const Slice& key, uint32_t hash,
-                                           Cache::ObjectPtr value,
-                                           const Cache::CacheItemHelper* helper,
-                                           size_t charge,
-                                           bool allow_uncharged) {
-  LRUHandle* e = CreateHandle(key, hash, value, helper, charge);
-  e->SetIsStandalone(true);
-  e->Ref();
-
-  autovector<LRUHandle*> last_reference_list;
-
-  {
-    DMutexLock l(mutex_);
-
-    EvictFromLRU(e->total_charge, &last_reference_list);
-
-    if (strict_capacity_limit_ && (usage_ + e->total_charge) > capacity_) {
-      if (allow_uncharged) {
-        e->total_charge = 0;
-      } else {
-        free(e);
-        e = nullptr;
-      }
-    } else {
-      usage_ += e->total_charge;
-    }
-  }
-
-  NotifyEvicted(last_reference_list);
-  return e;
+  return InsertItem(e, handle, /* free_handle_on_fail */ true);
 }
 
 void LRUCacheShard::Erase(const Slice& key, uint32_t hash) {
@@ -608,10 +731,22 @@ void LRUCacheShard::Erase(const Slice& key, uint32_t hash) {
   // Free the entry here outside of mutex for performance reasons.
   // last_reference will only be true if e != nullptr.
   if (last_reference) {
-    e->Free(table_.GetAllocator());
+    e->Free();
   }
 }
 
+bool LRUCacheShard::IsReady(Cache::Handle* handle) {
+  LRUHandle* e = reinterpret_cast<LRUHandle*>(handle);
+  DMutexLock l(mutex_);
+  bool ready = true;
+  if (e->IsPending()) {
+    assert(secondary_cache_);
+    assert(e->sec_handle);
+    ready = e->sec_handle->IsReady();
+  }
+  return ready;
+}
+
 size_t LRUCacheShard::GetUsage() const {
   DMutexLock l(mutex_);
   return usage_;
@@ -633,7 +768,7 @@ size_t LRUCacheShard::GetTableAddressCount() const {
   return size_t{1} << table_.GetLengthBits();
 }
 
-void LRUCacheShard::AppendPrintableOptions(std::string& str) const {
+std::string LRUCacheShard::GetPrintableOptions() const {
   const int kBufferSize = 200;
   char buffer[kBufferSize];
   {
@@ -643,7 +778,7 @@ void LRUCacheShard::AppendPrintableOptions(std::string& str) const {
     snprintf(buffer + strlen(buffer), kBufferSize - strlen(buffer),
              "    low_pri_pool_ratio: %.3lf\n", low_pri_pool_ratio_);
   }
-  str.append(buffer);
+  return std::string(buffer);
 }
 
 LRUCache::LRUCache(size_t capacity, int num_shard_bits,
@@ -651,43 +786,130 @@ LRUCache::LRUCache(size_t capacity, int num_shard_bits,
                    double low_pri_pool_ratio,
                    std::shared_ptr<MemoryAllocator> allocator,
                    bool use_adaptive_mutex,
-                   CacheMetadataChargePolicy metadata_charge_policy)
+                   CacheMetadataChargePolicy metadata_charge_policy,
+                   const std::shared_ptr<SecondaryCache>& secondary_cache)
     : ShardedCache(capacity, num_shard_bits, strict_capacity_limit,
                    std::move(allocator)) {
-  size_t per_shard = GetPerShardCapacity();
-  MemoryAllocator* alloc = memory_allocator();
-  const EvictionCallback* eviction_callback = &eviction_callback_;
-  InitShards([=](LRUCacheShard* cs) {
-    new (cs) LRUCacheShard(per_shard, strict_capacity_limit,
-                           high_pri_pool_ratio, low_pri_pool_ratio,
-                           use_adaptive_mutex, metadata_charge_policy,
-                           /* max_upper_hash_bits */ 32 - num_shard_bits, alloc,
-                           eviction_callback);
-  });
+  num_shards_ = 1 << num_shard_bits;
+  shards_ = reinterpret_cast<LRUCacheShard*>(
+      port::cacheline_aligned_alloc(sizeof(LRUCacheShard) * num_shards_));
+  size_t per_shard = (capacity + (num_shards_ - 1)) / num_shards_;
+  for (int i = 0; i < num_shards_; i++) {
+    new (&shards_[i]) LRUCacheShard(
+        per_shard, strict_capacity_limit, high_pri_pool_ratio,
+        low_pri_pool_ratio, use_adaptive_mutex, metadata_charge_policy,
+        /* max_upper_hash_bits */ 32 - num_shard_bits, secondary_cache);
+  }
+  secondary_cache_ = secondary_cache;
 }
 
-Cache::ObjectPtr LRUCache::Value(Handle* handle) {
-  auto h = reinterpret_cast<const LRUHandle*>(handle);
-  return h->value;
+LRUCache::~LRUCache() {
+  if (shards_ != nullptr) {
+    assert(num_shards_ > 0);
+    for (int i = 0; i < num_shards_; i++) {
+      shards_[i].~LRUCacheShard();
+    }
+    port::cacheline_aligned_free(shards_);
+  }
+}
+
+CacheShard* LRUCache::GetShard(uint32_t shard) {
+  return reinterpret_cast<CacheShard*>(&shards_[shard]);
+}
+
+const CacheShard* LRUCache::GetShard(uint32_t shard) const {
+  return reinterpret_cast<CacheShard*>(&shards_[shard]);
+}
+
+void* LRUCache::Value(Handle* handle) {
+  return reinterpret_cast<const LRUHandle*>(handle)->value;
 }
 
 size_t LRUCache::GetCharge(Handle* handle) const {
+  CacheMetadataChargePolicy metadata_charge_policy = kDontChargeCacheMetadata;
+  if (num_shards_ > 0) {
+    metadata_charge_policy = shards_[0].metadata_charge_policy_;
+  }
   return reinterpret_cast<const LRUHandle*>(handle)->GetCharge(
-      GetShard(0).metadata_charge_policy_);
+      metadata_charge_policy);
 }
 
-const Cache::CacheItemHelper* LRUCache::GetCacheItemHelper(
-    Handle* handle) const {
+Cache::DeleterFn LRUCache::GetDeleter(Handle* handle) const {
   auto h = reinterpret_cast<const LRUHandle*>(handle);
-  return h->helper;
+  if (h->IsSecondaryCacheCompatible()) {
+    return h->info_.helper->del_cb;
+  } else {
+    return h->info_.deleter;
+  }
+}
+
+uint32_t LRUCache::GetHash(Handle* handle) const {
+  return reinterpret_cast<const LRUHandle*>(handle)->hash;
+}
+
+void LRUCache::DisownData() {
+  // Leak data only if that won't generate an ASAN/valgrind warning.
+  if (!kMustFreeHeapAllocations) {
+    shards_ = nullptr;
+    num_shards_ = 0;
+  }
 }
 
 size_t LRUCache::TEST_GetLRUSize() {
-  return SumOverShards([](LRUCacheShard& cs) { return cs.TEST_GetLRUSize(); });
+  size_t lru_size_of_all_shards = 0;
+  for (int i = 0; i < num_shards_; i++) {
+    lru_size_of_all_shards += shards_[i].TEST_GetLRUSize();
+  }
+  return lru_size_of_all_shards;
 }
 
 double LRUCache::GetHighPriPoolRatio() {
-  return GetShard(0).GetHighPriPoolRatio();
+  double result = 0.0;
+  if (num_shards_ > 0) {
+    result = shards_[0].GetHighPriPoolRatio();
+  }
+  return result;
+}
+
+void LRUCache::WaitAll(std::vector<Handle*>& handles) {
+  if (secondary_cache_) {
+    std::vector<SecondaryCacheResultHandle*> sec_handles;
+    sec_handles.reserve(handles.size());
+    for (Handle* handle : handles) {
+      if (!handle) {
+        continue;
+      }
+      LRUHandle* lru_handle = reinterpret_cast<LRUHandle*>(handle);
+      if (!lru_handle->IsPending()) {
+        continue;
+      }
+      sec_handles.emplace_back(lru_handle->sec_handle);
+    }
+    secondary_cache_->WaitAll(sec_handles);
+    for (Handle* handle : handles) {
+      if (!handle) {
+        continue;
+      }
+      LRUHandle* lru_handle = reinterpret_cast<LRUHandle*>(handle);
+      if (!lru_handle->IsPending()) {
+        continue;
+      }
+      uint32_t hash = GetHash(handle);
+      LRUCacheShard* shard = static_cast<LRUCacheShard*>(GetShard(Shard(hash)));
+      shard->Promote(lru_handle);
+    }
+  }
+}
+
+std::string LRUCache::GetPrintableOptions() const {
+  std::string ret;
+  ret.reserve(20000);
+  ret.append(ShardedCache::GetPrintableOptions());
+  if (secondary_cache_) {
+    ret.append("  secondary_cache:\n");
+    ret.append(secondary_cache_->GetPrintableOptions());
+  }
+  return ret;
 }
 
 }  // namespace lru_cache
@@ -707,7 +929,7 @@ std::shared_ptr<Cache> NewLRUCache(
     return nullptr;
   }
   if (low_pri_pool_ratio < 0.0 || low_pri_pool_ratio > 1.0) {
-    // Invalid low_pri_pool_ratio
+    // Invalid high_pri_pool_ratio
     return nullptr;
   }
   if (low_pri_pool_ratio + high_pri_pool_ratio > 1.0) {
@@ -717,14 +939,10 @@ std::shared_ptr<Cache> NewLRUCache(
   if (num_shard_bits < 0) {
     num_shard_bits = GetDefaultCacheShardBits(capacity);
   }
-  std::shared_ptr<Cache> cache = std::make_shared<LRUCache>(
+  return std::make_shared<LRUCache>(
       capacity, num_shard_bits, strict_capacity_limit, high_pri_pool_ratio,
       low_pri_pool_ratio, std::move(memory_allocator), use_adaptive_mutex,
-      metadata_charge_policy);
-  if (secondary_cache) {
-    cache = std::make_shared<CacheWithSecondaryAdapter>(cache, secondary_cache);
-  }
-  return cache;
+      metadata_charge_policy, secondary_cache);
 }
 
 std::shared_ptr<Cache> NewLRUCache(const LRUCacheOptions& cache_opts) {

cache/lru_cache.h

@@ -13,9 +13,9 @@
 
 #include "cache/sharded_cache.h"
 #include "port/lang.h"
-#include "port/likely.h"
 #include "port/malloc.h"
 #include "port/port.h"
+#include "rocksdb/secondary_cache.h"
 #include "util/autovector.h"
 #include "util/distributed_mutex.h"
 
@@ -38,19 +38,36 @@ namespace lru_cache {
 //    (refs == 0 && in_cache == true)
 // 3. Referenced externally AND not in hash table.
 //    In that case the entry is not in the LRU list and not in hash table.
-//    The entry must be freed if refs becomes 0 in this state.
+//    The entry can be freed when refs becomes 0.
 //    (refs >= 1 && in_cache == false)
-// If you call LRUCacheShard::Release enough times on an entry in state 1, it
-// will go into state 2. To move from state 1 to state 3, either call
-// LRUCacheShard::Erase or LRUCacheShard::Insert with the same key (but
-// possibly different value). To move from state 2 to state 1, use
-// LRUCacheShard::Lookup.
-// While refs > 0, public properties like value and deleter must not change.
+// 4. The handle is never inserted into the LRUCache (both hash table and LRU
+//    list) and it doesn't experience the above three states.
+//    The entry can be freed when refs becomes 0.
+//    (refs >= 1 && in_cache == false && IS_STANDALONE == true)
+// All newly created LRUHandles are in state 1 or 4. If you call
+// LRUCacheShard::Release on entry in state 1, it will go into state 2.
+// To move from state 1 to state 3, either call LRUCacheShard::Erase or
+// LRUCacheShard::Insert with the same key (but possibly different value).
+// To move from state 2 to state 1, use LRUCacheShard::Lookup.
+// Before destruction, make sure that no handles are in state 1. This means
+// that any successful LRUCacheShard::Lookup/LRUCacheShard::Insert have a
+// matching LRUCache::Release (to move into state 2) or LRUCacheShard::Erase
+// (to move into state 3).
 
 struct LRUHandle {
-  Cache::ObjectPtr value;
-  const Cache::CacheItemHelper* helper;
-  LRUHandle* next_hash;
+  void* value;
+  union Info {
+    Info() {}
+    ~Info() {}
+    Cache::DeleterFn deleter;
+    const ShardedCache::CacheItemHelper* helper;
+  } info_;
+  // An entry is not added to the LRUHandleTable until the secondary cache
+  // lookup is complete, so its safe to have this union.
+  union {
+    LRUHandle* next_hash;
+    SecondaryCacheResultHandle* sec_handle;
+  };
   LRUHandle* next;
   LRUHandle* prev;
   size_t total_charge;  // TODO(opt): Only allow uint32_t?
@@ -60,41 +77,45 @@ struct LRUHandle {
   // The number of external refs to this entry. The cache itself is not counted.
   uint32_t refs;
 
-  // Mutable flags - access controlled by mutex
-  // The m_ and M_ prefixes (and im_ and IM_ later) are to hopefully avoid
-  // checking an M_ flag on im_flags or an IM_ flag on m_flags.
-  uint8_t m_flags;
-  enum MFlags : uint8_t {
+  enum Flags : uint16_t {
     // Whether this entry is referenced by the hash table.
-    M_IN_CACHE = (1 << 0),
-    // Whether this entry has had any lookups (hits).
-    M_HAS_HIT = (1 << 1),
+    IN_CACHE = (1 << 0),
+    // Whether this entry is high priority entry.
+    IS_HIGH_PRI = (1 << 1),
     // Whether this entry is in high-pri pool.
-    M_IN_HIGH_PRI_POOL = (1 << 2),
+    IN_HIGH_PRI_POOL = (1 << 2),
+    // Whether this entry has had any lookups (hits).
+    HAS_HIT = (1 << 3),
+    // Can this be inserted into the secondary cache.
+    IS_SECONDARY_CACHE_COMPATIBLE = (1 << 4),
+    // Is the handle still being read from a lower tier.
+    IS_PENDING = (1 << 5),
+    // Whether this handle is still in a lower tier
+    IS_IN_SECONDARY_CACHE = (1 << 6),
+    // Whether this entry is low priority entry.
+    IS_LOW_PRI = (1 << 7),
     // Whether this entry is in low-pri pool.
-    M_IN_LOW_PRI_POOL = (1 << 3),
+    IN_LOW_PRI_POOL = (1 << 8),
+    // Whether this entry is not inserted into the cache (both hash table and
+    // LRU list).
+    IS_STANDALONE = (1 << 9),
   };
 
-  // "Immutable" flags - only set in single-threaded context and then
-  // can be accessed without mutex
-  uint8_t im_flags;
-  enum ImFlags : uint8_t {
-    // Whether this entry is high priority entry.
-    IM_IS_HIGH_PRI = (1 << 0),
-    // Whether this entry is low priority entry.
-    IM_IS_LOW_PRI = (1 << 1),
-    // Marks result handles that should not be inserted into cache
-    IM_IS_STANDALONE = (1 << 2),
-  };
+  uint16_t flags;
+
+#ifdef __SANITIZE_THREAD__
+  // TSAN can report a false data race on flags, where one thread is writing
+  // to one of the mutable bits and another thread is reading this immutable
+  // bit. So precisely suppress that TSAN warning, we separate out this bit
+  // during TSAN runs.
+  bool is_secondary_cache_compatible_for_tsan;
+#endif  // __SANITIZE_THREAD__
 
   // Beginning of the key (MUST BE THE LAST FIELD IN THIS STRUCT!)
   char key_data[1];
 
   Slice key() const { return Slice(key_data, key_length); }
 
-  // For HandleImpl concept
-  uint32_t GetHash() const { return hash; }
-
   // Increase the reference count by 1.
   void Ref() { refs++; }
 
@@ -108,66 +129,117 @@ struct LRUHandle {
   // Return true if there are external refs, false otherwise.
   bool HasRefs() const { return refs > 0; }
 
-  bool InCache() const { return m_flags & M_IN_CACHE; }
-  bool IsHighPri() const { return im_flags & IM_IS_HIGH_PRI; }
-  bool InHighPriPool() const { return m_flags & M_IN_HIGH_PRI_POOL; }
-  bool IsLowPri() const { return im_flags & IM_IS_LOW_PRI; }
-  bool InLowPriPool() const { return m_flags & M_IN_LOW_PRI_POOL; }
-  bool HasHit() const { return m_flags & M_HAS_HIT; }
-  bool IsStandalone() const { return im_flags & IM_IS_STANDALONE; }
+  bool InCache() const { return flags & IN_CACHE; }
+  bool IsHighPri() const { return flags & IS_HIGH_PRI; }
+  bool InHighPriPool() const { return flags & IN_HIGH_PRI_POOL; }
+  bool IsLowPri() const { return flags & IS_LOW_PRI; }
+  bool InLowPriPool() const { return flags & IN_LOW_PRI_POOL; }
+  bool HasHit() const { return flags & HAS_HIT; }
+  bool IsSecondaryCacheCompatible() const {
+#ifdef __SANITIZE_THREAD__
+    return is_secondary_cache_compatible_for_tsan;
+#else
+    return flags & IS_SECONDARY_CACHE_COMPATIBLE;
+#endif  // __SANITIZE_THREAD__
+  }
+  bool IsPending() const { return flags & IS_PENDING; }
+  bool IsInSecondaryCache() const { return flags & IS_IN_SECONDARY_CACHE; }
+  bool IsStandalone() const { return flags & IS_STANDALONE; }
 
   void SetInCache(bool in_cache) {
     if (in_cache) {
-      m_flags |= M_IN_CACHE;
+      flags |= IN_CACHE;
     } else {
-      m_flags &= ~M_IN_CACHE;
+      flags &= ~IN_CACHE;
     }
   }
 
   void SetPriority(Cache::Priority priority) {
     if (priority == Cache::Priority::HIGH) {
-      im_flags |= IM_IS_HIGH_PRI;
-      im_flags &= ~IM_IS_LOW_PRI;
+      flags |= IS_HIGH_PRI;
+      flags &= ~IS_LOW_PRI;
     } else if (priority == Cache::Priority::LOW) {
-      im_flags &= ~IM_IS_HIGH_PRI;
-      im_flags |= IM_IS_LOW_PRI;
+      flags &= ~IS_HIGH_PRI;
+      flags |= IS_LOW_PRI;
     } else {
-      im_flags &= ~IM_IS_HIGH_PRI;
-      im_flags &= ~IM_IS_LOW_PRI;
+      flags &= ~IS_HIGH_PRI;
+      flags &= ~IS_LOW_PRI;
     }
   }
 
   void SetInHighPriPool(bool in_high_pri_pool) {
     if (in_high_pri_pool) {
-      m_flags |= M_IN_HIGH_PRI_POOL;
+      flags |= IN_HIGH_PRI_POOL;
     } else {
-      m_flags &= ~M_IN_HIGH_PRI_POOL;
+      flags &= ~IN_HIGH_PRI_POOL;
     }
   }
 
   void SetInLowPriPool(bool in_low_pri_pool) {
     if (in_low_pri_pool) {
-      m_flags |= M_IN_LOW_PRI_POOL;
+      flags |= IN_LOW_PRI_POOL;
     } else {
-      m_flags &= ~M_IN_LOW_PRI_POOL;
+      flags &= ~IN_LOW_PRI_POOL;
     }
   }
 
-  void SetHit() { m_flags |= M_HAS_HIT; }
+  void SetHit() { flags |= HAS_HIT; }
+
+  void SetSecondaryCacheCompatible(bool compat) {
+    if (compat) {
+      flags |= IS_SECONDARY_CACHE_COMPATIBLE;
+    } else {
+      flags &= ~IS_SECONDARY_CACHE_COMPATIBLE;
+    }
+#ifdef __SANITIZE_THREAD__
+    is_secondary_cache_compatible_for_tsan = compat;
+#endif  // __SANITIZE_THREAD__
+  }
+
+  void SetIncomplete(bool incomp) {
+    if (incomp) {
+      flags |= IS_PENDING;
+    } else {
+      flags &= ~IS_PENDING;
+    }
+  }
+
+  void SetIsInSecondaryCache(bool is_in_secondary_cache) {
+    if (is_in_secondary_cache) {
+      flags |= IS_IN_SECONDARY_CACHE;
+    } else {
+      flags &= ~IS_IN_SECONDARY_CACHE;
+    }
+  }
 
   void SetIsStandalone(bool is_standalone) {
     if (is_standalone) {
-      im_flags |= IM_IS_STANDALONE;
+      flags |= IS_STANDALONE;
     } else {
-      im_flags &= ~IM_IS_STANDALONE;
+      flags &= ~IS_STANDALONE;
     }
   }
 
-  void Free(MemoryAllocator* allocator) {
+  void Free() {
     assert(refs == 0);
-    assert(helper);
-    if (helper->del_cb) {
-      helper->del_cb(value, allocator);
+#ifdef __SANITIZE_THREAD__
+    // Here we can safely assert they are the same without a data race reported
+    assert(((flags & IS_SECONDARY_CACHE_COMPATIBLE) != 0) ==
+           is_secondary_cache_compatible_for_tsan);
+#endif  // __SANITIZE_THREAD__
+    if (!IsSecondaryCacheCompatible() && info_.deleter) {
+      (*info_.deleter)(key(), value);
+    } else if (IsSecondaryCacheCompatible()) {
+      if (IsPending()) {
+        assert(sec_handle != nullptr);
+        SecondaryCacheResultHandle* tmp_sec_handle = sec_handle;
+        tmp_sec_handle->Wait();
+        value = tmp_sec_handle->Value();
+        delete tmp_sec_handle;
+      }
+      if (value) {
+        (*info_.helper->del_cb)(key(), value);
+      }
     }
 
     free(this);
@@ -209,7 +281,10 @@ struct LRUHandle {
 // 4.4.3's builtin hashtable.
 class LRUHandleTable {
  public:
-  explicit LRUHandleTable(int max_upper_hash_bits, MemoryAllocator* allocator);
+  // If the table uses more hash bits than `max_upper_hash_bits`,
+  // it will eat into the bits used for sharding, which are constant
+  // for a given LRUHandleTable.
+  explicit LRUHandleTable(int max_upper_hash_bits);
   ~LRUHandleTable();
 
   LRUHandle* Lookup(const Slice& key, uint32_t hash);
@@ -217,8 +292,8 @@ class LRUHandleTable {
   LRUHandle* Remove(const Slice& key, uint32_t hash);
 
   template <typename T>
-  void ApplyToEntriesRange(T func, size_t index_begin, size_t index_end) {
-    for (size_t i = index_begin; i < index_end; i++) {
+  void ApplyToEntriesRange(T func, uint32_t index_begin, uint32_t index_end) {
+    for (uint32_t i = index_begin; i < index_end; i++) {
       LRUHandle* h = list_[i];
       while (h != nullptr) {
         auto n = h->next_hash;
@@ -233,8 +308,6 @@ class LRUHandleTable {
 
   size_t GetOccupancyCount() const { return elems_; }
 
-  MemoryAllocator* GetAllocator() const { return allocator_; }
-
  private:
   // Return a pointer to slot that points to a cache entry that
   // matches key/hash.  If there is no such cache entry, return a
@@ -256,40 +329,26 @@ class LRUHandleTable {
 
   // Set from max_upper_hash_bits (see constructor).
   const int max_length_bits_;
-
-  // From Cache, needed for delete
-  MemoryAllocator* const allocator_;
 };
 
 // A single shard of sharded cache.
-class ALIGN_AS(CACHE_LINE_SIZE) LRUCacheShard final : public CacheShardBase {
+class ALIGN_AS(CACHE_LINE_SIZE) LRUCacheShard final : public CacheShard {
  public:
-  // NOTE: the eviction_callback ptr is saved, as is it assumed to be kept
-  // alive in Cache.
   LRUCacheShard(size_t capacity, bool strict_capacity_limit,
                 double high_pri_pool_ratio, double low_pri_pool_ratio,
                 bool use_adaptive_mutex,
                 CacheMetadataChargePolicy metadata_charge_policy,
-                int max_upper_hash_bits, MemoryAllocator* allocator,
-                const Cache::EvictionCallback* eviction_callback);
-
- public:  // Type definitions expected as parameter to ShardedCache
-  using HandleImpl = LRUHandle;
-  using HashVal = uint32_t;
-  using HashCref = uint32_t;
-
- public:  // Function definitions expected as parameter to ShardedCache
-  static inline HashVal ComputeHash(const Slice& key) {
-    return Lower32of64(GetSliceNPHash64(key));
-  }
+                int max_upper_hash_bits,
+                const std::shared_ptr<SecondaryCache>& secondary_cache);
+  virtual ~LRUCacheShard() override = default;
 
   // Separate from constructor so caller can easily make an array of LRUCache
   // if current usage is more than new capacity, the function will attempt to
   // free the needed space.
-  void SetCapacity(size_t capacity);
+  virtual void SetCapacity(size_t capacity) override;
 
   // Set the flag to reject insertion if cache if full.
-  void SetStrictCapacityLimit(bool strict_capacity_limit);
+  virtual void SetStrictCapacityLimit(bool strict_capacity_limit) override;
 
   // Set percentage of capacity reserved for high-pri cache entries.
   void SetHighPriorityPoolRatio(double high_pri_pool_ratio);
@@ -298,42 +357,58 @@ class ALIGN_AS(CACHE_LINE_SIZE) LRUCacheShard final : public CacheShardBase {
   void SetLowPriorityPoolRatio(double low_pri_pool_ratio);
 
   // Like Cache methods, but with an extra "hash" parameter.
-  Status Insert(const Slice& key, uint32_t hash, Cache::ObjectPtr value,
-                const Cache::CacheItemHelper* helper, size_t charge,
-                LRUHandle** handle, Cache::Priority priority);
-
-  LRUHandle* CreateStandalone(const Slice& key, uint32_t hash,
-                              Cache::ObjectPtr obj,
-                              const Cache::CacheItemHelper* helper,
-                              size_t charge, bool allow_uncharged);
-
-  LRUHandle* Lookup(const Slice& key, uint32_t hash,
-                    const Cache::CacheItemHelper* helper,
-                    Cache::CreateContext* create_context,
-                    Cache::Priority priority, Statistics* stats);
-
-  bool Release(LRUHandle* handle, bool useful, bool erase_if_last_ref);
-  bool Ref(LRUHandle* handle);
-  void Erase(const Slice& key, uint32_t hash);
+  virtual Status Insert(const Slice& key, uint32_t hash, void* value,
+                        size_t charge, Cache::DeleterFn deleter,
+                        Cache::Handle** handle,
+                        Cache::Priority priority) override {
+    return Insert(key, hash, value, charge, deleter, nullptr, handle, priority);
+  }
+  virtual Status Insert(const Slice& key, uint32_t hash, void* value,
+                        const Cache::CacheItemHelper* helper, size_t charge,
+                        Cache::Handle** handle,
+                        Cache::Priority priority) override {
+    assert(helper);
+    return Insert(key, hash, value, charge, nullptr, helper, handle, priority);
+  }
+  // If helper_cb is null, the values of the following arguments don't matter.
+  virtual Cache::Handle* Lookup(const Slice& key, uint32_t hash,
+                                const ShardedCache::CacheItemHelper* helper,
+                                const ShardedCache::CreateCallback& create_cb,
+                                ShardedCache::Priority priority, bool wait,
+                                Statistics* stats) override;
+  virtual Cache::Handle* Lookup(const Slice& key, uint32_t hash) override {
+    return Lookup(key, hash, nullptr, nullptr, Cache::Priority::LOW, true,
+                  nullptr);
+  }
+  virtual bool Release(Cache::Handle* handle, bool /*useful*/,
+                       bool erase_if_last_ref) override {
+    return Release(handle, erase_if_last_ref);
+  }
+  virtual bool IsReady(Cache::Handle* /*handle*/) override;
+  virtual void Wait(Cache::Handle* /*handle*/) override {}
+  virtual bool Ref(Cache::Handle* handle) override;
+  virtual bool Release(Cache::Handle* handle,
+                       bool erase_if_last_ref = false) override;
+  virtual void Erase(const Slice& key, uint32_t hash) override;
 
   // Although in some platforms the update of size_t is atomic, to make sure
   // GetUsage() and GetPinnedUsage() work correctly under any platform, we'll
   // protect them with mutex_.
 
-  size_t GetUsage() const;
-  size_t GetPinnedUsage() const;
-  size_t GetOccupancyCount() const;
-  size_t GetTableAddressCount() const;
+  virtual size_t GetUsage() const override;
+  virtual size_t GetPinnedUsage() const override;
+  virtual size_t GetOccupancyCount() const override;
+  virtual size_t GetTableAddressCount() const override;
 
-  void ApplyToSomeEntries(
-      const std::function<void(const Slice& key, Cache::ObjectPtr value,
-                               size_t charge,
-                               const Cache::CacheItemHelper* helper)>& callback,
-      size_t average_entries_per_lock, size_t* state);
+  virtual void ApplyToSomeEntries(
+      const std::function<void(const Slice& key, void* value, size_t charge,
+                               DeleterFn deleter)>& callback,
+      uint32_t average_entries_per_lock, uint32_t* state) override;
 
-  void EraseUnRefEntries();
+  virtual void EraseUnRefEntries() override;
+
+  virtual std::string GetPrintableOptions() const override;
 
- public:  // other function definitions
   void TEST_GetLRUList(LRUHandle** lru, LRUHandle** lru_low_pri,
                        LRUHandle** lru_bottom_pri);
 
@@ -347,15 +422,26 @@ class ALIGN_AS(CACHE_LINE_SIZE) LRUCacheShard final : public CacheShardBase {
   // Retrieves low pri pool ratio
   double GetLowPriPoolRatio();
 
-  void AppendPrintableOptions(std::string& /*str*/) const;
-
  private:
   friend class LRUCache;
   // Insert an item into the hash table and, if handle is null, insert into
-  // the LRU list. Older items are evicted as necessary. Frees `item` on
-  // non-OK status.
-  Status InsertItem(LRUHandle* item, LRUHandle** handle);
-
+  // the LRU list. Older items are evicted as necessary. If the cache is full
+  // and free_handle_on_fail is true, the item is deleted and handle is set to
+  // nullptr.
+  Status InsertItem(LRUHandle* item, Cache::Handle** handle,
+                    bool free_handle_on_fail);
+  Status Insert(const Slice& key, uint32_t hash, void* value, size_t charge,
+                DeleterFn deleter, const Cache::CacheItemHelper* helper,
+                Cache::Handle** handle, Cache::Priority priority);
+  // Promote an item looked up from the secondary cache to the LRU cache.
+  // The item may be still in the secondary cache.
+  // It is only inserted into the hash table and not the LRU list, and only
+  // if the cache is not at full capacity, as is the case during Insert.  The
+  // caller should hold a reference on the LRUHandle. When the caller releases
+  // the last reference, the item is added to the LRU list.
+  // The item is promoted to the high pri or low pri pool as specified by the
+  // caller in Lookup.
+  void Promote(LRUHandle* e);
   void LRU_Remove(LRUHandle* e);
   void LRU_Insert(LRUHandle* e);
 
@@ -369,11 +455,8 @@ class ALIGN_AS(CACHE_LINE_SIZE) LRUCacheShard final : public CacheShardBase {
   // holding the mutex_.
   void EvictFromLRU(size_t charge, autovector<LRUHandle*>* deleted);
 
-  void NotifyEvicted(const autovector<LRUHandle*>& evicted_handles);
-
-  LRUHandle* CreateHandle(const Slice& key, uint32_t hash,
-                          Cache::ObjectPtr value,
-                          const Cache::CacheItemHelper* helper, size_t charge);
+  // Try to insert the evicted handles into the secondary cache.
+  void TryInsertIntoSecondaryCache(autovector<LRUHandle*> evicted_handles);
 
   // Initialized before use.
   size_t capacity_;
@@ -436,31 +519,43 @@ class ALIGN_AS(CACHE_LINE_SIZE) LRUCacheShard final : public CacheShardBase {
   // don't mind mutex_ invoking the non-const actions.
   mutable DMutex mutex_;
 
-  // A reference to Cache::eviction_callback_
-  const Cache::EvictionCallback& eviction_callback_;
+  std::shared_ptr<SecondaryCache> secondary_cache_;
 };
 
 class LRUCache
 #ifdef NDEBUG
     final
 #endif
-    : public ShardedCache<LRUCacheShard> {
+    : public ShardedCache {
  public:
   LRUCache(size_t capacity, int num_shard_bits, bool strict_capacity_limit,
            double high_pri_pool_ratio, double low_pri_pool_ratio,
            std::shared_ptr<MemoryAllocator> memory_allocator = nullptr,
            bool use_adaptive_mutex = kDefaultToAdaptiveMutex,
            CacheMetadataChargePolicy metadata_charge_policy =
-               kDontChargeCacheMetadata);
-  const char* Name() const override { return "LRUCache"; }
-  ObjectPtr Value(Handle* handle) override;
-  size_t GetCharge(Handle* handle) const override;
-  const CacheItemHelper* GetCacheItemHelper(Handle* handle) const override;
+               kDontChargeCacheMetadata,
+           const std::shared_ptr<SecondaryCache>& secondary_cache = nullptr);
+  virtual ~LRUCache();
+  virtual const char* Name() const override { return "LRUCache"; }
+  virtual CacheShard* GetShard(uint32_t shard) override;
+  virtual const CacheShard* GetShard(uint32_t shard) const override;
+  virtual void* Value(Handle* handle) override;
+  virtual size_t GetCharge(Handle* handle) const override;
+  virtual uint32_t GetHash(Handle* handle) const override;
+  virtual DeleterFn GetDeleter(Handle* handle) const override;
+  virtual void DisownData() override;
+  virtual void WaitAll(std::vector<Handle*>& handles) override;
+  std::string GetPrintableOptions() const override;
 
   // Retrieves number of elements in LRU, for unit test purpose only.
   size_t TEST_GetLRUSize();
   // Retrieves high pri pool ratio.
   double GetHighPriPoolRatio();
+
+ private:
+  LRUCacheShard* shards_ = nullptr;
+  int num_shards_ = 0;
+  std::shared_ptr<SecondaryCache> secondary_cache_;
 };
 
