Merge pull request #764 from dmittendorf/fix-java-static-packaging

Fix crossbuild jar packaging Conflicts: Makefile java/crossbuild/build-linux-centos.sh
Merge pull request #759 from jwlent55/statically-load-compression-libraries
2026-11-06 11:02:36 -05:00 · 2015-10-15 15:05:20 -07:00 · 2015-10-15 15:04:53 -07:00 · 2015-09-30 11:45:32 -07:00 · 2015-09-30 11:37:27 -07:00 · 2015-09-30 11:15:35 -07:00
390 changed files with 8035 additions and 31006 deletions
@@ -38,7 +38,7 @@ coverage/COVERAGE_REPORT
 .gdbhistory
 package/
 .phutil_module_cache
-unity.a
+unity
 tags
 rocksdb_dump
 rocksdb_undump
@@ -52,8 +52,6 @@ java/include/org_rocksdb_*.h
 .idea/
 *.iml

-rocksdb.cc
-rocksdb.h
 unity.cc
 java/crossbuild/.vagrant
 .vagrant/
@@ -33,8 +33,8 @@ before_script:
 # Lousy hack to disable use and testing of fallocate, which doesn't behave quite
 # as EnvPosixTest::AllocateTest expects within the Travis OpenVZ environment.
 script:
-  - if [[ "${TRAVIS_OS_NAME}" == 'linux' ]]; then OPT=-DTRAVIS CLANG_FORMAT_DIFF=/tmp/clang-format-diff.py make format || true; fi
-  - OPT=-DTRAVIS V=1 make -j4 check && OPT=-DTRAVIS V=1 make clean jclean rocksdbjava jtest && make clean && OPT="-DTRAVIS -DROCKSDB_LITE" V=1 make -j4 check
+  - if [[ "${TRAVIS_OS_NAME}" == 'linux' ]]; then OPT=-DTRAVIS CLANG_FORMAT_DIFF=/tmp/clang-format-diff.py make format; fi
+  - OPT=-DTRAVIS V=1 make -j4 check && OPT=-DTRAVIS V=1 make clean jclean rocksdbjava jtest

 notifications:
    email:
@@ -13,17 +13,11 @@
 #        cd build
 # 3. Run cmake to generate project files for Windows, add more options to enable required third-party libraries.
 #    See thirdparty.inc for more information.
-#        sample command: cmake -G "Visual Studio 12 Win64" -DGFLAGS=1 -DSNAPPY=1 -DJEMALLOC=1 ..
-#        OR for VS Studio 15 cmake -G "Visual Studio 14 Win64" -DGFLAGS=1 -DSNAPPY=1 -DJEMALLOC=1 ..
-# 4. Then build the project in debug mode (you may want to add /m[:<N>] flag to run msbuild in <N> parallel threads
-#                                          or simply /m ot use all avail cores)
-#        msbuild rocksdb.sln
-#
-#        rocksdb.sln build features exclusions of test only code in Release. If you build ALL_BUILD then everything
-#        will be attempted but test only code does not build in Release mode.
-#
-# 5. And release mode (/m[:<N>] is also supported)
-#        msbuild rocksdb.sln /p:Configuration=Release
+#        cmake -G "Visual Studio 12 Win64" .. <more options>
+# 4. Then build the project in debug mode (you may want to add /m:<N> flag to run msbuild in <N> parallel threads)
+#        msbuild ALL_BUILD.vcxproj
+# 5. And release mode (/m:<N> is also supported)
+#        msbuild ALL_BUILD.vcxproj /p:Configuration=Release
 #

 cmake_minimum_required(VERSION 2.6)
@@ -37,14 +31,7 @@ string(REGEX REPLACE "(..)/(..)/..(..).*" "\\1/\\2/\\3" DATE ${DATE})
 string(REGEX REPLACE "(..):(.....).*" " \\1:\\2" TIME ${TIME})
 string(CONCAT GIT_DATE_TIME ${DATE} ${TIME})

-find_package(Git)
-
-if (GIT_FOUND AND EXISTS "${CMAKE_CURRENT_SOURCE_DIR}/.git")
-    execute_process(COMMAND $ENV{COMSPEC} /C ${GIT_EXECUTABLE} -C ${CMAKE_CURRENT_SOURCE_DIR} rev-parse HEAD OUTPUT_VARIABLE GIT_SHA)
-else()
-    set(GIT_SHA 0)
-endif()
-
+execute_process(COMMAND $ENV{COMSPEC} " /C git rev-parse HEAD 2>nil" OUTPUT_VARIABLE GIT_SHA)
 string(REGEX REPLACE "[^0-9a-f]+" "" GIT_SHA ${GIT_SHA})

 set(BUILD_VERSION_CC ${CMAKE_CURRENT_SOURCE_DIR}/util/build_version.cc)
@@ -58,35 +45,19 @@ add_custom_command(OUTPUT ${BUILD_VERSION_CC}

 add_custom_target(GenerateBuildVersion DEPENDS ${BUILD_VERSION_CC})

-set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} /Zi /nologo  /EHsc /GS /Gd /GR /GF /fp:precise /Zc:wchar_t /Zc:forScope /errorReport:queue")
+set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} /Zi /nologo /W3 /WX /EHsc /GS /fp:precise /Zc:wchar_t /Zc:forScope /Gd /TP /errorReport:queue")
+set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} /FC /d2Zi+ /wd4018 /wd4100 /wd4101 /wd4127 /wd4189 /wd4200 /wd4244 /wd4267 /wd4296 /wd4305 /wd4307 /wd4309 /wd4512 /wd4701 /wd4702 /wd4800 /wd4804 /wd4996")

-set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} /FC /d2Zi+ /W3 /WX /wd4127 /wd4800 /wd4996 /wd4351")
+set(CMAKE_CXX_FLAGS_DEBUG "${CMAKE_CXX_FLAGS_DEBUG} /Od /RTC1 /Gm /MDd")
+set(CMAKE_CXX_FLAGS_RELEASE "${CMAKE_CXX_FLAGS_RELEASE} /O2 /Oi /Gm- /Gy /MD")

-# Used to run CI build and tests so we can run faster
-set(OPTIMIZE_DEBUG_DEFAULT 0)        # Debug build is unoptimized by default use -DOPTDBG=1 to optimize
-
-if(DEFINED OPTDBG)
-   set(OPTIMIZE_DEBUG ${OPTDBG})
-else()
-   set(OPTIMIZE_DEBUG ${OPTIMIZE_DEBUG_DEFAULT})
-endif()
-
-if((${OPTIMIZE_DEBUG} EQUAL 1))
-   message(STATUS "Debug optimization is enabled")
-   set(CMAKE_CXX_FLAGS_DEBUG "/Oxt /MDd")
-else()
-   set(CMAKE_CXX_FLAGS_DEBUG "${CMAKE_CXX_FLAGS_DEBUG} /Od /RTC1 /Gm /MDd")
-endif()
-
-set(CMAKE_CXX_FLAGS_RELEASE "${CMAKE_CXX_FLAGS_RELEASE} /Oxt /Zp8 /Gm- /Gy /MD")
-
-set(CMAKE_SHARED_LINKER_FLAGS "${CMAKE_SHARED_LINKER_FLAGS} /DEBUG")
 set(CMAKE_EXE_LINKER_FLAGS "${CMAKE_EXE_LINKER_FLAGS} /DEBUG")

 add_definitions(-DWIN32 -DOS_WIN -D_MBCS -DWIN64)

 include_directories(${PROJECT_SOURCE_DIR})
 include_directories(${PROJECT_SOURCE_DIR}/include)
+include_directories(${PROJECT_SOURCE_DIR}/port)
 include_directories(${PROJECT_SOURCE_DIR}/third-party/gtest-1.7.0/fused-src)

 set(ROCKSDB_LIBS rocksdblib${ARTIFACT_SUFFIX})
@@ -97,7 +68,6 @@ set(LIBS ${ROCKSDB_LIBS} ${THIRDPARTY_LIBS} ${SYSTEM_LIBS})

 add_subdirectory(third-party/gtest-1.7.0/fused-src/gtest)

-# Main library source code
 set(SOURCES
        db/builder.cc
        db/c.cc
@@ -145,9 +115,6 @@ set(SOURCES
        db/write_batch_base.cc
        db/write_controller.cc
        db/write_thread.cc
-        memtable/hash_cuckoo_rep.cc
-        memtable/hash_linklist_rep.cc
-        memtable/hash_skiplist_rep.cc
        port/stack_trace.cc
        port/win/env_win.cc
        port/win/port_win.cc
@@ -171,7 +138,6 @@ set(SOURCES
        table/get_context.cc
        table/iterator.cc
        table/merger.cc
-        table/sst_file_writer.cc
        table/meta_blocks.cc
        table/mock_table.cc
        table/plain_table_builder.cc
@@ -181,8 +147,6 @@ set(SOURCES
        table/plain_table_reader.cc
        table/table_properties.cc
        table/two_level_iterator.cc
-        tools/sst_dump_tool.cc
-        tools/dump/db_dump_tool.cc
        util/arena.cc
        util/auto_roll_logger.cc
        util/bloom.cc
@@ -191,10 +155,10 @@ set(SOURCES
        util/coding.cc
        util/compaction_job_stats_impl.cc
        util/comparator.cc
-        util/concurrent_arena.cc
        util/crc32c.cc
        util/db_info_dumper.cc
        util/delete_scheduler_impl.cc
+        util/db_test_util.cc
        util/dynamic_bloom.cc
        util/env.cc
        util/env_hdfs.cc
@@ -203,11 +167,14 @@ set(SOURCES
        util/file_reader_writer.cc
        util/filter_policy.cc
        util/hash.cc
+        util/hash_cuckoo_rep.cc
+        util/hash_linklist_rep.cc
+        util/hash_skiplist_rep.cc
        util/histogram.cc
        util/instrumented_mutex.cc
        util/iostats_context.cc
-        tools/ldb_cmd.cc
-        tools/ldb_tool.cc
+        util/ldb_cmd.cc
+        util/ldb_tool.cc
        util/logging.cc
        util/log_buffer.cc
        util/memenv.cc
@@ -217,14 +184,12 @@ set(SOURCES
        util/options.cc
        util/options_builder.cc
        util/options_helper.cc
-        util/options_parser.cc
-        util/options_sanity_check.cc
        util/perf_context.cc
        util/perf_level.cc
-        util/random.cc
        util/rate_limiter.cc
        util/skiplistrep.cc
        util/slice.cc
+        util/sst_dump_tool.cc
        util/statistics.cc
        util/status.cc
        util/status_message.cc
@@ -235,6 +200,7 @@ set(SOURCES
        util/thread_local.cc
        util/thread_status_impl.cc
        util/thread_status_updater.cc
+        util/thread_status_updater_debug.cc
        util/thread_status_util.cc
        util/thread_status_util_debug.cc
        util/vectorrep.cc
@@ -245,16 +211,13 @@ set(SOURCES
        utilities/document/document_db.cc
        utilities/document/json_document.cc
        utilities/document/json_document_builder.cc
-        utilities/env_mirror.cc
        utilities/flashcache/flashcache.cc
        utilities/geodb/geodb_impl.cc
        utilities/leveldb_options/leveldb_options.cc
-        utilities/memory/memory_util.cc
        utilities/merge_operators/string_append/stringappend.cc
        utilities/merge_operators/string_append/stringappend2.cc
        utilities/merge_operators/put.cc
        utilities/merge_operators/uint64add.cc
-        utilities/options/options_util.cc
        utilities/redis/redis_lists.cc
        utilities/spatialdb/spatial_db.cc
        utilities/table_properties_collectors/compact_on_deletion_collector.cc
@@ -271,13 +234,6 @@ set(SOURCES
        utilities/write_batch_with_index/write_batch_with_index_internal.cc
 )

-# For test util library that is build only in DEBUG mode
-# and linked to tests. Add test only code that is not #ifdefed for Release here.
-set(TESTUTIL_SOURCE
-    db/db_test_util.cc
-    util/thread_status_updater_debug.cc
-)
-
 add_library(rocksdblib${ARTIFACT_SUFFIX} ${SOURCES})
 set_target_properties(rocksdblib${ARTIFACT_SUFFIX} PROPERTIES COMPILE_FLAGS "/Fd${CMAKE_CFG_INTDIR}/rocksdblib${ARTIFACT_SUFFIX}.pdb")
 add_dependencies(rocksdblib${ARTIFACT_SUFFIX} GenerateBuildVersion)
@@ -292,7 +248,6 @@ set(APPS
        db/memtablerep_bench.cc
        table/table_reader_bench.cc
        tools/db_stress.cc
-        tools/write_stress.cc
        tools/db_repl_stress.cc
        tools/sst_dump.cc
        tools/dump/rocksdb_dump.cc
@@ -300,12 +255,10 @@ set(APPS
        util/cache_bench.cc
 )

-set(C_TESTS db/c_test.c)
-
 set(TESTS
+        db/c_test.c
        db/column_family_test.cc
        db/compact_files_test.cc
-        db/compaction_iterator_test.cc
        db/compaction_job_test.cc
        db/compaction_job_stats_test.cc
        db/compaction_picker_test.cc
@@ -322,21 +275,17 @@ set(TESTS
        db/db_universal_compaction_test.cc
        db/db_wal_test.cc
        db/db_tailing_iter_test.cc
-        db/db_table_properties_test.cc
        db/dbformat_test.cc
        db/deletefile_test.cc
        db/fault_injection_test.cc
        db/file_indexer_test.cc
        db/filename_test.cc
        db/flush_job_test.cc
-        db/inlineskiplist_test.cc
        db/listener_test.cc
        db/log_test.cc
-        db/manual_compaction_test.cc
        db/memtable_list_test.cc
        db/merge_test.cc
        db/merge_helper_test.cc
-        db/options_file_test.cc
        db/perf_context_test.cc
        db/plain_table_db_test.cc
        db/prefix_test.cc
@@ -358,9 +307,7 @@ set(TESTS
        table/merger_test.cc
        table/table_test.cc
        tools/db_sanity_test.cc
-        tools/ldb_cmd_test.cc
        tools/reduce_levels_test.cc
-        tools/sst_dump_test.cc
        util/arena_test.cc
        util/autovector_test.cc
        util/auto_roll_logger_test.cc
@@ -368,7 +315,6 @@ set(TESTS
        util/cache_test.cc
        util/coding_test.cc
        util/crc32c_test.cc
-        util/delete_scheduler_test.cc
        util/dynamic_bloom_test.cc
        util/env_test.cc
        util/event_logger_test.cc
@@ -376,11 +322,13 @@ set(TESTS
        util/file_reader_writer_test.cc
        util/heap_test.cc
        util/histogram_test.cc
+        util/manual_compaction_test.cc
        util/memenv_test.cc
        util/mock_env_test.cc
        util/options_test.cc
        util/rate_limiter_test.cc
        util/slice_transform_test.cc
+        util/sst_dump_test.cc
        util/thread_list_test.cc
        util/thread_local_test.cc
        utilities/backupable/backupable_db_test.cc
@@ -388,9 +336,7 @@ set(TESTS
        utilities/document/document_db_test.cc
        utilities/document/json_document_test.cc
        utilities/geodb/geodb_test.cc
-        utilities/memory/memory_test.cc
        utilities/merge_operators/string_append/stringappend_test.cc
-        utilities/options/options_util_test.cc
        utilities/redis/redis_lists_test.cc
        utilities/spatialdb/spatial_db_test.cc
        utilities/table_properties_collectors/compact_on_deletion_collector_test.cc
@@ -400,51 +346,12 @@ set(TESTS
        utilities/write_batch_with_index/write_batch_with_index_test.cc
 )

-set(EXES ${APPS})
+set(EXES ${APPS} ${TESTS})

 foreach(sourcefile ${EXES})
    string(REPLACE ".cc" "" exename ${sourcefile})
+    string(REPLACE ".c" "" exename ${exename})
    string(REGEX REPLACE "^((.+)/)+" "" exename ${exename})
    add_executable(${exename}${ARTIFACT_SUFFIX} ${sourcefile})
    target_link_libraries(${exename}${ARTIFACT_SUFFIX} ${LIBS})
 endforeach(sourcefile ${EXES})
-
-# test utilities are only build in debug
-set(TESTUTILLIB testutillib${ARTIFACT_SUFFIX})
-add_library(${TESTUTILLIB} STATIC ${TESTUTIL_SOURCE})
-set_target_properties(${TESTUTILLIB} PROPERTIES COMPILE_FLAGS "/Fd${CMAKE_CFG_INTDIR}/testutillib${ARTIFACT_SUFFIX}.pdb")
-set_target_properties(${TESTUTILLIB}
-      PROPERTIES EXCLUDE_FROM_DEFAULT_BUILD_RELEASE 1
-      EXCLUDE_FROM_DEFAULT_BUILD_MINRELEASE 1
-      EXCLUDE_FROM_DEFAULT_BUILD_RELWITHDEBINFO 1
-      )
-
-# Tests are excluded from Release builds
-set(TEST_EXES ${TESTS})
-
-foreach(sourcefile ${TEST_EXES})
-    string(REPLACE ".cc" "" exename ${sourcefile})
-    string(REGEX REPLACE "^((.+)/)+" "" exename ${exename})
-    add_executable(${exename}${ARTIFACT_SUFFIX} ${sourcefile})
-    set_target_properties(${exename}${ARTIFACT_SUFFIX}
-      PROPERTIES EXCLUDE_FROM_DEFAULT_BUILD_RELEASE 1
-      EXCLUDE_FROM_DEFAULT_BUILD_MINRELEASE 1
-      EXCLUDE_FROM_DEFAULT_BUILD_RELWITHDEBINFO 1
-      )
-    target_link_libraries(${exename}${ARTIFACT_SUFFIX} ${LIBS} testutillib${ARTIFACT_SUFFIX})
-endforeach(sourcefile ${TEST_EXES})
-
-# C executables must link to a shared object
-set(C_TEST_EXES ${C_TESTS})
-
-foreach(sourcefile ${C_TEST_EXES})
-    string(REPLACE ".c" "" exename ${sourcefile})
-    string(REGEX REPLACE "^((.+)/)+" "" exename ${exename})
-    add_executable(${exename}${ARTIFACT_SUFFIX} ${sourcefile})
-    set_target_properties(${exename}${ARTIFACT_SUFFIX}
-      PROPERTIES EXCLUDE_FROM_DEFAULT_BUILD_RELEASE 1
-      EXCLUDE_FROM_DEFAULT_BUILD_MINRELEASE 1
-      EXCLUDE_FROM_DEFAULT_BUILD_RELWITHDEBINFO 1
-      )
-    target_link_libraries(${exename}${ARTIFACT_SUFFIX} rocksdb${ARTIFACT_SUFFIX} testutillib${ARTIFACT_SUFFIX})
-endforeach(sourcefile ${C_TEST_EXES})
@@ -1,58 +1,9 @@
 # Rocksdb Change Log

-## 4.4.1 (2/18/2016)
-
-## 4.4.0 (1/14/2016)
-### Public API Changes
-* Change names in CompactionPri and add a new one.
-* Deprecate options.soft_rate_limit and add options.soft_pending_compaction_bytes_limit.
-* If options.max_write_buffer_number > 3, writes will be slowed down when writing to the last write buffer to delay a full stop.
-* Introduce CompactionJobInfo::compaction_reason, this field include the reason to trigger the compaction.
-* After slow down is triggered, if estimated pending compaction bytes keep increasing, slowdown more.
-* Increase default options.delayed_write_rate to 2MB/s.
-* Added a new parameter --path to ldb tool. --path accepts the name of either MANIFEST, SST or a WAL file. Either --db or --path can be used when calling ldb.
-
-## 4.3.0 (12/8/2015)
-### New Features
-* CompactionFilter has new member function called IgnoreSnapshots which allows CompactionFilter to be called even if there are snapshots later than the key.
-* RocksDB will now persist options under the same directory as the RocksDB database on successful DB::Open, CreateColumnFamily, DropColumnFamily, and SetOptions.
-* Introduce LoadLatestOptions() in rocksdb/utilities/options_util.h.  This function can construct the latest DBOptions / ColumnFamilyOptions used by the specified RocksDB intance.
-* Introduce CheckOptionsCompatibility() in rocksdb/utilities/options_util.h.  This function checks whether the input set of options is able to open the specified DB successfully.
-
-### Public API Changes
-* When options.db_write_buffer_size triggers, only the column family with the largest column family size will be flushed, not all the column families.
-
-## 4.2.0 (11/9/2015)
-### New Features
-* Introduce CreateLoggerFromOptions(), this function create a Logger for provided DBOptions.
-* Add GetAggregatedIntProperty(), which returns the sum of the GetIntProperty of all the column families.
-* Add MemoryUtil in rocksdb/utilities/memory.h.  It currently offers a way to get the memory usage by type from a list rocksdb instances.
-
-### Public API Changes
-* CompactionFilter::Context includes information of Column Family ID
-* The need-compaction hint given by TablePropertiesCollector::NeedCompact() will be persistent and recoverable after DB recovery. This introduces a breaking format change. If you use this experimental feature, including NewCompactOnDeletionCollectorFactory() in the new version, you may not be able to directly downgrade the DB back to version 4.0 or lower.
-* TablePropertiesCollectorFactory::CreateTablePropertiesCollector() now takes an option Context, containing the information of column family ID for the file being written.
-* Remove DefaultCompactionFilterFactory.
-
-
-## 4.1.0 (10/8/2015)
-### New Features
-* Added single delete operation as a more efficient way to delete keys that have not been overwritten.
-* Added experimental AddFile() to DB interface that allow users to add files created by SstFileWriter into an empty Database, see include/rocksdb/sst_file_writer.h and DB::AddFile() for more info.
-* Added support for opening SST files with .ldb suffix which enables opening LevelDB databases.
-* CompactionFilter now supports filtering of merge operands and merge results.
-
-### Public API Changes
-* Added SingleDelete() to the DB interface.
-* Added AddFile() to DB interface.
-* Added SstFileWriter class.
-* CompactionFilter has a new method FilterMergeOperand() that RocksDB applies to every merge operand during compaction to decide whether to filter the operand.
-* We removed CompactionFilterV2 interfaces from include/rocksdb/compaction_filter.h. The functionality was deprecated already in version 3.13.
-
 ## 4.0.0 (9/9/2015)
 ### New Features
 * Added support for transactions.  See include/rocksdb/utilities/transaction.h for more info.
-* DB::GetProperty() now accepts "rocksdb.aggregated-table-properties" and "rocksdb.aggregated-table-properties-at-levelN", in which case it returns aggregated table properties of the target column family, or the aggregated table properties of the specified level N if the "at-level" version is used.
+* DB::GetProperty() now accept "rocksdb.aggregated-table-properties" and "rocksdb.aggregated-table-properties-at-levelN", in which case it returns aggregated table properties of the target column family, or the aggregated table properties of the specified level N if the "at-level" version is used.
 * Add compression option kZSTDNotFinalCompression for people to experiment ZSTD although its format is not finalized.
 * We removed the need for LATEST_BACKUP file in BackupEngine. We still keep writing it when we create new backups (because of backward compatibility), but we don't read it anymore.

@@ -222,7 +173,7 @@
 * Support Multiple DB paths in universal style compactions
 * Add feature of storing plain table index and bloom filter in SST file.
 * CompactRange() will never output compacted files to level 0. This used to be the case when all the compaction input files were at level 0.
-* Added iterate_upper_bound to define the extent upto which the forward iterator will return entries. This will prevent iterating over delete markers and overwritten entries for edge cases where you want to break out the iterator anyways. This may improve performance in case there are a large number of delete markers or overwritten entries.
+* Added iterate_upper_bound to define the extent upto which the forward iterator will return entries. This will prevent iterating over delete markers and overwritten entries for edge cases where you want to break out the iterator anyways. This may improve perfomance in case there are a large number of delete markers or overwritten entries.

 ### Public API changes
 * DBOptions.db_paths now is a vector of a DBPath structure which indicates both of path and target size
@@ -1,24 +1,19 @@
 ## Compilation

-**Important**: If you plan to run RocksDB in production, don't compile using default 
-`make` or `make all`. That will compile RocksDB in debug mode, which is much slower
-than release mode.
-
 RocksDB's library should be able to compile without any dependency installed,
 although we recommend installing some compression libraries (see below).
 We do depend on newer gcc/clang with C++11 support.

 There are few options when compiling RocksDB:

-* [recommended] `make static_lib` will compile librocksdb.a, RocksDB static library. Compiles static library in release mode.
+* [recommended] `make static_lib` will compile librocksdb.a, RocksDB static library.

-* `make shared_lib` will compile librocksdb.so, RocksDB shared library. Compiles shared library in release mode.
+* `make shared_lib` will compile librocksdb.so, RocksDB shared library.

-* `make check` will compile and run all the unit tests. `make check` will compile RocksDB in debug mode.
+* `make check` will compile and run all the unit tests

 * `make all` will compile our static library, and all our tools and unit tests. Our tools
-depend on gflags. You will need to have gflags installed to run `make all`. This will compile RocksDB in debug mode. Don't
-use binaries compiled by `make all` in production.
+depend on gflags. You will need to have gflags installed to run `make all`.

 * By default the binary we produce is optimized for the platform you're compiling on
 (-march=native). If you want to build a portable binary, add 'PORTABLE=1' before
@@ -87,5 +82,4 @@ your make commands, like this: `PORTABLE=1 make static_lib`
  * 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**:
-  * For building with MS Visual Studio 13 you will need Update 4 installed.
  * Read and follow the instructions at CMakeLists.txt
@@ -33,22 +33,12 @@ quoted_perl_command = $(subst ','\'',$(perl_command))
 # with debug level 0. To compile with level 0, run `make shared_lib`,
 # `make install-shared`, `make static_lib`, `make install-static` or
 # `make install`
-
-# Set the default DEBUG_LEVEL to 1
-DEBUG_LEVEL?=1
+DEBUG_LEVEL=1

 ifeq ($(MAKECMDGOALS),dbg)
 	DEBUG_LEVEL=2
 endif

-ifeq ($(MAKECMDGOALS),clean)
-	DEBUG_LEVEL=0
-endif
-
-ifeq ($(MAKECMDGOALS),release)
-	DEBUG_LEVEL=0
-endif
-
 ifeq ($(MAKECMDGOALS),shared_lib)
 	DEBUG_LEVEL=0
 endif
@@ -73,14 +63,6 @@ ifeq ($(MAKECMDGOALS),rocksdbjavastatic)
 	DEBUG_LEVEL=0
 endif

-ifeq ($(MAKECMDGOALS),rocksdbjavastaticrelease)
-	DEBUG_LEVEL=0
-endif
-
-ifeq ($(MAKECMDGOALS),rocksdbjavastaticpublish)
-	DEBUG_LEVEL=0
-endif
-
 # compile with -O2 if debug level is not 2
 ifneq ($(DEBUG_LEVEL), 2)
 OPT += -O2 -fno-omit-frame-pointer
@@ -94,8 +76,6 @@ endif
 ifeq ($(DEBUG_LEVEL),0)
 OPT += -DNDEBUG
 DISABLE_WARNING_AS_ERROR=1
-else
-$(warning Warning: Compiling in debug mode. Don't use the resulting binary in production)
 endif

 #-----------------------------------------------
@@ -187,6 +167,10 @@ default: all
 WARNING_FLAGS = -W -Wextra -Wall -Wsign-compare -Wshadow \
  -Wno-unused-parameter

+ifndef DISABLE_WARNING_AS_ERROR
+	WARNING_FLAGS += -Werror
+endif
+
 CFLAGS += $(WARNING_FLAGS) -I. -I./include $(PLATFORM_CCFLAGS) $(OPT)
 CXXFLAGS += $(WARNING_FLAGS) -I. -I./include $(PLATFORM_CXXFLAGS) $(OPT) -Woverloaded-virtual -Wnon-virtual-dtor -Wno-missing-field-initializers

@@ -222,13 +206,13 @@ util/build_version.cc: FORCE
 	else mv -f $@-t $@; fi

 LIBOBJECTS = $(LIB_SOURCES:.cc=.o)
-LIBOBJECTS += $(TOOL_SOURCES:.cc=.o)
 MOCKOBJECTS = $(MOCK_SOURCES:.cc=.o)

 GTEST = $(GTEST_DIR)/gtest/gtest-all.o
 TESTUTIL = ./util/testutil.o
 TESTHARNESS = ./util/testharness.o $(TESTUTIL) $(MOCKOBJECTS) $(GTEST)
 VALGRIND_ERROR = 2
+VALGRIND_DIR = build_tools/VALGRIND_LOGS
 VALGRIND_VER := $(join $(VALGRIND_VER),valgrind)

 VALGRIND_OPTS = --error-exitcode=$(VALGRIND_ERROR) --leak-check=full
@@ -244,7 +228,6 @@ TESTS = \
 	db_tailing_iter_test \
 	db_universal_compaction_test \
 	db_wal_test \
-	db_table_properties_test \
 	block_hash_index_test \
 	autovector_test \
 	column_family_test \
@@ -270,17 +253,14 @@ TESTS = \
 	block_based_filter_block_test \
 	full_filter_block_test \
 	histogram_test \
-	inlineskiplist_test \
 	log_test \
 	manual_compaction_test \
 	memenv_test \
 	mock_env_test \
 	memtable_list_test \
 	merge_helper_test \
-	memory_test \
 	merge_test \
 	merger_test \
-	options_file_test \
 	redis_test \
 	reduce_levels_test \
 	plain_table_db_test \
@@ -308,7 +288,6 @@ TESTS = \
 	rate_limiter_test \
 	delete_scheduler_test \
 	options_test \
-	options_util_test \
 	event_logger_test \
 	cuckoo_table_builder_test \
 	cuckoo_table_reader_test \
@@ -316,7 +295,6 @@ TESTS = \
 	flush_job_test \
 	wal_manager_test \
 	listener_test \
-	compaction_iterator_test \
 	compaction_job_test \
 	thread_list_test \
 	sst_dump_test \
@@ -327,8 +305,7 @@ TESTS = \
 	heap_test \
 	compact_on_deletion_collector_test \
 	compaction_job_stats_test \
-	transaction_test \
-	ldb_cmd_test
+	transaction_test

 SUBSET :=  $(shell echo $(TESTS) |sed s/^.*$(ROCKSDBTESTS_START)/$(ROCKSDBTESTS_START)/)

@@ -336,23 +313,17 @@ TOOLS = \
 	sst_dump \
 	db_sanity_test \
 	db_stress \
-	write_stress \
 	ldb \
 	db_repl_stress \
 	rocksdb_dump \
 	rocksdb_undump

-# TODO: add back forward_iterator_bench, after making it build in all environemnts.
 BENCHMARKS = db_bench table_reader_bench cache_bench memtablerep_bench

-# if user didn't config LIBNAME, set the default
+# The library name is configurable since we are maintaining libraries of both
+# debug/release mode.
 ifeq ($(LIBNAME),)
-# we should only run rocksdb in production with DEBUG_LEVEL 0
-ifeq ($(DEBUG_LEVEL),0)
        LIBNAME=librocksdb
-else
-        LIBNAME=librocksdb_debug
-endif
 endif
 LIBRARY = ${LIBNAME}.a

@@ -398,7 +369,7 @@ $(SHARED3): $(SHARED4)
 endif

 $(SHARED4):
-	$(CXX) $(PLATFORM_SHARED_LDFLAGS)$(SHARED3) $(CXXFLAGS) $(PLATFORM_SHARED_CFLAGS) $(LIB_SOURCES) $(TOOL_SOURCES) \
+	$(CXX) $(PLATFORM_SHARED_LDFLAGS)$(SHARED3) $(CXXFLAGS) $(PLATFORM_SHARED_CFLAGS) $(LIB_SOURCES) \
 		$(LDFLAGS) -o $@

 endif  # PLATFORM_SHARED_EXT
@@ -422,7 +393,7 @@ dbg: $(LIBRARY) $(BENCHMARKS) tools $(TESTS)
 # creates static library and programs
 release:
 	$(MAKE) clean
-	DEBUG_LEVEL=0 $(MAKE) static_lib tools db_bench
+	OPT="-DNDEBUG -O2" $(MAKE) static_lib tools db_bench

 coverage:
 	$(MAKE) clean
@@ -571,10 +542,8 @@ check: all
 	      echo "===== Running $$t"; ./$$t || exit 1; done;          \
 	fi
 	rm -rf $(TMPD)
-ifeq ($(filter -DROCKSDB_LITE,$(OPT)),)
 	python tools/ldb_test.py
 	sh tools/rocksdb_dump_test.sh
-endif

 check_some: $(SUBSET) ldb_tests
 	for t in $(SUBSET); do echo "===== Running $$t"; ./$$t || exit 1; done
@@ -586,12 +555,12 @@ ldb_tests: ldb
 crash_test: whitebox_crash_test blackbox_crash_test

 blackbox_crash_test: db_stress
-	python -u tools/db_crashtest.py --simple blackbox 
-	python -u tools/db_crashtest.py blackbox
+	python -u tools/db_crashtest.py -s
+	python -u tools/db_crashtest.py

 whitebox_crash_test: db_stress
-	python -u tools/db_crashtest.py --simple whitebox
-	python -u tools/db_crashtest.py whitebox
+	python -u tools/db_crashtest2.py -s
+	python -u tools/db_crashtest2.py

 asan_check:
 	$(MAKE) clean
@@ -604,12 +573,20 @@ asan_crash_test:
 	$(MAKE) clean

 valgrind_check: $(TESTS)
-	for t in $(filter-out %skiplist_test,$(TESTS)); do \
+	mkdir -p $(VALGRIND_DIR)
+	echo TESTS THAT HAVE VALGRIND ERRORS > $(VALGRIND_DIR)/valgrind_failed_tests; \
+	echo TIMES in seconds TAKEN BY TESTS ON VALGRIND > $(VALGRIND_DIR)/valgrind_tests_times; \
+	for t in $(filter-out skiplist_test,$(TESTS)); do \
+		stime=`date '+%s'`; \
 		$(VALGRIND_VER) $(VALGRIND_OPTS) ./$$t; \
 		ret_code=$$?; \
-		if [ $$ret_code -ne 0 ]; then \
+		if [ $$ret_code -eq $(VALGRIND_ERROR) ] ; then \
+			echo $$t >> $(VALGRIND_DIR)/valgrind_failed_tests; \
+		elif [ $$ret_code -ne 0 ]; then \
 			exit $$ret_code; \
 		fi; \
+		etime=`date '+%s'`; \
+		echo $$t $$((etime - stime)) >> $(VALGRIND_DIR)/valgrind_tests_times; \
 	done

 analyze: clean
@@ -622,22 +599,14 @@ CLEAN_FILES += unity.cc
 unity.cc: Makefile
 	rm -f $@ $@-t
 	for source_file in $(LIB_SOURCES); do \
-		echo "#include \"$$source_file\"" >> $@-t; \
+		echo "#include <$$source_file>" >> $@-t; \
 	done
+	echo 'int main(int argc, char** argv){ return 0; }' >> $@-t
 	chmod a=r $@-t
 	mv $@-t $@

-unity.a: unity.o
-	$(AM_V_AR)rm -f $@
-	$(AM_V_at)$(AR) $(ARFLAGS) $@ unity.o
-
-# try compiling db_test with unity
-unity_test: db/db_test.o db/db_test_util.o $(TESTHARNESS) unity.a
+unity: unity.o
 	$(AM_LINK)
-	./unity_test
-
-rocksdb.h rocksdb.cc: build_tools/amalgamate.py Makefile $(LIB_SOURCES) unity.cc
-	build_tools/amalgamate.py -I. -i./include unity.cc -x include/rocksdb/c.h -H rocksdb.h -o rocksdb.cc

 clean:
 	rm -f $(BENCHMARKS) $(TOOLS) $(TESTS) $(LIBRARY) $(SHARED)
@@ -678,9 +647,6 @@ block_hash_index_test: table/block_hash_index_test.o $(LIBOBJECTS) $(TESTHARNESS
 db_stress: tools/db_stress.o $(LIBOBJECTS) $(TESTUTIL)
 	$(AM_LINK)

-write_stress: tools/write_stress.o $(LIBOBJECTS) $(TESTUTIL)
-	$(AM_LINK)
-
 db_sanity_test: tools/db_sanity_test.o $(LIBOBJECTS) $(TESTUTIL)
 	$(AM_LINK)

@@ -735,37 +701,34 @@ crc32c_test: util/crc32c_test.o $(LIBOBJECTS) $(TESTHARNESS)
 slice_transform_test: util/slice_transform_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

-db_test: db/db_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
+db_test: db/db_test.o util/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

-db_log_iter_test: db/db_log_iter_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
+db_log_iter_test: db/db_log_iter_test.o util/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

-db_compaction_filter_test: db/db_compaction_filter_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
+db_compaction_filter_test: db/db_compaction_filter_test.o util/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

-db_compaction_test: db/db_compaction_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
+db_compaction_test: db/db_compaction_test.o util/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

-db_dynamic_level_test: db/db_dynamic_level_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
+db_dynamic_level_test: db/db_dynamic_level_test.o util/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

-db_inplace_update_test: db/db_inplace_update_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
+db_inplace_update_test: db/db_inplace_update_test.o util/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

-db_tailing_iter_test: db/db_tailing_iter_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
+db_tailing_iter_test: db/db_tailing_iter_test.o util/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

 db_iter_test: db/db_iter_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

-db_universal_compaction_test: db/db_universal_compaction_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
+db_universal_compaction_test: db/db_universal_compaction_test.o util/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

-db_wal_test: db/db_wal_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
-	$(AM_LINK)
-
-db_table_properties_test: db/db_table_properties_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
+db_wal_test: db/db_wal_test.o util/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

 log_write_bench: util/log_write_bench.o $(LIBOBJECTS) $(TESTHARNESS)
@@ -801,9 +764,6 @@ json_document_test: utilities/document/json_document_test.o $(LIBOBJECTS) $(TEST
 spatial_db_test: utilities/spatialdb/spatial_db_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

-env_mirror_test: utilities/env_mirror_test.o $(LIBOBJECTS) $(TESTHARNESS)
-	$(AM_LINK)
-
 ttl_test: utilities/ttl/ttl_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

@@ -813,9 +773,6 @@ write_batch_with_index_test: utilities/write_batch_with_index/write_batch_with_i
 flush_job_test: db/flush_job_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

-compaction_iterator_test: db/compaction_iterator_test.o $(LIBOBJECTS) $(TESTHARNESS)
-	$(AM_LINK)
-
 compaction_job_test: db/compaction_job_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

@@ -864,9 +821,6 @@ table_test: table/table_test.o $(LIBOBJECTS) $(TESTHARNESS)
 block_test: table/block_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

-inlineskiplist_test: db/inlineskiplist_test.o $(LIBOBJECTS) $(TESTHARNESS)
-	$(AM_LINK)
-
 skiplist_test: db/skiplist_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

@@ -897,18 +851,12 @@ write_controller_test: db/write_controller_test.o $(LIBOBJECTS) $(TESTHARNESS)
 merge_helper_test: db/merge_helper_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

-memory_test: utilities/memory/memory_test.o $(LIBOBJECTS) $(TESTHARNESS)
-	$(AM_LINK)
-
 merge_test: db/merge_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

 merger_test: table/merger_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

-options_file_test: db/options_file_test.o $(LIBOBJECTS) $(TESTHARNESS)
-	$(AM_LINK)
-
 deletefile_test: db/deletefile_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

@@ -930,7 +878,7 @@ cuckoo_table_reader_test: table/cuckoo_table_reader_test.o $(LIBOBJECTS) $(TESTH
 cuckoo_table_db_test: db/cuckoo_table_db_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

-listener_test: db/listener_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
+listener_test: db/listener_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

 thread_list_test: util/thread_list_test.o $(LIBOBJECTS) $(TESTHARNESS)
@@ -942,13 +890,10 @@ compact_files_test: db/compact_files_test.o $(LIBOBJECTS) $(TESTHARNESS)
 options_test: util/options_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

-options_util_test: utilities/options/options_util_test.o $(LIBOBJECTS) $(TESTHARNESS)
-	$(AM_LINK)
-
 event_logger_test: util/event_logger_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

-sst_dump_test: tools/sst_dump_test.o $(LIBOBJECTS) $(TESTHARNESS)
+sst_dump_test: util/sst_dump_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

 memenv_test : util/memenv_test.o $(LIBOBJECTS) $(TESTHARNESS)
@@ -960,7 +905,7 @@ optimistic_transaction_test: utilities/transactions/optimistic_transaction_test.
 mock_env_test : util/mock_env_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

-manual_compaction_test: db/manual_compaction_test.o $(LIBOBJECTS) $(TESTHARNESS)
+manual_compaction_test: util/manual_compaction_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

 filelock_test: util/filelock_test.o $(LIBOBJECTS) $(TESTHARNESS)
@@ -984,9 +929,6 @@ transaction_test: utilities/transactions/transaction_test.o $(LIBOBJECTS) $(TEST
 sst_dump: tools/sst_dump.o $(LIBOBJECTS)
 	$(AM_LINK)

-ldb_cmd_test: tools/ldb_cmd_test.o $(LIBOBJECTS) $(TESTHARNESS)
-	$(AM_LINK)
-
 ldb: tools/ldb.o $(LIBOBJECTS)
 	$(AM_LINK)

@@ -1125,12 +1067,12 @@ $(java_libobjects): jl/%.o: %.cc
 	$(AM_V_CC)mkdir -p $(@D) && $(CXX) $(CXXFLAGS) -fPIC -c $< -o $@ $(COVERAGEFLAGS)

 rocksdbjava: $(java_libobjects)
-	$(AM_V_GEN)cd java;$(MAKE) javalib;
-	$(AM_V_at)rm -f ./java/target/$(ROCKSDBJNILIB)
-	$(AM_V_at)$(CXX) $(CXXFLAGS) -I./java/. $(JAVA_INCLUDE) -shared -fPIC -o ./java/target/$(ROCKSDBJNILIB) $(JNI_NATIVE_SOURCES) $(java_libobjects) $(JAVA_LDFLAGS) $(COVERAGEFLAGS)
-	$(AM_V_at)cd java;jar -cf target/$(ROCKSDB_JAR) HISTORY*.md
-	$(AM_V_at)cd java/target;jar -uf $(ROCKSDB_JAR) $(ROCKSDBJNILIB)
-	$(AM_V_at)cd java/target/classes;jar -uf ../$(ROCKSDB_JAR) org/rocksdb/*.class org/rocksdb/util/*.class
+	cd java;$(MAKE) javalib;
+	rm -f ./java/target/$(ROCKSDBJNILIB)
+	$(CXX) $(CXXFLAGS) -I./java/. $(JAVA_INCLUDE) -shared -fPIC -o ./java/target/$(ROCKSDBJNILIB) $(JNI_NATIVE_SOURCES) $(java_libobjects) $(JAVA_LDFLAGS) $(COVERAGEFLAGS)
+	cd java;jar -cf target/$(ROCKSDB_JAR) HISTORY*.md
+	cd java/target;jar -uf $(ROCKSDB_JAR) $(ROCKSDBJNILIB)
+	cd java/target/classes;jar -uf ../$(ROCKSDB_JAR) org/rocksdb/*.class org/rocksdb/util/*.class

 jclean:
 	cd java;$(MAKE) clean;
@@ -37,12 +37,3 @@ Check out our RocksDB Protobuf merge operator at: https://github.com/vladb38/roc

 ## Santanader UK/Cloudera Profession Services
 Check out their blog post: http://blog.cloudera.com/blog/2015/08/inside-santanders-near-real-time-data-ingest-architecture/
-
-## Airbnb
-Airbnb is using RocksDB as a storage engine for their personalized search service. You can learn more about it here: https://www.youtube.com/watch?v=ASQ6XMtogMs
-
-## Pinterest
-Pinterest's Object Retrieval System uses RocksDB for storage: https://www.youtube.com/watch?v=MtFEVEs_2Vo
-
-## Smyte
-[Smyte](https://www.smyte.com/) uses RocksDB as the storage layer for their core key-value storage, high-performance counters and time-windowed HyperLogLog services.
@@ -2,14 +2,9 @@ version: 1.0.{build}
 before_build:
 - md %APPVEYOR_BUILD_FOLDER%\build
 - cd %APPVEYOR_BUILD_FOLDER%\build
- cmake -G "Visual Studio 12 Win64" -DOPTDBG=1 ..
+- cmake -G "Visual Studio 12 Win64" ..
 - cd ..
 build:
-  project: build\rocksdb.sln
-  parallel: true
+  project: build\ALL_BUILD.vcxproj
  verbosity: minimal
-test:
-test_script:
- ps: build_tools\run_ci_db_test.ps1 -EnableRerun -Run db_test -Exclude DBTest.Randomized,DBTest.FileCreationRandomFailure -Concurrency 18
- ps: build_tools\run_ci_db_test.ps1 -Run env_test -Concurrency 1
-
+test: off
@@ -88,9 +88,6 @@ class FbcodeCppLinter extends ArcanistLinter {
  }

  private function getCppLintOutput($path) {
-     if (!array_key_exists($path, $this->rawLintOutput)) {
-       return array();
-     }
    list($output) = $this->rawLintOutput[$path];

    $msgs = array();
@@ -1,110 +0,0 @@
-#!/usr/bin/python
-
-# amalgamate.py creates an amalgamation from a unity build.
-# It can be run with either Python 2 or 3.
-# An amalgamation consists of a header that includes the contents of all public
-# headers and a source file that includes the contents of all source files and
-# private headers.
-#
-# This script works by starting with the unity build file and recursively expanding
-# #include directives. If the #include is found in a public include directory,
-# that header is expanded into the amalgamation header.
-#
-# A particular header is only expanded once, so this script will
-# break if there are multiple inclusions of the same header that are expected to
-# expand differently. Similarly, this type of code causes issues:
-#
-# #ifdef FOO
-#   #include "bar.h"
-#   // code here
-# #else
-#   #include "bar.h"            // oops, doesn't get expanded
-#   // different code here
-# #endif
-#
-# The solution is to move the include out of the #ifdef.
-
-from __future__ import print_function
-
-import argparse
-from os import path
-import re
-import sys
-
-include_re = re.compile('^[ \t]*#include[ \t]+"(.*)"[ \t]*$')
-included = set()
-excluded = set()
-
-def find_header(name, abs_path, include_paths):
-    samedir = path.join(path.dirname(abs_path), name)
-    if path.exists(samedir):
-        return samedir
-    for include_path in include_paths:
-        include_path = path.join(include_path, name)
-        if path.exists(include_path):
-            return include_path
-    return None
-
-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)
-    return True
-
-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:
-            filename = m.groups()[0]
-            # first check private headers
-            include_path = find_header(filename, abs_path, include_paths)
-            if include_path:
-                if include_path in excluded:
-                    source_out.write(text)
-                    expanded = False
-                else:
-                    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)
-                if include_path:
-                    # found public header
-                    expanded = False
-                    if include_path in excluded:
-                        source_out.write(text)
-                    else:
-                        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))
-
-            if expanded:
-                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.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("-o", dest="source_out", help="output C++ 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 []))
-    public_include_paths = list(map(path.abspath, args.public_include_paths or []))
-    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:
-        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)
-
-if __name__ == "__main__":
-    main()
@@ -45,14 +45,19 @@ fi
 # we depend on C++11
 PLATFORM_CXXFLAGS="-std=c++11"
 # we currently depend on POSIX platform
-COMMON_FLAGS="-DROCKSDB_PLATFORM_POSIX -DROCKSDB_LIB_IO_POSIX"
+COMMON_FLAGS="-DROCKSDB_PLATFORM_POSIX"

 # Default to fbcode gcc on internal fb machines
 if [ -z "$ROCKSDB_NO_FBCODE" -a -d /mnt/gvfs/third-party ]; then
    FBCODE_BUILD="true"
    # If we're compiling with TSAN we need pic build
    PIC_BUILD=$COMPILE_WITH_TSAN
-    source "$PWD/build_tools/fbcode_config.sh"
+    if [ -z "$ROCKSDB_FBCODE_BUILD_WITH_481" ]; then
+      source "$PWD/build_tools/fbcode_config.sh"
+    else
+      # we need this to build with MySQL. Don't use for other purposes.
+      source "$PWD/build_tools/fbcode_config4.8.1.sh"
+    fi
 fi

 # Delete existing output, if it exists
@@ -210,7 +215,7 @@ EOF
    fi

    # Test whether gflags library is installed
-    # http://gflags.github.io/gflags/
+    # http://code.google.com/p/gflags/
    # check if the namespace is gflags
    $CXX $CFLAGS -x c++ - -o /dev/null 2>/dev/null << EOF
      #include <gflags/gflags.h>
@@ -1,19 +0,0 @@
-# Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved.
-GCC_BASE=/mnt/gvfs/third-party2/gcc/7331085db891a2ef4a88a48a751d834e8d68f4cb/7.x/centos7-native/b2ef2b6
-CLANG_BASE=/mnt/gvfs/third-party2/llvm-fb/963d9aeda70cc4779885b1277484fe7544a04e3e/9.0.0/platform007/9e92d53/
-LIBGCC_BASE=/mnt/gvfs/third-party2/libgcc/6ace84e956873d53638c738b6f65f3f469cca74c/7.x/platform007/5620abc
-GLIBC_BASE=/mnt/gvfs/third-party2/glibc/192b0f42d63dcf6210d6ceae387b49af049e6e0c/2.26/platform007/f259413
-SNAPPY_BASE=/mnt/gvfs/third-party2/snappy/7f9bdaada18f59bc27ec2b0871eb8a6144343aef/1.1.3/platform007/ca4da3d
-ZLIB_BASE=/mnt/gvfs/third-party2/zlib/2d9f0b9a4274cc21f61272a9e89bdb859bce8f1f/1.2.8/platform007/ca4da3d
-BZIP2_BASE=/mnt/gvfs/third-party2/bzip2/dc49a21c5fceec6456a7a28a94dcd16690af1337/1.0.6/platform007/ca4da3d
-LZ4_BASE=/mnt/gvfs/third-party2/lz4/0f607f8fc442ea7d6b876931b1898bb573d5e5da/1.9.1/platform007/ca4da3d
-ZSTD_BASE=/mnt/gvfs/third-party2/zstd/ca22bc441a4eb709e9e0b1f9fec9750fed7b31c5/1.4.x/platform007/15a3614
-GFLAGS_BASE=/mnt/gvfs/third-party2/gflags/0b9929d2588991c65a57168bf88aff2db87c5d48/2.2.0/platform007/ca4da3d
-JEMALLOC_BASE=/mnt/gvfs/third-party2/jemalloc/c26f08f47ac35fc31da2633b7da92d6b863246eb/master/platform007/c26c002
-NUMA_BASE=/mnt/gvfs/third-party2/numa/3f3fb57a5ccc5fd21c66416c0b83e0aa76a05376/2.0.11/platform007/ca4da3d
-LIBUNWIND_BASE=/mnt/gvfs/third-party2/libunwind/40c73d874898b386a71847f1b99115d93822d11f/1.4/platform007/6f3e0a9
-TBB_BASE=/mnt/gvfs/third-party2/tbb/4ce8e8dba77cdbd81b75d6f0c32fd7a1b76a11ec/2018_U5/platform007/ca4da3d
-KERNEL_HEADERS_BASE=/mnt/gvfs/third-party2/kernel-headers/fb251ecd2f5ae16f8671f7014c246e52a748fe0b/fb/platform007/da39a3e
-BINUTILS_BASE=/mnt/gvfs/third-party2/binutils/ab9f09bba370e7066cafd4eb59752db93f2e8312/2.29.1/platform007/15a3614
-VALGRIND_BASE=/mnt/gvfs/third-party2/valgrind/d42d152a15636529b0861ec493927200ebebca8e/3.15.0/platform007/ca4da3d
-LUA_BASE=/mnt/gvfs/third-party2/lua/f0cd714433206d5139df61659eb7b28b1dea6683/5.3.4/platform007/5007832
@@ -1,16 +0,0 @@
-GCC_BASE=/mnt/gvfs/third-party2/gcc/ebc96bc2fb751b5a0300b8d91a95bdf24ac1d88b/4.8.1/centos6-native/cc6c9dc
-CLANG_BASE=/mnt/vol/engshare/fbcode/third-party2/clang/3.7.1/centos6-native/9d9ecb9/
-LIBGCC_BASE=/mnt/gvfs/third-party2/libgcc/ea2fd1278810d3af2ea52218d2767e09d786dbd0/4.8.1/gcc-4.8.1-glibc-2.17/8aac7fc
-GLIBC_BASE=/mnt/gvfs/third-party2/glibc/f5484f168c0e4d19823d41df052c5870c6e575a4/2.17/gcc-4.8.1-glibc-2.17/99df8fc
-SNAPPY_BASE=/mnt/gvfs/third-party2/snappy/cbf6f1f209e5bd160bdc5d971744e039f36b1566/1.1.3/gcc-4.8.1-glibc-2.17/c3f970a
-ZLIB_BASE=/mnt/gvfs/third-party2/zlib/6d39cb54708049f527e713ad19f2aadb9d3667e8/1.2.8/gcc-4.8.1-glibc-2.17/c3f970a
-BZIP2_BASE=/mnt/gvfs/third-party2/bzip2/2ddd45f0853bfc8bb1c27f0f447236a1a26c338a/1.0.6/gcc-4.8.1-glibc-2.17/c3f970a
-LZ4_BASE=/mnt/gvfs/third-party2/lz4/6858fac689e0f92e584224d91bdb0e39f6c8320d/r131/gcc-4.8.1-glibc-2.17/c3f970a
-ZSTD_BASE=/mnt/gvfs/third-party2/zstd/9c833012dea9ae6de1df2cbdcb7aebc57ba2cb39/0.5.0/gcc-4.8.1-glibc-2.17/c3f970a
-GFLAGS_BASE=/mnt/gvfs/third-party2/gflags/c7275a4ceae0aca0929e56964a31dafc53c1ee96/2.1.1/gcc-4.8.1-glibc-2.17/c3f970a
-JEMALLOC_BASE=/mnt/gvfs/third-party2/jemalloc/378c8f95cf28c833e121ca73432101e451c3e62a/4.0.3/gcc-4.8.1-glibc-2.17/8d31e51
-NUMA_BASE=/mnt/gvfs/third-party2/numa/ae54a5ed22cdabb1c6446dce4e8ffae5b4446d73/2.0.8/gcc-4.8.1-glibc-2.17/c3f970a
-LIBUNWIND_BASE=/mnt/gvfs/third-party2/libunwind/303048f72efc92ae079e62dfc84823401aecfd94/trunk/gcc-4.8.1-glibc-2.17/675d945
-KERNEL_HEADERS_BASE=/mnt/gvfs/third-party2/kernel-headers/1a48835975c66d30e47770ec419758ed3b9ba010/3.10.62-62_fbk17_03959_ge29cc63/gcc-4.8.1-glibc-2.17/da39a3e
-BINUTILS_BASE=/mnt/gvfs/third-party2/binutils/a5b8152b2a15ce8a98808cf954fbccec825a97bc/2.25/centos6-native/da39a3e
-VALGRIND_BASE=/mnt/gvfs/third-party2/valgrind/af85c56f424cd5edfc2c97588299b44ecdec96bb/3.8.1/gcc-4.8.1-glibc-2.17/c3f970a
@@ -6,147 +6,128 @@
 # 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.sh"
-
 CFLAGS=""

-# libgcc
-LIBGCC_INCLUDE="$LIBGCC_BASE/include/c++/7.3.0"
-LIBGCC_LIBS=" -L $LIBGCC_BASE/lib"
+# location of libgcc
+LIBGCC_BASE="/mnt/gvfs/third-party2/libgcc/0473c80518a10d6efcbe24c5eeca3fb4ec9b519c/4.9.x/gcc-4.9-glibc-2.20/e1a7e4e"
+LIBGCC_INCLUDE="$LIBGCC_BASE/include"
+LIBGCC_LIBS=" -L $LIBGCC_BASE/libs"

-# glibc
-GLIBC_INCLUDE="$GLIBC_BASE/include"
-GLIBC_LIBS=" -L $GLIBC_BASE/lib"
+# location of glibc
+GLIBC_REV=7397bed99280af5d9543439cdb7d018af7542720
+GLIBC_INCLUDE="/mnt/gvfs/third-party2/glibc/$GLIBC_REV/2.20/gcc-4.9-glibc-2.20/99df8fc/include"
+GLIBC_LIBS=" -L /mnt/gvfs/third-party2/glibc/$GLIBC_REV/2.20/gcc-4.9-glibc-2.20/99df8fc/lib"
+
+SNAPPY_INCLUDE=" -I /mnt/gvfs/third-party2/snappy/b0f269b3ca47770121aa159b99e1d8d2ab260e1f/1.0.3/gcc-4.9-glibc-2.20/c32916f/include/"

-# snappy
-SNAPPY_INCLUDE=" -I $SNAPPY_BASE/include/"
 if test -z $PIC_BUILD; then
-  SNAPPY_LIBS=" $SNAPPY_BASE/lib/libsnappy.a"
+  SNAPPY_LIBS=" /mnt/gvfs/third-party2/snappy/b0f269b3ca47770121aa159b99e1d8d2ab260e1f/1.0.3/gcc-4.9-glibc-2.20/c32916f/lib/libsnappy.a"
 else
-  SNAPPY_LIBS=" $SNAPPY_BASE/lib/libsnappy_pic.a"
+  SNAPPY_LIBS=" /mnt/gvfs/third-party2/snappy/b0f269b3ca47770121aa159b99e1d8d2ab260e1f/1.0.3/gcc-4.9-glibc-2.20/c32916f/lib/libsnappy_pic.a"
 fi
+
 CFLAGS+=" -DSNAPPY"

 if test -z $PIC_BUILD; then
  # location of zlib headers and libraries
-  ZLIB_INCLUDE=" -I $ZLIB_BASE/include/"
-  ZLIB_LIBS=" $ZLIB_BASE/lib/libz.a"
+  ZLIB_INCLUDE=" -I /mnt/gvfs/third-party2/zlib/feb983d9667f4cf5e9da07ce75abc824764b67a1/1.2.8/gcc-4.9-glibc-2.20/4230243/include/"
+  ZLIB_LIBS=" /mnt/gvfs/third-party2/zlib/feb983d9667f4cf5e9da07ce75abc824764b67a1/1.2.8/gcc-4.9-glibc-2.20/4230243/lib/libz.a"
  CFLAGS+=" -DZLIB"

  # location of bzip headers and libraries
-  BZIP_INCLUDE=" -I $BZIP2_BASE/include/"
-  BZIP_LIBS=" $BZIP2_BASE/lib/libbz2.a"
+  BZIP_INCLUDE=" -I /mnt/gvfs/third-party2/bzip2/af004cceebb2dfd173ca29933ea5915e727aad2f/1.0.6/gcc-4.9-glibc-2.20/4230243/include/"
+  BZIP_LIBS=" /mnt/gvfs/third-party2/bzip2/af004cceebb2dfd173ca29933ea5915e727aad2f/1.0.6/gcc-4.9-glibc-2.20/4230243/lib/libbz2.a"
  CFLAGS+=" -DBZIP2"

-  LZ4_INCLUDE=" -I $LZ4_BASE/include/"
-  LZ4_LIBS=" $LZ4_BASE/lib/liblz4.a"
+  LZ4_INCLUDE=" -I /mnt/gvfs/third-party2/lz4/79d2943e2dd7208a3e0b06cf95e9f85f05fe9e1b/r124/gcc-4.9-glibc-2.20/4230243/include/"
+  LZ4_LIBS=" /mnt/gvfs/third-party2/lz4/79d2943e2dd7208a3e0b06cf95e9f85f05fe9e1b/r124/gcc-4.9-glibc-2.20/4230243/lib/liblz4.a"
  CFLAGS+=" -DLZ4"
-fi

-ZSTD_INCLUDE=" -I $ZSTD_BASE/include/"
-if test -z $PIC_BUILD; then
-  ZSTD_LIBS=" $ZSTD_BASE/lib/libzstd.a"
-else
-  ZSTD_LIBS=" $ZSTD_BASE/lib/libzstd_pic.a"
+  ZSTD_REV=8df2d01673ae6afcc8c8d16fec862b2d67ecc1e9
+  ZSTD_INCLUDE=" -I /mnt/gvfs/third-party2/zstd/$ZSTD_REV/0.1.1/gcc-4.8.1-glibc-2.17/c3f970a/include"
+  ZSTD_LIBS=" /mnt/gvfs/third-party2/zstd/$ZSTD_REV/0.1.1/gcc-4.8.1-glibc-2.17/c3f970a/lib/libzstd.a"
+  CFLAGS+=" -DZSTD"
 fi
-CFLAGS+=" -DZSTD"

 # location of gflags headers and libraries
-GFLAGS_INCLUDE=" -I $GFLAGS_BASE/include/"
+GFLAGS_INCLUDE=" -I /mnt/gvfs/third-party2/gflags/0fa60e2b88de3e469db6c482d6e6dac72f5d65f9/1.6/gcc-4.9-glibc-2.20/4230243/include/"
 if test -z $PIC_BUILD; then
-  GFLAGS_LIBS=" $GFLAGS_BASE/lib/libgflags.a"
+  GFLAGS_LIBS=" /mnt/gvfs/third-party2/gflags/0fa60e2b88de3e469db6c482d6e6dac72f5d65f9/1.6/gcc-4.9-glibc-2.20/4230243/lib/libgflags.a"
 else
-  GFLAGS_LIBS=" $GFLAGS_BASE/lib/libgflags_pic.a"
+  GFLAGS_LIBS=" /mnt/gvfs/third-party2/gflags/0fa60e2b88de3e469db6c482d6e6dac72f5d65f9/1.6/gcc-4.9-glibc-2.20/4230243/lib/libgflags_pic.a"
 fi
-CFLAGS+=" -DGFLAGS=gflags"
+CFLAGS+=" -DGFLAGS=google"

 # location of jemalloc
-JEMALLOC_INCLUDE=" -I $JEMALLOC_BASE/include/"
-JEMALLOC_LIB=" $JEMALLOC_BASE/lib/libjemalloc.a"
+JEMALLOC_INCLUDE=" -I /mnt/gvfs/third-party2/jemalloc/bcd68e5e419efa4e61b9486d6854564d6d75a0b5/3.6.0/gcc-4.9-glibc-2.20/2aafc78/include/"
+JEMALLOC_LIB=" /mnt/gvfs/third-party2/jemalloc/bcd68e5e419efa4e61b9486d6854564d6d75a0b5/3.6.0/gcc-4.9-glibc-2.20/2aafc78/lib/libjemalloc.a"

 if test -z $PIC_BUILD; then
  # location of numa
-  NUMA_INCLUDE=" -I $NUMA_BASE/include/"
-  NUMA_LIB=" $NUMA_BASE/lib/libnuma.a"
+  NUMA_INCLUDE=" -I /mnt/gvfs/third-party2/numa/bbefc39ecbf31d0ca184168eb613ef8d397790ee/2.0.8/gcc-4.9-glibc-2.20/4230243/include/"
+  NUMA_LIB=" /mnt/gvfs/third-party2/numa/bbefc39ecbf31d0ca184168eb613ef8d397790ee/2.0.8/gcc-4.9-glibc-2.20/4230243/lib/libnuma.a"
  CFLAGS+=" -DNUMA"

  # location of libunwind
-  LIBUNWIND="$LIBUNWIND_BASE/lib/libunwind.a"
+  LIBUNWIND="/mnt/gvfs/third-party2/libunwind/1de3b75e0afedfe5585b231bbb340ec7a1542335/1.1/gcc-4.9-glibc-2.20/34235e8/lib/libunwind.a"
 fi

-# location of TBB
-TBB_INCLUDE=" -isystem $TBB_BASE/include/"
-if test -z $PIC_BUILD; then
-  TBB_LIBS="$TBB_BASE/lib/libtbb.a"
-else
-  TBB_LIBS="$TBB_BASE/lib/libtbb_pic.a"
-fi
-CFLAGS+=" -DTBB"
-
 # use Intel SSE support for checksum calculations
 export USE_SSE=1

-BINUTILS="$BINUTILS_BASE/bin"
+BINUTILS="/mnt/gvfs/third-party2/binutils/0b6ad0c88ddd903333a48ae8bff134efac468e4a/2.25/centos6-native/da39a3e/bin"
 AR="$BINUTILS/ar"

-DEPS_INCLUDE="$SNAPPY_INCLUDE $ZLIB_INCLUDE $BZIP_INCLUDE $LZ4_INCLUDE $ZSTD_INCLUDE $GFLAGS_INCLUDE $NUMA_INCLUDE $TBB_INCLUDE"
+DEPS_INCLUDE="$SNAPPY_INCLUDE $ZLIB_INCLUDE $BZIP_INCLUDE $LZ4_INCLUDE $ZSTD_INCLUDE $GFLAGS_INCLUDE $NUMA_INCLUDE"

+GCC_BASE="/mnt/gvfs/third-party2/gcc/1c67a0b88f64d4d9ced0382d141c76aaa7d62fba/4.9.x/centos6-native/1317bc4"
 STDLIBS="-L $GCC_BASE/lib64"

-CLANG_BIN="$CLANG_BASE/bin"
-CLANG_LIB="$CLANG_BASE/lib"
-CLANG_SRC="$CLANG_BASE/../../src"
-
+CLANG_BASE="/mnt/gvfs/third-party2/clang/d81444dd214df3d2466734de45bb264a0486acc3/dev"
+CLANG_BIN="$CLANG_BASE/centos6-native/af4b1a0/bin"
 CLANG_ANALYZER="$CLANG_BIN/clang++"
-CLANG_SCAN_BUILD="$CLANG_SRC/llvm/tools/clang/tools/scan-build/bin/scan-build"
+CLANG_SCAN_BUILD="$CLANG_BASE/src/clang/tools/scan-build/scan-build"

 if [ -z "$USE_CLANG" ]; then
  # gcc
  CC="$GCC_BASE/bin/gcc"
  CXX="$GCC_BASE/bin/g++"
-
+  
  CFLAGS+=" -B$BINUTILS/gold"
-  CFLAGS+=" -isystem $LIBGCC_INCLUDE"
  CFLAGS+=" -isystem $GLIBC_INCLUDE"
-  JEMALLOC=1
+  CFLAGS+=" -isystem $LIBGCC_INCLUDE"
 else
-  # clang
-  CLANG_INCLUDE="$CLANG_LIB/clang/stable/include"
+  # clang 
+  CLANG_INCLUDE="$CLANG_BASE/gcc-4.9-glibc-2.20/74c386f/lib/clang/dev/include/"
  CC="$CLANG_BIN/clang"
  CXX="$CLANG_BIN/clang++"

-  KERNEL_HEADERS_INCLUDE="$KERNEL_HEADERS_BASE/include"
+  KERNEL_HEADERS_INCLUDE="/mnt/gvfs/third-party2/kernel-headers/ffd14f660a43c4b92717986b1bba66722ef089d0/3.2.18_70_fbk11_00129_gc8882d0/gcc-4.9-glibc-2.20/da39a3e/include"

  CFLAGS+=" -B$BINUTILS/gold -nostdinc -nostdlib"
-  CFLAGS+=" -isystem $LIBGCC_BASE/include/c++/7.x "
-  CFLAGS+=" -isystem $LIBGCC_BASE/include/c++/7.x/x86_64-facebook-linux "
+  CFLAGS+=" -isystem $LIBGCC_BASE/include/c++/4.9.x "
+  CFLAGS+=" -isystem $LIBGCC_BASE/include/c++/4.9.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 -DROCKSDB_SUPPORT_THREAD_LOCAL -DHAVE_SSE42"
+CFLAGS+=" -DROCKSDB_PLATFORM_POSIX -DROCKSDB_FALLOCATE_PRESENT -DROCKSDB_MALLOC_USABLE_SIZE"
 CXXFLAGS+=" $CFLAGS"

-EXEC_LDFLAGS=" $SNAPPY_LIBS $ZLIB_LIBS $BZIP_LIBS $LZ4_LIBS $ZSTD_LIBS $GFLAGS_LIBS $NUMA_LIB $TBB_LIBS"
-EXEC_LDFLAGS+=" -B$BINUTILS/gold"
-EXEC_LDFLAGS+=" -Wl,--dynamic-linker,/usr/local/fbcode/platform007/lib/ld.so"
+EXEC_LDFLAGS=" $SNAPPY_LIBS $ZLIB_LIBS $BZIP_LIBS $LZ4_LIBS $ZSTD_LIBS $GFLAGS_LIBS $NUMA_LIB"
+EXEC_LDFLAGS+=" -Wl,--dynamic-linker,/usr/local/fbcode/gcc-4.9-glibc-2.20/lib/ld.so"
 EXEC_LDFLAGS+=" $LIBUNWIND"
-EXEC_LDFLAGS+=" -Wl,-rpath=/usr/local/fbcode/platform007/lib"
-# required by libtbb
-EXEC_LDFLAGS+=" -ldl"
+EXEC_LDFLAGS+=" -Wl,-rpath=/usr/local/fbcode/gcc-4.9-glibc-2.20/lib"

 PLATFORM_LDFLAGS="$LIBGCC_LIBS $GLIBC_LIBS $STDLIBS -lgcc -lstdc++"

-EXEC_LDFLAGS_SHARED="$SNAPPY_LIBS $ZLIB_LIBS $BZIP_LIBS $LZ4_LIBS $ZSTD_LIBS $GFLAGS_LIBS $TBB_LIBS"
+EXEC_LDFLAGS_SHARED="$SNAPPY_LIBS $ZLIB_LIBS $BZIP_LIBS $LZ4_LIBS $ZSTD_LIBS $GFLAGS_LIBS"

-VALGRIND_VER="$VALGRIND_BASE/bin/"
+VALGRIND_VER="/mnt/gvfs/third-party2/valgrind/6c45ef049cbf11c2df593addb712cd891049e737/3.10.0/gcc-4.9-glibc-2.20/4230243/bin/"

-export CC CXX AR CFLAGS CXXFLAGS EXEC_LDFLAGS EXEC_LDFLAGS_SHARED VALGRIND_VER JEMALLOC_LIB JEMALLOC_INCLUDE CLANG_ANALYZER CLANG_SCAN_BUILD 
+export CC CXX AR CFLAGS CXXFLAGS EXEC_LDFLAGS EXEC_LDFLAGS_SHARED VALGRIND_VER JEMALLOC_LIB JEMALLOC_INCLUDE CLANG_ANALYZER CLANG_SCAN_BUILD
@@ -4,77 +4,80 @@
 # fbcode settings.  It uses the latest g++ compiler and also
 # uses jemalloc

-BASEDIR=`dirname $BASH_SOURCE`
-source "$BASEDIR/dependencies_4.8.1.sh"
-
 # location of libgcc
+LIBGCC_BASE="/mnt/gvfs/third-party2/libgcc/7712e757d7355cb51292454ee0b7b46a467fdfed/4.8.1/gcc-4.8.1-glibc-2.17/8aac7fc"
 LIBGCC_INCLUDE="$LIBGCC_BASE/include"
 LIBGCC_LIBS=" -L $LIBGCC_BASE/libs"

 # location of glibc
-GLIBC_INCLUDE="$GLIBC_BASE/include"
-GLIBC_LIBS=" -L $GLIBC_BASE/lib"
+GLIBC_REV=6e40560b4e0b6d690fd1cf8c7a43ad7452b04cfa
+GLIBC_INCLUDE="/mnt/gvfs/third-party2/glibc/$GLIBC_REV/2.17/gcc-4.8.1-glibc-2.17/99df8fc/include"
+GLIBC_LIBS=" -L /mnt/gvfs/third-party2/glibc/$GLIBC_REV/2.17/gcc-4.8.1-glibc-2.17/99df8fc/lib"

 # location of snappy headers and libraries
-SNAPPY_INCLUDE=" -I $SNAPPY_BASE/include"
-SNAPPY_LIBS=" $SNAPPY_BASE/lib/libsnappy.a"
+SNAPPY_INCLUDE=" -I /mnt/gvfs/third-party2/snappy/aef17f6c0b44b4fe408bd06f67c93701ab0a6ceb/1.0.3/gcc-4.8.1-glibc-2.17/43d84e2/include"
+SNAPPY_LIBS=" /mnt/gvfs/third-party2/snappy/aef17f6c0b44b4fe408bd06f67c93701ab0a6ceb/1.0.3/gcc-4.8.1-glibc-2.17/43d84e2/lib/libsnappy.a"

 # location of zlib headers and libraries
-ZLIB_INCLUDE=" -I $ZLIB_BASE/include"
-ZLIB_LIBS=" $ZLIB_BASE/lib/libz.a"
+ZLIB_INCLUDE=" -I /mnt/gvfs/third-party2/zlib/25c6216928b4d77b59ddeca0990ff6fe9ac16b81/1.2.5/gcc-4.8.1-glibc-2.17/c3f970a/include"
+ZLIB_LIBS=" /mnt/gvfs/third-party2/zlib/25c6216928b4d77b59ddeca0990ff6fe9ac16b81/1.2.5/gcc-4.8.1-glibc-2.17/c3f970a/lib/libz.a"

 # location of bzip headers and libraries
-BZIP2_INCLUDE=" -I $BZIP2_BASE/include/"
-BZIP2_LIBS=" $BZIP2_BASE/lib/libbz2.a"
+BZIP_INCLUDE=" -I /mnt/gvfs/third-party2/bzip2/c9ef7629c2aa0024f7a416e87602f06eb88f5eac/1.0.6/gcc-4.8.1-glibc-2.17/c3f970a/include/"
+BZIP_LIBS=" /mnt/gvfs/third-party2/bzip2/c9ef7629c2aa0024f7a416e87602f06eb88f5eac/1.0.6/gcc-4.8.1-glibc-2.17/c3f970a/lib/libbz2.a"

-LZ4_INCLUDE=" -I $LZ4_BASE/include"
-LZ4_LIBS=" $LZ4_BASE/lib/liblz4.a"
+LZ4_REV=065ec7e38fe83329031f6668c43bef83eff5808b
+LZ4_INCLUDE=" -I /mnt/gvfs/third-party2/lz4/$LZ4_REV/r108/gcc-4.8.1-glibc-2.17/c3f970a/include"
+LZ4_LIBS=" /mnt/gvfs/third-party2/lz4/$LZ4_REV/r108/gcc-4.8.1-glibc-2.17/c3f970a/lib/liblz4.a"

-ZSTD_INCLUDE=" -I $ZSTD_BASE/include"
-ZSTD_LIBS=" $ZSTD_BASE/lib/libzstd.a"
+ZSTD_REV=8df2d01673ae6afcc8c8d16fec862b2d67ecc1e9
+ZSTD_INCLUDE=" -I /mnt/gvfs/third-party2/zstd/$ZSTD_REV/0.1.1/gcc-4.8.1-glibc-2.17/c3f970a/include"
+ZSTD_LIBS=" /mnt/gvfs/third-party2/zstd/$ZSTD_REV/0.1.1/gcc-4.8.1-glibc-2.17/c3f970a/lib/libzstd.a"

 # location of gflags headers and libraries
-GFLAGS_INCLUDE=" -I $GFLAGS_BASE/include/"
-GFLAGS_LIBS=" $GFLAGS_BASE/lib/libgflags.a"
+GFLAGS_INCLUDE=" -I /mnt/gvfs/third-party2/gflags/1ad047a6e6f6673991918ecadc670868205a243a/1.6/gcc-4.8.1-glibc-2.17/c3f970a/include/"
+GFLAGS_LIBS=" /mnt/gvfs/third-party2/gflags/1ad047a6e6f6673991918ecadc670868205a243a/1.6/gcc-4.8.1-glibc-2.17/c3f970a/lib/libgflags.a"

 # location of jemalloc
-JEMALLOC_INCLUDE=" -I $JEMALLOC_BASE/include"
-JEMALLOC_LIB="$JEMALLOC_BASE/lib/libjemalloc.a"
+JEMALLOC_INCLUDE=" -I /mnt/gvfs/third-party2/jemalloc/3691c776ac26dd8781e84f8888b6a0fbdbc0a9ed/dev/gcc-4.8.1-glibc-2.17/4d53c6f/include"
+JEMALLOC_LIB="/mnt/gvfs/third-party2/jemalloc/3691c776ac26dd8781e84f8888b6a0fbdbc0a9ed/dev/gcc-4.8.1-glibc-2.17/4d53c6f/lib/libjemalloc.a"

 # location of numa
-NUMA_INCLUDE=" -I $NUMA_BASE/include/"
-NUMA_LIB=" $NUMA_BASE/lib/libnuma.a"
+NUMA_REV=829d10dac0230f99cd7e1778869d2adf3da24b65
+NUMA_INCLUDE=" -I /mnt/gvfs/third-party2/numa/$NUMA_REV/2.0.8/gcc-4.8.1-glibc-2.17/c3f970a/include/"
+NUMA_LIB=" /mnt/gvfs/third-party2/numa/$NUMA_REV/2.0.8/gcc-4.8.1-glibc-2.17/c3f970a/lib/libnuma.a"

 # location of libunwind
-LIBUNWIND="$LIBUNWIND_BASE/lib/libunwind.a"
+LIBUNWIND_REV=2c060e64064559905d46fd194000d61592087bdc
+LIBUNWIND="/mnt/gvfs/third-party2/libunwind/$LIBUNWIND_REV/1.1/gcc-4.8.1-glibc-2.17/675d945/lib/libunwind.a"

 # use Intel SSE support for checksum calculations
 export USE_SSE=1

-BINUTILS="$BINUTILS_BASE/bin"
+BINUTILS="/mnt/gvfs/third-party2/binutils/2aff2e7b474cd3e6ab23495ad1224b7d214b9f8e/2.21.1/centos6-native/da39a3e/bin"
 AR="$BINUTILS/ar"

-DEPS_INCLUDE="$SNAPPY_INCLUDE $ZLIB_INCLUDE $BZIP2_INCLUDE $LZ4_INCLUDE $ZSTD_INCLUDE $GFLAGS_INCLUDE $NUMA_INCLUDE"
+DEPS_INCLUDE="$SNAPPY_INCLUDE $ZLIB_INCLUDE $BZIP_INCLUDE $LZ4_INCLUDE $ZSTD_INCLUDE $GFLAGS_INCLUDE $NUMA_INCLUDE"

+GCC_BASE="/mnt/gvfs/third-party2/gcc/1ec615e23800f0815d474478ba476a0adc3fe788/4.8.1/centos6-native/cc6c9dc"
 STDLIBS="-L $GCC_BASE/lib64"

 if [ -z "$USE_CLANG" ]; then
  # gcc
  CC="$GCC_BASE/bin/gcc"
  CXX="$GCC_BASE/bin/g++"
-
+  
  CFLAGS="-B$BINUTILS/gold -m64 -mtune=generic"
  CFLAGS+=" -isystem $GLIBC_INCLUDE"
  CFLAGS+=" -isystem $LIBGCC_INCLUDE"
 else
-  # clang
-  CLANG_BIN="$CLANG_BASE/bin"
-  CLANG_LIB="$CLANG_BASE/lib"
-  CLANG_INCLUDE="$CLANG_LIB/clang/*/include"
-  CC="$CLANG_BIN/clang"
-  CXX="$CLANG_BIN/clang++"
+  # clang 
+  CLANG_BASE="/mnt/gvfs/third-party2/clang/9ab68376f938992c4eb5946ca68f90c3185cffc8/3.4"
+  CLANG_INCLUDE="$CLANG_BASE/gcc-4.8.1-glibc-2.17/fb0f730/lib/clang/3.4/include"
+  CC="$CLANG_BASE/centos6-native/9cefd8a/bin/clang"
+  CXX="$CLANG_BASE/centos6-native/9cefd8a/bin/clang++"

-  KERNEL_HEADERS_INCLUDE="$KERNEL_HEADERS_BASE/include/"
+  KERNEL_HEADERS_INCLUDE="/mnt/gvfs/third-party2/kernel-headers/a683ed7135276731065a9d76d3016c9731f4e2f9/3.2.18_70_fbk11_00129_gc8882d0/gcc-4.8.1-glibc-2.17/da39a3e/include/"

  CFLAGS="-B$BINUTILS/gold -nostdinc -nostdlib"
  CFLAGS+=" -isystem $LIBGCC_BASE/include/c++/4.8.1 "
@@ -88,19 +91,20 @@ else
 fi

 CFLAGS+=" $DEPS_INCLUDE"
-CFLAGS+=" -DROCKSDB_PLATFORM_POSIX -DROCKSDB_LIB_IO_POSIX -DROCKSDB_FALLOCATE_PRESENT -DROCKSDB_MALLOC_USABLE_SIZE"
+CFLAGS+=" -DROCKSDB_PLATFORM_POSIX -DROCKSDB_FALLOCATE_PRESENT -DROCKSDB_MALLOC_USABLE_SIZE"
 CFLAGS+=" -DSNAPPY -DGFLAGS=google -DZLIB -DBZIP2 -DLZ4 -DZSTD -DNUMA"
 CXXFLAGS+=" $CFLAGS"

-EXEC_LDFLAGS=" $SNAPPY_LIBS $ZLIB_LIBS $BZIP2_LIBS $LZ4_LIBS $ZSTD_LIBS $GFLAGS_LIBS $NUMA_LIB"
+EXEC_LDFLAGS=" $SNAPPY_LIBS $ZLIB_LIBS $BZIP_LIBS $LZ4_LIBS $ZSTD_LIBS $GFLAGS_LIBS $NUMA_LIB"
 EXEC_LDFLAGS+=" -Wl,--dynamic-linker,/usr/local/fbcode/gcc-4.8.1-glibc-2.17/lib/ld.so"
 EXEC_LDFLAGS+=" $LIBUNWIND"
 EXEC_LDFLAGS+=" -Wl,-rpath=/usr/local/fbcode/gcc-4.8.1-glibc-2.17/lib"

 PLATFORM_LDFLAGS="$LIBGCC_LIBS $GLIBC_LIBS $STDLIBS -lgcc -lstdc++"

-EXEC_LDFLAGS_SHARED="$SNAPPY_LIBS $ZLIB_LIBS $BZIP2_LIBS $LZ4_LIBS $ZSTD_LIBS $GFLAGS_LIBS"
+EXEC_LDFLAGS_SHARED="$SNAPPY_LIBS $ZLIB_LIBS $BZIP_LIBS $LZ4_LIBS $ZSTD_LIBS $GFLAGS_LIBS"

-VALGRIND_VER="$VALGRIND_BASE/bin/"
+VALGRIND_REV=b2a9f85e4b70cd03abc85a7f3027fbc4cef35bd0
+VALGRIND_VER="/mnt/gvfs/third-party2/valgrind/$VALGRIND_REV/3.8.1/gcc-4.8.1-glibc-2.17/c3f970a/bin/"

 export CC CXX AR CFLAGS CXXFLAGS EXEC_LDFLAGS EXEC_LDFLAGS_SHARED VALGRIND_VER JEMALLOC_LIB JEMALLOC_INCLUDE
@@ -243,7 +243,7 @@ make release
    --bloom_bits=10 \
    --num=$((NUM / 4)) \
    --reads=$((NUM / 4)) \
-    --benchmark_write_rate_limit=$(( 110 * 1024 )) \
+    --writes_per_second=1000 \
    --write_buffer_size=100000000 \
    --cache_size=6442450944 \
    --cache_numshardbits=6 \
@@ -329,7 +329,7 @@ common_in_mem_args="--db=/dev/shm/rocksdb \
    --use_existing_db=1 \
    --duration=600 \
    --threads=32 \
-    --benchmark_write_rate_limit=9502720 > ${STAT_FILE}.readwhilewriting_in_ram
+    --writes_per_second=81920 > ${STAT_FILE}.readwhilewriting_in_ram

 # Seekrandomwhilewriting
 ./db_bench \
@@ -342,7 +342,7 @@ common_in_mem_args="--db=/dev/shm/rocksdb \
    --use_tailing_iterator=1 \
    --duration=600 \
    --threads=32 \
-    --benchmark_write_rate_limit=9502720 > ${STAT_FILE}.seekwhilewriting_in_ram
+    --writes_per_second=81920 > ${STAT_FILE}.seekwhilewriting_in_ram

 # measure fillseq with bunch of column families
 ./db_bench \
@@ -2,62 +2,22 @@
 # This script is executed by Sandcastle
 # to determine next steps to run

-# Usage:
-# EMAIL=<email> ONCALL=<email> TRIGGER=<trigger> SUBSCRIBER=<email> rocks_ci.py <test-name>
-#
-# Input         Value
-# -------------------------------------------------------------------------
-# EMAIL         Email address to report on trigger conditions
-# ONCAL         Email address to raise a task on failure
-# TRIGGER       Trigger conditions for email. Valid values are fail, warn, all
-# SUBSCRIBER    Email addresss to add as subscriber for task
-#
-
-#
-# Report configuration
-#
-REPORT_EMAIL=
-if [ ! -z $EMAIL ]; then
-  if [ -z $TRIGGER ]; then
-    TRIGGER="fail"
-  fi
-
-  REPORT_EMAIL="
-  {
-      'type':'email',
-      'triggers': [ '$TRIGGER' ],
-      'emails':['$EMAIL']
-  },"
-fi
-
-CREATE_TASK=
-if [ ! -z $ONCALL ]; then
-  CREATE_TASK="
-  {
-      'type':'task',
-      'triggers':[ 'fail' ],
-      'priority':0,
-      'subscribers':[ '$SUBSCRIBER' ],
-      'tags':[ 'rocksdb', 'ci' ],
-  },"
-fi
-
-REPORT=
-if [[ ! -z $REPORT_EMAIL || ! -z $CREATE_TASK ]]; then
-  REPORT="'report': [
-    $REPORT_EMAIL
-    $CREATE_TASK
-  ]"
-fi
-
 #
 # Helper variables
 #
+REPORT_EMAIL="'report': [
+    {
+        'type':'email',
+        'triggers':['fail'],
+        'emails':['leveldb@fb.com']
+    }
+]"
+
 CLEANUP_ENV="
 {
-    'name':'Cleanup environment',
-    'shell':'rm -rf /dev/shm/rocksdb && mkdir /dev/shm/rocksdb && make clean',
-    'user':'root'
+  'name':'Cleanup environment',
+  'shell':'rm -rf /dev/shm/rocksdb && mkdir /dev/shm/rocksdb && make clean',
+  'user':'root'
 }"

 DEBUG="OPT=-g"
@@ -67,32 +27,6 @@ ASAN="COMPILE_WITH_ASAN=1"
 CLANG="USE_CLANG=1"
 LITE="OPT=-DROCKSDB_LITE"
 TSAN="COMPILE_WITH_TSAN=1"
-DISABLE_JEMALLOC="DISABLE_JEMALLOC=1"
-PARSER="'parser':'egrep \'Failure|^#|Abort|Expected|Actual|GoogleTestFailure|^==\''"
-
-ARTIFACTS=" 'artifacts': [
-    {
-        'name':'database',
-        'paths':[ '/dev/shm/rocksdb' ],
-    }
-]"
-
-#
-# A mechanism to disable tests temporarily
-#
-DISABLE_COMMANDS="[
-    {
-        'name':'Disable test',
-        'oncall':'$ONCALL',
-        'steps': [
-            {
-              'name':'Job disabled. Please contact test owner',
-              'shell':'exit 1',
-              'user':'root'
-            },
-        ],
-    }
-]"

 #
 # RocksDB unit test in parallel
@@ -102,24 +36,21 @@ DISABLE_COMMANDS="[
 PARALLEL_UNIT_TEST_COMMANDS="[
    {
        'name':'Rocksdb Parallel Unit Test',
-        'oncall':'$ONCALL',
        'steps': [
            $CLEANUP_ENV,
            {
                'name':'Build and test RocksDB debug version',
-                'shell':'$DEBUG make -j$(nproc) all && $SHM make check > /dev/null 2>&1 || cat t/log-*',
-                'user':'root',
-                $PARSER
+                'shell':'$SHM $DEBUG make check 2>&1 > /dev/null || cat t/log-*',
+                'user':'root'
            },
            $CLEANUP_ENV,
            {
                'name':'Build and test RocksDB debug version under gcc-4.8.1',
-                'shell':'$GCC_481 $DEBUG make -j$(nproc) all && $SHM make check > /dev/null 2>&1 || cat t/log-*',
-                'user':'root',
-                $PARSER
+                'shell':'$SHM $GCC_481 $DEBUG make check 2>&1 > /dev/null || cat t/log-*',
+                'user':'root'
            },
        ],
-        $REPORT
+        $REPORT_EMAIL
    }
 ]"

@@ -129,17 +60,15 @@ PARALLEL_UNIT_TEST_COMMANDS="[
 UNIT_TEST_COMMANDS="[
    {
        'name':'Rocksdb Unit Test',
-        'oncall':'$ONCALL',
        'steps': [
            $CLEANUP_ENV,
            {
                'name':'Build and test RocksDB debug version',
                'shell':'$SHM $DEBUG make J=1 check',
-                'user':'root',
-                $PARSER
+                'user':'root'
            },
        ],
-        $REPORT
+        $REPORT_EMAIL
    }
 ]"

@@ -149,17 +78,15 @@ UNIT_TEST_COMMANDS="[
 UNIT_TEST_COMMANDS_481="[
    {
        'name':'Rocksdb Unit Test on GCC 4.8.1',
-        'oncall':'$ONCALL',
        'steps': [
            $CLEANUP_ENV,
            {
                'name':'Build and test RocksDB debug version',
                'shell':'$SHM $GCC_481 $DEBUG make J=1 check',
-                'user':'root',
-                $PARSER
+                'user':'root'
            },
        ],
-        $REPORT
+        $REPORT_EMAIL
    }
 ]"

@@ -169,77 +96,15 @@ UNIT_TEST_COMMANDS_481="[
 CLANG_UNIT_TEST_COMMANDS="[
    {
        'name':'Rocksdb Unit Test',
-        'oncall':'$ONCALL',
        'steps': [
            $CLEANUP_ENV,
            {
                'name':'Build and test RocksDB debug',
-                'shell':'$CLANG $SHM $DEBUG make J=1 check',
-                'user':'root',
-                $PARSER
+                'shell':'$CLANG $SHM $DEBUG make check',
+                'user':'root'
            },
        ],
-        $REPORT
-    }
-]"
-
-#
-# RocksDB analyze
-#
-CLANG_ANALYZE_COMMANDS="[
-    {
-        'name':'Rocksdb analyze',
-        'oncall':'$ONCALL',
-        'steps': [
-            $CLEANUP_ENV,
-            {
-                'name':'RocksDB build and analyze',
-                'shell':'$CLANG $SHM $DEBUG make J=1 analyze',
-                'user':'root',
-                $PARSER
-            },
-        ],
-        $REPORT
-    }
-]"
-
-#
-# RocksDB code coverage
-#
-CODE_COV_COMMANDS="[
-    {
-        'name':'Rocksdb Unit Test Code Coverage',
-        'oncall':'$ONCALL',
-        'steps': [
-            $CLEANUP_ENV,
-            {
-                'name':'Build, test and collect code coverage info',
-                'shell':'$SHM $DEBUG make J=1 coverage',
-                'user':'root',
-                $PARSER
-            },
-        ],
-        $REPORT
-    }
-]"
-
-#
-# RocksDB unity
-#
-UNITY_COMMANDS="[
-    {
-        'name':'Rocksdb Unity',
-        'oncall':'$ONCALL',
-        'steps': [
-            $CLEANUP_ENV,
-            {
-                'name':'Build, test unity test',
-                'shell':'$SHM $DEBUG V=1 make J=1 unity_test',
-                'user':'root',
-                $PARSER
-            },
-        ],
-        $REPORT
+        $REPORT_EMAIL
    }
 ]"

@@ -249,37 +114,15 @@ UNITY_COMMANDS="[
 LITE_BUILD_COMMANDS="[
    {
        'name':'Rocksdb Lite build',
-        'oncall':'$ONCALL',
        'steps': [
            $CLEANUP_ENV,
            {
                'name':'Build RocksDB debug version',
                'shell':'$LITE $DEBUG make J=1 static_lib',
-                'user':'root',
-                $PARSER
+                'user':'root'
            },
        ],
-        $REPORT
-    }
-]"
-
-#
-# RocksDB lite tests
-#
-LITE_UNIT_TEST_COMMANDS="[
-    {
-        'name':'Rocksdb Lite Unit Test',
-        'oncall':'$ONCALL',
-        'steps': [
-            $CLEANUP_ENV,
-            {
-                'name':'Build RocksDB debug version',
-                'shell':'$SHM $LITE $DEBUG make J=1 check',
-                'user':'root',
-                $PARSER
-            },
-        ],
-        $REPORT
+        $REPORT_EMAIL
    }
 ]"

@@ -289,47 +132,20 @@ LITE_UNIT_TEST_COMMANDS="[
 STRESS_CRASH_TEST_COMMANDS="[
    {
        'name':'Rocksdb Stress/Crash Test',
-        'oncall':'$ONCALL',
-        'timeout': 86400,
        'steps': [
            $CLEANUP_ENV,
            {
                'name':'Build and run RocksDB debug stress tests',
                'shell':'$SHM $DEBUG make J=1 db_stress',
-                'user':'root',
-                $PARSER
+                'user':'root'
            },
            {
                'name':'Build and run RocksDB debug crash tests',
-                'timeout': 86400,
                'shell':'$SHM $DEBUG make J=1 crash_test',
-                'user':'root',
-                $PARSER
+                'user':'root'
            }
        ],
-        $ARTIFACTS,
-        $REPORT
-    }
-]"
-
-# RocksDB write stress test.
-# We run on disk device on purpose (i.e. no $SHM)
-# because we want to add some randomness to fsync commands
-WRITE_STRESS_COMMANDS="[
-    {
-        'name':'Rocksdb Write Stress Test',
-        'oncall':'$ONCALL',
-        'steps': [
-            $CLEANUP_ENV,
-            {
-                'name':'Build and run RocksDB write stress tests',
-                'shell':'make write_stress && python tools/write_stress_runner.py --runtime_sec=3600 --db=/tmp/rocksdb_write_stress',
-                'user':'root',
-                $PARSER
-            }
-        ],
-        'artifacts': [{'name': 'database', 'paths': ['/tmp/rocksdb_write_stress']}],
-        $REPORT
+        $REPORT_EMAIL
    }
 ]"

@@ -340,17 +156,15 @@ WRITE_STRESS_COMMANDS="[
 ASAN_TEST_COMMANDS="[
    {
        'name':'Rocksdb Unit Test under ASAN',
-        'oncall':'$ONCALL',
        'steps': [
            $CLEANUP_ENV,
            {
                'name':'Test RocksDB debug under ASAN',
                'shell':'set -o pipefail && $SHM $ASAN $DEBUG make J=1 asan_check |& /usr/facebook/ops/scripts/asan_symbolize.py -d',
-                'user':'root',
-                $PARSER
+                'user':'root'
            }
        ],
-        $REPORT
+        $REPORT_EMAIL
    }
 ]"

@@ -360,19 +174,15 @@ ASAN_TEST_COMMANDS="[
 ASAN_CRASH_TEST_COMMANDS="[
    {
        'name':'Rocksdb crash test under ASAN',
-        'oncall':'$ONCALL',
-        'timeout': 86400,
        'steps': [
            $CLEANUP_ENV,
            {
                'name':'Build and run RocksDB debug asan_crash_test',
-                'timeout': 86400,
                'shell':'$SHM $DEBUG make J=1 asan_crash_test',
-                'user':'root',
-                $PARSER
+                'user':'root'
            },
        ],
-        $REPORT
+        $REPORT_EMAIL
    }
 ]"

@@ -382,213 +192,33 @@ ASAN_CRASH_TEST_COMMANDS="[
 VALGRIND_TEST_COMMANDS="[
    {
        'name':'Rocksdb Unit Test under valgrind',
-        'oncall':'$ONCALL',
        'steps': [
            $CLEANUP_ENV,
            {
                'name':'Run RocksDB debug unit tests',
-                'shell':'$DISABLE_JEMALLOC $SHM $DEBUG make valgrind_check',
-                'user':'root',
-                $PARSER
+                'shell':'$SHM $DEBUG ./build_tools/valgrind_test.sh',
+                'user':'root'
            },
        ],
-        $REPORT
+        $REPORT_EMAIL
    }
 ]"

 #
 # RocksDB test under TSAN
 #
-TSAN_UNIT_TEST_COMMANDS="[
+TSAN_TEST_COMMANDS="[
    {
        'name':'Rocksdb Unit Test under TSAN',
-        'oncall':'$ONCALL',
        'steps': [
            $CLEANUP_ENV,
            {
                'name':'Run RocksDB debug unit test',
                'shell':'set -o pipefail && $SHM $DEBUG $TSAN make J=1 check',
-                'user':'root',
-                $PARSER
+                'user':'root'
            },
        ],
-        $REPORT
-    }
-]"
-
-#
-# RocksDB crash test under TSAN
-#
-TSAN_CRASH_TEST_COMMANDS="[
-    {
-        'name':'Rocksdb Crash Test under TSAN',
-        'oncall':'$ONCALL',
-        'timeout': 86400,
-        'steps': [
-            $CLEANUP_ENV,
-            {
-                'name':'Compile and run',
-                'timeout': 86400,
-                'shell':'set -o pipefail && $SHM $DEBUG $TSAN make J=1 crash_test',
-                'user':'root',
-                $PARSER
-            },
-        ],
-        $REPORT
-    }
-]"
-
-TSAN_CRASH_TEST_COMMANDS=$DISABLE_COMMANDS
-
-#
-# RocksDB format compatible
-#
-
-run_format_compatible()
-{
-  export TEST_TMPDIR=/dev/shm/rocksdb
-  rm -rf /dev/shm/rocksdb
-  mkdir /dev/shm/rocksdb
-
-  echo '
-  if [ -e "build_tools/build_detect_platform" ]
-  then
-    sed "s/tcmalloc/nothingnothingnothing/g" build_tools/build_detect_platform > $TEST_TMPDIR/temp_build_file
-    rm -rf build_tools/build_detect_platform
-    cp $TEST_TMPDIR/temp_build_file build_tools/build_detect_platform
-    chmod +x build_tools/build_detect_platform
-  fi
-
-  if [ -e "build_detect_platform" ]
-  then
-    sed "s/tcmalloc/nothingnothingnothing/g" build_detect_platform > $TEST_TMPDIR/temp_build_file
-    rm -rf build_detect_platform 
-    cp $TEST_TMPDIR/temp_build_file build_detect_platform
-    chmod +x build_detect_platform 
-  fi
-
-  make ldb -j32
-
-  if [ -e "build_detect_platform" ]
-  then
-    git checkout -- build_detect_platform
-  fi
-
-  if [ -e "build_tools/build_detect_platform" ]
-  then
-    git checkout -- build_tools/build_detect_platform
-  fi
-  ' > temp_build_ldb.sh
-
-  sed "s/make ldb -j32/source temp_build_ldb.sh/g" tools/check_format_compatible.sh > tools/temp_check_format_compatible.sh
-  chmod +x tools/temp_check_format_compatible.sh
-  tools/temp_check_format_compatible.sh
-}
-
-FORMAT_COMPATIBLE_COMMANDS="[
-    {
-        'name':'Rocksdb Format Compatible tests',
-        'oncall':'$ONCALL',
-        'steps': [
-            $CLEANUP_ENV,
-            {
-                'name':'Run RocksDB debug unit test',
-                'shell':'build_tools/rocksdb-lego-determinator run_format_compatible',
-                'user':'root',
-                $PARSER
-            },
-        ],
-        $REPORT
-    }
-]"
-
-#
-# RocksDB no compression
-#
-run_no_compression()
-{
-  export TEST_TMPDIR=/dev/shm/rocksdb
-  rm -rf /dev/shm/rocksdb
-  mkdir /dev/shm/rocksdb
-  make clean
-  cat build_tools/fbcode_config.sh | grep -iv dzlib | grep -iv dlz4 | grep -iv dsnappy | grep -iv dbzip2 > .tmp.fbcode_config.sh
-  mv .tmp.fbcode_config.sh build_tools/fbcode_config.sh
-  cat Makefile | grep -v tools/ldb_test.py > .tmp.Makefile
-  mv .tmp.Makefile Makefile
-  make $DEBUG J=1 check
-}
-
-NO_COMPRESSION_COMMANDS="[
-    {
-        'name':'Rocksdb No Compression tests',
-        'oncall':'$ONCALL',
-        'steps': [
-            $CLEANUP_ENV,
-            {
-                'name':'Run RocksDB debug unit test',
-                'shell':'build_tools/rocksdb-lego-determinator run_no_compression',
-                'user':'root',
-                $PARSER
-            },
-        ],
-        $REPORT
-    }
-]"
-
-#
-# RocksDB regression
-#
-run_regression()
-{
-  time -v bash -vx ./build_tools/regression_build_test.sh $(mktemp -d  $WORKSPACE/leveldb.XXXX) $(mktemp leveldb_test_stats.XXXX)
-
-  # ======= report size to ODS ========
-
-  # parameters: $1 -- key, $2 -- value
-  function send_size_to_ods {
-    curl -s "https://www.intern.facebook.com/intern/agent/ods_set.php?entity=rocksdb_build&key=rocksdb.build_size.$1&value=$2" \
-      --connect-timeout 60
-  }
-
-  # === normal build ===
-  make clean
-  make -j$(nproc) static_lib
-  send_size_to_ods static_lib $(stat --printf="%s" librocksdb.a)
-  strip librocksdb.a
-  send_size_to_ods static_lib_stripped $(stat --printf="%s" librocksdb.a)
-
-  make -j$(nproc) shared_lib
-  send_size_to_ods shared_lib $(stat --printf="%s" `readlink -f librocksdb.so`)
-  strip `readlink -f librocksdb.so`
-  send_size_to_ods shared_lib_stripped $(stat --printf="%s" `readlink -f librocksdb.so`)
-
-  # === lite build ===
-  make clean
-  OPT=-DROCKSDB_LITE make -j$(nproc) static_lib
-  send_size_to_ods static_lib_lite $(stat --printf="%s" librocksdb.a)
-  strip librocksdb.a
-  send_size_to_ods static_lib_lite_stripped $(stat --printf="%s" librocksdb.a)
-
-  OPT=-DROCKSDB_LITE make -j$(nproc) shared_lib
-  send_size_to_ods shared_lib_lite $(stat --printf="%s" `readlink -f librocksdb.so`)
-  strip `readlink -f librocksdb.so`
-  send_size_to_ods shared_lib_lite_stripped $(stat --printf="%s" `readlink -f librocksdb.so`)
-}
-
-REGRESSION_COMMANDS="[
-    {
-        'name':'Rocksdb regression commands',
-        'oncall':'$ONCALL',
-        'steps': [
-            $CLEANUP_ENV,
-            {
-                'name':'Make and run script',
-                'shell':'build_tools/rocksdb-lego-determinator run_regression',
-                'user':'root',
-                $PARSER
-            },
-        ],
-        $REPORT
+        $REPORT_EMAIL
    }
 ]"

@@ -605,27 +235,12 @@ case $1 in
  clang_unit)
    echo $CLANG_UNIT_TEST_COMMANDS
    ;;
-  clang_analyze)
-    echo $CLANG_ANALYZE_COMMANDS
-    ;;
-  code_cov)
-    echo $CODE_COV_COMMANDS
-    ;;
-  unity)
-    echo $UNITY_COMMANDS
-    ;;
  lite)
    echo $LITE_BUILD_COMMANDS
    ;;
-  lite_test)
-    echo $LITE_UNIT_TEST_COMMANDS
-    ;;
  stress_crash)
    echo $STRESS_CRASH_TEST_COMMANDS
    ;;
-  write_stress)
-    echo $WRITE_STRESS_COMMANDS
-    ;;
  asan)
    echo $ASAN_TEST_COMMANDS
    ;;
@@ -636,28 +251,7 @@ case $1 in
    echo $VALGRIND_TEST_COMMANDS
    ;;
  tsan)
-    echo $TSAN_UNIT_TEST_COMMANDS
-    ;;
-  tsan_crash)
-    echo $TSAN_CRASH_TEST_COMMANDS
-    ;;
-  format_compatible)
-    echo $FORMAT_COMPATIBLE_COMMANDS
-    ;;
-  run_format_compatible)
-    run_format_compatible
-    ;;
-  no_compression)
-    echo $NO_COMPRESSION_COMMANDS
-    ;;
-  run_no_compression)
-    run_no_compression
-    ;;
-  regression)
-    echo $REGRESSION_COMMANDS
-    ;;
-  run_regression)
-    run_regression
+    echo $TSAN_TEST_COMMANDS
    ;;
  *)
    echo "Invalid determinator command"
@@ -1,318 +0,0 @@
-# This script enables you running RocksDB tests by running
-# All the tests in parallel and utilizing all the cores
-# For db_test the script first lists and parses the tests
-# and then fires them up in parallel using async PS Job functionality
-# Run the script from the enlistment
-Param(
-  [switch]$EnableJE = $false,  # Use je executable
-  [switch]$EnableRerun = $false, # Rerun failed tests sequentially at the end
-  [string]$WorkFolder = "",  # Direct tests to use that folder
-  [int]$Limit = -1, # -1 means run all otherwise limit for testing purposes
-  [string]$Exclude = "", # Expect a comma separated list, no spaces
-  [string]$Run = "db_test",  # Run db_test|tests|testname1,testname2...
-   # Number of async tasks that would run concurrently. Recommend a number below 64.
-   # However, CPU utlization really depends on the storage media. Recommend ram based disk.
-   # a value of 1 will run everything serially
-  [int]$Concurrency = 16
-)
-
-# Folders and commands must be fullpath to run assuming
-# the current folder is at the root of the git enlistment
-Get-Date
-
-# If running under Appveyor assume that root
-[string]$Appveyor = $Env:APPVEYOR_BUILD_FOLDER
-if($Appveyor -ne "") {
-    $RootFolder = $Appveyor
-} else {
-    $RootFolder = $PSScriptRoot -replace '\\build_tools', ''
-}
-
-$LogFolder = -Join($RootFolder, "\db_logs\")
-$BinariesFolder = -Join($RootFolder, "\build\Debug\")
-
-if($WorkFolder -eq "") {
-
-    # If TEST_TMPDIR is set use it    
-    [string]$var = $Env:TEST_TMPDIR
-    if($var -eq "") {
-        $WorkFolder = -Join($RootFolder, "\db_tests\")
-        $Env:TEST_TMPDIR = $WorkFolder
-    } else {
-        $WorkFolder = $var
-    }
-} else {
-# Override from a command line
-  $Env:TEST_TMPDIR = $WorkFolder
-}
-
-# Use JEMALLOC executables
-if($EnableJE) {
-    $db_test = -Join ($BinariesFolder, "db_test_je.exe")
-} else {
-    $db_test = -Join ($BinariesFolder, "db_test.exe")
-}
-
-Write-Output "Root: $RootFolder, WorkFolder: $WorkFolder"
-Write-Output "Binaries: $BinariesFolder db_test: $db_test"
-
-#Exclusions that we do not want to run
-$ExcludeTests = New-Object System.Collections.Generic.HashSet[string]
-
-
-if($Exclude -ne "") {
-    Write-Host "Exclude: $Exclude"
-    $l = $Exclude -split ' '
-    ForEach($t in $l) { $ExcludeTests.Add($t) | Out-Null }
-}
-
-# Create test directories in the current folder
-md -Path $WorkFolder -ErrorAction Ignore | Out-Null
-md -Path $LogFolder -ErrorAction Ignore | Out-Null
-
-# Extract the names of its tests by running db_test with --gtest_list_tests.
-# This filter removes the "#"-introduced comments, and expands to
-# fully-qualified names by changing input like this:
-#
-#   DBTest.
-#     Empty
-#     WriteEmptyBatch
-#   MultiThreaded/MultiThreadedDBTest.
-#     MultiThreaded/0  # GetParam() = 0
-#     MultiThreaded/1  # GetParam() = 1
-#
-# into this:
-#
-#   DBTest.Empty
-#   DBTest.WriteEmptyBatch
-#   MultiThreaded/MultiThreadedDBTest.MultiThreaded/0
-#   MultiThreaded/MultiThreadedDBTest.MultiThreaded/1
-# Output into the parameter in a form TestName -> Log File Name
-function Normalize-DbTests($HashTable) {
-
-    $Tests = @()
-# Run db_test to get a list of tests and store it into $a array
-    &$db_test --gtest_list_tests | tee -Variable Tests | Out-Null
-
-    # Current group
-    $Group=""
-
-    ForEach( $l in $Tests) {
-      # Trailing dot is a test group
-      if( $l -match "\.$") {
-        $Group = $l
-      }  else {
-        # Otherwise it is a test name, remove leading space
-        $test = $l -replace '^\s+',''
-        # remove trailing comment if any and create a log name
-        $test = $test -replace '\s+\#.*',''
-        $test = "$Group$test"
-
-        if($ExcludeTests.Contains($test)) {
-            continue
-        }
-
-        $test_log = $test -replace '[\./]','_'
-        $test_log += ".log"
-        $log_path = -join ($LogFolder, $test_log)
-
-        # Add to a hashtable
-        $HashTable.Add($test, $log_path);
-      }
-    }
-}
-
-# The function removes trailing .exe siffix if any,
-# creates a name for the log file
-function MakeAndAdd([string]$token, $HashTable) {
-    $test_name = $token -replace '.exe$', ''
-    $log_name =  -join ($test_name, ".log")
-    $log_path = -join ($LogFolder, $log_name)
-    if(!$ExcludeTests.Contains($test_name)) {
-        $HashTable.Add($test_name, $log_path)
-    } else {
-        Write-Warning "Test $test_name is excluded"
-    }
-}
-
-# The function scans build\Debug folder to discover
-# Test executables. It then populates a table with
-# Test executable name -> Log file
-function Discover-TestBinaries([string]$Pattern, $HashTable) {
-
-    $Exclusions = @("db_test*", "db_sanity_test*")
-
-    $p = -join ($BinariesFolder, $pattern)
-
-    Write-Host "Path: $p"
-
-    dir -Path $p -Exclude $Exclusions | ForEach-Object {
-       MakeAndAdd -token ($_.Name) -HashTable $HashTable
-    }
-}
-
-$TestsToRun = [ordered]@{}
-
-if($Run -ceq "db_test") {
-    Normalize-DbTests -HashTable $TestsToRun
-} elseif($Run -ceq "tests") {
-    if($EnableJE) {
-        $pattern = "*_test_je.exe"
-    } else {
-        $pattern = "*_test.exe"
-    }
-    Discover-TestBinaries -Pattern $pattern -HashTable $TestsToRun
-} else {
-
-    $test_list = $Run -split ' '
-
-    ForEach($t in $test_list) {
-       MakeAndAdd -token $t -HashTable $TestsToRun
-    }
-}
-
-$NumTestsToStart = $TestsToRun.Count
-if($Limit -ge 0 -and $NumTestsToStart -gt $Limit) {
-    $NumTestsToStart = $Limit
-}
-
-Write-Host "Attempting to start: $NumTestsToStart tests"
-
-# Invoke a test with a filter and redirect all output
-$InvokeTestCase = {
-    param($exe, $test, $log);
-    &$exe --gtest_filter=$test > $log 2>&1
-}
-
-# Invoke all tests and redirect output
-$InvokeTestAsync = {
-    param($exe, $log)
-    &$exe > $log 2>&1
-}
-
-# Hash that contains tests to rerun if any failed
-# Those tests will be rerun sequentially
-$Rerun = [ordered]@{}
-# Test limiting factor here
-$count = 0
-# Overall status
-[bool]$success = $true;
-
-function RunJobs($TestToLog, [int]$ConcurrencyVal, [bool]$AddForRerun)
-{
-    # Array to wait for any of the running jobs
-    $jobs = @()
-    # Hash JobToLog
-    $JobToLog = @{}
-
-    # Wait for all to finish and get the results
-    while(($JobToLog.Count -gt 0) -or
-          ($TestToLog.Count -gt 0)) {
-
-        # Make sure we have maximum concurrent jobs running if anything
-        # and the $Limit either not set or allows to proceed
-        while(($JobToLog.Count -lt $ConcurrencyVal) -and
-              (($TestToLog.Count -gt 0) -and
-              (($Limit -lt 0) -or ($count -lt $Limit)))) {
-
-
-            # We only need the first key
-            foreach($key in $TestToLog.keys) {
-                $k = $key
-                break
-            }
-
-            Write-Host "Starting $k"
-            $log_path = ($TestToLog.$k)
-
-            if($Run -ceq "db_test") {
-              $job = Start-Job -Name $k -ScriptBlock $InvokeTestCase -ArgumentList @($db_test,$k,$log_path)
-            } else {
-              [string]$Exe =  -Join ($BinariesFolder, $k)
-               $job = Start-Job -Name $k -ScriptBlock $InvokeTestAsync -ArgumentList @($exe,$log_path)
-            }
-
-            $JobToLog.Add($job, $log_path)
-            $TestToLog.Remove($k)
-
-            ++$count
-        }
-
-        if($JobToLog.Count -lt 1) {
-          break
-        }
-
-        $jobs = @()
-        foreach($k in $JobToLog.Keys) { $jobs += $k }
-
-        $completed = Wait-Job -Job $jobs -Any
-        $log = $JobToLog[$completed]
-        $JobToLog.Remove($completed)
-
-        $message = -join @($completed.Name, " State: ", ($completed.State))
-
-        $log_content = @(Get-Content $log)
-
-        if($completed.State -ne "Completed") {
-            $success = $false
-            Write-Warning $message
-            $log_content | Write-Warning
-        } else {
-            # Scan the log. If we find PASSED and no occurrence of FAILED
-            # then it is a success
-            [bool]$pass_found = $false
-            ForEach($l in $log_content) {
-
-                if(($l -match "^\[\s+FAILED") -or
-                   ($l -match "Assertion failed:")) {
-                    $pass_found = $false
-                    break
-                }
-
-                if(($l -match "^\[\s+PASSED") -or
-                   ($l -match " : PASSED$") -or
-                    ($l -match "^PASS$") -or   # Special c_test case
-                    ($l -match "Passed all tests!") ) {
-                    $pass_found = $true
-                }
-            }
-
-            if(!$pass_found) {
-                $success = $false;
-                Write-Warning $message
-                $log_content | Write-Warning
-                if($AddForRerun) {
-                    $Rerun.Add($completed.Name, $log)
-                }
-            } else {
-                Write-Host $message
-            }
-        }
-
-        # Remove cached job info from the system
-        # Should be no output
-        Receive-Job -Job $completed | Out-Null
-    }
-}
-
-RunJobs -TestToLog $TestsToRun -ConcurrencyVal $Concurrency -AddForRerun $EnableRerun
-
-if($Rerun.Count -gt 0) {
-    Write-Host "Rerunning " ($Rerun.Count) " tests sequentially"
-    $success = $true
-    $count = 0
-    RunJobs -TestToLog $Rerun -ConcurrencyVal 1 -AddForRerun $false
-}
-
-Get-Date
-
-
-if(!$success) {
-# This does not succeed killing off jobs quick
-# So we simply exit
-#    Remove-Job -Job $jobs -Force
-# indicate failure using this exit code
-    exit 12345
- }
-
- 
@@ -1,127 +0,0 @@
-#!/bin/sh
-#
-# Update dependencies.sh file with the latest avaliable versions
-
-BASEDIR=$(dirname $0)
-OUTPUT=""
-
-function log_variable()
-{
-  echo "$1=${!1}" >> "$OUTPUT"
-}
-
-
-TP2_LATEST="/mnt/vol/engshare/fbcode/third-party2"
-## $1 => lib name
-## $2 => lib version (if not provided, will try to pick latest)
-## $3 => platform (if not provided, will try to pick latest gcc)
-##
-## get_lib_base will set a variable named ${LIB_NAME}_BASE to the lib location
-function get_lib_base()
-{
-  local lib_name=$1
-  local lib_version=$2
-  local lib_platform=$3
-
-  local result="$TP2_LATEST/$lib_name/"
-  
-  # Lib Version
-  if [ -z "$lib_version" ] || [ "$lib_version" = "LATEST" ]; then
-    # version is not provided, use latest
-    result=`ls -dr1v $result/*/ | head -n1`
-  else
-    result="$result/$lib_version/"
-  fi
-  
-  # Lib Platform
-  if [ -z "$lib_platform" ]; then
-    # platform is not provided, use latest gcc
-    result=`ls -dr1v $result/gcc-*[^fb]/ | head -n1`
-  else
-    result="$result/$lib_platform/"
-  fi
-  
-  result=`ls -1d $result/*/ | head -n1`
-  
-  # lib_name => LIB_NAME_BASE
-  local __res_var=${lib_name^^}"_BASE"
-  __res_var=`echo $__res_var | tr - _`
-  # LIB_NAME_BASE=$result
-  eval $__res_var=`readlink -f $result`
-  
-  log_variable $__res_var
-}
-
-###########################################################
-#                   4.9.x dependencies                    #
-###########################################################
-
-OUTPUT="$BASEDIR/dependencies.sh"
-
-rm -f "$OUTPUT"
-touch "$OUTPUT"
-
-echo "Writing dependencies to $OUTPUT"
-
-# Compilers locations
-GCC_BASE=`readlink -f $TP2_LATEST/gcc/4.9.x/centos6-native/*/`
-CLANG_BASE=`readlink -f $TP2_LATEST/llvm-fb/stable/centos6-native/*/`
-
-log_variable GCC_BASE
-log_variable CLANG_BASE
-
-# Libraries locations
-get_lib_base libgcc     4.9.x
-get_lib_base glibc      2.20 
-get_lib_base snappy     LATEST
-get_lib_base zlib       LATEST
-get_lib_base bzip2      LATEST
-get_lib_base lz4        LATEST
-get_lib_base zstd       LATEST
-get_lib_base gflags     LATEST
-get_lib_base jemalloc   LATEST
-get_lib_base numa       LATEST
-get_lib_base libunwind  LATEST
-
-get_lib_base kernel-headers LATEST 
-get_lib_base binutils   LATEST centos6-native 
-get_lib_base valgrind   LATEST
-
-git diff $OUTPUT
-
-###########################################################
-#                   4.8.1 dependencies                    #
-###########################################################
-
-OUTPUT="$BASEDIR/dependencies_4.8.1.sh"
-
-rm -f "$OUTPUT"
-touch "$OUTPUT"
-
-echo "Writing 4.8.1 dependencies to $OUTPUT"
-
-# Compilers locations
-GCC_BASE=`readlink -f $TP2_LATEST/gcc/4.8.1/centos6-native/*/`
-CLANG_BASE=`readlink -f $TP2_LATEST/llvm-fb/stable/centos6-native/*/`
-
-log_variable GCC_BASE
-log_variable CLANG_BASE
-
-# Libraries locations
-get_lib_base libgcc     4.8.1  gcc-4.8.1-glibc-2.17
-get_lib_base glibc      2.17   gcc-4.8.1-glibc-2.17  
-get_lib_base snappy     LATEST gcc-4.8.1-glibc-2.17
-get_lib_base zlib       LATEST gcc-4.8.1-glibc-2.17
-get_lib_base bzip2      LATEST gcc-4.8.1-glibc-2.17
-get_lib_base lz4        LATEST gcc-4.8.1-glibc-2.17
-get_lib_base zstd       LATEST gcc-4.8.1-glibc-2.17
-get_lib_base gflags     LATEST gcc-4.8.1-glibc-2.17
-get_lib_base jemalloc   LATEST gcc-4.8.1-glibc-2.17
-get_lib_base numa       LATEST gcc-4.8.1-glibc-2.17
-get_lib_base libunwind  LATEST gcc-4.8.1-glibc-2.17
-
-get_lib_base kernel-headers LATEST gcc-4.8.1-glibc-2.17 
-get_lib_base binutils   LATEST centos6-native 
-get_lib_base valgrind   3.8.1  gcc-4.8.1-glibc-2.17
-
-git diff $OUTPUT
@@ -0,0 +1,15 @@
+#!/bin/bash
+#A shell script for Jenknis to run valgrind on rocksdb tests
+#Returns 0 on success when there are no failed tests 
+
+VALGRIND_DIR=build_tools/VALGRIND_LOGS
+make clean
+make -j$(nproc) valgrind_check
+NUM_FAILED_TESTS=$((`wc -l $VALGRIND_DIR/valgrind_failed_tests | awk '{print $1}'` - 1))
+if [ $NUM_FAILED_TESTS -lt 1 ]; then
+  echo No tests have valgrind errors
+  exit 0
+else
+  cat $VALGRIND_DIR/valgrind_failed_tests
+  exit 1
+fi
@@ -26,7 +26,6 @@
 #include "rocksdb/options.h"
 #include "rocksdb/table.h"
 #include "table/block_based_table_builder.h"
-#include "table/internal_iterator.h"
 #include "util/file_reader_writer.h"
 #include "util/iostats_context_imp.h"
 #include "util/stop_watch.h"
@@ -41,26 +40,22 @@ TableBuilder* NewTableBuilder(
    const InternalKeyComparator& internal_comparator,
    const std::vector<std::unique_ptr<IntTblPropCollectorFactory>>*
        int_tbl_prop_collector_factories,
-    uint32_t column_family_id, WritableFileWriter* file,
-    const CompressionType compression_type,
+    WritableFileWriter* file, const CompressionType compression_type,
    const CompressionOptions& compression_opts, const bool skip_filters) {
  return ioptions.table_factory->NewTableBuilder(
      TableBuilderOptions(ioptions, internal_comparator,
                          int_tbl_prop_collector_factories, compression_type,
                          compression_opts, skip_filters),
-      column_family_id, file);
+      file);
 }