 }  // namespace lru_cache

cache/secondary_cache.cc

@@ -1,44 +0,0 @@
-//  Copyright (c) 2011-present, Facebook, Inc.  All rights reserved.
-//  This source code is licensed under both the GPLv2 (found in the
-//  COPYING file in the root directory) and Apache 2.0 License
-//  (found in the LICENSE.Apache file in the root directory).
-
-#include "rocksdb/secondary_cache.h"
-
-#include "cache/cache_entry_roles.h"
-
-namespace ROCKSDB_NAMESPACE {
-
-namespace {
-
-void NoopDelete(Cache::ObjectPtr, MemoryAllocator*) {}
-
-size_t SliceSize(Cache::ObjectPtr obj) {
-  return static_cast<Slice*>(obj)->size();
-}
-
-Status SliceSaveTo(Cache::ObjectPtr from_obj, size_t from_offset, size_t length,
-                   char* out) {
-  const Slice& slice = *static_cast<Slice*>(from_obj);
-  std::memcpy(out, slice.data() + from_offset, length);
-  return Status::OK();
-}
-
-Status FailCreate(const Slice&, Cache::CreateContext*, MemoryAllocator*,
-                  Cache::ObjectPtr*, size_t*) {
-  return Status::NotSupported("Only for dumping data into SecondaryCache");
-}
-
-}  // namespace
-
-Status SecondaryCache::InsertSaved(const Slice& key, const Slice& saved) {
-  static Cache::CacheItemHelper helper_no_secondary{CacheEntryRole::kMisc,
-                                                    &NoopDelete};
-  static Cache::CacheItemHelper helper{
-      CacheEntryRole::kMisc, &NoopDelete, &SliceSize,
-      &SliceSaveTo,          &FailCreate, &helper_no_secondary};
-  // NOTE: depends on Insert() being synchronous, not keeping pointer `&saved`
-  return Insert(key, const_cast<Slice*>(&saved), &helper);
-}
-
-}  // namespace ROCKSDB_NAMESPACE

cache/secondary_cache_adapter.cc

@@ -1,295 +0,0 @@
-//  Copyright (c) Meta Platforms, Inc. and affiliates.
-//  This source code is licensed under both the GPLv2 (found in the
-//  COPYING file in the root directory) and Apache 2.0 License
-//  (found in the LICENSE.Apache file in the root directory).
-
-#include "cache/secondary_cache_adapter.h"
-
-#include "monitoring/perf_context_imp.h"
-
-namespace ROCKSDB_NAMESPACE {
-
-namespace {
-// A distinct pointer value for marking "dummy" cache entries
-struct Dummy {
-  char val[7] = "kDummy";
-};
-const Dummy kDummy{};
-Cache::ObjectPtr const kDummyObj = const_cast<Dummy*>(&kDummy);
-}  // namespace
-
-CacheWithSecondaryAdapter::CacheWithSecondaryAdapter(
-    std::shared_ptr<Cache> target,
-    std::shared_ptr<SecondaryCache> secondary_cache)
-    : CacheWrapper(std::move(target)),
-      secondary_cache_(std::move(secondary_cache)) {
-  target_->SetEvictionCallback([this](const Slice& key, Handle* handle) {
-    return EvictionHandler(key, handle);
-  });
-}
-
-CacheWithSecondaryAdapter::~CacheWithSecondaryAdapter() {
-  // `*this` will be destroyed before `*target_`, so we have to prevent
-  // use after free
-  target_->SetEvictionCallback({});
-}
-
-bool CacheWithSecondaryAdapter::EvictionHandler(const Slice& key,
-                                                Handle* handle) {
-  auto helper = GetCacheItemHelper(handle);
-  if (helper->IsSecondaryCacheCompatible()) {
-    auto obj = target_->Value(handle);
-    // Ignore dummy entry
-    if (obj != kDummyObj) {
-      // Spill into secondary cache.
-      secondary_cache_->Insert(key, obj, helper).PermitUncheckedError();
-    }
-  }
-  // Never takes ownership of obj
-  return false;
-}
-
-bool CacheWithSecondaryAdapter::ProcessDummyResult(Cache::Handle** handle,
-                                                   bool erase) {
-  if (*handle && target_->Value(*handle) == kDummyObj) {
-    target_->Release(*handle, erase);
-    *handle = nullptr;
-    return true;
-  } else {
-    return false;
-  }
-}
-
-void CacheWithSecondaryAdapter::CleanupCacheObject(
-    ObjectPtr obj, const CacheItemHelper* helper) {
-  if (helper->del_cb) {
-    helper->del_cb(obj, memory_allocator());
-  }
-}
-
-Cache::Handle* CacheWithSecondaryAdapter::Promote(
-    std::unique_ptr<SecondaryCacheResultHandle>&& secondary_handle,
-    const Slice& key, const CacheItemHelper* helper, Priority priority,
-    Statistics* stats, bool found_dummy_entry, bool kept_in_sec_cache) {
-  assert(secondary_handle->IsReady());
-
-  ObjectPtr obj = secondary_handle->Value();
-  if (!obj) {
-    // Nothing found.
-    return nullptr;
-  }
-  // Found something.
-  switch (helper->role) {
-    case CacheEntryRole::kFilterBlock:
-      RecordTick(stats, SECONDARY_CACHE_FILTER_HITS);
-      break;
-    case CacheEntryRole::kIndexBlock:
-      RecordTick(stats, SECONDARY_CACHE_INDEX_HITS);
-      break;
-    case CacheEntryRole::kDataBlock:
-      RecordTick(stats, SECONDARY_CACHE_DATA_HITS);
-      break;
-    default:
-      break;
-  }
-  PERF_COUNTER_ADD(secondary_cache_hit_count, 1);
-  RecordTick(stats, SECONDARY_CACHE_HITS);
-
-  // Note: SecondaryCache::Size() is really charge (from the CreateCallback)
-  size_t charge = secondary_handle->Size();
-  Handle* result = nullptr;
-  // Insert into primary cache, possibly as a standalone+dummy entries.
-  if (secondary_cache_->SupportForceErase() && !found_dummy_entry) {
-    // Create standalone and insert dummy
-    // Allow standalone to be created even if cache is full, to avoid
-    // reading the entry from storage.
-    result =
-        CreateStandalone(key, obj, helper, charge, /*allow_uncharged*/ true);
-    assert(result);
-    PERF_COUNTER_ADD(block_cache_standalone_handle_count, 1);
-
-    // Insert dummy to record recent use
-    // TODO: try to avoid case where inserting this dummy could overwrite a
-    // regular entry
-    Status s = Insert(key, kDummyObj, &kNoopCacheItemHelper, /*charge=*/0,
-                      /*handle=*/nullptr, priority);
-    s.PermitUncheckedError();
-    // Nothing to do or clean up on dummy insertion failure
-  } else {
-    // Insert regular entry into primary cache.
-    // Don't allow it to spill into secondary cache again if it was kept there.
-    Status s = Insert(
-        key, obj, kept_in_sec_cache ? helper->without_secondary_compat : helper,
-        charge, &result, priority);
-    if (s.ok()) {
-      assert(result);
-      PERF_COUNTER_ADD(block_cache_real_handle_count, 1);
-    } else {
-      // Create standalone result instead, even if cache is full, to avoid
-      // reading the entry from storage.
-      result =
-          CreateStandalone(key, obj, helper, charge, /*allow_uncharged*/ true);
-      assert(result);
-      PERF_COUNTER_ADD(block_cache_standalone_handle_count, 1);
-    }
-  }
-  return result;
-}
-
-Cache::Handle* CacheWithSecondaryAdapter::Lookup(const Slice& key,
-                                                 const CacheItemHelper* helper,
-                                                 CreateContext* create_context,
-                                                 Priority priority,
-                                                 Statistics* stats) {
-  // NOTE: we could just StartAsyncLookup() and Wait(), but this should be a bit
-  // more efficient
-  Handle* result =
-      target_->Lookup(key, helper, create_context, priority, stats);
-  bool secondary_compatible = helper && helper->IsSecondaryCacheCompatible();
-  bool found_dummy_entry =
-      ProcessDummyResult(&result, /*erase=*/secondary_compatible);
-  if (!result && secondary_compatible) {
-    // Try our secondary cache
-    bool kept_in_sec_cache = false;
-    std::unique_ptr<SecondaryCacheResultHandle> secondary_handle =
-        secondary_cache_->Lookup(key, helper, create_context, /*wait*/ true,
-                                 found_dummy_entry, /*out*/ kept_in_sec_cache);
-    if (secondary_handle) {
-      result = Promote(std::move(secondary_handle), key, helper, priority,
-                       stats, found_dummy_entry, kept_in_sec_cache);
-    }
-  }
-  return result;
-}
-
-Cache::ObjectPtr CacheWithSecondaryAdapter::Value(Handle* handle) {
-  ObjectPtr v = target_->Value(handle);
-  // TODO with stacked secondaries: might fail in EvictionHandler
-  assert(v != kDummyObj);
-  return v;
-}
-
-void CacheWithSecondaryAdapter::StartAsyncLookupOnMySecondary(
-    AsyncLookupHandle& async_handle) {
-  assert(!async_handle.IsPending());
-  assert(async_handle.result_handle == nullptr);
-
-  std::unique_ptr<SecondaryCacheResultHandle> secondary_handle =
-      secondary_cache_->Lookup(async_handle.key, async_handle.helper,
-                               async_handle.create_context, /*wait*/ false,
-                               async_handle.found_dummy_entry,
-                               /*out*/ async_handle.kept_in_sec_cache);
-  if (secondary_handle) {
-    // TODO with stacked secondaries: Check & process if already ready?
-    async_handle.pending_handle = secondary_handle.release();
-    async_handle.pending_cache = secondary_cache_.get();
-  }
-}
-
-void CacheWithSecondaryAdapter::StartAsyncLookup(
-    AsyncLookupHandle& async_handle) {
-  target_->StartAsyncLookup(async_handle);
-  if (!async_handle.IsPending()) {
-    bool secondary_compatible =
-        async_handle.helper &&
-        async_handle.helper->IsSecondaryCacheCompatible();
-    async_handle.found_dummy_entry |= ProcessDummyResult(
-        &async_handle.result_handle, /*erase=*/secondary_compatible);
-
-    if (async_handle.Result() == nullptr && secondary_compatible) {
-      // Not found and not pending on another secondary cache
-      StartAsyncLookupOnMySecondary(async_handle);
-    }
-  }
-}
-
-void CacheWithSecondaryAdapter::WaitAll(AsyncLookupHandle* async_handles,
-                                        size_t count) {
-  if (count == 0) {
-    // Nothing to do
-    return;
-  }
-  // Requests that are pending on *my* secondary cache, at the start of this
-  // function
-  std::vector<AsyncLookupHandle*> my_pending;
-  // Requests that are pending on an "inner" secondary cache (managed somewhere
-  // under target_), as of the start of this function
-  std::vector<AsyncLookupHandle*> inner_pending;
-
-  // Initial accounting of pending handles, excluding those already handled
-  // by "outer" secondary caches. (See cur->pending_cache = nullptr.)
-  for (size_t i = 0; i < count; ++i) {
-    AsyncLookupHandle* cur = async_handles + i;
-    if (cur->pending_cache) {
-      assert(cur->IsPending());
-      assert(cur->helper);
-      assert(cur->helper->IsSecondaryCacheCompatible());
-      if (cur->pending_cache == secondary_cache_.get()) {
-        my_pending.push_back(cur);
-        // Mark as "to be handled by this caller"
-        cur->pending_cache = nullptr;
-      } else {
-        // Remember as potentially needing a lookup in my secondary
-        inner_pending.push_back(cur);
-      }
-    }
-  }
-
-  // Wait on inner-most cache lookups first
-  // TODO with stacked secondaries: because we are not using proper
-  // async/await constructs here yet, there is a false synchronization point
-  // here where all the results at one level are needed before initiating
-  // any lookups at the next level. Probably not a big deal, but worth noting.
-  if (!inner_pending.empty()) {
-    target_->WaitAll(async_handles, count);
-  }
-
-  // For those that failed to find something, convert to lookup in my
-  // secondary cache.
-  for (AsyncLookupHandle* cur : inner_pending) {
-    if (cur->Result() == nullptr) {
-      // Not found, try my secondary
-      StartAsyncLookupOnMySecondary(*cur);
-      if (cur->IsPending()) {
-        assert(cur->pending_cache == secondary_cache_.get());
-        my_pending.push_back(cur);
-        // Mark as "to be handled by this caller"
-        cur->pending_cache = nullptr;
-      }
-    }
-  }
-
-  // Wait on all lookups on my secondary cache
-  {
-    std::vector<SecondaryCacheResultHandle*> my_secondary_handles;
-    for (AsyncLookupHandle* cur : my_pending) {
-      my_secondary_handles.push_back(cur->pending_handle);
-    }
-    secondary_cache_->WaitAll(my_secondary_handles);
-  }
-
-  // Process results
-  for (AsyncLookupHandle* cur : my_pending) {
-    std::unique_ptr<SecondaryCacheResultHandle> secondary_handle(
-        cur->pending_handle);
-    cur->pending_handle = nullptr;
-    cur->result_handle = Promote(
-        std::move(secondary_handle), cur->key, cur->helper, cur->priority,
-        cur->stats, cur->found_dummy_entry, cur->kept_in_sec_cache);
-    assert(cur->pending_cache == nullptr);
-  }
-}
-
-std::string CacheWithSecondaryAdapter::GetPrintableOptions() const {
-  std::string str = target_->GetPrintableOptions();
-  str.append("  secondary_cache:\n");
-  str.append(secondary_cache_->GetPrintableOptions());
-  return str;
-}
-
-const char* CacheWithSecondaryAdapter::Name() const {
-  // To the user, at least for now, configure the underlying cache with
-  // a secondary cache. So we pretend to be that cache
-  return target_->Name();
-}
-}  // namespace ROCKSDB_NAMESPACE

cache/secondary_cache_adapter.h

@@ -1,52 +0,0 @@
-//  Copyright (c) Meta Platforms, Inc. and affiliates.
-//  This source code is licensed under both the GPLv2 (found in the
-//  COPYING file in the root directory) and Apache 2.0 License
-//  (found in the LICENSE.Apache file in the root directory).
-
-#pragma once
-
-#include "rocksdb/secondary_cache.h"
-
-namespace ROCKSDB_NAMESPACE {
-
-class CacheWithSecondaryAdapter : public CacheWrapper {
- public:
-  explicit CacheWithSecondaryAdapter(
-      std::shared_ptr<Cache> target,
-      std::shared_ptr<SecondaryCache> secondary_cache);
-
-  ~CacheWithSecondaryAdapter() override;
-
-  Handle* Lookup(const Slice& key, const CacheItemHelper* helper,
-                 CreateContext* create_context,
-                 Priority priority = Priority::LOW,
-                 Statistics* stats = nullptr) override;
-
-  ObjectPtr Value(Handle* handle) override;
-
-  void StartAsyncLookup(AsyncLookupHandle& async_handle) override;
-
-  void WaitAll(AsyncLookupHandle* async_handles, size_t count) override;
-
-  std::string GetPrintableOptions() const override;
-
-  const char* Name() const override;
-
- private:
-  bool EvictionHandler(const Slice& key, Handle* handle);
-
-  void StartAsyncLookupOnMySecondary(AsyncLookupHandle& async_handle);
-
-  Handle* Promote(
-      std::unique_ptr<SecondaryCacheResultHandle>&& secondary_handle,
-      const Slice& key, const CacheItemHelper* helper, Priority priority,
-      Statistics* stats, bool found_dummy_entry, bool kept_in_sec_cache);
-
-  bool ProcessDummyResult(Cache::Handle** handle, bool erase);
-
-  void CleanupCacheObject(ObjectPtr obj, const CacheItemHelper* helper);
-
-  std::shared_ptr<SecondaryCache> secondary_cache_;
-};
-
-}  // namespace ROCKSDB_NAMESPACE

cache/sharded_cache.cc

@@ -19,49 +19,184 @@
 
 namespace ROCKSDB_NAMESPACE {
 
-ShardedCacheBase::ShardedCacheBase(size_t capacity, int num_shard_bits,
-                                   bool strict_capacity_limit,
-                                   std::shared_ptr<MemoryAllocator> allocator)
+namespace {
+
+inline uint32_t HashSlice(const Slice& s) {
+  return Lower32of64(GetSliceNPHash64(s));
+}
+
+}  // namespace
+
+ShardedCache::ShardedCache(size_t capacity, int num_shard_bits,
+                           bool strict_capacity_limit,
+                           std::shared_ptr<MemoryAllocator> allocator)
     : Cache(std::move(allocator)),
-      last_id_(1),
       shard_mask_((uint32_t{1} << num_shard_bits) - 1),
+      capacity_(capacity),
       strict_capacity_limit_(strict_capacity_limit),
-      capacity_(capacity) {}
+      last_id_(1) {}
 
-size_t ShardedCacheBase::ComputePerShardCapacity(size_t capacity) const {
+void ShardedCache::SetCapacity(size_t capacity) {
   uint32_t num_shards = GetNumShards();
-  return (capacity + (num_shards - 1)) / num_shards;
+  const size_t per_shard = (capacity + (num_shards - 1)) / num_shards;
+  MutexLock l(&capacity_mutex_);
+  for (uint32_t s = 0; s < num_shards; s++) {
+    GetShard(s)->SetCapacity(per_shard);
+  }
+  capacity_ = capacity;
 }
 
-size_t ShardedCacheBase::GetPerShardCapacity() const {
-  return ComputePerShardCapacity(GetCapacity());
+void ShardedCache::SetStrictCapacityLimit(bool strict_capacity_limit) {
+  uint32_t num_shards = GetNumShards();
+  MutexLock l(&capacity_mutex_);
+  for (uint32_t s = 0; s < num_shards; s++) {
+    GetShard(s)->SetStrictCapacityLimit(strict_capacity_limit);
+  }
+  strict_capacity_limit_ = strict_capacity_limit;
 }
 
-uint64_t ShardedCacheBase::NewId() {
+Status ShardedCache::Insert(const Slice& key, void* value, size_t charge,
+                            DeleterFn deleter, Handle** handle,
+                            Priority priority) {
+  uint32_t hash = HashSlice(key);
+  return GetShard(Shard(hash))
+      ->Insert(key, hash, value, charge, deleter, handle, priority);
+}
+
+Status ShardedCache::Insert(const Slice& key, void* value,
+                            const CacheItemHelper* helper, size_t charge,
+                            Handle** handle, Priority priority) {
+  uint32_t hash = HashSlice(key);
+  if (!helper) {
+    return Status::InvalidArgument();
+  }
+  return GetShard(Shard(hash))
+      ->Insert(key, hash, value, helper, charge, handle, priority);
+}
+
+Cache::Handle* ShardedCache::Lookup(const Slice& key, Statistics* /*stats*/) {
+  uint32_t hash = HashSlice(key);
+  return GetShard(Shard(hash))->Lookup(key, hash);
+}
+
+Cache::Handle* ShardedCache::Lookup(const Slice& key,
+                                    const CacheItemHelper* helper,
+                                    const CreateCallback& create_cb,
+                                    Priority priority, bool wait,
+                                    Statistics* stats) {
+  uint32_t hash = HashSlice(key);
+  return GetShard(Shard(hash))
+      ->Lookup(key, hash, helper, create_cb, priority, wait, stats);
+}
+
+bool ShardedCache::IsReady(Handle* handle) {
+  uint32_t hash = GetHash(handle);
+  return GetShard(Shard(hash))->IsReady(handle);
+}
+
+void ShardedCache::Wait(Handle* handle) {
+  uint32_t hash = GetHash(handle);
+  GetShard(Shard(hash))->Wait(handle);
+}
+
+bool ShardedCache::Ref(Handle* handle) {
+  uint32_t hash = GetHash(handle);
+  return GetShard(Shard(hash))->Ref(handle);
+}
+
+bool ShardedCache::Release(Handle* handle, bool erase_if_last_ref) {
+  uint32_t hash = GetHash(handle);
+  return GetShard(Shard(hash))->Release(handle, erase_if_last_ref);
+}
+
+bool ShardedCache::Release(Handle* handle, bool useful,
+                           bool erase_if_last_ref) {
+  uint32_t hash = GetHash(handle);
+  return GetShard(Shard(hash))->Release(handle, useful, erase_if_last_ref);
+}
+
+void ShardedCache::Erase(const Slice& key) {
+  uint32_t hash = HashSlice(key);
+  GetShard(Shard(hash))->Erase(key, hash);
+}
+
+uint64_t ShardedCache::NewId() {
   return last_id_.fetch_add(1, std::memory_order_relaxed);
 }
 
-size_t ShardedCacheBase::GetCapacity() const {
-  MutexLock l(&config_mutex_);
+size_t ShardedCache::GetCapacity() const {
+  MutexLock l(&capacity_mutex_);
   return capacity_;
 }
 
-bool ShardedCacheBase::HasStrictCapacityLimit() const {
-  MutexLock l(&config_mutex_);
+bool ShardedCache::HasStrictCapacityLimit() const {
+  MutexLock l(&capacity_mutex_);
   return strict_capacity_limit_;
 }
 
-size_t ShardedCacheBase::GetUsage(Handle* handle) const {
+size_t ShardedCache::GetUsage() const {
+  // We will not lock the cache when getting the usage from shards.
+  uint32_t num_shards = GetNumShards();
+  size_t usage = 0;
+  for (uint32_t s = 0; s < num_shards; s++) {
+    usage += GetShard(s)->GetUsage();
+  }
+  return usage;
+}
+
+size_t ShardedCache::GetUsage(Handle* handle) const {
   return GetCharge(handle);
 }
 
-std::string ShardedCacheBase::GetPrintableOptions() const {
+size_t ShardedCache::GetPinnedUsage() const {
+  // We will not lock the cache when getting the usage from shards.
+  uint32_t num_shards = GetNumShards();
+  size_t usage = 0;
+  for (uint32_t s = 0; s < num_shards; s++) {
+    usage += GetShard(s)->GetPinnedUsage();
+  }
+  return usage;
+}
+
+void ShardedCache::ApplyToAllEntries(
+    const std::function<void(const Slice& key, void* value, size_t charge,
+                             DeleterFn deleter)>& callback,
+    const ApplyToAllEntriesOptions& opts) {
+  uint32_t num_shards = GetNumShards();
+  // Iterate over part of each shard, rotating between shards, to
+  // minimize impact on latency of concurrent operations.
+  std::unique_ptr<uint32_t[]> states(new uint32_t[num_shards]{});
+
+  uint32_t aepl_in_32 = static_cast<uint32_t>(
+      std::min(size_t{UINT32_MAX}, opts.average_entries_per_lock));
+  aepl_in_32 = std::min(aepl_in_32, uint32_t{1});
+
+  bool remaining_work;
+  do {
+    remaining_work = false;
+    for (uint32_t s = 0; s < num_shards; s++) {
+      if (states[s] != UINT32_MAX) {
+        GetShard(s)->ApplyToSomeEntries(callback, aepl_in_32, &states[s]);
+        remaining_work |= states[s] != UINT32_MAX;
+      }
+    }
+  } while (remaining_work);
+}
+
+void ShardedCache::EraseUnRefEntries() {
+  uint32_t num_shards = GetNumShards();
+  for (uint32_t s = 0; s < num_shards; s++) {
+    GetShard(s)->EraseUnRefEntries();
+  }
+}
+
+std::string ShardedCache::GetPrintableOptions() const {
   std::string ret;
   ret.reserve(20000);
   const int kBufferSize = 200;
   char buffer[kBufferSize];
   {
-    MutexLock l(&config_mutex_);
+    MutexLock l(&capacity_mutex_);
     snprintf(buffer, kBufferSize, "    capacity : %" ROCKSDB_PRIszt "\n",
              capacity_);
     ret.append(buffer);
@@ -75,7 +210,7 @@ std::string ShardedCacheBase::GetPrintableOptions() const {
   snprintf(buffer, kBufferSize, "    memory_allocator : %s\n",
            memory_allocator() ? memory_allocator()->Name() : "None");
   ret.append(buffer);
-  AppendPrintableOptions(ret);
+  ret.append(GetShard(0)->GetPrintableOptions());
   return ret;
 }
 
@@ -91,10 +226,25 @@ int GetDefaultCacheShardBits(size_t capacity, size_t min_shard_size) {
   return num_shard_bits;
 }
 
-int ShardedCacheBase::GetNumShardBits() const {
-  return BitsSetToOne(shard_mask_);
+int ShardedCache::GetNumShardBits() const { return BitsSetToOne(shard_mask_); }
+
+uint32_t ShardedCache::GetNumShards() const { return shard_mask_ + 1; }
+
+size_t ShardedCache::GetOccupancyCount() const {
+  size_t oc = 0;
+  uint32_t num_shards = GetNumShards();
+  for (uint32_t s = 0; s < num_shards; s++) {
+    oc += GetShard(s)->GetOccupancyCount();
+  }
+  return oc;
+}
+size_t ShardedCache::GetTableAddressCount() const {
+  size_t tac = 0;
+  uint32_t num_shards = GetNumShards();
+  for (uint32_t s = 0; s < num_shards; s++) {
+    tac += GetShard(s)->GetTableAddressCount();
+  }
+  return tac;
 }
 
-uint32_t ShardedCacheBase::GetNumShards() const { return shard_mask_ + 1; }
-
 }  // namespace ROCKSDB_NAMESPACE

cache/sharded_cache.h

@@ -10,297 +10,122 @@
 #pragma once
 
 #include <atomic>
-#include <cstdint>
 #include <string>
 
-#include "port/lang.h"
 #include "port/port.h"
-#include "rocksdb/advanced_cache.h"
-#include "util/hash.h"
-#include "util/mutexlock.h"
+#include "rocksdb/cache.h"
 
 namespace ROCKSDB_NAMESPACE {
 
-// Optional base class for classes implementing the CacheShard concept
-class CacheShardBase {
+// Single cache shard interface.
+class CacheShard {
  public:
-  explicit CacheShardBase(CacheMetadataChargePolicy metadata_charge_policy)
+  explicit CacheShard(CacheMetadataChargePolicy metadata_charge_policy)
       : metadata_charge_policy_(metadata_charge_policy) {}
+  virtual ~CacheShard() = default;
 
   using DeleterFn = Cache::DeleterFn;
-
-  // Expected by concept CacheShard (TODO with C++20 support)
-  // Some Defaults
-  std::string GetPrintableOptions() const { return ""; }
-  using HashVal = uint64_t;
-  using HashCref = uint64_t;
-  static inline HashVal ComputeHash(const Slice& key) {
-    return GetSliceNPHash64(key);
-  }
-  static inline uint32_t HashPieceForSharding(HashCref hash) {
-    return Lower32of64(hash);
-  }
-  void AppendPrintableOptions(std::string& /*str*/) const {}
-
-  // Must be provided for concept CacheShard (TODO with C++20 support)
-  /*
-  struct HandleImpl {  // for concept HandleImpl
-    HashVal hash;
-    HashCref GetHash() const;
-    ...
-  };
-  Status Insert(const Slice& key, HashCref hash, Cache::ObjectPtr value,
-                const Cache::CacheItemHelper* helper, size_t charge,
-                HandleImpl** handle, Cache::Priority priority,
-                bool standalone) = 0;
-  Handle* CreateStandalone(const Slice& key, HashCref hash, ObjectPtr obj,
-                           const CacheItemHelper* helper,
-                           size_t charge, bool allow_uncharged) = 0;
-  HandleImpl* Lookup(const Slice& key, HashCref hash,
-                        const Cache::CacheItemHelper* helper,
-                        Cache::CreateContext* create_context,
-                        Cache::Priority priority,
-                        Statistics* stats) = 0;
-  bool Release(HandleImpl* handle, bool useful, bool erase_if_last_ref) = 0;
-  bool Ref(HandleImpl* handle) = 0;
-  void Erase(const Slice& key, HashCref hash) = 0;
-  void SetCapacity(size_t capacity) = 0;
-  void SetStrictCapacityLimit(bool strict_capacity_limit) = 0;
-  size_t GetUsage() const = 0;
-  size_t GetPinnedUsage() const = 0;
-  size_t GetOccupancyCount() const = 0;
-  size_t GetTableAddressCount() const = 0;
+  virtual Status Insert(const Slice& key, uint32_t hash, void* value,
+                        size_t charge, DeleterFn deleter,
+                        Cache::Handle** handle, Cache::Priority priority) = 0;
+  virtual Status Insert(const Slice& key, uint32_t hash, void* value,
+                        const Cache::CacheItemHelper* helper, size_t charge,
+                        Cache::Handle** handle, Cache::Priority priority) = 0;
+  virtual Cache::Handle* Lookup(const Slice& key, uint32_t hash) = 0;
+  virtual Cache::Handle* Lookup(const Slice& key, uint32_t hash,
+                                const Cache::CacheItemHelper* helper,
+                                const Cache::CreateCallback& create_cb,
+                                Cache::Priority priority, bool wait,
+                                Statistics* stats) = 0;
+  virtual bool Release(Cache::Handle* handle, bool useful,
+                       bool erase_if_last_ref) = 0;
+  virtual bool IsReady(Cache::Handle* handle) = 0;
+  virtual void Wait(Cache::Handle* handle) = 0;
+  virtual bool Ref(Cache::Handle* handle) = 0;
+  virtual bool Release(Cache::Handle* handle, bool erase_if_last_ref) = 0;
+  virtual void Erase(const Slice& key, uint32_t hash) = 0;
+  virtual void SetCapacity(size_t capacity) = 0;
+  virtual void SetStrictCapacityLimit(bool strict_capacity_limit) = 0;
+  virtual size_t GetUsage() const = 0;
+  virtual size_t GetPinnedUsage() const = 0;
+  virtual size_t GetOccupancyCount() const = 0;
+  virtual size_t GetTableAddressCount() const = 0;
   // Handles iterating over roughly `average_entries_per_lock` entries, using
   // `state` to somehow record where it last ended up. Caller initially uses
-  // *state == 0 and implementation sets *state = SIZE_MAX to indicate
+  // *state == 0 and implementation sets *state = UINT32_MAX to indicate
   // completion.
-  void ApplyToSomeEntries(
-      const std::function<void(const Slice& key, ObjectPtr value,
-                               size_t charge,
-                               const Cache::CacheItemHelper* helper)>& callback,
-      size_t average_entries_per_lock, size_t* state) = 0;
-  void EraseUnRefEntries() = 0;
-  */
+  virtual void ApplyToSomeEntries(
+      const std::function<void(const Slice& key, void* value, size_t charge,
+                               DeleterFn deleter)>& callback,
+      uint32_t average_entries_per_lock, uint32_t* state) = 0;
+  virtual void EraseUnRefEntries() = 0;
+  virtual std::string GetPrintableOptions() const { return ""; }
 
  protected:
   const CacheMetadataChargePolicy metadata_charge_policy_;
 };
 
-// Portions of ShardedCache that do not depend on the template parameter
-class ShardedCacheBase : public Cache {
+// Generic cache interface which shards cache by hash of keys. 2^num_shard_bits
+// shards will be created, with capacity split evenly to each of the shards.
+// Keys are sharded by the highest num_shard_bits bits of hash value.
+class ShardedCache : public Cache {
  public:
-  ShardedCacheBase(size_t capacity, int num_shard_bits,
-                   bool strict_capacity_limit,
-                   std::shared_ptr<MemoryAllocator> memory_allocator);
-  virtual ~ShardedCacheBase() = default;
+  ShardedCache(size_t capacity, int num_shard_bits, bool strict_capacity_limit,
+               std::shared_ptr<MemoryAllocator> memory_allocator = nullptr);
+  virtual ~ShardedCache() = default;
+  virtual CacheShard* GetShard(uint32_t shard) = 0;
+  virtual const CacheShard* GetShard(uint32_t shard) const = 0;
+
+  virtual uint32_t GetHash(Handle* handle) const = 0;
+
+  virtual void SetCapacity(size_t capacity) override;
+  virtual void SetStrictCapacityLimit(bool strict_capacity_limit) override;
+
+  virtual Status Insert(const Slice& key, void* value, size_t charge,
+                        DeleterFn deleter, Handle** handle,
+                        Priority priority) override;
+  virtual Status Insert(const Slice& key, void* value,
+                        const CacheItemHelper* helper, size_t charge,
+                        Handle** handle = nullptr,
+                        Priority priority = Priority::LOW) override;
+  virtual Handle* Lookup(const Slice& key, Statistics* stats) override;
+  virtual Handle* Lookup(const Slice& key, const CacheItemHelper* helper,
+                         const CreateCallback& create_cb, Priority priority,
+                         bool wait, Statistics* stats = nullptr) override;
+  virtual bool Release(Handle* handle, bool useful,
+                       bool erase_if_last_ref = false) override;
+  virtual bool IsReady(Handle* handle) override;
+  virtual void Wait(Handle* handle) override;
+  virtual bool Ref(Handle* handle) override;
+  virtual bool Release(Handle* handle, bool erase_if_last_ref = false) override;
+  virtual void Erase(const Slice& key) override;
+  virtual uint64_t NewId() override;
+  virtual size_t GetCapacity() const override;
+  virtual bool HasStrictCapacityLimit() const override;
+  virtual size_t GetUsage() const override;
+  virtual size_t GetUsage(Handle* handle) const override;
+  virtual size_t GetPinnedUsage() const override;
+  virtual size_t GetOccupancyCount() const override;
+  virtual size_t GetTableAddressCount() const override;
+  virtual void ApplyToAllEntries(
+      const std::function<void(const Slice& key, void* value, size_t charge,
+                               DeleterFn deleter)>& callback,
+      const ApplyToAllEntriesOptions& opts) override;
+  virtual void EraseUnRefEntries() override;
+  virtual std::string GetPrintableOptions() const override;
 
   int GetNumShardBits() const;
   uint32_t GetNumShards() const;
 
-  uint64_t NewId() override;
-
-  bool HasStrictCapacityLimit() const override;
-  size_t GetCapacity() const override;
-
-  using Cache::GetUsage;
-  size_t GetUsage(Handle* handle) const override;
-  std::string GetPrintableOptions() const override;
-
- protected:  // fns
-  virtual void AppendPrintableOptions(std::string& str) const = 0;
-  size_t GetPerShardCapacity() const;
-  size_t ComputePerShardCapacity(size_t capacity) const;
-
- protected:                        // data
-  std::atomic<uint64_t> last_id_;  // For NewId
-  const uint32_t shard_mask_;
-
-  // Dynamic configuration parameters, guarded by config_mutex_
-  bool strict_capacity_limit_;
-  size_t capacity_;
-  mutable port::Mutex config_mutex_;
-};
-
-// Generic cache interface that shards cache by hash of keys. 2^num_shard_bits
-// shards will be created, with capacity split evenly to each of the shards.
-// Keys are typically sharded by the lowest num_shard_bits bits of hash value
-// so that the upper bits of the hash value can keep a stable ordering of
-// table entries even as the table grows (using more upper hash bits).
-// See CacheShardBase above for what is expected of the CacheShard parameter.
-template <class CacheShard>
-class ShardedCache : public ShardedCacheBase {
- public:
-  using HashVal = typename CacheShard::HashVal;
-  using HashCref = typename CacheShard::HashCref;
-  using HandleImpl = typename CacheShard::HandleImpl;
-
-  ShardedCache(size_t capacity, int num_shard_bits, bool strict_capacity_limit,
-               std::shared_ptr<MemoryAllocator> allocator)
-      : ShardedCacheBase(capacity, num_shard_bits, strict_capacity_limit,
-                         allocator),
-        shards_(reinterpret_cast<CacheShard*>(port::cacheline_aligned_alloc(
-            sizeof(CacheShard) * GetNumShards()))),
-        destroy_shards_in_dtor_(false) {}
-
-  virtual ~ShardedCache() {
-    if (destroy_shards_in_dtor_) {
-      ForEachShard([](CacheShard* cs) { cs->~CacheShard(); });
-    }
-    port::cacheline_aligned_free(shards_);
-  }
-
-  CacheShard& GetShard(HashCref hash) {
-    return shards_[CacheShard::HashPieceForSharding(hash) & shard_mask_];
-  }
-
-  const CacheShard& GetShard(HashCref hash) const {
-    return shards_[CacheShard::HashPieceForSharding(hash) & shard_mask_];
-  }
-
-  void SetCapacity(size_t capacity) override {
-    MutexLock l(&config_mutex_);
-    capacity_ = capacity;
-    auto per_shard = ComputePerShardCapacity(capacity);
-    ForEachShard([=](CacheShard* cs) { cs->SetCapacity(per_shard); });
-  }
-
-  void SetStrictCapacityLimit(bool s_c_l) override {
-    MutexLock l(&config_mutex_);
-    strict_capacity_limit_ = s_c_l;
-    ForEachShard(
-        [s_c_l](CacheShard* cs) { cs->SetStrictCapacityLimit(s_c_l); });
-  }
-
-  Status Insert(const Slice& key, ObjectPtr obj, const CacheItemHelper* helper,
-                size_t charge, Handle** handle = nullptr,
-                Priority priority = Priority::LOW) override {
-    assert(helper);
-    HashVal hash = CacheShard::ComputeHash(key);
-    auto h_out = reinterpret_cast<HandleImpl**>(handle);
-    return GetShard(hash).Insert(key, hash, obj, helper, charge, h_out,
-                                 priority);
-  }
-
-  Handle* CreateStandalone(const Slice& key, ObjectPtr obj,
-                           const CacheItemHelper* helper, size_t charge,
-                           bool allow_uncharged) override {
-    assert(helper);
-    HashVal hash = CacheShard::ComputeHash(key);
-    HandleImpl* result = GetShard(hash).CreateStandalone(
-        key, hash, obj, helper, charge, allow_uncharged);
-    return reinterpret_cast<Handle*>(result);
-  }
-
-  Handle* Lookup(const Slice& key, const CacheItemHelper* helper = nullptr,
-                 CreateContext* create_context = nullptr,
-                 Priority priority = Priority::LOW,
-                 Statistics* stats = nullptr) override {
-    HashVal hash = CacheShard::ComputeHash(key);
-    HandleImpl* result = GetShard(hash).Lookup(key, hash, helper,
-                                               create_context, priority, stats);
-    return reinterpret_cast<Handle*>(result);
-  }
-
-  void Erase(const Slice& key) override {
-    HashVal hash = CacheShard::ComputeHash(key);
-    GetShard(hash).Erase(key, hash);
-  }
-
-  bool Release(Handle* handle, bool useful,
-               bool erase_if_last_ref = false) override {
-    auto h = reinterpret_cast<HandleImpl*>(handle);
-    return GetShard(h->GetHash()).Release(h, useful, erase_if_last_ref);
-  }
-  bool Ref(Handle* handle) override {
-    auto h = reinterpret_cast<HandleImpl*>(handle);
-    return GetShard(h->GetHash()).Ref(h);
-  }
-  bool Release(Handle* handle, bool erase_if_last_ref = false) override {
-    return Release(handle, true /*useful*/, erase_if_last_ref);
-  }
-  using ShardedCacheBase::GetUsage;
-  size_t GetUsage() const override {
-    return SumOverShards2(&CacheShard::GetUsage);
-  }
-  size_t GetPinnedUsage() const override {
-    return SumOverShards2(&CacheShard::GetPinnedUsage);
-  }
-  size_t GetOccupancyCount() const override {
-    return SumOverShards2(&CacheShard::GetOccupancyCount);
-  }
-  size_t GetTableAddressCount() const override {
-    return SumOverShards2(&CacheShard::GetTableAddressCount);
-  }
-  void ApplyToAllEntries(
-      const std::function<void(const Slice& key, ObjectPtr value, size_t charge,
-                               const CacheItemHelper* helper)>& callback,
-      const ApplyToAllEntriesOptions& opts) override {
-    uint32_t num_shards = GetNumShards();
-    // Iterate over part of each shard, rotating between shards, to
-    // minimize impact on latency of concurrent operations.
-    std::unique_ptr<size_t[]> states(new size_t[num_shards]{});
-
-    size_t aepl = opts.average_entries_per_lock;
-    aepl = std::min(aepl, size_t{1});
-
-    bool remaining_work;
-    do {
-      remaining_work = false;
-      for (uint32_t i = 0; i < num_shards; i++) {
-        if (states[i] != SIZE_MAX) {
-          shards_[i].ApplyToSomeEntries(callback, aepl, &states[i]);
-          remaining_work |= states[i] != SIZE_MAX;
-        }
-      }
-    } while (remaining_work);
-  }
-
-  virtual void EraseUnRefEntries() override {
-    ForEachShard([](CacheShard* cs) { cs->EraseUnRefEntries(); });
-  }
-
-  void DisownData() override {
-    // Leak data only if that won't generate an ASAN/valgrind warning.
-    if (!kMustFreeHeapAllocations) {
-      destroy_shards_in_dtor_ = false;
-    }
-  }
-
  protected:
-  inline void ForEachShard(const std::function<void(CacheShard*)>& fn) {
-    uint32_t num_shards = GetNumShards();
-    for (uint32_t i = 0; i < num_shards; i++) {
-      fn(shards_ + i);
-    }
-  }
-
-  inline size_t SumOverShards(
-      const std::function<size_t(CacheShard&)>& fn) const {
-    uint32_t num_shards = GetNumShards();
-    size_t result = 0;
-    for (uint32_t i = 0; i < num_shards; i++) {
-      result += fn(shards_[i]);
-    }
-    return result;
-  }
-
-  inline size_t SumOverShards2(size_t (CacheShard::*fn)() const) const {
-    return SumOverShards([fn](CacheShard& cs) { return (cs.*fn)(); });
-  }
-
-  // Must be called exactly once by derived class constructor
-  void InitShards(const std::function<void(CacheShard*)>& placement_new) {
-    ForEachShard(placement_new);
-    destroy_shards_in_dtor_ = true;
-  }
-
-  void AppendPrintableOptions(std::string& str) const override {
-    shards_[0].AppendPrintableOptions(str);
-  }
+  inline uint32_t Shard(uint32_t hash) { return hash & shard_mask_; }
 
  private:
-  CacheShard* const shards_;
-  bool destroy_shards_in_dtor_;
+  const uint32_t shard_mask_;
+  mutable port::Mutex capacity_mutex_;
+  size_t capacity_;
+  bool strict_capacity_limit_;
+  std::atomic<uint64_t> last_id_;
 };
 
 // 512KB is traditional minimum shard size.

cache/typed_cache.h

@@ -1,375 +0,0 @@
-//  Copyright (c) Meta Platforms, Inc. and affiliates.
-//  This source code is licensed under both the GPLv2 (found in the
-//  COPYING file in the root directory) and Apache 2.0 License
-//  (found in the LICENSE.Apache file in the root directory).
-
-// APIs for accessing Cache in a type-safe and convenient way. Cache is kept
-// at a low, thin level of abstraction so that different implementations can
-// be plugged in, but these wrappers provide clean, convenient access to the
-// most common operations.
-//
-// A number of template classes are needed for sharing common structure. The
-// key classes are these:
-//
-// * PlaceholderCacheInterface - Used for making cache reservations, with
-// entries that have a charge but no value.
-// * BasicTypedCacheInterface<TValue> - Used for primary cache storage of
-// objects of type TValue.
-// * FullTypedCacheHelper<TValue, TCreateContext> - Used for secondary cache
-// compatible storage of objects of type TValue.
-// * For each of these, there's a "Shared" version
-// (e.g. FullTypedSharedCacheInterface) that holds a shared_ptr to the Cache,
-// rather than assuming external ownership by holding only a raw `Cache*`.
-
-#pragma once
-
-#include <algorithm>
-#include <cstdint>
-#include <memory>
-#include <type_traits>
-
-#include "cache/cache_helpers.h"
-#include "rocksdb/advanced_cache.h"
-#include "rocksdb/advanced_options.h"
-
-namespace ROCKSDB_NAMESPACE {
-
-// For future consideration:
-// * Pass in value to Insert with std::unique_ptr& to simplify ownership
-//   transfer logic in callers
-// * Make key type a template parameter (e.g. useful for table cache)
-// * Closer integration with CacheHandleGuard (opt-in, so not always
-//   paying the extra overhead)
-
-#define CACHE_TYPE_DEFS()                     \
-  using Priority = Cache::Priority;           \
-  using Handle = Cache::Handle;               \
-  using ObjectPtr = Cache::ObjectPtr;         \
-  using CreateContext = Cache::CreateContext; \
-  using CacheItemHelper = Cache::CacheItemHelper /* caller ; */
-
-template <typename CachePtr>
-class BaseCacheInterface {
- public:
-  CACHE_TYPE_DEFS();
-
-  /*implicit*/ BaseCacheInterface(CachePtr cache) : cache_(std::move(cache)) {}
-
-  inline void Release(Handle* handle) { cache_->Release(handle); }
-
-  inline void ReleaseAndEraseIfLastRef(Handle* handle) {
-    cache_->Release(handle, /*erase_if_last_ref*/ true);
-  }
-
-  inline void RegisterReleaseAsCleanup(Handle* handle, Cleanable& cleanable) {
-    cleanable.RegisterCleanup(&ReleaseCacheHandleCleanup, get(), handle);
-  }
-
-  inline Cache* get() const { return &*cache_; }
-
-  explicit inline operator bool() const noexcept { return cache_ != nullptr; }
-
- protected:
-  CachePtr cache_;
-};
-
-// PlaceholderCacheInterface - Used for making cache reservations, with
-// entries that have a charge but no value. CacheEntryRole is required as
-// a template parameter.
-template <CacheEntryRole kRole, typename CachePtr = Cache*>
-class PlaceholderCacheInterface : public BaseCacheInterface<CachePtr> {
- public:
-  CACHE_TYPE_DEFS();
-  using BaseCacheInterface<CachePtr>::BaseCacheInterface;
-
-  inline Status Insert(const Slice& key, size_t charge, Handle** handle) {
-    return this->cache_->Insert(key, /*value=*/nullptr, GetHelper(), charge,
-                                handle);
-  }
-
-  static const Cache::CacheItemHelper* GetHelper() {
-    static const Cache::CacheItemHelper kHelper{kRole};
-    return &kHelper;
-  }
-};
-
-template <CacheEntryRole kRole>
-using PlaceholderSharedCacheInterface =
-    PlaceholderCacheInterface<kRole, std::shared_ptr<Cache>>;
-
-template <class TValue>
-class BasicTypedCacheHelperFns {
- public:
-  CACHE_TYPE_DEFS();
-  // E.g. char* for char[]
-  using TValuePtr = std::remove_extent_t<TValue>*;
-
- protected:
-  inline static ObjectPtr UpCastValue(TValuePtr value) { return value; }
-  inline static TValuePtr DownCastValue(ObjectPtr value) {
-    return static_cast<TValuePtr>(value);
-  }
-
-  static void Delete(ObjectPtr value, MemoryAllocator* allocator) {
-    // FIXME: Currently, no callers actually allocate the ObjectPtr objects
-    // using the custom allocator, just subobjects that keep a reference to
-    // the allocator themselves (with CacheAllocationPtr).
-    if (/*DISABLED*/ false && allocator) {
-      if constexpr (std::is_destructible_v<TValue>) {
-        DownCastValue(value)->~TValue();
-      }
-      allocator->Deallocate(value);
-    } else {
-      // Like delete but properly handles TValue=char[] etc.
-      std::default_delete<TValue>{}(DownCastValue(value));
-    }
-  }
-};
-
-// In its own class to try to minimize the number of distinct CacheItemHelper
-// instances (e.g. don't vary by CachePtr)
-template <class TValue, CacheEntryRole kRole>
-class BasicTypedCacheHelper : public BasicTypedCacheHelperFns<TValue> {
- public:
-  static const Cache::CacheItemHelper* GetBasicHelper() {
-    static const Cache::CacheItemHelper kHelper{kRole,
-                                                &BasicTypedCacheHelper::Delete};
-    return &kHelper;
-  }
-};
-
-// BasicTypedCacheInterface - Used for primary cache storage of objects of
-// type TValue, which can be cleaned up with std::default_delete<TValue>. The
-// role is provided by TValue::kCacheEntryRole or given in an optional
-// template parameter.
-template <class TValue, CacheEntryRole kRole = TValue::kCacheEntryRole,
-          typename CachePtr = Cache*>
-class BasicTypedCacheInterface : public BaseCacheInterface<CachePtr>,
-                                 public BasicTypedCacheHelper<TValue, kRole> {
- public:
-  CACHE_TYPE_DEFS();
-  using typename BasicTypedCacheHelperFns<TValue>::TValuePtr;
-  struct TypedHandle : public Handle {};
-  using BasicTypedCacheHelper<TValue, kRole>::GetBasicHelper;
-  // ctor
-  using BaseCacheInterface<CachePtr>::BaseCacheInterface;
-  struct TypedAsyncLookupHandle : public Cache::AsyncLookupHandle {
-    TypedHandle* Result() {
-      return reinterpret_cast<TypedHandle*>(Cache::AsyncLookupHandle::Result());
-    }
-  };
-
-  inline Status Insert(const Slice& key, TValuePtr value, size_t charge,
-                       TypedHandle** handle = nullptr,
-                       Priority priority = Priority::LOW) {
-    auto untyped_handle = reinterpret_cast<Handle**>(handle);
-    return this->cache_->Insert(
-        key, BasicTypedCacheHelperFns<TValue>::UpCastValue(value),
-        GetBasicHelper(), charge, untyped_handle, priority);
-  }
-
-  inline TypedHandle* Lookup(const Slice& key, Statistics* stats = nullptr) {
-    return reinterpret_cast<TypedHandle*>(
-        this->cache_->BasicLookup(key, stats));
-  }
-
-  inline void StartAsyncLookup(TypedAsyncLookupHandle& async_handle) {
-    assert(async_handle.helper == nullptr);
-    this->cache_->StartAsyncLookup(async_handle);
-  }
-
-  inline CacheHandleGuard<TValue> Guard(TypedHandle* handle) {
-    if (handle) {
-      return CacheHandleGuard<TValue>(&*this->cache_, handle);
-    } else {
-      return {};
-    }
-  }
-
-  inline std::shared_ptr<TValue> SharedGuard(TypedHandle* handle) {
-    if (handle) {
-      return MakeSharedCacheHandleGuard<TValue>(&*this->cache_, handle);
-    } else {
-      return {};
-    }
-  }
-
-  inline TValuePtr Value(TypedHandle* handle) {
-    return BasicTypedCacheHelperFns<TValue>::DownCastValue(
-        this->cache_->Value(handle));
-  }
-};
-
-// BasicTypedSharedCacheInterface - Like BasicTypedCacheInterface but with a
-// shared_ptr<Cache> for keeping Cache alive.
-template <class TValue, CacheEntryRole kRole = TValue::kCacheEntryRole>
-using BasicTypedSharedCacheInterface =
-    BasicTypedCacheInterface<TValue, kRole, std::shared_ptr<Cache>>;
-
-// TValue must implement ContentSlice() and ~TValue
-// TCreateContext must implement Create(std::unique_ptr<TValue>*, ...)
-template <class TValue, class TCreateContext>
-class FullTypedCacheHelperFns : public BasicTypedCacheHelperFns<TValue> {
- public:
-  CACHE_TYPE_DEFS();
-
- protected:
-  using typename BasicTypedCacheHelperFns<TValue>::TValuePtr;
-  using BasicTypedCacheHelperFns<TValue>::DownCastValue;
-  using BasicTypedCacheHelperFns<TValue>::UpCastValue;
-
-  static size_t Size(ObjectPtr v) {
-    TValuePtr value = DownCastValue(v);
-    auto slice = value->ContentSlice();
-    return slice.size();
-  }
-
-  static Status SaveTo(ObjectPtr v, size_t from_offset, size_t length,
-                       char* out) {
-    TValuePtr value = DownCastValue(v);
-    auto slice = value->ContentSlice();
-    assert(from_offset < slice.size());
-    assert(from_offset + length <= slice.size());
-    std::copy_n(slice.data() + from_offset, length, out);
-    return Status::OK();
-  }
-
-  static Status Create(const Slice& data, CreateContext* context,
-                       MemoryAllocator* allocator, ObjectPtr* out_obj,
-                       size_t* out_charge) {
-    std::unique_ptr<TValue> value = nullptr;
-    if constexpr (sizeof(TCreateContext) > 0) {
-      TCreateContext* tcontext = static_cast<TCreateContext*>(context);
-      tcontext->Create(&value, out_charge, data, allocator);
-    } else {
-      TCreateContext::Create(&value, out_charge, data, allocator);
-    }
-    *out_obj = UpCastValue(value.release());
-    return Status::OK();
-  }
-};
-
-// In its own class to try to minimize the number of distinct CacheItemHelper
-// instances (e.g. don't vary by CachePtr)
-template <class TValue, class TCreateContext, CacheEntryRole kRole>
-class FullTypedCacheHelper
-    : public FullTypedCacheHelperFns<TValue, TCreateContext> {
- public:
-  static const Cache::CacheItemHelper* GetFullHelper() {
-    static const Cache::CacheItemHelper kHelper{
-        kRole,
-        &FullTypedCacheHelper::Delete,
-        &FullTypedCacheHelper::Size,
-        &FullTypedCacheHelper::SaveTo,
-        &FullTypedCacheHelper::Create,
-        BasicTypedCacheHelper<TValue, kRole>::GetBasicHelper()};
-    return &kHelper;
-  }
-};
-
-// FullTypedCacheHelper - Used for secondary cache compatible storage of
-// objects of type TValue. In addition to BasicTypedCacheInterface constraints,
-// we require TValue::ContentSlice() to return persistable data. This
-// simplifies usage for the normal case of simple secondary cache compatibility
-// (can give you a Slice to the data already in memory). In addition to
-// TCreateContext performing the role of Cache::CreateContext, it is also
-// expected to provide a function Create(std::unique_ptr<TValue>* value,
-// size_t* out_charge, const Slice& data, MemoryAllocator* allocator) for
-// creating new TValue.
-template <class TValue, class TCreateContext,
-          CacheEntryRole kRole = TValue::kCacheEntryRole,
-          typename CachePtr = Cache*>
-class FullTypedCacheInterface
-    : public BasicTypedCacheInterface<TValue, kRole, CachePtr>,
-      public FullTypedCacheHelper<TValue, TCreateContext, kRole> {
- public:
-  CACHE_TYPE_DEFS();
-  using typename BasicTypedCacheInterface<TValue, kRole, CachePtr>::TypedHandle;
-  using typename BasicTypedCacheInterface<TValue, kRole,
-                                          CachePtr>::TypedAsyncLookupHandle;
-  using typename BasicTypedCacheHelperFns<TValue>::TValuePtr;
-  using BasicTypedCacheHelper<TValue, kRole>::GetBasicHelper;
-  using FullTypedCacheHelper<TValue, TCreateContext, kRole>::GetFullHelper;
-  using BasicTypedCacheHelperFns<TValue>::UpCastValue;
-  using BasicTypedCacheHelperFns<TValue>::DownCastValue;
-  // ctor
-  using BasicTypedCacheInterface<TValue, kRole,
-                                 CachePtr>::BasicTypedCacheInterface;
-
-  // Insert with SecondaryCache compatibility (subject to CacheTier).
-  // (Basic Insert() also inherited.)
-  inline Status InsertFull(
-      const Slice& key, TValuePtr value, size_t charge,
-      TypedHandle** handle = nullptr, Priority priority = Priority::LOW,
-      CacheTier lowest_used_cache_tier = CacheTier::kNonVolatileBlockTier) {
-    auto untyped_handle = reinterpret_cast<Handle**>(handle);
-    auto helper = lowest_used_cache_tier == CacheTier::kNonVolatileBlockTier
-                      ? GetFullHelper()
-                      : GetBasicHelper();
-    return this->cache_->Insert(key, UpCastValue(value), helper, charge,
-                                untyped_handle, priority);
-  }
-
-  // Like SecondaryCache::InsertSaved, with SecondaryCache compatibility
-  // (subject to CacheTier).
-  inline Status InsertSaved(
-      const Slice& key, const Slice& data, TCreateContext* create_context,
-      Priority priority = Priority::LOW,
-      CacheTier lowest_used_cache_tier = CacheTier::kNonVolatileBlockTier,
-      size_t* out_charge = nullptr) {
-    ObjectPtr value;
-    size_t charge;
-    Status st = GetFullHelper()->create_cb(data, create_context,
-                                           this->cache_->memory_allocator(),
-                                           &value, &charge);
-    if (out_charge) {
-      *out_charge = charge;
-    }
-    if (st.ok()) {
-      st = InsertFull(key, DownCastValue(value), charge, nullptr /*handle*/,
-                      priority, lowest_used_cache_tier);
-    } else {
-      GetFullHelper()->del_cb(value, this->cache_->memory_allocator());
-    }
-    return st;
-  }
-
-  // Lookup with SecondaryCache support (subject to CacheTier).
-  // (Basic Lookup() also inherited.)
-  inline TypedHandle* LookupFull(
-      const Slice& key, TCreateContext* create_context = nullptr,
-      Priority priority = Priority::LOW, Statistics* stats = nullptr,
-      CacheTier lowest_used_cache_tier = CacheTier::kNonVolatileBlockTier) {
-    if (lowest_used_cache_tier == CacheTier::kNonVolatileBlockTier) {
-      return reinterpret_cast<TypedHandle*>(this->cache_->Lookup(
-          key, GetFullHelper(), create_context, priority, stats));
-    } else {
-      return BasicTypedCacheInterface<TValue, kRole, CachePtr>::Lookup(key,
-                                                                       stats);
-    }
-  }
-
-  inline void StartAsyncLookupFull(
-      TypedAsyncLookupHandle& async_handle,
-      CacheTier lowest_used_cache_tier = CacheTier::kNonVolatileBlockTier) {
-    if (lowest_used_cache_tier == CacheTier::kNonVolatileBlockTier) {
-      async_handle.helper = GetFullHelper();
-      this->cache_->StartAsyncLookup(async_handle);
-    } else {
-      BasicTypedCacheInterface<TValue, kRole, CachePtr>::StartAsyncLookup(
-          async_handle);
-    }
-  }
-};
-
-// FullTypedSharedCacheInterface - Like FullTypedCacheInterface but with a
-// shared_ptr<Cache> for keeping Cache alive.
-template <class TValue, class TCreateContext,
-          CacheEntryRole kRole = TValue::kCacheEntryRole>
-using FullTypedSharedCacheInterface =
-    FullTypedCacheInterface<TValue, TCreateContext, kRole,
-                            std::shared_ptr<Cache>>;
-
-#undef CACHE_TYPE_DEFS
-
-}  // namespace ROCKSDB_NAMESPACE

coverage/coverage_test.sh

@@ -12,7 +12,7 @@ fi
 ROOT=".."
 # Fetch right version of gcov
 if [ -d /mnt/gvfs/third-party -a -z "$CXX" ]; then
-  source $ROOT/build_tools/fbcode_config_platform010.sh
+  source $ROOT/build_tools/fbcode_config_platform009.sh
   GCOV=$GCC_BASE/bin/gcov
 else
   GCOV=$(which gcov)

coverage/parse_gcov_output.py

@@ -47,39 +47,35 @@ def parse_gcov_report(gcov_input):
 
     return per_file_coverage, total_coverage
 
-
 def get_option_parser():
-    usage = (
-        "Parse the gcov output and generate more human-readable code "
-        + "coverage report."
-    )
+    usage = "Parse the gcov output and generate more human-readable code " +\
+            "coverage report."
     parser = optparse.OptionParser(usage)
 
     parser.add_option(
-        "--interested-files",
-        "-i",
+        "--interested-files", "-i",
         dest="filenames",
-        help="Comma separated files names. if specified, we will display "
-        + "the coverage report only for interested source files. "
-        + "Otherwise we will display the coverage report for all "
-        + "source files.",
+        help="Comma separated files names. if specified, we will display " +
+             "the coverage report only for interested source files. " +
+             "Otherwise we will display the coverage report for all " +
+             "source files."
     )
     return parser
 
-
 def display_file_coverage(per_file_coverage, total_coverage):
     # To print out auto-adjustable column, we need to know the longest
     # length of file names.
-    max_file_name_length = max(len(fname) for fname in per_file_coverage.keys())
+    max_file_name_length = max(
+        len(fname) for fname in per_file_coverage.keys()
+    )
 
     # -- Print header
     # size of separator is determined by 3 column sizes:
     # file name, coverage percentage and lines.
-    header_template = "%" + str(max_file_name_length) + "s\t%s\t%s"
+    header_template = \
+        "%" + str(max_file_name_length) + "s\t%s\t%s"
     separator = "-" * (max_file_name_length + 10 + 20)
-    print(
-        header_template % ("Filename", "Coverage", "Lines")
-    )  # noqa: E999 T25377293 Grandfathered in
+    print(header_template % ("Filename", "Coverage", "Lines"))  # noqa: E999 T25377293 Grandfathered in
     print(separator)
 
     # -- Print body
@@ -95,14 +91,13 @@ def display_file_coverage(per_file_coverage, total_coverage):
         print(separator)
         print(record_template % ("Total", total_coverage[0], total_coverage[1]))
 
-
 def report_coverage():
     parser = get_option_parser()
     (options, args) = parser.parse_args()
 
     interested_files = set()
     if options.filenames is not None:
-        interested_files = {f.strip() for f in options.filenames.split(",")}
+        interested_files = set(f.strip() for f in options.filenames.split(','))
 
     # To make things simple, right now we only read gcov report from the input
     per_file_coverage, total_coverage = parse_gcov_report(sys.stdin)
@@ -110,8 +105,7 @@ def report_coverage():
     # Check if we need to display coverage info for interested files.
     if len(interested_files):
         per_file_coverage = dict(
-            (fname, per_file_coverage[fname])
-            for fname in interested_files
+            (fname, per_file_coverage[fname]) for fname in interested_files
             if fname in per_file_coverage
         )
         # If we only interested in several files, it makes no sense to report
@@ -123,6 +117,5 @@ def report_coverage():
         return
     display_file_coverage(per_file_coverage, total_coverage)
 
-
 if __name__ == "__main__":
     report_coverage()

crash_test.mk

@@ -8,7 +8,7 @@ DB_STRESS_CMD?=./db_stress
 include common.mk
 
 CRASHTEST_MAKE=$(MAKE) -f crash_test.mk
-CRASHTEST_PY=$(PYTHON) -u tools/db_crashtest.py --stress_cmd=$(DB_STRESS_CMD) --cleanup_cmd='$(DB_CLEANUP_CMD)'
+CRASHTEST_PY=$(PYTHON) -u tools/db_crashtest.py --stress_cmd=$(DB_STRESS_CMD)
 
 .PHONY: crash_test crash_test_with_atomic_flush crash_test_with_txn \
 	crash_test_with_best_efforts_recovery crash_test_with_ts \

db/arena_wrapped_db_iter.cc

@@ -8,7 +8,6 @@
 // found in the LICENSE file. See the AUTHORS file for names of contributors.
 
 #include "db/arena_wrapped_db_iter.h"
-
 #include "memory/arena.h"
 #include "rocksdb/env.h"
 #include "rocksdb/iterator.h"
@@ -47,9 +46,6 @@ void ArenaWrappedDBIter::Init(
   read_options_ = read_options;
   allow_refresh_ = allow_refresh;
   memtable_range_tombstone_iter_ = nullptr;
-  if (!env->GetFileSystem()->use_async_io()) {
-    read_options_.async_io = false;
-  }
 }
 
 Status ArenaWrappedDBIter::Refresh() {

db/arena_wrapped_db_iter.h

@@ -9,9 +9,7 @@
 
 #pragma once
 #include <stdint.h>
-
 #include <string>
-
 #include "db/db_impl/db_impl.h"
 #include "db/db_iter.h"
 #include "db/range_del_aggregator.h"

db/blob/blob_contents.cc

@@ -13,6 +13,12 @@
 
 namespace ROCKSDB_NAMESPACE {
 
+std::unique_ptr<BlobContents> BlobContents::Create(
+    CacheAllocationPtr&& allocation, size_t size) {
+  return std::unique_ptr<BlobContents>(
+      new BlobContents(std::move(allocation), size));
+}
+
 size_t BlobContents::ApproximateMemoryUsage() const {
   size_t usage = 0;
 
@@ -39,4 +45,46 @@ size_t BlobContents::ApproximateMemoryUsage() const {
   return usage;
 }
 
+size_t BlobContents::SizeCallback(void* obj) {
+  assert(obj);
+
+  return static_cast<const BlobContents*>(obj)->size();
+}
+
+Status BlobContents::SaveToCallback(void* from_obj, size_t from_offset,
+                                    size_t length, void* out) {
+  assert(from_obj);
+
+  const BlobContents* buf = static_cast<const BlobContents*>(from_obj);
+  assert(buf->size() >= from_offset + length);
+
+  memcpy(out, buf->data().data() + from_offset, length);
+
+  return Status::OK();
+}
+
+Cache::CacheItemHelper* BlobContents::GetCacheItemHelper() {
+  static Cache::CacheItemHelper cache_helper(
+      &SizeCallback, &SaveToCallback,
+      GetCacheEntryDeleterForRole<BlobContents, CacheEntryRole::kBlobValue>());
+
+  return &cache_helper;
+}
+
+Status BlobContents::CreateCallback(CacheAllocationPtr&& allocation,
+                                    const void* buf, size_t size,
+                                    void** out_obj, size_t* charge) {
+  assert(allocation);
+
+  memcpy(allocation.get(), buf, size);
+
+  std::unique_ptr<BlobContents> obj = Create(std::move(allocation), size);
+  BlobContents* const contents = obj.release();
+
+  *out_obj = contents;
+  *charge = contents->ApproximateMemoryUsage();
+
+  return Status::OK();
+}
+
 }  // namespace ROCKSDB_NAMESPACE

db/blob/blob_contents.h

@@ -8,7 +8,7 @@
 #include <memory>
 
 #include "memory/memory_allocator.h"
-#include "rocksdb/advanced_cache.h"
+#include "rocksdb/cache.h"
 #include "rocksdb/rocksdb_namespace.h"
 #include "rocksdb/slice.h"
 #include "rocksdb/status.h"
@@ -18,8 +18,8 @@ namespace ROCKSDB_NAMESPACE {
 // A class representing a single uncompressed value read from a blob file.
 class BlobContents {
  public:
-  BlobContents(CacheAllocationPtr&& allocation, size_t size)
-      : allocation_(std::move(allocation)), data_(allocation_.get(), size) {}
+  static std::unique_ptr<BlobContents> Create(CacheAllocationPtr&& allocation,
+                                              size_t size);
 
   BlobContents(const BlobContents&) = delete;
   BlobContents& operator=(const BlobContents&) = delete;
@@ -34,26 +34,23 @@ class BlobContents {
 
   size_t ApproximateMemoryUsage() const;
 
-  // For TypedCacheInterface
-  const Slice& ContentSlice() const { return data_; }
-  static constexpr CacheEntryRole kCacheEntryRole = CacheEntryRole::kBlobValue;
+  // Callbacks for secondary cache
+  static size_t SizeCallback(void* obj);
+
+  static Status SaveToCallback(void* from_obj, size_t from_offset,
+                               size_t length, void* out);
+
+  static Cache::CacheItemHelper* GetCacheItemHelper();
+
+  static Status CreateCallback(CacheAllocationPtr&& allocation, const void* buf,
+                               size_t size, void** out_obj, size_t* charge);
 
  private:
+  BlobContents(CacheAllocationPtr&& allocation, size_t size)
+      : allocation_(std::move(allocation)), data_(allocation_.get(), size) {}
+
   CacheAllocationPtr allocation_;
   Slice data_;
 };
 
-class BlobContentsCreator : public Cache::CreateContext {
- public:
-  static void Create(std::unique_ptr<BlobContents>* out, size_t* out_charge,
-                     const Slice& contents, MemoryAllocator* alloc) {
-    auto raw = new BlobContents(AllocateAndCopyBlock(contents, alloc),
-                                contents.size());
-    out->reset(raw);
-    if (out_charge) {
-      *out_charge = raw->ApproximateMemoryUsage();
-    }
-  }
-};
-
 }  // namespace ROCKSDB_NAMESPACE

db/blob/blob_counting_iterator.h

@@ -123,10 +123,6 @@ class BlobCountingIterator : public InternalIterator {
     return iter_->GetProperty(prop_name, prop);
   }
 
-  bool IsDeleteRangeSentinelKey() const override {
-    return iter_->IsDeleteRangeSentinelKey();
-  }
-
  private:
   void UpdateAndCountBlobIfNeeded() {
     assert(!iter_->Valid() || iter_->status().ok());

db/blob/blob_counting_iterator_test.cc

@@ -321,7 +321,6 @@ TEST(BlobCountingIteratorTest, CorruptBlobIndex) {
 }  // namespace ROCKSDB_NAMESPACE
 
 int main(int argc, char** argv) {
-  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
   ::testing::InitGoogleTest(&argc, argv);
   return RUN_ALL_TESTS();
 }

db/blob/blob_file_addition_test.cc

@@ -205,7 +205,6 @@ TEST_F(BlobFileAdditionTest, ForwardIncompatibleCustomField) {
 }  // namespace ROCKSDB_NAMESPACE
 
 int main(int argc, char** argv) {
-  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
   ::testing::InitGoogleTest(&argc, argv);
   return RUN_ALL_TESTS();
 }

db/blob/blob_file_builder.cc

@@ -13,7 +13,6 @@
 #include "db/blob/blob_index.h"
 #include "db/blob/blob_log_format.h"
 #include "db/blob/blob_log_writer.h"
-#include "db/blob/blob_source.h"
 #include "db/event_helpers.h"
 #include "db/version_set.h"
 #include "file/filename.h"
@@ -259,7 +258,6 @@ Status BlobFileBuilder::CompressBlobIfNeeded(
     return Status::OK();
   }
 
-  // TODO: allow user CompressionOptions, including max_compressed_bytes_per_kb
   CompressionOptions opts;
   CompressionContext context(blob_compression_type_);
   constexpr uint64_t sample_for_compression = 0;
@@ -395,7 +393,7 @@ Status BlobFileBuilder::PutBlobIntoCacheIfNeeded(const Slice& blob,
                                                  uint64_t blob_offset) const {
   Status s = Status::OK();
 
-  BlobSource::SharedCacheInterface blob_cache{immutable_options_->blob_cache};
+  auto blob_cache = immutable_options_->blob_cache;
   auto statistics = immutable_options_->statistics.get();
   bool warm_cache =
       prepopulate_blob_cache_ == PrepopulateBlobCache::kFlushOnly &&
@@ -409,12 +407,34 @@ Status BlobFileBuilder::PutBlobIntoCacheIfNeeded(const Slice& blob,
 
     const Cache::Priority priority = Cache::Priority::BOTTOM;
 
-    s = blob_cache.InsertSaved(key, blob, nullptr /*context*/, priority,
-                               immutable_options_->lowest_used_cache_tier);
+    // Objects to be put into the cache have to be heap-allocated and
+    // self-contained, i.e. own their contents. The Cache has to be able to
+    // take unique ownership of them.
+    CacheAllocationPtr allocation =
+        AllocateBlock(blob.size(), blob_cache->memory_allocator());
+    memcpy(allocation.get(), blob.data(), blob.size());
+    std::unique_ptr<BlobContents> buf =
+        BlobContents::Create(std::move(allocation), blob.size());
+
+    Cache::CacheItemHelper* const cache_item_helper =
+        BlobContents::GetCacheItemHelper();
+    assert(cache_item_helper);
+
+    if (immutable_options_->lowest_used_cache_tier ==
+        CacheTier::kNonVolatileBlockTier) {
+      s = blob_cache->Insert(key, buf.get(), cache_item_helper,
+                             buf->ApproximateMemoryUsage(),
+                             nullptr /* cache_handle */, priority);
+    } else {
+      s = blob_cache->Insert(key, buf.get(), buf->ApproximateMemoryUsage(),
+                             cache_item_helper->del_cb,
+                             nullptr /* cache_handle */, priority);
+    }
 
     if (s.ok()) {
       RecordTick(statistics, BLOB_DB_CACHE_ADD);
-      RecordTick(statistics, BLOB_DB_CACHE_BYTES_WRITE, blob.size());
+      RecordTick(statistics, BLOB_DB_CACHE_BYTES_WRITE, buf->size());
+      buf.release();
     } else {
       RecordTick(statistics, BLOB_DB_CACHE_ADD_FAILURES);
     }

db/blob/blob_file_builder_test.cc

@@ -674,7 +674,6 @@ TEST_P(BlobFileBuilderIOErrorTest, IOError) {
 }  // namespace ROCKSDB_NAMESPACE
 
 int main(int argc, char** argv) {
-  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
   ::testing::InitGoogleTest(&argc, argv);
   return RUN_ALL_TESTS();
 }

db/blob/blob_file_cache.cc

@@ -37,18 +37,18 @@ BlobFileCache::BlobFileCache(Cache* cache,
 }
 
 Status BlobFileCache::GetBlobFileReader(
-    const ReadOptions& read_options, uint64_t blob_file_number,
+    uint64_t blob_file_number,
     CacheHandleGuard<BlobFileReader>* blob_file_reader) {
   assert(blob_file_reader);
   assert(blob_file_reader->IsEmpty());
 
-  const Slice key = GetSliceForKey(&blob_file_number);
+  const Slice key = GetSlice(&blob_file_number);
 
   assert(cache_);
 
-  TypedHandle* handle = cache_.Lookup(key);
+  Cache::Handle* handle = cache_->Lookup(key);
   if (handle) {
-    *blob_file_reader = cache_.Guard(handle);
+    *blob_file_reader = CacheHandleGuard<BlobFileReader>(cache_, handle);
     return Status::OK();
   }
 
@@ -57,9 +57,9 @@ Status BlobFileCache::GetBlobFileReader(
   // Check again while holding mutex
   MutexLock lock(mutex_.get(key));
 
-  handle = cache_.Lookup(key);
+  handle = cache_->Lookup(key);
   if (handle) {
-    *blob_file_reader = cache_.Guard(handle);
+    *blob_file_reader = CacheHandleGuard<BlobFileReader>(cache_, handle);
     return Status::OK();
   }
 
@@ -73,7 +73,7 @@ Status BlobFileCache::GetBlobFileReader(
   {
     assert(file_options_);
     const Status s = BlobFileReader::Create(
-        *immutable_options_, read_options, *file_options_, column_family_id_,
+        *immutable_options_, *file_options_, column_family_id_,
         blob_file_read_hist_, blob_file_number, io_tracer_, &reader);
     if (!s.ok()) {
       RecordTick(statistics, NO_FILE_ERRORS);
@@ -84,7 +84,8 @@ Status BlobFileCache::GetBlobFileReader(
   {
     constexpr size_t charge = 1;
 
-    const Status s = cache_.Insert(key, reader.get(), charge, &handle);
+    const Status s = cache_->Insert(key, reader.get(), charge,
+                                    &DeleteCacheEntry<BlobFileReader>, &handle);
     if (!s.ok()) {
       RecordTick(statistics, NO_FILE_ERRORS);
       return s;
@@ -93,7 +94,7 @@ Status BlobFileCache::GetBlobFileReader(
 
   reader.release();
 
-  *blob_file_reader = cache_.Guard(handle);
+  *blob_file_reader = CacheHandleGuard<BlobFileReader>(cache_, handle);
 
   return Status::OK();
 }

db/blob/blob_file_cache.h

@@ -7,8 +7,7 @@
 
 #include <cinttypes>
 
-#include "cache/typed_cache.h"
-#include "db/blob/blob_file_reader.h"
+#include "cache/cache_helpers.h"
 #include "rocksdb/rocksdb_namespace.h"
 #include "util/mutexlock.h"
 
@@ -19,6 +18,7 @@ struct ImmutableOptions;
 struct FileOptions;
 class HistogramImpl;
 class Status;
+class BlobFileReader;
 class Slice;
 class IOTracer;
 
@@ -32,15 +32,11 @@ class BlobFileCache {
   BlobFileCache(const BlobFileCache&) = delete;
   BlobFileCache& operator=(const BlobFileCache&) = delete;
 
-  Status GetBlobFileReader(const ReadOptions& read_options,
-                           uint64_t blob_file_number,
+  Status GetBlobFileReader(uint64_t blob_file_number,
                            CacheHandleGuard<BlobFileReader>* blob_file_reader);
 
  private:
-  using CacheInterface =
-      BasicTypedCacheInterface<BlobFileReader, CacheEntryRole::kMisc>;
-  using TypedHandle = CacheInterface::TypedHandle;
-  CacheInterface cache_;
+  Cache* cache_;
   // Note: mutex_ below is used to guard against multiple threads racing to open
   // the same file.
   Striped<port::Mutex, Slice> mutex_;

db/blob/blob_file_cache_test.cc

@@ -118,9 +118,7 @@ TEST_F(BlobFileCacheTest, GetBlobFileReader) {
   // First try: reader should be opened and put in cache
   CacheHandleGuard<BlobFileReader> first;
 
-  const ReadOptions read_options;
-  ASSERT_OK(blob_file_cache.GetBlobFileReader(read_options, blob_file_number,
-                                              &first));
+  ASSERT_OK(blob_file_cache.GetBlobFileReader(blob_file_number, &first));
   ASSERT_NE(first.GetValue(), nullptr);
   ASSERT_EQ(options.statistics->getTickerCount(NO_FILE_OPENS), 1);
   ASSERT_EQ(options.statistics->getTickerCount(NO_FILE_ERRORS), 0);
@@ -128,8 +126,7 @@ TEST_F(BlobFileCacheTest, GetBlobFileReader) {
   // Second try: reader should be served from cache
   CacheHandleGuard<BlobFileReader> second;
 
-  ASSERT_OK(blob_file_cache.GetBlobFileReader(read_options, blob_file_number,
-                                              &second));
+  ASSERT_OK(blob_file_cache.GetBlobFileReader(blob_file_number, &second));
   ASSERT_NE(second.GetValue(), nullptr);
   ASSERT_EQ(options.statistics->getTickerCount(NO_FILE_OPENS), 1);
   ASSERT_EQ(options.statistics->getTickerCount(NO_FILE_ERRORS), 0);
@@ -166,21 +163,19 @@ TEST_F(BlobFileCacheTest, GetBlobFileReader_Race) {
   CacheHandleGuard<BlobFileReader> first;
   CacheHandleGuard<BlobFileReader> second;
 
-  const ReadOptions read_options;
   SyncPoint::GetInstance()->SetCallBack(
       "BlobFileCache::GetBlobFileReader:DoubleCheck", [&](void* /* arg */) {
         // Disabling sync points to prevent infinite recursion
         SyncPoint::GetInstance()->DisableProcessing();
-        ASSERT_OK(blob_file_cache.GetBlobFileReader(read_options,
-                                                    blob_file_number, &second));
+
+        ASSERT_OK(blob_file_cache.GetBlobFileReader(blob_file_number, &second));
         ASSERT_NE(second.GetValue(), nullptr);
         ASSERT_EQ(options.statistics->getTickerCount(NO_FILE_OPENS), 1);
         ASSERT_EQ(options.statistics->getTickerCount(NO_FILE_ERRORS), 0);
       });
   SyncPoint::GetInstance()->EnableProcessing();
 
-  ASSERT_OK(blob_file_cache.GetBlobFileReader(read_options, blob_file_number,
-                                              &first));
+  ASSERT_OK(blob_file_cache.GetBlobFileReader(blob_file_number, &first));
   ASSERT_NE(first.GetValue(), nullptr);
   ASSERT_EQ(options.statistics->getTickerCount(NO_FILE_OPENS), 1);
   ASSERT_EQ(options.statistics->getTickerCount(NO_FILE_ERRORS), 0);
@@ -218,10 +213,8 @@ TEST_F(BlobFileCacheTest, GetBlobFileReader_IOError) {
 
   CacheHandleGuard<BlobFileReader> reader;
 
-  const ReadOptions read_options;
   ASSERT_TRUE(
-      blob_file_cache.GetBlobFileReader(read_options, blob_file_number, &reader)
-          .IsIOError());
+      blob_file_cache.GetBlobFileReader(blob_file_number, &reader).IsIOError());
   ASSERT_EQ(reader.GetValue(), nullptr);
   ASSERT_EQ(options.statistics->getTickerCount(NO_FILE_OPENS), 1);
   ASSERT_EQ(options.statistics->getTickerCount(NO_FILE_ERRORS), 1);
@@ -260,10 +253,8 @@ TEST_F(BlobFileCacheTest, GetBlobFileReader_CacheFull) {
   // strict_capacity_limit is set
   CacheHandleGuard<BlobFileReader> reader;
 
-  const ReadOptions read_options;
-  ASSERT_TRUE(
-      blob_file_cache.GetBlobFileReader(read_options, blob_file_number, &reader)
-          .IsMemoryLimit());
+  ASSERT_TRUE(blob_file_cache.GetBlobFileReader(blob_file_number, &reader)
+                  .IsMemoryLimit());
   ASSERT_EQ(reader.GetValue(), nullptr);
   ASSERT_EQ(options.statistics->getTickerCount(NO_FILE_OPENS), 1);
   ASSERT_EQ(options.statistics->getTickerCount(NO_FILE_ERRORS), 1);
@@ -272,7 +263,6 @@ TEST_F(BlobFileCacheTest, GetBlobFileReader_CacheFull) {
 }  // namespace ROCKSDB_NAMESPACE
 
 int main(int argc, char** argv) {
-  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
   ::testing::InitGoogleTest(&argc, argv);
   return RUN_ALL_TESTS();
 }

db/blob/blob_file_completion_callback.h

@@ -23,19 +23,32 @@ class BlobFileCompletionCallback {
       const std::vector<std::shared_ptr<EventListener>>& listeners,
       const std::string& dbname)
       : event_logger_(event_logger), listeners_(listeners), dbname_(dbname) {
+#ifndef ROCKSDB_LITE
     sst_file_manager_ = sst_file_manager;
     mutex_ = mutex;
     error_handler_ = error_handler;
+#else
+    (void)sst_file_manager;
+    (void)mutex;
+    (void)error_handler;
+#endif  // ROCKSDB_LITE
   }
 
   void OnBlobFileCreationStarted(const std::string& file_name,
                                  const std::string& column_family_name,
                                  int job_id,
                                  BlobFileCreationReason creation_reason) {
+#ifndef ROCKSDB_LITE
     // Notify the listeners.
     EventHelpers::NotifyBlobFileCreationStarted(listeners_, dbname_,
                                                 column_family_name, file_name,
                                                 job_id, creation_reason);
+#else
+    (void)file_name;
+    (void)column_family_name;
+    (void)job_id;
+    (void)creation_reason;
+#endif
   }
 
   Status OnBlobFileCompleted(const std::string& file_name,
@@ -48,6 +61,7 @@ class BlobFileCompletionCallback {
                              uint64_t blob_count, uint64_t blob_bytes) {
     Status s;
 
+#ifndef ROCKSDB_LITE
     auto sfm = static_cast<SstFileManagerImpl*>(sst_file_manager_);
     if (sfm) {
       // Report new blob files to SstFileManagerImpl
@@ -60,6 +74,7 @@ class BlobFileCompletionCallback {
         error_handler_->SetBGError(s, BackgroundErrorReason::kFlush);
       }
     }
+#endif  // !ROCKSDB_LITE
 
     // Notify the listeners.
     EventHelpers::LogAndNotifyBlobFileCreationFinished(
@@ -74,9 +89,11 @@ class BlobFileCompletionCallback {
   }
 
  private:
+#ifndef ROCKSDB_LITE
   SstFileManager* sst_file_manager_;
   InstrumentedMutex* mutex_;
   ErrorHandler* error_handler_;
+#endif  // ROCKSDB_LITE
   EventLogger* event_logger_;
   std::vector<std::shared_ptr<EventListener>> listeners_;
   std::string dbname_;

db/blob/blob_file_garbage_test.cc

@@ -168,7 +168,6 @@ TEST_F(BlobFileGarbageTest, ForwardIncompatibleCustomField) {
 }  // namespace ROCKSDB_NAMESPACE
 
 int main(int argc, char** argv) {
-  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
   ::testing::InitGoogleTest(&argc, argv);
   return RUN_ALL_TESTS();
 }

db/blob/blob_file_reader.cc

@@ -26,10 +26,9 @@
 namespace ROCKSDB_NAMESPACE {
 
 Status BlobFileReader::Create(
-    const ImmutableOptions& immutable_options, const ReadOptions& read_options,
-    const FileOptions& file_options, uint32_t column_family_id,
-    HistogramImpl* blob_file_read_hist, uint64_t blob_file_number,
-    const std::shared_ptr<IOTracer>& io_tracer,
+    const ImmutableOptions& immutable_options, const FileOptions& file_options,
+    uint32_t column_family_id, HistogramImpl* blob_file_read_hist,
+    uint64_t blob_file_number, const std::shared_ptr<IOTracer>& io_tracer,
     std::unique_ptr<BlobFileReader>* blob_file_reader) {
   assert(blob_file_reader);
   assert(!*blob_file_reader);
@@ -53,17 +52,15 @@ Status BlobFileReader::Create(
   CompressionType compression_type = kNoCompression;
 