 Status BuildTable(
    const std::string& dbname, Env* env, const ImmutableCFOptions& ioptions,
-    const EnvOptions& env_options, TableCache* table_cache,
-    InternalIterator* iter, FileMetaData* meta,
-    const InternalKeyComparator& internal_comparator,
+    const EnvOptions& env_options, TableCache* table_cache, Iterator* iter,
+    FileMetaData* meta, const InternalKeyComparator& internal_comparator,
    const std::vector<std::unique_ptr<IntTblPropCollectorFactory>>*
        int_tbl_prop_collector_factories,
-    uint32_t column_family_id, std::vector<SequenceNumber> snapshots,
-    SequenceNumber earliest_write_conflict_snapshot,
-    const CompressionType compression,
+    std::vector<SequenceNumber> snapshots, const CompressionType compression,
    const CompressionOptions& compression_opts, bool paranoid_file_checks,
    InternalStats* internal_stats, const Env::IOPriority io_priority,
    TableProperties* table_properties) {
@@ -77,7 +72,7 @@ Status BuildTable(
    unique_ptr<WritableFileWriter> file_writer;
    {
      unique_ptr<WritableFile> file;
-      s = NewWritableFile(env, fname, &file, env_options);
+      s = env->NewWritableFile(fname, &file, env_options);
      if (!s.ok()) {
        return s;
      }
@@ -87,18 +82,16 @@ Status BuildTable(

      builder = NewTableBuilder(
          ioptions, internal_comparator, int_tbl_prop_collector_factories,
-          column_family_id, file_writer.get(), compression, compression_opts);
+          file_writer.get(), compression, compression_opts);
    }

-    MergeHelper merge(env, internal_comparator.user_comparator(),
-                      ioptions.merge_operator, nullptr, ioptions.info_log,
+    MergeHelper merge(internal_comparator.user_comparator(),
+                      ioptions.merge_operator, ioptions.info_log,
                      ioptions.min_partial_merge_operands,
-                      true /* internal key corruption is not ok */,
-                      snapshots.empty() ? 0 : snapshots.back());
+                      true /* internal key corruption is not ok */);

    CompactionIterator c_iter(iter, internal_comparator.user_comparator(),
-                              &merge, kMaxSequenceNumber, &snapshots,
-                              earliest_write_conflict_snapshot, env,
+                              &merge, kMaxSequenceNumber, &snapshots, env,
                              true /* internal key corruption is not ok */);
    c_iter.SeekToFirst();
    for (; c_iter.Valid(); c_iter.Next()) {
@@ -116,15 +109,13 @@ Status BuildTable(
    }

    // Finish and check for builder errors
-    bool empty = builder->NumEntries() == 0;
    s = c_iter.status();
-    if (!s.ok() || empty) {
-      builder->Abandon();
-    } else {
+    if (s.ok()) {
      s = builder->Finish();
+    } else {
+      builder->Abandon();
    }
-
-    if (s.ok() && !empty) {
+    if (s.ok()) {
      meta->fd.file_size = builder->FileSize();
      meta->marked_for_compaction = builder->NeedCompact();
      assert(meta->fd.GetFileSize() > 0);
@@ -135,27 +126,28 @@ Status BuildTable(
    delete builder;

    // Finish and check for file errors
-    if (s.ok() && !empty && !ioptions.disable_data_sync) {
+    if (s.ok() && !ioptions.disable_data_sync) {
      StopWatch sw(env, ioptions.statistics, TABLE_SYNC_MICROS);
      file_writer->Sync(ioptions.use_fsync);
    }
-    if (s.ok() && !empty) {
+    if (s.ok()) {
      s = file_writer->Close();
    }

-    if (s.ok() && !empty) {
+    if (s.ok()) {
      // Verify that the table is usable
-      std::unique_ptr<InternalIterator> it(table_cache->NewIterator(
+      Iterator* it = table_cache->NewIterator(
          ReadOptions(), env_options, internal_comparator, meta->fd, nullptr,
          (internal_stats == nullptr) ? nullptr
                                      : internal_stats->GetFileReadHist(0),
-          false));
+          false);
      s = it->status();
      if (s.ok() && paranoid_file_checks) {
-        for (it->SeekToFirst(); it->Valid(); it->Next()) {
-        }
+        for (it->SeekToFirst(); it->Valid(); it->Next()) {}
        s = it->status();
      }
+
+      delete it;
    }
  }

@@ -164,7 +156,9 @@ Status BuildTable(
    s = iter->status();
  }

-  if (!s.ok() || meta->fd.GetFileSize() == 0) {
+  if (s.ok() && meta->fd.GetFileSize() > 0) {
+    // Keep it
+  } else {
    env->DeleteFile(fname);
  }
  return s;
@@ -31,15 +31,13 @@ class VersionEdit;
 class TableBuilder;
 class WritableFileWriter;
 class InternalStats;
-class InternalIterator;

 TableBuilder* NewTableBuilder(
    const ImmutableCFOptions& options,
    const InternalKeyComparator& internal_comparator,
    const std::vector<std::unique_ptr<IntTblPropCollectorFactory>>*
        int_tbl_prop_collector_factories,
-    uint32_t column_family_id, WritableFileWriter* file,
-    const CompressionType compression_type,
+    WritableFileWriter* file, const CompressionType compression_type,
    const CompressionOptions& compression_opts,
    const bool skip_filters = false);

@@ -50,14 +48,11 @@ TableBuilder* NewTableBuilder(
 // zero, and no Table file will be produced.
 extern Status BuildTable(
    const std::string& dbname, Env* env, const ImmutableCFOptions& options,
-    const EnvOptions& env_options, TableCache* table_cache,
-    InternalIterator* iter, FileMetaData* meta,
-    const InternalKeyComparator& internal_comparator,
+    const EnvOptions& env_options, TableCache* table_cache, Iterator* iter,
+    FileMetaData* meta, const InternalKeyComparator& internal_comparator,
    const std::vector<std::unique_ptr<IntTblPropCollectorFactory>>*
        int_tbl_prop_collector_factories,
-    uint32_t column_family_id, std::vector<SequenceNumber> snapshots,
-    SequenceNumber earliest_write_conflict_snapshot,
-    const CompressionType compression,
+    std::vector<SequenceNumber> snapshots, const CompressionType compression,
    const CompressionOptions& compression_opts, bool paranoid_file_checks,
    InternalStats* internal_stats,
    const Env::IOPriority io_priority = Env::IO_HIGH,
@@ -1288,11 +1288,6 @@ void rocksdb_block_based_options_set_cache_index_and_filter_blocks(
  options->rep.cache_index_and_filter_blocks = v;
 }

-void rocksdb_block_based_options_set_skip_table_builder_flush(
-    rocksdb_block_based_table_options_t* options, unsigned char v) {
-  options->rep.skip_table_builder_flush = v;
-}
-
 void rocksdb_options_set_block_based_table_factory(
    rocksdb_options_t *opt,
    rocksdb_block_based_table_options_t* table_options) {
@@ -1692,11 +1687,6 @@ void rocksdb_options_set_keep_log_file_num(rocksdb_options_t* opt, size_t v) {
  opt->rep.keep_log_file_num = v;
 }

-void rocksdb_options_set_recycle_log_file_num(rocksdb_options_t* opt,
-                                              size_t v) {
-  opt->rep.recycle_log_file_num = v;
-}
-
 void rocksdb_options_set_soft_rate_limit(rocksdb_options_t* opt, double v) {
  opt->rep.soft_rate_limit = v;
 }
@@ -1766,11 +1756,6 @@ void rocksdb_options_set_memtable_prefix_bloom_probes(
  opt->rep.memtable_prefix_bloom_probes = v;
 }

-void rocksdb_options_set_memtable_prefix_bloom_huge_page_tlb_size(
-    rocksdb_options_t* opt, size_t v) {
-  opt->rep.memtable_prefix_bloom_huge_page_tlb_size = v;
-}
-
 void rocksdb_options_set_hash_skip_list_rep(
    rocksdb_options_t *opt, size_t bucket_count,
    int32_t skiplist_height, int32_t skiplist_branching_factor) {
@@ -1970,7 +1955,7 @@ void rocksdb_filterpolicy_destroy(rocksdb_filterpolicy_t* filter) {
  delete filter;
 }

-rocksdb_filterpolicy_t* rocksdb_filterpolicy_create_bloom_format(int bits_per_key, bool original_format) {
+rocksdb_filterpolicy_t* rocksdb_filterpolicy_create_bloom(int bits_per_key) {
  // Make a rocksdb_filterpolicy_t, but override all of its methods so
  // they delegate to a NewBloomFilterPolicy() instead of user
  // supplied C functions.
@@ -1988,21 +1973,13 @@ rocksdb_filterpolicy_t* rocksdb_filterpolicy_create_bloom_format(int bits_per_ke
    static void DoNothing(void*) { }
  };
  Wrapper* wrapper = new Wrapper;
-  wrapper->rep_ = NewBloomFilterPolicy(bits_per_key, original_format);
+  wrapper->rep_ = NewBloomFilterPolicy(bits_per_key);
  wrapper->state_ = nullptr;
  wrapper->delete_filter_ = nullptr;
  wrapper->destructor_ = &Wrapper::DoNothing;
  return wrapper;
 }

-rocksdb_filterpolicy_t* rocksdb_filterpolicy_create_bloom_full(int bits_per_key) {
-  return rocksdb_filterpolicy_create_bloom_format(bits_per_key, false);
-}
-
-rocksdb_filterpolicy_t* rocksdb_filterpolicy_create_bloom(int bits_per_key) {
-  return rocksdb_filterpolicy_create_bloom_format(bits_per_key, true);
-}
-
 rocksdb_mergeoperator_t* rocksdb_mergeoperator_create(
    void* state, void (*destructor)(void*),
    char* (*full_merge)(void*, const char* key, size_t key_length,
@@ -27,9 +27,9 @@
 #include "db/version_set.h"
 #include "db/write_controller.h"
 #include "db/writebuffer.h"
-#include "memtable/hash_skiplist_rep.h"
 #include "util/autovector.h"
 #include "util/compression.h"
+#include "util/hash_skiplist_rep.h"
 #include "util/options_helper.h"
 #include "util/thread_status_util.h"
 #include "util/xfunc.h"
@@ -68,20 +68,6 @@ const std::string& ColumnFamilyHandleImpl::GetName() const {
  return cfd()->GetName();
 }

-Status ColumnFamilyHandleImpl::GetDescriptor(ColumnFamilyDescriptor* desc) {
-#ifndef ROCKSDB_LITE
-  // accessing mutable cf-options requires db mutex.
-  InstrumentedMutexLock l(mutex_);
-  *desc = ColumnFamilyDescriptor(
-      cfd()->GetName(),
-      BuildColumnFamilyOptions(*cfd()->options(),
-                               *cfd()->GetLatestMutableCFOptions()));
-  return Status::OK();
-#else
-  return Status::NotSupported();
-#endif  // !ROCKSDB_LITE
-}
-
 const Comparator* ColumnFamilyHandleImpl::user_comparator() const {
  return cfd()->user_comparator();
 }
@@ -124,24 +110,6 @@ Status CheckCompressionSupported(const ColumnFamilyOptions& cf_options) {
  return Status::OK();
 }

-Status CheckConcurrentWritesSupported(const ColumnFamilyOptions& cf_options) {
-  if (cf_options.inplace_update_support) {
-    return Status::InvalidArgument(
-        "In-place memtable updates (inplace_update_support) is not compatible "
-        "with concurrent writes (allow_concurrent_memtable_write)");
-  }
-  if (cf_options.filter_deletes) {
-    return Status::InvalidArgument(
-        "Delete filtering (filter_deletes) is not compatible with concurrent "
-        "memtable writes (allow_concurrent_memtable_writes)");
-  }
-  if (!cf_options.memtable_factory->IsInsertConcurrentlySupported()) {
-    return Status::InvalidArgument(
-        "Memtable doesn't concurrent writes (allow_concurrent_memtable_write)");
-  }
-  return Status::OK();
-}
-
 ColumnFamilyOptions SanitizeOptions(const DBOptions& db_options,
                                    const InternalKeyComparator* icmp,
                                    const ColumnFamilyOptions& src) {
@@ -207,12 +175,6 @@ ColumnFamilyOptions SanitizeOptions(const DBOptions& db_options,
    result.level0_stop_writes_trigger = std::numeric_limits<int>::max();
  }

-  if (result.level0_file_num_compaction_trigger == 0) {
-    Warn(db_options.info_log.get(),
-         "level0_file_num_compaction_trigger cannot be 0");
-    result.level0_file_num_compaction_trigger = 1;
-  }
-
  if (result.level0_stop_writes_trigger <
          result.level0_slowdown_writes_trigger ||
      result.level0_slowdown_writes_trigger <
@@ -347,8 +309,7 @@ ColumnFamilyData::ColumnFamilyData(
      log_number_(0),
      column_family_set_(column_family_set),
      pending_flush_(false),
-      pending_compaction_(false),
-      prev_compaction_needed_bytes_(0) {
+      pending_compaction_(false) {
  Ref();

  // Convert user defined table properties collector factories to internal ones.
@@ -466,66 +427,13 @@ void ColumnFamilyData::SetDropped() {
  column_family_set_->RemoveColumnFamily(this);
 }

-const double kSlowdownRatio = 1.2;
-
-namespace {
-std::unique_ptr<WriteControllerToken> SetupDelay(
-    uint64_t max_write_rate, WriteController* write_controller,
-    uint64_t compaction_needed_bytes, uint64_t prev_compaction_neeed_bytes,
-    bool auto_comapctions_disabled) {
-  const uint64_t kMinWriteRate = 1024u;  // Minimum write rate 1KB/s.
-
-  uint64_t write_rate = write_controller->delayed_write_rate();
-
-  if (auto_comapctions_disabled) {
-    // When auto compaction is disabled, always use the value user gave.
-    write_rate = max_write_rate;
-  } else if (write_controller->NeedsDelay() && max_write_rate > kMinWriteRate) {
-    // If user gives rate less than kMinWriteRate, don't adjust it.
-    //
-    // If already delayed, need to adjust based on previous compaction debt.
-    // When there are two or more column families require delay, we always
-    // increase or reduce write rate based on information for one single
-    // column family. It is likely to be OK but we can improve if there is a
-    // problem.
-    // Ignore compaction_needed_bytes = 0 case because compaction_needed_bytes
-    // is only available in level-based compaction
-    //
-    // If the compaction debt stays the same as previously, we also further slow
-    // down. It usually means a mem table is full. It's mainly for the case
-    // where both of flush and compaction are much slower than the speed we
-    // insert to mem tables, so we need to actively slow down before we get
-    // feedback signal from compaction and flushes to avoid the full stop
-    // because of hitting the max write buffer number.
-    if (prev_compaction_neeed_bytes > 0 &&
-        prev_compaction_neeed_bytes <= compaction_needed_bytes) {
-      write_rate = static_cast<uint64_t>(static_cast<double>(write_rate) /
-                                         kSlowdownRatio);
-      if (write_rate < kMinWriteRate) {
-        write_rate = kMinWriteRate;
-      }
-    } else if (prev_compaction_neeed_bytes > compaction_needed_bytes) {
-      // We are speeding up by ratio of kSlowdownRatio when we have paid
-      // compaction debt. But we'll never speed up to faster than the write rate
-      // given by users.
-      write_rate = static_cast<uint64_t>(static_cast<double>(write_rate) *
-                                         kSlowdownRatio);
-      if (write_rate > max_write_rate) {
-        write_rate = max_write_rate;
-      }
-    }
-  }
-  return write_controller->GetDelayToken(write_rate);
-}
-}  // namespace
-
 void ColumnFamilyData::RecalculateWriteStallConditions(
      const MutableCFOptions& mutable_cf_options) {
  if (current_ != nullptr) {
    auto* vstorage = current_->storage_info();
+    const double score = vstorage->max_compaction_score();
+    const int max_level = vstorage->max_compaction_score_level();
    auto write_controller = column_family_set_->write_controller_;
-    uint64_t compaction_needed_bytes =
-        vstorage->estimated_compaction_needed_bytes();

    if (imm()->NumNotFlushed() >= mutable_cf_options.max_write_buffer_number) {
      write_controller_token_ = write_controller->GetStopToken();
@@ -535,77 +443,31 @@ void ColumnFamilyData::RecalculateWriteStallConditions(
          "(waiting for flush), max_write_buffer_number is set to %d",
          name_.c_str(), imm()->NumNotFlushed(),
          mutable_cf_options.max_write_buffer_number);
-    } else if (mutable_cf_options.max_write_buffer_number > 3 &&
-               imm()->NumNotFlushed() >=
-                   mutable_cf_options.max_write_buffer_number - 1) {
-      write_controller_token_ =
-          SetupDelay(ioptions_.delayed_write_rate, write_controller,
-                     compaction_needed_bytes, prev_compaction_needed_bytes_,
-                     mutable_cf_options.disable_auto_compactions);
-      internal_stats_->AddCFStats(InternalStats::MEMTABLE_SLOWDOWN, 1);
-      Log(InfoLogLevel::WARN_LEVEL, ioptions_.info_log,
-          "[%s] Stalling writes because we have %d immutable memtables "
-          "(waiting for flush), max_write_buffer_number is set to %d "
-          "rate %" PRIu64,
-          name_.c_str(), imm()->NumNotFlushed(),
-          mutable_cf_options.max_write_buffer_number,
-          write_controller->delayed_write_rate());
    } else if (vstorage->l0_delay_trigger_count() >=
               mutable_cf_options.level0_stop_writes_trigger) {
      write_controller_token_ = write_controller->GetStopToken();
-      internal_stats_->AddCFStats(InternalStats::LEVEL0_NUM_FILES_TOTAL, 1);
-      if (compaction_picker_->IsLevel0CompactionInProgress()) {
-        internal_stats_->AddCFStats(
-            InternalStats::LEVEL0_NUM_FILES_WITH_COMPACTION, 1);
-      }
+      internal_stats_->AddCFStats(InternalStats::LEVEL0_NUM_FILES, 1);
      Log(InfoLogLevel::WARN_LEVEL, ioptions_.info_log,
          "[%s] Stopping writes because we have %d level-0 files",
          name_.c_str(), vstorage->l0_delay_trigger_count());
-    } else if (mutable_cf_options.hard_pending_compaction_bytes_limit > 0 &&
-               compaction_needed_bytes >=
-                   mutable_cf_options.hard_pending_compaction_bytes_limit) {
-      write_controller_token_ = write_controller->GetStopToken();
-      internal_stats_->AddCFStats(
-          InternalStats::HARD_PENDING_COMPACTION_BYTES_LIMIT, 1);
-      Log(InfoLogLevel::WARN_LEVEL, ioptions_.info_log,
-          "[%s] Stopping writes because of estimated pending compaction "
-          "bytes %" PRIu64,
-          name_.c_str(), compaction_needed_bytes);
    } else if (mutable_cf_options.level0_slowdown_writes_trigger >= 0 &&
               vstorage->l0_delay_trigger_count() >=
                   mutable_cf_options.level0_slowdown_writes_trigger) {
-      write_controller_token_ =
-          SetupDelay(ioptions_.delayed_write_rate, write_controller,
-                     compaction_needed_bytes, prev_compaction_needed_bytes_,
-                     mutable_cf_options.disable_auto_compactions);
-      internal_stats_->AddCFStats(InternalStats::LEVEL0_SLOWDOWN_TOTAL, 1);
-      if (compaction_picker_->IsLevel0CompactionInProgress()) {
-        internal_stats_->AddCFStats(
-            InternalStats::LEVEL0_SLOWDOWN_WITH_COMPACTION, 1);
-      }
+      write_controller_token_ = write_controller->GetDelayToken();
+      internal_stats_->AddCFStats(InternalStats::LEVEL0_SLOWDOWN, 1);
      Log(InfoLogLevel::WARN_LEVEL, ioptions_.info_log,
-          "[%s] Stalling writes because we have %d level-0 files "
-          "rate %" PRIu64,
-          name_.c_str(), vstorage->l0_delay_trigger_count(),
-          write_controller->delayed_write_rate());
-    } else if (mutable_cf_options.soft_pending_compaction_bytes_limit > 0 &&
-               vstorage->estimated_compaction_needed_bytes() >=
-                   mutable_cf_options.soft_pending_compaction_bytes_limit) {
-      write_controller_token_ =
-          SetupDelay(ioptions_.delayed_write_rate, write_controller,
-                     compaction_needed_bytes, prev_compaction_needed_bytes_,
-                     mutable_cf_options.disable_auto_compactions);
-      internal_stats_->AddCFStats(
-          InternalStats::SOFT_PENDING_COMPACTION_BYTES_LIMIT, 1);
+          "[%s] Stalling writes because we have %d level-0 files",
+          name_.c_str(), vstorage->l0_delay_trigger_count());
+    } else if (mutable_cf_options.soft_rate_limit > 0.0 &&
+               score > mutable_cf_options.soft_rate_limit) {
+      write_controller_token_ = write_controller->GetDelayToken();
+      internal_stats_->RecordLevelNSlowdown(max_level, true);
      Log(InfoLogLevel::WARN_LEVEL, ioptions_.info_log,
-          "[%s] Stalling writes because of estimated pending compaction "
-          "bytes %" PRIu64 " rate %" PRIu64,
-          name_.c_str(), vstorage->estimated_compaction_needed_bytes(),
-          write_controller->delayed_write_rate());
+          "[%s] Stalling writes because we hit soft limit on level %d",
+          name_.c_str(), max_level);
    } else {
      write_controller_token_.reset();
    }
-    prev_compaction_needed_bytes_ = compaction_needed_bytes;
  }
 }

@@ -659,12 +521,13 @@ const int ColumnFamilyData::kCompactAllLevels = -1;
 const int ColumnFamilyData::kCompactToBaseLevel = -2;

 Compaction* ColumnFamilyData::CompactRange(
-    const MutableCFOptions& mutable_cf_options, int input_level,
-    int output_level, uint32_t output_path_id, const InternalKey* begin,
-    const InternalKey* end, InternalKey** compaction_end, bool* conflict) {
+    const MutableCFOptions& mutable_cf_options,
+    int input_level, int output_level, uint32_t output_path_id,
+    const InternalKey* begin, const InternalKey* end,
+    InternalKey** compaction_end) {
  auto* result = compaction_picker_->CompactRange(
      GetName(), mutable_cf_options, current_->storage_info(), input_level,
-      output_level, output_path_id, begin, end, compaction_end, conflict);
+      output_level, output_path_id, begin, end, compaction_end);
  if (result != nullptr) {
    result->SetInputVersion(current_);
  }
@@ -950,6 +813,13 @@ ColumnFamilyHandle* ColumnFamilyMemTablesImpl::GetColumnFamilyHandle() {
  return &handle_;
 }

+void ColumnFamilyMemTablesImpl::CheckMemtableFull() {
+  if (current_ != nullptr && current_->mem()->ShouldScheduleFlush()) {
+    flush_scheduler_->ScheduleFlush(current_);
+    current_->mem()->MarkFlushScheduled();
+  }
+}
+
 uint32_t GetColumnFamilyID(ColumnFamilyHandle* column_family) {
  uint32_t column_family_id = 0;
  if (column_family != nullptr) {
@@ -19,10 +19,12 @@
 #include "db/write_controller.h"
 #include "db/table_cache.h"
 #include "db/table_properties_collector.h"
+#include "db/flush_scheduler.h"
 #include "rocksdb/compaction_job_stats.h"
 #include "rocksdb/db.h"
 #include "rocksdb/env.h"
 #include "rocksdb/options.h"
+#include "util/instrumented_mutex.h"
 #include "util/mutable_cf_options.h"
 #include "util/thread_local.h"

@@ -42,8 +44,6 @@ class LogBuffer;
 class InstrumentedMutex;
 class InstrumentedMutexLock;

-extern const double kSlowdownRatio;
-
 // ColumnFamilyHandleImpl is the class that clients use to access different
 // column families. It has non-trivial destructor, which gets called when client
 // is done using the column family
@@ -59,7 +59,6 @@ class ColumnFamilyHandleImpl : public ColumnFamilyHandle {

  virtual uint32_t GetID() const override;
  virtual const std::string& GetName() const override;
-  virtual Status GetDescriptor(ColumnFamilyDescriptor* desc) override;

 private:
  ColumnFamilyData* cfd_;
@@ -133,9 +132,6 @@ struct SuperVersion {

 extern Status CheckCompressionSupported(const ColumnFamilyOptions& cf_options);

-extern Status CheckConcurrentWritesSupported(
-    const ColumnFamilyOptions& cf_options);
-
 extern ColumnFamilyOptions SanitizeOptions(const DBOptions& db_options,
                                           const InternalKeyComparator* icmp,
                                           const ColumnFamilyOptions& src);
@@ -160,16 +156,14 @@ class ColumnFamilyData {
  // thread-safe
  const std::string& GetName() const { return name_; }

-  // Ref() can only be called from a context where the caller can guarantee
-  // that ColumnFamilyData is alive (while holding a non-zero ref already,
-  // holding a DB mutex, or as the leader in a write batch group).
+  // Ref() can only be called whily holding a DB mutex or during a
+  // single-threaded write.
  void Ref() { refs_.fetch_add(1, std::memory_order_relaxed); }
-
-  // Unref decreases the reference count, but does not handle deletion
-  // when the count goes to 0.  If this method returns true then the
-  // caller should delete the instance immediately, or later, by calling
-  // FreeDeadColumnFamilies().  Unref() can only be called while holding
-  // a DB mutex, or during single-threaded recovery.
+  // will just decrease reference count to 0, but will not delete it. returns
+  // true if the ref count was decreased to zero. in that case, it can be
+  // deleted by the caller immediately, or later, by calling
+  // FreeDeadColumnFamilies()
+  // Unref() can only be called while holding a DB mutex
  bool Unref() {
    int old_refs = refs_.fetch_sub(1, std::memory_order_relaxed);
    assert(old_refs > 0);
@@ -230,7 +224,7 @@ class ColumnFamilyData {
  MemTable* mem() { return mem_; }
  Version* current() { return current_; }
  Version* dummy_versions() { return dummy_versions_; }
-  void SetCurrent(Version* _current);
+  void SetCurrent(Version* current);
  uint64_t GetNumLiveVersions() const;  // REQUIRE: DB mutex held
  uint64_t GetTotalSstFilesSize() const;  // REQUIRE: DB mutex held
  void SetMemtable(MemTable* new_mem) { mem_ = new_mem; }
@@ -255,11 +249,11 @@ class ColumnFamilyData {
  // A flag to tell a manual compaction's output is base level.
  static const int kCompactToBaseLevel;
  // REQUIRES: DB mutex held
-  Compaction* CompactRange(const MutableCFOptions& mutable_cf_options,
-                           int input_level, int output_level,
-                           uint32_t output_path_id, const InternalKey* begin,
-                           const InternalKey* end, InternalKey** compaction_end,
-                           bool* manual_conflict);
+  Compaction* CompactRange(
+      const MutableCFOptions& mutable_cf_options,
+      int input_level, int output_level, uint32_t output_path_id,
+      const InternalKey* begin, const InternalKey* end,
+      InternalKey** compaction_end);

  CompactionPicker* compaction_picker() { return compaction_picker_.get(); }
  // thread-safe
@@ -311,14 +305,6 @@ class ColumnFamilyData {
  bool pending_flush() { return pending_flush_; }
  bool pending_compaction() { return pending_compaction_; }

-  // Recalculate some small conditions, which are changed only during
-  // compaction, adding new memtable and/or
-  // recalculation of compaction score. These values are used in
-  // DBImpl::MakeRoomForWrite function to decide, if it need to make
-  // a write stall
-  void RecalculateWriteStallConditions(
-      const MutableCFOptions& mutable_cf_options);
-
 private:
  friend class ColumnFamilySet;
  ColumnFamilyData(uint32_t id, const std::string& name,
@@ -328,6 +314,14 @@ class ColumnFamilyData {
                   const DBOptions* db_options, const EnvOptions& env_options,
                   ColumnFamilySet* column_family_set);

+  // Recalculate some small conditions, which are changed only during
+  // compaction, adding new memtable and/or
+  // recalculation of compaction score. These values are used in
+  // DBImpl::MakeRoomForWrite function to decide, if it need to make
+  // a write stall
+  void RecalculateWriteStallConditions(
+      const MutableCFOptions& mutable_cf_options);
+
  uint32_t id_;
  const std::string name_;
  Version* dummy_versions_;  // Head of circular doubly-linked list of versions.
@@ -388,8 +382,6 @@ class ColumnFamilyData {
  // If true --> this ColumnFamily is currently present in
  // DBImpl::compaction_queue_
  bool pending_compaction_;
-
-  uint64_t prev_compaction_needed_bytes_;
 };

 // ColumnFamilySet has interesting thread-safety requirements
@@ -501,18 +493,15 @@ class ColumnFamilySet {
 // memtables of different column families (specified by ID in the write batch)
 class ColumnFamilyMemTablesImpl : public ColumnFamilyMemTables {
 public:
-  explicit ColumnFamilyMemTablesImpl(ColumnFamilySet* column_family_set)
-      : column_family_set_(column_family_set), current_(nullptr) {}
-
-  // Constructs a ColumnFamilyMemTablesImpl equivalent to one constructed
-  // with the arguments used to construct *orig.
-  explicit ColumnFamilyMemTablesImpl(ColumnFamilyMemTablesImpl* orig)
-      : column_family_set_(orig->column_family_set_), current_(nullptr) {}
+  explicit ColumnFamilyMemTablesImpl(ColumnFamilySet* column_family_set,
+                                     FlushScheduler* flush_scheduler)
+      : column_family_set_(column_family_set),
+        current_(nullptr),
+        flush_scheduler_(flush_scheduler) {}

  // sets current_ to ColumnFamilyData with column_family_id
  // returns false if column family doesn't exist
-  // REQUIRES: use this function of DBImpl::column_family_memtables_ should be
-  //           under a DB mutex OR from a write thread
+  // REQUIRES: under a DB mutex OR from a write thread
  bool Seek(uint32_t column_family_id) override;

  // Returns log number of the selected column family
@@ -520,23 +509,20 @@ class ColumnFamilyMemTablesImpl : public ColumnFamilyMemTables {
  uint64_t GetLogNumber() const override;

  // REQUIRES: Seek() called first
-  // REQUIRES: use this function of DBImpl::column_family_memtables_ should be
-  //           under a DB mutex OR from a write thread
+  // REQUIRES: under a DB mutex OR from a write thread
  virtual MemTable* GetMemTable() const override;

  // Returns column family handle for the selected column family
-  // REQUIRES: use this function of DBImpl::column_family_memtables_ should be
-  //           under a DB mutex OR from a write thread
+  // REQUIRES: under a DB mutex OR from a write thread
  virtual ColumnFamilyHandle* GetColumnFamilyHandle() override;

-  // Cannot be called while another thread is calling Seek().
-  // REQUIRES: use this function of DBImpl::column_family_memtables_ should be
-  //           under a DB mutex OR from a write thread
-  virtual ColumnFamilyData* current() override { return current_; }
+  // REQUIRES: under a DB mutex OR from a write thread
+  virtual void CheckMemtableFull() override;

 private:
  ColumnFamilySet* column_family_set_;
  ColumnFamilyData* current_;
+  FlushScheduler* flush_scheduler_;
  ColumnFamilyHandleInternal handle_;
 };

@@ -12,7 +12,6 @@
 #include "rocksdb/db.h"
 #include "rocksdb/env.h"
 #include "util/string_util.h"
-#include "util/sync_point.h"
 #include "util/testharness.h"

 namespace rocksdb {
@@ -54,63 +53,6 @@ class FlushedFileCollector : public EventListener {
  std::mutex mutex_;
 };

-TEST_F(CompactFilesTest, L0ConflictsFiles) {
-  Options options;
-  // to trigger compaction more easily
-  const int kWriteBufferSize = 10000;
-  const int kLevel0Trigger = 2;
-  options.create_if_missing = true;
-  options.compaction_style = kCompactionStyleLevel;
-  // Small slowdown and stop trigger for experimental purpose.
-  options.level0_slowdown_writes_trigger = 20;
-  options.level0_stop_writes_trigger = 20;
-  options.level0_stop_writes_trigger = 20;
-  options.write_buffer_size = kWriteBufferSize;
-  options.level0_file_num_compaction_trigger = kLevel0Trigger;
-  options.compression = kNoCompression;
-
-  DB* db = nullptr;
-  DestroyDB(db_name_, options);
-  Status s = DB::Open(options, db_name_, &db);
-  assert(s.ok());
-  assert(db);
-
-  rocksdb::SyncPoint::GetInstance()->LoadDependency({
-      {"CompactFilesImpl:0", "BackgroundCallCompaction:0"},
-      {"BackgroundCallCompaction:1", "CompactFilesImpl:1"},
-  });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
-
-  // create couple files
-  // Background compaction starts and waits in BackgroundCallCompaction:0
-  for (int i = 0; i < kLevel0Trigger * 4; ++i) {
-    db->Put(WriteOptions(), ToString(i), "");
-    db->Put(WriteOptions(), ToString(100 - i), "");
-    db->Flush(FlushOptions());
-  }
-
-  rocksdb::ColumnFamilyMetaData meta;
-  db->GetColumnFamilyMetaData(&meta);
-  std::string file1;
-  for (auto& file : meta.levels[0].files) {
-    ASSERT_EQ(0, meta.levels[0].level);
-    if (file1 == "") {
-      file1 = file.db_path + "/" + file.name;
-    } else {
-      std::string file2 = file.db_path + "/" + file.name;
-      // Another thread starts a compact files and creates an L0 compaction
-      // The background compaction then notices that there is an L0 compaction
-      // already in progress and doesn't do an L0 compaction
-      // Once the background compaction finishes, the compact files finishes
-      ASSERT_OK(
-          db->CompactFiles(rocksdb::CompactionOptions(), {file1, file2}, 0));
-      break;
-    }
-  }
-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
-  delete db;
-}
-
 TEST_F(CompactFilesTest, ObsoleteFiles) {
  Options options;
  // to trigger compaction more easily
@@ -40,47 +40,6 @@ void Compaction::SetInputVersion(Version* _input_version) {
  edit_.SetColumnFamily(cfd_->GetID());
 }

-void Compaction::GetBoundaryKeys(
-    VersionStorageInfo* vstorage,
-    const std::vector<CompactionInputFiles>& inputs, Slice* smallest_user_key,
-    Slice* largest_user_key) {
-  bool initialized = false;
-  const Comparator* ucmp = vstorage->InternalComparator()->user_comparator();
-  for (size_t i = 0; i < inputs.size(); ++i) {
-    if (inputs[i].files.empty()) {
-      continue;
-    }
-    if (inputs[i].level == 0) {
-      // we need to consider all files on level 0
-      for (const auto* f : inputs[i].files) {
-        const Slice& start_user_key = f->smallest.user_key();
-        if (!initialized ||
-            ucmp->Compare(start_user_key, *smallest_user_key) < 0) {
-          *smallest_user_key = start_user_key;
-        }
-        const Slice& end_user_key = f->largest.user_key();
-        if (!initialized ||
-            ucmp->Compare(end_user_key, *largest_user_key) > 0) {
-          *largest_user_key = end_user_key;
-        }
-        initialized = true;
-      }
-    } else {
-      // we only need to consider the first and last file
-      const Slice& start_user_key = inputs[i].files[0]->smallest.user_key();
-      if (!initialized ||
-          ucmp->Compare(start_user_key, *smallest_user_key) < 0) {
-        *smallest_user_key = start_user_key;
-      }
-      const Slice& end_user_key = inputs[i].files.back()->largest.user_key();
-      if (!initialized || ucmp->Compare(end_user_key, *largest_user_key) > 0) {
-        *largest_user_key = end_user_key;
-      }
-      initialized = true;
-    }
-  }
-}
-
 // helper function to determine if compaction is creating files at the
 // bottommost level
 bool Compaction::IsBottommostLevel(
@@ -91,45 +50,20 @@ bool Compaction::IsBottommostLevel(
    return false;
  }

-  Slice smallest_key, largest_key;
-  GetBoundaryKeys(vstorage, inputs, &smallest_key, &largest_key);
-
-  // Checks whether there are files living beyond the output_level.
-  // If lower levels have files, it checks for overlap between files
-  // if the compaction process and those files.
-  // Bottomlevel optimizations can be made if there are no files in
-  // lower levels or if there is no overlap with the files in
-  // the lower levels.
+  // checks whether there are files living beyond the output_level.
  for (int i = output_level + 1; i < vstorage->num_levels(); i++) {
-    // It is not the bottommost level if there are files in higher
-    // levels when the output level is 0 or if there are files in
-    // higher levels which overlap with files to be compacted.
-    // output_level == 0 means that we want it to be considered
-    // s the bottommost level only if the last file on the level
-    // is a part of the files to be compacted - this is verified by
-    // the first if condition in this function
-    if (vstorage->NumLevelFiles(i) > 0 &&
-        (output_level == 0 ||
-         vstorage->OverlapInLevel(i, &smallest_key, &largest_key))) {
+    if (vstorage->NumLevelFiles(i) > 0) {
      return false;
    }
  }
  return true;
 }

-// test function to validate the functionality of IsBottommostLevel()
-// function -- determines if compaction with inputs and storage is bottommost
-bool Compaction::TEST_IsBottommostLevel(
-    int output_level, VersionStorageInfo* vstorage,
-    const std::vector<CompactionInputFiles>& inputs) {
-  return IsBottommostLevel(output_level, vstorage, inputs);
-}
-
 bool Compaction::IsFullCompaction(
    VersionStorageInfo* vstorage,
    const std::vector<CompactionInputFiles>& inputs) {
-  size_t num_files_in_compaction = 0;
-  size_t total_num_files = 0;
+  int num_files_in_compaction = 0;
+  int total_num_files = 0;
  for (int l = 0; l < vstorage->num_levels(); l++) {
    total_num_files += vstorage->NumLevelFiles(l);
  }
@@ -147,8 +81,7 @@ Compaction::Compaction(VersionStorageInfo* vstorage,
                       uint32_t _output_path_id, CompressionType _compression,
                       std::vector<FileMetaData*> _grandparents,
                       bool _manual_compaction, double _score,
-                       bool _deletion_compaction,
-                       CompactionReason _compaction_reason)
+                       bool _deletion_compaction)
    : start_level_(_inputs[0].level),
      output_level_(_output_level),
      max_output_file_size_(_target_file_size),
@@ -168,12 +101,8 @@ Compaction::Compaction(VersionStorageInfo* vstorage,
      score_(_score),
      bottommost_level_(IsBottommostLevel(output_level_, vstorage, inputs_)),
      is_full_compaction_(IsFullCompaction(vstorage, inputs_)),
-      is_manual_compaction_(_manual_compaction),
-      compaction_reason_(_compaction_reason) {
+      is_manual_compaction_(_manual_compaction) {
  MarkFilesBeingCompacted(true);
-  if (is_manual_compaction_) {
-    compaction_reason_ = CompactionReason::kManualCompaction;
-  }

 #ifndef NDEBUG
  for (size_t i = 1; i < inputs_.size(); ++i) {
@@ -189,9 +118,6 @@ Compaction::Compaction(VersionStorageInfo* vstorage,
                                &arena_);
    }
  }
-
-  Slice smallest_user_key;
-  GetBoundaryKeys(vstorage, inputs_, &smallest_user_key, &largest_user_key_);
 }

 Compaction::~Compaction() {
@@ -319,7 +245,7 @@ bool Compaction::ShouldStopBefore(const Slice& internal_key) {
 // Mark (or clear) each file that is being compacted
 void Compaction::MarkFilesBeingCompacted(bool mark_as_compacted) {
  for (size_t i = 0; i < num_input_levels(); i++) {
-    for (size_t j = 0; j < inputs_[i].size(); j++) {
+    for (unsigned int j = 0; j < inputs_[i].size(); j++) {
      assert(mark_as_compacted ? !inputs_[i][j]->being_compacted :
                                  inputs_[i][j]->being_compacted);
      inputs_[i][j]->being_compacted = mark_as_compacted;
@@ -379,7 +305,7 @@ int InputSummary(const std::vector<FileMetaData*>& files, char* output,
                 int len) {
  *output = '\0';
  int write = 0;
-  for (size_t i = 0; i < files.size(); i++) {
+  for (unsigned int i = 0; i < files.size(); i++) {
    int sz = len - write;
    int ret;
    char sztxt[16];
@@ -436,7 +362,7 @@ uint64_t Compaction::OutputFilePreallocationSize() {
  }
  // Over-estimate slightly so we don't end up just barely crossing
  // the threshold
-  return preallocation_size + (preallocation_size / 10);
+  return preallocation_size * 1.1;
 }

 std::unique_ptr<CompactionFilter> Compaction::CreateCompactionFilter() const {
@@ -447,7 +373,6 @@ std::unique_ptr<CompactionFilter> Compaction::CreateCompactionFilter() const {
  CompactionFilter::Context context;
  context.is_full_compaction = is_full_compaction_;
  context.is_manual_compaction = is_manual_compaction_;
-  context.column_family_id = cfd_->GetID();
  return cfd_->ioptions()->compaction_filter_factory->CreateCompactionFilter(
      context);
 }
@@ -41,8 +41,7 @@ class Compaction {
             uint32_t output_path_id, CompressionType compression,
             std::vector<FileMetaData*> grandparents,
             bool manual_compaction = false, double score = -1,
-             bool deletion_compaction = false,
-             CompactionReason compaction_reason = CompactionReason::kUnknown);
+             bool deletion_compaction = false);

  // No copying allowed
  Compaction(const Compaction&) = delete;
@@ -205,39 +204,15 @@ class Compaction {
  // Should this compaction be broken up into smaller ones run in parallel?
  bool ShouldFormSubcompactions() const;

-  // test function to validate the functionality of IsBottommostLevel()
-  // function -- determines if compaction with inputs and storage is bottommost
-  static bool TEST_IsBottommostLevel(
-      int output_level, VersionStorageInfo* vstorage,
-      const std::vector<CompactionInputFiles>& inputs);
-
-  TablePropertiesCollection GetOutputTableProperties() const {
-    return output_table_properties_;
-  }
-
-  void SetOutputTableProperties(TablePropertiesCollection tp) {
-    output_table_properties_ = std::move(tp);
-  }
-
-  Slice GetLargestUserKey() const { return largest_user_key_; }
-
-  CompactionReason compaction_reason() { return compaction_reason_; }
-
 private:
  // mark (or clear) all files that are being compacted
  void MarkFilesBeingCompacted(bool mark_as_compacted);

-  // get the smallest and largest key present in files to be compacted
-  static void GetBoundaryKeys(VersionStorageInfo* vstorage,
-                              const std::vector<CompactionInputFiles>& inputs,
-                              Slice* smallest_key, Slice* largest_key);
-
  // helper function to determine if compaction with inputs and storage is
  // bottommost
  static bool IsBottommostLevel(
      int output_level, VersionStorageInfo* vstorage,
      const std::vector<CompactionInputFiles>& inputs);
-
  static bool IsFullCompaction(VersionStorageInfo* vstorage,
                               const std::vector<CompactionInputFiles>& inputs);

@@ -286,15 +261,6 @@ class Compaction {

  // Does input compression match the output compression?
  bool InputCompressionMatchesOutput() const;
-
-  // table properties of output files
-  TablePropertiesCollection output_table_properties_;
-
-  // largest user keys in compaction
-  Slice largest_user_key_;
-
-  // Reason for compaction
-  CompactionReason compaction_reason_;
 };

 // Utility function
@@ -6,23 +6,22 @@
 //  of patent rights can be found in the PATENTS file in the same directory.

 #include "db/compaction_iterator.h"
-#include "table/internal_iterator.h"

 namespace rocksdb {

 CompactionIterator::CompactionIterator(
-    InternalIterator* input, const Comparator* cmp, MergeHelper* merge_helper,
+    Iterator* input, const Comparator* cmp, MergeHelper* merge_helper,
    SequenceNumber last_sequence, std::vector<SequenceNumber>* snapshots,
-    SequenceNumber earliest_write_conflict_snapshot, Env* env,
-    bool expect_valid_internal_key, Compaction* compaction,
-    const CompactionFilter* compaction_filter, LogBuffer* log_buffer)
+    Env* env, bool expect_valid_internal_key, Statistics* stats,
+    Compaction* compaction, const CompactionFilter* compaction_filter,
+    LogBuffer* log_buffer)
    : input_(input),
      cmp_(cmp),
      merge_helper_(merge_helper),
      snapshots_(snapshots),
-      earliest_write_conflict_snapshot_(earliest_write_conflict_snapshot),
      env_(env),
      expect_valid_internal_key_(expect_valid_internal_key),
+      stats_(stats),
      compaction_(compaction),
      compaction_filter_(compaction_filter),
      log_buffer_(log_buffer),
@@ -44,11 +43,6 @@ CompactionIterator::CompactionIterator(
    earliest_snapshot_ = snapshots_->at(0);
    latest_snapshot_ = snapshots_->back();
  }
-  if (compaction_filter_ != nullptr && compaction_filter_->IgnoreSnapshots()) {
-    ignore_snapshots_ = true;
-  } else {
-    ignore_snapshots_ = false;
-  }
 }

 void CompactionIterator::ResetRecordCounts() {
@@ -77,10 +71,6 @@ void CompactionIterator::Next() {
      // MergeUntil stops when it encounters a corrupt key and does not
      // include them in the result, so we expect the keys here to be valid.
      assert(valid_key);
-      // Keep current_key_ in sync.
-      current_key_.UpdateInternalKey(ikey_.sequence, ikey_.type);
-      key_ = current_key_.GetKey();
-      ikey_.user_key = current_key_.GetUserKey();
      valid_ = true;
    } else {
      // MergeHelper moves the iterator to the first record after the merged
@@ -89,27 +79,18 @@ void CompactionIterator::Next() {
      NextFromInput();
    }
  } else {
-    // Only advance the input iterator if there is no merge output and the
-    // iterator is not already at the next record.
-    if (!at_next_) {
-      input_->Next();
-    }
+    // Only advance the input iterator if there is no merge output.
+    input_->Next();
    NextFromInput();
  }

-  if (valid_) {
-    // Record that we've ouputted a record for the current key.
-    has_outputted_key_ = true;
-  }
-
  PrepareOutput();
 }

 void CompactionIterator::NextFromInput() {
-  at_next_ = false;
  valid_ = false;

-  while (!valid_ && input_->Valid()) {
+  while (input_->Valid()) {
    key_ = input_->key();
    value_ = input_->value();
    iter_stats_.num_input_records++;
@@ -119,11 +100,10 @@ void CompactionIterator::NextFromInput() {
      // and let the caller decide what to do with it.
      // TODO(noetzli): We should have a more elegant solution for this.
      if (expect_valid_internal_key_) {
-        assert(!"Corrupted internal key not expected.");
-        status_ = Status::Corruption("Corrupted internal key not expected.");
+        assert(!"corrupted internal key is not expected");
        break;
      }
-      key_ = current_key_.SetKey(key_);
+      current_user_key_.Clear();
      has_current_user_key_ = false;
      current_user_key_sequence_ = kMaxSequenceNumber;
      current_user_key_snapshot_ = 0;
@@ -133,29 +113,22 @@ void CompactionIterator::NextFromInput() {
    }

    // Update input statistics
-    if (ikey_.type == kTypeDeletion || ikey_.type == kTypeSingleDeletion) {
+    if (ikey_.type == kTypeDeletion) {
      iter_stats_.num_input_deletion_records++;
    }
    iter_stats_.total_input_raw_key_bytes += key_.size();
    iter_stats_.total_input_raw_value_bytes += value_.size();

-    // Check whether the user key changed. After this if statement current_key_
-    // is a copy of the current input key (maybe converted to a delete by the
-    // compaction filter). ikey_.user_key is pointing to the copy.
    if (!has_current_user_key_ ||
-        !cmp_->Equal(ikey_.user_key, current_user_key_)) {
+        cmp_->Compare(ikey_.user_key, current_user_key_.GetKey()) != 0) {
      // First occurrence of this user key
-      key_ = current_key_.SetKey(key_, &ikey_);
-      current_user_key_ = ikey_.user_key;
+      current_user_key_.SetKey(ikey_.user_key);
      has_current_user_key_ = true;
-      has_outputted_key_ = false;
      current_user_key_sequence_ = kMaxSequenceNumber;
      current_user_key_snapshot_ = 0;
-
      // apply the compaction filter to the first occurrence of the user key
      if (compaction_filter_ != nullptr && ikey_.type == kTypeValue &&
-          (visible_at_tip_ || ikey_.sequence > latest_snapshot_ ||
-           ignore_snapshots_)) {
+          (visible_at_tip_ || ikey_.sequence > latest_snapshot_)) {
        // 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,
@@ -172,9 +145,13 @@ void CompactionIterator::NextFromInput() {
              env_ != nullptr ? timer.ElapsedNanos() : 0;
        }
        if (to_delete) {
-          // convert the current key to a delete
-          ikey_.type = kTypeDeletion;
-          current_key_.UpdateInternalKey(ikey_.sequence, kTypeDeletion);
+          // make a copy of the original key and convert it to a delete
+          delete_key_.SetInternalKey(ExtractUserKey(key_), ikey_.sequence,
+                                     kTypeDeletion);
+          // anchor the key again
+          key_ = delete_key_.GetKey();
+          // needed because ikey_ is backed by key
+          ParseInternalKey(key_, &ikey_);
          // no value associated with delete
          value_.clear();
          iter_stats_.num_record_drop_user++;
@@ -182,15 +159,6 @@ void CompactionIterator::NextFromInput() {
          value_ = compaction_filter_value_;
        }
      }
-    } else {
-      // Update the current key to reflect the new sequence number/type without
-      // copying the user key.
-      // TODO(rven): Compaction filter does not process keys in this path
-      // Need to have the compaction filter process multiple versions
-      // if we have versions on both sides of a snapshot
-      current_key_.UpdateInternalKey(ikey_.sequence, ikey_.type);
-      key_ = current_key_.GetKey();
-      ikey_.user_key = current_key_.GetUserKey();
    }

    // If there are no snapshots, then this kv affect visibility at tip.
@@ -205,141 +173,14 @@ void CompactionIterator::NextFromInput() {
        visible_at_tip_ ? visible_at_tip_ : findEarliestVisibleSnapshot(
                                                ikey_.sequence, &prev_snapshot);

-    if (clear_and_output_next_key_) {
-      // In the previous iteration we encountered a single delete that we could
-      // not compact out.  We will keep this Put, but can drop it's data.
-      // (See Optimization 3, below.)
-      assert(ikey_.type == kTypeValue);
-      assert(current_user_key_snapshot_ == last_snapshot);
-
-      value_.clear();
-      valid_ = true;
-      clear_and_output_next_key_ = false;
-    } else if (ikey_.type == kTypeSingleDeletion) {
-      // We can compact out a SingleDelete if:
-      // 1) We encounter the corresponding PUT -OR- we know that this key
-      //    doesn't appear past this output level
-      // =AND=
-      // 2) We've already returned a record in this snapshot -OR-
-      //    there are no earlier earliest_write_conflict_snapshot.
-      //
-      // Rule 1 is needed for SingleDelete correctness.  Rule 2 is needed to
-      // allow Transactions to do write-conflict checking (if we compacted away
-      // all keys, then we wouldn't know that a write happened in this
-      // snapshot).  If there is no earlier snapshot, then we know that there
-      // are no active transactions that need to know about any writes.
-      //
-      // Optimization 3:
-      // If we encounter a SingleDelete followed by a PUT and Rule 2 is NOT
-      // true, then we must output a SingleDelete.  In this case, we will decide
-      // to also output the PUT.  While we are compacting less by outputting the
-      // PUT now, hopefully this will lead to better compaction in the future
-      // when Rule 2 is later true (Ie, We are hoping we can later compact out
-      // both the SingleDelete and the Put, while we couldn't if we only
-      // outputted the SingleDelete now).
-      // In this case, we can save space by removing the PUT's value as it will
-      // never be read.
-      //
-      // Deletes and Merges are not supported on the same key that has a
-      // SingleDelete as it is not possible to correctly do any partial
-      // compaction of such a combination of operations.  The result of mixing
-      // those operations for a given key is documented as being undefined.  So
-      // we can choose how to handle such a combinations of operations.  We will
-      // try to compact out as much as we can in these cases.
-
-      // The easiest way to process a SingleDelete during iteration is to peek
-      // ahead at the next key.
-      ParsedInternalKey next_ikey;
-      input_->Next();
-
-      // Check whether the next key exists, is not corrupt, and is the same key
-      // as the single delete.
-      if (input_->Valid() && ParseInternalKey(input_->key(), &next_ikey) &&
-          cmp_->Equal(ikey_.user_key, next_ikey.user_key)) {
-        // Check whether the next key belongs to the same snapshot as the
-        // SingleDelete.
-        if (prev_snapshot == 0 || next_ikey.sequence > prev_snapshot) {
-          if (next_ikey.type == kTypeSingleDeletion) {
-            // We encountered two SingleDeletes in a row.  This could be due to
-            // unexpected user input.
-            // Skip the first SingleDelete and let the next iteration decide how
-            // to handle the second SingleDelete
-
-            // First SingleDelete has been skipped since we already called
-            // input_->Next().
-            ++iter_stats_.num_record_drop_obsolete;
-          } else if ((ikey_.sequence <= earliest_write_conflict_snapshot_) ||
-                     has_outputted_key_) {
-            // Found a matching value, we can drop the single delete and the
-            // value.  It is safe to drop both records since we've already
-            // outputted a key in this snapshot, or there is no earlier
-            // snapshot (Rule 2 above).
-
-            // Note: it doesn't matter whether the second key is a Put or if it
-            // is an unexpected Merge or Delete.  We will compact it out
-            // either way.
-            ++iter_stats_.num_record_drop_hidden;
-            ++iter_stats_.num_record_drop_obsolete;
-            // Already called input_->Next() once.  Call it a second time to
-            // skip past the second key.
-            input_->Next();
-          } else {
-            // Found a matching value, but we cannot drop both keys since
-            // there is an earlier snapshot and we need to leave behind a record
-            // to know that a write happened in this snapshot (Rule 2 above).
-            // Clear the value and output the SingleDelete. (The value will be
-            // outputted on the next iteration.)
-            ++iter_stats_.num_record_drop_hidden;
-
-            // Setting valid_ to true will output the current SingleDelete
-            valid_ = true;
-
-            // Set up the Put to be outputted in the next iteration.
-            // (Optimization 3).
-            clear_and_output_next_key_ = true;
-          }
-        } else {
-          // We hit the next snapshot without hitting a put, so the iterator
-          // returns the single delete.
-          valid_ = true;
-        }
-      } else {
-        // We are at the end of the input, could not parse the next key, or hit
-        // the next key. The iterator returns the single delete if the key
-        // possibly exists beyond the current output level.  We set
-        // has_current_user_key to false so that if the iterator is at the next
-        // key, we do not compare it again against the previous key at the next
-        // iteration. If the next key is corrupt, we return before the
-        // comparison, so the value of has_current_user_key does not matter.
-        has_current_user_key_ = false;
-        if (compaction_ != nullptr && ikey_.sequence <= earliest_snapshot_ &&
-            compaction_->KeyNotExistsBeyondOutputLevel(ikey_.user_key,
-                                                       &level_ptrs_)) {
-          // Key doesn't exist outside of this range.
-          // Can compact out this SingleDelete.
-          ++iter_stats_.num_record_drop_obsolete;
-        } else {
-          // Output SingleDelete
-          valid_ = true;
-        }
-      }
-
-      if (valid_) {
-        at_next_ = true;
-      }
-    } else if (last_snapshot == current_user_key_snapshot_) {
+    if (last_snapshot == current_user_key_snapshot_) {
      // If the earliest snapshot is which this key is visible in
      // is the same as the visibility of a previous instance of the
      // same key, then this kv is not visible in any snapshot.
      // Hidden by an newer entry for same user key
      // TODO: why not > ?
-      //
-      // Note: Dropping this key will not affect TransactionDB write-conflict
-      // checking since there has already been a record returned for this key
-      // in this snapshot.
      assert(last_sequence >= current_user_key_sequence_);
      ++iter_stats_.num_record_drop_hidden;  // (A)
-      input_->Next();
    } else if (compaction_ != nullptr && ikey_.type == kTypeDeletion &&
               ikey_.sequence <= earliest_snapshot_ &&
               compaction_->KeyNotExistsBeyondOutputLevel(ikey_.user_key,
@@ -355,11 +196,7 @@ void CompactionIterator::NextFromInput() {
      //     smaller sequence numbers will be dropped in the next
      //     few iterations of this loop (by rule (A) above).
      // Therefore this deletion marker is obsolete and can be dropped.
-      //
-      // Note:  Dropping this Delete will not affect TransactionDB
-      // write-conflict checking since it is earlier than any snapshot.
      ++iter_stats_.num_record_drop_obsolete;
-      input_->Next();
    } else if (ikey_.type == kTypeMerge) {
      if (!merge_helper_->HasOperator()) {
        LogToBuffer(log_buffer_, "Options::merge_operator is null.");
@@ -372,33 +209,27 @@ 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_helper_->MergeUntil(input_, prev_snapshot, bottommost_level_);
+      merge_helper_->MergeUntil(input_, prev_snapshot, bottommost_level_,
+                                stats_, env_);
      merge_out_iter_.SeekToFirst();

-      if (merge_out_iter_.Valid()) {
-        // NOTE: key, value, and ikey_ refer to old entries.
-        //       These will be correctly set below.
-        key_ = merge_out_iter_.key();
-        value_ = merge_out_iter_.value();
-        bool valid_key __attribute__((__unused__)) =
-            ParseInternalKey(key_, &ikey_);
-        // MergeUntil stops when it encounters a corrupt key and does not
-        // include them in the result, so we expect the keys here to valid.
-        assert(valid_key);
-        // Keep current_key_ in sync.
-        current_key_.UpdateInternalKey(ikey_.sequence, ikey_.type);
-        key_ = current_key_.GetKey();
-        ikey_.user_key = current_key_.GetUserKey();
-        valid_ = true;
-      } else {
-        // all merge operands were filtered out. reset the user key, since the
-        // batch consumed by the merge operator should not shadow any keys
-        // coming after the merges
-        has_current_user_key_ = false;
-      }
+      // NOTE: key, value, and ikey_ refer to old entries.
+      //       These will be correctly set below.
+      key_ = merge_out_iter_.key();
+      value_ = merge_out_iter_.value();
+      bool valid_key __attribute__((__unused__)) =
+          ParseInternalKey(key_, &ikey_);
+      // MergeUntil stops when it encounters a corrupt key and does not
+      // include them in the result, so we expect the keys here to valid.
+      assert(valid_key);
+      valid_ = true;
+      break;
    } else {
      valid_ = true;
+      break;
    }
+
+    input_->Next();
  }
 }

@@ -406,17 +237,15 @@ void CompactionIterator::PrepareOutput() {
  // Zeroing out the sequence number leads to better compression.
  // If this is the bottommost level (no files in lower levels)
  // and the earliest snapshot is larger than this seqno
-  // and the userkey differs from the last userkey in compaction
  // then we can squash the seqno to zero.
-
-  // This is safe for TransactionDB write-conflict checking since transactions
-  // only care about sequence number larger than any active snapshots.
  if (bottommost_level_ && valid_ && ikey_.sequence < earliest_snapshot_ &&
-      ikey_.type != kTypeMerge &&
-      !cmp_->Equal(compaction_->GetLargestUserKey(), ikey_.user_key)) {
-    assert(ikey_.type != kTypeDeletion && ikey_.type != kTypeSingleDeletion);
-    ikey_.sequence = 0;
-    current_key_.UpdateInternalKey(0, ikey_.type);
+      ikey_.type != kTypeMerge) {
+    assert(ikey_.type != kTypeDeletion);
+    // make a copy because updating in place would cause problems
+    // with the priority queue that is managing the input key iterator
+    updated_key_.assign(key_.data(), key_.size());
+    UpdateInternalKey(&updated_key_, (uint64_t)0, ikey_.type);
+    key_ = Slice(updated_key_);
  }
 }

@@ -37,11 +37,11 @@ struct CompactionIteratorStats {

 class CompactionIterator {
 public:
-  CompactionIterator(InternalIterator* input, const Comparator* cmp,
+  CompactionIterator(Iterator* input, const Comparator* cmp,
                     MergeHelper* merge_helper, SequenceNumber last_sequence,
-                     std::vector<SequenceNumber>* snapshots,
-                     SequenceNumber earliest_write_conflict_snapshot, Env* env,
+                     std::vector<SequenceNumber>* snapshots, Env* env,
                     bool expect_valid_internal_key,
+                     Statistics* stats = nullptr,
                     Compaction* compaction = nullptr,
                     const CompactionFilter* compaction_filter = nullptr,
                     LogBuffer* log_buffer = nullptr);
@@ -64,7 +64,7 @@ class CompactionIterator {
  const Status& status() const { return status_; }
  const ParsedInternalKey& ikey() const { return ikey_; }
  bool Valid() const { return valid_; }
-  const Slice& user_key() const { return current_user_key_; }
+  Slice user_key() const { return current_user_key_.GetKey(); }
  const CompactionIteratorStats& iter_stats() const { return iter_stats_; }

 private:
@@ -85,13 +85,13 @@ class CompactionIterator {
  inline SequenceNumber findEarliestVisibleSnapshot(
      SequenceNumber in, SequenceNumber* prev_snapshot);

-  InternalIterator* input_;
+  Iterator* input_;
  const Comparator* cmp_;
  MergeHelper* merge_helper_;
  const std::vector<SequenceNumber>* snapshots_;
-  const SequenceNumber earliest_write_conflict_snapshot_;
  Env* env_;
  bool expect_valid_internal_key_;
+  Statistics* stats_;
  Compaction* compaction_;
  const CompactionFilter* compaction_filter_;
  LogBuffer* log_buffer_;
@@ -100,43 +100,20 @@ class CompactionIterator {
  SequenceNumber visible_at_tip_;
  SequenceNumber earliest_snapshot_;
  SequenceNumber latest_snapshot_;
-  bool ignore_snapshots_;