   {
-    const Status s =
-        ReadHeader(file_reader.get(), read_options, column_family_id,
-                   statistics, &compression_type);
+    const Status s = ReadHeader(file_reader.get(), column_family_id, statistics,
+                                &compression_type);
     if (!s.ok()) {
       return s;
     }
   }
 
   {
-    const Status s =
-        ReadFooter(file_reader.get(), read_options, file_size, statistics);
+    const Status s = ReadFooter(file_reader.get(), file_size, statistics);
     if (!s.ok()) {
       return s;
     }
@@ -137,7 +134,6 @@ Status BlobFileReader::OpenFile(
 }
 
 Status BlobFileReader::ReadHeader(const RandomAccessFileReader* file_reader,
-                                  const ReadOptions& read_options,
                                   uint32_t column_family_id,
                                   Statistics* statistics,
                                   CompressionType* compression_type) {
@@ -155,10 +151,9 @@ Status BlobFileReader::ReadHeader(const RandomAccessFileReader* file_reader,
     constexpr size_t read_size = BlobLogHeader::kSize;
 
     // TODO: rate limit reading headers from blob files.
-    const Status s =
-        ReadFromFile(file_reader, read_options, read_offset, read_size,
-                     statistics, &header_slice, &buf, &aligned_buf,
-                     Env::IO_TOTAL /* rate_limiter_priority */);
+    const Status s = ReadFromFile(file_reader, read_offset, read_size,
+                                  statistics, &header_slice, &buf, &aligned_buf,
+                                  Env::IO_TOTAL /* rate_limiter_priority */);
     if (!s.ok()) {
       return s;
     }
@@ -192,7 +187,6 @@ Status BlobFileReader::ReadHeader(const RandomAccessFileReader* file_reader,
 }
 
 Status BlobFileReader::ReadFooter(const RandomAccessFileReader* file_reader,
-                                  const ReadOptions& read_options,
                                   uint64_t file_size, Statistics* statistics) {
   assert(file_size >= BlobLogHeader::kSize + BlobLogFooter::kSize);
   assert(file_reader);
@@ -208,10 +202,9 @@ Status BlobFileReader::ReadFooter(const RandomAccessFileReader* file_reader,
     constexpr size_t read_size = BlobLogFooter::kSize;
 
     // TODO: rate limit reading footers from blob files.
-    const Status s =
-        ReadFromFile(file_reader, read_options, read_offset, read_size,
-                     statistics, &footer_slice, &buf, &aligned_buf,
-                     Env::IO_TOTAL /* rate_limiter_priority */);
+    const Status s = ReadFromFile(file_reader, read_offset, read_size,
+                                  statistics, &footer_slice, &buf, &aligned_buf,
+                                  Env::IO_TOTAL /* rate_limiter_priority */);
     if (!s.ok()) {
       return s;
     }
@@ -239,7 +232,6 @@ Status BlobFileReader::ReadFooter(const RandomAccessFileReader* file_reader,
 }
 
 Status BlobFileReader::ReadFromFile(const RandomAccessFileReader* file_reader,
-                                    const ReadOptions& read_options,
                                     uint64_t read_offset, size_t read_size,
                                     Statistics* statistics, Slice* slice,
                                     Buffer* buf, AlignedBuf* aligned_buf,
@@ -254,23 +246,17 @@ Status BlobFileReader::ReadFromFile(const RandomAccessFileReader* file_reader,
 
   Status s;
 
-  IOOptions io_options;
-  s = file_reader->PrepareIOOptions(read_options, io_options);
-  if (!s.ok()) {
-    return s;
-  }
-
   if (file_reader->use_direct_io()) {
     constexpr char* scratch = nullptr;
 
-    s = file_reader->Read(io_options, read_offset, read_size, slice, scratch,
+    s = file_reader->Read(IOOptions(), read_offset, read_size, slice, scratch,
                           aligned_buf, rate_limiter_priority);
   } else {
     buf->reset(new char[read_size]);
     constexpr AlignedBuf* aligned_scratch = nullptr;
 
-    s = file_reader->Read(io_options, read_offset, read_size, slice, buf->get(),
-                          aligned_scratch, rate_limiter_priority);
+    s = file_reader->Read(IOOptions(), read_offset, read_size, slice,
+                          buf->get(), aligned_scratch, rate_limiter_priority);
   }
 
   if (!s.ok()) {
@@ -338,13 +324,8 @@ Status BlobFileReader::GetBlob(
     Status s;
     constexpr bool for_compaction = true;
 
-    IOOptions io_options;
-    s = file_reader_->PrepareIOOptions(read_options, io_options);
-    if (!s.ok()) {
-      return s;
-    }
     prefetched = prefetch_buffer->TryReadFromCache(
-        io_options, file_reader_.get(), record_offset,
+        IOOptions(), file_reader_.get(), record_offset,
         static_cast<size_t>(record_size), &record_slice, &s,
         read_options.rate_limiter_priority, for_compaction);
     if (!s.ok()) {
@@ -357,10 +338,10 @@ Status BlobFileReader::GetBlob(
     PERF_COUNTER_ADD(blob_read_count, 1);
     PERF_COUNTER_ADD(blob_read_byte, record_size);
     PERF_TIMER_GUARD(blob_read_time);
-    const Status s = ReadFromFile(
-        file_reader_.get(), read_options, record_offset,
-        static_cast<size_t>(record_size), statistics_, &record_slice, &buf,
-        &aligned_buf, read_options.rate_limiter_priority);
+    const Status s = ReadFromFile(file_reader_.get(), record_offset,
+                                  static_cast<size_t>(record_size), statistics_,
+                                  &record_slice, &buf, &aligned_buf,
+                                  read_options.rate_limiter_priority);
     if (!s.ok()) {
       return s;
     }
@@ -588,7 +569,12 @@ Status BlobFileReader::UncompressBlobIfNeeded(
   assert(result);
 
   if (compression_type == kNoCompression) {
-    BlobContentsCreator::Create(result, nullptr, value_slice, allocator);
+    CacheAllocationPtr allocation =
+        AllocateBlock(value_slice.size(), allocator);
+    memcpy(allocation.get(), value_slice.data(), value_slice.size());
+
+    *result = BlobContents::Create(std::move(allocation), value_slice.size());
+
     return Status::OK();
   }
 
@@ -616,7 +602,7 @@ Status BlobFileReader::UncompressBlobIfNeeded(
     return Status::Corruption("Unable to uncompress blob");
   }
 
-  result->reset(new BlobContents(std::move(output), uncompressed_size));
+  *result = BlobContents::Create(std::move(output), uncompressed_size);
 
   return Status::OK();
 }

db/blob/blob_file_reader.h

@@ -29,7 +29,6 @@ class Statistics;
 class BlobFileReader {
  public:
   static Status Create(const ImmutableOptions& immutable_options,
-                       const ReadOptions& read_options,
                        const FileOptions& file_options,
                        uint32_t column_family_id,
                        HistogramImpl* blob_file_read_hist,
@@ -75,18 +74,15 @@ class BlobFileReader {
                          std::unique_ptr<RandomAccessFileReader>* file_reader);
 
   static Status ReadHeader(const RandomAccessFileReader* file_reader,
-                           const ReadOptions& read_options,
                            uint32_t column_family_id, Statistics* statistics,
                            CompressionType* compression_type);
 
   static Status ReadFooter(const RandomAccessFileReader* file_reader,
-                           const ReadOptions& read_options, uint64_t file_size,
-                           Statistics* statistics);
+                           uint64_t file_size, Statistics* statistics);
 
   using Buffer = std::unique_ptr<char[]>;
 
   static Status ReadFromFile(const RandomAccessFileReader* file_reader,
-                             const ReadOptions& read_options,
                              uint64_t read_offset, size_t read_size,
                              Statistics* statistics, Slice* slice, Buffer* buf,
                              AlignedBuf* aligned_buf,

db/blob/blob_file_reader_test.cc

@@ -172,12 +172,12 @@ TEST_F(BlobFileReaderTest, CreateReaderAndGetBlob) {
 
   std::unique_ptr<BlobFileReader> reader;
 
-  ReadOptions read_options;
   ASSERT_OK(BlobFileReader::Create(
-      immutable_options, read_options, FileOptions(), column_family_id,
-      blob_file_read_hist, blob_file_number, nullptr /*IOTracer*/, &reader));
+      immutable_options, FileOptions(), column_family_id, blob_file_read_hist,
+      blob_file_number, nullptr /*IOTracer*/, &reader));
 
   // Make sure the blob can be retrieved with and without checksum verification
+  ReadOptions read_options;
   read_options.verify_checksums = false;
 
   constexpr FilePrefetchBuffer* prefetch_buffer = nullptr;
@@ -479,11 +479,11 @@ TEST_F(BlobFileReaderTest, Malformed) {
   constexpr HistogramImpl* blob_file_read_hist = nullptr;
 
   std::unique_ptr<BlobFileReader> reader;
-  const ReadOptions read_options;
-  ASSERT_TRUE(BlobFileReader::Create(immutable_options, read_options,
-                                     FileOptions(), column_family_id,
-                                     blob_file_read_hist, blob_file_number,
-                                     nullptr /*IOTracer*/, &reader)
+
+  ASSERT_TRUE(BlobFileReader::Create(immutable_options, FileOptions(),
+                                     column_family_id, blob_file_read_hist,
+                                     blob_file_number, nullptr /*IOTracer*/,
+                                     &reader)
                   .IsCorruption());
 }
 
@@ -513,11 +513,11 @@ TEST_F(BlobFileReaderTest, TTL) {
   constexpr HistogramImpl* blob_file_read_hist = nullptr;
 
   std::unique_ptr<BlobFileReader> reader;
-  const ReadOptions read_options;
-  ASSERT_TRUE(BlobFileReader::Create(immutable_options, read_options,
-                                     FileOptions(), column_family_id,
-                                     blob_file_read_hist, blob_file_number,
-                                     nullptr /*IOTracer*/, &reader)
+
+  ASSERT_TRUE(BlobFileReader::Create(immutable_options, FileOptions(),
+                                     column_family_id, blob_file_read_hist,
+                                     blob_file_number, nullptr /*IOTracer*/,
+                                     &reader)
                   .IsCorruption());
 }
 
@@ -552,11 +552,11 @@ TEST_F(BlobFileReaderTest, ExpirationRangeInHeader) {
   constexpr HistogramImpl* blob_file_read_hist = nullptr;
 
   std::unique_ptr<BlobFileReader> reader;
-  const ReadOptions read_options;
-  ASSERT_TRUE(BlobFileReader::Create(immutable_options, read_options,
-                                     FileOptions(), column_family_id,
-                                     blob_file_read_hist, blob_file_number,
-                                     nullptr /*IOTracer*/, &reader)
+
+  ASSERT_TRUE(BlobFileReader::Create(immutable_options, FileOptions(),
+                                     column_family_id, blob_file_read_hist,
+                                     blob_file_number, nullptr /*IOTracer*/,
+                                     &reader)
                   .IsCorruption());
 }
 
@@ -591,11 +591,11 @@ TEST_F(BlobFileReaderTest, ExpirationRangeInFooter) {
   constexpr HistogramImpl* blob_file_read_hist = nullptr;
 
   std::unique_ptr<BlobFileReader> reader;
-  const ReadOptions read_options;
-  ASSERT_TRUE(BlobFileReader::Create(immutable_options, read_options,
-                                     FileOptions(), column_family_id,
-                                     blob_file_read_hist, blob_file_number,
-                                     nullptr /*IOTracer*/, &reader)
+
+  ASSERT_TRUE(BlobFileReader::Create(immutable_options, FileOptions(),
+                                     column_family_id, blob_file_read_hist,
+                                     blob_file_number, nullptr /*IOTracer*/,
+                                     &reader)
                   .IsCorruption());
 }
 
@@ -629,9 +629,9 @@ TEST_F(BlobFileReaderTest, IncorrectColumnFamily) {
   std::unique_ptr<BlobFileReader> reader;
 
   constexpr uint32_t incorrect_column_family_id = 2;
-  const ReadOptions read_options;
-  ASSERT_TRUE(BlobFileReader::Create(immutable_options, read_options,
-                                     FileOptions(), incorrect_column_family_id,
+
+  ASSERT_TRUE(BlobFileReader::Create(immutable_options, FileOptions(),
+                                     incorrect_column_family_id,
                                      blob_file_read_hist, blob_file_number,
                                      nullptr /*IOTracer*/, &reader)
                   .IsCorruption());
@@ -664,10 +664,10 @@ TEST_F(BlobFileReaderTest, BlobCRCError) {
   constexpr HistogramImpl* blob_file_read_hist = nullptr;
 
   std::unique_ptr<BlobFileReader> reader;
-  const ReadOptions read_options;
+
   ASSERT_OK(BlobFileReader::Create(
-      immutable_options, read_options, FileOptions(), column_family_id,
-      blob_file_read_hist, blob_file_number, nullptr /*IOTracer*/, &reader));
+      immutable_options, FileOptions(), column_family_id, blob_file_read_hist,
+      blob_file_number, nullptr /*IOTracer*/, &reader));
 
   SyncPoint::GetInstance()->SetCallBack(
       "BlobFileReader::VerifyBlob:CheckBlobCRC", [](void* arg) {
@@ -728,12 +728,13 @@ TEST_F(BlobFileReaderTest, Compression) {
   constexpr HistogramImpl* blob_file_read_hist = nullptr;
 
   std::unique_ptr<BlobFileReader> reader;
-  ReadOptions read_options;
+
   ASSERT_OK(BlobFileReader::Create(
-      immutable_options, read_options, FileOptions(), column_family_id,
-      blob_file_read_hist, blob_file_number, nullptr /*IOTracer*/, &reader));
+      immutable_options, FileOptions(), column_family_id, blob_file_read_hist,
+      blob_file_number, nullptr /*IOTracer*/, &reader));
 
   // Make sure the blob can be retrieved with and without checksum verification
+  ReadOptions read_options;
   read_options.verify_checksums = false;
 
   constexpr FilePrefetchBuffer* prefetch_buffer = nullptr;
@@ -802,10 +803,10 @@ TEST_F(BlobFileReaderTest, UncompressionError) {
   constexpr HistogramImpl* blob_file_read_hist = nullptr;
 
   std::unique_ptr<BlobFileReader> reader;
-  const ReadOptions read_options;
+
   ASSERT_OK(BlobFileReader::Create(
-      immutable_options, read_options, FileOptions(), column_family_id,
-      blob_file_read_hist, blob_file_number, nullptr /*IOTracer*/, &reader));
+      immutable_options, FileOptions(), column_family_id, blob_file_read_hist,
+      blob_file_number, nullptr /*IOTracer*/, &reader));
 
   SyncPoint::GetInstance()->SetCallBack(
       "BlobFileReader::UncompressBlobIfNeeded:TamperWithResult", [](void* arg) {
@@ -894,10 +895,10 @@ TEST_P(BlobFileReaderIOErrorTest, IOError) {
   constexpr HistogramImpl* blob_file_read_hist = nullptr;
 
   std::unique_ptr<BlobFileReader> reader;
-  const ReadOptions read_options;
+
   const Status s = BlobFileReader::Create(
-      immutable_options, read_options, FileOptions(), column_family_id,
-      blob_file_read_hist, blob_file_number, nullptr /*IOTracer*/, &reader);
+      immutable_options, FileOptions(), column_family_id, blob_file_read_hist,
+      blob_file_number, nullptr /*IOTracer*/, &reader);
 
   const bool fail_during_create =
       (sync_point_ != "BlobFileReader::GetBlob:ReadFromFile");
@@ -982,10 +983,10 @@ TEST_P(BlobFileReaderDecodingErrorTest, DecodingError) {
   constexpr HistogramImpl* blob_file_read_hist = nullptr;
 
   std::unique_ptr<BlobFileReader> reader;
-  const ReadOptions read_options;
+
   const Status s = BlobFileReader::Create(
-      immutable_options, read_options, FileOptions(), column_family_id,
-      blob_file_read_hist, blob_file_number, nullptr /*IOTracer*/, &reader);
+      immutable_options, FileOptions(), column_family_id, blob_file_read_hist,
+      blob_file_number, nullptr /*IOTracer*/, &reader);
 
   const bool fail_during_create =
       sync_point_ != "BlobFileReader::GetBlob:TamperWithResult";
@@ -1017,7 +1018,6 @@ TEST_P(BlobFileReaderDecodingErrorTest, DecodingError) {
 }  // namespace ROCKSDB_NAMESPACE
 
 int main(int argc, char** argv) {
-  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
   ::testing::InitGoogleTest(&argc, argv);
   return RUN_ALL_TESTS();
 }

db/blob/blob_garbage_meter_test.cc

@@ -191,7 +191,6 @@ TEST(BlobGarbageMeterTest, InlinedTTLBlobIndex) {
 }  // namespace ROCKSDB_NAMESPACE
 
 int main(int argc, char** argv) {
-  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
   ::testing::InitGoogleTest(&argc, argv);
   return RUN_ALL_TESTS();
 }

db/blob/blob_log_format.h

@@ -22,8 +22,6 @@ constexpr uint32_t kVersion1 = 1;
 
 using ExpirationRange = std::pair<uint64_t, uint64_t>;
 
-// clang-format off
-
 // Format of blob log file header (30 bytes):
 //
 //    +--------------+---------+---------+-------+-------------+-------------------+
@@ -37,9 +35,6 @@ using ExpirationRange = std::pair<uint64_t, uint64_t>;
 //
 // Expiration range in the header is a rough range based on
 // blob_db_options.ttl_range_secs.
-
-// clang-format on
-
 struct BlobLogHeader {
   static constexpr size_t kSize = 30;
 
@@ -62,8 +57,6 @@ struct BlobLogHeader {
   Status DecodeFrom(Slice slice);
 };
 
-// clang-format off
-
 // Format of blob log file footer (32 bytes):
 //
 //    +--------------+------------+-------------------+------------+
@@ -76,9 +69,6 @@ struct BlobLogHeader {
 //
 // Unlike the same field in file header, expiration range in the footer is the
 // range of smallest and largest expiration of the data in this file.
-
-// clang-format on
-
 struct BlobLogFooter {
   static constexpr size_t kSize = 32;
 
@@ -91,8 +81,6 @@ struct BlobLogFooter {
   Status DecodeFrom(Slice slice);
 };
 
-// clang-format off
-
 // Blob record format (32 bytes header + key + value):
 //
 //    +------------+--------------+------------+------------+----------+---------+-----------+
@@ -112,9 +100,6 @@ struct BlobLogFooter {
 //
 // We could use variable length encoding (Varint64) to save more space, but it
 // make reader more complicated.
-
-// clang-format on
-
 struct BlobLogRecord {
   // header include fields up to blob CRC
   static constexpr size_t kHeaderSize = 32;

db/blob/blob_source.cc

@@ -30,14 +30,16 @@ BlobSource::BlobSource(const ImmutableOptions* immutable_options,
       blob_file_cache_(blob_file_cache),
       blob_cache_(immutable_options->blob_cache),
       lowest_used_cache_tier_(immutable_options->lowest_used_cache_tier) {
+#ifndef ROCKSDB_LITE
   auto bbto =
       immutable_options->table_factory->GetOptions<BlockBasedTableOptions>();
   if (bbto &&
       bbto->cache_usage_options.options_overrides.at(CacheEntryRole::kBlobCache)
               .charged == CacheEntryRoleOptions::Decision::kEnabled) {
-    blob_cache_ = SharedCacheInterface{std::make_shared<ChargedCache>(
-        immutable_options->blob_cache, bbto->block_cache)};
+    blob_cache_ = std::make_shared<ChargedCache>(immutable_options->blob_cache,
+                                                 bbto->block_cache);
   }
+#endif  // ROCKSDB_LITE
 }
 
 BlobSource::~BlobSource() = default;
@@ -80,8 +82,9 @@ Status BlobSource::PutBlobIntoCache(
   assert(cached_blob);
   assert(cached_blob->IsEmpty());
 
-  TypedHandle* cache_handle = nullptr;
+  Cache::Handle* cache_handle = nullptr;
   const Status s = InsertEntryIntoCache(cache_key, blob->get(),
+                                        (*blob)->ApproximateMemoryUsage(),
                                         &cache_handle, Cache::Priority::BOTTOM);
   if (s.ok()) {
     blob->release();
@@ -103,10 +106,26 @@ Status BlobSource::PutBlobIntoCache(
   return s;
 }
 
-BlobSource::TypedHandle* BlobSource::GetEntryFromCache(const Slice& key) const {
-  return blob_cache_.LookupFull(key, nullptr /* context */,
-                                Cache::Priority::BOTTOM, statistics_,
-                                lowest_used_cache_tier_);
+Cache::Handle* BlobSource::GetEntryFromCache(const Slice& key) const {
+  Cache::Handle* cache_handle = nullptr;
+
+  if (lowest_used_cache_tier_ == CacheTier::kNonVolatileBlockTier) {
+    Cache::CreateCallback create_cb =
+        [allocator = blob_cache_->memory_allocator()](
+            const void* buf, size_t size, void** out_obj,
+            size_t* charge) -> Status {
+      return BlobContents::CreateCallback(AllocateBlock(size, allocator), buf,
+                                          size, out_obj, charge);
+    };
+
+    cache_handle = blob_cache_->Lookup(key, BlobContents::GetCacheItemHelper(),
+                                       create_cb, Cache::Priority::BOTTOM,
+                                       true /* wait_for_cache */, statistics_);
+  } else {
+    cache_handle = blob_cache_->Lookup(key, statistics_);
+  }
+
+  return cache_handle;
 }
 
 void BlobSource::PinCachedBlob(CacheHandleGuard<BlobContents>* cached_blob,
@@ -147,11 +166,24 @@ void BlobSource::PinOwnedBlob(std::unique_ptr<BlobContents>* owned_blob,
 }
 
 Status BlobSource::InsertEntryIntoCache(const Slice& key, BlobContents* value,
-                                        TypedHandle** cache_handle,
+                                        size_t charge,
+                                        Cache::Handle** cache_handle,
                                         Cache::Priority priority) const {
-  return blob_cache_.InsertFull(key, value, value->ApproximateMemoryUsage(),
-                                cache_handle, priority,
-                                lowest_used_cache_tier_);
+  Status s;
+
+  Cache::CacheItemHelper* const cache_item_helper =
+      BlobContents::GetCacheItemHelper();
+  assert(cache_item_helper);
+
+  if (lowest_used_cache_tier_ == CacheTier::kNonVolatileBlockTier) {
+    s = blob_cache_->Insert(key, value, cache_item_helper, charge, cache_handle,
+                            priority);
+  } else {
+    s = blob_cache_->Insert(key, value, charge, cache_item_helper->del_cb,
+                            cache_handle, priority);
+  }
+
+  return s;
 }
 
 Status BlobSource::GetBlob(const ReadOptions& read_options,
@@ -209,8 +241,7 @@ Status BlobSource::GetBlob(const ReadOptions& read_options,
 
   {
     CacheHandleGuard<BlobFileReader> blob_file_reader;
-    s = blob_file_cache_->GetBlobFileReader(read_options, file_number,
-                                            &blob_file_reader);
+    s = blob_file_cache_->GetBlobFileReader(file_number, &blob_file_reader);
     if (!s.ok()) {
       return s;
     }
@@ -221,10 +252,9 @@ Status BlobSource::GetBlob(const ReadOptions& read_options,
       return Status::Corruption("Compression type mismatch when reading blob");
     }
 
-    MemoryAllocator* const allocator =
-        (blob_cache_ && read_options.fill_cache)
-            ? blob_cache_.get()->memory_allocator()
-            : nullptr;
+    MemoryAllocator* const allocator = (blob_cache_ && read_options.fill_cache)
+                                           ? blob_cache_->memory_allocator()
+                                           : nullptr;
 
     uint64_t read_size = 0;
     s = blob_file_reader.GetValue()->GetBlob(
@@ -373,8 +403,8 @@ void BlobSource::MultiGetBlobFromOneFile(const ReadOptions& read_options,
     }
 
     CacheHandleGuard<BlobFileReader> blob_file_reader;
-    Status s = blob_file_cache_->GetBlobFileReader(read_options, file_number,
-                                                   &blob_file_reader);
+    Status s =
+        blob_file_cache_->GetBlobFileReader(file_number, &blob_file_reader);
     if (!s.ok()) {
       for (size_t i = 0; i < _blob_reqs.size(); ++i) {
         BlobReadRequest* const req = _blob_reqs[i].first;
@@ -388,10 +418,9 @@ void BlobSource::MultiGetBlobFromOneFile(const ReadOptions& read_options,
 
     assert(blob_file_reader.GetValue());
 
-    MemoryAllocator* const allocator =
-        (blob_cache_ && read_options.fill_cache)
-            ? blob_cache_.get()->memory_allocator()
-            : nullptr;
+    MemoryAllocator* const allocator = (blob_cache_ && read_options.fill_cache)
+                                           ? blob_cache_->memory_allocator()
+                                           : nullptr;
 
     blob_file_reader.GetValue()->MultiGetBlob(read_options, allocator,
                                               _blob_reqs, &_bytes_read);

db/blob/blob_source.h

@@ -8,9 +8,8 @@
 #include <cinttypes>
 #include <memory>
 
+#include "cache/cache_helpers.h"
 #include "cache/cache_key.h"
-#include "cache/typed_cache.h"
-#include "db/blob/blob_contents.h"
 #include "db/blob/blob_file_cache.h"
 #include "db/blob/blob_read_request.h"
 #include "rocksdb/cache.h"
@@ -24,6 +23,7 @@ struct ImmutableOptions;
 class Status;
 class FilePrefetchBuffer;
 class Slice;
+class BlobContents;
 
 // BlobSource is a class that provides universal access to blobs, regardless of
 // whether they are in the blob cache, secondary cache, or (remote) storage.
@@ -95,9 +95,9 @@ class BlobSource {
                                uint64_t* bytes_read);
 
   inline Status GetBlobFileReader(
-      const ReadOptions& read_options, uint64_t blob_file_number,
+      uint64_t blob_file_number,
       CacheHandleGuard<BlobFileReader>* blob_file_reader) {
-    return blob_file_cache_->GetBlobFileReader(read_options, blob_file_number,
+    return blob_file_cache_->GetBlobFileReader(blob_file_number,
                                                blob_file_reader);
   }
 
@@ -106,14 +106,6 @@ class BlobSource {
   bool TEST_BlobInCache(uint64_t file_number, uint64_t file_size,
                         uint64_t offset, size_t* charge = nullptr) const;
 
-  // For TypedSharedCacheInterface
-  void Create(BlobContents** out, const char* buf, size_t size,
-              MemoryAllocator* alloc);
-
-  using SharedCacheInterface =
-      FullTypedSharedCacheInterface<BlobContents, BlobContentsCreator>;
-  using TypedHandle = SharedCacheInterface::TypedHandle;
-
  private:
   Status GetBlobFromCache(const Slice& cache_key,
                           CacheHandleGuard<BlobContents>* cached_blob) const;
@@ -128,10 +120,10 @@ class BlobSource {
   static void PinOwnedBlob(std::unique_ptr<BlobContents>* owned_blob,
                            PinnableSlice* value);
 
-  TypedHandle* GetEntryFromCache(const Slice& key) const;
+  Cache::Handle* GetEntryFromCache(const Slice& key) const;
 
   Status InsertEntryIntoCache(const Slice& key, BlobContents* value,
-                              TypedHandle** cache_handle,
+                              size_t charge, Cache::Handle** cache_handle,
                               Cache::Priority priority) const;
 
   inline CacheKey GetCacheKey(uint64_t file_number, uint64_t /*file_size*/,
@@ -149,7 +141,7 @@ class BlobSource {
   BlobFileCache* blob_file_cache_;
 
   // A cache to store uncompressed blobs.
-  mutable SharedCacheInterface blob_cache_;
+  std::shared_ptr<Cache> blob_cache_;
 
   // The control option of how the cache tiers will be used. Currently rocksdb
   // support block/blob cache (volatile tier) and secondary cache (this tier

db/blob/blob_source_test.cc

@@ -517,8 +517,7 @@ TEST_F(BlobSourceTest, GetCompressedBlobs) {
                   compression, blob_offsets, blob_sizes);
 
     CacheHandleGuard<BlobFileReader> blob_file_reader;
-    ASSERT_OK(blob_source.GetBlobFileReader(read_options, file_number,
-                                            &blob_file_reader));
+    ASSERT_OK(blob_source.GetBlobFileReader(file_number, &blob_file_reader));
     ASSERT_NE(blob_file_reader.GetValue(), nullptr);
 
     const uint64_t file_size = blob_file_reader.GetValue()->GetFileSize();
@@ -1140,18 +1139,26 @@ TEST_F(BlobSecondaryCacheTest, GetBlobsFromSecondaryCache) {
                          blob_file_cache.get());
 
   CacheHandleGuard<BlobFileReader> file_reader;
-  ReadOptions read_options;
-  ASSERT_OK(
-      blob_source.GetBlobFileReader(read_options, file_number, &file_reader));
+  ASSERT_OK(blob_source.GetBlobFileReader(file_number, &file_reader));
   ASSERT_NE(file_reader.GetValue(), nullptr);
   const uint64_t file_size = file_reader.GetValue()->GetFileSize();
   ASSERT_EQ(file_reader.GetValue()->GetCompressionType(), kNoCompression);
 
+  ReadOptions read_options;
   read_options.verify_checksums = true;
 
   auto blob_cache = options_.blob_cache;
   auto secondary_cache = lru_cache_opts_.secondary_cache;
 
+  Cache::CreateCallback create_cb = [](const void* buf, size_t size,
+                                       void** out_obj,
+                                       size_t* charge) -> Status {
+    CacheAllocationPtr allocation(new char[size]);
+
+    return BlobContents::CreateCallback(std::move(allocation), buf, size,
+                                        out_obj, charge);
+  };
+
   {
     // GetBlob
     std::vector<PinnableSlice> values(keys.size());
@@ -1212,16 +1219,15 @@ TEST_F(BlobSecondaryCacheTest, GetBlobsFromSecondaryCache) {
     {
       CacheKey cache_key = base_cache_key.WithOffset(blob_offsets[0]);
       const Slice key0 = cache_key.AsSlice();
-      auto handle0 = blob_cache->BasicLookup(key0, statistics);
+      auto handle0 = blob_cache->Lookup(key0, statistics);
       ASSERT_EQ(handle0, nullptr);
 
       // key0's item should be in the secondary cache.
-      bool kept_in_sec_cache = false;
-      auto sec_handle0 = secondary_cache->Lookup(
-          key0, BlobSource::SharedCacheInterface::GetFullHelper(),
-          /*context*/ nullptr, true,
-          /*advise_erase=*/true, kept_in_sec_cache);
-      ASSERT_FALSE(kept_in_sec_cache);
+      bool is_in_sec_cache = false;
+      auto sec_handle0 =
+          secondary_cache->Lookup(key0, create_cb, true,
+                                  /*advise_erase=*/true, is_in_sec_cache);
+      ASSERT_FALSE(is_in_sec_cache);
       ASSERT_NE(sec_handle0, nullptr);
       ASSERT_TRUE(sec_handle0->IsReady());
       auto value = static_cast<BlobContents*>(sec_handle0->Value());
@@ -1240,16 +1246,15 @@ TEST_F(BlobSecondaryCacheTest, GetBlobsFromSecondaryCache) {
     {
       CacheKey cache_key = base_cache_key.WithOffset(blob_offsets[1]);
       const Slice key1 = cache_key.AsSlice();
-      auto handle1 = blob_cache->BasicLookup(key1, statistics);
+      auto handle1 = blob_cache->Lookup(key1, statistics);
       ASSERT_NE(handle1, nullptr);
       blob_cache->Release(handle1);
 
-      bool kept_in_sec_cache = false;
-      auto sec_handle1 = secondary_cache->Lookup(
-          key1, BlobSource::SharedCacheInterface::GetFullHelper(),
-          /*context*/ nullptr, true,
-          /*advise_erase=*/true, kept_in_sec_cache);
-      ASSERT_FALSE(kept_in_sec_cache);
+      bool is_in_sec_cache = false;
+      auto sec_handle1 =
+          secondary_cache->Lookup(key1, create_cb, true,
+                                  /*advise_erase=*/true, is_in_sec_cache);
+      ASSERT_FALSE(is_in_sec_cache);
       ASSERT_EQ(sec_handle1, nullptr);
 
       ASSERT_TRUE(blob_source.TEST_BlobInCache(file_number, file_size,
@@ -1271,7 +1276,7 @@ TEST_F(BlobSecondaryCacheTest, GetBlobsFromSecondaryCache) {
       // key0 should be in the primary cache.
       CacheKey cache_key0 = base_cache_key.WithOffset(blob_offsets[0]);
       const Slice key0 = cache_key0.AsSlice();
-      auto handle0 = blob_cache->BasicLookup(key0, statistics);
+      auto handle0 = blob_cache->Lookup(key0, statistics);
       ASSERT_NE(handle0, nullptr);
       auto value = static_cast<BlobContents*>(blob_cache->Value(handle0));
       ASSERT_NE(value, nullptr);
@@ -1281,12 +1286,12 @@ TEST_F(BlobSecondaryCacheTest, GetBlobsFromSecondaryCache) {
       // key1 is not in the primary cache and is in the secondary cache.
       CacheKey cache_key1 = base_cache_key.WithOffset(blob_offsets[1]);
       const Slice key1 = cache_key1.AsSlice();
-      auto handle1 = blob_cache->BasicLookup(key1, statistics);
+      auto handle1 = blob_cache->Lookup(key1, statistics);
       ASSERT_EQ(handle1, nullptr);
 
       // erase key0 from the primary cache.
       blob_cache->Erase(key0);
-      handle0 = blob_cache->BasicLookup(key0, statistics);
+      handle0 = blob_cache->Lookup(key0, statistics);
       ASSERT_EQ(handle0, nullptr);
 
       // key1 promotion should succeed due to the primary cache being empty. we
@@ -1299,20 +1304,12 @@ TEST_F(BlobSecondaryCacheTest, GetBlobsFromSecondaryCache) {
                                                blob_offsets[1]));
 
       // key1's dummy handle is in the primary cache and key1's item is still
-      // in the secondary cache. So, the primary cache's Lookup() without
-      // secondary cache support cannot see it. (NOTE: The dummy handle used
-      // to be a leaky abstraction but not anymore.)
-      handle1 = blob_cache->BasicLookup(key1, statistics);
-      ASSERT_EQ(handle1, nullptr);
-
-      // But after another access, it is promoted to primary cache
-      ASSERT_TRUE(blob_source.TEST_BlobInCache(file_number, file_size,
-                                               blob_offsets[1]));
-
-      // And Lookup() can find it (without secondary cache support)
-      handle1 = blob_cache->BasicLookup(key1, statistics);
+      // in the secondary cache. So, the primary cache's Lookup() can only
+      // get a dummy handle.
+      handle1 = blob_cache->Lookup(key1, statistics);
       ASSERT_NE(handle1, nullptr);
-      ASSERT_NE(blob_cache->Value(handle1), nullptr);
+      // handl1 is a dummy handle.
+      ASSERT_EQ(blob_cache->Value(handle1), nullptr);
       blob_cache->Release(handle1);
     }
   }
@@ -1374,7 +1371,7 @@ class BlobSourceCacheReservationTest : public DBTestBase {
 
   static constexpr std::size_t kSizeDummyEntry = CacheReservationManagerImpl<
       CacheEntryRole::kBlobCache>::GetDummyEntrySize();
-  static constexpr std::size_t kCacheCapacity = 2 * kSizeDummyEntry;
+  static constexpr std::size_t kCacheCapacity = 1 * kSizeDummyEntry;
   static constexpr int kNumShardBits = 0;  // 2^0 shard
 
   static constexpr uint32_t kColumnFamilyId = 1;
@@ -1393,6 +1390,7 @@ class BlobSourceCacheReservationTest : public DBTestBase {
   std::string db_session_id_;
 };
 
+#ifndef ROCKSDB_LITE
 TEST_F(BlobSourceCacheReservationTest, SimpleCacheReservation) {
   options_.cf_paths.emplace_back(
       test::PerThreadDBPath(
@@ -1507,10 +1505,11 @@ TEST_F(BlobSourceCacheReservationTest, SimpleCacheReservation) {
   }
 }
 
-TEST_F(BlobSourceCacheReservationTest, IncreaseCacheReservation) {
+TEST_F(BlobSourceCacheReservationTest, IncreaseCacheReservationOnFullCache) {
   options_.cf_paths.emplace_back(
       test::PerThreadDBPath(
-          env_, "BlobSourceCacheReservationTest_IncreaseCacheReservation"),
+          env_,
+          "BlobSourceCacheReservationTest_IncreaseCacheReservationOnFullCache"),
       0);
 
   GenerateKeysAndBlobs();
@@ -1518,7 +1517,7 @@ TEST_F(BlobSourceCacheReservationTest, IncreaseCacheReservation) {
   DestroyAndReopen(options_);
 
   ImmutableOptions immutable_options(options_);
-  constexpr size_t blob_size = 24 << 10;  // 24KB
+  constexpr size_t blob_size = kSizeDummyEntry / (kNumBlobs / 2);
   for (size_t i = 0; i < kNumBlobs; ++i) {
     blob_file_size_ -= blobs_[i].size();  // old blob size
     blob_strs_[i].resize(blob_size, '@');
@@ -1576,6 +1575,11 @@ TEST_F(BlobSourceCacheReservationTest, IncreaseCacheReservation) {
 
     std::vector<PinnableSlice> values(keys_.size());
 
+    // Since we resized each blob to be kSizeDummyEntry / (num_blobs / 2), we
+    // can't fit all the blobs in the cache at the same time, which means we
+    // should observe cache evictions once we reach the cache's capacity.
+    // Due to the overhead of the cache and the BlobContents objects, as well as
+    // jemalloc bin sizes, this happens after inserting seven blobs.
     uint64_t blob_bytes = 0;
     for (size_t i = 0; i < kNumBlobs; ++i) {
       ASSERT_OK(blob_source.GetBlob(
@@ -1586,26 +1590,26 @@ TEST_F(BlobSourceCacheReservationTest, IncreaseCacheReservation) {
       // Release cache handle
       values[i].Reset();
 
-      size_t charge = 0;
-      ASSERT_TRUE(blob_source.TEST_BlobInCache(kBlobFileNumber, blob_file_size_,
-                                               blob_offsets[i], &charge));
+      if (i < kNumBlobs / 2 - 1) {
+        size_t charge = 0;
+        ASSERT_TRUE(blob_source.TEST_BlobInCache(
+            kBlobFileNumber, blob_file_size_, blob_offsets[i], &charge));
 
-      blob_bytes += charge;
+        blob_bytes += charge;
+      }
 
-      ASSERT_EQ(cache_res_mgr->GetTotalReservedCacheSize(),
-                (blob_bytes <= kSizeDummyEntry) ? kSizeDummyEntry
-                                                : (2 * kSizeDummyEntry));
+      ASSERT_EQ(cache_res_mgr->GetTotalReservedCacheSize(), kSizeDummyEntry);
       ASSERT_EQ(cache_res_mgr->GetTotalMemoryUsed(), blob_bytes);
       ASSERT_EQ(cache_res_mgr->GetTotalMemoryUsed(),
                 options_.blob_cache->GetUsage());
     }
   }
 }
+#endif  // ROCKSDB_LITE
 