  // State
-  //
-  // Points to a copy of the current compaction iterator output (current_key_)
-  // if valid_.
  Slice key_;
-  // Points to the value in the underlying iterator that corresponds to the
-  // current output.
  Slice value_;
-  // The status is OK unless compaction iterator encounters a merge operand
-  // while not having a merge operator defined.
  Status status_;
-  // Stores the user key, sequence number and type of the current compaction
-  // iterator output (or current key in the underlying iterator during
-  // NextFromInput()).
  ParsedInternalKey ikey_;
-  // Stores whether ikey_.user_key is valid. If set to false, the user key is
-  // not compared against the current key in the underlying iterator.
  bool has_current_user_key_ = false;
-  bool at_next_ = false;  // If false, the iterator
-  // Holds a copy of the current compaction iterator output (or current key in
-  // the underlying iterator during NextFromInput()).
-  IterKey current_key_;
-  Slice current_user_key_;
+  IterKey current_user_key_;
  SequenceNumber current_user_key_sequence_;
  SequenceNumber current_user_key_snapshot_;
-
-  // True if the iterator has already returned a record for the current key.
-  bool has_outputted_key_ = false;
-
-  // truncated the value of the next key and output it without applying any
-  // compaction rules.  This is used for outputting a put after a single delete.
-  bool clear_and_output_next_key_ = false;
-
  MergeOutputIterator merge_out_iter_;
+  std::string updated_key_;
  std::string compaction_filter_value_;
+  IterKey delete_key_;
  // "level_ptrs" holds indices that remember which file of an associated
  // level we were last checking during the last call to compaction->
  // KeyNotExistsBeyondOutputLevel(). This allows future calls to the function
@@ -1,71 +0,0 @@
-//  Copyright (c) 2013, Facebook, Inc.  All rights reserved.
-//  This source code is licensed under the BSD-style license found in the
-//  LICENSE file in the root directory of this source tree. An additional grant
-//  of patent rights can be found in the PATENTS file in the same directory.
-
-#include "db/compaction_iterator.h"
-#include "util/testharness.h"
-#include "util/testutil.h"
-
-namespace rocksdb {
-
-class CompactionIteratorTest : public testing::Test {
- public:
-  CompactionIteratorTest() : cmp_(BytewiseComparator()), snapshots_({}) {}
-
-  void InitIterator(const std::vector<std::string>& ks,
-                    const std::vector<std::string>& vs,
-                    SequenceNumber last_sequence) {
-    merge_helper_.reset(new MergeHelper(Env::Default(), cmp_, nullptr, nullptr,
-                                        nullptr, 0U, false, 0));
-    iter_.reset(new test::VectorIterator(ks, vs));
-    iter_->SeekToFirst();
-    c_iter_.reset(new CompactionIterator(
-        iter_.get(), cmp_, merge_helper_.get(), last_sequence, &snapshots_,
-        kMaxSequenceNumber, Env::Default(), false));
-  }
-
-  const Comparator* cmp_;
-  std::vector<SequenceNumber> snapshots_;
-  std::unique_ptr<MergeHelper> merge_helper_;
-  std::unique_ptr<test::VectorIterator> iter_;
-  std::unique_ptr<CompactionIterator> c_iter_;
-};
-
-// It is possible that the output of the compaction iterator is empty even if
-// the input is not.
-TEST_F(CompactionIteratorTest, EmptyResult) {
-  InitIterator({test::KeyStr("a", 5, kTypeSingleDeletion),
-                test::KeyStr("a", 3, kTypeValue)},
-               {"", "val"}, 5);
-  c_iter_->SeekToFirst();
-  ASSERT_FALSE(c_iter_->Valid());
-}
-
-// If there is a corruption after a single deletion, the corrupted key should
-// be preserved.
-TEST_F(CompactionIteratorTest, CorruptionAfterSingleDeletion) {
-  InitIterator({test::KeyStr("a", 5, kTypeSingleDeletion),
-                test::KeyStr("a", 3, kTypeValue, true),
-                test::KeyStr("b", 10, kTypeValue)},
-               {"", "val", "val2"}, 10);
-  c_iter_->SeekToFirst();
-  ASSERT_TRUE(c_iter_->Valid());
-  ASSERT_EQ(test::KeyStr("a", 5, kTypeSingleDeletion),
-            c_iter_->key().ToString());
-  c_iter_->Next();
-  ASSERT_TRUE(c_iter_->Valid());
-  ASSERT_EQ(test::KeyStr("a", 3, kTypeValue, true), c_iter_->key().ToString());
-  c_iter_->Next();
-  ASSERT_TRUE(c_iter_->Valid());
-  ASSERT_EQ(test::KeyStr("b", 10, kTypeValue), c_iter_->key().ToString());
-  c_iter_->Next();
-  ASSERT_FALSE(c_iter_->Valid());
-}
-
-}  // namespace rocksdb
-
-int main(int argc, char** argv) {
-  ::testing::InitGoogleTest(&argc, argv);
-  return RUN_ALL_TESTS();
-}
@@ -77,7 +77,6 @@ struct CompactionJob::SubcompactionState {
  struct Output {
    FileMetaData meta;
    bool finished;
-    std::shared_ptr<const TableProperties> table_properties;
  };

  // State kept for output being generated
@@ -212,7 +211,6 @@ CompactionJob::CompactionJob(
    std::atomic<bool>* shutting_down, LogBuffer* log_buffer,
    Directory* db_directory, Directory* output_directory, Statistics* stats,
    std::vector<SequenceNumber> existing_snapshots,
-    SequenceNumber earliest_write_conflict_snapshot,
    std::shared_ptr<Cache> table_cache, EventLogger* event_logger,
    bool paranoid_file_checks, bool measure_io_stats, const std::string& dbname,
    CompactionJobStats* compaction_job_stats)
@@ -231,7 +229,6 @@ CompactionJob::CompactionJob(
      output_directory_(output_directory),
      stats_(stats),
      existing_snapshots_(std::move(existing_snapshots)),
-      earliest_write_conflict_snapshot_(earliest_write_conflict_snapshot),
      table_cache_(std::move(table_cache)),
      event_logger_(event_logger),
      paranoid_file_checks_(paranoid_file_checks),
@@ -319,8 +316,6 @@ void CompactionJob::Prepare() {
      Slice* end = i == boundaries_.size() ? nullptr : &boundaries_[i];
      compact_->sub_compact_states.emplace_back(c, start, end, sizes_[i]);
    }
-    MeasureTime(stats_, NUM_SUBCOMPACTIONS_SCHEDULED,
-                compact_->sub_compact_states.size());
  } else {
    compact_->sub_compact_states.emplace_back(c, nullptr, nullptr);
  }
@@ -334,6 +329,11 @@ struct RangeWithSize {
      : range(a, b), size(s) {}
 };

+bool SliceCompare(const Comparator* cmp, const Slice& a, const Slice& b) {
+  // Returns true if a < b
+  return cmp->Compare(ExtractUserKey(a), ExtractUserKey(b)) < 0;
+}
+
 // Generates a histogram representing potential divisions of key ranges from
 // the input. It adds the starting and/or ending keys of certain input files
 // to the working set and then finds the approximate size of data in between
@@ -342,13 +342,14 @@ struct RangeWithSize {
 void CompactionJob::GenSubcompactionBoundaries() {
  auto* c = compact_->compaction;
  auto* cfd = c->column_family_data();
-  const Comparator* cfd_comparator = cfd->user_comparator();
-  std::vector<Slice> bounds;
+  std::set<Slice, std::function<bool(const Slice& a, const Slice& b)> > bounds(
+      std::bind(&SliceCompare, cfd->user_comparator(), std::placeholders::_1,
+                std::placeholders::_2));
  int start_lvl = c->start_level();
  int out_lvl = c->output_level();

  // Add the starting and/or ending key of certain input files as a potential
-  // boundary
+  // boundary (because we're inserting into a set, it avoids duplicates)
  for (size_t lvl_idx = 0; lvl_idx < c->num_input_levels(); lvl_idx++) {
    int lvl = c->level(lvl_idx);
    if (lvl >= start_lvl && lvl <= out_lvl) {
@@ -363,37 +364,27 @@ void CompactionJob::GenSubcompactionBoundaries() {
        // For level 0 add the starting and ending key of each file since the
        // files may have greatly differing key ranges (not range-partitioned)
        for (size_t i = 0; i < num_files; i++) {
-          bounds.emplace_back(flevel->files[i].smallest_key);
-          bounds.emplace_back(flevel->files[i].largest_key);
+          bounds.emplace(flevel->files[i].smallest_key);
+          bounds.emplace(flevel->files[i].largest_key);
        }
      } else {
        // For all other levels add the smallest/largest key in the level to
        // encompass the range covered by that level
-        bounds.emplace_back(flevel->files[0].smallest_key);
-        bounds.emplace_back(flevel->files[num_files - 1].largest_key);
+        bounds.emplace(flevel->files[0].smallest_key);
+        bounds.emplace(flevel->files[num_files - 1].largest_key);
        if (lvl == out_lvl) {
          // For the last level include the starting keys of all files since
          // the last level is the largest and probably has the widest key
          // range. Since it's range partitioned, the ending key of one file
          // and the starting key of the next are very close (or identical).
          for (size_t i = 1; i < num_files; i++) {
-            bounds.emplace_back(flevel->files[i].smallest_key);
+            bounds.emplace(flevel->files[i].smallest_key);
          }
        }
      }
    }
  }

-  std::sort(bounds.begin(), bounds.end(),
-    [cfd_comparator] (const Slice& a, const Slice& b) -> bool {
-      return cfd_comparator->Compare(ExtractUserKey(a), ExtractUserKey(b)) < 0;
-    });
-  // Remove duplicated entries from bounds
-  bounds.erase(std::unique(bounds.begin(), bounds.end(),
-    [cfd_comparator] (const Slice& a, const Slice& b) -> bool {
-      return cfd_comparator->Compare(ExtractUserKey(a), ExtractUserKey(b)) == 0;
-    }), bounds.end());
-
  // Combine consecutive pairs of boundaries into ranges with an approximate
  // size of data covered by keys in that range
  uint64_t sum = 0;
@@ -415,12 +406,9 @@ void CompactionJob::GenSubcompactionBoundaries() {

  // Group the ranges into subcompactions
  const double min_file_fill_percent = 4.0 / 5;
-  uint64_t max_output_files = 
-    static_cast<uint64_t>(
-        std::ceil(
-          sum / min_file_fill_percent /
-          cfd->GetCurrentMutableCFOptions()->MaxFileSizeForLevel(out_lvl))
-          );
+  uint64_t max_output_files = std::ceil(
+      sum / min_file_fill_percent /
+      cfd->GetCurrentMutableCFOptions()->MaxFileSizeForLevel(out_lvl));
  uint64_t subcompactions =
      std::min({static_cast<uint64_t>(ranges.size()),
                static_cast<uint64_t>(db_options_.max_subcompactions),
@@ -433,12 +421,12 @@ void CompactionJob::GenSubcompactionBoundaries() {
    // sizes becomes >= the expected mean size of a subcompaction
    sum = 0;
    for (size_t i = 0; i < ranges.size() - 1; i++) {
-      sum += ranges[i].size;
      if (subcompactions == 1) {
        // If there's only one left to schedule then it goes to the end so no
        // need to put an end boundary
-        continue;
+        break;
      }
+      sum += ranges[i].size;
      if (sum >= mean) {
        boundaries_.emplace_back(ExtractUserKey(ranges[i].range.limit));
        sizes_.emplace_back(sum);
@@ -497,16 +485,6 @@ Status CompactionJob::Run() {
    }
  }

-  TablePropertiesCollection tp;
-  for (const auto& state : compact_->sub_compact_states) {
-    for (const auto& output : state.outputs) {
-      auto fn = TableFileName(db_options_.db_paths, output.meta.fd.GetNumber(),
-                              output.meta.fd.GetPathId());
-      tp[fn] = output.table_properties;
-    }
-  }
-  compact_->compaction->SetOutputTableProperties(std::move(tp));
-
  // Finish up all book-keeping to unify the subcompaction results
  AggregateStatistics();
  UpdateCompactionStats();
@@ -562,9 +540,8 @@ Status CompactionJob::Install(const MutableCFOptions& mutable_cf_options,
  UpdateCompactionJobStats(stats);

  auto stream = event_logger_->LogToBuffer(log_buffer_);
-  stream << "job" << job_id_
-         << "event" << "compaction_finished"
-         << "compaction_time_micros" << compaction_stats_.micros
+  stream << "job" << job_id_ << "event"
+         << "compaction_finished"
         << "output_level" << compact_->compaction->output_level()
         << "num_output_files" << compact_->NumOutputFiles()
         << "total_output_size" << compact_->total_bytes
@@ -594,7 +571,7 @@ Status CompactionJob::Install(const MutableCFOptions& mutable_cf_options,

 void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
  assert(sub_compact != nullptr);
-  std::unique_ptr<InternalIterator> input(
+  std::unique_ptr<Iterator> input(
      versions_->MakeInputIterator(sub_compact->compaction));

  AutoThreadOperationStageUpdater stage_updater(
@@ -610,13 +587,17 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
  if (measure_io_stats_) {
    prev_perf_level = GetPerfLevel();
    SetPerfLevel(PerfLevel::kEnableTime);
-    prev_write_nanos = IOSTATS(write_nanos);
-    prev_fsync_nanos = IOSTATS(fsync_nanos);
-    prev_range_sync_nanos = IOSTATS(range_sync_nanos);
-    prev_prepare_write_nanos = IOSTATS(prepare_write_nanos);
+    prev_write_nanos = iostats_context.write_nanos;
+    prev_fsync_nanos = iostats_context.fsync_nanos;
+    prev_range_sync_nanos = iostats_context.range_sync_nanos;
+    prev_prepare_write_nanos = iostats_context.prepare_write_nanos;
  }

  ColumnFamilyData* cfd = sub_compact->compaction->column_family_data();
+  MergeHelper merge(cfd->user_comparator(), cfd->ioptions()->merge_operator,
+                    db_options_.info_log.get(),
+                    cfd->ioptions()->min_partial_merge_operands,
+                    false /* internal key corruption is expected */);
  auto compaction_filter = cfd->ioptions()->compaction_filter;
  std::unique_ptr<CompactionFilter> compaction_filter_from_factory = nullptr;
  if (compaction_filter == nullptr) {
@@ -624,13 +605,6 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
        sub_compact->compaction->CreateCompactionFilter();
    compaction_filter = compaction_filter_from_factory.get();
  }
-  MergeHelper merge(
-      env_, cfd->user_comparator(), cfd->ioptions()->merge_operator,
-      compaction_filter, db_options_.info_log.get(),
-      cfd->ioptions()->min_partial_merge_operands,
-      false /* internal key corruption is expected */,
-      existing_snapshots_.empty() ? 0 : existing_snapshots_.back(),
-      compact_->compaction->level(), db_options_.statistics.get());

  TEST_SYNC_POINT("CompactionJob::Run():Inprogress");

@@ -647,7 +621,7 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
  Status status;
  sub_compact->c_iter.reset(new CompactionIterator(
      input.get(), cfd->user_comparator(), &merge, versions_->LastSequence(),
-      &existing_snapshots_, earliest_write_conflict_snapshot_, env_, false,
+      &existing_snapshots_, env_, false, db_options_.statistics.get(),
      sub_compact->compaction, compaction_filter));
  auto c_iter = sub_compact->c_iter.get();
  c_iter->SeekToFirst();
@@ -737,13 +711,13 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {

  if (measure_io_stats_) {
    sub_compact->compaction_job_stats.file_write_nanos +=
-        IOSTATS(write_nanos) - prev_write_nanos;
+        iostats_context.write_nanos - prev_write_nanos;
    sub_compact->compaction_job_stats.file_fsync_nanos +=
-        IOSTATS(fsync_nanos) - prev_fsync_nanos;
+        iostats_context.fsync_nanos - prev_fsync_nanos;
    sub_compact->compaction_job_stats.file_range_sync_nanos +=
-        IOSTATS(range_sync_nanos) - prev_range_sync_nanos;
+        iostats_context.range_sync_nanos - prev_range_sync_nanos;
    sub_compact->compaction_job_stats.file_prepare_write_nanos +=
-        IOSTATS(prepare_write_nanos) - prev_prepare_write_nanos;
+        iostats_context.prepare_write_nanos - prev_prepare_write_nanos;
    if (prev_perf_level != PerfLevel::kEnableTime) {
      SetPerfLevel(prev_perf_level);
    }
@@ -820,7 +794,7 @@ Status CompactionJob::FinishCompactionOutputFile(
  if (s.ok() && current_entries > 0) {
    // Verify that the table is usable
    ColumnFamilyData* cfd = sub_compact->compaction->column_family_data();
-    InternalIterator* iter = cfd->table_cache()->NewIterator(
+    Iterator* iter = cfd->table_cache()->NewIterator(
        ReadOptions(), env_options_, cfd->internal_comparator(), meta->fd,
        nullptr, cfd->internal_stats()->GetFileReadHist(
                     compact_->compaction->output_level()),
@@ -834,10 +808,7 @@ Status CompactionJob::FinishCompactionOutputFile(

    delete iter;
    if (s.ok()) {
-      auto tp = sub_compact->builder->GetTableProperties();
-      sub_compact->current_output()->table_properties =
-          std::make_shared<TableProperties>(tp);
-      TableFileCreationInfo info(std::move(tp));
+      TableFileCreationInfo info(sub_compact->builder->GetTableProperties());
      info.db_name = dbname_;
      info.cf_name = cfd->GetName();
      info.file_path =
@@ -920,7 +891,7 @@ Status CompactionJob::OpenCompactionOutputFile(
  unique_ptr<WritableFile> writable_file;
  std::string fname = TableFileName(db_options_.db_paths, file_number,
                                    sub_compact->compaction->output_path_id());
-  Status s = NewWritableFile(env_, fname, &writable_file, env_options_);
+  Status s = env_->NewWritableFile(fname, &writable_file, env_options_);
  if (!s.ok()) {
    Log(InfoLogLevel::ERROR_LEVEL, db_options_.info_log,
        "[%s] [JOB %d] OpenCompactionOutputFiles for table #%" PRIu64
@@ -950,8 +921,8 @@ Status CompactionJob::OpenCompactionOutputFile(
      cfd->ioptions()->optimize_filters_for_hits && bottommost_level_;
  sub_compact->builder.reset(NewTableBuilder(
      *cfd->ioptions(), cfd->internal_comparator(),
-      cfd->int_tbl_prop_collector_factories(), cfd->GetID(),
-      sub_compact->outfile.get(), sub_compact->compaction->output_compression(),
+      cfd->int_tbl_prop_collector_factories(), sub_compact->outfile.get(),
+      sub_compact->compaction->output_compression(),
      cfd->ioptions()->compression_opts, skip_filters));
  LogFlush(db_options_.info_log);
  return s;
@@ -35,9 +35,9 @@
 #include "rocksdb/env.h"
 #include "rocksdb/memtablerep.h"
 #include "rocksdb/transaction_log.h"
-#include "table/scoped_arena_iterator.h"
 #include "util/autovector.h"
 #include "util/event_logger.h"
+#include "util/scoped_arena_iterator.h"
 #include "util/stop_watch.h"
 #include "util/thread_local.h"

@@ -58,7 +58,6 @@ class CompactionJob {
                Directory* db_directory, Directory* output_directory,
                Statistics* stats,
                std::vector<SequenceNumber> existing_snapshots,
-                SequenceNumber earliest_write_conflict_snapshot,
                std::shared_ptr<Cache> table_cache, EventLogger* event_logger,
                bool paranoid_file_checks, bool measure_io_stats,
                const std::string& dbname,
@@ -135,12 +134,6 @@ class CompactionJob {
  // entirely within s1 and s2, then the earlier version of k1 can be safely
  // deleted because that version is not visible in any snapshot.
  std::vector<SequenceNumber> existing_snapshots_;
-
-  // This is the earliest snapshot that could be used for write-conflict
-  // checking by a transaction.  For any user-key newer than this snapshot, we
-  // should make sure not to remove evidence that a write occurred.
-  SequenceNumber earliest_write_conflict_snapshot_;
-
  std::shared_ptr<Cache> table_cache_;

  EventLogger* event_logger_;
@@ -27,7 +27,6 @@
 #include "db/job_context.h"
 #include "db/version_set.h"
 #include "db/write_batch_internal.h"
-#include "memtable/hash_linklist_rep.h"
 #include "port/stack_trace.h"
 #include "rocksdb/cache.h"
 #include "rocksdb/compaction_filter.h"
@@ -48,13 +47,14 @@
 #include "table/block_based_table_factory.h"
 #include "table/mock_table.h"
 #include "table/plain_table_factory.h"
-#include "table/scoped_arena_iterator.h"
 #include "util/compression.h"
 #include "util/hash.h"
+#include "util/hash_linklist_rep.h"
 #include "util/logging.h"
 #include "util/mock_env.h"
 #include "util/mutexlock.h"
 #include "util/rate_limiter.h"
+#include "util/scoped_arena_iterator.h"
 #include "util/statistics.h"
 #include "util/string_util.h"
 #include "util/sync_point.h"
@@ -64,7 +64,7 @@
 #include "util/xfunc.h"
 #include "utilities/merge_operators.h"

-#if !defined(IOS_CROSS_COMPILE)
+#if !defined(IOS_CROSS_COMPILE) && (!defined(NDEBUG) || !defined(OS_WIN))
 #ifndef ROCKSDB_LITE
 namespace rocksdb {

@@ -551,9 +551,8 @@ uint64_t EstimatedFileSize(
  const size_t kFooterSize = 512;

  uint64_t data_size =
-    static_cast<uint64_t>(
      num_records * (key_size + value_size * compression_ratio +
-                     kPerKeyOverhead));
+                     kPerKeyOverhead);

  return data_size + kFooterSize
         + num_records * bloom_bits_per_key / 8      // filter block
@@ -1030,14 +1029,6 @@ int main(int argc, char** argv) {
  return RUN_ALL_TESTS();
 }

-#else
-#include <stdio.h>
-
-int main(int argc, char** argv) {
-  fprintf(stderr, "SKIPPED, not supported in ROCKSDB_LITE\n");
-  return 0;
-}
-
 #endif  // !ROCKSDB_LITE

 #else
@@ -3,8 +3,6 @@
 //  LICENSE file in the root directory of this source tree. An additional grant
 //  of patent rights can be found in the PATENTS file in the same directory.

-#ifndef ROCKSDB_LITE
-
 #include <algorithm>
 #include <map>
 #include <string>
@@ -162,12 +160,9 @@ class CompactionJobTest : public testing::Test {
        auto key = ToString(i * kMatchingKeys + k);
        auto value = ToString(i * kKeysPerFile + k);
        InternalKey internal_key(key, ++sequence_number, kTypeValue);
-
        // This is how the key will look like once it's written in bottommost
        // file
-        InternalKey bottommost_internal_key(
-            key, (key == "9999") ? sequence_number : 0, kTypeValue);
-
+        InternalKey bottommost_internal_key(key, 0, kTypeValue);
        if (corrupt_id(k)) {
          test::CorruptKeyType(&internal_key);
          test::CorruptKeyType(&bottommost_internal_key);
@@ -187,7 +182,7 @@ class CompactionJobTest : public testing::Test {
    return expected_results;
  }

-  void NewDB() {
+  void NewDB(std::shared_ptr<MergeOperator> merge_operator = nullptr) {
    VersionEdit new_db;
    new_db.SetLogNumber(0);
    new_db.SetNextFile(2);
@@ -201,7 +196,7 @@ class CompactionJobTest : public testing::Test {
    unique_ptr<WritableFileWriter> file_writer(
        new WritableFileWriter(std::move(file), env_options_));
    {
-      log::Writer log(std::move(file_writer), 0, false);
+      log::Writer log(std::move(file_writer));
      std::string record;
      new_db.EncodeTo(&record);
      s = log.AddRecord(record);
@@ -212,19 +207,15 @@ class CompactionJobTest : public testing::Test {

    std::vector<ColumnFamilyDescriptor> column_families;
    cf_options_.table_factory = mock_table_factory_;
-    cf_options_.merge_operator = merge_op_;
-    cf_options_.compaction_filter = compaction_filter_.get();
+    cf_options_.merge_operator = merge_operator;
    column_families.emplace_back(kDefaultColumnFamilyName, cf_options_);

    EXPECT_OK(versions_->Recover(column_families, false));
    cfd_ = versions_->GetColumnFamilySet()->GetDefault();
  }

-  void RunCompaction(
-      const std::vector<std::vector<FileMetaData*>>& input_files,
-      const stl_wrappers::KVMap& expected_results,
-      const std::vector<SequenceNumber>& snapshots = {},
-      SequenceNumber earliest_write_conflict_snapshot = kMaxSequenceNumber) {
+  void RunCompaction(const std::vector<std::vector<FileMetaData*>>& input_files,
+                     const stl_wrappers::KVMap& expected_results) {
    auto cfd = versions_->GetColumnFamilySet()->GetDefault();

    size_t num_input_files = 0;
@@ -248,11 +239,11 @@ class CompactionJobTest : public testing::Test {
    LogBuffer log_buffer(InfoLogLevel::INFO_LEVEL, db_options_.info_log.get());
    mutex_.Lock();
    EventLogger event_logger(db_options_.info_log.get());
-    CompactionJob compaction_job(
-        0, &compaction, db_options_, env_options_, versions_.get(),
-        &shutting_down_, &log_buffer, nullptr, nullptr, nullptr, snapshots,
-        earliest_write_conflict_snapshot, table_cache_, &event_logger, false,
-        false, dbname_, &compaction_job_stats_);
+    CompactionJob compaction_job(0, &compaction, db_options_, env_options_,
+                                 versions_.get(), &shutting_down_, &log_buffer,
+                                 nullptr, nullptr, nullptr, {}, table_cache_,
+                                 &event_logger, false, false, dbname_,
+                                 &compaction_job_stats_);

    VerifyInitializationOfCompactionJobStats(compaction_job_stats_);

@@ -266,16 +257,10 @@ class CompactionJobTest : public testing::Test {
                                     &mutex_));
    mutex_.Unlock();

-    if (expected_results.size() == 0) {
-      ASSERT_GE(compaction_job_stats_.elapsed_micros, 0U);
-      ASSERT_EQ(compaction_job_stats_.num_input_files, num_input_files);
-      ASSERT_EQ(compaction_job_stats_.num_output_files, 0U);
-    } else {
-      ASSERT_GE(compaction_job_stats_.elapsed_micros, 0U);
-      ASSERT_EQ(compaction_job_stats_.num_input_files, num_input_files);
-      ASSERT_EQ(compaction_job_stats_.num_output_files, 1U);
-      mock_table_factory_->AssertLatestFile(expected_results);
-    }
+    ASSERT_GE(compaction_job_stats_.elapsed_micros, 0U);
+    ASSERT_EQ(compaction_job_stats_.num_input_files, num_input_files);
+    ASSERT_EQ(compaction_job_stats_.num_output_files, 1U);
+    mock_table_factory_->AssertLatestFile(expected_results);
  }

  Env* env_;
@@ -293,8 +278,6 @@ class CompactionJobTest : public testing::Test {
  std::shared_ptr<mock::MockTableFactory> mock_table_factory_;
  CompactionJobStats compaction_job_stats_;
  ColumnFamilyData* cfd_;
-  std::unique_ptr<CompactionFilter> compaction_filter_;
-  std::shared_ptr<MergeOperator> merge_op_;
 };

 TEST_F(CompactionJobTest, Simple) {
@@ -313,7 +296,7 @@ TEST_F(CompactionJobTest, SimpleCorrupted) {
  auto expected_results = CreateTwoFiles(true);
  auto cfd = versions_->GetColumnFamilySet()->GetDefault();
  auto files = cfd->current()->storage_info()->LevelFiles(0);
-  RunCompaction({files}, expected_results);
+  RunCompaction({ files }, expected_results);
  ASSERT_EQ(compaction_job_stats_.num_corrupt_keys, 400U);
 }

@@ -333,7 +316,7 @@ TEST_F(CompactionJobTest, SimpleDeletion) {

  SetLastSequence(4U);
  auto files = cfd_->current()->storage_info()->LevelFiles(0);
-  RunCompaction({files}, expected_results);
+  RunCompaction({ files }, expected_results);
 }

 TEST_F(CompactionJobTest, SimpleOverwrite) {
@@ -351,11 +334,11 @@ TEST_F(CompactionJobTest, SimpleOverwrite) {

  auto expected_results =
      mock::MakeMockFile({{KeyStr("a", 0U, kTypeValue), "val2"},
-                          {KeyStr("b", 4U, kTypeValue), "val3"}});
+                          {KeyStr("b", 0U, kTypeValue), "val3"}});

  SetLastSequence(4U);
  auto files = cfd_->current()->storage_info()->LevelFiles(0);
-  RunCompaction({files}, expected_results);
+  RunCompaction({ files }, expected_results);
 }

 TEST_F(CompactionJobTest, SimpleNonLastLevel) {
@@ -384,12 +367,12 @@ TEST_F(CompactionJobTest, SimpleNonLastLevel) {
  SetLastSequence(6U);
  auto lvl0_files = cfd_->current()->storage_info()->LevelFiles(0);
  auto lvl1_files = cfd_->current()->storage_info()->LevelFiles(1);
-  RunCompaction({lvl0_files, lvl1_files}, expected_results);
+  RunCompaction({ lvl0_files, lvl1_files }, expected_results);
 }

 TEST_F(CompactionJobTest, SimpleMerge) {
-  merge_op_ = MergeOperators::CreateStringAppendOperator();
-  NewDB();
+  auto merge_op = MergeOperators::CreateStringAppendOperator();
+  NewDB(merge_op);

  auto file1 = mock::MakeMockFile({
      {KeyStr("a", 5U, kTypeMerge), "5"},
@@ -404,16 +387,16 @@ TEST_F(CompactionJobTest, SimpleMerge) {

  auto expected_results =
      mock::MakeMockFile({{KeyStr("a", 0U, kTypeValue), "3,4,5"},
-                          {KeyStr("b", 2U, kTypeValue), "1,2"}});
+                          {KeyStr("b", 0U, kTypeValue), "1,2"}});

  SetLastSequence(5U);
  auto files = cfd_->current()->storage_info()->LevelFiles(0);
-  RunCompaction({files}, expected_results);
+  RunCompaction({ files }, expected_results);
 }

 TEST_F(CompactionJobTest, NonAssocMerge) {
-  merge_op_ = MergeOperators::CreateStringAppendTESTOperator();
-  NewDB();
+  auto merge_op = MergeOperators::CreateStringAppendTESTOperator();
+  NewDB(merge_op);

  auto file1 = mock::MakeMockFile({
      {KeyStr("a", 5U, kTypeMerge), "5"},
@@ -433,481 +416,7 @@ TEST_F(CompactionJobTest, NonAssocMerge) {

  SetLastSequence(5U);
  auto files = cfd_->current()->storage_info()->LevelFiles(0);
-  RunCompaction({files}, expected_results);
-}
-
-// Filters merge operands with value 10.
-TEST_F(CompactionJobTest, MergeOperandFilter) {
-  merge_op_ = MergeOperators::CreateUInt64AddOperator();
-  compaction_filter_.reset(new test::FilterNumber(10U));
-  NewDB();
-
-  auto file1 = mock::MakeMockFile(
-      {{KeyStr("a", 5U, kTypeMerge), test::EncodeInt(5U)},
-       {KeyStr("a", 4U, kTypeMerge), test::EncodeInt(10U)},  // Filtered
-       {KeyStr("a", 3U, kTypeMerge), test::EncodeInt(3U)}});
-  AddMockFile(file1);
-
-  auto file2 = mock::MakeMockFile({
-      {KeyStr("b", 2U, kTypeMerge), test::EncodeInt(2U)},
-      {KeyStr("b", 1U, kTypeMerge), test::EncodeInt(10U)}  // Filtered
-  });
-  AddMockFile(file2);
-
-  auto expected_results =
-      mock::MakeMockFile({{KeyStr("a", 0U, kTypeValue), test::EncodeInt(8U)},
-                          {KeyStr("b", 2U, kTypeMerge), test::EncodeInt(2U)}});
-
-  SetLastSequence(5U);
-  auto files = cfd_->current()->storage_info()->LevelFiles(0);
-  RunCompaction({files}, expected_results);
-}
-
-TEST_F(CompactionJobTest, FilterSomeMergeOperands) {
-  merge_op_ = MergeOperators::CreateUInt64AddOperator();
-  compaction_filter_.reset(new test::FilterNumber(10U));
-  NewDB();
-
-  auto file1 = mock::MakeMockFile(
-      {{KeyStr("a", 5U, kTypeMerge), test::EncodeInt(5U)},
-       {KeyStr("a", 4U, kTypeMerge), test::EncodeInt(10U)},  // Filtered
-       {KeyStr("a", 3U, kTypeValue), test::EncodeInt(5U)},
-       {KeyStr("d", 8U, kTypeMerge), test::EncodeInt(10U)}});
-  AddMockFile(file1);
-
-  auto file2 =
-      mock::MakeMockFile({{KeyStr("b", 2U, kTypeMerge), test::EncodeInt(10U)},
-                          {KeyStr("b", 1U, kTypeMerge), test::EncodeInt(10U)},
-                          {KeyStr("c", 2U, kTypeMerge), test::EncodeInt(3U)},
-                          {KeyStr("c", 1U, kTypeValue), test::EncodeInt(7U)},
-                          {KeyStr("d", 1U, kTypeValue), test::EncodeInt(6U)}});
-  AddMockFile(file2);
-
-  auto file3 =
-      mock::MakeMockFile({{KeyStr("a", 1U, kTypeMerge), test::EncodeInt(3U)}});
-  AddMockFile(file3, 2);
-
-  auto expected_results = mock::MakeMockFile({
-      {KeyStr("a", 5U, kTypeValue), test::EncodeInt(10U)},
-      {KeyStr("c", 2U, kTypeValue), test::EncodeInt(10U)},
-      {KeyStr("d", 1U, kTypeValue), test::EncodeInt(6U)}
-      // b does not appear because the operands are filtered
-  });
-
-  SetLastSequence(5U);
-  auto files = cfd_->current()->storage_info()->LevelFiles(0);
-  RunCompaction({files}, expected_results);
-}
-
-// Test where all operands/merge results are filtered out.
-TEST_F(CompactionJobTest, FilterAllMergeOperands) {
-  merge_op_ = MergeOperators::CreateUInt64AddOperator();
-  compaction_filter_.reset(new test::FilterNumber(10U));
-  NewDB();
-
-  auto file1 =
-      mock::MakeMockFile({{KeyStr("a", 11U, kTypeMerge), test::EncodeInt(10U)},
-                          {KeyStr("a", 10U, kTypeMerge), test::EncodeInt(10U)},
-                          {KeyStr("a", 9U, kTypeMerge), test::EncodeInt(10U)}});
-  AddMockFile(file1);
-
-  auto file2 =
-      mock::MakeMockFile({{KeyStr("b", 8U, kTypeMerge), test::EncodeInt(10U)},
-                          {KeyStr("b", 7U, kTypeMerge), test::EncodeInt(10U)},
-                          {KeyStr("b", 6U, kTypeMerge), test::EncodeInt(10U)},
-                          {KeyStr("b", 5U, kTypeMerge), test::EncodeInt(10U)},
-                          {KeyStr("b", 4U, kTypeMerge), test::EncodeInt(10U)},
-                          {KeyStr("b", 3U, kTypeMerge), test::EncodeInt(10U)},
-                          {KeyStr("b", 2U, kTypeMerge), test::EncodeInt(10U)},
-                          {KeyStr("c", 2U, kTypeMerge), test::EncodeInt(10U)},
-                          {KeyStr("c", 1U, kTypeMerge), test::EncodeInt(10U)}});
-  AddMockFile(file2);
-
-  auto file3 =
-      mock::MakeMockFile({{KeyStr("a", 2U, kTypeMerge), test::EncodeInt(10U)},
-                          {KeyStr("b", 1U, kTypeMerge), test::EncodeInt(10U)}});
-  AddMockFile(file3, 2);
-
-  SetLastSequence(11U);
-  auto files = cfd_->current()->storage_info()->LevelFiles(0);
-
-  stl_wrappers::KVMap empty_map;
-  RunCompaction({files}, empty_map);
-}
-
-TEST_F(CompactionJobTest, SimpleSingleDelete) {
-  NewDB();
-
-  auto file1 = mock::MakeMockFile({
-      {KeyStr("a", 5U, kTypeDeletion), ""},
-      {KeyStr("b", 6U, kTypeSingleDeletion), ""},
-  });
-  AddMockFile(file1);
-
-  auto file2 = mock::MakeMockFile({{KeyStr("a", 3U, kTypeValue), "val"},
-                                   {KeyStr("b", 4U, kTypeValue), "val"}});
-  AddMockFile(file2);
-
-  auto file3 = mock::MakeMockFile({
-      {KeyStr("a", 1U, kTypeValue), "val"},
-  });
-  AddMockFile(file3, 2);
-
-  auto expected_results =
-      mock::MakeMockFile({{KeyStr("a", 5U, kTypeDeletion), ""}});
-
-  SetLastSequence(6U);
-  auto files = cfd_->current()->storage_info()->LevelFiles(0);
-  RunCompaction({files}, expected_results);
-}
-
-TEST_F(CompactionJobTest, SingleDeleteSnapshots) {
-  NewDB();
-
-  auto file1 = mock::MakeMockFile({
-      {KeyStr("A", 12U, kTypeSingleDeletion), ""},
-      {KeyStr("a", 12U, kTypeSingleDeletion), ""},
-      {KeyStr("b", 21U, kTypeSingleDeletion), ""},
-      {KeyStr("c", 22U, kTypeSingleDeletion), ""},
-      {KeyStr("d", 9U, kTypeSingleDeletion), ""},
-      {KeyStr("f", 21U, kTypeSingleDeletion), ""},
-      {KeyStr("j", 11U, kTypeSingleDeletion), ""},
-      {KeyStr("j", 9U, kTypeSingleDeletion), ""},
-      {KeyStr("k", 12U, kTypeSingleDeletion), ""},
-      {KeyStr("k", 11U, kTypeSingleDeletion), ""},
-      {KeyStr("l", 3U, kTypeSingleDeletion), ""},
-      {KeyStr("l", 2U, kTypeSingleDeletion), ""},
-  });
-  AddMockFile(file1);
-
-  auto file2 = mock::MakeMockFile({
-      {KeyStr("0", 2U, kTypeSingleDeletion), ""},
-      {KeyStr("a", 11U, kTypeValue), "val1"},
-      {KeyStr("b", 11U, kTypeValue), "val2"},
-      {KeyStr("c", 21U, kTypeValue), "val3"},
-      {KeyStr("d", 8U, kTypeValue), "val4"},
-      {KeyStr("e", 2U, kTypeSingleDeletion), ""},
-      {KeyStr("f", 1U, kTypeValue), "val1"},
-      {KeyStr("g", 11U, kTypeSingleDeletion), ""},
-      {KeyStr("h", 2U, kTypeSingleDeletion), ""},
-      {KeyStr("m", 12U, kTypeValue), "val1"},
-      {KeyStr("m", 11U, kTypeSingleDeletion), ""},
-      {KeyStr("m", 8U, kTypeValue), "val2"},
-  });
-  AddMockFile(file2);
-
-  auto file3 = mock::MakeMockFile({
-      {KeyStr("A", 1U, kTypeValue), "val"},
-      {KeyStr("e", 1U, kTypeValue), "val"},
-  });
-  AddMockFile(file3, 2);
-
-  auto expected_results = mock::MakeMockFile({
-      {KeyStr("A", 12U, kTypeSingleDeletion), ""},
-      {KeyStr("a", 12U, kTypeSingleDeletion), ""},
-      {KeyStr("a", 11U, kTypeValue), ""},
-      {KeyStr("b", 21U, kTypeSingleDeletion), ""},
-      {KeyStr("b", 11U, kTypeValue), "val2"},
-      {KeyStr("c", 22U, kTypeSingleDeletion), ""},
-      {KeyStr("c", 21U, kTypeValue), ""},
-      {KeyStr("e", 2U, kTypeSingleDeletion), ""},
-      {KeyStr("f", 21U, kTypeSingleDeletion), ""},
-      {KeyStr("f", 1U, kTypeValue), "val1"},
-      {KeyStr("g", 11U, kTypeSingleDeletion), ""},
-      {KeyStr("j", 11U, kTypeSingleDeletion), ""},
-      {KeyStr("k", 11U, kTypeSingleDeletion), ""},
-      {KeyStr("m", 12U, kTypeValue), "val1"},
-      {KeyStr("m", 11U, kTypeSingleDeletion), ""},
-      {KeyStr("m", 8U, kTypeValue), "val2"},
-  });
-
-  SetLastSequence(22U);
-  auto files = cfd_->current()->storage_info()->LevelFiles(0);
-  RunCompaction({files}, expected_results, {10U, 20U}, 10U);
-}
-
-TEST_F(CompactionJobTest, EarliestWriteConflictSnapshot) {
-  NewDB();
-
-  // Test multiple snapshots where the earliest snapshot is not a
-  // write-conflic-snapshot.
-
-  auto file1 = mock::MakeMockFile({
-      {KeyStr("A", 24U, kTypeSingleDeletion), ""},
-      {KeyStr("A", 23U, kTypeValue), "val"},
-      {KeyStr("B", 24U, kTypeSingleDeletion), ""},
-      {KeyStr("B", 23U, kTypeValue), "val"},
-      {KeyStr("D", 24U, kTypeSingleDeletion), ""},
-      {KeyStr("G", 32U, kTypeSingleDeletion), ""},
-      {KeyStr("G", 31U, kTypeValue), "val"},
-      {KeyStr("G", 24U, kTypeSingleDeletion), ""},
-      {KeyStr("G", 23U, kTypeValue), "val2"},
-      {KeyStr("H", 31U, kTypeValue), "val"},
-      {KeyStr("H", 24U, kTypeSingleDeletion), ""},
-      {KeyStr("H", 23U, kTypeValue), "val"},
-      {KeyStr("I", 35U, kTypeSingleDeletion), ""},
-      {KeyStr("I", 34U, kTypeValue), "val2"},
-      {KeyStr("I", 33U, kTypeSingleDeletion), ""},
-      {KeyStr("I", 32U, kTypeValue), "val3"},
-      {KeyStr("I", 31U, kTypeSingleDeletion), ""},
-      {KeyStr("J", 34U, kTypeValue), "val"},
-      {KeyStr("J", 33U, kTypeSingleDeletion), ""},
-      {KeyStr("J", 25U, kTypeValue), "val2"},
-      {KeyStr("J", 24U, kTypeSingleDeletion), ""},
-  });
-  AddMockFile(file1);
-
-  auto file2 = mock::MakeMockFile({
-      {KeyStr("A", 14U, kTypeSingleDeletion), ""},
-      {KeyStr("A", 13U, kTypeValue), "val2"},
-      {KeyStr("C", 14U, kTypeSingleDeletion), ""},
-      {KeyStr("C", 13U, kTypeValue), "val"},
-      {KeyStr("E", 12U, kTypeSingleDeletion), ""},
-      {KeyStr("F", 4U, kTypeSingleDeletion), ""},
-      {KeyStr("F", 3U, kTypeValue), "val"},
-      {KeyStr("G", 14U, kTypeSingleDeletion), ""},
-      {KeyStr("G", 13U, kTypeValue), "val3"},
-      {KeyStr("H", 14U, kTypeSingleDeletion), ""},
-      {KeyStr("H", 13U, kTypeValue), "val2"},
-      {KeyStr("I", 13U, kTypeValue), "val4"},
-      {KeyStr("I", 12U, kTypeSingleDeletion), ""},
-      {KeyStr("I", 11U, kTypeValue), "val5"},
-      {KeyStr("J", 15U, kTypeValue), "val3"},
-      {KeyStr("J", 14U, kTypeSingleDeletion), ""},
-  });
-  AddMockFile(file2);
-
-  auto expected_results = mock::MakeMockFile({
-      {KeyStr("A", 24U, kTypeSingleDeletion), ""},
-      {KeyStr("A", 23U, kTypeValue), ""},
-      {KeyStr("B", 24U, kTypeSingleDeletion), ""},
-      {KeyStr("B", 23U, kTypeValue), ""},
-      {KeyStr("D", 24U, kTypeSingleDeletion), ""},
-      {KeyStr("E", 12U, kTypeSingleDeletion), ""},
-      {KeyStr("G", 32U, kTypeSingleDeletion), ""},
-      {KeyStr("G", 31U, kTypeValue), ""},
-      {KeyStr("H", 31U, kTypeValue), "val"},
-      {KeyStr("I", 35U, kTypeSingleDeletion), ""},
-      {KeyStr("I", 34U, kTypeValue), ""},
-      {KeyStr("I", 31U, kTypeSingleDeletion), ""},
-      {KeyStr("I", 13U, kTypeValue), "val4"},
-      {KeyStr("J", 34U, kTypeValue), "val"},
-      {KeyStr("J", 33U, kTypeSingleDeletion), ""},
-      {KeyStr("J", 25U, kTypeValue), "val2"},
-      {KeyStr("J", 24U, kTypeSingleDeletion), ""},
-      {KeyStr("J", 15U, kTypeValue), "val3"},
-      {KeyStr("J", 14U, kTypeSingleDeletion), ""},
-  });
-
-  SetLastSequence(24U);
-  auto files = cfd_->current()->storage_info()->LevelFiles(0);
-  RunCompaction({files}, expected_results, {10U, 20U, 30U}, 20U);
-}
-
-TEST_F(CompactionJobTest, SingleDeleteZeroSeq) {
-  NewDB();
-
-  auto file1 = mock::MakeMockFile({
-      {KeyStr("A", 10U, kTypeSingleDeletion), ""},
-      {KeyStr("dummy", 5U, kTypeValue), "val2"},
-  });
-  AddMockFile(file1);
-
-  auto file2 = mock::MakeMockFile({
-      {KeyStr("A", 0U, kTypeValue), "val"},
-  });
-  AddMockFile(file2);
-
-  auto expected_results = mock::MakeMockFile({
-      {KeyStr("dummy", 5U, kTypeValue), "val2"},
-  });
-
-  SetLastSequence(22U);
-  auto files = cfd_->current()->storage_info()->LevelFiles(0);
-  RunCompaction({files}, expected_results, {});
-}
-
-TEST_F(CompactionJobTest, MultiSingleDelete) {
-  // Tests three scenarios involving multiple single delete/put pairs:
-  //
-  // A: Put Snapshot SDel Put SDel -> Put Snapshot SDel
-  // B: Snapshot Put SDel Put SDel Snapshot -> Snapshot SDel Snapshot
-  // C: SDel Put SDel Snapshot Put -> Snapshot Put
-  // D: (Put) SDel Snapshot Put SDel -> (Put) SDel Snapshot SDel
-  // E: Put SDel Snapshot Put SDel -> Snapshot SDel
-  // F: Put SDel Put Sdel Snapshot -> removed
-  // G: Snapshot SDel Put SDel Put -> Snapshot Put SDel
-  // H: (Put) Put SDel Put Sdel Snapshot -> Removed
-  // I: (Put) Snapshot Put SDel Put SDel -> SDel
-  // J: Put Put SDel Put SDel SDel Snapshot Put Put SDel SDel Put
-  //      -> Snapshot Put
-  // K: SDel SDel Put SDel Put Put Snapshot SDel Put SDel SDel Put SDel
-  //      -> Snapshot Put Snapshot SDel
-  // L: SDel Put Del Put SDel Snapshot Del Put Del SDel Put SDel
-  //      -> Snapshot SDel
-  // M: (Put) SDel Put Del Put SDel Snapshot Put Del SDel Put SDel Del
-  //      -> SDel Snapshot Del
-  NewDB();
-
-  auto file1 = mock::MakeMockFile({
-      {KeyStr("A", 14U, kTypeSingleDeletion), ""},
-      {KeyStr("A", 13U, kTypeValue), "val5"},
-      {KeyStr("A", 12U, kTypeSingleDeletion), ""},
-      {KeyStr("B", 14U, kTypeSingleDeletion), ""},
-      {KeyStr("B", 13U, kTypeValue), "val2"},
-      {KeyStr("C", 14U, kTypeValue), "val3"},
-      {KeyStr("D", 12U, kTypeSingleDeletion), ""},
-      {KeyStr("D", 11U, kTypeValue), "val4"},
-      {KeyStr("G", 15U, kTypeValue), "val"},
-      {KeyStr("G", 14U, kTypeSingleDeletion), ""},
-      {KeyStr("G", 13U, kTypeValue), "val"},
-      {KeyStr("I", 14U, kTypeSingleDeletion), ""},
-      {KeyStr("I", 13U, kTypeValue), "val"},
-      {KeyStr("J", 15U, kTypeValue), "val"},
-      {KeyStr("J", 14U, kTypeSingleDeletion), ""},
-      {KeyStr("J", 13U, kTypeSingleDeletion), ""},
-      {KeyStr("J", 12U, kTypeValue), "val"},
-      {KeyStr("J", 11U, kTypeValue), "val"},
-      {KeyStr("K", 16U, kTypeSingleDeletion), ""},
-      {KeyStr("K", 15U, kTypeValue), "val1"},
-      {KeyStr("K", 14U, kTypeSingleDeletion), ""},
-      {KeyStr("K", 13U, kTypeSingleDeletion), ""},
-      {KeyStr("K", 12U, kTypeValue), "val2"},
-      {KeyStr("K", 11U, kTypeSingleDeletion), ""},
-      {KeyStr("L", 16U, kTypeSingleDeletion), ""},
-      {KeyStr("L", 15U, kTypeValue), "val"},
-      {KeyStr("L", 14U, kTypeSingleDeletion), ""},
-      {KeyStr("L", 13U, kTypeDeletion), ""},
-      {KeyStr("L", 12U, kTypeValue), "val"},
-      {KeyStr("L", 11U, kTypeDeletion), ""},
-      {KeyStr("M", 16U, kTypeDeletion), ""},
-      {KeyStr("M", 15U, kTypeSingleDeletion), ""},
-      {KeyStr("M", 14U, kTypeValue), "val"},
-      {KeyStr("M", 13U, kTypeSingleDeletion), ""},
-      {KeyStr("M", 12U, kTypeDeletion), ""},
-      {KeyStr("M", 11U, kTypeValue), "val"},
-  });
-  AddMockFile(file1);
-
-  auto file2 = mock::MakeMockFile({
-      {KeyStr("A", 10U, kTypeValue), "val"},
-      {KeyStr("B", 12U, kTypeSingleDeletion), ""},
-      {KeyStr("B", 11U, kTypeValue), "val2"},
-      {KeyStr("C", 10U, kTypeSingleDeletion), ""},
-      {KeyStr("C", 9U, kTypeValue), "val6"},
-      {KeyStr("C", 8U, kTypeSingleDeletion), ""},
-      {KeyStr("D", 10U, kTypeSingleDeletion), ""},
-      {KeyStr("E", 12U, kTypeSingleDeletion), ""},
-      {KeyStr("E", 11U, kTypeValue), "val"},
-      {KeyStr("E", 5U, kTypeSingleDeletion), ""},
-      {KeyStr("E", 4U, kTypeValue), "val"},
-      {KeyStr("F", 6U, kTypeSingleDeletion), ""},
-      {KeyStr("F", 5U, kTypeValue), "val"},
-      {KeyStr("F", 4U, kTypeSingleDeletion), ""},
-      {KeyStr("F", 3U, kTypeValue), "val"},
-      {KeyStr("G", 12U, kTypeSingleDeletion), ""},
-      {KeyStr("H", 6U, kTypeSingleDeletion), ""},
-      {KeyStr("H", 5U, kTypeValue), "val"},
-      {KeyStr("H", 4U, kTypeSingleDeletion), ""},
-      {KeyStr("H", 3U, kTypeValue), "val"},
-      {KeyStr("I", 12U, kTypeSingleDeletion), ""},
-      {KeyStr("I", 11U, kTypeValue), "val"},
-      {KeyStr("J", 6U, kTypeSingleDeletion), ""},
-      {KeyStr("J", 5U, kTypeSingleDeletion), ""},
-      {KeyStr("J", 4U, kTypeValue), "val"},
-      {KeyStr("J", 3U, kTypeSingleDeletion), ""},
-      {KeyStr("J", 2U, kTypeValue), "val"},
-      {KeyStr("K", 8U, kTypeValue), "val3"},
-      {KeyStr("K", 7U, kTypeValue), "val4"},
-      {KeyStr("K", 6U, kTypeSingleDeletion), ""},
-      {KeyStr("K", 5U, kTypeValue), "val5"},
-      {KeyStr("K", 2U, kTypeSingleDeletion), ""},
-      {KeyStr("K", 1U, kTypeSingleDeletion), ""},
-      {KeyStr("L", 5U, kTypeSingleDeletion), ""},
-      {KeyStr("L", 4U, kTypeValue), "val"},
-      {KeyStr("L", 3U, kTypeDeletion), ""},
-      {KeyStr("L", 2U, kTypeValue), "val"},
-      {KeyStr("L", 1U, kTypeSingleDeletion), ""},
-      {KeyStr("M", 10U, kTypeSingleDeletion), ""},
-      {KeyStr("M", 7U, kTypeValue), "val"},
-      {KeyStr("M", 5U, kTypeDeletion), ""},
-      {KeyStr("M", 4U, kTypeValue), "val"},
-      {KeyStr("M", 3U, kTypeSingleDeletion), ""},
-  });
-  AddMockFile(file2);
-
-  auto file3 = mock::MakeMockFile({
-      {KeyStr("D", 1U, kTypeValue), "val"},
-      {KeyStr("H", 1U, kTypeValue), "val"},
-      {KeyStr("I", 2U, kTypeValue), "val"},
-  });
-  AddMockFile(file3, 2);
-
-  auto file4 = mock::MakeMockFile({
-      {KeyStr("M", 1U, kTypeValue), "val"},
-  });
-  AddMockFile(file4, 2);
-
-  auto expected_results =
-      mock::MakeMockFile({{KeyStr("A", 14U, kTypeSingleDeletion), ""},
-                          {KeyStr("A", 13U, kTypeValue), ""},
-                          {KeyStr("A", 12U, kTypeSingleDeletion), ""},
-                          {KeyStr("A", 10U, kTypeValue), "val"},
-                          {KeyStr("B", 14U, kTypeSingleDeletion), ""},
-                          {KeyStr("B", 13U, kTypeValue), ""},
-                          {KeyStr("C", 14U, kTypeValue), "val3"},
-                          {KeyStr("D", 12U, kTypeSingleDeletion), ""},
-                          {KeyStr("D", 11U, kTypeValue), ""},
-                          {KeyStr("D", 10U, kTypeSingleDeletion), ""},
-                          {KeyStr("E", 12U, kTypeSingleDeletion), ""},
-                          {KeyStr("E", 11U, kTypeValue), ""},
-                          {KeyStr("G", 15U, kTypeValue), "val"},
-                          {KeyStr("G", 12U, kTypeSingleDeletion), ""},
-                          {KeyStr("I", 14U, kTypeSingleDeletion), ""},
-                          {KeyStr("I", 13U, kTypeValue), ""},
-                          {KeyStr("J", 15U, kTypeValue), "val"},
-                          {KeyStr("K", 16U, kTypeSingleDeletion), ""},
-                          {KeyStr("K", 15U, kTypeValue), ""},
-                          {KeyStr("K", 11U, kTypeSingleDeletion), ""},
-                          {KeyStr("K", 8U, kTypeValue), "val3"},
-                          {KeyStr("L", 16U, kTypeSingleDeletion), ""},
-                          {KeyStr("L", 15U, kTypeValue), ""},
-                          {KeyStr("M", 16U, kTypeDeletion), ""},
-                          {KeyStr("M", 3U, kTypeSingleDeletion), ""}});
-
-  SetLastSequence(22U);
-  auto files = cfd_->current()->storage_info()->LevelFiles(0);
-  RunCompaction({files}, expected_results, {10U}, 10U);
-}
-
-// This test documents the behavior where a corrupt key follows a deletion or a
-// single deletion and the (single) deletion gets removed while the corrupt key
-// gets written out. TODO(noetzli): We probably want a better way to treat
-// corrupt keys.
-TEST_F(CompactionJobTest, CorruptionAfterDeletion) {
-  NewDB();
-
-  auto file1 =
-      mock::MakeMockFile({{test::KeyStr("A", 6U, kTypeValue), "val3"},
-                          {test::KeyStr("a", 5U, kTypeDeletion), ""},
-                          {test::KeyStr("a", 4U, kTypeValue, true), "val"}});
-  AddMockFile(file1);
-
-  auto file2 =
-      mock::MakeMockFile({{test::KeyStr("b", 3U, kTypeSingleDeletion), ""},
-                          {test::KeyStr("b", 2U, kTypeValue, true), "val"},
-                          {test::KeyStr("c", 1U, kTypeValue), "val2"}});
-  AddMockFile(file2);
-
-  auto expected_results =
-      mock::MakeMockFile({{test::KeyStr("A", 0U, kTypeValue), "val3"},
-                          {test::KeyStr("a", 0U, kTypeValue, true), "val"},
-                          {test::KeyStr("b", 0U, kTypeValue, true), "val"},
-                          {test::KeyStr("c", 1U, kTypeValue), "val2"}});
-
-  SetLastSequence(6U);
-  auto files = cfd_->current()->storage_info()->LevelFiles(0);
-  RunCompaction({files}, expected_results);
+  RunCompaction({ files }, expected_results);
 }

 }  // namespace rocksdb
@@ -916,14 +425,3 @@ int main(int argc, char** argv) {
  ::testing::InitGoogleTest(&argc, argv);
  return RUN_ALL_TESTS();
 }
-
-#else
-#include <stdio.h>
-
-int main(int argc, char** argv) {
-  fprintf(stderr,
-          "SKIPPED as CompactionJobStats is not supported in ROCKSDB_LITE\n");
-  return 0;
-}
-
-#endif  // ROCKSDB_LITE
@@ -243,7 +243,7 @@ bool CompactionPicker::ExpandWhileOverlapping(const std::string& cf_name,
 // Returns true if any one of specified files are being compacted
 bool CompactionPicker::FilesInCompaction(
    const std::vector<FileMetaData*>& files) {
-  for (size_t i = 0; i < files.size(); i++) {
+  for (unsigned int i = 0; i < files.size(); i++) {
    if (files[i]->being_compacted) {
      return true;
    }
@@ -255,30 +255,16 @@ Compaction* CompactionPicker::FormCompaction(
    const CompactionOptions& compact_options,
    const std::vector<CompactionInputFiles>& input_files, int output_level,
    VersionStorageInfo* vstorage, const MutableCFOptions& mutable_cf_options,
-    uint32_t output_path_id) {
+    uint32_t output_path_id) const {
  uint64_t max_grandparent_overlap_bytes =
      output_level + 1 < vstorage->num_levels() ?
          mutable_cf_options.MaxGrandParentOverlapBytes(output_level + 1) :
          std::numeric_limits<uint64_t>::max();
  assert(input_files.size());
-
-  // TODO(rven ): we might be able to run concurrent level 0 compaction
-  // if the key ranges of the two compactions do not overlap, but for now
-  // we do not allow it.
-  if ((input_files[0].level == 0) && !level0_compactions_in_progress_.empty()) {
-    return nullptr;
-  }
-  auto c = new Compaction(
+  return new Compaction(
      vstorage, mutable_cf_options, input_files, output_level,
      compact_options.output_file_size_limit, max_grandparent_overlap_bytes,
      output_path_id, compact_options.compression, /* grandparents */ {}, true);
-
-  // If it's level 0 compaction, make sure we don't execute any other level 0
-  // compactions in parallel
-  if ((c != nullptr) && (input_files[0].level == 0)) {
-    level0_compactions_in_progress_.insert(c);
-  }
-  return c;
 }

 Status CompactionPicker::GetCompactionInputsFromFileNumbers(
@@ -455,7 +441,7 @@ Compaction* CompactionPicker::CompactRange(
    const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
    VersionStorageInfo* vstorage, int input_level, int output_level,
    uint32_t output_path_id, const InternalKey* begin, const InternalKey* end,
-    InternalKey** compaction_end, bool* manual_conflict) {
+    InternalKey** compaction_end) {
  // CompactionPickerFIFO has its own implementation of compact range
  assert(ioptions_.compaction_style != kCompactionStyleFIFO);

@@ -481,12 +467,6 @@ Compaction* CompactionPicker::CompactRange(
      return nullptr;
    }