 }  // namespace ROCKSDB_NAMESPACE
 
 int main(int argc, char** argv) {
-  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
   ::testing::InitGoogleTest(&argc, argv);
   return RUN_ALL_TESTS();
 }

db/blob/db_blob_basic_test.cc

@@ -11,7 +11,6 @@
 #include "db/blob/blob_index.h"
 #include "db/blob/blob_log_format.h"
 #include "db/db_test_util.h"
-#include "db/db_with_timestamp_test_util.h"
 #include "port/stack_trace.h"
 #include "test_util/sync_point.h"
 #include "utilities/fault_injection_env.h"
@@ -585,6 +584,7 @@ TEST_F(DBBlobBasicTest, MultiGetBlobsFromCache) {
   }
 }
 
+#ifndef ROCKSDB_LITE
 TEST_F(DBBlobBasicTest, MultiGetWithDirectIO) {
   Options options = GetDefaultOptions();
 
@@ -773,12 +773,13 @@ TEST_F(DBBlobBasicTest, MultiGetWithDirectIO) {
     ASSERT_EQ(values[2], second_blob);
   }
 }
+#endif  // !ROCKSDB_LITE
 
 TEST_F(DBBlobBasicTest, MultiGetBlobsFromMultipleFiles) {
   Options options = GetDefaultOptions();
 
   LRUCacheOptions co;
-  co.capacity = 2 << 20;  // 2MB
+  co.capacity = 2 << 20; // 2MB
   co.num_shard_bits = 2;
   co.metadata_charge_policy = kDontChargeCacheMetadata;
   auto backing_cache = NewLRUCache(co);
@@ -1061,6 +1062,7 @@ TEST_F(DBBlobBasicTest, GetBlob_IndexWithInvalidFileNumber) {
                   .IsCorruption());
 }
 
+#ifndef ROCKSDB_LITE
 TEST_F(DBBlobBasicTest, GenerateIOTracing) {
   Options options = GetDefaultOptions();
   options.enable_blob_files = true;
@@ -1115,6 +1117,7 @@ TEST_F(DBBlobBasicTest, GenerateIOTracing) {
     ASSERT_GT(blob_files_op_count, 2);
   }
 }
+#endif  // !ROCKSDB_LITE
 
 TEST_F(DBBlobBasicTest, BestEffortsRecovery_MissingNewestBlobFile) {
   Options options = GetDefaultOptions();
@@ -1216,6 +1219,7 @@ TEST_F(DBBlobBasicTest, MultiGetMergeBlobWithPut) {
   ASSERT_EQ(values[2], "v2_0");
 }
 
+#ifndef ROCKSDB_LITE
 TEST_F(DBBlobBasicTest, Properties) {
   Options options = GetDefaultOptions();
   options.enable_blob_files = true;
@@ -1378,6 +1382,7 @@ TEST_F(DBBlobBasicTest, PropertiesMultiVersion) {
                  BlobLogRecord::CalculateAdjustmentForRecordHeader(key_size) +
                  blob_size + BlobLogFooter::kSize));
 }
+#endif  // !ROCKSDB_LITE
 
 class DBBlobBasicIOErrorTest : public DBBlobBasicTest,
                                public testing::WithParamInterface<std::string> {
@@ -1627,6 +1632,7 @@ TEST_F(DBBlobBasicTest, WarmCacheWithBlobsDuringFlush) {
             options.statistics->getTickerCount(BLOB_DB_CACHE_ADD));
 }
 
+#ifndef ROCKSDB_LITE
 TEST_F(DBBlobBasicTest, DynamicallyWarmCacheDuringFlush) {
   Options options = GetDefaultOptions();
 
@@ -1694,6 +1700,7 @@ TEST_F(DBBlobBasicTest, DynamicallyWarmCacheDuringFlush) {
                               /*end=*/nullptr));
   EXPECT_EQ(0, options.statistics->getTickerCount(BLOB_DB_CACHE_ADD));
 }
+#endif  // !ROCKSDB_LITE
 
 TEST_F(DBBlobBasicTest, WarmCacheWithBlobsSecondary) {
   CompressedSecondaryCacheOptions secondary_cache_opts;
@@ -1772,461 +1779,6 @@ TEST_F(DBBlobBasicTest, WarmCacheWithBlobsSecondary) {
             1);
 }
 
-TEST_F(DBBlobBasicTest, GetEntityBlob) {
-  Options options = GetDefaultOptions();
-  options.enable_blob_files = true;
-  options.min_blob_size = 0;
-
-  Reopen(options);
-
-  constexpr char key[] = "key";
-  constexpr char blob_value[] = "blob_value";
-
-  constexpr char other_key[] = "other_key";
-  constexpr char other_blob_value[] = "other_blob_value";
-
-  ASSERT_OK(Put(key, blob_value));
-  ASSERT_OK(Put(other_key, other_blob_value));
-
-  ASSERT_OK(Flush());
-
-  WideColumns expected_columns{{kDefaultWideColumnName, blob_value}};
-  WideColumns other_expected_columns{
-      {kDefaultWideColumnName, other_blob_value}};
-
-  {
-    PinnableWideColumns result;
-    ASSERT_OK(db_->GetEntity(ReadOptions(), db_->DefaultColumnFamily(), key,
-                             &result));
-    ASSERT_EQ(result.columns(), expected_columns);
-  }
-
-  {
-    PinnableWideColumns result;
-    ASSERT_OK(db_->GetEntity(ReadOptions(), db_->DefaultColumnFamily(),
-                             other_key, &result));
-
-    ASSERT_EQ(result.columns(), other_expected_columns);
-  }
-
-  {
-    constexpr size_t num_keys = 2;
-
-    std::array<Slice, num_keys> keys{{key, other_key}};
-    std::array<PinnableWideColumns, num_keys> results;
-    std::array<Status, num_keys> statuses;
-
-    db_->MultiGetEntity(ReadOptions(), db_->DefaultColumnFamily(), num_keys,
-                        &keys[0], &results[0], &statuses[0]);
-
-    ASSERT_OK(statuses[0]);
-    ASSERT_EQ(results[0].columns(), expected_columns);
-
-    ASSERT_OK(statuses[1]);
-    ASSERT_EQ(results[1].columns(), other_expected_columns);
-  }
-}
-
-class DBBlobWithTimestampTest : public DBBasicTestWithTimestampBase {
- protected:
-  DBBlobWithTimestampTest()
-      : DBBasicTestWithTimestampBase("db_blob_with_timestamp_test") {}
-};
-
-TEST_F(DBBlobWithTimestampTest, GetBlob) {
-  Options options = GetDefaultOptions();
-  options.create_if_missing = true;
-  options.enable_blob_files = true;
-  options.min_blob_size = 0;
-  const size_t kTimestampSize = Timestamp(0, 0).size();
-  TestComparator test_cmp(kTimestampSize);
-  options.comparator = &test_cmp;
-
-  DestroyAndReopen(options);
-  WriteOptions write_opts;
-  const std::string ts = Timestamp(1, 0);
-  constexpr char key[] = "key";
-  constexpr char blob_value[] = "blob_value";
-
-  ASSERT_OK(db_->Put(write_opts, key, ts, blob_value));
-
-  ASSERT_OK(Flush());
-
-  const std::string read_ts = Timestamp(2, 0);
-  Slice read_ts_slice(read_ts);
-  ReadOptions read_opts;
-  read_opts.timestamp = &read_ts_slice;
-  std::string value;
-  ASSERT_OK(db_->Get(read_opts, key, &value));
-  ASSERT_EQ(value, blob_value);
-}
-
-TEST_F(DBBlobWithTimestampTest, MultiGetBlobs) {
-  constexpr size_t min_blob_size = 6;
-
-  Options options = GetDefaultOptions();
-  options.enable_blob_files = true;
-  options.min_blob_size = min_blob_size;
-  options.create_if_missing = true;
-  const size_t kTimestampSize = Timestamp(0, 0).size();
-  TestComparator test_cmp(kTimestampSize);
-  options.comparator = &test_cmp;
-
-  DestroyAndReopen(options);
-
-  // Put then retrieve three key-values. The first value is below the size limit
-  // and is thus stored inline; the other two are stored separately as blobs.
-  constexpr size_t num_keys = 3;
-
-  constexpr char first_key[] = "first_key";
-  constexpr char first_value[] = "short";
-  static_assert(sizeof(first_value) - 1 < min_blob_size,
-                "first_value too long to be inlined");
-
-  DestroyAndReopen(options);
-  WriteOptions write_opts;
-  const std::string ts = Timestamp(1, 0);
-  ASSERT_OK(db_->Put(write_opts, first_key, ts, first_value));
-
-  constexpr char second_key[] = "second_key";
-  constexpr char second_value[] = "long_value";
-  static_assert(sizeof(second_value) - 1 >= min_blob_size,
-                "second_value too short to be stored as blob");
-
-  ASSERT_OK(db_->Put(write_opts, second_key, ts, second_value));
-
-  constexpr char third_key[] = "third_key";
-  constexpr char third_value[] = "other_long_value";
-  static_assert(sizeof(third_value) - 1 >= min_blob_size,
-                "third_value too short to be stored as blob");
-
-  ASSERT_OK(db_->Put(write_opts, third_key, ts, third_value));
-
-  ASSERT_OK(Flush());
-
-  ReadOptions read_options;
-  const std::string read_ts = Timestamp(2, 0);
-  Slice read_ts_slice(read_ts);
-  read_options.timestamp = &read_ts_slice;
-  std::array<Slice, num_keys> keys{{first_key, second_key, third_key}};
-
-  {
-    std::array<PinnableSlice, num_keys> values;
-    std::array<Status, num_keys> statuses;
-
-    db_->MultiGet(read_options, db_->DefaultColumnFamily(), num_keys, &keys[0],
-                  &values[0], &statuses[0]);
-
-    ASSERT_OK(statuses[0]);
-    ASSERT_EQ(values[0], first_value);
-
-    ASSERT_OK(statuses[1]);
-    ASSERT_EQ(values[1], second_value);
-
-    ASSERT_OK(statuses[2]);
-    ASSERT_EQ(values[2], third_value);
-  }
-}
-
-TEST_F(DBBlobWithTimestampTest, GetMergeBlobWithPut) {
-  Options options = GetDefaultOptions();
-  options.merge_operator = MergeOperators::CreateStringAppendOperator();
-  options.enable_blob_files = true;
-  options.min_blob_size = 0;
-  options.create_if_missing = true;
-  const size_t kTimestampSize = Timestamp(0, 0).size();
-  TestComparator test_cmp(kTimestampSize);
-  options.comparator = &test_cmp;
-
-  DestroyAndReopen(options);
-
-  WriteOptions write_opts;
-  const std::string ts = Timestamp(1, 0);
-  ASSERT_OK(db_->Put(write_opts, "Key1", ts, "v1"));
-  ASSERT_OK(Flush());
-  ASSERT_OK(
-      db_->Merge(write_opts, db_->DefaultColumnFamily(), "Key1", ts, "v2"));
-  ASSERT_OK(Flush());
-  ASSERT_OK(
-      db_->Merge(write_opts, db_->DefaultColumnFamily(), "Key1", ts, "v3"));
-  ASSERT_OK(Flush());
-
-  std::string value;
-  const std::string read_ts = Timestamp(2, 0);
-  Slice read_ts_slice(read_ts);
-  ReadOptions read_opts;
-  read_opts.timestamp = &read_ts_slice;
-  ASSERT_OK(db_->Get(read_opts, "Key1", &value));
-  ASSERT_EQ(value, "v1,v2,v3");
-}
-
-TEST_F(DBBlobWithTimestampTest, MultiGetMergeBlobWithPut) {
-  constexpr size_t num_keys = 3;
-
-  Options options = GetDefaultOptions();
-  options.merge_operator = MergeOperators::CreateStringAppendOperator();
-  options.enable_blob_files = true;
-  options.min_blob_size = 0;
-  options.create_if_missing = true;
-  const size_t kTimestampSize = Timestamp(0, 0).size();
-  TestComparator test_cmp(kTimestampSize);
-  options.comparator = &test_cmp;
-
-  DestroyAndReopen(options);
-
-  WriteOptions write_opts;
-  const std::string ts = Timestamp(1, 0);
-
-  ASSERT_OK(db_->Put(write_opts, "Key0", ts, "v0_0"));
-  ASSERT_OK(db_->Put(write_opts, "Key1", ts, "v1_0"));
-  ASSERT_OK(db_->Put(write_opts, "Key2", ts, "v2_0"));
-  ASSERT_OK(Flush());
-  ASSERT_OK(
-      db_->Merge(write_opts, db_->DefaultColumnFamily(), "Key0", ts, "v0_1"));
-  ASSERT_OK(
-      db_->Merge(write_opts, db_->DefaultColumnFamily(), "Key1", ts, "v1_1"));
-  ASSERT_OK(Flush());
-  ASSERT_OK(
-      db_->Merge(write_opts, db_->DefaultColumnFamily(), "Key0", ts, "v0_2"));
-  ASSERT_OK(Flush());
-
-  const std::string read_ts = Timestamp(2, 0);
-  Slice read_ts_slice(read_ts);
-  ReadOptions read_opts;
-  read_opts.timestamp = &read_ts_slice;
-  std::array<Slice, num_keys> keys{{"Key0", "Key1", "Key2"}};
-  std::array<PinnableSlice, num_keys> values;
-  std::array<Status, num_keys> statuses;
-
-  db_->MultiGet(read_opts, db_->DefaultColumnFamily(), num_keys, &keys[0],
-                &values[0], &statuses[0]);
-
-  ASSERT_OK(statuses[0]);
-  ASSERT_EQ(values[0], "v0_0,v0_1,v0_2");
-
-  ASSERT_OK(statuses[1]);
-  ASSERT_EQ(values[1], "v1_0,v1_1");
-
-  ASSERT_OK(statuses[2]);
-  ASSERT_EQ(values[2], "v2_0");
-}
-
-TEST_F(DBBlobWithTimestampTest, IterateBlobs) {
-  Options options = GetDefaultOptions();
-  options.enable_blob_files = true;
-  options.create_if_missing = true;
-  const size_t kTimestampSize = Timestamp(0, 0).size();
-  TestComparator test_cmp(kTimestampSize);
-  options.comparator = &test_cmp;
-
-  DestroyAndReopen(options);
-
-  int num_blobs = 5;
-  std::vector<std::string> keys;
-  std::vector<std::string> blobs;
-
-  WriteOptions write_opts;
-  std::vector<std::string> write_timestamps = {Timestamp(1, 0),
-                                               Timestamp(2, 0)};
-
-  // For each key in ["key0", ... "keyi", ...], write two versions:
-  // Timestamp(1, 0), "blobi0"
-  // Timestamp(2, 0), "blobi1"
-  for (int i = 0; i < num_blobs; i++) {
-    keys.push_back("key" + std::to_string(i));
-    blobs.push_back("blob" + std::to_string(i));
-    for (size_t j = 0; j < write_timestamps.size(); j++) {
-      ASSERT_OK(db_->Put(write_opts, keys[i], write_timestamps[j],
-                         blobs[i] + std::to_string(j)));
-    }
-  }
-  ASSERT_OK(Flush());
-
-  ReadOptions read_options;
-  std::vector<std::string> read_timestamps = {Timestamp(0, 0), Timestamp(3, 0)};
-  Slice ts_upper_bound(read_timestamps[1]);
-  read_options.timestamp = &ts_upper_bound;
-
-  auto check_iter_entry =
-      [](const Iterator* iter, const std::string& expected_key,
-         const std::string& expected_ts, const std::string& expected_value,
-         bool key_is_internal = true) {
-        ASSERT_OK(iter->status());
-        if (key_is_internal) {
-          std::string expected_ukey_and_ts;
-          expected_ukey_and_ts.assign(expected_key.data(), expected_key.size());
-          expected_ukey_and_ts.append(expected_ts.data(), expected_ts.size());
-
-          ParsedInternalKey parsed_ikey;
-          ASSERT_OK(ParseInternalKey(iter->key(), &parsed_ikey,
-                                     true /* log_err_key */));
-          ASSERT_EQ(parsed_ikey.user_key, expected_ukey_and_ts);
-        } else {
-          ASSERT_EQ(iter->key(), expected_key);
-        }
-        ASSERT_EQ(iter->timestamp(), expected_ts);
-        ASSERT_EQ(iter->value(), expected_value);
-      };
-
-  // Forward iterating one version of each key, get in this order:
-  // [("key0", Timestamp(2, 0), "blob01"),
-  //  ("key1", Timestamp(2, 0), "blob11")...]
-  {
-    std::unique_ptr<Iterator> iter(db_->NewIterator(read_options));
-    ASSERT_OK(iter->status());
-
-    iter->SeekToFirst();
-    for (int i = 0; i < num_blobs; i++) {
-      check_iter_entry(iter.get(), keys[i], write_timestamps[1],
-                       blobs[i] + std::to_string(1), /*key_is_internal*/ false);
-      iter->Next();
-    }
-  }
-
-  // Forward iteration, then reverse to backward.
-  {
-    std::unique_ptr<Iterator> iter(db_->NewIterator(read_options));
-    ASSERT_OK(iter->status());
-
-    iter->SeekToFirst();
-    for (int i = 0; i < num_blobs * 2 - 1; i++) {
-      if (i < num_blobs) {
-        check_iter_entry(iter.get(), keys[i], write_timestamps[1],
-                         blobs[i] + std::to_string(1),
-                         /*key_is_internal*/ false);
-        if (i != num_blobs - 1) {
-          iter->Next();
-        }
-      } else {
-        if (i != num_blobs) {
-          check_iter_entry(iter.get(), keys[num_blobs * 2 - 1 - i],
-                           write_timestamps[1],
-                           blobs[num_blobs * 2 - 1 - i] + std::to_string(1),
-                           /*key_is_internal*/ false);
-        }
-        iter->Prev();
-      }
-    }
-  }
-
-  // Backward iterating one versions of each key, get in this order:
-  // [("key4", Timestamp(2, 0), "blob41"),
-  //  ("key3", Timestamp(2, 0), "blob31")...]
-  {
-    std::unique_ptr<Iterator> iter(db_->NewIterator(read_options));
-    ASSERT_OK(iter->status());
-
-    iter->SeekToLast();
-    for (int i = 0; i < num_blobs; i++) {
-      check_iter_entry(iter.get(), keys[num_blobs - 1 - i], write_timestamps[1],
-                       blobs[num_blobs - 1 - i] + std::to_string(1),
-                       /*key_is_internal*/ false);
-      iter->Prev();
-    }
-  }
-
-  // Backward iteration, then reverse to forward.
-  {
-    std::unique_ptr<Iterator> iter(db_->NewIterator(read_options));
-    ASSERT_OK(iter->status());
-
-    iter->SeekToLast();
-    for (int i = 0; i < num_blobs * 2 - 1; i++) {
-      if (i < num_blobs) {
-        check_iter_entry(iter.get(), keys[num_blobs - 1 - i],
-                         write_timestamps[1],
-                         blobs[num_blobs - 1 - i] + std::to_string(1),
-                         /*key_is_internal*/ false);
-        if (i != num_blobs - 1) {
-          iter->Prev();
-        }
-      } else {
-        if (i != num_blobs) {
-          check_iter_entry(iter.get(), keys[i - num_blobs], write_timestamps[1],
-                           blobs[i - num_blobs] + std::to_string(1),
-                           /*key_is_internal*/ false);
-        }
-        iter->Next();
-      }
-    }
-  }
-
-  Slice ts_lower_bound(read_timestamps[0]);
-  read_options.iter_start_ts = &ts_lower_bound;
-  // Forward iterating multiple versions of the same key, get in this order:
-  // [("key0", Timestamp(2, 0), "blob01"),
-  //  ("key0", Timestamp(1, 0), "blob00"),
-  //  ("key1", Timestamp(2, 0), "blob11")...]
-  {
-    std::unique_ptr<Iterator> iter(db_->NewIterator(read_options));
-    ASSERT_OK(iter->status());
-
-    iter->SeekToFirst();
-    for (int i = 0; i < num_blobs; i++) {
-      for (size_t j = write_timestamps.size(); j > 0; --j) {
-        check_iter_entry(iter.get(), keys[i], write_timestamps[j - 1],
-                         blobs[i] + std::to_string(j - 1));
-        iter->Next();
-      }
-    }
-  }
-
-  // Backward iterating multiple versions of the same key, get in this order:
-  // [("key4", Timestamp(1, 0), "blob00"),
-  //  ("key4", Timestamp(2, 0), "blob01"),
-  //  ("key3", Timestamp(1, 0), "blob10")...]
-  {
-    std::unique_ptr<Iterator> iter(db_->NewIterator(read_options));
-    ASSERT_OK(iter->status());
-
-    iter->SeekToLast();
-    for (int i = num_blobs; i > 0; i--) {
-      for (size_t j = 0; j < write_timestamps.size(); j++) {
-        check_iter_entry(iter.get(), keys[i - 1], write_timestamps[j],
-                         blobs[i - 1] + std::to_string(j));
-        iter->Prev();
-      }
-    }
-  }
-
-  int upper_bound_idx = num_blobs - 2;
-  int lower_bound_idx = 1;
-  Slice upper_bound_slice(keys[upper_bound_idx]);
-  Slice lower_bound_slice(keys[lower_bound_idx]);
-  read_options.iterate_upper_bound = &upper_bound_slice;
-  read_options.iterate_lower_bound = &lower_bound_slice;
-
-  // Forward iteration with upper and lower bound.
-  {
-    std::unique_ptr<Iterator> iter(db_->NewIterator(read_options));
-    ASSERT_OK(iter->status());
-
-    iter->SeekToFirst();
-    for (int i = lower_bound_idx; i < upper_bound_idx; i++) {
-      for (size_t j = write_timestamps.size(); j > 0; --j) {
-        check_iter_entry(iter.get(), keys[i], write_timestamps[j - 1],
-                         blobs[i] + std::to_string(j - 1));
-        iter->Next();
-      }
-    }
-  }
-
-  // Backward iteration with upper and lower bound.
-  {
-    std::unique_ptr<Iterator> iter(db_->NewIterator(read_options));
-    ASSERT_OK(iter->status());
-
-    iter->SeekToLast();
-    for (int i = upper_bound_idx; i > lower_bound_idx; i--) {
-      for (size_t j = 0; j < write_timestamps.size(); j++) {
-        check_iter_entry(iter.get(), keys[i - 1], write_timestamps[j],
-                         blobs[i - 1] + std::to_string(j));
-        iter->Prev();
-      }
-    }
-  }
-}
-
 }  // namespace ROCKSDB_NAMESPACE
 
 int main(int argc, char** argv) {

db/blob/db_blob_compaction_test.cc

@@ -16,6 +16,7 @@ class DBBlobCompactionTest : public DBTestBase {
   explicit DBBlobCompactionTest()
       : DBTestBase("db_blob_compaction_test", /*env_do_fsync=*/false) {}
 
+#ifndef ROCKSDB_LITE
   const std::vector<InternalStats::CompactionStats>& GetCompactionStats() {
     VersionSet* const versions = dbfull()->GetVersionSet();
     assert(versions);
@@ -29,6 +30,7 @@ class DBBlobCompactionTest : public DBTestBase {
 
     return internal_stats->TEST_GetCompactionStats();
   }
+#endif  // ROCKSDB_LITE
 };
 
 namespace {
@@ -248,6 +250,7 @@ TEST_F(DBBlobCompactionTest, FilterByKeyLength) {
   ASSERT_OK(db_->Get(ReadOptions(), long_key, &value));
   ASSERT_EQ("value", value);
 
+#ifndef ROCKSDB_LITE
   const auto& compaction_stats = GetCompactionStats();
   ASSERT_GE(compaction_stats.size(), 2);
 
@@ -255,6 +258,7 @@ TEST_F(DBBlobCompactionTest, FilterByKeyLength) {
   // this involves neither reading nor writing blobs
   ASSERT_EQ(compaction_stats[1].bytes_read_blob, 0);
   ASSERT_EQ(compaction_stats[1].bytes_written_blob, 0);
+#endif  // ROCKSDB_LITE
 
   Close();
 }
@@ -295,6 +299,7 @@ TEST_F(DBBlobCompactionTest, FilterByValueLength) {
     ASSERT_EQ(long_value, value);
   }
 
+#ifndef ROCKSDB_LITE
   const auto& compaction_stats = GetCompactionStats();
   ASSERT_GE(compaction_stats.size(), 2);
 
@@ -302,10 +307,12 @@ TEST_F(DBBlobCompactionTest, FilterByValueLength) {
   // this involves reading but not writing blobs
   ASSERT_GT(compaction_stats[1].bytes_read_blob, 0);
   ASSERT_EQ(compaction_stats[1].bytes_written_blob, 0);
+#endif  // ROCKSDB_LITE
 
   Close();
 }
 
+#ifndef ROCKSDB_LITE
 TEST_F(DBBlobCompactionTest, BlobCompactWithStartingLevel) {
   Options options = GetDefaultOptions();
 
@@ -381,6 +388,7 @@ TEST_F(DBBlobCompactionTest, BlobCompactWithStartingLevel) {
 
   Close();
 }
+#endif
 
 TEST_F(DBBlobCompactionTest, BlindWriteFilter) {
   Options options = GetDefaultOptions();
@@ -405,6 +413,7 @@ TEST_F(DBBlobCompactionTest, BlindWriteFilter) {
     ASSERT_EQ(new_blob_value, Get(key));
   }
 
+#ifndef ROCKSDB_LITE
   const auto& compaction_stats = GetCompactionStats();
   ASSERT_GE(compaction_stats.size(), 2);
 
@@ -412,6 +421,7 @@ TEST_F(DBBlobCompactionTest, BlindWriteFilter) {
   // this involves writing but not reading blobs
   ASSERT_EQ(compaction_stats[1].bytes_read_blob, 0);
   ASSERT_GT(compaction_stats[1].bytes_written_blob, 0);
+#endif  // ROCKSDB_LITE
 
   Close();
 }
@@ -530,6 +540,7 @@ TEST_F(DBBlobCompactionTest, CompactionFilter) {
     ASSERT_EQ(kv.second + std::string(padding), Get(kv.first));
   }
 
+#ifndef ROCKSDB_LITE
   const auto& compaction_stats = GetCompactionStats();
   ASSERT_GE(compaction_stats.size(), 2);
 
@@ -537,6 +548,7 @@ TEST_F(DBBlobCompactionTest, CompactionFilter) {
   // this involves reading and writing blobs
   ASSERT_GT(compaction_stats[1].bytes_read_blob, 0);
   ASSERT_GT(compaction_stats[1].bytes_written_blob, 0);
+#endif  // ROCKSDB_LITE
 
   Close();
 }
@@ -594,6 +606,7 @@ TEST_F(DBBlobCompactionTest, CompactionFilterReadBlobAndKeep) {
                               /*end=*/nullptr));
   ASSERT_EQ(blob_files, GetBlobFileNumbers());
 
+#ifndef ROCKSDB_LITE
   const auto& compaction_stats = GetCompactionStats();
   ASSERT_GE(compaction_stats.size(), 2);
 
@@ -601,6 +614,7 @@ TEST_F(DBBlobCompactionTest, CompactionFilterReadBlobAndKeep) {
   // this involves reading but not writing blobs
   ASSERT_GT(compaction_stats[1].bytes_read_blob, 0);
   ASSERT_EQ(compaction_stats[1].bytes_written_blob, 0);
+#endif  // ROCKSDB_LITE
 
   Close();
 }

db/blob/db_blob_corruption_test.cc

@@ -34,6 +34,7 @@ class DBBlobCorruptionTest : public DBTestBase {
   }
 };
 
+#ifndef ROCKSDB_LITE
 TEST_F(DBBlobCorruptionTest, VerifyWholeBlobFileChecksum) {
   Options options = GetDefaultOptions();
   options.enable_blob_files = true;
@@ -70,6 +71,7 @@ TEST_F(DBBlobCorruptionTest, VerifyWholeBlobFileChecksum) {
   ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
   ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
 }
+#endif  // !ROCKSDB_LITE
 }  // namespace ROCKSDB_NAMESPACE
 
 int main(int argc, char** argv) {

db/blob/db_blob_index_test.cc

@@ -131,7 +131,9 @@ class DBBlobIndexTest : public DBTestBase {
         ASSERT_OK(Flush());
         ASSERT_OK(
             dbfull()->CompactRange(CompactRangeOptions(), nullptr, nullptr));
+#ifndef ROCKSDB_LITE
         ASSERT_EQ("0,1", FilesPerLevel());
+#endif  // !ROCKSDB_LITE
         break;
     }
   }
@@ -457,6 +459,7 @@ TEST_F(DBBlobIndexTest, Iterate) {
     verify(15, Status::kOk, get_value(16, 0), get_value(14, 0),
            create_blob_iterator, check_is_blob(false));
 
+#ifndef ROCKSDB_LITE
     // Iterator with blob support and using seek.
     ASSERT_OK(dbfull()->SetOptions(
         cfh(), {{"max_sequential_skip_in_iterations", "0"}}));
@@ -481,6 +484,7 @@ TEST_F(DBBlobIndexTest, Iterate) {
            create_blob_iterator, check_is_blob(false));
     verify(15, Status::kOk, get_value(16, 0), get_value(14, 0),
            create_blob_iterator, check_is_blob(false));
+#endif  // !ROCKSDB_LITE
 
     for (auto* snapshot : snapshots) {
       dbfull()->ReleaseSnapshot(snapshot);
@@ -580,10 +584,12 @@ TEST_F(DBBlobIndexTest, IntegratedBlobIterate) {
   Status expected_status;
   verify(1, expected_status, expected_value);
 
+#ifndef ROCKSDB_LITE
   // Test DBIter::FindValueForCurrentKeyUsingSeek flow.
   ASSERT_OK(dbfull()->SetOptions(cfh(),
                                  {{"max_sequential_skip_in_iterations", "0"}}));
   verify(1, expected_status, expected_value);
+#endif  // !ROCKSDB_LITE
 }
 
 }  // namespace ROCKSDB_NAMESPACE

db/builder.cc

@@ -56,8 +56,8 @@ TableBuilder* NewTableBuilder(const TableBuilderOptions& tboptions,
 Status BuildTable(
     const std::string& dbname, VersionSet* versions,
     const ImmutableDBOptions& db_options, const TableBuilderOptions& tboptions,
-    const FileOptions& file_options, const ReadOptions& read_options,
-    TableCache* table_cache, InternalIterator* iter,
+    const FileOptions& file_options, TableCache* table_cache,
+    InternalIterator* iter,
     std::vector<std::unique_ptr<FragmentedRangeTombstoneIterator>>
         range_del_iters,
     FileMetaData* meta, std::vector<BlobFileAddition>* blob_file_additions,
@@ -71,9 +71,8 @@ Status BuildTable(
     int job_id, const Env::IOPriority io_priority,
     TableProperties* table_properties, Env::WriteLifeTimeHint write_hint,
     const std::string* full_history_ts_low,
-    BlobFileCompletionCallback* blob_callback, Version* version,
-    uint64_t* num_input_entries, uint64_t* memtable_payload_bytes,
-    uint64_t* memtable_garbage_bytes) {
+    BlobFileCompletionCallback* blob_callback, uint64_t* num_input_entries,
+    uint64_t* memtable_payload_bytes, uint64_t* memtable_garbage_bytes) {
   assert((tboptions.column_family_id ==
           TablePropertiesCollectorFactory::Context::kUnknownColumnFamily) ==
          tboptions.column_family_name.empty());
@@ -91,7 +90,7 @@ Status BuildTable(
   iter->SeekToFirst();
   std::unique_ptr<CompactionRangeDelAggregator> range_del_agg(
       new CompactionRangeDelAggregator(&tboptions.internal_comparator,
-                                       snapshots, full_history_ts_low));
+                                       snapshots));
   uint64_t num_unfragmented_tombstones = 0;
   uint64_t total_tombstone_payload_bytes = 0;
   for (auto& range_del_iter : range_del_iters) {
@@ -107,9 +106,11 @@ Status BuildTable(
   std::vector<std::string> blob_file_paths;
   std::string file_checksum = kUnknownFileChecksum;
   std::string file_checksum_func_name = kUnknownFileChecksumFuncName;
+#ifndef ROCKSDB_LITE
   EventHelpers::NotifyTableFileCreationStarted(ioptions.listeners, dbname,
                                                tboptions.column_family_name,
                                                fname, job_id, tboptions.reason);
+#endif  // !ROCKSDB_LITE
   Env* env = db_options.env;
   assert(env);
   FileSystem* fs = db_options.fs.get();
@@ -174,10 +175,10 @@ Status BuildTable(
       builder = NewTableBuilder(tboptions, file_writer.get());
     }
 
-    auto ucmp = tboptions.internal_comparator.user_comparator();
     MergeHelper merge(
-        env, ucmp, ioptions.merge_operator.get(), compaction_filter.get(),
-        ioptions.logger, true /* internal key corruption is not ok */,
+        env, tboptions.internal_comparator.user_comparator(),
+        ioptions.merge_operator.get(), compaction_filter.get(), ioptions.logger,
+        true /* internal key corruption is not ok */,
         snapshots.empty() ? 0 : snapshots.back(), snapshot_checker);
 
     std::unique_ptr<BlobFileBuilder> blob_file_builder(
@@ -195,8 +196,9 @@ Status BuildTable(
 
     const std::atomic<bool> kManualCompactionCanceledFalse{false};
     CompactionIterator c_iter(
-        iter, ucmp, &merge, kMaxSequenceNumber, &snapshots,
-        earliest_write_conflict_snapshot, job_snapshot, snapshot_checker, env,
+        iter, tboptions.internal_comparator.user_comparator(), &merge,
+        kMaxSequenceNumber, &snapshots, earliest_write_conflict_snapshot,
+        job_snapshot, snapshot_checker, env,
         ShouldReportDetailedTime(env, ioptions.stats),
         true /* internal key corruption is not ok */, range_del_agg.get(),
         blob_file_builder.get(), ioptions.allow_data_in_errors,
@@ -239,28 +241,14 @@ Status BuildTable(
 
     if (s.ok()) {
       auto range_del_it = range_del_agg->NewIterator();
-      Slice last_tombstone_start_user_key{};
       for (range_del_it->SeekToFirst(); range_del_it->Valid();
            range_del_it->Next()) {
         auto tombstone = range_del_it->Tombstone();
         auto kv = tombstone.Serialize();
         builder->Add(kv.first.Encode(), kv.second);
-        InternalKey tombstone_end = tombstone.SerializeEndKey();
-        meta->UpdateBoundariesForRange(kv.first, tombstone_end, tombstone.seq_,
+        meta->UpdateBoundariesForRange(kv.first, tombstone.SerializeEndKey(),
+                                       tombstone.seq_,
                                        tboptions.internal_comparator);
-        if (version) {
-          if (last_tombstone_start_user_key.empty() ||
-              ucmp->CompareWithoutTimestamp(last_tombstone_start_user_key,
-                                            range_del_it->start_key()) < 0) {
-            SizeApproximationOptions approx_opts;
-            approx_opts.files_size_error_margin = 0.1;
-            meta->compensated_range_deletion_size += versions->ApproximateSize(
-                approx_opts, read_options, version, kv.first.Encode(),
-                tombstone_end.Encode(), 0 /* start_level */, -1 /* end_level */,
-                TableReaderCaller::kFlush);
-          }
-          last_tombstone_start_user_key = range_del_it->start_key();
-        }
       }
     }
 
@@ -293,8 +281,7 @@ Status BuildTable(
       meta->fd.file_size = file_size;
       meta->marked_for_compaction = builder->NeedCompact();
       assert(meta->fd.GetFileSize() > 0);
-      tp = builder
-               ->GetTableProperties();  // refresh now that builder is finished
+      tp = builder->GetTableProperties(); // refresh now that builder is finished
       if (memtable_payload_bytes != nullptr &&
           memtable_garbage_bytes != nullptr) {
         const CompactionIterationStats& ci_stats = c_iter.iter_stats();
@@ -369,6 +356,7 @@ Status BuildTable(
       // here because this is a special case after we finish the table building.
       // No matter whether use_direct_io_for_flush_and_compaction is true,
       // the goal is to cache it here for further user reads.
+      ReadOptions read_options;
       std::unique_ptr<InternalIterator> it(table_cache->NewIterator(
           read_options, file_options, tboptions.internal_comparator, *meta,
           nullptr /* range_del_agg */, mutable_cf_options.prefix_extractor,

db/builder.h

@@ -13,7 +13,6 @@
 #include "db/range_tombstone_fragmenter.h"
 #include "db/seqno_to_time_mapping.h"
 #include "db/table_properties_collector.h"
-#include "db/version_set.h"
 #include "logging/event_logger.h"
 #include "options/cf_options.h"
 #include "rocksdb/comparator.h"
@@ -53,8 +52,8 @@ TableBuilder* NewTableBuilder(const TableBuilderOptions& tboptions,
 extern Status BuildTable(
     const std::string& dbname, VersionSet* versions,
     const ImmutableDBOptions& db_options, const TableBuilderOptions& tboptions,
-    const FileOptions& file_options, const ReadOptions& read_options,
-    TableCache* table_cache, InternalIterator* iter,
+    const FileOptions& file_options, TableCache* table_cache,
+    InternalIterator* iter,
     std::vector<std::unique_ptr<FragmentedRangeTombstoneIterator>>
         range_del_iters,
     FileMetaData* meta, std::vector<BlobFileAddition>* blob_file_additions,
@@ -71,7 +70,7 @@ extern Status BuildTable(
     Env::WriteLifeTimeHint write_hint = Env::WLTH_NOT_SET,
     const std::string* full_history_ts_low = nullptr,
     BlobFileCompletionCallback* blob_callback = nullptr,
-    Version* version = nullptr, uint64_t* num_input_entries = nullptr,
+    uint64_t* num_input_entries = nullptr,
     uint64_t* memtable_payload_bytes = nullptr,
     uint64_t* memtable_garbage_bytes = nullptr);
 

db/c_test.c

@@ -5,6 +5,8 @@
 
 #include <stdio.h>
 
+#ifndef ROCKSDB_LITE  // Lite does not support C API
+
 #include <assert.h>
 #include <stddef.h>
 #include <stdlib.h>
@@ -47,32 +49,32 @@ static void StartPhase(const char* name) {
 }
 #ifdef _MSC_VER
 #pragma warning(push)
-#pragma warning(disable : 4996)  // getenv security warning
+#pragma warning (disable: 4996) // getenv security warning
 #endif
 static const char* GetTempDir(void) {
-  const char* ret = getenv("TEST_TMPDIR");
-  if (ret == NULL || ret[0] == '\0')
+    const char* ret = getenv("TEST_TMPDIR");
+    if (ret == NULL || ret[0] == '\0')
 #ifdef OS_WIN
-    ret = getenv("TEMP");
+      ret = getenv("TEMP");
 #else
-    ret = "/tmp";
+      ret = "/tmp";
 #endif
-  return ret;
+    return ret;
 }
 #ifdef _MSC_VER
 #pragma warning(pop)
 #endif
 
-#define CheckNoError(err)                                                 \
-  if ((err) != NULL) {                                                    \
+#define CheckNoError(err)                                               \
+  if ((err) != NULL) {                                                  \
     fprintf(stderr, "%s:%d: %s: %s\n", __FILE__, __LINE__, phase, (err)); \
-    abort();                                                              \
+    abort();                                                            \
   }
 
-#define CheckCondition(cond)                                              \
-  if (!(cond)) {                                                          \
+#define CheckCondition(cond)                                            \
+  if (!(cond)) {                                                        \
     fprintf(stderr, "%s:%d: %s: %s\n", __FILE__, __LINE__, phase, #cond); \
-    abort();                                                              \
+    abort();                                                            \
   }
 
 static void CheckEqual(const char* expected, const char* v, size_t n) {
@@ -96,15 +98,21 @@ static void Free(char** ptr) {
   }
 }
 