-    if ((start_level == 0) && (!level0_compactions_in_progress_.empty())) {
-      *manual_conflict = true;
-      // Only one level 0 compaction allowed
-      return nullptr;
-    }
-
    std::vector<CompactionInputFiles> inputs(vstorage->num_levels() -
                                             start_level);
    for (int level = start_level; level < vstorage->num_levels(); level++) {
@@ -495,21 +475,13 @@ Compaction* CompactionPicker::CompactRange(
      for (FileMetaData* f : vstorage->LevelFiles(level)) {
        files.push_back(f);
      }
-      if (FilesInCompaction(files)) {
-        *manual_conflict = true;
-        return nullptr;
-      }
    }
-    Compaction* c = new Compaction(
+    return new Compaction(
        vstorage, mutable_cf_options, std::move(inputs), output_level,
        mutable_cf_options.MaxFileSizeForLevel(output_level),
        /* max_grandparent_overlap_bytes */ LLONG_MAX, output_path_id,
        GetCompressionType(ioptions_, output_level, 1),
        /* grandparents */ {}, /* is manual */ true);
-    if (start_level == 0) {
-      level0_compactions_in_progress_.insert(c);
-    }
-    return c;
  }

  CompactionInputFiles inputs;
@@ -528,13 +500,6 @@ Compaction* CompactionPicker::CompactRange(
    return nullptr;
  }

-  if ((input_level == 0) && (!level0_compactions_in_progress_.empty())) {
-    // Only one level 0 compaction allowed
-    TEST_SYNC_POINT("CompactionPicker::CompactRange:Conflict");
-    *manual_conflict = true;
-    return nullptr;
-  }
-
  // Avoid compacting too much in one shot in case the range is large.
  // But we cannot do this for level-0 since level-0 files can overlap
  // and we must not pick one file and drop another older file if the
@@ -557,10 +522,9 @@ Compaction* CompactionPicker::CompactRange(
  assert(output_path_id < static_cast<uint32_t>(ioptions_.db_paths.size()));

  if (ExpandWhileOverlapping(cf_name, vstorage, &inputs) == false) {
-    // manual compaction is now multi-threaded, so it can
+    // manual compaction is currently single-threaded, so it should never
    // happen that ExpandWhileOverlapping fails
-    // we handle it higher in RunManualCompaction
-    *manual_conflict = true;
+    assert(false);
    return nullptr;
  }

@@ -579,10 +543,9 @@ Compaction* CompactionPicker::CompactRange(
    int parent_index = -1;
    if (!SetupOtherInputs(cf_name, mutable_cf_options, vstorage, &inputs,
                          &output_level_inputs, &parent_index, -1)) {
-      // manual compaction is now multi-threaded, so it can
+      // manual compaction is currently single-threaded, so it should never
      // happen that SetupOtherInputs fails
-      // we handle it higher in RunManualCompaction
-      *manual_conflict = true;
+      assert(false);
      return nullptr;
    }
  }
@@ -591,12 +554,6 @@ Compaction* CompactionPicker::CompactRange(
  if (!output_level_inputs.empty()) {
    compaction_inputs.push_back(output_level_inputs);
  }
-  for (size_t i = 0; i < compaction_inputs.size(); i++) {
-    if (FilesInCompaction(compaction_inputs[i].files)) {
-      *manual_conflict = true;
-      return nullptr;
-    }
-  }

  std::vector<FileMetaData*> grandparents;
  GetGrandparents(vstorage, inputs, output_level_inputs, &grandparents);
@@ -609,9 +566,6 @@ Compaction* CompactionPicker::CompactRange(
      std::move(grandparents), /* is manual compaction */ true);

  TEST_SYNC_POINT_CALLBACK("CompactionPicker::CompactRange:Return", compaction);
-  if (input_level == 0) {
-    level0_compactions_in_progress_.insert(compaction);
-  }
  return compaction;
 }

@@ -916,40 +870,22 @@ Compaction* LevelCompactionPicker::PickCompaction(
  int base_index = -1;
  CompactionInputFiles inputs;
  double score = 0;
-  CompactionReason compaction_reason = CompactionReason::kUnknown;

  // Find the compactions by size on all levels.
-  bool skipped_l0 = false;
  for (int i = 0; i < NumberLevels() - 1; i++) {
    score = vstorage->CompactionScore(i);
    level = vstorage->CompactionScoreLevel(i);
    assert(i == 0 || score <= vstorage->CompactionScore(i - 1));
    if (score >= 1) {
-      if (skipped_l0 && level == vstorage->base_level()) {
-        // If L0->base_level compaction is pending, don't schedule further
-        // compaction from base level. Otherwise L0->base_level compaction
-        // may starve.
-        continue;
-      }
      output_level = (level == 0) ? vstorage->base_level() : level + 1;
      if (PickCompactionBySize(vstorage, level, output_level, &inputs,
                               &parent_index, &base_index) &&
          ExpandWhileOverlapping(cf_name, vstorage, &inputs)) {
        // found the compaction!
-        if (level == 0) {
-          // L0 score = `num L0 files` / `level0_file_num_compaction_trigger`
-          compaction_reason = CompactionReason::kLevelL0FilesNum;
-        } else {
-          // L1+ score = `Level files size` / `MaxBytesForLevel`
-          compaction_reason = CompactionReason::kLevelMaxLevelSize;
-        }
        break;
      } else {
        // didn't find the compaction, clear the inputs
        inputs.clear();
-        if (level == 0) {
-          skipped_l0 = true;
-        }
      }
    }
  }
@@ -962,9 +898,6 @@ Compaction* LevelCompactionPicker::PickCompaction(
    parent_index = base_index = -1;
    PickFilesMarkedForCompactionExperimental(cf_name, vstorage, &inputs, &level,
                                             &output_level);
-    if (!inputs.empty()) {
-      compaction_reason = CompactionReason::kFilesMarkedForCompaction;
-    }
  }
  if (inputs.empty()) {
    return nullptr;
@@ -1015,8 +948,7 @@ Compaction* LevelCompactionPicker::PickCompaction(
      mutable_cf_options.MaxGrandParentOverlapBytes(level),
      GetPathId(ioptions_, mutable_cf_options, output_level),
      GetCompressionType(ioptions_, output_level, vstorage->base_level()),
-      std::move(grandparents), is_manual, score,
-      false /* deletion_compaction */, compaction_reason);
+      std::move(grandparents), is_manual, score);

  // If it's level 0 compaction, make sure we don't execute any other level 0
  // compactions in parallel
@@ -1087,7 +1019,6 @@ bool LevelCompactionPicker::PickCompactionBySize(VersionStorageInfo* vstorage,
  // could be made better by looking at key-ranges that are
  // being compacted at level 0.
  if (level == 0 && !level0_compactions_in_progress_.empty()) {
-    TEST_SYNC_POINT("LevelCompactionPicker::PickCompactionBySize:0");
    return false;
  }

@@ -1097,7 +1028,7 @@ bool LevelCompactionPicker::PickCompactionBySize(VersionStorageInfo* vstorage,

  // Pick the largest file in this level that is not already
  // being compacted
-  const std::vector<int>& file_size = vstorage->FilesByCompactionPri(level);
+  const std::vector<int>& file_size = vstorage->FilesBySize(level);
  const std::vector<FileMetaData*>& level_files = vstorage->LevelFiles(level);

  // record the first file that is not yet compacted
@@ -1108,6 +1039,11 @@ bool LevelCompactionPicker::PickCompactionBySize(VersionStorageInfo* vstorage,
    int index = file_size[i];
    auto* f = level_files[index];

+    assert((i == file_size.size() - 1) ||
+           (i >= VersionStorageInfo::kNumberFilesToSort - 1) ||
+           (f->compensated_file_size >=
+            level_files[file_size[i + 1]]->compensated_file_size));
+
    // do not pick a file to compact if it is being compacted
    // from n-1 level.
    if (f->being_compacted) {
@@ -1164,19 +1100,18 @@ void UniversalCompactionPicker::SortedRun::Dump(char* out_buf,
 }

 void UniversalCompactionPicker::SortedRun::DumpSizeInfo(
-    char* out_buf, size_t out_buf_size, size_t sorted_run_count) const {
+    char* out_buf, size_t out_buf_size, int sorted_run_count) const {
  if (level == 0) {
    assert(file != nullptr);
    snprintf(out_buf, out_buf_size,
-             "file %" PRIu64 "[%" ROCKSDB_PRIszt
-             "] "
+             "file %" PRIu64
+             "[%d] "
             "with size %" PRIu64 " (compensated size %" PRIu64 ")",
             file->fd.GetNumber(), sorted_run_count, file->fd.GetFileSize(),
             file->compensated_file_size);
  } else {
    snprintf(out_buf, out_buf_size,
-             "level %d[%" ROCKSDB_PRIszt
-             "] "
+             "level %d[%d] "
             "with size %" PRIu64 " (compensated size %" PRIu64 ")",
             level, sorted_run_count, size, compensated_file_size);
  }
@@ -1304,9 +1239,8 @@ Compaction* UniversalCompactionPicker::PickCompaction(
  std::vector<SortedRun> sorted_runs =
      CalculateSortedRuns(*vstorage, ioptions_);

-  if (sorted_runs.size() == 0 ||
-      sorted_runs.size() <
-          (unsigned int)mutable_cf_options.level0_file_num_compaction_trigger) {
+  if (sorted_runs.size() <
+      (unsigned int)mutable_cf_options.level0_file_num_compaction_trigger) {
    LogToBuffer(log_buffer, "[%s] Universal: nothing to do\n", cf_name.c_str());
    return nullptr;
  }
@@ -1388,13 +1322,7 @@ Compaction* UniversalCompactionPicker::PickCompaction(
                              &largest_seqno);
      if (is_first) {
        is_first = false;
-      } else if (prev_smallest_seqno > 0) {
-        // A level is considered as the bottommost level if there are
-        // no files in higher levels or if files in higher levels do
-        // not overlap with the files being compacted. Sequence numbers
-        // of files in bottommost level can be set to 0 to help
-        // compression. As a result, the following assert may not hold
-        // if the prev_smallest_seqno is 0.
+      } else {
        assert(prev_smallest_seqno > largest_seqno);
      }
      prev_smallest_seqno = smallest_seqno;
@@ -1458,21 +1386,16 @@ Compaction* UniversalCompactionPicker::PickCompactionUniversalReadAmp(

  const SortedRun* sr = nullptr;
  bool done = false;
-  size_t start_index = 0;
+  int start_index = 0;
  unsigned int candidate_count = 0;

  unsigned int max_files_to_compact = std::min(max_merge_width,
                                       max_number_of_files_to_compact);
  min_merge_width = std::max(min_merge_width, 2U);

-  // Caller checks the size before executing this function. This invariant is
-  // important because otherwise we may have a possible integer underflow when
-  // dealing with unsigned types.
-  assert(sorted_runs.size() > 0);
-
  // Considers a candidate file only if it is smaller than the
  // total size accumulated so far.
-  for (size_t loop = 0; loop < sorted_runs.size(); loop++) {
+  for (unsigned int loop = 0; loop < sorted_runs.size(); loop++) {
    candidate_count = 0;

    // Skip files that are already being compacted
@@ -1504,7 +1427,7 @@ Compaction* UniversalCompactionPicker::PickCompactionUniversalReadAmp(
    }

    // Check if the succeeding files need compaction.
-    for (size_t i = loop + 1;
+    for (unsigned int i = loop + 1;
         candidate_count < max_files_to_compact && i < sorted_runs.size();
         i++) {
      const SortedRun* succeeding_sr = &sorted_runs[i];
@@ -1546,7 +1469,7 @@ Compaction* UniversalCompactionPicker::PickCompactionUniversalReadAmp(
      done = true;
      break;
    } else {
-      for (size_t i = loop;
+      for (unsigned int i = loop;
           i < loop + candidate_count && i < sorted_runs.size(); i++) {
        const SortedRun* skipping_sr = &sorted_runs[i];
        char file_num_buf[256];
@@ -1559,7 +1482,7 @@ Compaction* UniversalCompactionPicker::PickCompactionUniversalReadAmp(
  if (!done || candidate_count <= 1) {
    return nullptr;
  }
-  size_t first_index_after = start_index + candidate_count;
+  unsigned int first_index_after = start_index + candidate_count;
  // Compression is enabled if files compacted earlier already reached
  // size ratio of compression.
  bool enable_compression = true;
@@ -1600,7 +1523,7 @@ Compaction* UniversalCompactionPicker::PickCompactionUniversalReadAmp(
  for (size_t i = 0; i < inputs.size(); ++i) {
    inputs[i].level = start_level + static_cast<int>(i);
  }
-  for (size_t i = start_index; i < first_index_after; i++) {
+  for (unsigned int i = start_index; i < first_index_after; i++) {
    auto& picking_sr = sorted_runs[i];
    if (picking_sr.level == 0) {
      FileMetaData* picking_file = picking_sr.file;
@@ -1617,18 +1540,11 @@ Compaction* UniversalCompactionPicker::PickCompactionUniversalReadAmp(
                file_num_buf);
  }

-  CompactionReason compaction_reason;
-  if (max_number_of_files_to_compact == UINT_MAX) {
-    compaction_reason = CompactionReason::kUniversalSortedRunNum;
-  } else {
-    compaction_reason = CompactionReason::kUniversalSizeRatio;
-  }
  return new Compaction(
      vstorage, mutable_cf_options, std::move(inputs), output_level,
      mutable_cf_options.MaxFileSizeForLevel(output_level), LLONG_MAX, path_id,
      GetCompressionType(ioptions_, start_level, 1, enable_compression),
-      /* grandparents */ {}, /* is manual */ false, score,
-      false /* deletion_compaction */, compaction_reason);
+      /* grandparents */ {}, /* is manual */ false, score);
 }

 // Look at overall size amplification. If size amplification
@@ -1647,11 +1563,11 @@ Compaction* UniversalCompactionPicker::PickCompactionUniversalSizeAmp(

  unsigned int candidate_count = 0;
  uint64_t candidate_size = 0;
-  size_t start_index = 0;
+  unsigned int start_index = 0;
  const SortedRun* sr = nullptr;

  // Skip files that are already being compacted
-  for (size_t loop = 0; loop < sorted_runs.size() - 1; loop++) {
+  for (unsigned int loop = 0; loop < sorted_runs.size() - 1; loop++) {
    sr = &sorted_runs[loop];
    if (!sr->being_compacted) {
      start_index = loop;         // Consider this as the first candidate.
@@ -1671,14 +1587,13 @@ Compaction* UniversalCompactionPicker::PickCompactionUniversalSizeAmp(
  {
    char file_num_buf[kFormatFileNumberBufSize];
    sr->Dump(file_num_buf, sizeof(file_num_buf), true);
-    LogToBuffer(log_buffer,
-                "[%s] Universal: First candidate %s[%" ROCKSDB_PRIszt "] %s",
+    LogToBuffer(log_buffer, "[%s] Universal: First candidate %s[%d] %s",
                cf_name.c_str(), file_num_buf, start_index,
                " to reduce size amp.\n");
  }

  // keep adding up all the remaining files
-  for (size_t loop = start_index; loop < sorted_runs.size() - 1; loop++) {
+  for (unsigned int loop = start_index; loop < sorted_runs.size() - 1; loop++) {
    sr = &sorted_runs[loop];
    if (sr->being_compacted) {
      char file_num_buf[kFormatFileNumberBufSize];
@@ -1718,7 +1633,7 @@ Compaction* UniversalCompactionPicker::PickCompactionUniversalSizeAmp(

  // Estimate total file size
  uint64_t estimated_total_size = 0;
-  for (size_t loop = start_index; loop < sorted_runs.size(); loop++) {
+  for (unsigned int loop = start_index; loop < sorted_runs.size(); loop++) {
    estimated_total_size += sorted_runs[loop].size;
  }
  uint32_t path_id = GetPathId(ioptions_, estimated_total_size);
@@ -1729,7 +1644,7 @@ Compaction* UniversalCompactionPicker::PickCompactionUniversalSizeAmp(
    inputs[i].level = start_level + static_cast<int>(i);
  }
  // We always compact all the files, so always compress.
-  for (size_t loop = start_index; loop < sorted_runs.size(); loop++) {
+  for (unsigned int loop = start_index; loop < sorted_runs.size(); loop++) {
    auto& picking_sr = sorted_runs[loop];
    if (picking_sr.level == 0) {
      FileMetaData* f = picking_sr.file;
@@ -1741,7 +1656,7 @@ Compaction* UniversalCompactionPicker::PickCompactionUniversalSizeAmp(
      }
    }
    char file_num_buf[256];
-    picking_sr.DumpSizeInfo(file_num_buf, sizeof(file_num_buf), loop);
+    sr->DumpSizeInfo(file_num_buf, sizeof(file_num_buf), loop);
    LogToBuffer(log_buffer, "[%s] Universal: size amp picking %s",
                cf_name.c_str(), file_num_buf);
  }
@@ -1752,9 +1667,7 @@ Compaction* UniversalCompactionPicker::PickCompactionUniversalSizeAmp(
      mutable_cf_options.MaxFileSizeForLevel(vstorage->num_levels() - 1),
      /* max_grandparent_overlap_bytes */ LLONG_MAX, path_id,
      GetCompressionType(ioptions_, vstorage->num_levels() - 1, 1),
-      /* grandparents */ {}, /* is manual */ false, score,
-      false /* deletion_compaction */,
-      CompactionReason::kUniversalSizeAmplification);
+      /* grandparents */ {}, /* is manual */ false, score);
 }

 bool FIFOCompactionPicker::NeedsCompaction(const VersionStorageInfo* vstorage)
@@ -1813,7 +1726,7 @@ Compaction* FIFOCompactionPicker::PickCompaction(
  Compaction* c = new Compaction(
      vstorage, mutable_cf_options, std::move(inputs), 0, 0, 0, 0,
      kNoCompression, {}, /* is manual */ false, vstorage->CompactionScore(0),
-      /* is deletion compaction */ true, CompactionReason::kFIFOMaxSize);
+      /* is deletion compaction */ true);
  level0_compactions_in_progress_.insert(c);
  return c;
 }
@@ -1822,7 +1735,7 @@ Compaction* FIFOCompactionPicker::CompactRange(
    const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
    VersionStorageInfo* vstorage, int input_level, int output_level,
    uint32_t output_path_id, const InternalKey* begin, const InternalKey* end,
-    InternalKey** compaction_end, bool* manual_conflict) {
+    InternalKey** compaction_end) {
  assert(input_level == 0);
  assert(output_level == 0);
  *compaction_end = nullptr;
@@ -60,7 +60,7 @@ class CompactionPicker {
      const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
      VersionStorageInfo* vstorage, int input_level, int output_level,
      uint32_t output_path_id, const InternalKey* begin, const InternalKey* end,
-      InternalKey** compaction_end, bool* manual_conflict);
+      InternalKey** compaction_end);

  // The maximum allowed output level.  Default value is NumberLevels() - 1.
  virtual int MaxOutputLevel() const {
@@ -94,7 +94,7 @@ class CompactionPicker {
      const CompactionOptions& compact_options,
      const std::vector<CompactionInputFiles>& input_files, int output_level,
      VersionStorageInfo* vstorage, const MutableCFOptions& mutable_cf_options,
-      uint32_t output_path_id);
+      uint32_t output_path_id) const;

  // Converts a set of compaction input file numbers into
  // a list of CompactionInputFiles.
@@ -110,11 +110,6 @@ class CompactionPicker {
  // overlapping.
  bool IsInputNonOverlapping(Compaction* c);

-  // Is there currently a compaction involving level 0 taking place
-  bool IsLevel0CompactionInProgress() const {
-    return !level0_compactions_in_progress_.empty();
-  }
-
 protected:
  int NumberLevels() const { return ioptions_.num_levels; }

@@ -251,7 +246,7 @@ class UniversalCompactionPicker : public CompactionPicker {

    // sorted_run_count is added into the string to print
    void DumpSizeInfo(char* out_buf, size_t out_buf_size,
-                      size_t sorted_run_count) const;
+                      int sorted_run_count) const;

    int level;
    // `file` Will be null for level > 0. For level = 0, the sorted run is
@@ -302,7 +297,7 @@ class FIFOCompactionPicker : public CompactionPicker {
      const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
      VersionStorageInfo* vstorage, int input_level, int output_level,
      uint32_t output_path_id, const InternalKey* begin, const InternalKey* end,
-      InternalKey** compaction_end, bool* manual_conflict) override;
+      InternalKey** compaction_end) override;

  // The maximum allowed output level.  Always returns 0.
  virtual int MaxOutputLevel() const override {
@@ -329,13 +324,11 @@ class NullCompactionPicker : public CompactionPicker {
  }

  // Always return "nullptr"
-  Compaction* CompactRange(const std::string& cf_name,
-                           const MutableCFOptions& mutable_cf_options,
-                           VersionStorageInfo* vstorage, int input_level,
-                           int output_level, uint32_t output_path_id,
-                           const InternalKey* begin, const InternalKey* end,
-                           InternalKey** compaction_end,
-                           bool* manual_conflict) override {
+  Compaction* CompactRange(
+      const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
+      VersionStorageInfo* vstorage, int input_level, int output_level,
+      uint32_t output_path_id, const InternalKey* begin, const InternalKey* end,
+      InternalKey** compaction_end) override {
    return nullptr;
  }

@@ -3,12 +3,9 @@
 //  LICENSE file in the root directory of this source tree. An additional grant
 //  of patent rights can be found in the PATENTS file in the same directory.

-#include "db/compaction.h"
 #include "db/compaction_picker.h"
 #include <limits>
 #include <string>
-#include <utility>
-
 #include "util/logging.h"
 #include "util/string_util.h"
 #include "util/testharness.h"
@@ -38,11 +35,6 @@ class CompactionPickerTest : public testing::Test {
  CompactionOptionsFIFO fifo_options_;
  std::unique_ptr<VersionStorageInfo> vstorage_;
  std::vector<std::unique_ptr<FileMetaData>> files_;
-  // does not own FileMetaData
-  std::unordered_map<uint32_t, std::pair<FileMetaData*, int>> file_map_;
-  // input files to compaction process.
-  std::vector<CompactionInputFiles> input_files_;
-  int compaction_level_start_;

  CompactionPickerTest()
      : ucmp_(BytewiseComparator()),
@@ -74,8 +66,6 @@ class CompactionPickerTest : public testing::Test {
  void DeleteVersionStorage() {
    vstorage_.reset();
    files_.clear();
-    file_map_.clear();
-    input_files_.clear();
  }

  void Add(int level, uint32_t file_number, const char* smallest,
@@ -93,29 +83,11 @@ class CompactionPickerTest : public testing::Test {
    f->refs = 0;
    vstorage_->AddFile(level, f);
    files_.emplace_back(f);
-    file_map_.insert({file_number, {f, level}});
-  }
-
-  void SetCompactionInputFilesLevels(int level_count, int start_level) {
-    input_files_.resize(level_count);
-    for (int i = 0; i < level_count; ++i) {
-      input_files_[i].level = start_level + i;
-    }
-    compaction_level_start_ = start_level;
-  }
-
-  void AddToCompactionFiles(uint32_t file_number) {
-    auto iter = file_map_.find(file_number);
-    assert(iter != file_map_.end());
-    int level = iter->second.second;
-    assert(level < vstorage_->num_levels());
-    input_files_[level - compaction_level_start_].files.emplace_back(
-        iter->second.first);
  }

  void UpdateVersionStorageInfo() {
    vstorage_->CalculateBaseBytes(ioptions_, mutable_cf_options_);
-    vstorage_->UpdateFilesByCompactionPri(mutable_cf_options_);
+    vstorage_->UpdateFilesBySize();
    vstorage_->UpdateNumNonEmptyLevels();
    vstorage_->GenerateFileIndexer();
    vstorage_->GenerateLevelFilesBrief();
@@ -580,101 +552,6 @@ TEST_F(CompactionPickerTest, OverlappingUserKeys3) {
  ASSERT_EQ(7U, compaction->input(1, 1)->fd.GetNumber());
 }

-TEST_F(CompactionPickerTest, NotScheduleL1IfL0WithHigherPri1) {
-  NewVersionStorage(6, kCompactionStyleLevel);
-  mutable_cf_options_.level0_file_num_compaction_trigger = 2;
-  mutable_cf_options_.max_bytes_for_level_base = 900000000U;
-
-  // 6 L0 files, score 3.
-  Add(0, 1U, "000", "400", 1U);
-  Add(0, 2U, "001", "400", 1U, 0, 0);
-  Add(0, 3U, "001", "400", 1000000000U, 0, 0);
-  Add(0, 31U, "001", "400", 1000000000U, 0, 0);
-  Add(0, 32U, "001", "400", 1000000000U, 0, 0);
-  Add(0, 33U, "001", "400", 1000000000U, 0, 0);
-
-  // L1 total size 2GB, score 2.2. If one file being comapcted, score 1.1.
-  Add(1, 4U, "050", "300", 1000000000U, 0, 0);
-  file_map_[4u].first->being_compacted = true;
-  Add(1, 5U, "301", "350", 1000000000U, 0, 0);
-
-  // Output level overlaps with the beginning and the end of the chain
-  Add(2, 6U, "050", "100", 1U);
-  Add(2, 7U, "300", "400", 1U);
-
-  // No compaction should be scheduled, if L0 has higher priority than L1
-  // but L0->L1 compaction is blocked by a file in L1 being compacted.
-  UpdateVersionStorageInfo();
-  ASSERT_EQ(0, vstorage_->CompactionScoreLevel(0));
-  ASSERT_EQ(1, vstorage_->CompactionScoreLevel(1));
-  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
-  ASSERT_TRUE(compaction.get() == nullptr);
-}
-
-TEST_F(CompactionPickerTest, NotScheduleL1IfL0WithHigherPri2) {
-  NewVersionStorage(6, kCompactionStyleLevel);
-  mutable_cf_options_.level0_file_num_compaction_trigger = 2;
-  mutable_cf_options_.max_bytes_for_level_base = 900000000U;
-
-  // 6 L0 files, score 3.
-  Add(0, 1U, "000", "400", 1U);
-  Add(0, 2U, "001", "400", 1U, 0, 0);
-  Add(0, 3U, "001", "400", 1000000000U, 0, 0);
-  Add(0, 31U, "001", "400", 1000000000U, 0, 0);
-  Add(0, 32U, "001", "400", 1000000000U, 0, 0);
-  Add(0, 33U, "001", "400", 1000000000U, 0, 0);
-
-  // L1 total size 2GB, score 2.2. If one file being comapcted, score 1.1.
-  Add(1, 4U, "050", "300", 1000000000U, 0, 0);
-  Add(1, 5U, "301", "350", 1000000000U, 0, 0);
-
-  // Output level overlaps with the beginning and the end of the chain
-  Add(2, 6U, "050", "100", 1U);
-  Add(2, 7U, "300", "400", 1U);
-
-  // If no file in L1 being compacted, L0->L1 compaction will be scheduled.
-  UpdateVersionStorageInfo();  // being_compacted flag is cleared here.
-  ASSERT_EQ(0, vstorage_->CompactionScoreLevel(0));
-  ASSERT_EQ(1, vstorage_->CompactionScoreLevel(1));
-  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
-  ASSERT_TRUE(compaction.get() != nullptr);
-}
-
-TEST_F(CompactionPickerTest, NotScheduleL1IfL0WithHigherPri3) {
-  NewVersionStorage(6, kCompactionStyleLevel);
-  mutable_cf_options_.level0_file_num_compaction_trigger = 2;
-  mutable_cf_options_.max_bytes_for_level_base = 900000000U;
-
-  // 6 L0 files, score 3.
-  Add(0, 1U, "000", "400", 1U);
-  Add(0, 2U, "001", "400", 1U, 0, 0);
-  Add(0, 3U, "001", "400", 1000000000U, 0, 0);
-  Add(0, 31U, "001", "400", 1000000000U, 0, 0);
-  Add(0, 32U, "001", "400", 1000000000U, 0, 0);
-  Add(0, 33U, "001", "400", 1000000000U, 0, 0);
-
-  // L1 score more than 6.
-  Add(1, 4U, "050", "300", 1000000000U, 0, 0);
-  file_map_[4u].first->being_compacted = true;
-  Add(1, 5U, "301", "350", 1000000000U, 0, 0);
-  Add(1, 51U, "351", "400", 6000000000U, 0, 0);
-
-  // Output level overlaps with the beginning and the end of the chain
-  Add(2, 6U, "050", "100", 1U);
-  Add(2, 7U, "300", "400", 1U);
-
-  // If score in L1 is larger than L0, L1 compaction goes through despite
-  // there is pending L0 compaction.
-  UpdateVersionStorageInfo();
-  ASSERT_EQ(1, vstorage_->CompactionScoreLevel(0));
-  ASSERT_EQ(0, vstorage_->CompactionScoreLevel(1));
-  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
-  ASSERT_TRUE(compaction.get() != nullptr);
-}
-
 TEST_F(CompactionPickerTest, EstimateCompactionBytesNeeded1) {
  int num_levels = ioptions_.num_levels;
  ioptions_.level_compaction_dynamic_level_bytes = false;
@@ -760,164 +637,6 @@ TEST_F(CompactionPickerTest, EstimateCompactionBytesNeededDynamicLevel) {
            vstorage_->estimated_compaction_needed_bytes());
 }

-TEST_F(CompactionPickerTest, IsBottommostLevelTest) {
-  // case 1: Higher levels are empty
-  NewVersionStorage(6, kCompactionStyleLevel);
-  Add(0, 1U, "a", "m");
-  Add(0, 2U, "c", "z");
-  Add(1, 3U, "d", "e");
-  Add(1, 4U, "l", "p");
-  Add(2, 5U, "g", "i");
-  Add(2, 6U, "x", "z");
-  UpdateVersionStorageInfo();
-  SetCompactionInputFilesLevels(2, 1);
-  AddToCompactionFiles(3U);
-  AddToCompactionFiles(5U);
-  bool result =
-      Compaction::TEST_IsBottommostLevel(2, vstorage_.get(), input_files_);
-  ASSERT_TRUE(result);
-
-  // case 2: Higher levels have no overlap
-  NewVersionStorage(6, kCompactionStyleLevel);
-  Add(0, 1U, "a", "m");
-  Add(0, 2U, "c", "z");
-  Add(1, 3U, "d", "e");
-  Add(1, 4U, "l", "p");
-  Add(2, 5U, "g", "i");
-  Add(2, 6U, "x", "z");
-  Add(3, 7U, "k", "p");
-  Add(3, 8U, "t", "w");
-  Add(4, 9U, "a", "b");
-  Add(5, 10U, "c", "cc");
-  UpdateVersionStorageInfo();
-  SetCompactionInputFilesLevels(2, 1);
-  AddToCompactionFiles(3U);
-  AddToCompactionFiles(5U);
-  result = Compaction::TEST_IsBottommostLevel(2, vstorage_.get(), input_files_);
-  ASSERT_TRUE(result);
-
-  // case 3.1: Higher levels (level 3) have overlap
-  NewVersionStorage(6, kCompactionStyleLevel);
-  Add(0, 1U, "a", "m");
-  Add(0, 2U, "c", "z");
-  Add(1, 3U, "d", "e");
-  Add(1, 4U, "l", "p");
-  Add(2, 5U, "g", "i");
-  Add(2, 6U, "x", "z");
-  Add(3, 7U, "e", "g");
-  Add(3, 8U, "h", "k");
-  Add(4, 9U, "a", "b");
-  Add(5, 10U, "c", "cc");
-  UpdateVersionStorageInfo();
-  SetCompactionInputFilesLevels(2, 1);
-  AddToCompactionFiles(3U);
-  AddToCompactionFiles(5U);
-  result = Compaction::TEST_IsBottommostLevel(2, vstorage_.get(), input_files_);
-  ASSERT_FALSE(result);
-
-  // case 3.2: Higher levels (level 5) have overlap
-  DeleteVersionStorage();
-  NewVersionStorage(6, kCompactionStyleLevel);
-  Add(0, 1U, "a", "m");
-  Add(0, 2U, "c", "z");
-  Add(1, 3U, "d", "e");
-  Add(1, 4U, "l", "p");
-  Add(2, 5U, "g", "i");
-  Add(2, 6U, "x", "z");
-  Add(3, 7U, "j", "k");
-  Add(3, 8U, "l", "m");
-  Add(4, 9U, "a", "b");
-  Add(5, 10U, "c", "cc");
-  Add(5, 11U, "h", "k");
-  Add(5, 12U, "y", "yy");
-  Add(5, 13U, "z", "zz");
-  UpdateVersionStorageInfo();
-  SetCompactionInputFilesLevels(2, 1);
-  AddToCompactionFiles(3U);
-  AddToCompactionFiles(5U);
-  result = Compaction::TEST_IsBottommostLevel(2, vstorage_.get(), input_files_);
-  ASSERT_FALSE(result);
-
-  // case 3.3: Higher levels (level 5) have overlap, but it's only overlapping
-  // one key ("d")
-  NewVersionStorage(6, kCompactionStyleLevel);
-  Add(0, 1U, "a", "m");
-  Add(0, 2U, "c", "z");
-  Add(1, 3U, "d", "e");
-  Add(1, 4U, "l", "p");
-  Add(2, 5U, "g", "i");
-  Add(2, 6U, "x", "z");
-  Add(3, 7U, "j", "k");
-  Add(3, 8U, "l", "m");
-  Add(4, 9U, "a", "b");
-  Add(5, 10U, "c", "cc");
-  Add(5, 11U, "ccc", "d");
-  Add(5, 12U, "y", "yy");
-  Add(5, 13U, "z", "zz");
-  UpdateVersionStorageInfo();
-  SetCompactionInputFilesLevels(2, 1);
-  AddToCompactionFiles(3U);
-  AddToCompactionFiles(5U);
-  result = Compaction::TEST_IsBottommostLevel(2, vstorage_.get(), input_files_);
-  ASSERT_FALSE(result);
-
-  // Level 0 files overlap
-  NewVersionStorage(6, kCompactionStyleLevel);
-  Add(0, 1U, "s", "t");
-  Add(0, 2U, "a", "m");
-  Add(0, 3U, "b", "z");
-  Add(0, 4U, "e", "f");
-  Add(5, 10U, "y", "z");
-  UpdateVersionStorageInfo();
-  SetCompactionInputFilesLevels(1, 0);
-  AddToCompactionFiles(1U);
-  AddToCompactionFiles(2U);
-  AddToCompactionFiles(3U);
-  AddToCompactionFiles(4U);
-  result = Compaction::TEST_IsBottommostLevel(2, vstorage_.get(), input_files_);
-  ASSERT_FALSE(result);
-
-  // Level 0 files don't overlap
-  NewVersionStorage(6, kCompactionStyleLevel);
-  Add(0, 1U, "s", "t");
-  Add(0, 2U, "a", "m");
-  Add(0, 3U, "b", "k");
-  Add(0, 4U, "e", "f");
-  Add(5, 10U, "y", "z");
-  UpdateVersionStorageInfo();
-  SetCompactionInputFilesLevels(1, 0);
-  AddToCompactionFiles(1U);
-  AddToCompactionFiles(2U);
-  AddToCompactionFiles(3U);
-  AddToCompactionFiles(4U);
-  result = Compaction::TEST_IsBottommostLevel(2, vstorage_.get(), input_files_);
-  ASSERT_TRUE(result);
-
-  // Level 1 files overlap
-  NewVersionStorage(6, kCompactionStyleLevel);
-  Add(0, 1U, "s", "t");
-  Add(0, 2U, "a", "m");
-  Add(0, 3U, "b", "k");
-  Add(0, 4U, "e", "f");
-  Add(1, 5U, "a", "m");
-  Add(1, 6U, "n", "o");
-  Add(1, 7U, "w", "y");
-  Add(5, 10U, "y", "z");
-  UpdateVersionStorageInfo();
-  SetCompactionInputFilesLevels(2, 0);
-  AddToCompactionFiles(1U);
-  AddToCompactionFiles(2U);
-  AddToCompactionFiles(3U);
-  AddToCompactionFiles(4U);
-  AddToCompactionFiles(5U);
-  AddToCompactionFiles(6U);
-  AddToCompactionFiles(7U);
-  result = Compaction::TEST_IsBottommostLevel(2, vstorage_.get(), input_files_);
-  ASSERT_FALSE(result);
-
-  DeleteVersionStorage();
-}
-
 }  // namespace rocksdb

 int main(int argc, char** argv) {
@@ -7,11 +7,10 @@
 #include <map>
 #include <string>

-#include "memtable/stl_wrappers.h"
 #include "rocksdb/db.h"
 #include "rocksdb/env.h"
 #include "util/hash.h"
-#include "util/kv_map.h"
+#include "util/stl_wrappers.h"
 #include "util/string_util.h"
 #include "util/testharness.h"
 #include "util/testutil.h"
@@ -18,13 +18,6 @@ namespace rocksdb {
 void CancelAllBackgroundWork(DB* db, bool wait) {
  (dynamic_cast<DBImpl*>(db))->CancelAllBackgroundWork(wait);
 }
-
-Status DeleteFilesInRange(DB* db, ColumnFamilyHandle* column_family,
-                          const Slice* begin, const Slice* end) {
-  return (dynamic_cast<DBImpl*>(db))
-      ->DeleteFilesInRange(column_family, begin, end);
-}
-
 }  // namespace rocksdb

 #endif  // ROCKSDB_LITE
@@ -7,11 +7,8 @@
 // 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.

-#ifndef ROCKSDB_LITE
-
 #include "rocksdb/db.h"

-#include <inttypes.h>
 #include <errno.h>
 #include <fcntl.h>
 #include <sys/stat.h>
@@ -88,14 +85,10 @@ class CorruptionTest : public testing::Test {
    ASSERT_OK(::rocksdb::RepairDB(dbname_, options_));
  }

-  void Build(int n, int flush_every = 0) {
+  void Build(int n) {
    std::string key_space, value_space;
    WriteBatch batch;
    for (int i = 0; i < n; i++) {
-      if (flush_every != 0 && i != 0 && i % flush_every == 0) {
-        DBImpl* dbi = reinterpret_cast<DBImpl*>(db_);
-        dbi->TEST_FlushMemTable();
-      }
      //if ((i % 100) == 0) fprintf(stderr, "@ %d of %d\n", i, n);
      Slice key = Key(i, &key_space);
      batch.Clear();
@@ -105,8 +98,8 @@ class CorruptionTest : public testing::Test {
  }

  void Check(int min_expected, int max_expected) {
-    uint64_t next_expected = 0;
-    uint64_t missed = 0;
+    unsigned int next_expected = 0;
+    int missed = 0;
    int bad_keys = 0;
    int bad_values = 0;
    int correct = 0;
@@ -127,7 +120,7 @@ class CorruptionTest : public testing::Test {
        continue;
      }
      missed += (key - next_expected);
-      next_expected = key + 1;
+      next_expected = static_cast<unsigned int>(key + 1);
      if (iter->value() != Value(static_cast<int>(key), &value_space)) {
        bad_values++;
      } else {
@@ -137,9 +130,8 @@ class CorruptionTest : public testing::Test {
    delete iter;

    fprintf(stderr,
-      "expected=%d..%d; got=%d; bad_keys=%d; bad_values=%d; missed=%llu\n",
-            min_expected, max_expected, correct, bad_keys, bad_values,
-            static_cast<unsigned long long>(missed));
+            "expected=%d..%d; got=%d; bad_keys=%d; bad_values=%d; missed=%d\n",
+            min_expected, max_expected, correct, bad_keys, bad_values, missed);
    ASSERT_LE(min_expected, correct);
    ASSERT_GE(max_expected, correct);
  }
@@ -185,7 +177,7 @@ class CorruptionTest : public testing::Test {
    FileType type;
    std::string fname;
    int picked_number = -1;
-    for (size_t i = 0; i < filenames.size(); i++) {
+    for (unsigned int i = 0; i < filenames.size(); i++) {
      if (ParseFileName(filenames[i], &number, &type) &&
          type == filetype &&
          static_cast<int>(number) > picked_number) {  // Pick latest file
@@ -234,16 +226,8 @@ class CorruptionTest : public testing::Test {

  // Return the value to associate with the specified key
  Slice Value(int k, std::string* storage) {
-    if (k == 0) {
-      // Ugh.  Random seed of 0 used to produce no entropy.  This code
-      // preserves the implementation that was in place when all of the
-      // magic values in this file were picked.
-      *storage = std::string(kValueSize, ' ');
-      return Slice(*storage);
-    } else {
-      Random r(k);
-      return test::RandomString(&r, kValueSize, storage);
-    }
+    Random r(k);
+    return test::RandomString(&r, kValueSize, storage);
  }
 };

@@ -311,21 +295,13 @@ TEST_F(CorruptionTest, TableFile) {
 }

 TEST_F(CorruptionTest, TableFileIndexData) {
-  Options options;
-  // very big, we'll trigger flushes manually
-  options.write_buffer_size = 100 * 1024 * 1024;
-  Reopen(&options);
-  // build 2 tables, flush at 5000
-  Build(10000, 5000);
+  Build(10000);  // Enough to build multiple Tables
  DBImpl* dbi = reinterpret_cast<DBImpl*>(db_);
  dbi->TEST_FlushMemTable();

-  // corrupt an index block of an entire file
  Corrupt(kTableFile, -2000, 500);
  Reopen();
-  // one full file should be readable, since only one was corrupted
-  // the other file should be fully non-readable, since index was corrupted
-  Check(5000, 5000);
+  Check(5000, 9999);
 }

 TEST_F(CorruptionTest, MissingDescriptor) {
@@ -491,13 +467,3 @@ int main(int argc, char** argv) {
  ::testing::InitGoogleTest(&argc, argv);
  return RUN_ALL_TESTS();
 }
-
-#else
-#include <stdio.h>
-
-int main(int argc, char** argv) {
-  fprintf(stderr, "SKIPPED as RepairDB() is not supported in ROCKSDB_LITE\n");
-  return 0;
-}
-
-#endif  // !ROCKSDB_LITE
@@ -7,8 +7,8 @@
 // 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 "db/db_test_util.h"
 #include "port/stack_trace.h"
+#include "util/db_test_util.h"

 namespace rocksdb {

@@ -47,24 +47,6 @@ class DeleteFilter : public CompactionFilter {
  virtual const char* Name() const override { return "DeleteFilter"; }
 };

-class DeleteISFilter : public CompactionFilter {
- public:
-  virtual bool Filter(int level, const Slice& key, const Slice& value,
-                      std::string* new_value,
-                      bool* value_changed) const override {
-    cfilter_count++;
-    int i = std::stoi(key.ToString());
-    if (i > 5 && i <= 105) {
-      return true;
-    }
-    return false;
-  }
-
-  virtual bool IgnoreSnapshots() const override { return true; }
-
-  virtual const char* Name() const override { return "DeleteFilter"; }
-};
-
 class DelayFilter : public CompactionFilter {
 public:
  explicit DelayFilter(DBTestBase* d) : db_test(d) {}
@@ -115,11 +97,8 @@ class ChangeFilter : public CompactionFilter {

 class KeepFilterFactory : public CompactionFilterFactory {
 public:
-  explicit KeepFilterFactory(bool check_context = false,
-                             bool check_context_cf_id = false)
-      : check_context_(check_context),
-        check_context_cf_id_(check_context_cf_id),
-        compaction_filter_created_(false) {}
+  explicit KeepFilterFactory(bool check_context = false)
+      : check_context_(check_context) {}

  virtual std::unique_ptr<CompactionFilter> CreateCompactionFilter(
      const CompactionFilter::Context& context) override {
@@ -127,22 +106,13 @@ class KeepFilterFactory : public CompactionFilterFactory {
      EXPECT_EQ(expect_full_compaction_.load(), context.is_full_compaction);
      EXPECT_EQ(expect_manual_compaction_.load(), context.is_manual_compaction);
    }
-    if (check_context_cf_id_) {
-      EXPECT_EQ(expect_cf_id_.load(), context.column_family_id);
-    }
-    compaction_filter_created_ = true;
    return std::unique_ptr<CompactionFilter>(new KeepFilter());
  }

-  bool compaction_filter_created() const { return compaction_filter_created_; }
-
  virtual const char* Name() const override { return "KeepFilterFactory"; }
  bool check_context_;
-  bool check_context_cf_id_;
  std::atomic_bool expect_full_compaction_;
  std::atomic_bool expect_manual_compaction_;
-  std::atomic<uint32_t> expect_cf_id_;
-  bool compaction_filter_created_;
 };

 class DeleteFilterFactory : public CompactionFilterFactory {
@@ -159,21 +129,6 @@ class DeleteFilterFactory : public CompactionFilterFactory {
  virtual const char* Name() const override { return "DeleteFilterFactory"; }
 };

-// Delete Filter Factory which ignores snapshots
-class DeleteISFilterFactory : public CompactionFilterFactory {
- public:
-  virtual std::unique_ptr<CompactionFilter> CreateCompactionFilter(
-      const CompactionFilter::Context& context) override {
-    if (context.is_manual_compaction) {
-      return std::unique_ptr<CompactionFilter>(new DeleteISFilter());
-    } else {
-      return std::unique_ptr<CompactionFilter>(nullptr);
-    }
-  }
-
-  virtual const char* Name() const override { return "DeleteFilterFactory"; }
-};
-
 class DelayFilterFactory : public CompactionFilterFactory {
 public:
  explicit DelayFilterFactory(DBTestBase* d) : db_test(d) {}
@@ -219,7 +174,6 @@ class ChangeFilterFactory : public CompactionFilterFactory {
  virtual const char* Name() const override { return "ChangeFilterFactory"; }
 };

-#ifndef ROCKSDB_LITE
 TEST_F(DBTestCompactionFilter, CompactionFilter) {
  Options options = CurrentOptions();
  options.max_open_files = -1;
@@ -262,7 +216,7 @@ TEST_F(DBTestCompactionFilter, CompactionFilter) {
  Arena arena;
  {
    ScopedArenaIterator iter(
-        dbfull()->NewInternalIterator(&arena, handles_[1]));
+        dbfull()->TEST_NewInternalIterator(&arena, handles_[1]));
    iter->SeekToFirst();
    ASSERT_OK(iter->status());
    while (iter->Valid()) {
@@ -350,7 +304,7 @@ TEST_F(DBTestCompactionFilter, CompactionFilter) {
  count = 0;
  {
    ScopedArenaIterator iter(
-        dbfull()->NewInternalIterator(&arena, handles_[1]));
+        dbfull()->TEST_NewInternalIterator(&arena, handles_[1]));
    iter->SeekToFirst();
    ASSERT_OK(iter->status());
    while (iter->Valid()) {
@@ -396,7 +350,6 @@ TEST_F(DBTestCompactionFilter, CompactionFilterDeletesAll) {

  delete itr;
 }
-#endif  // ROCKSDB_LITE

 TEST_F(DBTestCompactionFilter, CompactionFilterWithValueChange) {
  do {
@@ -528,9 +481,8 @@ TEST_F(DBTestCompactionFilter, CompactionFilterWithMergeOperator) {
  ASSERT_EQ(newvalue, four);
 }

-#ifndef ROCKSDB_LITE
 TEST_F(DBTestCompactionFilter, CompactionFilterContextManual) {
-  KeepFilterFactory* filter = new KeepFilterFactory(true, true);
+  KeepFilterFactory* filter = new KeepFilterFactory();

  Options options = CurrentOptions();
  options.compaction_style = kCompactionStyleUniversal;
@@ -552,24 +504,22 @@ TEST_F(DBTestCompactionFilter, CompactionFilterContextManual) {
    // be triggered.
    num_keys_per_file /= 2;
  }
-  dbfull()->TEST_WaitForCompact();

  // Force a manual compaction
  cfilter_count = 0;
  filter->expect_manual_compaction_.store(true);
-  filter->expect_full_compaction_.store(true);
-  filter->expect_cf_id_.store(0);
+  filter->expect_full_compaction_.store(false);  // Manual compaction always
+                                                 // set this flag.
  dbfull()->CompactRange(CompactRangeOptions(), nullptr, nullptr);
  ASSERT_EQ(cfilter_count, 700);
  ASSERT_EQ(NumSortedRuns(0), 1);
-  ASSERT_TRUE(filter->compaction_filter_created());

  // Verify total number of keys is correct after manual compaction.
  {
    int count = 0;
    int total = 0;
    Arena arena;
-    ScopedArenaIterator iter(dbfull()->NewInternalIterator(&arena));
+    ScopedArenaIterator iter(dbfull()->TEST_NewInternalIterator(&arena));
    iter->SeekToFirst();
    ASSERT_OK(iter->status());
    while (iter->Valid()) {
@@ -583,41 +533,10 @@ TEST_F(DBTestCompactionFilter, CompactionFilterContextManual) {
      iter->Next();
    }
    ASSERT_EQ(total, 700);
-    ASSERT_EQ(count, 2);
+    ASSERT_EQ(count, 1);
  }
 }
-#endif  // ROCKSDB_LITE

-TEST_F(DBTestCompactionFilter, CompactionFilterContextCfId) {
-  KeepFilterFactory* filter = new KeepFilterFactory(false, true);
-  filter->expect_cf_id_.store(1);
-
-  Options options = CurrentOptions();
-  options.compaction_filter_factory.reset(filter);
-  options.compression = kNoCompression;
-  options.level0_file_num_compaction_trigger = 2;
-  CreateAndReopenWithCF({"pikachu"}, options);
-
-  int num_keys_per_file = 400;
-  for (int j = 0; j < 3; j++) {
-    // Write several keys.
-    const std::string value(10, 'x');
-    for (int i = 0; i < num_keys_per_file; i++) {
-      char key[100];
-      snprintf(key, sizeof(key), "B%08d%02d", i, j);
-      Put(1, key, value);
-    }
-    Flush(1);
-    // Make sure next file is much smaller so automatic compaction will not
-    // be triggered.
-    num_keys_per_file /= 2;
-  }
-  dbfull()->TEST_WaitForCompact();
-
-  ASSERT_TRUE(filter->compaction_filter_created());
-}
-
-#ifndef ROCKSDB_LITE
 // Compaction filters should only be applied to records that are newer than the
 // latest snapshot. This test inserts records and applies a delete filter.
 TEST_F(DBTestCompactionFilter, CompactionFilterSnapshot) {
@@ -654,73 +573,14 @@ TEST_F(DBTestCompactionFilter, CompactionFilterSnapshot) {
  ASSERT_EQ(0U, CountLiveFiles());
 }

-// Compaction filters should only be applied to records that are newer than the
-// latest snapshot. However, if the compaction filter asks to ignore snapshots
-// records newer than the snapshot will also be processed
-TEST_F(DBTestCompactionFilter, CompactionFilterIgnoreSnapshot) {
-  std::string five = ToString(5);
-  Options options;
-  options.compaction_filter_factory = std::make_shared<DeleteISFilterFactory>();
-  options.disable_auto_compactions = true;
-  options.create_if_missing = true;
-  options = CurrentOptions(options);
-  DestroyAndReopen(options);
-
-  // Put some data.
-  const Snapshot* snapshot = nullptr;
-  for (int table = 0; table < 4; ++table) {
-    for (int i = 0; i < 10; ++i) {
-      Put(ToString(table * 100 + i), "val");
-    }
-    Flush();
-
-    if (table == 0) {
-      snapshot = db_->GetSnapshot();
-    }
-  }
-  assert(snapshot != nullptr);
-
-  cfilter_count = 0;
-  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
-  // The filter should delete 40 records.
-  ASSERT_EQ(40U, cfilter_count);
-
-  {
-    // Scan the entire database as of the snapshot to ensure
-    // that nothing is left
-    ReadOptions read_options;
-    read_options.snapshot = snapshot;
-    std::unique_ptr<Iterator> iter(db_->NewIterator(read_options));
-    iter->SeekToFirst();
-    int count = 0;
-    while (iter->Valid()) {
-      count++;
-      iter->Next();
-    }
-    ASSERT_EQ(count, 6);
-    read_options.snapshot = 0;
-    std::unique_ptr<Iterator> iter1(db_->NewIterator(read_options));
-    iter1->SeekToFirst();
-    count = 0;
-    while (iter1->Valid()) {
-      count++;
-      iter1->Next();
-    }
-    // We have deleted 10 keys from 40 using the compaction filter
-    //  Keys 6-9 before the snapshot and 100-105 after the snapshot
-    ASSERT_EQ(count, 30);
-  }
-
-  // Release the snapshot and compact again -> now all records should be
-  // removed.
-  db_->ReleaseSnapshot(snapshot);
-}
-#endif  // ROCKSDB_LITE
-
 }  // namespace rocksdb

 int main(int argc, char** argv) {
+#if !(defined NDEBUG) || !defined(OS_WIN)
  rocksdb::port::InstallStackTraceHandler();
  ::testing::InitGoogleTest(&argc, argv);
  return RUN_ALL_TESTS();
+#else
+  return 0;
+#endif
 }
@@ -7,28 +7,26 @@
 // 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 "db/db_test_util.h"
 #include "port/stack_trace.h"
 #include "rocksdb/experimental.h"
-#include "rocksdb/utilities/convenience.h"
+#include "util/db_test_util.h"
 #include "util/sync_point.h"
 namespace rocksdb {

 // SYNC_POINT is not supported in released Windows mode.
-#if !defined(ROCKSDB_LITE)
+#if !(defined NDEBUG) || !defined(OS_WIN)
+

 class DBCompactionTest : public DBTestBase {
 public:
  DBCompactionTest() : DBTestBase("/db_compaction_test") {}
 };

-class DBCompactionTestWithParam
-    : public DBTestBase,
-      public testing::WithParamInterface<std::tuple<uint32_t, bool>> {
+class DBCompactionTestWithParam : public DBTestBase,
+                                public testing::WithParamInterface<uint32_t> {
 public:
  DBCompactionTestWithParam() : DBTestBase("/db_compaction_test") {
-    max_subcompactions_ = std::get<0>(GetParam());
-    exclusive_manual_compaction_ = std::get<1>(GetParam());
+    max_subcompactions_ = GetParam();
  }

  // Required if inheriting from testing::WithParamInterface<>
@@ -36,7 +34,6 @@ class DBCompactionTestWithParam
  static void TearDownTestCase() {}

  uint32_t max_subcompactions_;
-  bool exclusive_manual_compaction_;
 };

 namespace {
@@ -69,30 +66,49 @@ class OnFileDeletionListener : public EventListener {
  std::string expected_file_name_;
 };

-class FlushedFileCollector : public EventListener {
+class SleepingBackgroundTask {
 public:
-  FlushedFileCollector() {}
-  ~FlushedFileCollector() {}
-
-  virtual void OnFlushCompleted(DB* db, const FlushJobInfo& info) override {
-    std::lock_guard<std::mutex> lock(mutex_);
-    flushed_files_.push_back(info.file_path);
-  }
-
-  std::vector<std::string> GetFlushedFiles() {
-    std::lock_guard<std::mutex> lock(mutex_);
-    std::vector<std::string> result;
-    for (auto fname : flushed_files_) {
-      result.push_back(fname);
+  SleepingBackgroundTask()
+      : bg_cv_(&mutex_), should_sleep_(true), done_with_sleep_(false) {}
+  void DoSleep() {
+    MutexLock l(&mutex_);
+    while (should_sleep_) {
+      bg_cv_.Wait();
    }
-    return result;
+    done_with_sleep_ = true;
+    bg_cv_.SignalAll();
+  }
+  void WakeUp() {
+    MutexLock l(&mutex_);
+    should_sleep_ = false;
+    bg_cv_.SignalAll();
+  }
+  void WaitUntilDone() {
+    MutexLock l(&mutex_);
+    while (!done_with_sleep_) {
+      bg_cv_.Wait();
+    }
+  }
+  bool WokenUp() {
+    MutexLock l(&mutex_);
+    return should_sleep_ == false;
  }

-  void ClearFlushedFiles() { flushed_files_.clear(); }
+  void Reset() {
+    MutexLock l(&mutex_);
+    should_sleep_ = true;
+    done_with_sleep_ = false;
+  }
+
+  static void DoSleepTask(void* arg) {
+    reinterpret_cast<SleepingBackgroundTask*>(arg)->DoSleep();
+  }

 private:
-  std::vector<std::string> flushed_files_;
-  std::mutex mutex_;
+  port::Mutex mutex_;
+  port::CondVar bg_cv_;  // Signalled when background work finishes
+  bool should_sleep_;
+  bool done_with_sleep_;
 };

 static const int kCDTValueSize = 1000;
@@ -190,7 +206,7 @@ const SstFileMetaData* PickFileRandomly(
      auto result = rand->Uniform(file_id);
      return &(level_meta.files[result]);
    }
-    file_id -= static_cast<uint32_t>(level_meta.files.size());
+    file_id -= level_meta.files.size();
  }
  assert(false);
  return nullptr;
@@ -269,12 +285,9 @@ TEST_F(DBCompactionTest, SkipStatsUpdateTest) {
  env_->random_file_open_counter_.store(0);
  Reopen(options);

-  // As stats-update is disabled, we expect a very low number of
-  // random file open.
-  // Note that this number must be changed accordingly if we change
-  // the number of files needed to be opened in the DB::Open process.
-  const int kMaxFileOpenCount = 10;
-  ASSERT_LT(env_->random_file_open_counter_.load(), kMaxFileOpenCount);
+  // As stats-update is disabled, we expect a very low
+  // number of random file open.
+  ASSERT_LT(env_->random_file_open_counter_.load(), 5);

  // Repeat the reopen process, but this time we enable
  // stats-update.
@@ -284,7 +297,7 @@ TEST_F(DBCompactionTest, SkipStatsUpdateTest) {

  // Since we do a normal stats update on db-open, there
  // will be more random open files.
-  ASSERT_GT(env_->random_file_open_counter_.load(), kMaxFileOpenCount);
+  ASSERT_GT(env_->random_file_open_counter_.load(), 5);
 }

 TEST_F(DBCompactionTest, TestTableReaderForCompaction) {
@@ -491,15 +504,12 @@ TEST_F(DBCompactionTest, DisableStatsUpdateReopen) {


 TEST_P(DBCompactionTestWithParam, CompactionTrigger) {
-  const int kNumKeysPerFile = 100;
-
  Options options;
  options.write_buffer_size = 110 << 10;  // 110KB
  options.arena_block_size = 4 << 10;
  options.num_levels = 3;
  options.level0_file_num_compaction_trigger = 3;
  options.max_subcompactions = max_subcompactions_;
-  options.memtable_factory.reset(new SpecialSkipListFactory(kNumKeysPerFile));
  options = CurrentOptions(options);
  CreateAndReopenWithCF({"pikachu"}, options);

@@ -509,24 +519,20 @@ TEST_P(DBCompactionTestWithParam, CompactionTrigger) {
       num++) {
    std::vector<std::string> values;
    // Write 100KB (100 values, each 1K)
-    for (int i = 0; i < kNumKeysPerFile; i++) {
+    for (int i = 0; i < 100; i++) {
      values.push_back(RandomString(&rnd, 990));
      ASSERT_OK(Put(1, Key(i), values[i]));
    }
-    // put extra key to trigger flush
-    ASSERT_OK(Put(1, "", ""));
    dbfull()->TEST_WaitForFlushMemTable(handles_[1]);
    ASSERT_EQ(NumTableFilesAtLevel(0, 1), num + 1);
  }

  // generate one more file in level-0, and should trigger level-0 compaction
  std::vector<std::string> values;
-  for (int i = 0; i < kNumKeysPerFile; i++) {
+  for (int i = 0; i < 100; i++) {
    values.push_back(RandomString(&rnd, 990));
    ASSERT_OK(Put(1, Key(i), values[i]));
  }
-  // put extra key to trigger flush
-  ASSERT_OK(Put(1, "", ""));
  dbfull()->TEST_WaitForCompact();

  ASSERT_EQ(NumTableFilesAtLevel(0, 1), 0);
@@ -590,61 +596,6 @@ TEST_F(DBCompactionTest, MinorCompactionsHappen) {
  } while (ChangeCompactOptions());
 }

-TEST_F(DBCompactionTest, ZeroSeqIdCompaction) {
-  Options options;
-  options.compaction_style = kCompactionStyleLevel;
-  options.level0_file_num_compaction_trigger = 3;
-
-  FlushedFileCollector* collector = new FlushedFileCollector();
-  options.listeners.emplace_back(collector);
-
-  // compaction options
-  CompactionOptions compact_opt;
-  compact_opt.compression = kNoCompression;
-  compact_opt.output_file_size_limit = 4096;
-  const size_t key_len =
-    static_cast<size_t>(compact_opt.output_file_size_limit) / 5;
-
-  options = CurrentOptions(options);
-  DestroyAndReopen(options);
-
-  std::vector<const Snapshot*> snaps;
-
-  // create first file and flush to l0
-  for (auto& key : {"1", "2", "3", "3", "3", "3"}) {
-    Put(key, std::string(key_len, 'A'));
-    snaps.push_back(dbfull()->GetSnapshot());
-  }
-  Flush();
-  dbfull()->TEST_WaitForFlushMemTable();
-
-  // create second file and flush to l0
-  for (auto& key : {"3", "4", "5", "6", "7", "8"}) {
-    Put(key, std::string(key_len, 'A'));
-    snaps.push_back(dbfull()->GetSnapshot());
-  }
-  Flush();
-  dbfull()->TEST_WaitForFlushMemTable();
-
-  // move both files down to l1
-  dbfull()->CompactFiles(compact_opt, collector->GetFlushedFiles(), 1);
-
-  // release snap so that first instance of key(3) can have seqId=0
-  for (auto snap : snaps) {
-    dbfull()->ReleaseSnapshot(snap);
-  }
-
-  // create 3 files in l0 so to trigger compaction
-  for (int i = 0; i < options.level0_file_num_compaction_trigger; i++) {
-    Put("2", std::string(1, 'A'));
-    Flush();
-    dbfull()->TEST_WaitForFlushMemTable();
-  }
-
-  dbfull()->TEST_WaitForCompact();
-  ASSERT_OK(Put("", ""));
-}
-
 // Check that writes done during a memtable compaction are recovered
 // if the database is shutdown during the memtable compaction.
 TEST_F(DBCompactionTest, RecoverDuringMemtableCompaction) {
@@ -702,11 +653,8 @@ TEST_P(DBCompactionTestWithParam, TrivialMoveOneFile) {
  ASSERT_EQ(metadata.size(), 1U);
  LiveFileMetaData level0_file = metadata[0];  // L0 file meta

-  CompactRangeOptions cro;
-  cro.exclusive_manual_compaction = exclusive_manual_compaction_;
-
  // Compaction will initiate a trivial move from L0 to L1
-  dbfull()->CompactRange(cro, nullptr, nullptr);
+  dbfull()->CompactRange(CompactRangeOptions(), nullptr, nullptr);

  // File moved From L0 to L1
  ASSERT_EQ(NumTableFilesAtLevel(0, 0), 0);  // 0 files in L0
@@ -758,7 +706,7 @@ TEST_P(DBCompactionTestWithParam, TrivialMoveNonOverlappingFiles) {

  Random rnd(301);
  std::map<int32_t, std::string> values;
-  for (size_t i = 0; i < ranges.size(); i++) {
+  for (uint32_t i = 0; i < ranges.size(); i++) {
    for (int32_t j = ranges[i].first; j <= ranges[i].second; j++) {
      values[j] = RandomString(&rnd, value_size);
      ASSERT_OK(Put(Key(j), values[j]));
@@ -770,17 +718,14 @@ TEST_P(DBCompactionTestWithParam, TrivialMoveNonOverlappingFiles) {
  ASSERT_EQ(level0_files, ranges.size());    // Multiple files in L0
  ASSERT_EQ(NumTableFilesAtLevel(1, 0), 0);  // No files in L1

-  CompactRangeOptions cro;
-  cro.exclusive_manual_compaction = exclusive_manual_compaction_;
-
  // Since data is non-overlapping we expect compaction to initiate
  // a trivial move
-  db_->CompactRange(cro, nullptr, nullptr);
+  db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
  // We expect that all the files were trivially moved from L0 to L1
  ASSERT_EQ(NumTableFilesAtLevel(0, 0), 0);
  ASSERT_EQ(NumTableFilesAtLevel(1, 0) /* level1_files */, level0_files);

-  for (size_t i = 0; i < ranges.size(); i++) {
+  for (uint32_t i = 0; i < ranges.size(); i++) {
    for (int32_t j = ranges[i].first; j <= ranges[i].second; j++) {
      ASSERT_EQ(Get(Key(j)), values[j]);
    }
@@ -804,7 +749,7 @@ TEST_P(DBCompactionTestWithParam, TrivialMoveNonOverlappingFiles) {
    {500, 560},  // this range overlap with the next one
    {551, 599},
  };
-  for (size_t i = 0; i < ranges.size(); i++) {
+  for (uint32_t i = 0; i < ranges.size(); i++) {
    for (int32_t j = ranges[i].first; j <= ranges[i].second; j++) {
      values[j] = RandomString(&rnd, value_size);
      ASSERT_OK(Put(Key(j), values[j]));
@@ -812,9 +757,9 @@ TEST_P(DBCompactionTestWithParam, TrivialMoveNonOverlappingFiles) {
    ASSERT_OK(Flush());
  }

-  db_->CompactRange(cro, nullptr, nullptr);
+  db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);

-  for (size_t i = 0; i < ranges.size(); i++) {
+  for (uint32_t i = 0; i < ranges.size(); i++) {
    for (int32_t j = ranges[i].first; j <= ranges[i].second; j++) {
      ASSERT_EQ(Get(Key(j)), values[j]);
    }
@@ -868,7 +813,6 @@ TEST_P(DBCompactionTestWithParam, TrivialMoveTargetLevel) {
  CompactRangeOptions compact_options;
  compact_options.change_level = true;
  compact_options.target_level = 6;
-  compact_options.exclusive_manual_compaction = exclusive_manual_compaction_;
  ASSERT_OK(db_->CompactRange(compact_options, nullptr, nullptr));
  // 2 files in L6
  ASSERT_EQ("0,0,0,0,0,0,2", FilesPerLevel(0));
@@ -884,382 +828,6 @@ TEST_P(DBCompactionTestWithParam, TrivialMoveTargetLevel) {
  }
 }

-TEST_P(DBCompactionTestWithParam, ManualCompactionPartial) {
-  int32_t trivial_move = 0;
-  int32_t non_trivial_move = 0;
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
-      "DBImpl::BackgroundCompaction:TrivialMove",
-      [&](void* arg) { trivial_move++; });
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
-      "DBImpl::BackgroundCompaction:NonTrivial",
-      [&](void* arg) { non_trivial_move++; });
-  bool first = true;
-  bool second = true;
-  rocksdb::SyncPoint::GetInstance()->LoadDependency(
-      {{"DBCompaction::ManualPartial:4", "DBCompaction::ManualPartial:1"},
-       {"DBCompaction::ManualPartial:2", "DBCompaction::ManualPartial:3"}});
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
-      "DBImpl::BackgroundCompaction:NonTrivial:AfterRun", [&](void* arg) {
-        if (first) {
-          TEST_SYNC_POINT("DBCompaction::ManualPartial:4");
-          first = false;
-          TEST_SYNC_POINT("DBCompaction::ManualPartial:3");
-        } else if (second) {
-          TEST_SYNC_POINT("DBCompaction::ManualPartial:2");
-        }
-      });
-
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
-
-  Options options = CurrentOptions();
-  options.write_buffer_size = 10 * 1024 * 1024;
-  options.num_levels = 7;
-  options.max_subcompactions = max_subcompactions_;
-  options.level0_file_num_compaction_trigger = 3;
-  options.max_background_compactions = 3;
-  options.target_file_size_base = 1 << 23;  // 8 MB
-
-  DestroyAndReopen(options);
-  int32_t value_size = 10 * 1024;  // 10 KB
-
-  // Add 2 non-overlapping files
-  Random rnd(301);
-  std::map<int32_t, std::string> values;
-
-  // file 1 [0 => 100]
-  for (int32_t i = 0; i < 100; i++) {
-    values[i] = RandomString(&rnd, value_size);
-    ASSERT_OK(Put(Key(i), values[i]));
-  }
-  ASSERT_OK(Flush());
-
-  // file 2 [100 => 300]
-  for (int32_t i = 100; i < 300; i++) {
-    values[i] = RandomString(&rnd, value_size);
-    ASSERT_OK(Put(Key(i), values[i]));
-  }
-  ASSERT_OK(Flush());
-
-  // 2 files in L0
-  ASSERT_EQ("2", FilesPerLevel(0));
-  CompactRangeOptions compact_options;
-  compact_options.change_level = true;
-  compact_options.target_level = 6;
-  compact_options.exclusive_manual_compaction = exclusive_manual_compaction_;
-  // Trivial move the two non-overlapping files to level 6
-  ASSERT_OK(db_->CompactRange(compact_options, nullptr, nullptr));
-  // 2 files in L6
-  ASSERT_EQ("0,0,0,0,0,0,2", FilesPerLevel(0));
-
-  ASSERT_EQ(trivial_move, 1);
-  ASSERT_EQ(non_trivial_move, 0);
-
-  // file 3 [ 0 => 200]
-  for (int32_t i = 0; i < 200; i++) {
-    values[i] = RandomString(&rnd, value_size);
-    ASSERT_OK(Put(Key(i), values[i]));
-  }
-  ASSERT_OK(Flush());
-
-  // 1 files in L0
-  ASSERT_EQ("1,0,0,0,0,0,2", FilesPerLevel(0));
-  ASSERT_OK(dbfull()->TEST_CompactRange(0, nullptr, nullptr, nullptr, false));
-  ASSERT_OK(dbfull()->TEST_CompactRange(1, nullptr, nullptr, nullptr, false));
-  ASSERT_OK(dbfull()->TEST_CompactRange(2, nullptr, nullptr, nullptr, false));
-  ASSERT_OK(dbfull()->TEST_CompactRange(3, nullptr, nullptr, nullptr, false));
-  ASSERT_OK(dbfull()->TEST_CompactRange(4, nullptr, nullptr, nullptr, false));
-  // 2 files in L6, 1 file in L5
-  ASSERT_EQ("0,0,0,0,0,1,2", FilesPerLevel(0));
-
-  ASSERT_EQ(trivial_move, 6);
-  ASSERT_EQ(non_trivial_move, 0);
-
-  std::thread threads([&] {
-    compact_options.change_level = false;
-    compact_options.exclusive_manual_compaction = false;
-    std::string begin_string = Key(0);
-    std::string end_string = Key(199);
-    Slice begin(begin_string);
-    Slice end(end_string);
-    ASSERT_OK(db_->CompactRange(compact_options, &begin, &end));
-  });
-
-  TEST_SYNC_POINT("DBCompaction::ManualPartial:1");
-  // file 4 [300 => 400)
-  for (int32_t i = 300; i <= 400; i++) {
-    values[i] = RandomString(&rnd, value_size);
-    ASSERT_OK(Put(Key(i), values[i]));
-  }
-  ASSERT_OK(Flush());
-
-  // file 5 [400 => 500)
-  for (int32_t i = 400; i <= 500; i++) {
-    values[i] = RandomString(&rnd, value_size);
-    ASSERT_OK(Put(Key(i), values[i]));
-  }
-  ASSERT_OK(Flush());
-
-  // file 6 [500 => 600)
-  for (int32_t i = 500; i <= 600; i++) {
-    values[i] = RandomString(&rnd, value_size);
-    ASSERT_OK(Put(Key(i), values[i]));
-  }
-  ASSERT_OK(Flush());
-
-  // 3 files in L0
-  ASSERT_EQ("3,0,0,0,0,1,2", FilesPerLevel(0));
-  // 1 file in L6, 1 file in L1
-  dbfull()->TEST_WaitForFlushMemTable();
-  dbfull()->TEST_WaitForCompact();
-  ASSERT_EQ("0,1,0,0,0,0,1", FilesPerLevel(0));
-  threads.join();
-
-  for (int32_t i = 0; i < 600; i++) {
-    ASSERT_EQ(Get(Key(i)), values[i]);
-  }
-}
-
-TEST_F(DBCompactionTest, ManualPartialFill) {
-  int32_t trivial_move = 0;
-  int32_t non_trivial_move = 0;
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
-      "DBImpl::BackgroundCompaction:TrivialMove",
-      [&](void* arg) { trivial_move++; });
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
-      "DBImpl::BackgroundCompaction:NonTrivial",
-      [&](void* arg) { non_trivial_move++; });
-  bool first = true;
-  bool second = true;
-  rocksdb::SyncPoint::GetInstance()->LoadDependency(
-      {{"DBCompaction::PartialFill:4", "DBCompaction::PartialFill:1"},
-       {"DBCompaction::PartialFill:2", "DBCompaction::PartialFill:3"}});
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
-      "DBImpl::BackgroundCompaction:NonTrivial:AfterRun", [&](void* arg) {
-        if (first) {
-          TEST_SYNC_POINT("DBCompaction::PartialFill:4");
-          first = false;
-          TEST_SYNC_POINT("DBCompaction::PartialFill:3");
-        } else if (second) {
-        }
-      });
-
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
-
-  Options options = CurrentOptions();
-  options.write_buffer_size = 10 * 1024 * 1024;
-  options.max_bytes_for_level_multiplier = 2;
-  options.num_levels = 4;
-  options.level0_file_num_compaction_trigger = 3;
-  options.max_background_compactions = 3;
-
-  DestroyAndReopen(options);
-  int32_t value_size = 10 * 1024;  // 10 KB
-
-  // Add 2 non-overlapping files
-  Random rnd(301);
-  std::map<int32_t, std::string> values;
-
-  // file 1 [0 => 100]
-  for (int32_t i = 0; i < 100; i++) {
-    values[i] = RandomString(&rnd, value_size);
-    ASSERT_OK(Put(Key(i), values[i]));
-  }
-  ASSERT_OK(Flush());
-
-  // file 2 [100 => 300]
-  for (int32_t i = 100; i < 300; i++) {
-    values[i] = RandomString(&rnd, value_size);
-    ASSERT_OK(Put(Key(i), values[i]));
-  }
-  ASSERT_OK(Flush());
-
-  // 2 files in L0
-  ASSERT_EQ("2", FilesPerLevel(0));
-  CompactRangeOptions compact_options;
-  compact_options.change_level = true;
-  compact_options.target_level = 2;
-  ASSERT_OK(db_->CompactRange(compact_options, nullptr, nullptr));
-  // 2 files in L2
-  ASSERT_EQ("0,0,2", FilesPerLevel(0));
-
-  ASSERT_EQ(trivial_move, 1);
-  ASSERT_EQ(non_trivial_move, 0);
-
-  // file 3 [ 0 => 200]
-  for (int32_t i = 0; i < 200; i++) {
-    values[i] = RandomString(&rnd, value_size);
-    ASSERT_OK(Put(Key(i), values[i]));
-  }
-  ASSERT_OK(Flush());
-
-  // 2 files in L2, 1 in L0
-  ASSERT_EQ("1,0,2", FilesPerLevel(0));
-  ASSERT_OK(dbfull()->TEST_CompactRange(0, nullptr, nullptr, nullptr, false));
-  // 2 files in L2, 1 in L1
-  ASSERT_EQ("0,1,2", FilesPerLevel(0));
-
-  ASSERT_EQ(trivial_move, 2);
-  ASSERT_EQ(non_trivial_move, 0);
-
-  std::thread threads([&] {
-    compact_options.change_level = false;
-    compact_options.exclusive_manual_compaction = false;
-    std::string begin_string = Key(0);
-    std::string end_string = Key(199);
-    Slice begin(begin_string);
-    Slice end(end_string);
-    ASSERT_OK(db_->CompactRange(compact_options, &begin, &end));
-  });
-
-  TEST_SYNC_POINT("DBCompaction::PartialFill:1");
-  // Many files 4 [300 => 4300)
-  for (int32_t i = 0; i <= 5; i++) {
-    for (int32_t j = 300; j < 4300; j++) {
-      if (j == 2300) {
-        ASSERT_OK(Flush());
-        dbfull()->TEST_WaitForFlushMemTable();
-      }
-      values[j] = RandomString(&rnd, value_size);
-      ASSERT_OK(Put(Key(j), values[j]));
-    }
-  }
-
-  // Verify level sizes
-  uint64_t target_size = 4 * options.max_bytes_for_level_base;
-  for (int32_t i = 1; i < options.num_levels; i++) {
-    ASSERT_LE(SizeAtLevel(i), target_size);
-    target_size *= options.max_bytes_for_level_multiplier;
-  }
-
-  TEST_SYNC_POINT("DBCompaction::PartialFill:2");
-  dbfull()->TEST_WaitForFlushMemTable();
-  dbfull()->TEST_WaitForCompact();
-  threads.join();
-
-  for (int32_t i = 0; i < 4300; i++) {
-    ASSERT_EQ(Get(Key(i)), values[i]);
-  }
-}
-
-TEST_F(DBCompactionTest, DeleteFileRange) {
-  Options options = CurrentOptions();
-  options.write_buffer_size = 10 * 1024 * 1024;
-  options.max_bytes_for_level_multiplier = 2;
-  options.num_levels = 4;
-  options.level0_file_num_compaction_trigger = 3;
-  options.max_background_compactions = 3;
-
-  DestroyAndReopen(options);
-  int32_t value_size = 10 * 1024;  // 10 KB
-
-  // Add 2 non-overlapping files
-  Random rnd(301);
-  std::map<int32_t, std::string> values;
-
-  // file 1 [0 => 100]
-  for (int32_t i = 0; i < 100; i++) {
-    values[i] = RandomString(&rnd, value_size);
-    ASSERT_OK(Put(Key(i), values[i]));
-  }
-  ASSERT_OK(Flush());
-
-  // file 2 [100 => 300]
-  for (int32_t i = 100; i < 300; i++) {
-    values[i] = RandomString(&rnd, value_size);
-    ASSERT_OK(Put(Key(i), values[i]));
-  }
-  ASSERT_OK(Flush());
-
-  // 2 files in L0
-  ASSERT_EQ("2", FilesPerLevel(0));
-  CompactRangeOptions compact_options;
-  compact_options.change_level = true;
-  compact_options.target_level = 2;
-  ASSERT_OK(db_->CompactRange(compact_options, nullptr, nullptr));
-  // 2 files in L2
-  ASSERT_EQ("0,0,2", FilesPerLevel(0));
-
-  // file 3 [ 0 => 200]
-  for (int32_t i = 0; i < 200; i++) {
-    values[i] = RandomString(&rnd, value_size);
-    ASSERT_OK(Put(Key(i), values[i]));
-  }
-  ASSERT_OK(Flush());
-
-  // Many files 4 [300 => 4300)
-  for (int32_t i = 0; i <= 5; i++) {
-    for (int32_t j = 300; j < 4300; j++) {
-      if (j == 2300) {
-        ASSERT_OK(Flush());
-        dbfull()->TEST_WaitForFlushMemTable();
-      }
-      values[j] = RandomString(&rnd, value_size);
-      ASSERT_OK(Put(Key(j), values[j]));
-    }
-  }
-  ASSERT_OK(Flush());
-  dbfull()->TEST_WaitForFlushMemTable();
-  dbfull()->TEST_WaitForCompact();
-
-  // Verify level sizes
-  uint64_t target_size = 4 * options.max_bytes_for_level_base;
-  for (int32_t i = 1; i < options.num_levels; i++) {
-    ASSERT_LE(SizeAtLevel(i), target_size);
-    target_size *= options.max_bytes_for_level_multiplier;
-  }
-
-  size_t old_num_files = CountFiles();
-  std::string begin_string = Key(1000);
-  std::string end_string = Key(2000);
-  Slice begin(begin_string);
-  Slice end(end_string);
-  ASSERT_OK(DeleteFilesInRange(db_, db_->DefaultColumnFamily(), &begin, &end));
-
-  int32_t deleted_count = 0;
-  for (int32_t i = 0; i < 4300; i++) {
-    if (i < 1000 || i > 2000) {
-      ASSERT_EQ(Get(Key(i)), values[i]);
-    } else {
-      ReadOptions roptions;
-      std::string result;
-      Status s = db_->Get(roptions, Key(i), &result);
-      ASSERT_TRUE(s.IsNotFound() || s.ok());
-      if (s.IsNotFound()) {
-        deleted_count++;
-      }
-    }
-  }
-  ASSERT_GT(deleted_count, 0);
-  begin_string = Key(5000);
-  end_string = Key(6000);
-  Slice begin1(begin_string);
-  Slice end1(end_string);
-  // Try deleting files in range which contain no keys
-  ASSERT_OK(
-      DeleteFilesInRange(db_, db_->DefaultColumnFamily(), &begin1, &end1));
-
-  // Push data from level 0 to level 1 to force all data to be deleted
-  // Note that we don't delete level 0 files
-  compact_options.change_level = true;
-  compact_options.target_level = 1;
-  ASSERT_OK(dbfull()->TEST_CompactRange(0, nullptr, nullptr));
-
-  ASSERT_OK(
-      DeleteFilesInRange(db_, db_->DefaultColumnFamily(), nullptr, nullptr));
-
-  int32_t deleted_count2 = 0;
-  for (int32_t i = 0; i < 4300; i++) {
-    ReadOptions roptions;
-    std::string result;
-    Status s = db_->Get(roptions, Key(i), &result);
-    ASSERT_TRUE(s.IsNotFound());
-    deleted_count2++;
-  }
-  ASSERT_GT(deleted_count2, deleted_count);
-  size_t new_num_files = CountFiles();
-  ASSERT_GT(old_num_files, new_num_files);
-}
-
 TEST_P(DBCompactionTestWithParam, TrivialMoveToLastLevelWithFiles) {
  int32_t trivial_move = 0;
  int32_t non_trivial_move = 0;
@@ -1293,7 +861,6 @@ TEST_P(DBCompactionTestWithParam, TrivialMoveToLastLevelWithFiles) {
  CompactRangeOptions compact_options;
  compact_options.change_level = true;
  compact_options.target_level = 3;
-  compact_options.exclusive_manual_compaction = exclusive_manual_compaction_;
  ASSERT_OK(db_->CompactRange(compact_options, nullptr, nullptr));
  ASSERT_EQ("0,0,0,1", FilesPerLevel(0));
  ASSERT_EQ(trivial_move, 1);
@@ -1307,10 +874,8 @@ TEST_P(DBCompactionTestWithParam, TrivialMoveToLastLevelWithFiles) {
  ASSERT_OK(Flush());

  ASSERT_EQ("1,0,0,1", FilesPerLevel(0));
-  CompactRangeOptions cro;
-  cro.exclusive_manual_compaction = exclusive_manual_compaction_;
  // Compaction will do L0=>L1 L1=>L2 L2=>L3 (3 trivial moves)
-  ASSERT_OK(db_->CompactRange(cro, nullptr, nullptr));
+  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
  ASSERT_EQ("0,0,0,2", FilesPerLevel(0));
  ASSERT_EQ(trivial_move, 4);
  ASSERT_EQ(non_trivial_move, 0);
@@ -1327,8 +892,6 @@ TEST_P(DBCompactionTestWithParam, LevelCompactionThirdPath) {
  options.db_paths.emplace_back(dbname_, 500 * 1024);
  options.db_paths.emplace_back(dbname_ + "_2", 4 * 1024 * 1024);
  options.db_paths.emplace_back(dbname_ + "_3", 1024 * 1024 * 1024);
-  options.memtable_factory.reset(
-      new SpecialSkipListFactory(KNumKeysByGenerateNewFile - 1));
  options.compaction_style = kCompactionStyleLevel;
  options.write_buffer_size = 110 << 10;  // 110KB
  options.arena_block_size = 4 << 10;
@@ -1444,8 +1007,6 @@ TEST_P(DBCompactionTestWithParam, LevelCompactionPathUse) {
  options.db_paths.emplace_back(dbname_, 500 * 1024);
  options.db_paths.emplace_back(dbname_ + "_2", 4 * 1024 * 1024);
  options.db_paths.emplace_back(dbname_ + "_3", 1024 * 1024 * 1024);
-  options.memtable_factory.reset(
-      new SpecialSkipListFactory(KNumKeysByGenerateNewFile - 1));
  options.compaction_style = kCompactionStyleLevel;
  options.write_buffer_size = 110 << 10;  // 110KB
  options.arena_block_size = 4 << 10;
@@ -1614,7 +1175,6 @@ TEST_P(DBCompactionTestWithParam, ConvertCompactionStyle) {
  compact_options.target_level = 0;
  compact_options.bottommost_level_compaction =
      BottommostLevelCompaction::kForce;
-  compact_options.exclusive_manual_compaction = exclusive_manual_compaction_;
  dbfull()->CompactRange(compact_options, handles_[1], nullptr, nullptr);

  // Only 1 file in L0
@@ -1713,7 +1273,6 @@ TEST_F(DBCompactionTest, L0_CompactionBug_Issue44_b) {
 TEST_P(DBCompactionTestWithParam, ManualCompaction) {
  Options options = CurrentOptions();
  options.max_subcompactions = max_subcompactions_;
-  options.statistics = rocksdb::CreateDBStatistics();
  CreateAndReopenWithCF({"pikachu"}, options);

  // iter - 0 with 7 levels
@@ -1745,16 +1304,7 @@ TEST_P(DBCompactionTestWithParam, ManualCompaction) {
    // Compact all
    MakeTables(1, "a", "z", 1);
    ASSERT_EQ("1,0,2", FilesPerLevel(1));
-
-    uint64_t prev_block_cache_add =
-        options.statistics->getTickerCount(BLOCK_CACHE_ADD);
-    CompactRangeOptions cro;
-    cro.exclusive_manual_compaction = exclusive_manual_compaction_;
-    db_->CompactRange(cro, handles_[1], nullptr, nullptr);
-    // Verify manual compaction doesn't fill block cache
-    ASSERT_EQ(prev_block_cache_add,
-              options.statistics->getTickerCount(BLOCK_CACHE_ADD));
-
+    db_->CompactRange(CompactRangeOptions(), handles_[1], nullptr, nullptr);
    ASSERT_EQ("0,0,1", FilesPerLevel(1));

    if (iter == 0) {
@@ -1762,7 +1312,6 @@ TEST_P(DBCompactionTestWithParam, ManualCompaction) {
      options.max_background_flushes = 0;
      options.num_levels = 3;
      options.create_if_missing = true;
-      options.statistics = rocksdb::CreateDBStatistics();
      DestroyAndReopen(options);
      CreateAndReopenWithCF({"pikachu"}, options);
    }
@@ -1829,7 +1378,6 @@ TEST_P(DBCompactionTestWithParam, ManualLevelCompactionOutputPathId) {
    ASSERT_EQ(1, GetSstFileCount(options.db_paths[0].path));
    CompactRangeOptions compact_options;
    compact_options.target_path_id = 1;
-    compact_options.exclusive_manual_compaction = exclusive_manual_compaction_;
    db_->CompactRange(compact_options, handles_[1], nullptr, nullptr);

    ASSERT_EQ("0,1", FilesPerLevel(1));
@@ -1945,8 +1493,8 @@ TEST_P(DBCompactionTestWithParam, PartialCompactionFailure) {
  env_->SetBackgroundThreads(1, Env::HIGH);
  env_->SetBackgroundThreads(1, Env::LOW);
  // stop the compaction thread until we simulate the file creation failure.
-  test::SleepingBackgroundTask sleeping_task_low;
-  env_->Schedule(&test::SleepingBackgroundTask::DoSleepTask, &sleeping_task_low,
+  SleepingBackgroundTask sleeping_task_low;
+  env_->Schedule(&SleepingBackgroundTask::DoSleepTask, &sleeping_task_low,
                 Env::Priority::LOW);

  options.env = env_;
@@ -2038,8 +1586,8 @@ TEST_P(DBCompactionTestWithParam, DeleteMovedFileAfterCompaction) {
    ASSERT_EQ("0,1", FilesPerLevel(0));

    // block compactions
-    test::SleepingBackgroundTask sleeping_task;
-    env_->Schedule(&test::SleepingBackgroundTask::DoSleepTask, &sleeping_task,
+    SleepingBackgroundTask sleeping_task;
+    env_->Schedule(&SleepingBackgroundTask::DoSleepTask, &sleeping_task,
                   Env::Priority::LOW);

    options.max_bytes_for_level_base = 1024 * 1024;  // 1 MB
@@ -2090,8 +1638,6 @@ TEST_P(DBCompactionTestWithParam, CompressLevelCompaction) {
    return;
  }
  Options options = CurrentOptions();
-  options.memtable_factory.reset(
-      new SpecialSkipListFactory(KNumKeysByGenerateNewFile - 1));
  options.compaction_style = kCompactionStyleLevel;
  options.write_buffer_size = 110 << 10;  // 110KB
  options.arena_block_size = 4 << 10;
@@ -2342,62 +1888,12 @@ TEST_P(DBCompactionTestWithParam, ForceBottommostLevelCompaction) {
 }

 INSTANTIATE_TEST_CASE_P(DBCompactionTestWithParam, DBCompactionTestWithParam,
-                        ::testing::Values(std::make_tuple(1, true),
-                                          std::make_tuple(1, false),
-                                          std::make_tuple(4, true),
-                                          std::make_tuple(4, false)));
-
-class CompactionPriTest : public DBTestBase,
-                          public testing::WithParamInterface<uint32_t> {
- public:
-  CompactionPriTest() : DBTestBase("/compaction_pri_test") {
-    compaction_pri_ = GetParam();
-  }
-
-  // Required if inheriting from testing::WithParamInterface<>
-  static void SetUpTestCase() {}
-  static void TearDownTestCase() {}
-
-  uint32_t compaction_pri_;
-};
-
-TEST_P(CompactionPriTest, Test) {
-  Options options;
-  options.write_buffer_size = 16 * 1024;
-  options.compaction_pri = static_cast<CompactionPri>(compaction_pri_);
-  options.hard_pending_compaction_bytes_limit = 256 * 1024;
-  options.max_bytes_for_level_base = 64 * 1024;
-  options.max_bytes_for_level_multiplier = 4;
-  options.compression = kNoCompression;
-  options = CurrentOptions(options);
-  DestroyAndReopen(options);
-
-  Random rnd(301);
-  const int kNKeys = 5000;
-  int keys[kNKeys];
-  for (int i = 0; i < kNKeys; i++) {
-    keys[i] = i;
-  }
-  std::random_shuffle(std::begin(keys), std::end(keys));
-
-  for (int i = 0; i < kNKeys; i++) {
-    ASSERT_OK(Put(Key(keys[i]), RandomString(&rnd, 102)));
-  }
-
-  dbfull()->TEST_WaitForCompact();
-  for (int i = 0; i < kNKeys; i++) {
-    ASSERT_NE("NOT_FOUND", Get(Key(i)));
-  }
-}
-
-INSTANTIATE_TEST_CASE_P(CompactionPriTest, CompactionPriTest,
-                        ::testing::Values(0, 1, 2));
-
-#endif // !defined(ROCKSDB_LITE)
+                        ::testing::Values(1, 4));
+#endif  // !(defined NDEBUG) || !defined(OS_WIN)
 }  // namespace rocksdb

 int main(int argc, char** argv) {
-#if !defined(ROCKSDB_LITE)
+#if !(defined NDEBUG) || !defined(OS_WIN)
  rocksdb::port::InstallStackTraceHandler();
  ::testing::InitGoogleTest(&argc, argv);
  return RUN_ALL_TESTS();
@@ -10,10 +10,10 @@
 // Introduction of SyncPoint effectively disabled building and running this test
 // in Release build.
 // which is a pity, it is a good test
-#if !defined(ROCKSDB_LITE)
+#if !(defined NDEBUG) || !defined(OS_WIN)

-#include "db/db_test_util.h"
 #include "port/stack_trace.h"
+#include "util/db_test_util.h"

 namespace rocksdb {
 class DBTestDynamicLevel : public DBTestBase {
@@ -484,10 +484,10 @@ TEST_F(DBTestDynamicLevel, MigrateToDynamicLevelMaxBytesBase) {
 }
 }  // namespace rocksdb

-#endif  // !defined(ROCKSDB_LITE)
+#endif  // !(defined NDEBUG) || !defined(OS_WIN)

 int main(int argc, char** argv) {
-#if !defined(ROCKSDB_LITE)
+#if !(defined NDEBUG) || !defined(OS_WIN)
  rocksdb::port::InstallStackTraceHandler();
  ::testing::InitGoogleTest(&argc, argv);
  return RUN_ALL_TESTS();
@@ -12,6 +12,7 @@
 #include <deque>
 #include <limits>
 #include <list>
+#include <list>
 #include <set>
 #include <string>
 #include <utility>
@@ -36,11 +37,12 @@
 #include "rocksdb/env.h"
 #include "rocksdb/memtablerep.h"
 #include "rocksdb/transaction_log.h"
-#include "table/scoped_arena_iterator.h"
 #include "util/autovector.h"
 #include "util/event_logger.h"
 #include "util/hash.h"
+#include "util/hash.h"
 #include "util/instrumented_mutex.h"
+#include "util/scoped_arena_iterator.h"
 #include "util/stop_watch.h"
 #include "util/thread_local.h"

@@ -51,10 +53,10 @@ class TableCache;
 class Version;
 class VersionEdit;
 class VersionSet;
+class CompactionFilterV2;
 class Arena;
 class WriteCallback;
 struct JobContext;
-struct ExternalSstFileInfo;

 class DBImpl : public DB {
 public:
@@ -74,10 +76,6 @@ class DBImpl : public DB {
  virtual Status Delete(const WriteOptions& options,
                        ColumnFamilyHandle* column_family,
                        const Slice& key) override;
-  using DB::SingleDelete;
-  virtual Status SingleDelete(const WriteOptions& options,
-                              ColumnFamilyHandle* column_family,
-                              const Slice& key) override;
  using DB::Write;
  virtual Status Write(const WriteOptions& options,
                       WriteBatch* updates) override;
@@ -122,9 +120,6 @@ class DBImpl : public DB {
  using DB::GetIntProperty;
  virtual bool GetIntProperty(ColumnFamilyHandle* column_family,
                              const Slice& property, uint64_t* value) override;
-  using DB::GetAggregatedIntProperty;
-  virtual bool GetAggregatedIntProperty(const Slice& property,
-                                        uint64_t* aggregated_value) override;
  using DB::GetApproximateSizes;
  virtual void GetApproximateSizes(ColumnFamilyHandle* column_family,
                                   const Range* range, int n, uint64_t* sizes,
@@ -141,12 +136,6 @@ class DBImpl : public DB {
                              const int output_level,
                              const int output_path_id = -1) override;

-  virtual Status PauseBackgroundWork() override;
-  virtual Status ContinueBackgroundWork() override;
-
-  virtual Status EnableAutoCompaction(
-      const std::vector<ColumnFamilyHandle*>& column_family_handles) override;
-
  using DB::SetOptions;
  Status SetOptions(
      ColumnFamilyHandle* column_family,
@@ -188,8 +177,6 @@ class DBImpl : public DB {
      const TransactionLogIterator::ReadOptions&
          read_options = TransactionLogIterator::ReadOptions()) override;
  virtual Status DeleteFile(std::string name) override;
-  Status DeleteFilesInRange(ColumnFamilyHandle* column_family,
-                            const Slice* begin, const Slice* end);

  virtual void GetLiveFilesMetaData(
      std::vector<LiveFileMetaData>* metadata) override;
@@ -228,46 +215,16 @@ class DBImpl : public DB {
                                                   bool include_history);

  // For a given key, check to see if there are any records for this key
-  // in the memtables, including memtable history.  If cache_only is false,
-  // SST files will also be checked.
-  //
-  // If a key is found, *found_record_for_key will be set to true and
-  // *seq will will be set to the stored sequence number for the latest
-  // operation on this key or kMaxSequenceNumber if unknown.
-  // If no key is found, *found_record_for_key will be set to false.
-  //
-  // Note: If cache_only=false, it is possible for *seq to be set to 0 if
-  // the sequence number has been cleared from the record.  If the caller is
-  // holding an active db snapshot, we know the missing sequence must be less
-  // than the snapshot's sequence number (sequence numbers are only cleared
-  // when there are no earlier active snapshots).
-  //
-  // If NotFound is returned and found_record_for_key is set to false, then no
-  // record for this key was found.  If the caller is holding an active db
-  // snapshot, we know that no key could have existing after this snapshot
-  // (since we do not compact keys that have an earlier snapshot).
-  //
-  // Returns OK or NotFound on success,
-  // other status on unexpected error.
-  Status GetLatestSequenceForKey(SuperVersion* sv, const Slice& key,
-                                 bool cache_only, SequenceNumber* seq,
-                                 bool* found_record_for_key);
+  // in the memtables, including memtable history.

-  using DB::AddFile;
-  virtual Status AddFile(ColumnFamilyHandle* column_family,
-                         const ExternalSstFileInfo* file_info,
-                         bool move_file) override;
-  virtual Status AddFile(ColumnFamilyHandle* column_family,
-                         const std::string& file_path, bool move_file) override;
+  // On success, *seq will contain the sequence number for the
+  // latest such change or kMaxSequenceNumber if no records were present.
+  // Returns OK on success, other status on error reading memtables.
+  Status GetLatestSequenceForKeyFromMemtable(SuperVersion* sv, const Slice& key,
+                                             SequenceNumber* seq);

 #endif  // ROCKSDB_LITE

-  // Similar to GetSnapshot(), but also lets the db know that this snapshot
-  // will be used for transaction write-conflict checking.  The DB can then
-  // make sure not to compact any keys that would prevent a write-conflict from
-  // being detected.
-  const Snapshot* GetSnapshotForWriteConflictBoundary();
-
  // checks if all live files exist on file system and that their file sizes
  // match to our in-memory records
  virtual Status CheckConsistency();
@@ -277,16 +234,9 @@ class DBImpl : public DB {
  Status RunManualCompaction(ColumnFamilyData* cfd, int input_level,
                             int output_level, uint32_t output_path_id,
                             const Slice* begin, const Slice* end,
-                             bool exclusive,
                             bool disallow_trivial_move = false);

-  // Return an internal iterator over the current state of the database.
-  // The keys of this iterator are internal keys (see format.h).
-  // The returned iterator should be deleted when no longer needed.
-  InternalIterator* NewInternalIterator(
-      Arena* arena, ColumnFamilyHandle* column_family = nullptr);
-
-#ifndef NDEBUG
+#ifndef ROCKSDB_LITE
  // Extra methods (for testing) that are not in the public DB interface
  // Implemented in db_impl_debug.cc

@@ -304,6 +254,12 @@ class DBImpl : public DB {
  // Wait for any compaction
  Status TEST_WaitForCompact();

+  // Return an internal iterator over the current state of the database.
+  // The keys of this iterator are internal keys (see format.h).
+  // The returned iterator should be deleted when no longer needed.
+  Iterator* TEST_NewInternalIterator(
+      Arena* arena, ColumnFamilyHandle* column_family = nullptr);
+
  // Return the maximum overlapping data (in bytes) at next level for any
  // file at a level >= 1.
  int64_t TEST_MaxNextLevelOverlappingBytes(ColumnFamilyHandle* column_family =
@@ -337,15 +293,7 @@ class DBImpl : public DB {

  uint64_t TEST_LogfileNumber();

-  // Returns column family name to ImmutableCFOptions map.
-  Status TEST_GetAllImmutableCFOptions(
-      std::unordered_map<std::string, const ImmutableCFOptions*>* iopts_map);
-
-  Cache* TEST_table_cache() { return table_cache_.get(); }
-
-  WriteController& TEST_write_controler() { return write_controller_; }
-
-#endif  // NDEBUG
+#endif  // ROCKSDB_LITE

  // Returns the list of live files in 'live' and the list
  // of all files in the filesystem in 'candidate_files'.
@@ -403,20 +351,6 @@ class DBImpl : public DB {
  // Same as above, should called without mutex held and not on write thread.
  ColumnFamilyHandle* GetColumnFamilyHandleUnlocked(uint32_t column_family_id);

-  // Returns the number of currently running flushes.
-  // REQUIREMENT: mutex_ must be held when calling this function.
-  int num_running_flushes() {
-    mutex_.AssertHeld();
-    return num_running_flushes_;
-  }
-
-  // Returns the number of currently running compactions.
-  // REQUIREMENT: mutex_ must be held when calling this function.
-  int num_running_compactions() {
-    mutex_.AssertHeld();
-    return num_running_compactions_;
-  }
-
 protected:
  Env* const env_;
  const std::string dbname_;
@@ -424,25 +358,12 @@ class DBImpl : public DB {
  const DBOptions db_options_;
  Statistics* stats_;

-  InternalIterator* NewInternalIterator(const ReadOptions&,
-                                        ColumnFamilyData* cfd,
-                                        SuperVersion* super_version,
-                                        Arena* arena);
-
-  // Except in DB::Open(), WriteOptionsFile can only be called when:
-  // 1. WriteThread::Writer::EnterUnbatched() is used.
-  // 2. db_mutex is held
-  Status WriteOptionsFile();
-
-  // The following two functions can only be called when:
-  // 1. WriteThread::Writer::EnterUnbatched() is used.
-  // 2. db_mutex is NOT held
-  Status RenameTempFileToOptionsFile(const std::string& file_name);
-  Status DeleteObsoleteOptionsFiles();
+  Iterator* NewInternalIterator(const ReadOptions&, ColumnFamilyData* cfd,
+                                SuperVersion* super_version, Arena* arena);

  void NotifyOnFlushCompleted(ColumnFamilyData* cfd, FileMetaData* file_meta,
                              const MutableCFOptions& mutable_cf_options,
-                              int job_id, TableProperties prop);
+                              int job_id);

  void NotifyOnCompactionCompleted(ColumnFamilyData* cfd,
                                   Compaction *c, const Status &st,
@@ -555,13 +476,12 @@ class DBImpl : public DB {
  void MaybeScheduleFlushOrCompaction();
  void SchedulePendingFlush(ColumnFamilyData* cfd);
  void SchedulePendingCompaction(ColumnFamilyData* cfd);
-  static void BGWorkCompaction(void* arg);
+  static void BGWorkCompaction(void* db);
  static void BGWorkFlush(void* db);
-  static void UnscheduleCallback(void* arg);
-  void BackgroundCallCompaction(void* arg);
+  void BackgroundCallCompaction();
  void BackgroundCallFlush();
  Status BackgroundCompaction(bool* madeProgress, JobContext* job_context,
-                              LogBuffer* log_buffer, void* m = 0);
+                              LogBuffer* log_buffer);
  Status BackgroundFlush(bool* madeProgress, JobContext* job_context,
                         LogBuffer* log_buffer);

@@ -589,25 +509,18 @@ class DBImpl : public DB {
  // helper function to call after some of the logs_ were synced
  void MarkLogsSynced(uint64_t up_to, bool synced_dir, const Status& status);

-  const Snapshot* GetSnapshotImpl(bool is_write_conflict_boundary);
-
  // table_cache_ provides its own synchronization
  std::shared_ptr<Cache> table_cache_;

  // Lock over the persistent DB state.  Non-nullptr iff successfully acquired.
  FileLock* db_lock_;

-  // The mutex for options file related operations.
-  // NOTE: should never acquire options_file_mutex_ and mutex_ at the
-  //       same time.
-  InstrumentedMutex options_files_mutex_;
  // State below is protected by mutex_
  InstrumentedMutex mutex_;
-
  std::atomic<bool> shutting_down_;
  // This condition variable is signaled on these conditions:
  // * whenever bg_compaction_scheduled_ goes down to 0
-  // * if AnyManualCompaction, whenever a compaction finishes, even if it hasn't
+  // * if bg_manual_only_ > 0, whenever a compaction finishes, even if it hasn't
  // made any progress
  // * whenever a compaction made any progress
  // * whenever bg_flush_scheduled_ value decreases (i.e. whenever a flush is
@@ -615,8 +528,6 @@ class DBImpl : public DB {
  // * whenever there is an error in background flush or compaction
  InstrumentedCondVar bg_cv_;
  uint64_t logfile_number_;
-  std::deque<uint64_t>
-      log_recycle_files;  // a list of log files that we can recycle
  bool log_dir_synced_;
  bool log_empty_;
  ColumnFamilyHandleImpl* default_cf_handle_;
@@ -762,40 +673,29 @@ class DBImpl : public DB {
  // count how many background compactions are running or have been scheduled
  int bg_compaction_scheduled_;

-  // stores the number of compactions are currently running
-  int num_running_compactions_;
+  // If non-zero, MaybeScheduleFlushOrCompaction() will only schedule manual
+  // compactions (if manual_compaction_ is not null). This mechanism enables
+  // manual compactions to wait until all other compactions are finished.
+  int bg_manual_only_;

  // number of background memtable flush jobs, submitted to the HIGH pool
  int bg_flush_scheduled_;

-  // stores the number of flushes are currently running
-  int num_running_flushes_;
-
  // Information for a manual compaction
  struct ManualCompaction {
    ColumnFamilyData* cfd;
    int input_level;
    int output_level;
    uint32_t output_path_id;
-    Status status;
    bool done;
+    Status status;
    bool in_progress;             // compaction request being processed?
-    bool incomplete;              // only part of requested range compacted
-    bool exclusive;               // current behavior of only one manual
-    bool disallow_trivial_move;   // Force actual compaction to run
    const InternalKey* begin;     // nullptr means beginning of key range
    const InternalKey* end;       // nullptr means end of key range
-    InternalKey* manual_end;      // how far we are compacting
    InternalKey tmp_storage;      // Used to keep track of compaction progress
-    InternalKey tmp_storage1;     // Used to keep track of compaction progress
-    Compaction* compaction;
-  };
-  std::deque<ManualCompaction*> manual_compaction_dequeue_;
-
-  struct CompactionArg {
-    DBImpl* db;
-    ManualCompaction* m;
+    bool disallow_trivial_move;   // Force actual compaction to run
  };
+  ManualCompaction* manual_compaction_;

  // Have we encountered a background error in paranoid mode?
  Status bg_error_;
@@ -836,8 +736,8 @@ class DBImpl : public DB {
  // Unified interface for logging events
  EventLogger event_logger_;

-  // A value of >0 temporarily disables scheduling of background work
-  int bg_work_paused_;
+  // A value of true temporarily disables scheduling of background work
+  bool bg_work_gate_closed_;

  // Guard against multiple concurrent refitting
  bool refitting_level_;
@@ -876,10 +776,6 @@ class DBImpl : public DB {
  virtual Status GetPropertiesOfAllTables(ColumnFamilyHandle* column_family,
                                          TablePropertiesCollection* props)
      override;
-  virtual Status GetPropertiesOfTablesInRange(
-      ColumnFamilyHandle* column_family, const Range* range, std::size_t n,
-      TablePropertiesCollection* props) override;
-
 #endif  // ROCKSDB_LITE

  // Function that Get and KeyMayExist call with no_io true or false
@@ -888,18 +784,9 @@ class DBImpl : public DB {
                 const Slice& key, std::string* value,
                 bool* value_found = nullptr);

-  bool GetIntPropertyInternal(ColumnFamilyData* cfd,
+  bool GetIntPropertyInternal(ColumnFamilyHandle* column_family,
                              DBPropertyType property_type,
-                              bool need_out_of_mutex, bool is_locked,
-                              uint64_t* value);
-
-  bool HasPendingManualCompaction();
-  bool HasExclusiveManualCompaction();
-  void AddManualCompaction(ManualCompaction* m);
-  void RemoveManualCompaction(ManualCompaction* m);
-  bool ShouldntRunManualCompaction(ManualCompaction* m);
-  bool HaveManualCompaction(ColumnFamilyData* cfd);
-  bool MCOverlap(ManualCompaction* m, ManualCompaction* m1);
+                              bool need_out_of_mutex, uint64_t* value);
 };

 // Sanitize db options.  The caller should delete result.info_log if
@@ -7,7 +7,7 @@
 // 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.

-#ifndef NDEBUG
+#ifndef ROCKSDB_LITE

 #include "db/db_impl.h"
 #include "util/thread_status_updater.h"
@@ -19,6 +19,23 @@ uint64_t DBImpl::TEST_GetLevel0TotalSize() {
  return default_cf_handle_->cfd()->current()->storage_info()->NumLevelBytes(0);
 }

+Iterator* DBImpl::TEST_NewInternalIterator(Arena* arena,
+                                           ColumnFamilyHandle* column_family) {
+  ColumnFamilyData* cfd;
+  if (column_family == nullptr) {
+    cfd = default_cf_handle_->cfd();
+  } else {
+    auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
+    cfd = cfh->cfd();
+  }
+
+  mutex_.Lock();
+  SuperVersion* super_version = cfd->GetSuperVersion()->Ref();
+  mutex_.Unlock();
+  ReadOptions roptions;
+  return NewInternalIterator(roptions, cfd, super_version, arena);
+}
+
 int64_t DBImpl::TEST_MaxNextLevelOverlappingBytes(
    ColumnFamilyHandle* column_family) {
  ColumnFamilyData* cfd;
@@ -70,7 +87,7 @@ Status DBImpl::TEST_CompactRange(int level, const Slice* begin,
       cfd->ioptions()->compaction_style == kCompactionStyleFIFO)
          ? level
          : level + 1;
-  return RunManualCompaction(cfd, level, output_level, 0, begin, end, true,
+  return RunManualCompaction(cfd, level, output_level, 0, begin, end,
                             disallow_trivial_move);
 }

@@ -135,24 +152,5 @@ uint64_t DBImpl::TEST_LogfileNumber() {
  return logfile_number_;
 }

-Status DBImpl::TEST_GetAllImmutableCFOptions(
-    std::unordered_map<std::string, const ImmutableCFOptions*>* iopts_map) {
-  std::vector<std::string> cf_names;
-  std::vector<const ImmutableCFOptions*> iopts;
-  {
-    InstrumentedMutexLock l(&mutex_);
-    for (auto cfd : *versions_->GetColumnFamilySet()) {
-      cf_names.push_back(cfd->GetName());
-      iopts.push_back(cfd->ioptions());
-    }
-  }
-  iopts_map->clear();
-  for (size_t i = 0; i < cf_names.size(); ++i) {
-    iopts_map->insert({cf_names[i], iopts[i]});
-  }
-
-  return Status::OK();
-}
-
 }  // namespace rocksdb
-#endif  // NDEBUG
+#endif  // ROCKSDB_LITE
@@ -53,12 +53,6 @@ class DBImplReadOnly : public DBImpl {
                        const Slice& key) override {
    return Status::NotSupported("Not supported operation in read only mode.");
  }
-  using DBImpl::SingleDelete;
-  virtual Status SingleDelete(const WriteOptions& options,
-                              ColumnFamilyHandle* column_family,
-                              const Slice& key) override {
-    return Status::NotSupported("Not supported operation in read only mode.");
-  }
  virtual Status Write(const WriteOptions& options,
                       WriteBatch* updates) override {
    return Status::NotSupported("Not supported operation in read only mode.");
@@ -100,11 +94,6 @@ class DBImplReadOnly : public DBImpl {
    return Status::NotSupported("Not supported operation in read only mode.");
  }

-  using DBImpl::SyncWAL;
-  virtual Status SyncWAL() override {
-    return Status::NotSupported("Not supported operation in read only mode.");
-  }
-
 private:
  friend class DB;

@@ -6,8 +6,8 @@
 // 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 "db/db_test_util.h"
 #include "port/stack_trace.h"
+#include "util/db_test_util.h"

 namespace rocksdb {

@@ -15,12 +15,11 @@

 #include "db/filename.h"
 #include "db/dbformat.h"
-#include "port/port.h"
 #include "rocksdb/env.h"
 #include "rocksdb/options.h"
 #include "rocksdb/iterator.h"
 #include "rocksdb/merge_operator.h"
-#include "table/internal_iterator.h"
+#include "port/port.h"
 #include "util/arena.h"
 #include "util/logging.h"
 #include "util/mutexlock.h"
@@ -59,11 +58,10 @@ class DBIter: public Iterator {
    kReverse
  };

-  DBIter(Env* env, const ImmutableCFOptions& ioptions, const Comparator* cmp,
-         InternalIterator* iter, SequenceNumber s, bool arena_mode,
-         uint64_t max_sequential_skip_in_iterations,
-         const Slice* iterate_upper_bound = nullptr,
-         bool prefix_same_as_start = false)
+  DBIter(Env* env, const ImmutableCFOptions& ioptions,
+         const Comparator* cmp, Iterator* iter, SequenceNumber s,
+         bool arena_mode, uint64_t max_sequential_skip_in_iterations,
+         const Slice* iterate_upper_bound = nullptr)
      : arena_mode_(arena_mode),
        env_(env),
        logger_(ioptions.info_log),
@@ -75,9 +73,7 @@ class DBIter: public Iterator {
        valid_(false),
        current_entry_is_merged_(false),
        statistics_(ioptions.statistics),
-        iterate_upper_bound_(iterate_upper_bound),
-        prefix_same_as_start_(prefix_same_as_start),
-        iter_pinned_(false) {
+        iterate_upper_bound_(iterate_upper_bound) {
    RecordTick(statistics_, NO_ITERATORS);
    prefix_extractor_ = ioptions.prefix_extractor;
    max_skip_ = max_sequential_skip_in_iterations;
@@ -87,15 +83,12 @@ class DBIter: public Iterator {
    if (!arena_mode_) {
      delete iter_;
    } else {
-      iter_->~InternalIterator();
+      iter_->~Iterator();
    }
  }
-  virtual void SetIter(InternalIterator* iter) {
+  virtual void SetIter(Iterator* iter) {
    assert(iter_ == nullptr);
    iter_ = iter;
-    if (iter_ && iter_pinned_) {
-      iter_->PinData();
-    }
  }
  virtual bool Valid() const override { return valid_; }
  virtual Slice key() const override {
@@ -114,32 +107,6 @@ class DBIter: public Iterator {
      return status_;
    }
  }
-  virtual Status PinData() {
-    Status s;
-    if (iter_) {
-      s = iter_->PinData();
-    }
-    if (s.ok()) {
-      // Even if iter_ is nullptr, we set iter_pinned_ to true so that when
-      // iter_ is updated using SetIter, we Pin it.
-      iter_pinned_ = true;
-    }
-    return s;
-  }
-  virtual Status ReleasePinnedData() {
-    Status s;
-    if (iter_) {
-      s = iter_->ReleasePinnedData();
-    }
-    if (s.ok()) {
-      iter_pinned_ = false;
-    }
-    return s;
-  }
-  virtual bool IsKeyPinned() const override {
-    assert(valid_);
-    return iter_pinned_ && saved_key_.IsKeyPinned();
-  }

  virtual void Next() override;
  virtual void Prev() override;
@@ -175,7 +142,7 @@ class DBIter: public Iterator {
  Logger* logger_;
  const Comparator* const user_comparator_;
  const MergeOperator* const user_merge_operator_;
-  InternalIterator* iter_;
+  Iterator* iter_;
  SequenceNumber const sequence_;

  Status status_;
@@ -187,11 +154,6 @@ class DBIter: public Iterator {
  Statistics* statistics_;
  uint64_t max_skip_;
  const Slice* iterate_upper_bound_;
-  IterKey prefix_start_;
-  bool prefix_same_as_start_;
-  bool iter_pinned_;
-  // List of operands for merge operator.
-  std::deque<std::string> merge_operands_;

  // No copying allowed
  DBIter(const DBIter&);
@@ -219,35 +181,14 @@ void DBIter::Next() {
    if (!iter_->Valid()) {
      iter_->SeekToFirst();
    }
-  } else if (iter_->Valid() && !current_entry_is_merged_) {
-    // If the current value is not a merge, the iter position is the
-    // current key, which is already returned. We can safely issue a
-    // Next() without checking the current key.
-    // If the current key is a merge, very likely iter already points
-    // to the next internal position.
-    iter_->Next();
-    PERF_COUNTER_ADD(internal_key_skipped_count, 1);
  }

-  // Now we point to the next internal position, for both of merge and
-  // not merge cases.
+  // If the current value is merged, we might already hit end of iter_
  if (!iter_->Valid()) {
    valid_ = false;
    return;
  }
  FindNextUserEntry(true /* skipping the current user key */);
-  if (statistics_ != nullptr) {
-    RecordTick(statistics_, NUMBER_DB_NEXT);
-    if (valid_) {
-      RecordTick(statistics_, NUMBER_DB_NEXT_FOUND);
-      RecordTick(statistics_, ITER_BYTES_READ, key().size() + value().size());
-    }
-  }
-  if (valid_ && prefix_extractor_ && prefix_same_as_start_ &&
-      prefix_extractor_->Transform(saved_key_.GetKey())
-              .compare(prefix_start_.GetKey()) != 0) {
-    valid_ = false;
-  }
 }

 // PRE: saved_key_ has the current user key if skipping
@@ -287,24 +228,20 @@ void DBIter::FindNextUserEntryInternal(bool skipping) {
        } else {
          switch (ikey.type) {
            case kTypeDeletion:
-            case kTypeSingleDeletion:
              // Arrange to skip all upcoming entries for this key since
              // they are hidden by this deletion.
-              saved_key_.SetKey(ikey.user_key,
-                                !iter_->IsKeyPinned() /* copy */);
+              saved_key_.SetKey(ikey.user_key);
              skipping = true;
              num_skipped = 0;
              PERF_COUNTER_ADD(internal_delete_skipped_count, 1);
              break;
            case kTypeValue:
              valid_ = true;
-              saved_key_.SetKey(ikey.user_key,
-                                !iter_->IsKeyPinned() /* copy */);
+              saved_key_.SetKey(ikey.user_key);
              return;
            case kTypeMerge:
              // By now, we are sure the current ikey is going to yield a value
-              saved_key_.SetKey(ikey.user_key,
-                                !iter_->IsKeyPinned() /* copy */);
+              saved_key_.SetKey(ikey.user_key);
              current_entry_is_merged_ = true;
              valid_ = true;
              MergeValuesNewToOld();  // Go to a different state machine
@@ -364,12 +301,16 @@ void DBIter::MergeValuesNewToOld() {
    if (!user_comparator_->Equal(ikey.user_key, saved_key_.GetKey())) {
      // hit the next user key, stop right here
      break;
-    } else if (kTypeDeletion == ikey.type || kTypeSingleDeletion == ikey.type) {
+    }
+
+    if (kTypeDeletion == ikey.type) {
      // hit a delete with the same user key, stop right here
      // iter_ is positioned after delete
      iter_->Next();
      break;
-    } else if (kTypeValue == ikey.type) {
+    }
+
+    if (kTypeValue == ikey.type) {
      // hit a put, merge the put value with operands and store the
      // final result in saved_value_. We are done!
      // ignore corruption if there is any.
@@ -385,13 +326,13 @@ void DBIter::MergeValuesNewToOld() {
      // iter_ is positioned after put
      iter_->Next();
      return;
-    } else if (kTypeMerge == ikey.type) {
+    }
+
+    if (kTypeMerge == ikey.type) {
      // hit a merge, add the value as an operand and run associative merge.
      // when complete, add result to operands and continue.
      const Slice& val = iter_->value();
      operands.push_front(val.ToString());
-    } else {
-      assert(false);
    }
  }

@@ -414,18 +355,6 @@ void DBIter::Prev() {
    ReverseToBackward();
  }
  PrevInternal();
-  if (statistics_ != nullptr) {
-    RecordTick(statistics_, NUMBER_DB_PREV);
-    if (valid_) {
-      RecordTick(statistics_, NUMBER_DB_PREV_FOUND);
-      RecordTick(statistics_, ITER_BYTES_READ, key().size() + value().size());
-    }
-  }
-  if (valid_ && prefix_extractor_ && prefix_same_as_start_ &&
-      prefix_extractor_->Transform(saved_key_.GetKey())
-              .compare(prefix_start_.GetKey()) != 0) {
-    valid_ = false;
-  }
 }

 void DBIter::ReverseToBackward() {
@@ -464,8 +393,7 @@ void DBIter::PrevInternal() {
  ParsedInternalKey ikey;

  while (iter_->Valid()) {
-    saved_key_.SetKey(ExtractUserKey(iter_->key()),
-                      !iter_->IsKeyPinned() /* copy */);
+    saved_key_.SetKey(ExtractUserKey(iter_->key()));
    if (FindValueForCurrentKey()) {
      valid_ = true;
      if (!iter_->Valid()) {
@@ -491,13 +419,12 @@ void DBIter::PrevInternal() {
 }

 // This function checks, if the entry with biggest sequence_number <= sequence_
-// is non kTypeDeletion or kTypeSingleDeletion. If it's not, we save value in
-// saved_value_
+// is non kTypeDeletion. If it's not, we save value in saved_value_
 bool DBIter::FindValueForCurrentKey() {
  assert(iter_->Valid());
-  merge_operands_.clear();
-  // last entry before merge (could be kTypeDeletion, kTypeSingleDeletion or
-  // kTypeValue)
+  // Contains operands for merge operator.
+  std::deque<std::string> operands;
+  // last entry before merge (could be kTypeDeletion or kTypeValue)
  ValueType last_not_merge_type = kTypeDeletion;
  ValueType last_key_entry_type = kTypeDeletion;

@@ -515,19 +442,18 @@ bool DBIter::FindValueForCurrentKey() {
    last_key_entry_type = ikey.type;
    switch (last_key_entry_type) {
      case kTypeValue:
-        merge_operands_.clear();
+        operands.clear();
        saved_value_ = iter_->value().ToString();
        last_not_merge_type = kTypeValue;
        break;
      case kTypeDeletion:
-      case kTypeSingleDeletion:
-        merge_operands_.clear();
-        last_not_merge_type = last_key_entry_type;
+        operands.clear();
+        last_not_merge_type = kTypeDeletion;
        PERF_COUNTER_ADD(internal_delete_skipped_count, 1);
        break;
      case kTypeMerge:
        assert(user_merge_operator_ != nullptr);
-        merge_operands_.push_back(iter_->value().ToString());
+        operands.push_back(iter_->value().ToString());
        break;
      default:
        assert(false);
@@ -542,16 +468,14 @@ bool DBIter::FindValueForCurrentKey() {

  switch (last_key_entry_type) {
    case kTypeDeletion:
-    case kTypeSingleDeletion:
      valid_ = false;
      return false;
    case kTypeMerge:
      if (last_not_merge_type == kTypeDeletion) {
        StopWatchNano timer(env_, statistics_ != nullptr);
        PERF_TIMER_GUARD(merge_operator_time_nanos);
-        user_merge_operator_->FullMerge(saved_key_.GetKey(), nullptr,
-                                        merge_operands_, &saved_value_,
-                                        logger_);
+        user_merge_operator_->FullMerge(saved_key_.GetKey(), nullptr, operands,
+                                        &saved_value_, logger_);
        RecordTick(statistics_, MERGE_OPERATION_TOTAL_TIME,
                   timer.ElapsedNanos());
      } else {
@@ -562,8 +486,7 @@ bool DBIter::FindValueForCurrentKey() {
          StopWatchNano timer(env_, statistics_ != nullptr);
          PERF_TIMER_GUARD(merge_operator_time_nanos);
          user_merge_operator_->FullMerge(saved_key_.GetKey(), &temp_slice,
-                                          merge_operands_, &saved_value_,
-                                          logger_);
+                                          operands, &saved_value_, logger_);
          RecordTick(statistics_, MERGE_OPERATION_TOTAL_TIME,
                     timer.ElapsedNanos());
        }
@@ -593,8 +516,7 @@ bool DBIter::FindValueForCurrentKeyUsingSeek() {
  ParsedInternalKey ikey;
  FindParseableKey(&ikey, kForward);