-static void CheckValue(char* err, const char* expected, char** actual,
-                       size_t actual_length) {
+static void CheckValue(
+    char* err,
+    const char* expected,
+    char** actual,
+    size_t actual_length) {
   CheckNoError(err);
   CheckEqual(expected, *actual, actual_length);
   Free(actual);
 }
 
-static void CheckGet(rocksdb_t* db, const rocksdb_readoptions_t* options,
-                     const char* key, const char* expected) {
+static void CheckGet(
+    rocksdb_t* db,
+    const rocksdb_readoptions_t* options,
+    const char* key,
+    const char* expected) {
   char* err = NULL;
   size_t val_len;
   char* val;
@@ -114,9 +122,12 @@ static void CheckGet(rocksdb_t* db, const rocksdb_readoptions_t* options,
   Free(&val);
 }
 
-static void CheckGetCF(rocksdb_t* db, const rocksdb_readoptions_t* options,
-                       rocksdb_column_family_handle_t* handle, const char* key,
-                       const char* expected) {
+static void CheckGetCF(
+    rocksdb_t* db,
+    const rocksdb_readoptions_t* options,
+    rocksdb_column_family_handle_t* handle,
+    const char* key,
+    const char* expected) {
   char* err = NULL;
   size_t val_len;
   char* val;
@@ -163,8 +174,8 @@ static void CheckMultiGetValues(size_t num_keys, char** values,
   }
 }
 
-static void CheckIter(rocksdb_iterator_t* iter, const char* key,
-                      const char* val) {
+static void CheckIter(rocksdb_iterator_t* iter,
+                      const char* key, const char* val) {
   size_t len;
   const char* str;
   str = rocksdb_iter_key(iter, &len);
@@ -174,9 +185,10 @@ static void CheckIter(rocksdb_iterator_t* iter, const char* key,
 }
 
 // Callback from rocksdb_writebatch_iterate()
-static void CheckPut(void* ptr, const char* k, size_t klen, const char* v,
-                     size_t vlen) {
-  int* state = (int*)ptr;
+static void CheckPut(void* ptr,
+                     const char* k, size_t klen,
+                     const char* v, size_t vlen) {
+  int* state = (int*) ptr;
   CheckCondition(*state < 2);
   switch (*state) {
     case 0:
@@ -193,7 +205,7 @@ static void CheckPut(void* ptr, const char* k, size_t klen, const char* v,
 
 // Callback from rocksdb_writebatch_iterate()
 static void CheckDel(void* ptr, const char* k, size_t klen) {
-  int* state = (int*)ptr;
+  int* state = (int*) ptr;
   CheckCondition(*state == 2);
   CheckEqual("bar", k, klen);
   (*state)++;
@@ -201,16 +213,14 @@ static void CheckDel(void* ptr, const char* k, size_t klen) {
 
 static void CmpDestroy(void* arg) { (void)arg; }
 
-static int CmpCompare(void* arg, const char* a, size_t alen, const char* b,
-                      size_t blen) {
+static int CmpCompare(void* arg, const char* a, size_t alen,
+                      const char* b, size_t blen) {
   (void)arg;
   size_t n = (alen < blen) ? alen : blen;
   int r = memcmp(a, b, n);
   if (r == 0) {
-    if (alen < blen)
-      r = -1;
-    else if (alen > blen)
-      r = +1;
+    if (alen < blen) r = -1;
+    else if (alen > blen) r = +1;
   }
   return r;
 }
@@ -395,9 +405,11 @@ static const char* MergeOperatorName(void* arg) {
   return "TestMergeOperator";
 }
 static char* MergeOperatorFullMerge(
-    void* arg, const char* key, size_t key_length, const char* existing_value,
-    size_t existing_value_length, const char* const* operands_list,
-    const size_t* operands_list_length, int num_operands,
+    void* arg,
+    const char* key, size_t key_length,
+    const char* existing_value, size_t existing_value_length,
+    const char* const* operands_list, const size_t* operands_list_length,
+    int num_operands,
     unsigned char* success, size_t* new_value_length) {
   (void)arg;
   (void)key;
@@ -413,12 +425,12 @@ static char* MergeOperatorFullMerge(
   memcpy(result, "fake", 4);
   return result;
 }
-static char* MergeOperatorPartialMerge(void* arg, const char* key,
-                                       size_t key_length,
-                                       const char* const* operands_list,
-                                       const size_t* operands_list_length,
-                                       int num_operands, unsigned char* success,
-                                       size_t* new_value_length) {
+static char* MergeOperatorPartialMerge(
+    void* arg,
+    const char* key, size_t key_length,
+    const char* const* operands_list, const size_t* operands_list_length,
+    int num_operands,
+    unsigned char* success, size_t* new_value_length) {
   (void)arg;
   (void)key;
   (void)key_length;
@@ -432,16 +444,18 @@ static char* MergeOperatorPartialMerge(void* arg, const char* key,
   return result;
 }
 
-static void CheckTxnGet(rocksdb_transaction_t* txn,
-                        const rocksdb_readoptions_t* options, const char* key,
-                        const char* expected) {
-  char* err = NULL;
-  size_t val_len;
-  char* val;
-  val = rocksdb_transaction_get(txn, options, key, strlen(key), &val_len, &err);
-  CheckNoError(err);
-  CheckEqual(expected, val, val_len);
-  Free(&val);
+static void CheckTxnGet(
+        rocksdb_transaction_t* txn,
+        const rocksdb_readoptions_t* options,
+        const char* key,
+        const char* expected) {
+        char* err = NULL;
+        size_t val_len;
+        char* val;
+        val = rocksdb_transaction_get(txn, options, key, strlen(key), &val_len, &err);
+        CheckNoError(err);
+        CheckEqual(expected, val, val_len);
+        Free(&val);
 }
 
 static void CheckTxnGetCF(rocksdb_transaction_t* txn,
@@ -488,22 +502,11 @@ static void CheckTxnPinGetCF(rocksdb_transaction_t* txn,
   rocksdb_pinnableslice_destroy(p);
 }
 
-static void CheckTxnGetForUpdate(rocksdb_transaction_t* txn,
-                                 const rocksdb_readoptions_t* options,
-                                 const char* key, const char* expected) {
-  char* err = NULL;
-  size_t val_len;
-  char* val;
-  val = rocksdb_transaction_get_for_update(txn, options, key, strlen(key),
-                                           &val_len, true, &err);
-  CheckNoError(err);
-  CheckEqual(expected, val, val_len);
-  Free(&val);
-}
-
-static void CheckTxnDBGet(rocksdb_transactiondb_t* txn_db,
-                          const rocksdb_readoptions_t* options, const char* key,
-                          const char* expected) {
+static void CheckTxnDBGet(
+        rocksdb_transactiondb_t* txn_db,
+        const rocksdb_readoptions_t* options,
+        const char* key,
+        const char* expected) {
   char* err = NULL;
   size_t val_len;
   char* val;
@@ -528,20 +531,6 @@ static void CheckTxnDBGetCF(rocksdb_transactiondb_t* txn_db,
   Free(&val);
 }
 
-static void CheckTxnGetForUpdateCF(
-    rocksdb_transaction_t* txn, const rocksdb_readoptions_t* options,
-    rocksdb_column_family_handle_t* column_family, const char* key,
-    const char* expected) {
-  char* err = NULL;
-  size_t val_len;
-  char* val;
-  val = rocksdb_transaction_get_for_update_cf(
-      txn, options, column_family, key, strlen(key), &val_len, true, &err);
-  CheckNoError(err);
-  CheckEqual(expected, val, val_len);
-  Free(&val);
-}
-
 static void CheckTxnDBPinGet(rocksdb_transactiondb_t* txn_db,
                              const rocksdb_readoptions_t* options,
                              const char* key, const char* expected) {
@@ -643,7 +632,7 @@ int main(int argc, char** argv) {
   rocksdb_t* db;
   rocksdb_comparator_t* cmp;
   rocksdb_cache_t* cache;
-  rocksdb_dbpath_t* dbpath;
+  rocksdb_dbpath_t *dbpath;
   rocksdb_env_t* env;
   rocksdb_options_t* options;
   rocksdb_compactoptions_t* coptions;
@@ -660,20 +649,30 @@ int main(int argc, char** argv) {
   char* err = NULL;
   int run = -1;
 
-  snprintf(dbname, sizeof(dbname), "%s/rocksdb_c_test-%d", GetTempDir(),
-           ((int)geteuid()));
+  snprintf(dbname, sizeof(dbname),
+           "%s/rocksdb_c_test-%d",
+           GetTempDir(),
+           ((int) geteuid()));
 
-  snprintf(dbbackupname, sizeof(dbbackupname), "%s/rocksdb_c_test-%d-backup",
-           GetTempDir(), ((int)geteuid()));
+  snprintf(dbbackupname, sizeof(dbbackupname),
+           "%s/rocksdb_c_test-%d-backup",
+           GetTempDir(),
+           ((int) geteuid()));
 
   snprintf(dbcheckpointname, sizeof(dbcheckpointname),
-           "%s/rocksdb_c_test-%d-checkpoint", GetTempDir(), ((int)geteuid()));
+           "%s/rocksdb_c_test-%d-checkpoint",
+           GetTempDir(),
+           ((int) geteuid()));
 
-  snprintf(sstfilename, sizeof(sstfilename), "%s/rocksdb_c_test-%d-sst",
-           GetTempDir(), ((int)geteuid()));
+  snprintf(sstfilename, sizeof(sstfilename),
+           "%s/rocksdb_c_test-%d-sst",
+           GetTempDir(),
+           ((int)geteuid()));
 
-  snprintf(dbpathname, sizeof(dbpathname), "%s/rocksdb_c_test-%d-dbpath",
-           GetTempDir(), ((int)geteuid()));
+  snprintf(dbpathname, sizeof(dbpathname),
+           "%s/rocksdb_c_test-%d-dbpath",
+           GetTempDir(),
+           ((int) geteuid()));
 
   StartPhase("create_objects");
   cmp = rocksdb_comparator_create(NULL, CmpDestroy, CmpCompare, CmpName);
@@ -747,8 +746,7 @@ int main(int argc, char** argv) {
     rocksdb_destroy_db(options, dbbackupname, &err);
     CheckNoError(err);
 
-    rocksdb_backup_engine_t* be =
-        rocksdb_backup_engine_open(options, dbbackupname, &err);
+    rocksdb_backup_engine_t *be = rocksdb_backup_engine_open(options, dbbackupname, &err);
     CheckNoError(err);
 
     rocksdb_backup_engine_create_new_backup(be, db, &err);
@@ -761,8 +759,7 @@ int main(int argc, char** argv) {
     rocksdb_backup_engine_create_new_backup(be, db, &err);
     CheckNoError(err);
 
-    const rocksdb_backup_engine_info_t* bei =
-        rocksdb_backup_engine_get_backup_info(be);
+    const rocksdb_backup_engine_info_t* bei = rocksdb_backup_engine_get_backup_info(be);
     CheckCondition(rocksdb_backup_engine_info_count(bei) > 1);
     rocksdb_backup_engine_info_destroy(bei);
 
@@ -781,11 +778,9 @@ int main(int argc, char** argv) {
     rocksdb_destroy_db(options, dbname, &err);
     CheckNoError(err);
 
-    rocksdb_restore_options_t* restore_options =
-        rocksdb_restore_options_create();
+    rocksdb_restore_options_t *restore_options = rocksdb_restore_options_create();
     rocksdb_restore_options_set_keep_log_files(restore_options, 0);
-    rocksdb_backup_engine_restore_db_from_latest_backup(be, dbname, dbname,
-                                                        restore_options, &err);
+    rocksdb_backup_engine_restore_db_from_latest_backup(be, dbname, dbname, restore_options, &err);
     CheckNoError(err);
     rocksdb_restore_options_destroy(restore_options);
 
@@ -804,8 +799,7 @@ int main(int argc, char** argv) {
     rocksdb_destroy_db(options, dbcheckpointname, &err);
     CheckNoError(err);
 
-    rocksdb_checkpoint_t* checkpoint =
-        rocksdb_checkpoint_object_create(db, &err);
+    rocksdb_checkpoint_t* checkpoint = rocksdb_checkpoint_object_create(db, &err);
     CheckNoError(err);
 
     rocksdb_checkpoint_create(checkpoint, dbcheckpointname, 0, &err);
@@ -982,10 +976,10 @@ int main(int argc, char** argv) {
   StartPhase("writebatch_vectors");
   {
     rocksdb_writebatch_t* wb = rocksdb_writebatch_create();
-    const char* k_list[2] = {"z", "ap"};
-    const size_t k_sizes[2] = {1, 2};
-    const char* v_list[3] = {"x", "y", "z"};
-    const size_t v_sizes[3] = {1, 1, 1};
+    const char* k_list[2] = { "z", "ap" };
+    const size_t k_sizes[2] = { 1, 2 };
+    const char* v_list[3] = { "x", "y", "z" };
+    const size_t v_sizes[3] = { 1, 1, 1 };
     rocksdb_writebatch_putv(wb, 2, k_list, k_sizes, 3, v_list, v_sizes);
     rocksdb_write(db, woptions, wb, &err);
     CheckNoError(err);
@@ -1047,17 +1041,13 @@ int main(int argc, char** argv) {
     CheckCondition(count == 3);
     size_t size;
     char* value;
-    value = rocksdb_writebatch_wi_get_from_batch(wbi, options, "box", 3, &size,
-                                                 &err);
+    value = rocksdb_writebatch_wi_get_from_batch(wbi, options, "box", 3, &size, &err);
     CheckValue(err, "c", &value, size);
-    value = rocksdb_writebatch_wi_get_from_batch(wbi, options, "bar", 3, &size,
-                                                 &err);
+    value = rocksdb_writebatch_wi_get_from_batch(wbi, options, "bar", 3, &size, &err);
     CheckValue(err, NULL, &value, size);
-    value = rocksdb_writebatch_wi_get_from_batch_and_db(wbi, db, roptions,
-                                                        "foo", 3, &size, &err);
+    value = rocksdb_writebatch_wi_get_from_batch_and_db(wbi, db, roptions, "foo", 3, &size, &err);
     CheckValue(err, "hello", &value, size);
-    value = rocksdb_writebatch_wi_get_from_batch_and_db(wbi, db, roptions,
-                                                        "box", 3, &size, &err);
+    value = rocksdb_writebatch_wi_get_from_batch_and_db(wbi, db, roptions, "box", 3, &size, &err);
     CheckValue(err, "c", &value, size);
     rocksdb_write_writebatch_wi(db, woptions, wbi, &err);
     CheckNoError(err);
@@ -1074,10 +1064,10 @@ int main(int argc, char** argv) {
   StartPhase("writebatch_wi_vectors");
   {
     rocksdb_writebatch_wi_t* wb = rocksdb_writebatch_wi_create(0, 1);
-    const char* k_list[2] = {"z", "ap"};
-    const size_t k_sizes[2] = {1, 2};
-    const char* v_list[3] = {"x", "y", "z"};
-    const size_t v_sizes[3] = {1, 1, 1};
+    const char* k_list[2] = { "z", "ap" };
+    const size_t k_sizes[2] = { 1, 2 };
+    const char* v_list[3] = { "x", "y", "z" };
+    const size_t v_sizes[3] = { 1, 1, 1 };
     rocksdb_writebatch_wi_putv(wb, 2, k_list, k_sizes, 3, v_list, v_sizes);
     rocksdb_write_writebatch_wi(db, woptions, wb, &err);
     CheckNoError(err);
@@ -1166,14 +1156,13 @@ int main(int argc, char** argv) {
 
   StartPhase("multiget");
   {
-    const char* keys[3] = {"box", "foo", "notfound"};
-    const size_t keys_sizes[3] = {3, 3, 8};
+    const char* keys[3] = { "box", "foo", "notfound" };
+    const size_t keys_sizes[3] = { 3, 3, 8 };
     char* vals[3];
     size_t vals_sizes[3];
     char* errs[3];
     const char* expected[3] = {"c", "hello", NULL};
-    rocksdb_multi_get(db, roptions, 3, keys, keys_sizes, vals, vals_sizes,
-                      errs);
+    rocksdb_multi_get(db, roptions, 3, keys, keys_sizes, vals, vals_sizes, errs);
     CheckMultiGetValues(3, vals, vals_sizes, errs, expected);
   }
 
@@ -1191,10 +1180,10 @@ int main(int argc, char** argv) {
     char keybuf[100];
     char valbuf[100];
     uint64_t sizes[2];
-    const char* start[2] = {"a", "k00000000000000010000"};
-    size_t start_len[2] = {1, 21};
-    const char* limit[2] = {"k00000000000000010000", "z"};
-    size_t limit_len[2] = {21, 1};
+    const char* start[2] = { "a", "k00000000000000010000" };
+    size_t start_len[2] = { 1, 21 };
+    const char* limit[2] = { "k00000000000000010000", "z" };
+    size_t limit_len[2] = { 21, 1 };
     rocksdb_writeoptions_set_sync(woptions, 0);
     for (i = 0; i < n; i++) {
       snprintf(keybuf, sizeof(keybuf), "k%020d", i);
@@ -1404,8 +1393,8 @@ int main(int argc, char** argv) {
                                                   factory);
     db = CheckCompaction(db, options_with_filter_factory, roptions, woptions);
 
-    rocksdb_options_set_compaction_filter_factory(options_with_filter_factory,
-                                                  NULL);
+    rocksdb_options_set_compaction_filter_factory(
+        options_with_filter_factory, NULL);
     rocksdb_options_destroy(options_with_filter_factory);
   }
 
@@ -1460,8 +1449,7 @@ int main(int argc, char** argv) {
     rocksdb_close(db);
 
     size_t cflen;
-    char** column_fams =
-        rocksdb_list_column_families(db_options, dbname, &cflen, &err);
+    char** column_fams = rocksdb_list_column_families(db_options, dbname, &cflen, &err);
     CheckNoError(err);
     CheckEqual("default", column_fams[0], 7);
     CheckEqual("cf1", column_fams[1], 3);
@@ -1477,8 +1465,7 @@ int main(int argc, char** argv) {
     LoadAndCheckLatestOptions(dbname, env, false, cache, NULL, 2, cf_names,
                               NULL);
 
-    db = rocksdb_open_column_families(db_options, dbname, 2, cf_names, cf_opts,
-                                      handles, &err);
+    db = rocksdb_open_column_families(db_options, dbname, 2, cf_names, cf_opts, handles, &err);
     CheckNoError(err);
 
     rocksdb_put_cf(db, woptions, handles[1], "foo", 3, "hello", 5, &err);
@@ -1496,10 +1483,11 @@ int main(int argc, char** argv) {
                                      &err);
     CheckNoError(err);
 
-    rocksdb_flushoptions_t* flush_options = rocksdb_flushoptions_create();
+    rocksdb_flushoptions_t *flush_options = rocksdb_flushoptions_create();
     rocksdb_flushoptions_set_wait(flush_options, 1);
     rocksdb_flush_cf(db, flush_options, handles[1], &err);
-    CheckNoError(err) rocksdb_flushoptions_destroy(flush_options);
+    CheckNoError(err)
+    rocksdb_flushoptions_destroy(flush_options);
 
     CheckGetCF(db, roptions, handles[1], "foo", "hello");
     CheckPinGetCF(db, roptions, handles[1], "foo", "hello");
@@ -1536,29 +1524,27 @@ int main(int argc, char** argv) {
     rocksdb_flush_wal(db, 1, &err);
     CheckNoError(err);
 
-    const char* keys[3] = {"box", "box", "barfooxx"};
-    const rocksdb_column_family_handle_t* get_handles[3] = {
-        handles[0], handles[1], handles[1]};
-    const size_t keys_sizes[3] = {3, 3, 8};
+    const char* keys[3] = { "box", "box", "barfooxx" };
+    const rocksdb_column_family_handle_t* get_handles[3] = { handles[0], handles[1], handles[1] };
+    const size_t keys_sizes[3] = { 3, 3, 8 };
     char* vals[3];
     size_t vals_sizes[3];
     char* errs[3];
-    rocksdb_multi_get_cf(db, roptions, get_handles, 3, keys, keys_sizes, vals,
-                         vals_sizes, errs);
+    rocksdb_multi_get_cf(db, roptions, get_handles, 3, keys, keys_sizes, vals, vals_sizes, errs);
 
     int i;
     for (i = 0; i < 3; i++) {
       CheckEqual(NULL, errs[i], 0);
       switch (i) {
-        case 0:
-          CheckEqual(NULL, vals[i], vals_sizes[i]);  // wrong cf
-          break;
-        case 1:
-          CheckEqual("c", vals[i], vals_sizes[i]);  // bingo
-          break;
-        case 2:
-          CheckEqual(NULL, vals[i], vals_sizes[i]);  // normal not found
-          break;
+      case 0:
+        CheckEqual(NULL, vals[i], vals_sizes[i]); // wrong cf
+        break;
+      case 1:
+        CheckEqual("c", vals[i], vals_sizes[i]); // bingo
+        break;
+      case 2:
+        CheckEqual(NULL, vals[i], vals_sizes[i]); // normal not found
+        break;
       }
       Free(&vals[i]);
     }
@@ -1606,8 +1592,7 @@ int main(int argc, char** argv) {
       }
     }
 
-    rocksdb_iterator_t* iter =
-        rocksdb_create_iterator_cf(db, roptions, handles[1]);
+    rocksdb_iterator_t* iter = rocksdb_create_iterator_cf(db, roptions, handles[1]);
     CheckCondition(!rocksdb_iter_valid(iter));
     rocksdb_iter_seek_to_first(iter);
     CheckCondition(rocksdb_iter_valid(iter));
@@ -1620,11 +1605,9 @@ int main(int argc, char** argv) {
     CheckNoError(err);
     rocksdb_iter_destroy(iter);
 
-    rocksdb_column_family_handle_t* iters_cf_handles[2] = {handles[0],
-                                                           handles[1]};
+    rocksdb_column_family_handle_t* iters_cf_handles[2] = { handles[0], handles[1] };
     rocksdb_iterator_t* iters_handles[2];
-    rocksdb_create_iterators(db, roptions, iters_cf_handles, iters_handles, 2,
-                             &err);
+    rocksdb_create_iterators(db, roptions, iters_cf_handles, iters_handles, 2, &err);
     CheckNoError(err);
 
     iter = iters_handles[0];
@@ -1669,8 +1652,7 @@ int main(int argc, char** argv) {
   {
     // Create new database
     rocksdb_options_set_allow_mmap_reads(options, 1);
-    rocksdb_options_set_prefix_extractor(
-        options, rocksdb_slicetransform_create_fixed_prefix(3));
+    rocksdb_options_set_prefix_extractor(options, rocksdb_slicetransform_create_fixed_prefix(3));
     rocksdb_options_set_hash_skip_list_rep(options, 5000, 4, 4);
     rocksdb_options_set_plain_table_factory(options, 4, 10, 0.75, 16);
     rocksdb_options_set_allow_concurrent_memtable_write(options, 0);
@@ -1765,9 +1747,8 @@ int main(int argc, char** argv) {
     // amount of memory used within memtables should grow
     CheckCondition(rocksdb_approximate_memory_usage_get_mem_table_total(mu2) >=
                    rocksdb_approximate_memory_usage_get_mem_table_total(mu1));
-    CheckCondition(
-        rocksdb_approximate_memory_usage_get_mem_table_unflushed(mu2) >=
-        rocksdb_approximate_memory_usage_get_mem_table_unflushed(mu1));
+    CheckCondition(rocksdb_approximate_memory_usage_get_mem_table_unflushed(mu2) >=
+                   rocksdb_approximate_memory_usage_get_mem_table_unflushed(mu1));
 
     rocksdb_memory_consumers_destroy(consumers);
     rocksdb_approximate_memory_usage_destroy(mu1);
@@ -2597,9 +2578,6 @@ int main(int argc, char** argv) {
     rocksdb_readoptions_set_io_timeout(ro, 400);
     CheckCondition(400 == rocksdb_readoptions_get_io_timeout(ro));
 
-    rocksdb_readoptions_set_async_io(ro, 1);
-    CheckCondition(1 == rocksdb_readoptions_get_async_io(ro));
-
     rocksdb_readoptions_destroy(ro);
   }
 
@@ -2861,57 +2839,53 @@ int main(int argc, char** argv) {
     db = rocksdb_open(options, dbname, &err);
     CheckNoError(err);
 
-    rocksdb_put(db, woptions, "a", 1, "0", 1, &err);
-    CheckNoError(err);
-    rocksdb_put(db, woptions, "foo", 3, "bar", 3, &err);
-    CheckNoError(err);
-    rocksdb_put(db, woptions, "foo1", 4, "bar1", 4, &err);
-    CheckNoError(err);
-    rocksdb_put(db, woptions, "g1", 2, "0", 1, &err);
-    CheckNoError(err);
+    rocksdb_put(db, woptions, "a",    1, "0",    1, &err); CheckNoError(err);
+    rocksdb_put(db, woptions, "foo",  3, "bar",  3, &err); CheckNoError(err);
+    rocksdb_put(db, woptions, "foo1", 4, "bar1", 4, &err); CheckNoError(err);
+    rocksdb_put(db, woptions, "g1",   2, "0",    1, &err); CheckNoError(err);
 
     // testing basic case with no iterate_upper_bound and no prefix_extractor
     {
-      rocksdb_readoptions_set_iterate_upper_bound(roptions, NULL, 0);
-      rocksdb_iterator_t* iter = rocksdb_create_iterator(db, roptions);
+       rocksdb_readoptions_set_iterate_upper_bound(roptions, NULL, 0);
+       rocksdb_iterator_t* iter = rocksdb_create_iterator(db, roptions);
 
-      rocksdb_iter_seek(iter, "foo", 3);
-      CheckCondition(rocksdb_iter_valid(iter));
-      CheckIter(iter, "foo", "bar");
+       rocksdb_iter_seek(iter, "foo", 3);
+       CheckCondition(rocksdb_iter_valid(iter));
+       CheckIter(iter, "foo", "bar");
 
-      rocksdb_iter_next(iter);
-      CheckCondition(rocksdb_iter_valid(iter));
-      CheckIter(iter, "foo1", "bar1");
+       rocksdb_iter_next(iter);
+       CheckCondition(rocksdb_iter_valid(iter));
+       CheckIter(iter, "foo1", "bar1");
 
-      rocksdb_iter_next(iter);
-      CheckCondition(rocksdb_iter_valid(iter));
-      CheckIter(iter, "g1", "0");
+       rocksdb_iter_next(iter);
+       CheckCondition(rocksdb_iter_valid(iter));
+       CheckIter(iter, "g1", "0");
 
-      rocksdb_iter_destroy(iter);
+       rocksdb_iter_destroy(iter);
     }
 
     // testing iterate_upper_bound and forward iterator
     // to make sure it stops at bound
     {
-      // iterate_upper_bound points beyond the last expected entry
-      rocksdb_readoptions_set_iterate_upper_bound(roptions, "foo2", 4);
+       // iterate_upper_bound points beyond the last expected entry
+       rocksdb_readoptions_set_iterate_upper_bound(roptions, "foo2", 4);
 
-      rocksdb_iterator_t* iter = rocksdb_create_iterator(db, roptions);
+       rocksdb_iterator_t* iter = rocksdb_create_iterator(db, roptions);
 
-      rocksdb_iter_seek(iter, "foo", 3);
-      CheckCondition(rocksdb_iter_valid(iter));
-      CheckIter(iter, "foo", "bar");
+       rocksdb_iter_seek(iter, "foo", 3);
+       CheckCondition(rocksdb_iter_valid(iter));
+       CheckIter(iter, "foo", "bar");
 
-      rocksdb_iter_next(iter);
-      CheckCondition(rocksdb_iter_valid(iter));
-      CheckIter(iter, "foo1", "bar1");
+       rocksdb_iter_next(iter);
+       CheckCondition(rocksdb_iter_valid(iter));
+       CheckIter(iter, "foo1", "bar1");
 
-      rocksdb_iter_next(iter);
-      // should stop here...
-      CheckCondition(!rocksdb_iter_valid(iter));
+       rocksdb_iter_next(iter);
+       // should stop here...
+       CheckCondition(!rocksdb_iter_valid(iter));
 
-      rocksdb_iter_destroy(iter);
-      rocksdb_readoptions_set_iterate_upper_bound(roptions, NULL, 0);
+       rocksdb_iter_destroy(iter);
+       rocksdb_readoptions_set_iterate_upper_bound(roptions, NULL, 0);
     }
   }
 
@@ -3035,7 +3009,7 @@ int main(int argc, char** argv) {
     snapshot = rocksdb_transactiondb_create_snapshot(txn_db);
     rocksdb_readoptions_set_snapshot(roptions, snapshot);
 
-    rocksdb_transactiondb_put(txn_db, woptions, "foo", 3, "hey", 3, &err);
+    rocksdb_transactiondb_put(txn_db, woptions, "foo", 3, "hey", 3,  &err);
     CheckNoError(err);
 
     CheckTxnDBGet(txn_db, roptions, "foo", "hello");
@@ -3047,8 +3021,7 @@ int main(int argc, char** argv) {
 
     // iterate
     rocksdb_transaction_put(txn, "bar", 3, "hi", 2, &err);
-    rocksdb_iterator_t* iter =
-        rocksdb_transaction_create_iterator(txn, roptions);
+    rocksdb_iterator_t* iter = rocksdb_transaction_create_iterator(txn, roptions);
     CheckCondition(!rocksdb_iter_valid(iter));
     rocksdb_iter_seek_to_first(iter);
     CheckCondition(rocksdb_iter_valid(iter));
@@ -3231,120 +3204,6 @@ int main(int argc, char** argv) {
     rocksdb_transactiondb_options_destroy(txn_db_options);
   }
 
-  StartPhase("transactions_multi_get_for_update");
-  {
-    // open a TransactionDB
-    txn_db_options = rocksdb_transactiondb_options_create();
-    rocksdb_transactiondb_options_set_transaction_lock_timeout(txn_db_options,
-                                                               0);
-    txn_options = rocksdb_transaction_options_create();
-    rocksdb_options_set_create_if_missing(options, 1);
-    txn_db = rocksdb_transactiondb_open(options, txn_db_options, dbname, &err);
-    CheckNoError(err);
-
-    rocksdb_transactiondb_put(txn_db, woptions, "foo", 3, "hey", 3, &err);
-    CheckNoError(err);
-    rocksdb_transactiondb_put(txn_db, woptions, "bar", 3, "hello", 5, &err);
-    CheckNoError(err);
-
-    // begin transactions
-    txn = rocksdb_transaction_begin(txn_db, woptions, txn_options, NULL);
-    rocksdb_transaction_t* txn2 =
-        rocksdb_transaction_begin(txn_db, woptions, txn_options, NULL);
-
-    // multi get
-    {
-      const char* keys[2] = {"foo", "bar"};
-      const size_t keys_sizes[2] = {3, 3};
-      char* vals[2];
-      size_t vals_sizes[2];
-      char* errs[2];
-      const char* expected[2] = {"hey", "hello"};
-      rocksdb_transaction_multi_get_for_update(
-          txn, roptions, 2, keys, keys_sizes, vals, vals_sizes, errs);
-      CheckMultiGetValues(2, vals, vals_sizes, errs, expected);
-    }
-
-    char* conflict_err = NULL;
-    size_t val_len;
-    rocksdb_transaction_get_for_update(txn2, roptions, "foo", 3, &val_len, true,
-                                       &conflict_err);
-    // get-for-update conflict
-    CheckCondition(conflict_err != NULL);
-    Free(&conflict_err);
-
-    // commit
-    rocksdb_transaction_commit(txn, &err);
-    CheckNoError(err);
-
-    // should work after first tx is commited
-    CheckTxnGetForUpdate(txn2, roptions, "foo", "hey");
-
-    // commit the second one
-    rocksdb_transaction_commit(txn2, &err);
-    CheckNoError(err);
-
-    // destroy txns
-    rocksdb_transaction_destroy(txn);
-    rocksdb_transaction_destroy(txn2);
-
-    // same for column families
-
-    rocksdb_column_family_handle_t* cfh;
-    cfh = rocksdb_transactiondb_create_column_family(txn_db, options,
-                                                     "txn_db_cf", &err);
-    CheckNoError(err);
-
-    rocksdb_transactiondb_put_cf(txn_db, woptions, cfh, "cf_foo", 6, "cf_hello",
-                                 8, &err);
-    CheckNoError(err);
-    rocksdb_transactiondb_put_cf(txn_db, woptions, cfh, "cf_bar", 6, "cf_hey",
-                                 6, &err);
-    CheckNoError(err);
-
-    txn = rocksdb_transaction_begin(txn_db, woptions, txn_options, NULL);
-    txn2 = rocksdb_transaction_begin(txn_db, woptions, txn_options, NULL);
-
-    {
-      const rocksdb_column_family_handle_t* get_handles[2] = {cfh, cfh};
-      const char* keys[2] = {"cf_foo", "cf_bar"};
-      const size_t keys_sizes[2] = {6, 6};
-      char* vals[2];
-      size_t vals_sizes[2];
-      char* errs[2];
-      const char* expected[2] = {"cf_hello", "cf_hey"};
-      rocksdb_transaction_multi_get_for_update_cf(txn, roptions, get_handles, 2,
-                                                  keys, keys_sizes, vals,
-                                                  vals_sizes, errs);
-      CheckMultiGetValues(2, vals, vals_sizes, errs, expected);
-    }
-
-    char* conflict_err_cf = NULL;
-    size_t val_len_cf;
-    rocksdb_transaction_get_for_update_cf(txn2, roptions, cfh, "cf_foo", 6,
-                                          &val_len_cf, true, &conflict_err_cf);
-    CheckCondition(conflict_err_cf != NULL);
-    Free(&conflict_err_cf);
-
-    rocksdb_transaction_commit(txn, &err);
-    CheckNoError(err);
-
-    CheckTxnGetForUpdateCF(txn2, roptions, cfh, "cf_foo", "cf_hello");
-
-    rocksdb_transaction_commit(txn2, &err);
-    CheckNoError(err);
-
-    // close and destroy
-    rocksdb_column_family_handle_destroy(cfh);
-    rocksdb_transaction_destroy(txn);
-    rocksdb_transaction_destroy(txn2);
-    rocksdb_transactiondb_close(txn_db);
-    rocksdb_destroy_db(options, dbname, &err);
-    CheckNoError(err);
-    rocksdb_transaction_options_destroy(txn_options);
-    rocksdb_transactiondb_options_destroy(txn_db_options);
-  }
-
   StartPhase("optimistic_transactions");
   {
     rocksdb_options_t* db_options = rocksdb_options_create();
@@ -3607,3 +3466,12 @@ int main(int argc, char** argv) {
   fprintf(stderr, "PASS\n");
   return 0;
 }
+
+#else
+
+int main(void) {
+  fprintf(stderr, "SKIPPED\n");
+  return 0;
+}
+
+#endif  // !ROCKSDB_LITE

db/column_family.cc

@@ -53,9 +53,11 @@ ColumnFamilyHandleImpl::ColumnFamilyHandleImpl(
 
 ColumnFamilyHandleImpl::~ColumnFamilyHandleImpl() {
   if (cfd_ != nullptr) {
+#ifndef ROCKSDB_LITE
     for (auto& listener : cfd_->ioptions()->listeners) {
       listener->OnColumnFamilyHandleDeletionStarted(this);
     }
+#endif  // ROCKSDB_LITE
     // Job id == 0 means that this is not our background process, but rather
     // user thread
     // Need to hold some shared pointers owned by the initial_cf_options
@@ -86,10 +88,15 @@ const std::string& ColumnFamilyHandleImpl::GetName() const {
 }
 
 Status ColumnFamilyHandleImpl::GetDescriptor(ColumnFamilyDescriptor* desc) {
+#ifndef ROCKSDB_LITE
   // accessing mutable cf-options requires db mutex.
   InstrumentedMutexLock l(mutex_);
   *desc = ColumnFamilyDescriptor(cfd()->GetName(), cfd()->GetLatestCFOptions());
   return Status::OK();
+#else
+  (void)desc;
+  return Status::NotSupported();
+#endif  // !ROCKSDB_LITE
 }
 
 const Comparator* ColumnFamilyHandleImpl::GetComparator() const {
@@ -185,7 +192,8 @@ Status CheckCFPathsSupported(const DBOptions& db_options,
       return Status::NotSupported(
           "More than one CF paths are only supported in "
           "universal and level compaction styles. ");
-    } else if (cf_options.cf_paths.empty() && db_options.db_paths.size() > 1) {
+    } else if (cf_options.cf_paths.empty() &&
+               db_options.db_paths.size() > 1) {
       return Status::NotSupported(
           "More than one DB paths are only supported in "
           "universal and level compaction styles. ");
@@ -197,7 +205,7 @@ Status CheckCFPathsSupported(const DBOptions& db_options,
 namespace {
 const uint64_t kDefaultTtl = 0xfffffffffffffffe;
 const uint64_t kDefaultPeriodicCompSecs = 0xfffffffffffffffe;
-}  // anonymous namespace
+}  // namespace
 
 ColumnFamilyOptions SanitizeOptions(const ImmutableDBOptions& db_options,
                                     const ColumnFamilyOptions& src) {
@@ -283,6 +291,7 @@ ColumnFamilyOptions SanitizeOptions(const ImmutableDBOptions& db_options,
   }
 
   if (result.compaction_style == kCompactionStyleFIFO) {
+    result.num_levels = 1;
     // since we delete level0 files in FIFO compaction when there are too many
     // of them, these options don't really mean anything
     result.level0_slowdown_writes_trigger = std::numeric_limits<int>::max();
@@ -340,17 +349,18 @@ ColumnFamilyOptions SanitizeOptions(const ImmutableDBOptions& db_options,
         result.hard_pending_compaction_bytes_limit;
   }
 
+#ifndef ROCKSDB_LITE
   // When the DB is stopped, it's possible that there are some .trash files that
   // were not deleted yet, when we open the DB we will find these .trash files
   // and schedule them to be deleted (or delete immediately if SstFileManager
   // was not used)
-  auto sfm =
-      static_cast<SstFileManagerImpl*>(db_options.sst_file_manager.get());
+  auto sfm = static_cast<SstFileManagerImpl*>(db_options.sst_file_manager.get());
   for (size_t i = 0; i < result.cf_paths.size(); i++) {
     DeleteScheduler::CleanupDirectory(db_options.env, sfm,
                                       result.cf_paths[i].path)
         .PermitUncheckedError();
   }
+#endif
 
   if (result.cf_paths.empty()) {
     result.cf_paths = db_options.db_paths;
@@ -548,6 +558,7 @@ ColumnFamilyData::ColumnFamilyData(
       next_(nullptr),
       prev_(nullptr),
       log_number_(0),
+      flush_reason_(FlushReason::kOthers),
       column_family_set_(column_family_set),
       queued_for_flush_(false),
       queued_for_compaction_(false),
@@ -555,8 +566,7 @@ ColumnFamilyData::ColumnFamilyData(
       allow_2pc_(db_options.allow_2pc),
       last_memtable_id_(0),
       db_paths_registered_(false),
-      mempurge_used_(false),
-      next_epoch_number_(1) {
+      mempurge_used_(false) {
   if (id_ != kDummyColumnFamilyDataId) {
     // TODO(cc): RegisterDbPaths can be expensive, considering moving it
     // outside of this constructor which might be called with db mutex held.
@@ -593,6 +603,7 @@ ColumnFamilyData::ColumnFamilyData(
     if (ioptions_.compaction_style == kCompactionStyleLevel) {
       compaction_picker_.reset(
           new LevelCompactionPicker(ioptions_, &internal_comparator_));
+#ifndef ROCKSDB_LITE
     } else if (ioptions_.compaction_style == kCompactionStyleUniversal) {
       compaction_picker_.reset(
           new UniversalCompactionPicker(ioptions_, &internal_comparator_));
@@ -600,12 +611,13 @@ ColumnFamilyData::ColumnFamilyData(
       compaction_picker_.reset(
           new FIFOCompactionPicker(ioptions_, &internal_comparator_));
     } else if (ioptions_.compaction_style == kCompactionStyleNone) {
-      compaction_picker_.reset(
-          new NullCompactionPicker(ioptions_, &internal_comparator_));
+      compaction_picker_.reset(new NullCompactionPicker(
+          ioptions_, &internal_comparator_));
       ROCKS_LOG_WARN(ioptions_.logger,
                      "Column family %s does not use any background compaction. "
                      "Compactions can only be done via CompactFiles\n",
                      GetName().c_str());
+#endif  // !ROCKSDB_LITE
     } else {
       ROCKS_LOG_ERROR(ioptions_.logger,
                       "Unable to recognize the specified compaction style %d. "
@@ -867,9 +879,9 @@ int GetL0ThresholdSpeedupCompaction(int level0_file_num_compaction_trigger,
     return static_cast<int>(res);
   }
 }
-}  // anonymous namespace
+}  // namespace
 
-std::pair<WriteStallCondition, WriteStallCause>
+std::pair<WriteStallCondition, ColumnFamilyData::WriteStallCause>
 ColumnFamilyData::GetWriteStallConditionAndCause(
     int num_unflushed_memtables, int num_l0_files,
     uint64_t num_compaction_needed_bytes,
@@ -908,7 +920,7 @@ ColumnFamilyData::GetWriteStallConditionAndCause(
 }
 
 WriteStallCondition ColumnFamilyData::RecalculateWriteStallConditions(
-    const MutableCFOptions& mutable_cf_options) {
+      const MutableCFOptions& mutable_cf_options) {
   auto write_stall_condition = WriteStallCondition::kNormal;
   if (current_ != nullptr) {
     auto* vstorage = current_->storage_info();
@@ -942,8 +954,7 @@ WriteStallCondition ColumnFamilyData::RecalculateWriteStallConditions(
       internal_stats_->AddCFStats(InternalStats::L0_FILE_COUNT_LIMIT_STOPS, 1);
       if (compaction_picker_->IsLevel0CompactionInProgress()) {
         internal_stats_->AddCFStats(
-            InternalStats::L0_FILE_COUNT_LIMIT_STOPS_WITH_ONGOING_COMPACTION,
-            1);
+            InternalStats::LOCKED_L0_FILE_COUNT_LIMIT_STOPS, 1);
       }
       ROCKS_LOG_WARN(ioptions_.logger,
                      "[%s] Stopping writes because we have %d level-0 files",
@@ -964,7 +975,7 @@ WriteStallCondition ColumnFamilyData::RecalculateWriteStallConditions(
           SetupDelay(write_controller, compaction_needed_bytes,
                      prev_compaction_needed_bytes_, was_stopped,
                      mutable_cf_options.disable_auto_compactions);
-      internal_stats_->AddCFStats(InternalStats::MEMTABLE_LIMIT_DELAYS, 1);
+      internal_stats_->AddCFStats(InternalStats::MEMTABLE_LIMIT_SLOWDOWNS, 1);
       ROCKS_LOG_WARN(
           ioptions_.logger,
           "[%s] Stalling writes because we have %d immutable memtables "
@@ -982,11 +993,11 @@ WriteStallCondition ColumnFamilyData::RecalculateWriteStallConditions(
           SetupDelay(write_controller, compaction_needed_bytes,
                      prev_compaction_needed_bytes_, was_stopped || near_stop,
                      mutable_cf_options.disable_auto_compactions);
-      internal_stats_->AddCFStats(InternalStats::L0_FILE_COUNT_LIMIT_DELAYS, 1);
+      internal_stats_->AddCFStats(InternalStats::L0_FILE_COUNT_LIMIT_SLOWDOWNS,
+                                  1);
       if (compaction_picker_->IsLevel0CompactionInProgress()) {
         internal_stats_->AddCFStats(
-            InternalStats::L0_FILE_COUNT_LIMIT_DELAYS_WITH_ONGOING_COMPACTION,
-            1);
+            InternalStats::LOCKED_L0_FILE_COUNT_LIMIT_SLOWDOWNS, 1);
       }
       ROCKS_LOG_WARN(ioptions_.logger,
                      "[%s] Stalling writes because we have %d level-0 files "
@@ -1002,8 +1013,7 @@ WriteStallCondition ColumnFamilyData::RecalculateWriteStallConditions(
           mutable_cf_options.hard_pending_compaction_bytes_limit > 0 &&
           (compaction_needed_bytes -
            mutable_cf_options.soft_pending_compaction_bytes_limit) >
-              3 *
-                  (mutable_cf_options.hard_pending_compaction_bytes_limit -
+              3 * (mutable_cf_options.hard_pending_compaction_bytes_limit -
                    mutable_cf_options.soft_pending_compaction_bytes_limit) /
                   4;
 
@@ -1012,7 +1022,7 @@ WriteStallCondition ColumnFamilyData::RecalculateWriteStallConditions(
                      prev_compaction_needed_bytes_, was_stopped || near_stop,
                      mutable_cf_options.disable_auto_compactions);
       internal_stats_->AddCFStats(
-          InternalStats::PENDING_COMPACTION_BYTES_LIMIT_DELAYS, 1);
+          InternalStats::PENDING_COMPACTION_BYTES_LIMIT_SLOWDOWNS, 1);
       ROCKS_LOG_WARN(
           ioptions_.logger,
           "[%s] Stalling writes because of estimated pending compaction "
@@ -1118,9 +1128,12 @@ bool ColumnFamilyData::NeedsCompaction() const {
 Compaction* ColumnFamilyData::PickCompaction(
     const MutableCFOptions& mutable_options,
     const MutableDBOptions& mutable_db_options, LogBuffer* log_buffer) {
+  SequenceNumber earliest_mem_seqno =
+      std::min(mem_->GetEarliestSequenceNumber(),
+               imm_.current()->GetEarliestSequenceNumber(false));
   auto* result = compaction_picker_->PickCompaction(
       GetName(), mutable_options, mutable_db_options, current_->storage_info(),
-      log_buffer);
+      log_buffer, earliest_mem_seqno);
   if (result != nullptr) {
     result->SetInputVersion(current_);
   }
@@ -1141,7 +1154,6 @@ Status ColumnFamilyData::RangesOverlapWithMemtables(
   *overlap = false;
   // Create an InternalIterator over all unflushed memtables
   Arena arena;
-  // TODO: plumb Env::IOActivity
   ReadOptions read_opts;
   read_opts.total_order_seek = true;
   MergeIteratorBuilder merge_iter_builder(&internal_comparator_, &arena);
@@ -1208,7 +1220,6 @@ Compaction* ColumnFamilyData::CompactRange(
   if (result != nullptr) {
     result->SetInputVersion(current_);
   }
-  TEST_SYNC_POINT("ColumnFamilyData::CompactRange:Return");
   return result;
 }
 
@@ -1292,8 +1303,8 @@ bool ColumnFamilyData::ReturnThreadLocalSuperVersion(SuperVersion* sv) {
   return false;
 }
 
-void ColumnFamilyData::InstallSuperVersion(SuperVersionContext* sv_context,
-                                           InstrumentedMutex* db_mutex) {
+void ColumnFamilyData::InstallSuperVersion(
+    SuperVersionContext* sv_context, InstrumentedMutex* db_mutex) {
   db_mutex->AssertHeld();
   return InstallSuperVersion(sv_context, mutable_cf_options_);
 }
@@ -1431,6 +1442,7 @@ Status ColumnFamilyData::ValidateOptions(
   return s;
 }
 
+#ifndef ROCKSDB_LITE
 Status ColumnFamilyData::SetOptions(
     const DBOptions& db_opts,
     const std::unordered_map<std::string, std::string>& options_map) {
@@ -1449,6 +1461,7 @@ Status ColumnFamilyData::SetOptions(
   }
   return s;
 }
+#endif  // ROCKSDB_LITE
 
 // REQUIRES: DB mutex held
 Env::WriteLifeTimeHint ColumnFamilyData::CalculateSSTWriteHint(int level) {
@@ -1468,8 +1481,8 @@ Env::WriteLifeTimeHint ColumnFamilyData::CalculateSSTWriteHint(int level) {
     // than base_level.
     return Env::WLTH_MEDIUM;
   }
-  return static_cast<Env::WriteLifeTimeHint>(
-      level - base_level + static_cast<int>(Env::WLTH_MEDIUM));
+  return static_cast<Env::WriteLifeTimeHint>(level - base_level +
+                            static_cast<int>(Env::WLTH_MEDIUM));
 }
 
 Status ColumnFamilyData::AddDirectories(
@@ -1507,13 +1520,6 @@ FSDirectory* ColumnFamilyData::GetDataDir(size_t path_id) const {
   return data_dirs_[path_id].get();
 }
 
-void ColumnFamilyData::RecoverEpochNumbers() {
-  assert(current_);
-  auto* vstorage = current_->storage_info();
-  assert(vstorage);
-  vstorage->RecoverEpochNumbers(this);
-}
-
 ColumnFamilySet::ColumnFamilySet(const std::string& dbname,
                                  const ImmutableDBOptions* db_options,
                                  const FileOptions& file_options,
@@ -1572,8 +1578,8 @@ ColumnFamilyData* ColumnFamilySet::GetColumnFamily(uint32_t id) const {
   }
 }
 
-ColumnFamilyData* ColumnFamilySet::GetColumnFamily(
-    const std::string& name) const {
+ColumnFamilyData* ColumnFamilySet::GetColumnFamily(const std::string& name)
+    const {
   auto cfd_iter = column_families_.find(name);
   if (cfd_iter != column_families_.end()) {
     auto cfd = GetColumnFamily(cfd_iter->second);

db/column_family.h

@@ -163,8 +163,8 @@ extern const double kIncSlowdownRatio;
 class ColumnFamilyHandleImpl : public ColumnFamilyHandle {
  public:
   // create while holding the mutex
-  ColumnFamilyHandleImpl(ColumnFamilyData* cfd, DBImpl* db,
-                         InstrumentedMutex* mutex);
+  ColumnFamilyHandleImpl(
+      ColumnFamilyData* cfd, DBImpl* db, InstrumentedMutex* mutex);
   // destroy without mutex
   virtual ~ColumnFamilyHandleImpl();
   virtual ColumnFamilyData* cfd() const { return cfd_; }
@@ -189,8 +189,7 @@ class ColumnFamilyHandleImpl : public ColumnFamilyHandle {
 class ColumnFamilyHandleInternal : public ColumnFamilyHandleImpl {
  public:
   ColumnFamilyHandleInternal()
-      : ColumnFamilyHandleImpl(nullptr, nullptr, nullptr),
-        internal_cfd_(nullptr) {}
+      : ColumnFamilyHandleImpl(nullptr, nullptr, nullptr), internal_cfd_(nullptr) {}
 
   void SetCFD(ColumnFamilyData* _cfd) { internal_cfd_ = _cfd; }
   virtual ColumnFamilyData* cfd() const override { return internal_cfd_; }
@@ -310,6 +309,10 @@ class ColumnFamilyData {
   void SetLogNumber(uint64_t log_number) { log_number_ = log_number; }
   uint64_t GetLogNumber() const { return log_number_; }
 
+  void SetFlushReason(FlushReason flush_reason) {
+    flush_reason_ = flush_reason;
+  }
+  FlushReason GetFlushReason() const { return flush_reason_; }
   // thread-safe
   const FileOptions* soptions() const;
   const ImmutableOptions* ioptions() const { return &ioptions_; }
@@ -335,10 +338,12 @@ class ColumnFamilyData {
   // Validate CF options against DB options
   static Status ValidateOptions(const DBOptions& db_options,
                                 const ColumnFamilyOptions& cf_options);
+#ifndef ROCKSDB_LITE
   // REQUIRES: DB mutex held
   Status SetOptions(
       const DBOptions& db_options,
       const std::unordered_map<std::string, std::string>& options_map);
+#endif  // ROCKSDB_LITE
 
   InternalStats* internal_stats() { return internal_stats_.get(); }
 
@@ -352,7 +357,7 @@ class ColumnFamilyData {
   Version* current() { return current_; }
   Version* dummy_versions() { return dummy_versions_; }
   void SetCurrent(Version* _current);
-  uint64_t GetNumLiveVersions() const;    // REQUIRE: DB mutex held
+  uint64_t GetNumLiveVersions() const;  // REQUIRE: DB mutex held
   uint64_t GetTotalSstFilesSize() const;  // REQUIRE: DB mutex held
   uint64_t GetLiveSstFilesSize() const;   // REQUIRE: DB mutex held
   uint64_t GetTotalBlobFileSize() const;  // REQUIRE: DB mutex held
@@ -462,6 +467,12 @@ class ColumnFamilyData {
   bool queued_for_flush() { return queued_for_flush_; }
   bool queued_for_compaction() { return queued_for_compaction_; }
 
+  enum class WriteStallCause {
+    kNone,
+    kMemtableLimit,
+    kL0FileCountLimit,
+    kPendingCompactionBytes,
+  };
   static std::pair<WriteStallCondition, WriteStallCause>
   GetWriteStallConditionAndCause(
       int num_unflushed_memtables, int num_l0_files,
@@ -521,24 +532,6 @@ class ColumnFamilyData {
   void SetMempurgeUsed() { mempurge_used_ = true; }
   bool GetMempurgeUsed() { return mempurge_used_; }
 
-  // Allocate and return a new epoch number
-  uint64_t NewEpochNumber() { return next_epoch_number_.fetch_add(1); }
-
-  // Get the next epoch number to be assigned
-  uint64_t GetNextEpochNumber() const { return next_epoch_number_.load(); }
-
-  // Set the next epoch number to be assigned
-  void SetNextEpochNumber(uint64_t next_epoch_number) {
-    next_epoch_number_.store(next_epoch_number);
-  }
-
-  // Reset the next epoch number to be assigned
-  void ResetNextEpochNumber() { next_epoch_number_.store(1); }
-
-  // Recover the next epoch number of this CF and epoch number
-  // of its files (if missing)
-  void RecoverEpochNumbers();
-
  private:
   friend class ColumnFamilySet;
   ColumnFamilyData(uint32_t id, const std::string& name,
@@ -559,7 +552,7 @@ class ColumnFamilyData {
   Version* dummy_versions_;  // Head of circular doubly-linked list of versions.
   Version* current_;         // == dummy_versions->prev_
 
-  std::atomic<int> refs_;  // outstanding references to ColumnFamilyData
+  std::atomic<int> refs_;      // outstanding references to ColumnFamilyData
   std::atomic<bool> initialized_;
   std::atomic<bool> dropped_;  // true if client dropped it
 
@@ -604,6 +597,8 @@ class ColumnFamilyData {
   // recovered from
   uint64_t log_number_;
 
+  std::atomic<FlushReason> flush_reason_;
+
   // An object that keeps all the compaction stats
   // and picks the next compaction
   std::unique_ptr<CompactionPicker> compaction_picker_;
@@ -638,8 +633,6 @@ class ColumnFamilyData {
   // a Version associated with this CFD
   std::shared_ptr<CacheReservationManager> file_metadata_cache_res_mgr_;
   bool mempurge_used_;
-
-  std::atomic<uint64_t> next_epoch_number_;
 };
 
 // ColumnFamilySet has interesting thread-safety requirements
@@ -663,7 +656,8 @@ class ColumnFamilySet {
   // ColumnFamilySet supports iteration
   class iterator {
    public:
-    explicit iterator(ColumnFamilyData* cfd) : current_(cfd) {}
+    explicit iterator(ColumnFamilyData* cfd)
+        : current_(cfd) {}
     // NOTE: minimum operators for for-loop iteration
     iterator& operator++() {
       current_ = current_->next_;

db/column_family_test.cc

@@ -39,7 +39,9 @@ class EnvCounter : public SpecialEnv {
  public:
   explicit EnvCounter(Env* base)
       : SpecialEnv(base), num_new_writable_file_(0) {}
-  int GetNumberOfNewWritableFileCalls() { return num_new_writable_file_; }
+  int GetNumberOfNewWritableFileCalls() {
+    return num_new_writable_file_;
+  }
   Status NewWritableFile(const std::string& f, std::unique_ptr<WritableFile>* r,
                          const EnvOptions& soptions) override {
     ++num_new_writable_file_;
@@ -71,7 +73,11 @@ class ColumnFamilyTestBase : public testing::Test {
     for (auto h : handles_) {
       ColumnFamilyDescriptor cfdescriptor;
       Status s = h->GetDescriptor(&cfdescriptor);
+#ifdef ROCKSDB_LITE
+      EXPECT_TRUE(s.IsNotSupported());
+#else
       EXPECT_OK(s);
+#endif  // ROCKSDB_LITE
       column_families.push_back(cfdescriptor);
     }
     ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
@@ -181,7 +187,7 @@ class ColumnFamilyTestBase : public testing::Test {
   }
 
   Status OpenReadOnly(std::vector<std::string> cf,
-                      std::vector<ColumnFamilyOptions> options = {}) {
+                         std::vector<ColumnFamilyOptions> options = {}) {
     std::vector<ColumnFamilyDescriptor> column_families;
     names_.clear();
     for (size_t i = 0; i < cf.size(); ++i) {
@@ -193,17 +199,22 @@ class ColumnFamilyTestBase : public testing::Test {
                                &db_);
   }
 
+#ifndef ROCKSDB_LITE  // ReadOnlyDB is not supported
   void AssertOpenReadOnly(std::vector<std::string> cf,
-                          std::vector<ColumnFamilyOptions> options = {}) {
+                    std::vector<ColumnFamilyOptions> options = {}) {
     ASSERT_OK(OpenReadOnly(cf, options));
   }
+#endif  // !ROCKSDB_LITE
+
 
   void Open(std::vector<std::string> cf,
             std::vector<ColumnFamilyOptions> options = {}) {
     ASSERT_OK(TryOpen(cf, options));
   }
 
-  void Open() { Open({"default"}); }
+  void Open() {
+    Open({"default"});
+  }
 
   DBImpl* dbfull() { return static_cast_with_check<DBImpl>(db_); }
 
@@ -218,20 +229,31 @@ class ColumnFamilyTestBase : public testing::Test {
   }
 
   bool IsDbWriteStopped() {
+#ifndef ROCKSDB_LITE
     uint64_t v;
     EXPECT_TRUE(dbfull()->GetIntProperty("rocksdb.is-write-stopped", &v));
     return (v == 1);
+#else
+    return dbfull()->TEST_write_controler().IsStopped();
+#endif  // !ROCKSDB_LITE
   }
 
   uint64_t GetDbDelayedWriteRate() {
+#ifndef ROCKSDB_LITE
     uint64_t v;
     EXPECT_TRUE(
         dbfull()->GetIntProperty("rocksdb.actual-delayed-write-rate", &v));
     return v;
+#else
+    if (!dbfull()->TEST_write_controler().NeedsDelay()) {
+      return 0;
+    }
+    return dbfull()->TEST_write_controler().delayed_write_rate();
+#endif  // !ROCKSDB_LITE
   }
 
   void Destroy(const std::vector<ColumnFamilyDescriptor>& column_families =
-                   std::vector<ColumnFamilyDescriptor>()) {
+                  std::vector<ColumnFamilyDescriptor>()) {
     Close();
     ASSERT_OK(DestroyDB(dbname_, Options(db_options_, column_family_options_),
                         column_families));
@@ -250,6 +272,7 @@ class ColumnFamilyTestBase : public testing::Test {
           db_->CreateColumnFamily(current_cf_opt, cfs[i], &handles_[cfi]));
       names_[cfi] = cfs[i];
 
+#ifndef ROCKSDB_LITE  // RocksDBLite does not support GetDescriptor
       // Verify the CF options of the returned CF handle.
       ColumnFamilyDescriptor desc;
       ASSERT_OK(handles_[cfi]->GetDescriptor(&desc));
@@ -258,6 +281,7 @@ class ColumnFamilyTestBase : public testing::Test {
       ASSERT_OK(RocksDBOptionsParser::VerifyCFOptions(
           ConfigOptions(), desc.options,
           SanitizeOptions(dbfull()->immutable_db_options(), current_cf_opt)));
+#endif  // !ROCKSDB_LITE
       cfi++;
     }
   }
@@ -306,11 +330,14 @@ class ColumnFamilyTestBase : public testing::Test {
     ASSERT_OK(db_->FlushWAL(/*sync=*/false));
   }
 
+#ifndef ROCKSDB_LITE  // TEST functions in DB are not supported in lite
   void WaitForFlush(int cf) {
     ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable(handles_[cf]));
   }
 
-  void WaitForCompaction() { ASSERT_OK(dbfull()->TEST_WaitForCompact()); }
+  void WaitForCompaction() {
+    ASSERT_OK(dbfull()->TEST_WaitForCompact());
+  }
 
   uint64_t MaxTotalInMemoryState() {
     return dbfull()->TEST_MaxTotalInMemoryState();
@@ -319,6 +346,7 @@ class ColumnFamilyTestBase : public testing::Test {
   void AssertMaxTotalInMemoryState(uint64_t value) {
     ASSERT_EQ(value, MaxTotalInMemoryState());
   }
+#endif  // !ROCKSDB_LITE
 
   Status Put(int cf, const std::string& key, const std::string& value) {
     return db_->Put(WriteOptions(), handles_[cf], Slice(key), Slice(value));
@@ -326,7 +354,9 @@ class ColumnFamilyTestBase : public testing::Test {
   Status Merge(int cf, const std::string& key, const std::string& value) {
     return db_->Merge(WriteOptions(), handles_[cf], Slice(key), Slice(value));
   }
-  Status Flush(int cf) { return db_->Flush(FlushOptions(), handles_[cf]); }
+  Status Flush(int cf) {
+    return db_->Flush(FlushOptions(), handles_[cf]);
+  }
 
   std::string Get(int cf, const std::string& key) {
     ReadOptions options;
@@ -356,6 +386,7 @@ class ColumnFamilyTestBase : public testing::Test {
                        "rocksdb.num-files-at-level" + std::to_string(level));
   }
 
+#ifndef ROCKSDB_LITE
   // Return spread of files per level
   std::string FilesPerLevel(int cf) {
     std::string result;
@@ -372,19 +403,31 @@ class ColumnFamilyTestBase : public testing::Test {
     result.resize(last_non_zero_offset);
     return result;
   }
+#endif
 
   void AssertFilesPerLevel(const std::string& value, int cf) {
+#ifndef ROCKSDB_LITE
     ASSERT_EQ(value, FilesPerLevel(cf));
+#else
+    (void) value;
+    (void) cf;
+#endif
   }
 
+#ifndef ROCKSDB_LITE  // GetLiveFilesMetaData is not supported
   int CountLiveFiles() {
     std::vector<LiveFileMetaData> metadata;
     db_->GetLiveFilesMetaData(&metadata);
     return static_cast<int>(metadata.size());
   }
+#endif  // !ROCKSDB_LITE
 
   void AssertCountLiveFiles(int expected_value) {
+#ifndef ROCKSDB_LITE
     ASSERT_EQ(expected_value, CountLiveFiles());
+#else
+    (void) expected_value;
+#endif
   }
 
   // Do n memtable flushes, each of which produces an sstable
@@ -398,6 +441,7 @@ class ColumnFamilyTestBase : public testing::Test {
     }
   }
 
+#ifndef ROCKSDB_LITE  // GetSortedWalFiles is not supported
   int CountLiveLogFiles() {
     int micros_wait_for_log_deletion = 20000;
     env_->SleepForMicroseconds(micros_wait_for_log_deletion);
@@ -426,18 +470,25 @@ class ColumnFamilyTestBase : public testing::Test {
     return ret;
     return 0;
   }
+#endif  // !ROCKSDB_LITE
 
   void AssertCountLiveLogFiles(int value) {
+#ifndef ROCKSDB_LITE  // GetSortedWalFiles is not supported
     ASSERT_EQ(value, CountLiveLogFiles());
+#else
+    (void) value;
+#endif  // !ROCKSDB_LITE
   }
 
   void AssertNumberOfImmutableMemtables(std::vector<int> num_per_cf) {
     assert(num_per_cf.size() == handles_.size());
 
+#ifndef ROCKSDB_LITE  // GetProperty is not supported in lite
     for (size_t i = 0; i < num_per_cf.size(); ++i) {
       ASSERT_EQ(num_per_cf[i], GetProperty(static_cast<int>(i),
                                            "rocksdb.num-immutable-mem-table"));
     }
+#endif  // !ROCKSDB_LITE
   }
 
   void CopyFile(const std::string& source, const std::string& destination,
@@ -470,14 +521,14 @@ class ColumnFamilyTestBase : public testing::Test {
     return static_cast<int>(files.size());
   }
 
-  void RecalculateWriteStallConditions(
-      ColumnFamilyData* cfd, const MutableCFOptions& mutable_cf_options) {
+  void RecalculateWriteStallConditions(ColumnFamilyData* cfd,
+      const MutableCFOptions& mutable_cf_options)  {
     // add lock to avoid race condition between
     // `RecalculateWriteStallConditions` which writes to CFStats and
     // background `DBImpl::DumpStats()` threads which read CFStats
     dbfull()->TEST_LockMutex();
     cfd->RecalculateWriteStallConditions(mutable_cf_options);
-    dbfull()->TEST_UnlockMutex();
+    dbfull()-> TEST_UnlockMutex();
   }
 
   std::vector<ColumnFamilyHandle*> handles_;
@@ -533,6 +584,7 @@ TEST_P(ColumnFamilyTest, DontReuseColumnFamilyID) {
   }
 }
 
+#ifndef ROCKSDB_LITE
 TEST_P(ColumnFamilyTest, CreateCFRaceWithGetAggProperty) {
   Open();
 
@@ -555,6 +607,7 @@ TEST_P(ColumnFamilyTest, CreateCFRaceWithGetAggProperty) {
 
   ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
 }
+#endif  // !ROCKSDB_LITE
 
 class FlushEmptyCFTestWithParam
     : public ColumnFamilyTestBase,
@@ -898,6 +951,7 @@ TEST_P(ColumnFamilyTest, IgnoreRecoveredLog) {
   }
 }
 
+#ifndef ROCKSDB_LITE  // TEST functions used are not supported
 TEST_P(ColumnFamilyTest, FlushTest) {
   Open();
   CreateColumnFamiliesAndReopen({"one", "two"});
@@ -916,7 +970,8 @@ TEST_P(ColumnFamilyTest, FlushTest) {
     }
 
     for (int i = 0; i < 3; ++i) {
-      uint64_t max_total_in_memory_state = MaxTotalInMemoryState();
+      uint64_t max_total_in_memory_state =
+          MaxTotalInMemoryState();
       ASSERT_OK(Flush(i));
       AssertMaxTotalInMemoryState(max_total_in_memory_state);
     }
@@ -1012,6 +1067,7 @@ TEST_P(ColumnFamilyTest, LogDeletionTest) {
   AssertCountLiveLogFiles(4);
   Close();
 }
+#endif  // !ROCKSDB_LITE
 
 TEST_P(ColumnFamilyTest, CrashAfterFlush) {
   std::unique_ptr<FaultInjectionTestEnv> fault_env(
@@ -1051,6 +1107,7 @@ TEST_P(ColumnFamilyTest, OpenNonexistentColumnFamily) {
   ASSERT_TRUE(TryOpen({"default", "dne"}).IsInvalidArgument());
 }
 
+#ifndef ROCKSDB_LITE  // WaitForFlush() is not supported
 // Makes sure that obsolete log files get deleted
 TEST_P(ColumnFamilyTest, DifferentWriteBufferSizes) {
   // disable flushing stale column families
@@ -1152,18 +1209,20 @@ TEST_P(ColumnFamilyTest, DifferentWriteBufferSizes) {
   WaitForFlush(2);
   AssertNumberOfImmutableMemtables({0, 0, 0, 0});
   AssertCountLiveLogFiles(12);
-  PutRandomData(1, 2 * 200, 1000);
+  PutRandomData(1, 2*200, 1000);
   WaitForFlush(1);
   AssertNumberOfImmutableMemtables({0, 0, 0, 0});
   AssertCountLiveLogFiles(7);
   Close();
 }
+#endif  // !ROCKSDB_LITE
 
 // The test is commented out because we want to test that snapshot is
 // not created for memtables not supported it, but There isn't a memtable
 // that doesn't support snapshot right now. If we have one later, we can
 // re-enable the test.
 //
+// #ifndef ROCKSDB_LITE  // Cuckoo is not supported in lite
 //   TEST_P(ColumnFamilyTest, MemtableNotSupportSnapshot) {
 //   db_options_.allow_concurrent_memtable_write = false;
 //   Open();
@@ -1183,6 +1242,7 @@ TEST_P(ColumnFamilyTest, DifferentWriteBufferSizes) {
 //   {second}); auto* s3 = dbfull()->GetSnapshot(); ASSERT_TRUE(s3 == nullptr);
 //   Close();
 // }
+// #endif  // !ROCKSDB_LITE
 
 class TestComparator : public Comparator {
   int Compare(const ROCKSDB_NAMESPACE::Slice& /*a*/,
@@ -1249,6 +1309,7 @@ TEST_P(ColumnFamilyTest, DifferentMergeOperators) {
   Close();
 }
 
+#ifndef ROCKSDB_LITE  // WaitForFlush() is not supported
 TEST_P(ColumnFamilyTest, DifferentCompactionStyles) {
   Open();
   CreateColumnFamilies({"one", "two"});
@@ -1316,7 +1377,9 @@ TEST_P(ColumnFamilyTest, DifferentCompactionStyles) {
 
   Close();
 }
+#endif  // !ROCKSDB_LITE
 
+#ifndef ROCKSDB_LITE
 // Sync points not supported in RocksDB Lite
 
 TEST_P(ColumnFamilyTest, MultipleManualCompactions) {
@@ -1980,7 +2043,9 @@ TEST_P(ColumnFamilyTest, SameCFAutomaticManualCompactions) {
   ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
   ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
 }
+#endif  // !ROCKSDB_LITE
 
+#ifndef ROCKSDB_LITE  // Tailing iterator not supported
 namespace {
 std::string IterStatus(Iterator* iter) {
   std::string result;
@@ -2038,7 +2103,9 @@ TEST_P(ColumnFamilyTest, NewIteratorsTest) {
     Destroy();
   }
 }
+#endif  // !ROCKSDB_LITE
 
+#ifndef ROCKSDB_LITE  // ReadOnlyDB is not supported
 TEST_P(ColumnFamilyTest, ReadOnlyDBTest) {
   Open();
   CreateColumnFamiliesAndReopen({"one", "two", "three", "four"});
@@ -2056,6 +2123,7 @@ TEST_P(ColumnFamilyTest, ReadOnlyDBTest) {
   ASSERT_EQ("bla", Get(1, "foo"));
   ASSERT_EQ("blablablabla", Get(2, "foo"));
 
+
   // test newiterators
   {
     std::vector<Iterator*> iterators;
@@ -2087,7 +2155,9 @@ TEST_P(ColumnFamilyTest, ReadOnlyDBTest) {
   s = OpenReadOnly({"one", "four"});
   ASSERT_TRUE(!s.ok());
 }
+#endif  // !ROCKSDB_LITE
 
+#ifndef ROCKSDB_LITE  //  WaitForFlush() is not supported in lite
 TEST_P(ColumnFamilyTest, DontRollEmptyLogs) {
   Open();
   CreateColumnFamiliesAndReopen({"one", "two", "three", "four"});
@@ -2109,7 +2179,9 @@ TEST_P(ColumnFamilyTest, DontRollEmptyLogs) {
   ASSERT_EQ(static_cast<size_t>(total_new_writable_files), handles_.size() + 1);
   Close();
 }
+#endif  // !ROCKSDB_LITE
 
+#ifndef ROCKSDB_LITE  //  WaitForCompaction() is not supported in lite
 TEST_P(ColumnFamilyTest, FlushStaleColumnFamilies) {
   Open();
   CreateColumnFamilies({"one", "two"});
@@ -2156,6 +2228,7 @@ TEST_P(ColumnFamilyTest, FlushStaleColumnFamilies) {
   ASSERT_EQ(0, dbfull()->TEST_total_log_size());
   Close();
 }
+#endif  // !ROCKSDB_LITE
 
 TEST_P(ColumnFamilyTest, CreateMissingColumnFamilies) {
   Status s = TryOpen({"one", "two"});
@@ -2395,6 +2468,8 @@ TEST_P(ColumnFamilyTest, FlushAndDropRaceCondition) {
   Destroy();
 }
 
+#ifndef ROCKSDB_LITE
+// skipped as persisting options is not supported in ROCKSDB_LITE
 namespace {
 std::atomic<int> test_stage(0);
 std::atomic<bool> ordered_by_writethread(false);
@@ -2413,7 +2488,7 @@ void DropSingleColumnFamily(ColumnFamilyTest* cf_test, int cf_id,
   }
   test_stage = kChildThreadFinishDroppingColumnFamily;
 }
-}  // anonymous namespace
+}  // namespace
 
 TEST_P(ColumnFamilyTest, CreateAndDropRace) {
   const int kCfCount = 5;
@@ -2476,6 +2551,7 @@ TEST_P(ColumnFamilyTest, CreateAndDropRace) {
   ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
   ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
 }
+#endif  // !ROCKSDB_LITE
 
 TEST_P(ColumnFamilyTest, WriteStallSingleColumnFamily) {
   const uint64_t kBaseRate = 800000u;
@@ -2885,6 +2961,7 @@ TEST_P(ColumnFamilyTest, CreateDropAndDestroy) {
   ASSERT_OK(db_->DestroyColumnFamilyHandle(cfh));
 }
 
+#ifndef ROCKSDB_LITE
 TEST_P(ColumnFamilyTest, CreateDropAndDestroyWithoutFileDeletion) {
   ColumnFamilyHandle* cfh;
   Open();
@@ -2939,7 +3016,9 @@ TEST_P(ColumnFamilyTest, FlushCloseWALFiles) {
   db_options_.env = env_;
   Close();
 }
+#endif  // !ROCKSDB_LITE
 
+#ifndef ROCKSDB_LITE  // WaitForFlush() is not supported
 TEST_P(ColumnFamilyTest, IteratorCloseWALFile1) {
   SpecialEnv env(Env::Default());
   db_options_.env = &env;
@@ -3046,7 +3125,9 @@ TEST_P(ColumnFamilyTest, IteratorCloseWALFile2) {
   db_options_.env = env_;
   Close();
 }
+#endif  // !ROCKSDB_LITE
 
+#ifndef ROCKSDB_LITE  // TEST functions are not supported in lite
 TEST_P(ColumnFamilyTest, ForwardIteratorCloseWALFile) {
   SpecialEnv env(Env::Default());
   // Allow both of flush and purge job to schedule.
@@ -3122,6 +3203,7 @@ TEST_P(ColumnFamilyTest, ForwardIteratorCloseWALFile) {
   db_options_.env = env_;
   Close();
 }
+#endif  // !ROCKSDB_LITE
 
 // Disable on windows because SyncWAL requires env->IsSyncThreadSafe()
 // to return true which is not so in unbuffered mode.

db/compact_files_test.cc

@@ -3,6 +3,7 @@
 //  COPYING file in the root directory) and Apache 2.0 License
 //  (found in the LICENSE.Apache file in the root directory).
 