-  if (ikey.type == kTypeValue || ikey.type == kTypeDeletion ||
-      ikey.type == kTypeSingleDeletion) {
+  if (ikey.type == kTypeValue || ikey.type == kTypeDeletion) {
    if (ikey.type == kTypeValue) {
      saved_value_ = iter_->value().ToString();
      valid_ = true;
@@ -617,7 +539,7 @@ bool DBIter::FindValueForCurrentKeyUsingSeek() {

  if (!iter_->Valid() ||
      !user_comparator_->Equal(ikey.user_key, saved_key_.GetKey()) ||
-      ikey.type == kTypeDeletion || ikey.type == kTypeSingleDeletion) {
+      ikey.type == kTypeDeletion) {
    {
      StopWatchNano timer(env_, statistics_ != nullptr);
      PERF_TIMER_GUARD(merge_operator_time_nanos);
@@ -715,23 +637,13 @@ void DBIter::Seek(const Slice& target) {
    iter_->Seek(saved_key_.GetKey());
  }

-  RecordTick(statistics_, NUMBER_DB_SEEK);
  if (iter_->Valid()) {
    direction_ = kForward;
    ClearSavedValue();
    FindNextUserEntry(false /* not skipping */);
-    if (statistics_ != nullptr) {
-      if (valid_) {
-        RecordTick(statistics_, NUMBER_DB_SEEK_FOUND);
-        RecordTick(statistics_, ITER_BYTES_READ, key().size() + value().size());
-      }
-    }
  } else {
    valid_ = false;
  }
-  if (valid_ && prefix_extractor_ && prefix_same_as_start_) {
-    prefix_start_.SetKey(prefix_extractor_->Transform(target));
-  }
 }

 void DBIter::SeekToFirst() {
@@ -748,21 +660,11 @@ void DBIter::SeekToFirst() {
    iter_->SeekToFirst();
  }

-  RecordTick(statistics_, NUMBER_DB_SEEK);
  if (iter_->Valid()) {
    FindNextUserEntry(false /* not skipping */);
-    if (statistics_ != nullptr) {
-      if (valid_) {
-        RecordTick(statistics_, NUMBER_DB_SEEK_FOUND);
-        RecordTick(statistics_, ITER_BYTES_READ, key().size() + value().size());
-      }
-    }
  } else {
    valid_ = false;
  }
-  if (valid_ && prefix_extractor_ && prefix_same_as_start_) {
-    prefix_start_.SetKey(prefix_extractor_->Transform(saved_key_.GetKey()));
-  }
 }

 void DBIter::SeekToLast() {
@@ -782,7 +684,7 @@ void DBIter::SeekToLast() {
  // it will seek to the last key before the
  // ReadOptions.iterate_upper_bound
  if (iter_->Valid() && iterate_upper_bound_ != nullptr) {
-    saved_key_.SetKey(*iterate_upper_bound_, false /* copy */);
+    saved_key_.SetKey(*iterate_upper_bound_);
    std::string last_key;
    AppendInternalKey(&last_key,
                      ParsedInternalKey(saved_key_.GetKey(), kMaxSequenceNumber,
@@ -801,40 +703,24 @@ void DBIter::SeekToLast() {
    }
  }
  PrevInternal();
-  if (statistics_ != nullptr) {
-    RecordTick(statistics_, NUMBER_DB_SEEK);
-    if (valid_) {
-      RecordTick(statistics_, NUMBER_DB_SEEK_FOUND);
-      RecordTick(statistics_, ITER_BYTES_READ, key().size() + value().size());
-    }
-  }
-  if (valid_ && prefix_extractor_ && prefix_same_as_start_) {
-    prefix_start_.SetKey(prefix_extractor_->Transform(saved_key_.GetKey()));
-  }
 }

 Iterator* NewDBIterator(Env* env, const ImmutableCFOptions& ioptions,
                        const Comparator* user_key_comparator,
-                        InternalIterator* internal_iter,
+                        Iterator* internal_iter,
                        const SequenceNumber& sequence,
                        uint64_t max_sequential_skip_in_iterations,
-                        const Slice* iterate_upper_bound,
-                        bool prefix_same_as_start, bool pin_data) {
-  DBIter* db_iter =
-      new DBIter(env, ioptions, user_key_comparator, internal_iter, sequence,
-                 false, max_sequential_skip_in_iterations, iterate_upper_bound,
-                 prefix_same_as_start);
-  if (pin_data) {
-    db_iter->PinData();
-  }
-  return db_iter;
+                        const Slice* iterate_upper_bound) {
+  return new DBIter(env, ioptions, user_key_comparator, internal_iter, sequence,
+                    false, max_sequential_skip_in_iterations,
+                    iterate_upper_bound);
 }

 ArenaWrappedDBIter::~ArenaWrappedDBIter() { db_iter_->~DBIter(); }

 void ArenaWrappedDBIter::SetDBIter(DBIter* iter) { db_iter_ = iter; }

-void ArenaWrappedDBIter::SetIterUnderDBIter(InternalIterator* iter) {
+void ArenaWrappedDBIter::SetIterUnderDBIter(Iterator* iter) {
  static_cast<DBIter*>(db_iter_)->SetIter(iter);
 }

@@ -849,13 +735,6 @@ inline void ArenaWrappedDBIter::Prev() { db_iter_->Prev(); }
 inline Slice ArenaWrappedDBIter::key() const { return db_iter_->key(); }
 inline Slice ArenaWrappedDBIter::value() const { return db_iter_->value(); }
 inline Status ArenaWrappedDBIter::status() const { return db_iter_->status(); }
-inline Status ArenaWrappedDBIter::PinData() { return db_iter_->PinData(); }
-inline Status ArenaWrappedDBIter::ReleasePinnedData() {
-  return db_iter_->ReleasePinnedData();
-}
-inline bool ArenaWrappedDBIter::IsKeyPinned() const {
-  return db_iter_->IsKeyPinned();
-}
 void ArenaWrappedDBIter::RegisterCleanup(CleanupFunction function, void* arg1,
                                         void* arg2) {
  db_iter_->RegisterCleanup(function, arg1, arg2);
@@ -863,22 +742,18 @@ void ArenaWrappedDBIter::RegisterCleanup(CleanupFunction function, void* arg1,

 ArenaWrappedDBIter* NewArenaWrappedDbIterator(
    Env* env, const ImmutableCFOptions& ioptions,
-    const Comparator* user_key_comparator, const SequenceNumber& sequence,
+    const Comparator* user_key_comparator,
+    const SequenceNumber& sequence,
    uint64_t max_sequential_skip_in_iterations,
-    const Slice* iterate_upper_bound, bool prefix_same_as_start,
-    bool pin_data) {
+    const Slice* iterate_upper_bound) {
  ArenaWrappedDBIter* iter = new ArenaWrappedDBIter();
  Arena* arena = iter->GetArena();
  auto mem = arena->AllocateAligned(sizeof(DBIter));
-  DBIter* db_iter =
-      new (mem) DBIter(env, ioptions, user_key_comparator, nullptr, sequence,
-                       true, max_sequential_skip_in_iterations,
-                       iterate_upper_bound, prefix_same_as_start);
+  DBIter* db_iter = new (mem) DBIter(env, ioptions, user_key_comparator,
+      nullptr, sequence, true, max_sequential_skip_in_iterations,
+      iterate_upper_bound);

  iter->SetDBIter(db_iter);
-  if (pin_data) {
-    iter->PinData();
-  }

  return iter;
 }
@@ -10,7 +10,6 @@
 #pragma once
 #include <stdint.h>
 #include "rocksdb/db.h"
-#include "rocksdb/iterator.h"
 #include "db/dbformat.h"
 #include "util/arena.h"
 #include "util/autovector.h"
@@ -19,19 +18,18 @@ namespace rocksdb {

 class Arena;
 class DBIter;
-class InternalIterator;

 // Return a new iterator that converts internal keys (yielded by
 // "*internal_iter") that were live at the specified "sequence" number
 // into appropriate user keys.
-extern Iterator* NewDBIterator(Env* env, const ImmutableCFOptions& options,
-                               const Comparator* user_key_comparator,
-                               InternalIterator* internal_iter,
-                               const SequenceNumber& sequence,
-                               uint64_t max_sequential_skip_in_iterations,
-                               const Slice* iterate_upper_bound = nullptr,
-                               bool prefix_same_as_start = false,
-                               bool pin_data = false);
+extern Iterator* NewDBIterator(
+    Env* env,
+    const ImmutableCFOptions& options,
+    const Comparator *user_key_comparator,
+    Iterator* internal_iter,
+    const SequenceNumber& sequence,
+    uint64_t max_sequential_skip_in_iterations,
+    const Slice* iterate_upper_bound = nullptr);

 // A wrapper iterator which wraps DB Iterator and the arena, with which the DB
 // iterator is supposed be allocated. This class is used as an entry point of
@@ -52,7 +50,7 @@ class ArenaWrappedDBIter : public Iterator {

  // Set the internal iterator wrapped inside the DB Iterator. Usually it is
  // a merging iterator.
-  virtual void SetIterUnderDBIter(InternalIterator* iter);
+  virtual void SetIterUnderDBIter(Iterator* iter);
  virtual bool Valid() const override;
  virtual void SeekToFirst() override;
  virtual void SeekToLast() override;
@@ -62,11 +60,7 @@ class ArenaWrappedDBIter : public Iterator {
  virtual Slice key() const override;
  virtual Slice value() const override;
  virtual Status status() const override;
-
  void RegisterCleanup(CleanupFunction function, void* arg1, void* arg2);
-  virtual Status PinData();
-  virtual Status ReleasePinnedData();
-  virtual bool IsKeyPinned() const override;

 private:
  DBIter* db_iter_;
@@ -76,9 +70,8 @@ class ArenaWrappedDBIter : public Iterator {
 // Generate the arena wrapped iterator class.
 extern ArenaWrappedDBIter* NewArenaWrappedDbIterator(
    Env* env, const ImmutableCFOptions& options,
-    const Comparator* user_key_comparator, const SequenceNumber& sequence,
-    uint64_t max_sequential_skip_in_iterations,
-    const Slice* iterate_upper_bound = nullptr,
-    bool prefix_same_as_start = false, bool pin_data = false);
+    const Comparator* user_key_comparator,
+    const SequenceNumber& sequence, uint64_t max_sequential_skip_in_iterations,
+    const Slice* iterate_upper_bound = nullptr);

 }  // namespace rocksdb
@@ -29,7 +29,7 @@ static uint64_t TestGetTickerCount(const Options& options,
  return options.statistics->getTickerCount(ticker_type);
 }

-class TestIterator : public InternalIterator {
+class TestIterator : public Iterator {
 public:
  explicit TestIterator(const Comparator* comparator)
      : initialized_(false),
@@ -38,20 +38,16 @@ class TestIterator : public InternalIterator {
        iter_(0),
        cmp(comparator) {}

-  void AddPut(std::string argkey, std::string argvalue) {
-    Add(argkey, kTypeValue, argvalue);
+  void AddMerge(std::string argkey, std::string argvalue) {
+    Add(argkey, kTypeMerge, argvalue);
  }

  void AddDeletion(std::string argkey) {
    Add(argkey, kTypeDeletion, std::string());
  }

-  void AddSingleDeletion(std::string argkey) {
-    Add(argkey, kTypeSingleDeletion, std::string());
-  }
-
-  void AddMerge(std::string argkey, std::string argvalue) {
-    Add(argkey, kTypeMerge, argvalue);
+  void AddPut(std::string argkey, std::string argvalue) {
+    Add(argkey, kTypeValue, argvalue);
  }

  void Add(std::string argkey, ValueType type, std::string argvalue) {
@@ -895,8 +891,6 @@ TEST_F(DBIteratorTest, DBIterator1) {
  db_iter->Next();
  ASSERT_TRUE(db_iter->Valid());
  ASSERT_EQ(db_iter->key().ToString(), "b");
-  db_iter->Next();
-  ASSERT_FALSE(db_iter->Valid());
 }

 TEST_F(DBIteratorTest, DBIterator2) {
@@ -1793,58 +1787,6 @@ TEST_F(DBIteratorTest, SeekToLastOccurrenceSeq0) {
  ASSERT_FALSE(db_iter->Valid());
 }

-TEST_F(DBIteratorTest, DBIterator11) {
-  Options options;
-  options.merge_operator = MergeOperators::CreateFromStringId("stringappend");
-
-  TestIterator* internal_iter = new TestIterator(BytewiseComparator());
-  internal_iter->AddPut("a", "0");
-  internal_iter->AddPut("b", "0");
-  internal_iter->AddSingleDeletion("b");
-  internal_iter->AddMerge("a", "1");
-  internal_iter->AddMerge("b", "2");
-  internal_iter->Finish();
-
-  std::unique_ptr<Iterator> db_iter(NewDBIterator(
-      env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter, 1,
-      options.max_sequential_skip_in_iterations));
-  db_iter->SeekToFirst();
-  ASSERT_TRUE(db_iter->Valid());
-  ASSERT_EQ(db_iter->key().ToString(), "a");
-  ASSERT_EQ(db_iter->value().ToString(), "0");
-  db_iter->Next();
-  ASSERT_TRUE(db_iter->Valid());
-  ASSERT_EQ(db_iter->key().ToString(), "b");
-  db_iter->Next();
-  ASSERT_FALSE(db_iter->Valid());
-}
-
-TEST_F(DBIteratorTest, DBIterator12) {
-  Options options;
-  options.merge_operator = nullptr;
-
-  TestIterator* internal_iter = new TestIterator(BytewiseComparator());
-  internal_iter->AddPut("a", "1");
-  internal_iter->AddPut("b", "2");
-  internal_iter->AddPut("c", "3");
-  internal_iter->AddSingleDeletion("b");
-  internal_iter->Finish();
-
-  std::unique_ptr<Iterator> db_iter(
-      NewDBIterator(env_, ImmutableCFOptions(options), BytewiseComparator(),
-                    internal_iter, 10, 0));
-  db_iter->SeekToLast();
-  ASSERT_TRUE(db_iter->Valid());
-  ASSERT_EQ(db_iter->key().ToString(), "c");
-  ASSERT_EQ(db_iter->value().ToString(), "3");
-  db_iter->Prev();
-  ASSERT_TRUE(db_iter->Valid());
-  ASSERT_EQ(db_iter->key().ToString(), "a");
-  ASSERT_EQ(db_iter->value().ToString(), "1");
-  db_iter->Prev();
-  ASSERT_FALSE(db_iter->Valid());
-}
-
 class DBIterWithMergeIterTest : public testing::Test {
 public:
  DBIterWithMergeIterTest()
@@ -1864,12 +1806,11 @@ class DBIterWithMergeIterTest : public testing::Test {
    internal_iter2_->Add("d", kTypeValue, "7", 3u);
    internal_iter2_->Finish();

-    std::vector<InternalIterator*> child_iters;
+    std::vector<Iterator*> child_iters;
    child_iters.push_back(internal_iter1_);
    child_iters.push_back(internal_iter2_);
    InternalKeyComparator icomp(BytewiseComparator());
-    InternalIterator* merge_iter =
-        NewMergingIterator(&icomp_, &child_iters[0], 2u);
+    Iterator* merge_iter = NewMergingIterator(&icomp_, &child_iters[0], 2u);

    db_iter_.reset(NewDBIterator(env_, ImmutableCFOptions(options_),
                                 BytewiseComparator(), merge_iter,
@@ -10,10 +10,10 @@
 // Introduction of SyncPoint effectively disabled building and running this test
 // in Release build.
 // which is a pity, it is a good test
-#if !defined(ROCKSDB_LITE)
+#if !(defined NDEBUG) || !defined(OS_WIN)

-#include "db/db_test_util.h"
 #include "port/stack_trace.h"
+#include "util/db_test_util.h"

 namespace rocksdb {

@@ -27,7 +27,7 @@ class DBTestXactLogIterator : public DBTestBase {
    Status status = dbfull()->GetUpdatesSince(seq, &iter);
    EXPECT_OK(status);
    EXPECT_TRUE(iter->Valid());
-    return iter;
+    return std::move(iter);
  }
 };

@@ -277,10 +277,10 @@ TEST_F(DBTestXactLogIterator, TransactionLogIteratorBlobs) {
 }
 }  // namespace rocksdb

-#endif  // !defined(ROCKSDB_LITE)
+#endif  // !(defined NDEBUG) || !defined(OS_WIN)

 int main(int argc, char** argv) {
-#if !defined(ROCKSDB_LITE)
+#if !(defined NDEBUG) || !defined(OS_WIN)
  rocksdb::port::InstallStackTraceHandler();
  ::testing::InitGoogleTest(&argc, argv);
  return RUN_ALL_TESTS();
@@ -1,226 +0,0 @@
-//  Copyright (c) 2013, Facebook, Inc.  All rights reserved.
-//  This source code is licensed under the BSD-style license found in the
-//  LICENSE file in the root directory of this source tree. An additional grant
-//  of patent rights can be found in the PATENTS file in the same 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 <unordered_set>
-#include <vector>
-
-#include "db/db_test_util.h"
-#include "port/stack_trace.h"
-#include "rocksdb/db.h"
-#include "util/testharness.h"
-#include "util/testutil.h"
-
-#ifndef ROCKSDB_LITE
-
-namespace rocksdb {
-
-// A helper function that ensures the table properties returned in
-// `GetPropertiesOfAllTablesTest` is correct.
-// This test assumes entries size is different for each of the tables.
-namespace {
-
-void VerifyTableProperties(DB* db, uint64_t expected_entries_size) {
-  TablePropertiesCollection props;
-  ASSERT_OK(db->GetPropertiesOfAllTables(&props));
-
-  ASSERT_EQ(4U, props.size());
-  std::unordered_set<uint64_t> unique_entries;
-
-  // Indirect test
-  uint64_t sum = 0;
-  for (const auto& item : props) {
-    unique_entries.insert(item.second->num_entries);
-    sum += item.second->num_entries;
-  }
-
-  ASSERT_EQ(props.size(), unique_entries.size());
-  ASSERT_EQ(expected_entries_size, sum);
-}
-}  // namespace
-
-class DBTablePropertiesTest : public DBTestBase {
- public:
-  DBTablePropertiesTest() : DBTestBase("/db_table_properties_test") {}
-  TablePropertiesCollection TestGetPropertiesOfTablesInRange(
-      std::vector<Range> ranges, std::size_t* num_properties = nullptr,
-      std::size_t* num_files = nullptr);
-};
-
-TEST_F(DBTablePropertiesTest, GetPropertiesOfAllTablesTest) {
-  Options options = CurrentOptions();
-  options.level0_file_num_compaction_trigger = 8;
-  Reopen(options);
-  // Create 4 tables
-  for (int table = 0; table < 4; ++table) {
-    for (int i = 0; i < 10 + table; ++i) {
-      db_->Put(WriteOptions(), ToString(table * 100 + i), "val");
-    }
-    db_->Flush(FlushOptions());
-  }
-
-  // 1. Read table properties directly from file
-  Reopen(options);
-  VerifyTableProperties(db_, 10 + 11 + 12 + 13);
-
-  // 2. Put two tables to table cache and
-  Reopen(options);
-  // fetch key from 1st and 2nd table, which will internally place that table to
-  // the table cache.
-  for (int i = 0; i < 2; ++i) {
-    Get(ToString(i * 100 + 0));
-  }
-
-  VerifyTableProperties(db_, 10 + 11 + 12 + 13);
-
-  // 3. Put all tables to table cache
-  Reopen(options);
-  // fetch key from 1st and 2nd table, which will internally place that table to
-  // the table cache.
-  for (int i = 0; i < 4; ++i) {
-    Get(ToString(i * 100 + 0));
-  }
-  VerifyTableProperties(db_, 10 + 11 + 12 + 13);
-}
-
-TablePropertiesCollection
-DBTablePropertiesTest::TestGetPropertiesOfTablesInRange(
-    std::vector<Range> ranges, std::size_t* num_properties,
-    std::size_t* num_files) {
-
-  // Since we deref zero element in the vector it can not be empty
-  // otherwise we pass an address to some random memory
-  EXPECT_GT(ranges.size(), 0U);
-  // run the query
-  TablePropertiesCollection props;
-  EXPECT_OK(db_->GetPropertiesOfTablesInRange(
-      db_->DefaultColumnFamily(), &ranges[0], ranges.size(), &props));
-
-  // Make sure that we've received properties for those and for those files
-  // only which fall within requested ranges
-  std::vector<LiveFileMetaData> vmd;
-  db_->GetLiveFilesMetaData(&vmd);
-  for (auto& md : vmd) {
-    std::string fn = md.db_path + md.name;
-    bool in_range = false;
-    for (auto& r : ranges) {
-      // smallestkey < limit && largestkey >= start
-      if (r.limit.compare(md.smallestkey) >= 0 &&
-          r.start.compare(md.largestkey) <= 0) {
-        in_range = true;
-        EXPECT_GT(props.count(fn), 0);
-      }
-    }
-    if (!in_range) {
-      EXPECT_EQ(props.count(fn), 0);
-    }
-  }
-
-  if (num_properties) {
-    *num_properties = props.size();
-  }
-
-  if (num_files) {
-    *num_files = vmd.size();
-  }
-  return props;
-}
-
-TEST_F(DBTablePropertiesTest, GetPropertiesOfTablesInRange) {
-  // Fixed random sead
-  Random rnd(301);
-
-  Options options;
-  options.create_if_missing = true;
-  options.write_buffer_size = 4096;
-  options.max_write_buffer_number = 8;
-  options.level0_file_num_compaction_trigger = 2;
-  options.level0_slowdown_writes_trigger = 2;
-  options.level0_stop_writes_trigger = 4;
-  options.target_file_size_base = 2048;
-  options.max_bytes_for_level_base = 10240;
-  options.max_bytes_for_level_multiplier = 4;
-  options.soft_rate_limit = 1.1;
-  options.num_levels = 8;
-
-  DestroyAndReopen(options);
-
-  // build a decent LSM
-  for (int i = 0; i < 10000; i++) {
-    ASSERT_OK(Put(test::RandomKey(&rnd, 5), RandomString(&rnd, 102)));
-  }
-  Flush();
-  db_->PauseBackgroundWork();
-
-  // Ensure that we have at least L0, L1 and L2
-  ASSERT_GT(NumTableFilesAtLevel(0), 0);
-  ASSERT_GT(NumTableFilesAtLevel(1), 0);
-  ASSERT_GT(NumTableFilesAtLevel(2), 0);
-
-  // Query the largest range
-  std::size_t num_properties, num_files;
-  TestGetPropertiesOfTablesInRange(
-      {Range(test::RandomKey(&rnd, 5, test::RandomKeyType::SMALLEST),
-             test::RandomKey(&rnd, 5, test::RandomKeyType::LARGEST))},
-      &num_properties, &num_files);
-  ASSERT_EQ(num_properties, num_files);
-
-  // Query the empty range
-  TestGetPropertiesOfTablesInRange(
-      {Range(test::RandomKey(&rnd, 5, test::RandomKeyType::LARGEST),
-             test::RandomKey(&rnd, 5, test::RandomKeyType::SMALLEST))},
-      &num_properties, &num_files);
-  ASSERT_GT(num_files, 0);
-  ASSERT_EQ(num_properties, 0);
-
-  // Query the middle rangee
-  TestGetPropertiesOfTablesInRange(
-      {Range(test::RandomKey(&rnd, 5, test::RandomKeyType::MIDDLE),
-             test::RandomKey(&rnd, 5, test::RandomKeyType::LARGEST))},
-      &num_properties, &num_files);
-  ASSERT_GT(num_files, 0);
-  ASSERT_GT(num_files, num_properties);
-  ASSERT_GT(num_properties, 0);
-
-  // Query a bunch of random ranges
-  for (int j = 0; j < 100; j++) {
-    // create a bunch of ranges
-    std::vector<std::string> random_keys;
-    // Random returns numbers with zero included
-    // when we pass empty ranges TestGetPropertiesOfTablesInRange()
-    // derefs random memory in the empty ranges[0]
-    // so want to be greater than zero and even since
-    // the below loop requires that random_keys.size() to be even.
-    auto n = 2 * (rnd.Uniform(50) + 1);
-
-    for (uint32_t i = 0; i < n; ++i) {
-      random_keys.push_back(test::RandomKey(&rnd, 5));
-    }
-
-    ASSERT_GT(random_keys.size(), 0U);
-    ASSERT_EQ((random_keys.size() % 2), 0U);
-
-    std::vector<Range> ranges;
-    auto it = random_keys.begin();
-    while (it != random_keys.end()) {
-      ranges.push_back(Range(*it, *(it + 1)));
-      it += 2;
-    }
-
-    TestGetPropertiesOfTablesInRange(std::move(ranges));
-  }
-}
-}  // namespace rocksdb
-
-#endif  // ROCKSDB_LITE
-
-int main(int argc, char** argv) {
-  rocksdb::port::InstallStackTraceHandler();
-  ::testing::InitGoogleTest(&argc, argv);
-  return RUN_ALL_TESTS();
-}
@@ -10,11 +10,11 @@
 // Introduction of SyncPoint effectively disabled building and running this test
 // in Release build.
 // which is a pity, it is a good test
-#if !defined(ROCKSDB_LITE)
+#if !(defined NDEBUG) || !defined(OS_WIN)

-#include "db/db_test_util.h"
 #include "db/forward_iterator.h"
 #include "port/stack_trace.h"
+#include "util/db_test_util.h"

 namespace rocksdb {

@@ -140,8 +140,6 @@ TEST_F(DBTestTailingIterator, TailingIteratorTrimSeekToNext) {
  std::unique_ptr<Iterator> iterh(db_->NewIterator(read_options, handles_[1]));
  std::string value(1024, 'a');
  bool file_iters_deleted = false;
-  bool file_iters_renewed_null = false;
-  bool file_iters_renewed_copy = false;
  rocksdb::SyncPoint::GetInstance()->SetCallBack(
      "ForwardIterator::SeekInternal:Return", [&](void* arg) {
        ForwardIterator* fiter = reinterpret_cast<ForwardIterator*>(arg);
@@ -154,12 +152,6 @@ TEST_F(DBTestTailingIterator, TailingIteratorTrimSeekToNext) {
        ASSERT_TRUE(!file_iters_deleted ||
                    fiter->TEST_CheckDeletedIters(&deleted_iters, &num_iters));
      });
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
-      "ForwardIterator::RenewIterators:Null",
-      [&](void* arg) { file_iters_renewed_null = true; });
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
-      "ForwardIterator::RenewIterators:Copy",
-      [&](void* arg) { file_iters_renewed_copy = true; });
  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
  const int num_records = 1000;
  for (int i = 1; i < num_records; ++i) {
@@ -168,7 +160,7 @@ TEST_F(DBTestTailingIterator, TailingIteratorTrimSeekToNext) {
    char buf3[32];
    char buf4[32];
    snprintf(buf1, sizeof(buf1), "00a0%016d", i * 5);
-    snprintf(buf3, sizeof(buf3), "00b0%016d", i * 5);
+    snprintf(buf3, sizeof(buf1), "00b0%016d", i * 5);

    Slice key(buf1, 20);
    ASSERT_OK(Put(1, key, value));
@@ -181,7 +173,7 @@ TEST_F(DBTestTailingIterator, TailingIteratorTrimSeekToNext) {
      if (i == 299) {
        file_iters_deleted = true;
      }
-      snprintf(buf4, sizeof(buf4), "00a0%016d", i * 5 / 2);
+      snprintf(buf4, sizeof(buf1), "00a0%016d", i * 5 / 2);
      Slice target(buf4, 20);
      iterh->Seek(target);
      ASSERT_TRUE(iter->Valid());
@@ -211,8 +203,6 @@ TEST_F(DBTestTailingIterator, TailingIteratorTrimSeekToNext) {
    ASSERT_LE(num_iters, 1);
    file_iters_deleted = false;
  }
-  ASSERT_TRUE(file_iters_renewed_null);
-  ASSERT_TRUE(file_iters_renewed_copy);
  iter = 0;
  itern = 0;
  iterh = 0;
@@ -656,10 +646,10 @@ TEST_F(DBTestTailingIterator, ManagedTailingIteratorSeekToSame) {

 }  // namespace rocksdb

-#endif  // !defined(ROCKSDB_LITE)
+#endif  // !(defined NDEBUG) || !defined(OS_WIN)

 int main(int argc, char** argv) {
-#if !defined(ROCKSDB_LITE)
+#if !(defined NDEBUG) || !defined(OS_WIN)
  rocksdb::port::InstallStackTraceHandler();
  ::testing::InitGoogleTest(&argc, argv);
  return RUN_ALL_TESTS();
@@ -7,18 +7,13 @@
 // 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 "db/db_test_util.h"
 #include "port/stack_trace.h"
-#if !defined(ROCKSDB_LITE)
+#include "util/db_test_util.h"
+#if !(defined NDEBUG) || !defined(OS_WIN)
 #include "util/sync_point.h"

 namespace rocksdb {

-static uint64_t TestGetTickerCount(const Options& options,
-                                   Tickers ticker_type) {
-  return options.statistics->getTickerCount(ticker_type);
-}
-
 static std::string CompressibleString(Random* rnd, int len) {
  std::string r;
  test::CompressibleString(rnd, 0.8, len, &r);
@@ -27,16 +22,12 @@ static std::string CompressibleString(Random* rnd, int len) {

 class DBTestUniversalCompactionBase
    : public DBTestBase,
-      public ::testing::WithParamInterface<std::tuple<int, bool>> {
+      public ::testing::WithParamInterface<int> {
 public:
  explicit DBTestUniversalCompactionBase(
      const std::string& path) : DBTestBase(path) {}
-  virtual void SetUp() override {
-    num_levels_ = std::get<0>(GetParam());
-    exclusive_manual_compaction_ = std::get<1>(GetParam());
-  }
+  virtual void SetUp() override { num_levels_ = GetParam(); }
  int num_levels_;
-  bool exclusive_manual_compaction_;
 };

 class DBTestUniversalCompaction : public DBTestUniversalCompactionBase {
@@ -121,101 +112,6 @@ class DelayFilterFactory : public CompactionFilterFactory {
 };
 }  // namespace

-// Make sure we don't trigger a problem if the trigger conditon is given
-// to be 0, which is invalid.
-TEST_P(DBTestUniversalCompaction, UniversalCompactionSingleSortedRun) {
-  Options options;
-  options = CurrentOptions(options);
-
-  options.compaction_style = kCompactionStyleUniversal;
-  options.num_levels = num_levels_;
-  // Config universal compaction to always compact to one single sorted run.
-  options.level0_file_num_compaction_trigger = 0;
-  options.compaction_options_universal.size_ratio = 10;
-  options.compaction_options_universal.min_merge_width = 2;
-  options.compaction_options_universal.max_size_amplification_percent = 1;
-
-  options.write_buffer_size = 105 << 10;  // 105KB
-  options.arena_block_size = 4 << 10;
-  options.target_file_size_base = 32 << 10;  // 32KB
-  // trigger compaction if there are >= 4 files
-  KeepFilterFactory* filter = new KeepFilterFactory(true);
-  filter->expect_manual_compaction_.store(false);
-  options.compaction_filter_factory.reset(filter);
-
-  DestroyAndReopen(options);
-  ASSERT_EQ(1, db_->GetOptions().level0_file_num_compaction_trigger);
-
-  Random rnd(301);
-  int key_idx = 0;
-
-  filter->expect_full_compaction_.store(true);
-
-  for (int num = 0; num < 16; num++) {
-    // Write 100KB file. And immediately it should be compacted to one file.
-    GenerateNewFile(&rnd, &key_idx);
-    dbfull()->TEST_WaitForCompact();
-    ASSERT_EQ(NumSortedRuns(0), 1);
-  }
-}
-
-TEST_P(DBTestUniversalCompaction, OptimizeFiltersForHits) {
-  Options options;
-  options = CurrentOptions(options);
-  options.compaction_style = kCompactionStyleUniversal;
-  options.compaction_options_universal.size_ratio = 5;
-  options.num_levels = num_levels_;
-  options.write_buffer_size = 105 << 10;  // 105KB
-  options.arena_block_size = 4 << 10;
-  options.target_file_size_base = 32 << 10;  // 32KB
-  // trigger compaction if there are >= 4 files
-  options.level0_file_num_compaction_trigger = 4;
-  BlockBasedTableOptions bbto;
-  bbto.cache_index_and_filter_blocks = true;
-  bbto.filter_policy.reset(NewBloomFilterPolicy(10, false));
-  bbto.whole_key_filtering = true;
-  options.table_factory.reset(NewBlockBasedTableFactory(bbto));
-  options.optimize_filters_for_hits = true;
-  options.statistics = rocksdb::CreateDBStatistics();
-  options.memtable_factory.reset(new SpecialSkipListFactory(3));
-
-  DestroyAndReopen(options);
-
-  // block compaction from happening
-  env_->SetBackgroundThreads(1, Env::LOW);
-  test::SleepingBackgroundTask sleeping_task_low;
-  env_->Schedule(&test::SleepingBackgroundTask::DoSleepTask, &sleeping_task_low,
-                 Env::Priority::LOW);
-
-  Put("", "");
-  for (int num = 0; num < options.level0_file_num_compaction_trigger; num++) {
-    Put(Key(num * 10), "val");
-    Put(Key(30 + num * 10), "val");
-    Put(Key(60 + num * 10), "val");
-
-    dbfull()->TEST_WaitForFlushMemTable();
-  }
-
-  // Query set of non existing keys
-  for (int i = 5; i < 90; i += 10) {
-    ASSERT_EQ(Get(Key(i)), "NOT_FOUND");
-  }
-
-  // Make sure bloom filter is used at least once.
-  ASSERT_GT(TestGetTickerCount(options, BLOOM_FILTER_USEFUL), 0);
-  auto prev_counter = TestGetTickerCount(options, BLOOM_FILTER_USEFUL);
-
-  // Unblock compaction and wait it for happening.
-  sleeping_task_low.WakeUp();
-  dbfull()->TEST_WaitForCompact();
-
-  // The same queries will not trigger bloom filter
-  for (int i = 5; i < 90; i += 10) {
-    ASSERT_EQ(Get(Key(i)), "NOT_FOUND");
-  }
-  ASSERT_EQ(prev_counter, TestGetTickerCount(options, BLOOM_FILTER_USEFUL));
-}
-
 // TODO(kailiu) The tests on UniversalCompaction has some issues:
 //  1. A lot of magic numbers ("11" or "12").
 //  2. Made assumption on the memtable flush conditions, which may change from
@@ -223,7 +119,6 @@ TEST_P(DBTestUniversalCompaction, OptimizeFiltersForHits) {
 TEST_P(DBTestUniversalCompaction, UniversalCompactionTrigger) {
  Options options;
  options.compaction_style = kCompactionStyleUniversal;
-  options.compaction_options_universal.size_ratio = 5;
  options.num_levels = num_levels_;
  options.write_buffer_size = 105 << 10;  // 105KB
  options.arena_block_size = 4 << 10;
@@ -472,7 +367,6 @@ TEST_P(DBTestUniversalCompaction, UniversalCompactionTargetLevel) {
  CompactRangeOptions compact_options;
  compact_options.change_level = true;
  compact_options.target_level = 4;
-  compact_options.exclusive_manual_compaction = exclusive_manual_compaction_;
  db_->CompactRange(compact_options, nullptr, nullptr);
  ASSERT_EQ("0,0,0,0,1", FilesPerLevel(0));
 }
@@ -565,8 +459,7 @@ TEST_P(DBTestUniversalCompactionMultiLevels, UniversalCompactionTrivialMove) {

 INSTANTIATE_TEST_CASE_P(DBTestUniversalCompactionMultiLevels,
                        DBTestUniversalCompactionMultiLevels,
-                        ::testing::Combine(::testing::Values(3, 20),
-                                           ::testing::Bool()));
+                        ::testing::Values(3, 20));

 class DBTestUniversalCompactionParallel :
    public DBTestUniversalCompactionBase {
@@ -639,8 +532,7 @@ TEST_P(DBTestUniversalCompactionParallel, UniversalCompactionParallel) {

 INSTANTIATE_TEST_CASE_P(DBTestUniversalCompactionParallel,
                        DBTestUniversalCompactionParallel,
-                        ::testing::Combine(::testing::Values(1, 10),
-                                           ::testing::Bool()));
+                        ::testing::Values(1, 10));

 TEST_P(DBTestUniversalCompaction, UniversalCompactionOptions) {
  Options options;
@@ -959,10 +851,7 @@ TEST_P(DBTestUniversalCompaction, UniversalCompactionFourPaths) {
  options.db_paths.emplace_back(dbname_ + "_2", 300 * 1024);
  options.db_paths.emplace_back(dbname_ + "_3", 500 * 1024);
  options.db_paths.emplace_back(dbname_ + "_4", 1024 * 1024 * 1024);
-  options.memtable_factory.reset(
-      new SpecialSkipListFactory(KNumKeysByGenerateNewFile - 1));
  options.compaction_style = kCompactionStyleUniversal;
-  options.compaction_options_universal.size_ratio = 5;
  options.write_buffer_size = 110 << 10;  // 105KB
  options.arena_block_size = 4 << 10;
  options.level0_file_num_compaction_trigger = 2;
@@ -1079,15 +968,13 @@ TEST_P(DBTestUniversalCompaction, IncreaseUniversalCompactionNumLevels) {
  int max_key1 = 200;
  int max_key2 = 600;
  int max_key3 = 800;
-  const int KNumKeysPerFile = 10;

  // Stage 1: open a DB with universal compaction, num_levels=1
  Options options = CurrentOptions();
  options.compaction_style = kCompactionStyleUniversal;
  options.num_levels = 1;
-  options.write_buffer_size = 200 << 10;  // 200KB
+  options.write_buffer_size = 100 << 10;  // 100KB
  options.level0_file_num_compaction_trigger = 3;
-  options.memtable_factory.reset(new SpecialSkipListFactory(KNumKeysPerFile));
  options = CurrentOptions(options);
  CreateAndReopenWithCF({"pikachu"}, options);

@@ -1132,7 +1019,6 @@ TEST_P(DBTestUniversalCompaction, IncreaseUniversalCompactionNumLevels) {
  CompactRangeOptions compact_options;
  compact_options.change_level = true;
  compact_options.target_level = 0;
-  compact_options.exclusive_manual_compaction = exclusive_manual_compaction_;
  dbfull()->CompactRange(compact_options, handles_[1], nullptr, nullptr);
  // Need to restart it once to remove higher level records in manifest.
  ReopenWithColumnFamilies({"default", "pikachu"}, options);
@@ -1161,14 +1047,11 @@ TEST_P(DBTestUniversalCompaction, UniversalCompactionSecondPathRatio) {
  options.db_paths.emplace_back(dbname_, 500 * 1024);
  options.db_paths.emplace_back(dbname_ + "_2", 1024 * 1024 * 1024);
  options.compaction_style = kCompactionStyleUniversal;
-  options.compaction_options_universal.size_ratio = 5;
  options.write_buffer_size = 110 << 10;  // 105KB
  options.arena_block_size = 4 * 1024;
  options.arena_block_size = 4 << 10;
  options.level0_file_num_compaction_trigger = 2;
  options.num_levels = 1;
-  options.memtable_factory.reset(
-      new SpecialSkipListFactory(KNumKeysByGenerateNewFile - 1));
  options = CurrentOptions(options);

  std::vector<std::string> filenames;
@@ -1256,8 +1139,7 @@ TEST_P(DBTestUniversalCompaction, UniversalCompactionSecondPathRatio) {
 }

 INSTANTIATE_TEST_CASE_P(UniversalCompactionNumLevels, DBTestUniversalCompaction,
-                        ::testing::Combine(::testing::Values(1, 3, 5),
-                                           ::testing::Bool()));
+                        ::testing::Values(1, 3, 5));

 class DBTestUniversalManualCompactionOutputPathId
    : public DBTestUniversalCompactionBase {
@@ -1289,7 +1171,6 @@ TEST_P(DBTestUniversalManualCompactionOutputPathId,
  // Full compaction to DB path 0
  CompactRangeOptions compact_options;
  compact_options.target_path_id = 1;
-  compact_options.exclusive_manual_compaction = exclusive_manual_compaction_;
  db_->CompactRange(compact_options, handles_[1], nullptr, nullptr);
  ASSERT_EQ(1, TotalLiveFiles(1));
  ASSERT_EQ(0, GetSstFileCount(options.db_paths[0].path));
@@ -1312,7 +1193,6 @@ TEST_P(DBTestUniversalManualCompactionOutputPathId,

  // Full compaction to DB path 0
  compact_options.target_path_id = 0;
-  compact_options.exclusive_manual_compaction = exclusive_manual_compaction_;
  db_->CompactRange(compact_options, handles_[1], nullptr, nullptr);
  ASSERT_EQ(1, TotalLiveFiles(1));
  ASSERT_EQ(1, GetSstFileCount(options.db_paths[0].path));
@@ -1320,22 +1200,20 @@ TEST_P(DBTestUniversalManualCompactionOutputPathId,

  // Fail when compacting to an invalid path ID
  compact_options.target_path_id = 2;
-  compact_options.exclusive_manual_compaction = exclusive_manual_compaction_;
  ASSERT_TRUE(db_->CompactRange(compact_options, handles_[1], nullptr, nullptr)
                  .IsInvalidArgument());
 }

 INSTANTIATE_TEST_CASE_P(DBTestUniversalManualCompactionOutputPathId,
                        DBTestUniversalManualCompactionOutputPathId,
-                        ::testing::Combine(::testing::Values(1, 8),
-                                           ::testing::Bool()));
+                        ::testing::Values(1, 8));

 }  // namespace rocksdb

-#endif  // !defined(ROCKSDB_LITE)
+#endif  // !(defined NDEBUG) || !defined(OS_WIN)

 int main(int argc, char** argv) {
-#if !defined(ROCKSDB_LITE)
+#if !(defined NDEBUG) || !defined(OS_WIN)
  rocksdb::port::InstallStackTraceHandler();
  ::testing::InitGoogleTest(&argc, argv);
  return RUN_ALL_TESTS();
@@ -7,9 +7,11 @@
 // 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 "db/db_test_util.h"
 #include "port/stack_trace.h"
+#include "util/db_test_util.h"
+#if !(defined NDEBUG) || !defined(OS_WIN)
 #include "util/sync_point.h"
+#endif

 namespace rocksdb {
 class DBWALTest : public DBTestBase {
@@ -68,6 +70,7 @@ TEST_F(DBWALTest, RollLog) {
  } while (ChangeOptions());
 }

+#if !(defined NDEBUG) || !defined(OS_WIN)
 TEST_F(DBWALTest, SyncWALNotBlockWrite) {
  Options options = CurrentOptions();
  options.max_write_buffer_number = 4;
@@ -127,10 +130,15 @@ TEST_F(DBWALTest, SyncWALNotWaitWrite) {
  ASSERT_EQ(Get("foo2"), "bar2");
  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
 }
+#endif
 }  // namespace rocksdb

 int main(int argc, char** argv) {
+#if !(defined NDEBUG) || !defined(OS_WIN)
  rocksdb::port::InstallStackTraceHandler();
  ::testing::InitGoogleTest(&argc, argv);
  return RUN_ALL_TESTS();
+#else
+  return 0;
+#endif
 }
@@ -22,7 +22,7 @@ namespace rocksdb {

 uint64_t PackSequenceAndType(uint64_t seq, ValueType t) {
  assert(seq <= kMaxSequenceNumber);
-  assert(IsValueType(t));
+  assert(t <= kValueTypeForSeek);
  return (seq << 8) | t;
 }

@@ -31,7 +31,7 @@ void UnPackSequenceAndType(uint64_t packed, uint64_t* seq, ValueType* t) {
  *t = static_cast<ValueType>(packed & 0xff);

  assert(*seq <= kMaxSequenceNumber);
-  assert(IsValueType(*t));
+  assert(*t <= kValueTypeForSeek);
 }

 void AppendInternalKey(std::string* result, const ParsedInternalKey& key) {
@@ -33,13 +33,13 @@ enum ValueType : unsigned char {
  kTypeDeletion = 0x0,
  kTypeValue = 0x1,
  kTypeMerge = 0x2,
-  kTypeLogData = 0x3,               // WAL only.
-  kTypeColumnFamilyDeletion = 0x4,  // WAL only.
-  kTypeColumnFamilyValue = 0x5,     // WAL only.
-  kTypeColumnFamilyMerge = 0x6,     // WAL only.
-  kTypeSingleDeletion = 0x7,
-  kTypeColumnFamilySingleDeletion = 0x8,  // WAL only.
-  kMaxValue = 0x7F                        // Not used for storing records.
+  // Following types are used only in write ahead logs. They are not used in
+  // memtables or sst files:
+  kTypeLogData = 0x3,
+  kTypeColumnFamilyDeletion = 0x4,
+  kTypeColumnFamilyValue = 0x5,
+  kTypeColumnFamilyMerge = 0x6,
+  kMaxValue = 0x7F
 };

 // kValueTypeForSeek defines the ValueType that should be passed when
@@ -48,13 +48,7 @@ enum ValueType : unsigned char {
 // and the value type is embedded as the low 8 bits in the sequence
 // number in internal keys, we need to use the highest-numbered
 // ValueType, not the lowest).
-static const ValueType kValueTypeForSeek = kTypeSingleDeletion;
-
-// Checks whether a type is a value type (i.e. a type used in memtables and sst
-// files).
-inline bool IsValueType(ValueType t) {
-  return t <= kTypeMerge || t == kTypeSingleDeletion;
-}
+static const ValueType kValueTypeForSeek = kTypeMerge;

 // We leave eight bits empty at the bottom so a type and sequence#
 // can be packed together into 64-bits.
@@ -199,12 +193,13 @@ inline bool ParseInternalKey(const Slice& internal_key,
  result->type = static_cast<ValueType>(c);
  assert(result->type <= ValueType::kMaxValue);
  result->user_key = Slice(internal_key.data(), n - 8);
-  return IsValueType(result->type);
+  return (c <= static_cast<unsigned char>(kValueTypeForSeek));
 }

 // Update the sequence number in the internal key.
 // Guarantees not to invalidate ikey.data().
-inline void UpdateInternalKey(std::string* ikey, uint64_t seq, ValueType t) {
+inline void UpdateInternalKey(std::string* ikey,
+                              uint64_t seq, ValueType t) {
  size_t ikey_sz = ikey->size();
  assert(ikey_sz >= 8);
  uint64_t newval = (seq << 8) | t;
@@ -271,19 +266,13 @@ inline LookupKey::~LookupKey() {

 class IterKey {
 public:
-  IterKey()
-      : buf_(space_), buf_size_(sizeof(space_)), key_(buf_), key_size_(0) {}
+  IterKey() : key_(space_), buf_size_(sizeof(space_)), key_size_(0) {}

  ~IterKey() { ResetBuffer(); }

  Slice GetKey() const { return Slice(key_, key_size_); }

-  Slice GetUserKey() const {
-    assert(key_size_ >= 8);
-    return Slice(key_, key_size_ - 8);
-  }
-
-  size_t Size() const { return key_size_; }
+  size_t Size() { return key_size_; }

  void Clear() { key_size_ = 0; }

@@ -294,66 +283,34 @@ class IterKey {
  void TrimAppend(const size_t shared_len, const char* non_shared_data,
                  const size_t non_shared_len) {
    assert(shared_len <= key_size_);
-    size_t total_size = shared_len + non_shared_len;

-    if (IsKeyPinned() /* key is not in buf_ */) {
-      // Copy the key from external memory to buf_ (copy shared_len bytes)
-      EnlargeBufferIfNeeded(total_size);
-      memcpy(buf_, key_, shared_len);
-    } else if (total_size > buf_size_) {
+    size_t total_size = shared_len + non_shared_len;
+    if (total_size <= buf_size_) {
+      key_size_ = total_size;
+    } else {
      // Need to allocate space, delete previous space
      char* p = new char[total_size];
      memcpy(p, key_, shared_len);

-      if (buf_ != space_) {
-        delete[] buf_;
+      if (key_ != nullptr && key_ != space_) {
+        delete[] key_;
      }

-      buf_ = p;
+      key_ = p;
+      key_size_ = total_size;
      buf_size_ = total_size;
    }

-    memcpy(buf_ + shared_len, non_shared_data, non_shared_len);
-    key_ = buf_;
-    key_size_ = total_size;
+    memcpy(key_ + shared_len, non_shared_data, non_shared_len);
  }

-  Slice SetKey(const Slice& key, bool copy = true) {
+  void SetKey(const Slice& key) {
    size_t size = key.size();
-    if (copy) {
-      // Copy key to buf_
-      EnlargeBufferIfNeeded(size);
-      memcpy(buf_, key.data(), size);
-      key_ = buf_;
-    } else {
-      // Update key_ to point to external memory
-      key_ = key.data();
-    }
+    EnlargeBufferIfNeeded(size);
+    memcpy(key_, key.data(), size);
    key_size_ = size;
-    return Slice(key_, key_size_);
  }

-  // Copies the content of key, updates the reference to the user key in ikey
-  // and returns a Slice referencing the new copy.
-  Slice SetKey(const Slice& key, ParsedInternalKey* ikey) {
-    size_t key_n = key.size();
-    assert(key_n >= 8);
-    SetKey(key);
-    ikey->user_key = Slice(key_, key_n - 8);
-    return Slice(key_, key_n);
-  }
-
-  // Update the sequence number in the internal key.  Guarantees not to
-  // invalidate slices to the key (and the user key).
-  void UpdateInternalKey(uint64_t seq, ValueType t) {
-    assert(!IsKeyPinned());
-    assert(key_size_ >= 8);
-    uint64_t newval = (seq << 8) | t;
-    EncodeFixed64(&buf_[key_size_ - 8], newval);
-  }
-
-  bool IsKeyPinned() const { return (key_ != buf_); }
-
  void SetInternalKey(const Slice& key_prefix, const Slice& user_key,
                      SequenceNumber s,
                      ValueType value_type = kValueTypeForSeek) {
@@ -361,12 +318,10 @@ class IterKey {
    size_t usize = user_key.size();
    EnlargeBufferIfNeeded(psize + usize + sizeof(uint64_t));
    if (psize > 0) {
-      memcpy(buf_, key_prefix.data(), psize);
+      memcpy(key_, key_prefix.data(), psize);
    }
-    memcpy(buf_ + psize, user_key.data(), usize);
-    EncodeFixed64(buf_ + usize + psize, PackSequenceAndType(s, value_type));
-
-    key_ = buf_;
+    memcpy(key_ + psize, user_key.data(), usize);
+    EncodeFixed64(key_ + usize + psize, PackSequenceAndType(s, value_type));
    key_size_ = psize + usize + sizeof(uint64_t);
  }

@@ -393,23 +348,21 @@ class IterKey {
  void EncodeLengthPrefixedKey(const Slice& key) {
    auto size = key.size();
    EnlargeBufferIfNeeded(size + static_cast<size_t>(VarintLength(size)));
-    char* ptr = EncodeVarint32(buf_, static_cast<uint32_t>(size));
+    char* ptr = EncodeVarint32(key_, static_cast<uint32_t>(size));
    memcpy(ptr, key.data(), size);
-    key_ = buf_;
  }

 private:
-  char* buf_;
+  char* key_;
  size_t buf_size_;
-  const char* key_;
  size_t key_size_;
  char space_[32];  // Avoid allocation for short keys

  void ResetBuffer() {
-    if (buf_ != space_) {
-      delete[] buf_;
-      buf_ = space_;
+    if (key_ != nullptr && key_ != space_) {
+      delete[] key_;
    }
+    key_ = space_;
    buf_size_ = sizeof(space_);
    key_size_ = 0;
  }
@@ -425,7 +378,7 @@ class IterKey {
    if (key_size > buf_size_) {
      // Need to enlarge the buffer.
      ResetBuffer();
-      buf_ = new char[key_size];
+      key_ = new char[key_size];
      buf_size_ = key_size;
    }
  }
@@ -74,7 +74,6 @@ class DeleteFileTest : public testing::Test {

  void CloseDB() {
    delete db_;
-    db_ = nullptr;
  }

  void AddKeys(int numkeys, int startkey = 0) {
@@ -8,10 +8,7 @@
 namespace rocksdb {

 namespace {
-template<class T>
-inline T SafeDivide(T a, T b) {
-  return b == 0 ? 0 : a / b;
-}
+inline double SafeDivide(double a, double b) { return b == 0.0 ? 0 : a / b; }
 }  // namespace

 void EventHelpers::AppendCurrentTime(JSONWriter* jwriter) {
@@ -56,7 +53,7 @@ void EventHelpers::LogAndNotifyTableFileCreation(
                info.table_properties.filter_policy_name;

    // user collected properties
-    for (const auto& prop : info.table_properties.readable_properties) {
+    for (const auto& prop : info.table_properties.user_collected_properties) {
      jwriter << prop.first << prop.second;
    }
    jwriter.EndObject();
@@ -11,6 +11,8 @@
 // the last "sync". It then checks for data loss errors by purposely dropping
 // file data (or entire files) not protected by a "sync".