+#ifndef ROCKSDB_LITE
 
 #include <mutex>
 #include <string>
@@ -120,9 +121,7 @@ TEST_F(CompactFilesTest, L0ConflictsFiles) {
 TEST_F(CompactFilesTest, MultipleLevel) {
   Options options;
   options.create_if_missing = true;
-  // Otherwise background compaction can happen to
-  // drain unnecessary level
-  options.level_compaction_dynamic_level_bytes = false;
+  options.level_compaction_dynamic_level_bytes = true;
   options.num_levels = 6;
   // Add listener
   FlushedFileCollector* collector = new FlushedFileCollector();
@@ -183,6 +182,7 @@ TEST_F(CompactFilesTest, MultipleLevel) {
   for (int invalid_output_level = 0; invalid_output_level < 5;
        invalid_output_level++) {
     s = db->CompactFiles(CompactionOptions(), files, invalid_output_level);
+    std::cout << s.ToString() << std::endl;
     ASSERT_TRUE(s.IsInvalidArgument());
   }
 
@@ -348,7 +348,9 @@ TEST_F(CompactFilesTest, CompactionFilterWithGetSv) {
       return true;
     }
 
-    void SetDB(DB* db) { db_ = db; }
+    void SetDB(DB* db) {
+      db_ = db;
+    }
 
     const char* Name() const override { return "FilterWithGet"; }
 
@@ -356,6 +358,7 @@ TEST_F(CompactFilesTest, CompactionFilterWithGetSv) {
     DB* db_;
   };
 
+
   std::shared_ptr<FilterWithGet> cf(new FilterWithGet());
 
   Options options;
@@ -382,6 +385,7 @@ TEST_F(CompactFilesTest, CompactionFilterWithGetSv) {
         db->CompactFiles(ROCKSDB_NAMESPACE::CompactionOptions(), {fname}, 0));
   }
 
+
   delete db;
 }
 
@@ -396,9 +400,10 @@ TEST_F(CompactFilesTest, SentinelCompressionType) {
   }
   // Check that passing `CompressionType::kDisableCompressionOption` to
   // `CompactFiles` causes it to use the column family compression options.
-  for (auto compaction_style : {CompactionStyle::kCompactionStyleLevel,
-                                CompactionStyle::kCompactionStyleUniversal,
-                                CompactionStyle::kCompactionStyleNone}) {
+  for (auto compaction_style :
+       {CompactionStyle::kCompactionStyleLevel,
+        CompactionStyle::kCompactionStyleUniversal,
+        CompactionStyle::kCompactionStyleNone}) {
     ASSERT_OK(DestroyDB(db_name_, Options()));
     Options options;
     options.compaction_style = compaction_style;
@@ -485,8 +490,17 @@ TEST_F(CompactFilesTest, GetCompactionJobInfo) {
 }  // namespace ROCKSDB_NAMESPACE
 
 int main(int argc, char** argv) {
-  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
   ::testing::InitGoogleTest(&argc, argv);
   return RUN_ALL_TESTS();
 }
 
+#else
+#include <stdio.h>
+
+int main(int /*argc*/, char** /*argv*/) {
+  fprintf(stderr,
+          "SKIPPED as DBImpl::CompactFiles is not supported in ROCKSDB_LITE\n");
+  return 0;
+}
+
+#endif  // !ROCKSDB_LITE

db/compaction/clipping_iterator.h

@@ -188,11 +188,6 @@ class ClippingIterator : public InternalIterator {
     return iter_->GetProperty(prop_name, prop);
   }
 
-  bool IsDeleteRangeSentinelKey() const override {
-    assert(valid_);
-    return iter_->IsDeleteRangeSentinelKey();
-  }
-
  private:
   void UpdateValid() {
     assert(!iter_->Valid() || iter_->status().ok());

db/compaction/clipping_iterator_test.cc

@@ -253,7 +253,6 @@ INSTANTIATE_TEST_CASE_P(
 }  // namespace ROCKSDB_NAMESPACE
 
 int main(int argc, char** argv) {
-  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
   ::testing::InitGoogleTest(&argc, argv);
   return RUN_ALL_TESTS();
 }

db/compaction/compaction.cc

@@ -221,7 +221,7 @@ Compaction::Compaction(
     : input_vstorage_(vstorage),
       start_level_(_inputs[0].level),
       output_level_(_output_level),
-      target_output_file_size_(_target_file_size),
+      max_output_file_size_(_target_file_size),
       max_compaction_bytes_(_max_compaction_bytes),
       max_subcompactions_(_max_subcompactions),
       immutable_options_(_immutable_options),
@@ -238,19 +238,12 @@ Compaction::Compaction(
       inputs_(PopulateWithAtomicBoundaries(vstorage, std::move(_inputs))),
       grandparents_(std::move(_grandparents)),
       score_(_score),
-      bottommost_level_(
-          // For simplicity, we don't support the concept of "bottommost level"
-          // with
-          // `CompactionReason::kExternalSstIngestion` and
-          // `CompactionReason::kRefitLevel`
-          (_compaction_reason == CompactionReason::kExternalSstIngestion ||
-           _compaction_reason == CompactionReason::kRefitLevel)
-              ? false
-              : IsBottommostLevel(output_level_, vstorage, inputs_)),
+      bottommost_level_(IsBottommostLevel(output_level_, vstorage, inputs_)),
       is_full_compaction_(IsFullCompaction(vstorage, inputs_)),
       is_manual_compaction_(_manual_compaction),
       trim_ts_(_trim_ts),
       is_trivial_move_(false),
+
       compaction_reason_(_compaction_reason),
       notify_on_compaction_completion_(false),
       enable_blob_garbage_collection_(
@@ -265,15 +258,8 @@ Compaction::Compaction(
                   _blob_garbage_collection_age_cutoff > 1
               ? mutable_cf_options()->blob_garbage_collection_age_cutoff
               : _blob_garbage_collection_age_cutoff),
-      penultimate_level_(
-          // For simplicity, we don't support the concept of "penultimate level"
-          // with `CompactionReason::kExternalSstIngestion` and
-          // `CompactionReason::kRefitLevel`
-          _compaction_reason == CompactionReason::kExternalSstIngestion ||
-                  _compaction_reason == CompactionReason::kRefitLevel
-              ? Compaction::kInvalidLevel
-              : EvaluatePenultimateLevel(vstorage, immutable_options_,
-                                         start_level_, output_level_)) {
+      penultimate_level_(EvaluatePenultimateLevel(
+          immutable_options_, start_level_, output_level_)) {
   MarkFilesBeingCompacted(true);
   if (is_manual_compaction_) {
     compaction_reason_ = CompactionReason::kManualCompaction;
@@ -282,14 +268,6 @@ Compaction::Compaction(
     max_subcompactions_ = _mutable_db_options.max_subcompactions;
   }
 
-  // for the non-bottommost levels, it tries to build files match the target
-  // file size, but not guaranteed. It could be 2x the size of the target size.
-  max_output_file_size_ =
-      bottommost_level_ || grandparents_.empty() ||
-              !_immutable_options.level_compaction_dynamic_file_size
-          ? target_output_file_size_
-          : 2 * target_output_file_size_;
-
 #ifndef NDEBUG
   for (size_t i = 1; i < inputs_.size(); ++i) {
     assert(inputs_[i].level > inputs_[i - 1].level);
@@ -336,40 +314,9 @@ void Compaction::PopulatePenultimateLevelOutputRange() {
     return;
   }
 
-  // exclude the last level, the range of all input levels is the safe range
-  // of keys that can be moved up.
-  int exclude_level = number_levels_ - 1;
-  penultimate_output_range_type_ = PenultimateOutputRangeType::kNonLastRange;
-
-  // For universal compaction, the penultimate_output_range could be extended if
-  // all penultimate level files are included in the compaction (which includes
-  // the case that the penultimate level is empty).
-  if (immutable_options_.compaction_style == kCompactionStyleUniversal) {
-    exclude_level = kInvalidLevel;
-    penultimate_output_range_type_ = PenultimateOutputRangeType::kFullRange;
-    std::set<uint64_t> penultimate_inputs;
-    for (const auto& input_lvl : inputs_) {
-      if (input_lvl.level == penultimate_level_) {
-        for (const auto& file : input_lvl.files) {
-          penultimate_inputs.emplace(file->fd.GetNumber());
-        }
-      }
-    }
-    auto penultimate_files = input_vstorage_->LevelFiles(penultimate_level_);
-    for (const auto& file : penultimate_files) {
-      if (penultimate_inputs.find(file->fd.GetNumber()) ==
-          penultimate_inputs.end()) {
-        exclude_level = number_levels_ - 1;
-        penultimate_output_range_type_ =
-            PenultimateOutputRangeType::kNonLastRange;
-        break;
-      }
-    }
-  }
-
   GetBoundaryKeys(input_vstorage_, inputs_,
                   &penultimate_level_smallest_user_key_,
-                  &penultimate_level_largest_user_key_, exclude_level);
+                  &penultimate_level_largest_user_key_, number_levels_ - 1);
 }
 
 Compaction::~Compaction() {
@@ -387,8 +334,6 @@ bool Compaction::SupportsPerKeyPlacement() const {
 
 int Compaction::GetPenultimateLevel() const { return penultimate_level_; }
 
-// smallest_key and largest_key include timestamps if user-defined timestamp is
-// enabled.
 bool Compaction::OverlapPenultimateLevelOutputRange(
     const Slice& smallest_key, const Slice& largest_key) const {
   if (!SupportsPerKeyPlacement()) {
@@ -397,30 +342,21 @@ bool Compaction::OverlapPenultimateLevelOutputRange(
   const Comparator* ucmp =
       input_vstorage_->InternalComparator()->user_comparator();
 
-  return ucmp->CompareWithoutTimestamp(
-             smallest_key, penultimate_level_largest_user_key_) <= 0 &&
-         ucmp->CompareWithoutTimestamp(
-             largest_key, penultimate_level_smallest_user_key_) >= 0;
+  return ucmp->Compare(smallest_key, penultimate_level_largest_user_key_) <=
+             0 &&
+         ucmp->Compare(largest_key, penultimate_level_smallest_user_key_) >= 0;
 }
 
-// key includes timestamp if user-defined timestamp is enabled.
 bool Compaction::WithinPenultimateLevelOutputRange(const Slice& key) const {
   if (!SupportsPerKeyPlacement()) {
     return false;
   }
 
-  if (penultimate_level_smallest_user_key_.empty() ||
-      penultimate_level_largest_user_key_.empty()) {
-    return false;
-  }
-
   const Comparator* ucmp =
       input_vstorage_->InternalComparator()->user_comparator();
 
-  return ucmp->CompareWithoutTimestamp(
-             key, penultimate_level_smallest_user_key_) >= 0 &&
-         ucmp->CompareWithoutTimestamp(
-             key, penultimate_level_largest_user_key_) <= 0;
+  return ucmp->Compare(key, penultimate_level_smallest_user_key_) >= 0 &&
+         ucmp->Compare(key, penultimate_level_largest_user_key_) <= 0;
 }
 
 bool Compaction::InputCompressionMatchesOutput() const {
@@ -794,18 +730,7 @@ uint64_t Compaction::MinInputFileOldestAncesterTime(
   return min_oldest_ancester_time;
 }
 
-uint64_t Compaction::MinInputFileEpochNumber() const {
-  uint64_t min_epoch_number = std::numeric_limits<uint64_t>::max();
-  for (const auto& inputs_per_level : inputs_) {
-    for (const auto& file : inputs_per_level.files) {
-      min_epoch_number = std::min(min_epoch_number, file->epoch_number);
-    }
-  }
-  return min_epoch_number;
-}
-
 int Compaction::EvaluatePenultimateLevel(
-    const VersionStorageInfo* vstorage,
     const ImmutableOptions& immutable_options, const int start_level,
     const int output_level) {
   // TODO: currently per_key_placement feature only support level and universal
@@ -820,19 +745,7 @@ int Compaction::EvaluatePenultimateLevel(
 
   int penultimate_level = output_level - 1;
   assert(penultimate_level < immutable_options.num_levels);
-  if (penultimate_level <= 0) {
-    return kInvalidLevel;
-  }
-
-  // If the penultimate level is not within input level -> output level range
-  // check if the penultimate output level is empty, if it's empty, it could
-  // also be locked for the penultimate output.
-  // TODO: ideally, it only needs to check if there's a file within the
-  //  compaction output key range. For simplicity, it just check if there's any
-  //  file on the penultimate level.
-  if (start_level == immutable_options.num_levels - 1 &&
-      (immutable_options.compaction_style != kCompactionStyleUniversal ||
-       !vstorage->LevelFiles(penultimate_level).empty())) {
+  if (penultimate_level <= 0 || penultimate_level < start_level) {
     return kInvalidLevel;
   }
 

db/compaction/compaction.h

@@ -87,15 +87,6 @@ class Compaction {
                  BlobGarbageCollectionPolicy::kUseDefault,
              double blob_garbage_collection_age_cutoff = -1);
 
-  // The type of the penultimate level output range
-  enum class PenultimateOutputRangeType : int {
-    kNotSupported,  // it cannot output to the penultimate level
-    kFullRange,     // any data could be output to the penultimate level
-    kNonLastRange,  // only the keys within non_last_level compaction inputs can
-                    // be outputted to the penultimate level
-    kDisabled,      // no data can be outputted to the penultimate level
-  };
-
   // No copying allowed
   Compaction(const Compaction&) = delete;
   void operator=(const Compaction&) = delete;
@@ -172,9 +163,6 @@ class Compaction {
   // Maximum size of files to build during this compaction.
   uint64_t max_output_file_size() const { return max_output_file_size_; }
 
-  // Target output file size for this compaction
-  uint64_t target_output_file_size() const { return target_output_file_size_; }
-
   // What compression for output
   CompressionType output_compression() const { return output_compression_; }
 
@@ -319,18 +307,6 @@ class Compaction {
 
   Slice GetLargestUserKey() const { return largest_user_key_; }
 
-  Slice GetPenultimateLevelSmallestUserKey() const {
-    return penultimate_level_smallest_user_key_;
-  }
-
-  Slice GetPenultimateLevelLargestUserKey() const {
-    return penultimate_level_largest_user_key_;
-  }
-
-  PenultimateOutputRangeType GetPenultimateOutputRangeType() const {
-    return penultimate_output_range_type_;
-  }
-
   // Return true if the compaction supports per_key_placement
   bool SupportsPerKeyPlacement() const;
 
@@ -340,8 +316,6 @@ class Compaction {
 
   // Return true if the given range is overlap with penultimate level output
   // range.
-  // Both smallest_key and largest_key include timestamps if user-defined
-  // timestamp is enabled.
   bool OverlapPenultimateLevelOutputRange(const Slice& smallest_key,
                                           const Slice& largest_key) const;
 
@@ -351,7 +325,6 @@ class Compaction {
   // If per_key_placement is not supported, always return false.
   // TODO: currently it doesn't support moving data from the last level to the
   //  penultimate level
-  //  key includes timestamp if user-defined timestamp is enabled.
   bool WithinPenultimateLevelOutputRange(const Slice& key) const;
 
   CompactionReason compaction_reason() const { return compaction_reason_; }
@@ -378,9 +351,6 @@ class Compaction {
   // This is used to filter out some input files' ancester's time range.
   uint64_t MinInputFileOldestAncesterTime(const InternalKey* start,
                                           const InternalKey* end) const;
-  // Return the minimum epoch number among
-  // input files' associated with this compaction
-  uint64_t MinInputFileEpochNumber() const;
 
   // Called by DBImpl::NotifyOnCompactionCompleted to make sure number of
   // compaction begin and compaction completion callbacks match.
@@ -393,18 +363,11 @@ class Compaction {
   }
 
   static constexpr int kInvalidLevel = -1;
-
   // Evaluate penultimate output level. If the compaction supports
   // per_key_placement feature, it returns the penultimate level number.
   // Otherwise, it's set to kInvalidLevel (-1), which means
   // output_to_penultimate_level is not supported.
-  // Note: even the penultimate level output is supported (PenultimateLevel !=
-  // kInvalidLevel), some key range maybe unsafe to be outputted to the
-  // penultimate level. The safe key range is populated by
-  // `PopulatePenultimateLevelOutputRange()`.
-  // Which could potentially disable all penultimate level output.
-  static int EvaluatePenultimateLevel(const VersionStorageInfo* vstorage,
-                                      const ImmutableOptions& immutable_options,
+  static int EvaluatePenultimateLevel(const ImmutableOptions& immutable_options,
                                       const int start_level,
                                       const int output_level);
 
@@ -421,6 +384,11 @@ class Compaction {
   // populate penultimate level output range, which will be used to determine if
   // a key is safe to output to the penultimate level (details see
   // `Compaction::WithinPenultimateLevelOutputRange()`.
+  // TODO: Currently the penultimate level output range is the min/max keys of
+  //  non-last-level input files. Which is only good if there's no key moved
+  //  from the last level to the penultimate level. For a more complicated per
+  //  key placement which may move data from the last level to the penultimate
+  //  level, it needs extra check.
   void PopulatePenultimateLevelOutputRange();
 
   // Get the atomic file boundaries for all files in the compaction. Necessary
@@ -444,7 +412,6 @@ class Compaction {
 
   const int start_level_;   // the lowest level to be compacted
   const int output_level_;  // levels to which output files are stored
-  uint64_t target_output_file_size_;
   uint64_t max_output_file_size_;
   uint64_t max_compaction_bytes_;
   uint32_t max_subcompactions_;
@@ -503,11 +470,9 @@ class Compaction {
   TablePropertiesCollection output_table_properties_;
 
   // smallest user keys in compaction
-  // includes timestamp if user-defined timestamp is enabled.
   Slice smallest_user_key_;
 
   // largest user keys in compaction
-  // includes timestamp if user-defined timestamp is enabled.
   Slice largest_user_key_;
 
   // Reason for compaction
@@ -528,12 +493,8 @@ class Compaction {
   const int penultimate_level_;
 
   // Key range for penultimate level output
-  // includes timestamp if user-defined timestamp is enabled.
-  // penultimate_output_range_type_ shows the range type
   Slice penultimate_level_smallest_user_key_;
   Slice penultimate_level_largest_user_key_;
-  PenultimateOutputRangeType penultimate_output_range_type_ =
-      PenultimateOutputRangeType::kNotSupported;
 };
 
 #ifndef NDEBUG
@@ -546,16 +507,13 @@ struct PerKeyPlacementContext {
   const Slice value;
   const SequenceNumber seq_num;
 
-  bool& output_to_penultimate_level;
+  bool output_to_penultimate_level;
 
   PerKeyPlacementContext(int _level, Slice _key, Slice _value,
-                         SequenceNumber _seq_num,
-                         bool& _output_to_penultimate_level)
-      : level(_level),
-        key(_key),
-        value(_value),
-        seq_num(_seq_num),
-        output_to_penultimate_level(_output_to_penultimate_level) {}
+                         SequenceNumber _seq_num)
+      : level(_level), key(_key), value(_value), seq_num(_seq_num) {
+    output_to_penultimate_level = false;
+  }
 };
 #endif /* !NDEBUG */
 

db/compaction/compaction_iterator.cc

@@ -13,7 +13,6 @@
 #include "db/blob/blob_index.h"
 #include "db/blob/prefetch_buffer_collection.h"
 #include "db/snapshot_checker.h"
-#include "db/wide/wide_column_serialization.h"
 #include "logging/logging.h"
 #include "port/likely.h"
 #include "rocksdb/listener.h"
@@ -35,8 +34,7 @@ CompactionIterator::CompactionIterator(
     const std::atomic<bool>* shutting_down,
     const std::shared_ptr<Logger> info_log,
     const std::string* full_history_ts_low,
-    const SequenceNumber preserve_time_min_seqno,
-    const SequenceNumber preclude_last_level_min_seqno)
+    const SequenceNumber penultimate_level_cutoff_seqno)
     : CompactionIterator(
           input, cmp, merge_helper, last_sequence, snapshots,
           earliest_write_conflict_snapshot, job_snapshot, snapshot_checker, env,
@@ -46,7 +44,7 @@ CompactionIterator::CompactionIterator(
           std::unique_ptr<CompactionProxy>(
               compaction ? new RealCompaction(compaction) : nullptr),
           compaction_filter, shutting_down, info_log, full_history_ts_low,
-          preserve_time_min_seqno, preclude_last_level_min_seqno) {}
+          penultimate_level_cutoff_seqno) {}
 
 CompactionIterator::CompactionIterator(
     InternalIterator* input, const Comparator* cmp, MergeHelper* merge_helper,
@@ -63,8 +61,7 @@ CompactionIterator::CompactionIterator(
     const std::atomic<bool>* shutting_down,
     const std::shared_ptr<Logger> info_log,
     const std::string* full_history_ts_low,
-    const SequenceNumber preserve_time_min_seqno,
-    const SequenceNumber preclude_last_level_min_seqno)
+    const SequenceNumber penultimate_level_cutoff_seqno)
     : input_(input, cmp,
              !compaction || compaction->DoesInputReferenceBlobFiles()),
       cmp_(cmp),
@@ -108,10 +105,8 @@ CompactionIterator::CompactionIterator(
       current_key_committed_(false),
       cmp_with_history_ts_low_(0),
       level_(compaction_ == nullptr ? 0 : compaction_->level()),
-      preserve_time_min_seqno_(preserve_time_min_seqno),
-      preclude_last_level_min_seqno_(preclude_last_level_min_seqno) {
+      penultimate_level_cutoff_seqno_(penultimate_level_cutoff_seqno) {
   assert(snapshots_ != nullptr);
-  assert(preserve_time_min_seqno_ <= preclude_last_level_min_seqno_);
 
   if (compaction_ != nullptr) {
     level_ptrs_ = std::vector<size_t>(compaction_->number_levels(), 0);
@@ -125,9 +120,6 @@ CompactionIterator::CompactionIterator(
          timestamp_size_ == full_history_ts_low_->size());
 #endif
   input_.SetPinnedItersMgr(&pinned_iters_mgr_);
-  // The default `merge_until_status_` does not need to be checked since it is
-  // overwritten as soon as `MergeUntil()` is called
-  merge_until_status_.PermitUncheckedError();
   TEST_SYNC_POINT_CALLBACK("CompactionIterator:AfterInit", compaction_.get());
 }
 
@@ -172,30 +164,11 @@ void CompactionIterator::Next() {
       }
 
       // Keep current_key_ in sync.
-      if (0 == timestamp_size_) {
-        current_key_.UpdateInternalKey(ikey_.sequence, ikey_.type);
-      } else {
-        Slice ts = ikey_.GetTimestamp(timestamp_size_);
-        current_key_.UpdateInternalKey(ikey_.sequence, ikey_.type, &ts);
-      }
+      current_key_.UpdateInternalKey(ikey_.sequence, ikey_.type);
       key_ = current_key_.GetInternalKey();
       ikey_.user_key = current_key_.GetUserKey();
       validity_info_.SetValid(ValidContext::kMerge1);
     } else {
-      if (merge_until_status_.IsMergeInProgress()) {
-        // `Status::MergeInProgress()` tells us that the previous `MergeUntil()`
-        // produced only merge operands. Those merge operands were accessed and
-        // written out using `merge_out_iter_`. Since `merge_out_iter_` is
-        // exhausted at this point, all merge operands have been written out.
-        //
-        // Still, there may be a base value (PUT, DELETE, SINGLEDEL, etc.) that
-        // needs to be written out. Normally, `CompactionIterator` would skip it
-        // on the basis that it has already output something in the same
-        // snapshot stripe. To prevent this, we reset `has_current_user_key_` to
-        // trick the future iteration from finding out the snapshot stripe is
-        // unchanged.
-        has_current_user_key_ = false;
-      }
       // We consumed all pinned merge operands, release pinned iterators
       pinned_iters_mgr_.ReleasePinnedData();
       // MergeHelper moves the iterator to the first record after the merged
@@ -222,47 +195,39 @@ void CompactionIterator::Next() {
 
 bool CompactionIterator::InvokeFilterIfNeeded(bool* need_skip,
                                               Slice* skip_until) {
-  if (!compaction_filter_) {
+  // TODO: support compaction filter for wide-column entities
+  if (!compaction_filter_ ||
+      (ikey_.type != kTypeValue && ikey_.type != kTypeBlobIndex)) {
     return true;
   }
-
-  if (ikey_.type != kTypeValue && ikey_.type != kTypeBlobIndex &&
-      ikey_.type != kTypeWideColumnEntity) {
-    return true;
-  }
-
-  CompactionFilter::Decision decision =
-      CompactionFilter::Decision::kUndetermined;
+  bool error = false;
+  // If the user has specified a compaction filter and the sequence
+  // number is greater than any external snapshot, then invoke the
+  // filter. If the return value of the compaction filter is true,
+  // replace the entry with a deletion marker.
+  CompactionFilter::Decision filter = CompactionFilter::Decision::kUndetermined;
+  compaction_filter_value_.clear();
+  compaction_filter_skip_until_.Clear();
   CompactionFilter::ValueType value_type =
       ikey_.type == kTypeValue ? CompactionFilter::ValueType::kValue
-      : ikey_.type == kTypeBlobIndex
-          ? CompactionFilter::ValueType::kBlobIndex
-          : CompactionFilter::ValueType::kWideColumnEntity;
-
+                               : CompactionFilter::ValueType::kBlobIndex;
   // Hack: pass internal key to BlobIndexCompactionFilter since it needs
   // to get sequence number.
   assert(compaction_filter_);
-  const Slice& filter_key =
-      (ikey_.type != kTypeBlobIndex ||
+  Slice& filter_key =
+      (ikey_.type == kTypeValue ||
        !compaction_filter_->IsStackedBlobDbInternalCompactionFilter())
           ? ikey_.user_key
           : key_;
-
-  compaction_filter_value_.clear();
-  compaction_filter_skip_until_.Clear();
-
-  std::vector<std::pair<std::string, std::string>> new_columns;
-
   {
     StopWatchNano timer(clock_, report_detailed_time_);
-
-    if (ikey_.type == kTypeBlobIndex) {
-      decision = compaction_filter_->FilterBlobByKey(
+    if (kTypeBlobIndex == ikey_.type) {
+      filter = compaction_filter_->FilterBlobByKey(
           level_, filter_key, &compaction_filter_value_,
           compaction_filter_skip_until_.rep());
-      if (decision == CompactionFilter::Decision::kUndetermined &&
+      if (CompactionFilter::Decision::kUndetermined == filter &&
           !compaction_filter_->IsStackedBlobDbInternalCompactionFilter()) {
-        if (!compaction_) {
+        if (compaction_ == nullptr) {
           status_ =
               Status::Corruption("Unexpected blob index outside of compaction");
           validity_info_.Invalidate();
@@ -308,61 +273,33 @@ bool CompactionIterator::InvokeFilterIfNeeded(bool* need_skip,
         value_type = CompactionFilter::ValueType::kValue;
       }
     }
-
-    if (decision == CompactionFilter::Decision::kUndetermined) {
-      const Slice* existing_val = nullptr;
-      const WideColumns* existing_col = nullptr;
-
-      WideColumns existing_columns;
-
-      if (ikey_.type != kTypeWideColumnEntity) {
-        if (!blob_value_.empty()) {
-          existing_val = &blob_value_;
-        } else {
-          existing_val = &value_;
-        }
-      } else {
-        Slice value_copy = value_;
-        const Status s =
-            WideColumnSerialization::Deserialize(value_copy, existing_columns);
-
-        if (!s.ok()) {
-          status_ = s;
-          validity_info_.Invalidate();
-          return false;
-        }
-
-        existing_col = &existing_columns;
-      }
-
-      decision = compaction_filter_->FilterV3(
-          level_, filter_key, value_type, existing_val, existing_col,
-          &compaction_filter_value_, &new_columns,
+    if (CompactionFilter::Decision::kUndetermined == filter) {
+      filter = compaction_filter_->FilterV2(
+          level_, filter_key, value_type,
+          blob_value_.empty() ? value_ : blob_value_, &compaction_filter_value_,
           compaction_filter_skip_until_.rep());
     }
-
     iter_stats_.total_filter_time +=
         env_ != nullptr && report_detailed_time_ ? timer.ElapsedNanos() : 0;
   }
 
-  if (decision == CompactionFilter::Decision::kUndetermined) {
-    // Should not reach here, since FilterV2/FilterV3 should never return
-    // kUndetermined.
-    status_ = Status::NotSupported(
-        "FilterV2/FilterV3 should never return kUndetermined");
+  if (CompactionFilter::Decision::kUndetermined == filter) {
+    // Should not reach here, since FilterV2 should never return kUndetermined.
+    status_ =
+        Status::NotSupported("FilterV2() should never return kUndetermined");
     validity_info_.Invalidate();
     return false;
   }
 
-  if (decision == CompactionFilter::Decision::kRemoveAndSkipUntil &&
+  if (filter == CompactionFilter::Decision::kRemoveAndSkipUntil &&
       cmp_->Compare(*compaction_filter_skip_until_.rep(), ikey_.user_key) <=
           0) {
     // Can't skip to a key smaller than the current one.
-    // Keep the key as per FilterV2/FilterV3 documentation.
-    decision = CompactionFilter::Decision::kKeep;
+    // Keep the key as per FilterV2 documentation.
+    filter = CompactionFilter::Decision::kKeep;
   }
 
-  if (decision == CompactionFilter::Decision::kRemove) {
+  if (filter == CompactionFilter::Decision::kRemove) {
     // convert the current key to a delete; key_ is pointing into
     // current_key_ at this point, so updating current_key_ updates key()
     ikey_.type = kTypeDeletion;
@@ -370,7 +307,7 @@ bool CompactionIterator::InvokeFilterIfNeeded(bool* need_skip,
     // no value associated with delete
     value_.clear();
     iter_stats_.num_record_drop_user++;
-  } else if (decision == CompactionFilter::Decision::kPurge) {
+  } else if (filter == CompactionFilter::Decision::kPurge) {
     // convert the current key to a single delete; key_ is pointing into
     // current_key_ at this point, so updating current_key_ updates key()
     ikey_.type = kTypeSingleDeletion;
@@ -378,19 +315,19 @@ bool CompactionIterator::InvokeFilterIfNeeded(bool* need_skip,
     // no value associated with single delete
     value_.clear();
     iter_stats_.num_record_drop_user++;
-  } else if (decision == CompactionFilter::Decision::kChangeValue) {
-    if (ikey_.type != kTypeValue) {
+  } else if (filter == CompactionFilter::Decision::kChangeValue) {
+    if (ikey_.type == kTypeBlobIndex) {
+      // value transfer from blob file to inlined data
       ikey_.type = kTypeValue;
-      current_key_.UpdateInternalKey(ikey_.sequence, kTypeValue);
+      current_key_.UpdateInternalKey(ikey_.sequence, ikey_.type);
     }
-
     value_ = compaction_filter_value_;
-  } else if (decision == CompactionFilter::Decision::kRemoveAndSkipUntil) {
+  } else if (filter == CompactionFilter::Decision::kRemoveAndSkipUntil) {
     *need_skip = true;
     compaction_filter_skip_until_.ConvertFromUserKey(kMaxSequenceNumber,
                                                      kValueTypeForSeek);
     *skip_until = compaction_filter_skip_until_.Encode();
-  } else if (decision == CompactionFilter::Decision::kChangeBlobIndex) {
+  } else if (filter == CompactionFilter::Decision::kChangeBlobIndex) {
     // Only the StackableDB-based BlobDB impl's compaction filter should return
     // kChangeBlobIndex. Decision about rewriting blob and changing blob index
     // in the integrated BlobDB impl is made in subsequent call to
@@ -402,56 +339,23 @@ bool CompactionIterator::InvokeFilterIfNeeded(bool* need_skip,
       validity_info_.Invalidate();
       return false;
     }
-
-    if (ikey_.type != kTypeBlobIndex) {
+    if (ikey_.type == kTypeValue) {
+      // value transfer from inlined data to blob file
       ikey_.type = kTypeBlobIndex;
-      current_key_.UpdateInternalKey(ikey_.sequence, kTypeBlobIndex);
+      current_key_.UpdateInternalKey(ikey_.sequence, ikey_.type);
     }
-
     value_ = compaction_filter_value_;
-  } else if (decision == CompactionFilter::Decision::kIOError) {
+  } else if (filter == CompactionFilter::Decision::kIOError) {
     if (!compaction_filter_->IsStackedBlobDbInternalCompactionFilter()) {
       status_ = Status::NotSupported(
           "CompactionFilter for integrated BlobDB should not return kIOError");
       validity_info_.Invalidate();
       return false;
     }
-
     status_ = Status::IOError("Failed to access blob during compaction filter");
-    validity_info_.Invalidate();
-    return false;
-  } else if (decision == CompactionFilter::Decision::kChangeWideColumnEntity) {
-    WideColumns sorted_columns;
-
-    sorted_columns.reserve(new_columns.size());
-    for (const auto& column : new_columns) {
-      sorted_columns.emplace_back(column.first, column.second);
-    }
-
-    std::sort(sorted_columns.begin(), sorted_columns.end(),
-              [](const WideColumn& lhs, const WideColumn& rhs) {
-                return lhs.name().compare(rhs.name()) < 0;
-              });
-
-    {
-      const Status s = WideColumnSerialization::Serialize(
-          sorted_columns, compaction_filter_value_);
-      if (!s.ok()) {
-        status_ = s;
-        validity_info_.Invalidate();
-        return false;
-      }
-    }
-
-    if (ikey_.type != kTypeWideColumnEntity) {
-      ikey_.type = kTypeWideColumnEntity;
-      current_key_.UpdateInternalKey(ikey_.sequence, kTypeWideColumnEntity);
-    }
-
-    value_ = compaction_filter_value_;
+    error = true;
   }
-
-  return true;
+  return !error;
 }
 
 void CompactionIterator::NextFromInput() {
@@ -464,7 +368,6 @@ void CompactionIterator::NextFromInput() {
     value_ = input_.value();
     blob_value_.Reset();
     iter_stats_.num_input_records++;
-    is_range_del_ = input_.IsDeleteRangeSentinelKey();
 
     Status pik_status = ParseInternalKey(key_, &ikey_, allow_data_in_errors_);
     if (!pik_status.ok()) {
@@ -484,10 +387,7 @@ void CompactionIterator::NextFromInput() {
       break;
     }
     TEST_SYNC_POINT_CALLBACK("CompactionIterator:ProcessKV", &ikey_);
-    if (is_range_del_) {
-      validity_info_.SetValid(kRangeDeletion);
-      break;
-    }
+
     // Update input statistics
     if (ikey_.type == kTypeDeletion || ikey_.type == kTypeSingleDeletion ||
         ikey_.type == kTypeDeletionWithTimestamp) {
@@ -709,14 +609,6 @@ void CompactionIterator::NextFromInput() {
 
       ParsedInternalKey next_ikey;
       AdvanceInputIter();
-      while (input_.Valid() && input_.IsDeleteRangeSentinelKey() &&
-             ParseInternalKey(input_.key(), &next_ikey, allow_data_in_errors_)
-                 .ok() &&
-             cmp_->EqualWithoutTimestamp(ikey_.user_key, next_ikey.user_key)) {
-        // skip range tombstone start keys with the same user key
-        // since they are not "real" point keys.
-        AdvanceInputIter();
-      }
 
       // Check whether the next key exists, is not corrupt, and is the same key
       // as the single delete.
@@ -724,7 +616,6 @@ void CompactionIterator::NextFromInput() {
           ParseInternalKey(input_.key(), &next_ikey, allow_data_in_errors_)
               .ok() &&
           cmp_->EqualWithoutTimestamp(ikey_.user_key, next_ikey.user_key)) {
-        assert(!input_.IsDeleteRangeSentinelKey());
 #ifndef NDEBUG
         const Compaction* c =
             compaction_ ? compaction_->real_compaction() : nullptr;
@@ -930,8 +821,8 @@ void CompactionIterator::NextFromInput() {
                  cmp_with_history_ts_low_ < 0)) &&
                bottommost_level_) {
       // Handle the case where we have a delete key at the bottom most level
-      // We can skip outputting the key iff there are no subsequent puts for
-      // this key
+      // We can skip outputting the key iff there are no subsequent puts for this
+      // key
       assert(!compaction_ || compaction_->KeyNotExistsBeyondOutputLevel(
                                  ikey_.user_key, &level_ptrs_));
       ParsedInternalKey next_ikey;
@@ -949,19 +840,17 @@ void CompactionIterator::NextFromInput() {
       // Note that a deletion marker of type kTypeDeletionWithTimestamp will be
       // considered to have a different user key unless the timestamp is older
       // than *full_history_ts_low_.
-      //
-      // Range tombstone start keys are skipped as they are not "real" keys.
       while (!IsPausingManualCompaction() && !IsShuttingDown() &&
              input_.Valid() &&
              (ParseInternalKey(input_.key(), &next_ikey, allow_data_in_errors_)
                   .ok()) &&
              cmp_->EqualWithoutTimestamp(ikey_.user_key, next_ikey.user_key) &&
-             (prev_snapshot == 0 || input_.IsDeleteRangeSentinelKey() ||
+             (prev_snapshot == 0 ||
               DefinitelyNotInSnapshot(next_ikey.sequence, prev_snapshot))) {
         AdvanceInputIter();
       }
-      // If you find you still need to output a row with this key, we need to
-      // output the delete too
+      // If you find you still need to output a row with this key, we need to output the
+      // delete too
       if (input_.Valid() &&
           (ParseInternalKey(input_.key(), &next_ikey, allow_data_in_errors_)
                .ok()) &&
@@ -982,15 +871,14 @@ void CompactionIterator::NextFromInput() {
       // have hit (A)
       // We encapsulate the merge related state machine in a different
       // object to minimize change to the existing flow.
-      merge_until_status_ = merge_helper_->MergeUntil(
+      Status s = merge_helper_->MergeUntil(
           &input_, range_del_agg_, prev_snapshot, bottommost_level_,
-          allow_data_in_errors_, blob_fetcher_.get(), full_history_ts_low_,
-          prefetch_buffers_.get(), &iter_stats_);
+          allow_data_in_errors_, blob_fetcher_.get(), prefetch_buffers_.get(),
+          &iter_stats_);
       merge_out_iter_.SeekToFirst();
 
-      if (!merge_until_status_.ok() &&
-          !merge_until_status_.IsMergeInProgress()) {
-        status_ = merge_until_status_;
+      if (!s.ok() && !s.IsMergeInProgress()) {
+        status_ = s;
         return;
       } else if (merge_out_iter_.Valid()) {
         // NOTE: key, value, and ikey_ refer to old entries.
@@ -1026,17 +914,8 @@ void CompactionIterator::NextFromInput() {
     } else {
       // 1. new user key -OR-
       // 2. different snapshot stripe
-      // If user-defined timestamp is enabled, we consider keys for GC if they
-      // are below history_ts_low_. CompactionRangeDelAggregator::ShouldDelete()
-      // only considers range deletions that are at or below history_ts_low_ and
-      // trim_ts_. We drop keys here that are below history_ts_low_ and are
-      // covered by a range tombstone that is at or below history_ts_low_ and
-      // trim_ts.
-      bool should_delete = false;
-      if (!timestamp_size_ || cmp_with_history_ts_low_ < 0) {
-        should_delete = range_del_agg_->ShouldDelete(
-            key_, RangeDelPositioningMode::kForwardTraversal);
-      }
+      bool should_delete = range_del_agg_->ShouldDelete(
+          key_, RangeDelPositioningMode::kForwardTraversal);
       if (should_delete) {
         ++iter_stats_.num_record_drop_hidden;
         ++iter_stats_.num_record_drop_range_del;
@@ -1200,26 +1079,24 @@ void CompactionIterator::GarbageCollectBlobIfNeeded() {
 }
 
 void CompactionIterator::DecideOutputLevel() {
-  assert(compaction_->SupportsPerKeyPlacement());
+#ifndef NDEBUG
+  // Could be overridden by unittest
+  PerKeyPlacementContext context(level_, ikey_.user_key, value_,
+                                 ikey_.sequence);
+  TEST_SYNC_POINT_CALLBACK("CompactionIterator::PrepareOutput.context",
+                           &context);
+  output_to_penultimate_level_ = context.output_to_penultimate_level;
+#else
   output_to_penultimate_level_ = false;
+#endif  // NDEBUG
+
   // if the key is newer than the cutoff sequence or within the earliest
   // snapshot, it should output to the penultimate level.
-  if (ikey_.sequence > preclude_last_level_min_seqno_ ||
+  if (ikey_.sequence > penultimate_level_cutoff_seqno_ ||
       ikey_.sequence > earliest_snapshot_) {
     output_to_penultimate_level_ = true;
   }
 
-#ifndef NDEBUG
-  // Could be overridden by unittest
-  PerKeyPlacementContext context(level_, ikey_.user_key, value_, ikey_.sequence,
-                                 output_to_penultimate_level_);
-  TEST_SYNC_POINT_CALLBACK("CompactionIterator::PrepareOutput.context",
-                           &context);
-  if (ikey_.sequence > earliest_snapshot_) {
-    output_to_penultimate_level_ = true;
-  }
-#endif  // NDEBUG
-
   if (output_to_penultimate_level_) {
     // If it's decided to output to the penultimate level, but unsafe to do so,
     // still output to the last level. For example, moving the data from a lower
@@ -1251,12 +1128,10 @@ void CompactionIterator::DecideOutputLevel() {
 
 void CompactionIterator::PrepareOutput() {
   if (Valid()) {
-    if (LIKELY(!is_range_del_)) {
-      if (ikey_.type == kTypeValue) {
-        ExtractLargeValueIfNeeded();
-      } else if (ikey_.type == kTypeBlobIndex) {
-        GarbageCollectBlobIfNeeded();
-      }
+    if (ikey_.type == kTypeValue) {
+      ExtractLargeValueIfNeeded();
+    } else if (ikey_.type == kTypeBlobIndex) {
+      GarbageCollectBlobIfNeeded();
     }
 
     if (compaction_ != nullptr && compaction_->SupportsPerKeyPlacement()) {
@@ -1279,7 +1154,7 @@ void CompactionIterator::PrepareOutput() {
         DefinitelyInSnapshot(ikey_.sequence, earliest_snapshot_) &&
         ikey_.type != kTypeMerge && current_key_committed_ &&
         !output_to_penultimate_level_ &&
-        ikey_.sequence < preserve_time_min_seqno_ && !is_range_del_) {
+        ikey_.sequence < penultimate_level_cutoff_seqno_) {
       if (ikey_.type == kTypeDeletion ||
           (ikey_.type == kTypeSingleDeletion && timestamp_size_ == 0)) {
         ROCKS_LOG_FATAL(
@@ -1325,8 +1200,8 @@ inline SequenceNumber CompactionIterator::findEarliestVisibleSnapshot(
     ROCKS_LOG_FATAL(info_log_,
                     "No snapshot left in findEarliestVisibleSnapshot");
   }
-  auto snapshots_iter =
-      std::lower_bound(snapshots_->begin(), snapshots_->end(), in);
+  auto snapshots_iter = std::lower_bound(
+      snapshots_->begin(), snapshots_->end(), in);
   assert(prev_snapshot != nullptr);
   if (snapshots_iter == snapshots_->begin()) {
     *prev_snapshot = 0;
@@ -1341,8 +1216,8 @@ inline SequenceNumber CompactionIterator::findEarliestVisibleSnapshot(
     }
   }
   if (snapshot_checker_ == nullptr) {
-    return snapshots_iter != snapshots_->end() ? *snapshots_iter
-                                               : kMaxSequenceNumber;
+    return snapshots_iter != snapshots_->end()
+      ? *snapshots_iter : kMaxSequenceNumber;
   }
   bool has_released_snapshot = !released_snapshots_.empty();
   for (; snapshots_iter != snapshots_->end(); ++snapshots_iter) {
@@ -1412,7 +1287,6 @@ std::unique_ptr<BlobFetcher> CompactionIterator::CreateBlobFetcherIfNeeded(
   }
 
   ReadOptions read_options;
-  read_options.io_activity = Env::IOActivity::kCompaction;
   read_options.fill_cache = false;
 
   return std::unique_ptr<BlobFetcher>(new BlobFetcher(version, read_options));