+#if !(defined NDEBUG) || !defined(OS_WIN)
+
 #include <map>
 #include <set>
 #include "db/db_impl.h"
@@ -139,7 +141,6 @@ class TestWritableFile : public WritableFile {
                            FaultInjectionTestEnv* env);
  virtual ~TestWritableFile();
  virtual Status Append(const Slice& data) override;
-  virtual Status Truncate(uint64_t size) override { return target_->Truncate(size); }
  virtual Status Close() override;
  virtual Status Flush() override;
  virtual Status Sync() override;
@@ -645,17 +646,18 @@ class FaultInjectionTest : public testing::Test,
    return test::RandomString(&r, kValueSize, storage);
  }

+  Status OpenDB() {
+    delete db_;
+    db_ = NULL;
+    env_->ResetState();
+    return DB::Open(options_, dbname_, &db_);
+  }
+
  void CloseDB() {
    delete db_;
    db_ = NULL;
  }

-  Status OpenDB() {
-    CloseDB();
-    env_->ResetState();
-    return DB::Open(options_, dbname_, &db_);
-  }
-
  void DeleteAllData() {
    Iterator* iter = db_->NewIterator(ReadOptions());
    WriteOptions options;
@@ -776,14 +778,45 @@ TEST_P(FaultInjectionTest, FaultTest) {
  } while (ChangeOptions());
 }

+class SleepingBackgroundTask {
+ public:
+  SleepingBackgroundTask()
+      : bg_cv_(&mutex_), should_sleep_(true), done_with_sleep_(false) {}
+  void DoSleep() {
+    MutexLock l(&mutex_);
+    while (should_sleep_) {
+      bg_cv_.Wait();
+    }
+    done_with_sleep_ = true;
+    bg_cv_.SignalAll();
+  }
+  void WakeUp() {
+    MutexLock l(&mutex_);
+    should_sleep_ = false;
+    bg_cv_.SignalAll();
+    while (!done_with_sleep_) {
+      bg_cv_.Wait();
+    }
+  }
+
+  static void DoSleepTask(void* arg) {
+    reinterpret_cast<SleepingBackgroundTask*>(arg)->DoSleep();
+  }
+
+ private:
+  port::Mutex mutex_;
+  port::CondVar bg_cv_;  // Signalled when background work finishes
+  bool should_sleep_;
+  bool done_with_sleep_;
+};
+
 // Previous log file is not fsynced if sync is forced after log rolling.
 TEST_P(FaultInjectionTest, WriteOptionSyncTest) {
-  test::SleepingBackgroundTask sleeping_task_low;
+  SleepingBackgroundTask sleeping_task_low;
  env_->SetBackgroundThreads(1, Env::HIGH);
  // Block the job queue to prevent flush job from running.
-  env_->Schedule(&test::SleepingBackgroundTask::DoSleepTask, &sleeping_task_low,
+  env_->Schedule(&SleepingBackgroundTask::DoSleepTask, &sleeping_task_low,
                 Env::Priority::HIGH);
-  sleeping_task_low.WaitUntilSleeping();

  WriteOptions write_options;
  write_options.sync = false;
@@ -862,12 +895,11 @@ TEST_P(FaultInjectionTest, UninstalledCompaction) {
 }

 TEST_P(FaultInjectionTest, ManualLogSyncTest) {
-  test::SleepingBackgroundTask sleeping_task_low;
+  SleepingBackgroundTask sleeping_task_low;
  env_->SetBackgroundThreads(1, Env::HIGH);
  // Block the job queue to prevent flush job from running.
-  env_->Schedule(&test::SleepingBackgroundTask::DoSleepTask, &sleeping_task_low,
+  env_->Schedule(&SleepingBackgroundTask::DoSleepTask, &sleeping_task_low,
                 Env::Priority::HIGH);
-  sleeping_task_low.WaitUntilSleeping();

  WriteOptions write_options;
  write_options.sync = false;
@@ -901,7 +933,13 @@ INSTANTIATE_TEST_CASE_P(FaultTest, FaultInjectionTest, ::testing::Bool());

 }  // namespace rocksdb

+#endif // #if !(defined NDEBUG) || !defined(OS_WIN)
+
 int main(int argc, char** argv) {
+#if !(defined NDEBUG) || !defined(OS_WIN)
  ::testing::InitGoogleTest(&argc, argv);
  return RUN_ALL_TESTS();
+#else
+  return 0;
+#endif
 }
@@ -26,9 +26,9 @@ struct FileLevel;
 // The file tree structure in Version is prebuilt and the range of each file
 // is known. On Version::Get(), it uses binary search to find a potential file
 // and then check if a target key can be found in the file by comparing the key
-// to each file's smallest and largest key. The results of these comparisons
+// to each file's smallest and largest key. The results of these comparisions
 // can be reused beyond checking if a key falls into a file's range.
-// With some pre-calculated knowledge, each key comparison that has been done
+// With some pre-calculated knowledge, each key comparision that has been done
 // can serve as a hint to narrow down further searches: if a key compared to
 // be smaller than a file's smallest or largest, that comparison can be used
 // to find out the right bound of next binary search. Similarly, if a key
@@ -48,7 +48,7 @@ class FileIndexer {
  size_t LevelIndexSize(size_t level) const;

  // Return a file index range in the next level to search for a key based on
-  // smallest and largest key comparison for the current file specified by
+  // smallest and largest key comparision for the current file specified by
  // level and file_index. When *left_index < *right_index, both index should
  // be valid and fit in the vector size.
  void GetNextLevelIndex(const size_t level, const size_t file_index,
@@ -21,14 +21,9 @@
 #include "util/file_reader_writer.h"
 #include "util/logging.h"
 #include "util/stop_watch.h"
-#include "util/string_util.h"
-#include "util/sync_point.h"

 namespace rocksdb {

-static const std::string kRocksDbTFileExt = "sst";
-static const std::string kLevelDbTFileExt = "ldb";
-
 // Given a path, flatten the path name by replacing all chars not in
 // {[0-9,a-z,A-Z,-,_,.]} with _. And append '_LOG\0' at the end.
 // Return the number of chars stored in dest not including the trailing '\0'.
@@ -48,9 +43,8 @@ static size_t GetInfoLogPrefix(const std::string& path, char* dest, int len) {
        path[i] == '_'){
      dest[write_idx++] = path[i];
    } else {
-      if (i > 0) {
+      if (i > 0)
        dest[write_idx++] = '_';
-      }
    }
    i++;
  }
@@ -84,16 +78,7 @@ std::string ArchivedLogFileName(const std::string& name, uint64_t number) {
 }

 std::string MakeTableFileName(const std::string& path, uint64_t number) {
-  return MakeFileName(path, number, kRocksDbTFileExt.c_str());
-}
-
-std::string Rocks2LevelTableFileName(const std::string& fullname) {
-  assert(fullname.size() > kRocksDbTFileExt.size() + 1);
-  if (fullname.size() <= kRocksDbTFileExt.size() + 1) {
-    return "";
-  }
-  return fullname.substr(0, fullname.size() - kRocksDbTFileExt.size()) +
-         kLevelDbTFileExt;
+  return MakeFileName(path, number, "sst");
 }

 uint64_t TableFileNameToNumber(const std::string& name) {
@@ -148,7 +133,7 @@ std::string LockFileName(const std::string& dbname) {
 }

 std::string TempFileName(const std::string& dbname, uint64_t number) {
-  return MakeFileName(dbname, number, kTempFileNameSuffix.c_str());
+  return MakeFileName(dbname, number, "dbtmp");
 }

 InfoLogPrefix::InfoLogPrefix(bool has_log_dir,
@@ -166,9 +151,8 @@ InfoLogPrefix::InfoLogPrefix(bool has_log_dir,

 std::string InfoLogFileName(const std::string& dbname,
    const std::string& db_path, const std::string& log_dir) {
-  if (log_dir.empty()) {
+  if (log_dir.empty())
    return dbname + "/LOG";
-  }

  InfoLogPrefix info_log_prefix(true, db_path);
  return log_dir + "/" + info_log_prefix.buf;
@@ -180,29 +164,13 @@ std::string OldInfoLogFileName(const std::string& dbname, uint64_t ts,
  char buf[50];
  snprintf(buf, sizeof(buf), "%llu", static_cast<unsigned long long>(ts));

-  if (log_dir.empty()) {
+  if (log_dir.empty())
    return dbname + "/LOG.old." + buf;
-  }

  InfoLogPrefix info_log_prefix(true, db_path);
  return log_dir + "/" + info_log_prefix.buf + ".old." + buf;
 }

-std::string OptionsFileName(const std::string& dbname, uint64_t file_num) {
-  char buffer[256];
-  snprintf(buffer, sizeof(buffer), "%s%06" PRIu64,
-           kOptionsFileNamePrefix.c_str(), file_num);
-  return dbname + "/" + buffer;
-}
-
-std::string TempOptionsFileName(const std::string& dbname, uint64_t file_num) {
-  char buffer[256];
-  snprintf(buffer, sizeof(buffer), "%s%06" PRIu64 ".%s",
-           kOptionsFileNamePrefix.c_str(), file_num,
-           kTempFileNameSuffix.c_str());
-  return dbname + "/" + buffer;
-}
-
 std::string MetaDatabaseName(const std::string& dbname, uint64_t number) {
  char buf[100];
  snprintf(buf, sizeof(buf), "/METADB-%llu",
@@ -223,8 +191,6 @@ std::string IdentityFileName(const std::string& dbname) {
 //    dbname/MANIFEST-[0-9]+
 //    dbname/[0-9]+.(log|sst)
 //    dbname/METADB-[0-9]+
-//    dbname/OPTIONS-[0-9]+
-//    dbname/OPTIONS-[0-9]+.dbtmp
 //    Disregards / at the beginning
 bool ParseFileName(const std::string& fname,
                   uint64_t* number,
@@ -287,21 +253,6 @@ bool ParseFileName(const std::string& fname, uint64_t* number,
    }
    *type = kMetaDatabase;
    *number = num;
-  } else if (rest.starts_with(kOptionsFileNamePrefix)) {
-    uint64_t ts_suffix;
-    bool is_temp_file = false;
-    rest.remove_prefix(kOptionsFileNamePrefix.size());
-    const std::string kTempFileNameSuffixWithDot =
-        std::string(".") + kTempFileNameSuffix;
-    if (rest.ends_with(kTempFileNameSuffixWithDot)) {
-      rest.remove_suffix(kTempFileNameSuffixWithDot.size());
-      is_temp_file = true;
-    }
-    if (!ConsumeDecimalNumber(&rest, &ts_suffix)) {
-      return false;
-    }
-    *number = ts_suffix;
-    *type = is_temp_file ? kTempFile : kOptionsFile;
  } else {
    // Avoid strtoull() to keep filename format independent of the
    // current locale
@@ -320,23 +271,17 @@ bool ParseFileName(const std::string& fname, uint64_t* number,
    if (!ConsumeDecimalNumber(&rest, &num)) {
      return false;
    }
-    if (rest.size() <= 1 || rest[0] != '.') {
-      return false;
-    }
-    rest.remove_prefix(1);
-
    Slice suffix = rest;
-    if (suffix == Slice("log")) {
+    if (suffix == Slice(".log")) {
      *type = kLogFile;
      if (log_type && !archive_dir_found) {
        *log_type = kAliveLogFile;
      }
    } else if (archive_dir_found) {
      return false; // Archive dir can contain only log files
-    } else if (suffix == Slice(kRocksDbTFileExt) ||
-               suffix == Slice(kLevelDbTFileExt)) {
+    } else if (suffix == Slice(".sst")) {
      *type = kTableFile;
-    } else if (suffix == Slice(kTempFileNameSuffix)) {
+    } else if (suffix == Slice(".dbtmp")) {
      *type = kTempFile;
    } else {
      return false;
@@ -357,9 +302,7 @@ Status SetCurrentFile(Env* env, const std::string& dbname,
  std::string tmp = TempFileName(dbname, descriptor_number);
  Status s = WriteStringToFile(env, contents.ToString() + "\n", tmp, true);
  if (s.ok()) {
-    TEST_KILL_RANDOM("SetCurrentFile:0", rocksdb_kill_odds * REDUCE_ODDS2);
    s = env->RenameFile(tmp, CurrentFileName(dbname));
-    TEST_KILL_RANDOM("SetCurrentFile:1", rocksdb_kill_odds * REDUCE_ODDS2);
  }
  if (s.ok()) {
    if (directory_to_fsync != nullptr) {
@@ -388,7 +331,6 @@ Status SetIdentityFile(Env* env, const std::string& dbname) {

 Status SyncManifest(Env* env, const DBOptions* db_options,
                    WritableFileWriter* file) {
-  TEST_KILL_RANDOM("SyncManifest:0", rocksdb_kill_odds * REDUCE_ODDS2);
  if (db_options->disableDataSync) {
    return Status::OK();
  } else {
@@ -36,8 +36,7 @@ enum FileType {
  kTempFile,
  kInfoLogFile,  // Either the current one, or an old one
  kMetaDatabase,
-  kIdentityFile,
-  kOptionsFile
+  kIdentityFile
 };

 // Return the name of the log file with the specified number
@@ -56,10 +55,6 @@ extern std::string ArchivedLogFileName(const std::string& dbname,

 extern std::string MakeTableFileName(const std::string& name, uint64_t number);

-// Return the name of sstable with LevelDB suffix
-// created from RocksDB sstable suffixed name
-extern std::string Rocks2LevelTableFileName(const std::string& fullname);
-
 // the reverse function of MakeTableFileName
 // TODO(yhchiang): could merge this function with ParseFileName()
 extern uint64_t TableFileNameToNumber(const std::string& name);
@@ -107,26 +102,11 @@ struct InfoLogPrefix {

 // Return the name of the info log file for "dbname".
 extern std::string InfoLogFileName(const std::string& dbname,
-                                   const std::string& db_path = "",
-                                   const std::string& log_dir = "");
+    const std::string& db_path="", const std::string& log_dir="");

 // Return the name of the old info log file for "dbname".
 extern std::string OldInfoLogFileName(const std::string& dbname, uint64_t ts,
-                                      const std::string& db_path = "",
-                                      const std::string& log_dir = "");
-
-static const std::string kOptionsFileNamePrefix = "OPTIONS-";
-static const std::string kTempFileNameSuffix = "dbtmp";
-
-// Return a options file name given the "dbname" and file number.
-// Format:  OPTIONS-[number].dbtmp
-extern std::string OptionsFileName(const std::string& dbname,
-                                   uint64_t file_num);
-
-// Return a temp options file name given the "dbname" and file number.
-// Format:  OPTIONS-[number]
-extern std::string TempOptionsFileName(const std::string& dbname,
-                                       uint64_t file_num);
+    const std::string& db_path="", const std::string& log_dir="");

 // Return the name to use for a metadatabase. The result will be prefixed with
 // "dbname".
@@ -62,7 +62,6 @@ FlushJob::FlushJob(const std::string& dbname, ColumnFamilyData* cfd,
                   InstrumentedMutex* db_mutex,
                   std::atomic<bool>* shutting_down,
                   std::vector<SequenceNumber> existing_snapshots,
-                   SequenceNumber earliest_write_conflict_snapshot,
                   JobContext* job_context, LogBuffer* log_buffer,
                   Directory* db_directory, Directory* output_file_directory,
                   CompressionType output_compression, Statistics* stats,
@@ -76,7 +75,6 @@ FlushJob::FlushJob(const std::string& dbname, ColumnFamilyData* cfd,
      db_mutex_(db_mutex),
      shutting_down_(shutting_down),
      existing_snapshots_(std::move(existing_snapshots)),
-      earliest_write_conflict_snapshot_(earliest_write_conflict_snapshot),
      job_context_(job_context),
      log_buffer_(log_buffer),
      db_directory_(db_directory),
@@ -199,7 +197,7 @@ Status FlushJob::WriteLevel0Table(const autovector<MemTable*>& mems,
    if (log_buffer_) {
      log_buffer_->FlushBufferToLog();
    }
-    std::vector<InternalIterator*> memtables;
+    std::vector<Iterator*> memtables;
    ReadOptions ro;
    ro.total_order_seek = true;
    Arena arena;
@@ -233,15 +231,13 @@ Status FlushJob::WriteLevel0Table(const autovector<MemTable*>& mems,

      TEST_SYNC_POINT_CALLBACK("FlushJob::WriteLevel0Table:output_compression",
                               &output_compression_);
-      s = BuildTable(dbname_, db_options_.env, *cfd_->ioptions(), env_options_,
-                     cfd_->table_cache(), iter.get(), meta,
-                     cfd_->internal_comparator(),
-                     cfd_->int_tbl_prop_collector_factories(), cfd_->GetID(),
-                     existing_snapshots_, earliest_write_conflict_snapshot_,
-                     output_compression_, cfd_->ioptions()->compression_opts,
-                     mutable_cf_options_.paranoid_file_checks,
-                     cfd_->internal_stats(), Env::IO_HIGH, &table_properties_);
-      info.table_properties = table_properties_;
+      s = BuildTable(
+          dbname_, db_options_.env, *cfd_->ioptions(), env_options_,
+          cfd_->table_cache(), iter.get(), meta, cfd_->internal_comparator(),
+          cfd_->int_tbl_prop_collector_factories(), existing_snapshots_,
+          output_compression_, cfd_->ioptions()->compression_opts,
+          mutable_cf_options_.paranoid_file_checks, cfd_->internal_stats(),
+          Env::IO_HIGH, &info.table_properties);
      LogFlush(db_options_.info_log);
    }
    Log(InfoLogLevel::INFO_LEVEL, db_options_.info_log,
@@ -27,12 +27,12 @@
 #include "rocksdb/env.h"
 #include "rocksdb/memtablerep.h"
 #include "rocksdb/transaction_log.h"
-#include "table/scoped_arena_iterator.h"
 #include "util/autovector.h"
 #include "util/event_logger.h"
 #include "util/instrumented_mutex.h"
 #include "util/stop_watch.h"
 #include "util/thread_local.h"
+#include "util/scoped_arena_iterator.h"
 #include "db/internal_stats.h"
 #include "db/write_controller.h"
 #include "db/flush_scheduler.h"
@@ -58,7 +58,6 @@ class FlushJob {
           const EnvOptions& env_options, VersionSet* versions,
           InstrumentedMutex* db_mutex, std::atomic<bool>* shutting_down,
           std::vector<SequenceNumber> existing_snapshots,
-           SequenceNumber earliest_write_conflict_snapshot,
           JobContext* job_context, LogBuffer* log_buffer,
           Directory* db_directory, Directory* output_file_directory,
           CompressionType output_compression, Statistics* stats,
@@ -67,7 +66,6 @@ class FlushJob {
  ~FlushJob();

  Status Run(FileMetaData* file_meta = nullptr);
-  TableProperties GetTableProperties() const { return table_properties_; }

 private:
  void ReportStartedFlush();
@@ -84,7 +82,6 @@ class FlushJob {
  InstrumentedMutex* db_mutex_;
  std::atomic<bool>* shutting_down_;
  std::vector<SequenceNumber> existing_snapshots_;
-  SequenceNumber earliest_write_conflict_snapshot_;
  JobContext* job_context_;
  LogBuffer* log_buffer_;
  Directory* db_directory_;
@@ -92,7 +89,6 @@ class FlushJob {
  CompressionType output_compression_;
  Statistics* stats_;
  EventLogger* event_logger_;
-  TableProperties table_properties_;
 };

 }  // namespace rocksdb
@@ -61,7 +61,7 @@ class FlushJobTest : public testing::Test {
    unique_ptr<WritableFileWriter> file_writer(
        new WritableFileWriter(std::move(file), EnvOptions()));
    {
-      log::Writer log(std::move(file_writer), 0, false);
+      log::Writer log(std::move(file_writer));
      std::string record;
      new_db.EncodeTo(&record);
      s = log.AddRecord(record);
@@ -92,8 +92,8 @@ TEST_F(FlushJobTest, Empty) {
  FlushJob flush_job(dbname_, versions_->GetColumnFamilySet()->GetDefault(),
                     db_options_, *cfd->GetLatestMutableCFOptions(),
                     env_options_, versions_.get(), &mutex_, &shutting_down_,
-                     {}, kMaxSequenceNumber, &job_context, nullptr, nullptr,
-                     nullptr, kNoCompression, nullptr, &event_logger);
+                     {}, &job_context, nullptr, nullptr, nullptr,
+                     kNoCompression, nullptr, &event_logger);
  ASSERT_OK(flush_job.Run());
  job_context.Clean();
 }
@@ -131,8 +131,8 @@ TEST_F(FlushJobTest, NonEmpty) {
  FlushJob flush_job(dbname_, versions_->GetColumnFamilySet()->GetDefault(),
                     db_options_, *cfd->GetLatestMutableCFOptions(),
                     env_options_, versions_.get(), &mutex_, &shutting_down_,
-                     {}, kMaxSequenceNumber, &job_context, nullptr, nullptr,
-                     nullptr, kNoCompression, nullptr, &event_logger);
+                     {}, &job_context, nullptr, nullptr, nullptr,
+                     kNoCompression, nullptr, &event_logger);
  FileMetaData fd;
  mutex_.Lock();
  ASSERT_OK(flush_job.Run(&fd));
@@ -195,8 +195,8 @@ TEST_F(FlushJobTest, Snapshots) {
  FlushJob flush_job(dbname_, versions_->GetColumnFamilySet()->GetDefault(),
                     db_options_, *cfd->GetLatestMutableCFOptions(),
                     env_options_, versions_.get(), &mutex_, &shutting_down_,
-                     snapshots, kMaxSequenceNumber, &job_context, nullptr,
-                     nullptr, nullptr, kNoCompression, nullptr, &event_logger);
+                     snapshots, &job_context, nullptr, nullptr, nullptr,
+                     kNoCompression, nullptr, &event_logger);
  mutex_.Lock();
  ASSERT_OK(flush_job.Run());
  mutex_.Unlock();
@@ -13,69 +13,51 @@ namespace rocksdb {

 void FlushScheduler::ScheduleFlush(ColumnFamilyData* cfd) {
 #ifndef NDEBUG
-  {
-    std::lock_guard<std::mutex> lock(checking_mutex_);
-    assert(checking_set_.count(cfd) == 0);
-    checking_set_.insert(cfd);
-  }
+  assert(column_families_set_.find(cfd) == column_families_set_.end());
+  column_families_set_.insert(cfd);
 #endif  // NDEBUG
  cfd->Ref();
-  Node* node = new Node{cfd, head_.load(std::memory_order_relaxed)};
-  while (!head_.compare_exchange_strong(
-      node->next, node, std::memory_order_relaxed, std::memory_order_relaxed)) {
-    // failing CAS updates the first param, so we are already set for
-    // retry.  TakeNextColumnFamily won't happen until after another
-    // inter-thread synchronization, so we don't even need release
-    // semantics for this CAS
-  }
+  column_families_.push_back(cfd);
 }

-ColumnFamilyData* FlushScheduler::TakeNextColumnFamily() {
-  while (true) {
-    if (Empty()) {
-      return nullptr;
+ColumnFamilyData* FlushScheduler::GetNextColumnFamily() {
+  ColumnFamilyData* cfd = nullptr;
+  while (column_families_.size() > 0) {
+    cfd = column_families_.front();
+    column_families_.pop_front();
+    if (cfd->IsDropped()) {
+      if (cfd->Unref()) {
+        delete cfd;
+        cfd = nullptr;
+      }
+    } else {
+      break;
    }
-
-    // dequeue the head
-    Node* node = head_.load(std::memory_order_relaxed);
-    head_.store(node->next, std::memory_order_relaxed);
-    ColumnFamilyData* cfd = node->column_family;
-    delete node;
-
+  }
 #ifndef NDEBUG
-    {
-      auto iter = checking_set_.find(cfd);
-      assert(iter != checking_set_.end());
-      checking_set_.erase(iter);
-    }
-#endif  // NDEBUG
-
-    if (!cfd->IsDropped()) {
-      // success
-      return cfd;
-    }
-
-    // no longer relevant, retry
-    if (cfd->Unref()) {
-      delete cfd;
-    }
+  if (cfd != nullptr) {
+    auto itr = column_families_set_.find(cfd);
+    assert(itr != column_families_set_.end());
+    column_families_set_.erase(itr);
  }
+#endif  // NDEBUG
+  return cfd;
 }

-bool FlushScheduler::Empty() {
-  auto rv = head_.load(std::memory_order_relaxed) == nullptr;
-  assert(rv == checking_set_.empty());
-  return rv;
-}
+bool FlushScheduler::Empty() { return column_families_.empty(); }

 void FlushScheduler::Clear() {
-  ColumnFamilyData* cfd;
-  while ((cfd = TakeNextColumnFamily()) != nullptr) {
+  for (auto cfd : column_families_) {
+#ifndef NDEBUG
+    auto itr = column_families_set_.find(cfd);
+    assert(itr != column_families_set_.end());
+    column_families_set_.erase(itr);
+#endif  // NDEBUG
    if (cfd->Unref()) {
      delete cfd;
    }
  }
-  assert(Empty());
+  column_families_.clear();
 }

 }  // namespace rocksdb
@@ -6,42 +6,34 @@
 #pragma once

 #include <stdint.h>
-#include <atomic>
-#include <mutex>
+#include <deque>
 #include <set>
+#include <vector>

 namespace rocksdb {

 class ColumnFamilyData;

-// Unless otherwise noted, all methods on FlushScheduler should be called
-// only with the DB mutex held or from a single-threaded recovery context.
+// This class is thread-compatible. It's should only be accessed from single
+// write thread (between BeginWrite() and EndWrite())
 class FlushScheduler {
 public:
-  FlushScheduler() : head_(nullptr) {}
+  FlushScheduler() = default;
+  ~FlushScheduler() = default;

-  // May be called from multiple threads at once, but not concurrent with
-  // any other method calls on this instance
  void ScheduleFlush(ColumnFamilyData* cfd);
-
-  // Removes and returns Ref()-ed column family. Client needs to Unref().
-  // Filters column families that have been dropped.
-  ColumnFamilyData* TakeNextColumnFamily();
+  // Returns Ref()-ed column family. Client needs to Unref()
+  // REQUIRES: db mutex is held (exception is single-threaded recovery)
+  ColumnFamilyData* GetNextColumnFamily();

  bool Empty();

  void Clear();

 private:
-  struct Node {
-    ColumnFamilyData* column_family;
-    Node* next;
-  };
-
-  std::atomic<Node*> head_;
+  std::deque<ColumnFamilyData*> column_families_;
 #ifndef NDEBUG
-  std::mutex checking_mutex_;
-  std::set<ColumnFamilyData*> checking_set_;
+  std::set<ColumnFamilyData*> column_families_set_;
 #endif  // NDEBUG
 };

@@ -28,7 +28,7 @@ namespace rocksdb {
 //     iter.SetFileIndex(file_index);
 //     iter.Seek(target);
 //     iter.Next()
-class LevelIterator : public InternalIterator {
+class LevelIterator : public Iterator {
 public:
  LevelIterator(const ColumnFamilyData* const cfd,
      const ReadOptions& read_options,
@@ -113,7 +113,7 @@ class LevelIterator : public InternalIterator {
  bool valid_;
  uint32_t file_index_;
  Status status_;
-  std::unique_ptr<InternalIterator> file_iter_;
+  std::unique_ptr<Iterator> file_iter_;
 };

 ForwardIterator::ForwardIterator(DBImpl* db, const ReadOptions& read_options,
@@ -144,29 +144,12 @@ ForwardIterator::~ForwardIterator() {
  Cleanup(true);
 }

-void ForwardIterator::SVCleanup() {
-  if (sv_ != nullptr && sv_->Unref()) {
-    // Job id == 0 means that this is not our background process, but rather
-    // user thread
-    JobContext job_context(0);
-    db_->mutex_.Lock();
-    sv_->Cleanup();
-    db_->FindObsoleteFiles(&job_context, false, true);
-    db_->mutex_.Unlock();
-    delete sv_;
-    if (job_context.HaveSomethingToDelete()) {
-      db_->PurgeObsoleteFiles(job_context);
-    }
-    job_context.Clean();
-  }
-}
-
 void ForwardIterator::Cleanup(bool release_sv) {
  if (mutable_iter_ != nullptr) {
-    mutable_iter_->~InternalIterator();
+    mutable_iter_->~Iterator();
  }
  for (auto* m : imm_iters_) {
-    m->~InternalIterator();
+    m->~Iterator();
  }
  imm_iters_.clear();
  for (auto* f : l0_iters_) {
@@ -179,7 +162,20 @@ void ForwardIterator::Cleanup(bool release_sv) {
  level_iters_.clear();

  if (release_sv) {
-    SVCleanup();
+    if (sv_ != nullptr && sv_->Unref()) {
+      // Job id == 0 means that this is not our background process, but rather
+      // user thread
+      JobContext job_context(0);
+      db_->mutex_.Lock();
+      sv_->Cleanup();
+      db_->FindObsoleteFiles(&job_context, false, true);
+      db_->mutex_.Unlock();
+      delete sv_;
+      if (job_context.HaveSomethingToDelete()) {
+        db_->PurgeObsoleteFiles(job_context);
+      }
+      job_context.Clean();
+    }
  }
 }

@@ -189,10 +185,9 @@ bool ForwardIterator::Valid() const {
 }

 void ForwardIterator::SeekToFirst() {
-  if (sv_ == nullptr) {
+  if (sv_ == nullptr ||
+      sv_ ->version_number != cfd_->GetSuperVersionNumber()) {
    RebuildIterators(true);
-  } else if (sv_->version_number != cfd_->GetSuperVersionNumber()) {
-    RenewIterators();
  } else if (immutable_status_.IsIncomplete()) {
    ResetIncompleteIterators();
  }
@@ -210,10 +205,9 @@ void ForwardIterator::Seek(const Slice& internal_key) {
  if (IsOverUpperBound(internal_key)) {
    valid_ = false;
  }
-  if (sv_ == nullptr) {
+  if (sv_ == nullptr ||
+      sv_ ->version_number != cfd_->GetSuperVersionNumber()) {
    RebuildIterators(true);
-  } else if (sv_->version_number != cfd_->GetSuperVersionNumber()) {
-    RenewIterators();
  } else if (immutable_status_.IsIncomplete()) {
    ResetIncompleteIterators();
  }
@@ -233,9 +227,7 @@ void ForwardIterator::SeekInternal(const Slice& internal_key,
  // an option to turn it off.
  if (seek_to_first || NeedToSeekImmutable(internal_key)) {
    immutable_status_ = Status::OK();
-    if ((has_iter_trimmed_for_upper_bound_) &&
-        (cfd_->internal_comparator().InternalKeyComparator::Compare(
-             prev_key_.GetKey(), internal_key) > 0)) {
+    if (has_iter_trimmed_for_upper_bound_) {
      // Some iterators are trimmed. Need to rebuild.
      RebuildIterators(true);
      // Already seeked mutable iter, so seek again
@@ -262,7 +254,7 @@ void ForwardIterator::SeekInternal(const Slice& internal_key,
    }
    const VersionStorageInfo* vstorage = sv_->current->storage_info();
    const std::vector<FileMetaData*>& l0 = vstorage->LevelFiles(0);
-    for (size_t i = 0; i < l0.size(); ++i) {
+    for (uint32_t i = 0; i < l0.size(); ++i) {
      if (!l0_iters_[i]) {
        continue;
      }
@@ -401,11 +393,7 @@ void ForwardIterator::Next() {
    std::string current_key = key().ToString();
    Slice old_key(current_key.data(), current_key.size());

-    if (sv_ == nullptr) {
-      RebuildIterators(true);
-    } else {
-      RenewIterators();
-    }
+    RebuildIterators(true);
    SeekInternal(old_key, false);
    if (!valid_ || key().compare(old_key) != 0) {
      return;
@@ -496,81 +484,10 @@ void ForwardIterator::RebuildIterators(bool refresh_sv) {
    l0_iters_.push_back(cfd_->table_cache()->NewIterator(
        read_options_, *cfd_->soptions(), cfd_->internal_comparator(), l0->fd));
  }
-  BuildLevelIterators(vstorage);
-  current_ = nullptr;
-  is_prev_set_ = false;
-}
-
-void ForwardIterator::RenewIterators() {
-  SuperVersion* svnew;
-  assert(sv_);
-  svnew = cfd_->GetReferencedSuperVersion(&(db_->mutex_));
-
-  if (mutable_iter_ != nullptr) {
-    mutable_iter_->~InternalIterator();
-  }
-  for (auto* m : imm_iters_) {
-    m->~InternalIterator();
-  }
-  imm_iters_.clear();
-
-  mutable_iter_ = svnew->mem->NewIterator(read_options_, &arena_);
-  svnew->imm->AddIterators(read_options_, &imm_iters_, &arena_);
-
-  const auto* vstorage = sv_->current->storage_info();
-  const auto& l0_files = vstorage->LevelFiles(0);
-  const auto* vstorage_new = svnew->current->storage_info();
-  const auto& l0_files_new = vstorage_new->LevelFiles(0);
-  size_t iold, inew;
-  bool found;
-  std::vector<InternalIterator*> l0_iters_new;
-  l0_iters_new.reserve(l0_files_new.size());
-
-  for (inew = 0; inew < l0_files_new.size(); inew++) {
-    found = false;
-    for (iold = 0; iold < l0_files.size(); iold++) {
-      if (l0_files[iold] == l0_files_new[inew]) {
-        found = true;
-        break;
-      }
-    }
-    if (found) {
-      if (l0_iters_[iold] == nullptr) {
-        l0_iters_new.push_back(nullptr);
-        TEST_SYNC_POINT_CALLBACK("ForwardIterator::RenewIterators:Null", this);
-      } else {
-        l0_iters_new.push_back(l0_iters_[iold]);
-        l0_iters_[iold] = nullptr;
-        TEST_SYNC_POINT_CALLBACK("ForwardIterator::RenewIterators:Copy", this);
-      }
-      continue;
-    }
-    l0_iters_new.push_back(cfd_->table_cache()->NewIterator(
-        read_options_, *cfd_->soptions(), cfd_->internal_comparator(),
-        l0_files_new[inew]->fd));
-  }
-
-  for (auto* f : l0_iters_) {
-    delete f;
-  }
-  l0_iters_.clear();
-  l0_iters_ = l0_iters_new;
-
-  for (auto* l : level_iters_) {
-    delete l;
-  }
-  level_iters_.clear();
-  BuildLevelIterators(vstorage_new);
-  current_ = nullptr;
-  is_prev_set_ = false;
-  SVCleanup();
-  sv_ = svnew;
-}
-
-void ForwardIterator::BuildLevelIterators(const VersionStorageInfo* vstorage) {
  level_iters_.reserve(vstorage->num_levels() - 1);
  for (int32_t level = 1; level < vstorage->num_levels(); ++level) {
    const auto& level_files = vstorage->LevelFiles(level);
+
    if ((level_files.empty()) ||
        ((read_options_.iterate_upper_bound != nullptr) &&
         (user_comparator_->Compare(*read_options_.iterate_upper_bound,
@@ -585,11 +502,14 @@ void ForwardIterator::BuildLevelIterators(const VersionStorageInfo* vstorage) {
          new LevelIterator(cfd_, read_options_, level_files));
    }
  }
+
+  current_ = nullptr;
+  is_prev_set_ = false;
 }

 void ForwardIterator::ResetIncompleteIterators() {
  const auto& l0_files = sv_->current->storage_info()->LevelFiles(0);
-  for (size_t i = 0; i < l0_iters_.size(); ++i) {
+  for (uint32_t i = 0; i < l0_iters_.size(); ++i) {
    assert(i < l0_files.size());
    if (!l0_iters_[i] || !l0_iters_[i]->status().IsIncomplete()) {
      continue;
@@ -680,7 +600,7 @@ bool ForwardIterator::NeedToSeekImmutable(const Slice& target) {
 void ForwardIterator::DeleteCurrentIter() {
  const VersionStorageInfo* vstorage = sv_->current->storage_info();
  const std::vector<FileMetaData*>& l0 = vstorage->LevelFiles(0);
-  for (size_t i = 0; i < l0.size(); ++i) {
+  for (uint32_t i = 0; i < l0.size(); ++i) {
    if (!l0_iters_[i]) {
      continue;
    }
@@ -712,7 +632,7 @@ bool ForwardIterator::TEST_CheckDeletedIters(int* pdeleted_iters,

  const VersionStorageInfo* vstorage = sv_->current->storage_info();
  const std::vector<FileMetaData*>& l0 = vstorage->LevelFiles(0);
-  for (size_t i = 0; i < l0.size(); ++i) {
+  for (uint32_t i = 0; i < l0.size(); ++i) {
    if (!l0_iters_[i]) {
      retval = true;
      deleted_iters++;
@@ -14,7 +14,6 @@
 #include "rocksdb/iterator.h"
 #include "rocksdb/options.h"
 #include "db/dbformat.h"
-#include "table/internal_iterator.h"
 #include "util/arena.h"

 namespace rocksdb {
@@ -24,7 +23,6 @@ class Env;
 struct SuperVersion;
 class ColumnFamilyData;
 class LevelIterator;
-class VersionStorageInfo;
 struct FileMetaData;

 class MinIterComparator {
@@ -32,15 +30,16 @@ class MinIterComparator {
  explicit MinIterComparator(const Comparator* comparator) :
    comparator_(comparator) {}

-  bool operator()(InternalIterator* a, InternalIterator* b) {
+  bool operator()(Iterator* a, Iterator* b) {
    return comparator_->Compare(a->key(), b->key()) > 0;
  }
 private:
  const Comparator* comparator_;
 };

-typedef std::priority_queue<InternalIterator*, std::vector<InternalIterator*>,
-                            MinIterComparator> MinIterHeap;
+typedef std::priority_queue<Iterator*,
+          std::vector<Iterator*>,
+          MinIterComparator> MinIterHeap;

 /**
 * ForwardIterator is a special type of iterator that only supports Seek()
@@ -49,7 +48,7 @@ typedef std::priority_queue<InternalIterator*, std::vector<InternalIterator*>,
 * the iterator. At the current implementation, snapshot is taken at the
 * time Seek() is called. The Next() followed do not see new values after.
 */
-class ForwardIterator : public InternalIterator {
+class ForwardIterator : public Iterator {
 public:
  ForwardIterator(DBImpl* db, const ReadOptions& read_options,
                  ColumnFamilyData* cfd, SuperVersion* current_sv = nullptr);
@@ -75,10 +74,7 @@ class ForwardIterator : public InternalIterator {

 private:
  void Cleanup(bool release_sv);
-  void SVCleanup();
  void RebuildIterators(bool refresh_sv);
-  void RenewIterators();
-  void BuildLevelIterators(const VersionStorageInfo* vstorage);
  void ResetIncompleteIterators();
  void SeekInternal(const Slice& internal_key, bool seek_to_first);
  void UpdateCurrent();
@@ -98,11 +94,11 @@ class ForwardIterator : public InternalIterator {
  MinIterHeap immutable_min_heap_;

  SuperVersion* sv_;
-  InternalIterator* mutable_iter_;
-  std::vector<InternalIterator*> imm_iters_;
-  std::vector<InternalIterator*> l0_iters_;
+  Iterator* mutable_iter_;
+  std::vector<Iterator*> imm_iters_;
+  std::vector<Iterator*> l0_iters_;
  std::vector<LevelIterator*> level_iters_;
-  InternalIterator* current_;
+  Iterator* current_;
  bool valid_;

  // Internal iterator status; set only by one of the unsupported methods.
@@ -1,374 +0,0 @@
-//  Copyright (c) 2013, Facebook, Inc.  All rights reserved.
-//  This source code is licensed under the BSD-style license found in the
-//  LICENSE file in the root directory of this source tree. An additional grant
-//  of patent rights can be found in the PATENTS file in the same directory.
-
-#ifndef __STDC_FORMAT_MACROS
-#define __STDC_FORMAT_MACROS
-#endif
-
-#if !defined(GFLAGS) || defined(ROCKSDB_LITE)
-#include <cstdio>
-int main() {
-  fprintf(stderr, "Please install gflags to run rocksdb tools\n");
-  return 1;
-}
-#elif defined(OS_MACOSX) || defined(OS_WIN)
-// Block forward_iterator_bench under MAC and Windows
-int main() { return 0; }
-#else
-#include <gflags/gflags.h>
-#include <semaphore.h>
-#include <atomic>
-#include <bitset>
-#include <chrono>
-#include <climits>
-#include <condition_variable>
-#include <limits>
-#include <mutex>
-#include <queue>
-#include <random>
-#include <thread>
-
-#include "rocksdb/cache.h"
-#include "rocksdb/db.h"
-#include "rocksdb/status.h"
-#include "rocksdb/table.h"
-#include "util/testharness.h"
-
-const int MAX_SHARDS = 100000;
-
-DEFINE_int32(writers, 8, "");
-DEFINE_int32(readers, 8, "");
-DEFINE_int64(rate, 100000, "");
-DEFINE_int64(value_size, 300, "");
-DEFINE_int64(shards, 1000, "");
-DEFINE_int64(memtable_size, 500000000, "");
-DEFINE_int64(block_cache_size, 300000000, "");
-DEFINE_int64(block_size, 65536, "");
-DEFINE_double(runtime, 300.0, "");
-DEFINE_bool(cache_only_first, true, "");
-DEFINE_bool(iterate_upper_bound, true, "");
-
-struct Stats {
-  char pad1[128] __attribute__((__unused__));
-  std::atomic<uint64_t> written{0};
-  char pad2[128] __attribute__((__unused__));
-  std::atomic<uint64_t> read{0};
-  std::atomic<uint64_t> cache_misses{0};
-  char pad3[128] __attribute__((__unused__));
-} stats;
-
-struct Key {
-  Key() {}
-  Key(uint64_t shard_in, uint64_t seqno_in)
-      : shard_be(htobe64(shard_in)), seqno_be(htobe64(seqno_in)) {}
-
-  uint64_t shard() const { return be64toh(shard_be); }
-  uint64_t seqno() const { return be64toh(seqno_be); }
-
- private:
-  uint64_t shard_be;
-  uint64_t seqno_be;
-} __attribute__((__packed__));
-
-struct Reader;
-struct Writer;
-
-struct ShardState {
-  char pad1[128] __attribute__((__unused__));
-  std::atomic<uint64_t> last_written{0};
-  Writer* writer;
-  Reader* reader;
-  char pad2[128] __attribute__((__unused__));
-  std::atomic<uint64_t> last_read{0};
-  std::unique_ptr<rocksdb::Iterator> it;
-  std::unique_ptr<rocksdb::Iterator> it_cacheonly;
-  Key upper_bound;
-  rocksdb::Slice upper_bound_slice;
-  char pad3[128] __attribute__((__unused__));
-};
-
-struct Reader {
- public:
-  explicit Reader(std::vector<ShardState>* shard_states, rocksdb::DB* db)
-      : shard_states_(shard_states), db_(db) {
-    sem_init(&sem_, 0, 0);
-    thread_ = std::thread(&Reader::run, this);
-  }
-
-  void run() {
-    while (1) {
-      sem_wait(&sem_);
-      if (done_.load()) {
-        break;
-      }
-
-      uint64_t shard;
-      {
-        std::lock_guard<std::mutex> guard(queue_mutex_);
-        assert(!shards_pending_queue_.empty());
-        shard = shards_pending_queue_.front();
-        shards_pending_queue_.pop();
-        shards_pending_set_.reset(shard);
-      }
-      readOnceFromShard(shard);
-    }
-  }
-
-  void readOnceFromShard(uint64_t shard) {
-    ShardState& state = (*shard_states_)[shard];
-    if (!state.it) {
-      // Initialize iterators
-      rocksdb::ReadOptions options;
-      options.tailing = true;
-      if (FLAGS_iterate_upper_bound) {
-        state.upper_bound = Key(shard, std::numeric_limits<uint64_t>::max());
-        state.upper_bound_slice = rocksdb::Slice(
-            (const char*)&state.upper_bound, sizeof(state.upper_bound));
-        options.iterate_upper_bound = &state.upper_bound_slice;
-      }
-
-      state.it.reset(db_->NewIterator(options));
-
-      if (FLAGS_cache_only_first) {
-        options.read_tier = rocksdb::ReadTier::kBlockCacheTier;
-        state.it_cacheonly.reset(db_->NewIterator(options));
-      }
-    }
-
-    const uint64_t upto = state.last_written.load();
-    for (rocksdb::Iterator* it : {state.it_cacheonly.get(), state.it.get()}) {
-      if (it == nullptr) {
-        continue;
-      }
-      if (state.last_read.load() >= upto) {
-        break;
-      }
-      bool need_seek = true;
-      for (uint64_t seq = state.last_read.load() + 1; seq <= upto; ++seq) {
-        if (need_seek) {
-          Key from(shard, state.last_read.load() + 1);
-          it->Seek(rocksdb::Slice((const char*)&from, sizeof(from)));
-          need_seek = false;
-        } else {
-          it->Next();
-        }
-        if (it->status().IsIncomplete()) {
-          ++::stats.cache_misses;
-          break;
-        }
-        assert(it->Valid());
-        assert(it->key().size() == sizeof(Key));
-        Key key;
-        memcpy(&key, it->key().data(), it->key().size());
-        // fprintf(stderr, "Expecting (%ld, %ld) read (%ld, %ld)\n",
-        //         shard, seq, key.shard(), key.seqno());
-        assert(key.shard() == shard);
-        assert(key.seqno() == seq);
-        state.last_read.store(seq);
-        ++::stats.read;
-      }
-    }
-  }
-
-  void onWrite(uint64_t shard) {
-    {
-      std::lock_guard<std::mutex> guard(queue_mutex_);
-      if (!shards_pending_set_.test(shard)) {
-        shards_pending_queue_.push(shard);
-        shards_pending_set_.set(shard);
-        sem_post(&sem_);
-      }
-    }
-  }
-
-  ~Reader() {
-    done_.store(true);
-    sem_post(&sem_);
-    thread_.join();
-  }
-
- private:
-  char pad1[128] __attribute__((__unused__));
-  std::vector<ShardState>* shard_states_;
-  rocksdb::DB* db_;
-  std::thread thread_;
-  sem_t sem_;
-  std::mutex queue_mutex_;
-  std::bitset<MAX_SHARDS + 1> shards_pending_set_;
-  std::queue<uint64_t> shards_pending_queue_;
-  std::atomic<bool> done_{false};
-  char pad2[128] __attribute__((__unused__));
-};
-
-struct Writer {
-  explicit Writer(std::vector<ShardState>* shard_states, rocksdb::DB* db)
-      : shard_states_(shard_states), db_(db) {}
-
-  void start() { thread_ = std::thread(&Writer::run, this); }
-
-  void run() {
-    std::queue<std::chrono::steady_clock::time_point> workq;
-    std::chrono::steady_clock::time_point deadline(
-        std::chrono::steady_clock::now() +
-        std::chrono::nanoseconds((uint64_t)(1000000000 * FLAGS_runtime)));
-    std::vector<uint64_t> my_shards;
-    for (int i = 1; i <= FLAGS_shards; ++i) {
-      if ((*shard_states_)[i].writer == this) {
-        my_shards.push_back(i);
-      }
-    }
-
-    std::mt19937 rng{std::random_device()()};
-    std::uniform_int_distribution<int> shard_dist(
-        0, static_cast<int>(my_shards.size()) - 1);
-    std::string value(FLAGS_value_size, '*');
-
-    while (1) {
-      auto now = std::chrono::steady_clock::now();
-      if (FLAGS_runtime >= 0 && now >= deadline) {
-        break;
-      }
-      if (workq.empty()) {
-        for (int i = 0; i < FLAGS_rate; i += FLAGS_writers) {
-          std::chrono::nanoseconds offset(1000000000LL * i / FLAGS_rate);
-          workq.push(now + offset);
-        }
-      }
-      while (!workq.empty() && workq.front() < now) {
-        workq.pop();
-        uint64_t shard = my_shards[shard_dist(rng)];
-        ShardState& state = (*shard_states_)[shard];
-        uint64_t seqno = state.last_written.load() + 1;
-        Key key(shard, seqno);
-        // fprintf(stderr, "Writing (%ld, %ld)\n", shard, seqno);
-        rocksdb::Status status =
-            db_->Put(rocksdb::WriteOptions(),
-                     rocksdb::Slice((const char*)&key, sizeof(key)),
-                     rocksdb::Slice(value));
-        assert(status.ok());
-        state.last_written.store(seqno);
-        state.reader->onWrite(shard);
-        ++::stats.written;
-      }
-      std::this_thread::sleep_for(std::chrono::milliseconds(1));
-    }
-    // fprintf(stderr, "Writer done\n");
-  }
-
-  ~Writer() { thread_.join(); }
-
- private:
-  char pad1[128] __attribute__((__unused__));
-  std::vector<ShardState>* shard_states_;
-  rocksdb::DB* db_;
-  std::thread thread_;
-  char pad2[128] __attribute__((__unused__));
-};
-
-struct StatsThread {
-  explicit StatsThread(rocksdb::DB* db)
-      : db_(db), thread_(&StatsThread::run, this) {}
-
-  void run() {
-    //    using namespace std::chrono;
-    auto tstart = std::chrono::steady_clock::now(), tlast = tstart;
-    uint64_t wlast = 0, rlast = 0;
-    while (!done_.load()) {
-      {
-        std::unique_lock<std::mutex> lock(cvm_);
-        cv_.wait_for(lock, std::chrono::seconds(1));
-      }
-      auto now = std::chrono::steady_clock::now();
-      double elapsed =
-          std::chrono::duration_cast<std::chrono::duration<double> >(
-              now - tlast).count();
-      uint64_t w = ::stats.written.load();
-      uint64_t r = ::stats.read.load();
-      fprintf(stderr,
-              "%s elapsed %4lds | written %10ld | w/s %10.0f | read %10ld | "
-              "r/s %10.0f | cache misses %10ld\n",
-              db_->GetEnv()->TimeToString(time(nullptr)).c_str(),
-              std::chrono::duration_cast<std::chrono::seconds>(now - tstart)
-                  .count(),
-              w, (w - wlast) / elapsed, r, (r - rlast) / elapsed,
-              ::stats.cache_misses.load());
-      wlast = w;
-      rlast = r;
-      tlast = now;
-    }
-  }
-
-  ~StatsThread() {
-    {
-      std::lock_guard<std::mutex> guard(cvm_);
-      done_.store(true);
-    }
-    cv_.notify_all();
-    thread_.join();
-  }
-
- private:
-  rocksdb::DB* db_;
-  std::mutex cvm_;
-  std::condition_variable cv_;
-  std::thread thread_;
-  std::atomic<bool> done_{false};
-};
-
-int main(int argc, char** argv) {
-  GFLAGS::ParseCommandLineFlags(&argc, &argv, true);
-
-  std::mt19937 rng{std::random_device()()};
-  rocksdb::Status status;
-  std::string path = rocksdb::test::TmpDir() + "/forward_iterator_test";
-  fprintf(stderr, "db path is %s\n", path.c_str());
-  rocksdb::Options options;
-  options.create_if_missing = true;
-  options.compression = rocksdb::CompressionType::kNoCompression;
-  options.compaction_style = rocksdb::CompactionStyle::kCompactionStyleNone;
-  options.level0_slowdown_writes_trigger = 99999;
-  options.level0_stop_writes_trigger = 99999;
-  options.allow_os_buffer = false;
-  options.write_buffer_size = FLAGS_memtable_size;
-  rocksdb::BlockBasedTableOptions table_options;
-  table_options.block_cache = rocksdb::NewLRUCache(FLAGS_block_cache_size);
-  table_options.block_size = FLAGS_block_size;
-  options.table_factory.reset(
-      rocksdb::NewBlockBasedTableFactory(table_options));
-
-  status = rocksdb::DestroyDB(path, options);
-  assert(status.ok());
-  rocksdb::DB* db_raw;
-  status = rocksdb::DB::Open(options, path, &db_raw);
-  assert(status.ok());
-  std::unique_ptr<rocksdb::DB> db(db_raw);
-
-  std::vector<ShardState> shard_states(FLAGS_shards + 1);
-  std::deque<Reader> readers;
-  while (static_cast<int>(readers.size()) < FLAGS_readers) {
-    readers.emplace_back(&shard_states, db_raw);
-  }
-  std::deque<Writer> writers;
-  while (static_cast<int>(writers.size()) < FLAGS_writers) {
-    writers.emplace_back(&shard_states, db_raw);
-  }
-
-  // Each shard gets a random reader and random writer assigned to it
-  for (int i = 1; i <= FLAGS_shards; ++i) {
-    std::uniform_int_distribution<int> reader_dist(0, FLAGS_readers - 1);
-    std::uniform_int_distribution<int> writer_dist(0, FLAGS_writers - 1);
-    shard_states[i].reader = &readers[reader_dist(rng)];
-    shard_states[i].writer = &writers[writer_dist(rng)];
-  }
-
-  StatsThread stats_thread(db_raw);
-  for (Writer& w : writers) {
-    w.start();
-  }
-
-  writers.clear();
-  readers.clear();
-}
-#endif  // !defined(GFLAGS) || defined(ROCKSDB_LITE)
@@ -1,657 +0,0 @@
-//  Copyright (c) 2013, Facebook, Inc.  All rights reserved.
-//  This source code is licensed under the BSD-style license found in the
-//  LICENSE file in the root directory of this source tree. An additional
-//  grant of patent rights can be found in the PATENTS file in the same
-//  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.
-//
-// InlineSkipList is derived from SkipList (skiplist.h), but it optimizes
-// the memory layout by requiring that the key storage be allocated through
-// the skip list instance.  For the common case of SkipList<const char*,
-// Cmp> this saves 1 pointer per skip list node and gives better cache
-// locality, at the expense of wasted padding from using AllocateAligned
-// instead of Allocate for the keys.  The unused padding will be from
-// 0 to sizeof(void*)-1 bytes, and the space savings are sizeof(void*)
-// bytes, so despite the padding the space used is always less than
-// SkipList<const char*, ..>.
-//
-// Thread safety -------------
-//
-// Writes via Insert require external synchronization, most likely a mutex.
-// InsertConcurrently can be safely called concurrently with reads and
-// with other concurrent inserts.  Reads require a guarantee that the
-// InlineSkipList will not be destroyed while the read is in progress.
-// Apart from that, reads progress without any internal locking or
-// synchronization.
-//
-// Invariants:
-//
-// (1) Allocated nodes are never deleted until the InlineSkipList is
-// destroyed.  This is trivially guaranteed by the code since we never
-// delete any skip list nodes.
-//
-// (2) The contents of a Node except for the next/prev pointers are
-// immutable after the Node has been linked into the InlineSkipList.
-// Only Insert() modifies the list, and it is careful to initialize a
-// node and use release-stores to publish the nodes in one or more lists.
-//
-// ... prev vs. next pointer ordering ...
-//
-
-#pragma once
-#include <assert.h>
-#include <stdlib.h>
-#include <atomic>
-#include "port/port.h"
-#include "util/allocator.h"
-#include "util/random.h"
-
-namespace rocksdb {
-
-template <class Comparator>
-class InlineSkipList {
- private:
-  struct Node;
-
- public:
-  // Create a new InlineSkipList object that will use "cmp" for comparing
-  // keys, and will allocate memory using "*allocator".  Objects allocated
-  // in the allocator must remain allocated for the lifetime of the
-  // skiplist object.
-  explicit InlineSkipList(Comparator cmp, Allocator* allocator,
-                          int32_t max_height = 12,
-                          int32_t branching_factor = 4);
-
-  // Allocates a key and a skip-list node, returning a pointer to the key
-  // portion of the node.  This method is thread-safe if the allocator
-  // is thread-safe.
-  char* AllocateKey(size_t key_size);
-
-  // Inserts a key allocated by AllocateKey, after the actual key value
-  // has been filled in.
-  //
-  // REQUIRES: nothing that compares equal to key is currently in the list.
-  // REQUIRES: no concurrent calls to INSERT
-  void Insert(const char* key);
-
-  // Like Insert, but external synchronization is not required.
-  void InsertConcurrently(const char* key);
-
-  // Returns true iff an entry that compares equal to key is in the list.
-  bool Contains(const char* key) const;
-
-  // Return estimated number of entries smaller than `key`.
-  uint64_t EstimateCount(const char* key) const;
-
-  // Iteration over the contents of a skip list
-  class Iterator {
-   public:
-    // Initialize an iterator over the specified list.
-    // The returned iterator is not valid.
-    explicit Iterator(const InlineSkipList* list);
-
-    // Change the underlying skiplist used for this iterator
-    // This enables us not changing the iterator without deallocating
-    // an old one and then allocating a new one
-    void SetList(const InlineSkipList* list);
-
-    // Returns true iff the iterator is positioned at a valid node.
-    bool Valid() const;
-
-    // Returns the key at the current position.
-    // REQUIRES: Valid()
-    const char* key() const;
-
-    // Advances to the next position.
-    // REQUIRES: Valid()
-    void Next();
-
-    // Advances to the previous position.
-    // REQUIRES: Valid()
-    void Prev();
-
-    // Advance to the first entry with a key >= target
-    void Seek(const char* target);
-
-    // Position at the first entry in list.
-    // Final state of iterator is Valid() iff list is not empty.
-    void SeekToFirst();
-
-    // Position at the last entry in list.
-    // Final state of iterator is Valid() iff list is not empty.
-    void SeekToLast();
-
-   private:
-    const InlineSkipList* list_;
-    Node* node_;
-    // Intentionally copyable
-  };
-
- private:
-  enum MaxPossibleHeightEnum : uint16_t { kMaxPossibleHeight = 32 };
-
-  const uint16_t kMaxHeight_;
-  const uint16_t kBranching_;
-  const uint32_t kScaledInverseBranching_;
-
-  // Immutable after construction
-  Comparator const compare_;
-  Allocator* const allocator_;  // Allocator used for allocations of nodes
-
-  Node* const head_;
-
-  // Modified only by Insert().  Read racily by readers, but stale
-  // values are ok.
-  std::atomic<int> max_height_;  // Height of the entire list
-
-  // Used for optimizing sequential insert patterns.  Tricky.  prev_height_
-  // of zero means prev_ is undefined.  Otherwise: prev_[i] for i up
-  // to max_height_ - 1 (inclusive) is the predecessor of prev_[0], and
-  // prev_height_ is the height of prev_[0].  prev_[0] can only be equal
-  // to head when max_height_ and prev_height_ are both 1.
-  Node** prev_;
-  std::atomic<int32_t> prev_height_;
-
-  inline int GetMaxHeight() const {
-    return max_height_.load(std::memory_order_relaxed);
-  }
-
-  int RandomHeight();
-
-  Node* AllocateNode(size_t key_size, int height);
-
-  bool Equal(const char* a, const char* b) const {
-    return (compare_(a, b) == 0);
-  }
-
-  // Return true if key is greater than the data stored in "n".  Null n
-  // is considered infinite.
-  bool KeyIsAfterNode(const char* key, Node* n) const;
-
-  // Returns the earliest node with a key >= key.
-  // Return nullptr if there is no such node.
-  Node* FindGreaterOrEqual(const char* key) const;
-
-  // Return the latest node with a key < key.
-  // Return head_ if there is no such node.
-  // Fills prev[level] with pointer to previous node at "level" for every
-  // level in [0..max_height_-1], if prev is non-null.
-  Node* FindLessThan(const char* key, Node** prev = nullptr) const;
-
-  // Return the last node in the list.
-  // Return head_ if list is empty.
-  Node* FindLast() const;
-
-  // Traverses a single level of the list, setting *out_prev to the last
-  // node before the key and *out_next to the first node after. Assumes
-  // that the key is not present in the skip list. On entry, before should
-  // point to a node that is before the key, and after should point to
-  // a node that is after the key.  after should be nullptr if a good after
-  // node isn't conveniently available.
-  void FindLevelSplice(const char* key, Node* before, Node* after, int level,
-                       Node** out_prev, Node** out_next);
-
-  // No copying allowed
-  InlineSkipList(const InlineSkipList&);
-  InlineSkipList& operator=(const InlineSkipList&);
-};
-
-// Implementation details follow
-
-// The Node data type is more of a pointer into custom-managed memory than
-// a traditional C++ struct.  The key is stored in the bytes immediately
-// after the struct, and the next_ pointers for nodes with height > 1 are
-// stored immediately _before_ the struct.  This avoids the need to include
-// any pointer or sizing data, which reduces per-node memory overheads.
-template <class Comparator>
-struct InlineSkipList<Comparator>::Node {
-  // Stores the height of the node in the memory location normally used for
-  // next_[0].  This is used for passing data from AllocateKey to Insert.
-  void StashHeight(const int height) {
-    assert(sizeof(int) <= sizeof(next_[0]));
-    memcpy(&next_[0], &height, sizeof(int));
-  }
-
-  // Retrieves the value passed to StashHeight.  Undefined after a call
-  // to SetNext or NoBarrier_SetNext.
-  int UnstashHeight() const {
-    int rv;
-    memcpy(&rv, &next_[0], sizeof(int));
-    return rv;
-  }
-
-  const char* Key() const { return reinterpret_cast<const char*>(&next_[1]); }
-
-  // Accessors/mutators for links.  Wrapped in methods so we can add
-  // the appropriate barriers as necessary, and perform the necessary
-  // addressing trickery for storing links below the Node in memory.
-  Node* Next(int n) {
-    assert(n >= 0);
-    // Use an 'acquire load' so that we observe a fully initialized
-    // version of the returned Node.
-    return (next_[-n].load(std::memory_order_acquire));
-  }
-
-  void SetNext(int n, Node* x) {
-    assert(n >= 0);
-    // Use a 'release store' so that anybody who reads through this
-    // pointer observes a fully initialized version of the inserted node.
-    next_[-n].store(x, std::memory_order_release);
-  }
-
-  bool CASNext(int n, Node* expected, Node* x) {
-    assert(n >= 0);
-    return next_[-n].compare_exchange_strong(expected, x);
-  }
-
-  // No-barrier variants that can be safely used in a few locations.
-  Node* NoBarrier_Next(int n) {
-    assert(n >= 0);
-    return next_[-n].load(std::memory_order_relaxed);
-  }
-
-  void NoBarrier_SetNext(int n, Node* x) {
-    assert(n >= 0);
-    next_[-n].store(x, std::memory_order_relaxed);
-  }
-
- private:
-  // next_[0] is the lowest level link (level 0).  Higher levels are
-  // stored _earlier_, so level 1 is at next_[-1].
-  std::atomic<Node*> next_[1];
-};
-
-template <class Comparator>
-inline InlineSkipList<Comparator>::Iterator::Iterator(
-    const InlineSkipList* list) {
-  SetList(list);
-}
-
-template <class Comparator>
-inline void InlineSkipList<Comparator>::Iterator::SetList(
-    const InlineSkipList* list) {
-  list_ = list;
-  node_ = nullptr;
-}
-
-template <class Comparator>
-inline bool InlineSkipList<Comparator>::Iterator::Valid() const {
-  return node_ != nullptr;
-}
-
-template <class Comparator>
-inline const char* InlineSkipList<Comparator>::Iterator::key() const {
-  assert(Valid());
-  return node_->Key();
-}
-
-template <class Comparator>
-inline void InlineSkipList<Comparator>::Iterator::Next() {
-  assert(Valid());
-  node_ = node_->Next(0);
-}
-
-template <class Comparator>
-inline void InlineSkipList<Comparator>::Iterator::Prev() {
-  // Instead of using explicit "prev" links, we just search for the
-  // last node that falls before key.
-  assert(Valid());
-  node_ = list_->FindLessThan(node_->Key());
-  if (node_ == list_->head_) {
-    node_ = nullptr;
-  }
-}
-
-template <class Comparator>
-inline void InlineSkipList<Comparator>::Iterator::Seek(const char* target) {
-  node_ = list_->FindGreaterOrEqual(target);
-}
-
-template <class Comparator>
-inline void InlineSkipList<Comparator>::Iterator::SeekToFirst() {
-  node_ = list_->head_->Next(0);
-}
-
-template <class Comparator>
-inline void InlineSkipList<Comparator>::Iterator::SeekToLast() {
-  node_ = list_->FindLast();
-  if (node_ == list_->head_) {
-    node_ = nullptr;
-  }
-}
-
-template <class Comparator>
-int InlineSkipList<Comparator>::RandomHeight() {
-  auto rnd = Random::GetTLSInstance();
-
-  // Increase height with probability 1 in kBranching
-  int height = 1;
-  while (height < kMaxHeight_ && height < kMaxPossibleHeight &&
-         rnd->Next() < kScaledInverseBranching_) {
-    height++;
-  }
-  assert(height > 0);
-  assert(height <= kMaxHeight_);
-  assert(height <= kMaxPossibleHeight);
-  return height;
-}
-
-template <class Comparator>
-bool InlineSkipList<Comparator>::KeyIsAfterNode(const char* key,
-                                                Node* n) const {
-  // nullptr n is considered infinite
-  return (n != nullptr) && (compare_(n->Key(), key) < 0);
-}
-
-template <class Comparator>
-typename InlineSkipList<Comparator>::Node*
-InlineSkipList<Comparator>::FindGreaterOrEqual(const char* key) const {
-  // Note: It looks like we could reduce duplication by implementing
-  // this function as FindLessThan(key)->Next(0), but we wouldn't be able
-  // to exit early on equality and the result wouldn't even be correct.
-  // A concurrent insert might occur after FindLessThan(key) but before
-  // we get a chance to call Next(0).
-  Node* x = head_;
-  int level = GetMaxHeight() - 1;
-  Node* last_bigger = nullptr;
-  while (true) {
-    Node* next = x->Next(level);
-    // Make sure the lists are sorted
-    assert(x == head_ || next == nullptr || KeyIsAfterNode(next->Key(), x));
-    // Make sure we haven't overshot during our search
-    assert(x == head_ || KeyIsAfterNode(key, x));
-    int cmp = (next == nullptr || next == last_bigger)
-                  ? 1
-                  : compare_(next->Key(), key);
-    if (cmp == 0 || (cmp > 0 && level == 0)) {
-      return next;
-    } else if (cmp < 0) {
-      // Keep searching in this list
-      x = next;
-    } else {
-      // Switch to next list, reuse compare_() result
-      last_bigger = next;
-      level--;
-    }
-  }
-}
-
-template <class Comparator>
-typename InlineSkipList<Comparator>::Node*
-InlineSkipList<Comparator>::FindLessThan(const char* key, Node** prev) const {
-  Node* x = head_;
-  int level = GetMaxHeight() - 1;
-  // KeyIsAfter(key, last_not_after) is definitely false
-  Node* last_not_after = nullptr;
-  while (true) {
-    Node* next = x->Next(level);
-    assert(x == head_ || next == nullptr || KeyIsAfterNode(next->Key(), x));
-    assert(x == head_ || KeyIsAfterNode(key, x));
-    if (next != last_not_after && KeyIsAfterNode(key, next)) {
-      // Keep searching in this list
-      x = next;
-    } else {
-      if (prev != nullptr) {
-        prev[level] = x;
-      }
-      if (level == 0) {
-        return x;
-      } else {
-        // Switch to next list, reuse KeyIUsAfterNode() result
-        last_not_after = next;
-        level--;
-      }
-    }
-  }
-}
-
-template <class Comparator>
-typename InlineSkipList<Comparator>::Node*
-InlineSkipList<Comparator>::FindLast() const {
-  Node* x = head_;
-  int level = GetMaxHeight() - 1;
-  while (true) {
-    Node* next = x->Next(level);
-    if (next == nullptr) {
-      if (level == 0) {
-        return x;
-      } else {
-        // Switch to next list
-        level--;
-      }
-    } else {
-      x = next;
-    }
-  }
-}
-
-template <class Comparator>
-uint64_t InlineSkipList<Comparator>::EstimateCount(const char* key) const {
-  uint64_t count = 0;
-
-  Node* x = head_;
-  int level = GetMaxHeight() - 1;
-  while (true) {
-    assert(x == head_ || compare_(x->Key(), key) < 0);
-    Node* next = x->Next(level);
-    if (next == nullptr || compare_(next->Key(), key) >= 0) {
-      if (level == 0) {
-        return count;
-      } else {
-        // Switch to next list
-        count *= kBranching_;
-        level--;
-      }
-    } else {
-      x = next;
-      count++;
-    }
-  }
-}
-
-template <class Comparator>
-InlineSkipList<Comparator>::InlineSkipList(const Comparator cmp,
-                                           Allocator* allocator,
-                                           int32_t max_height,
-                                           int32_t branching_factor)
-    : kMaxHeight_(max_height),
-      kBranching_(branching_factor),
-      kScaledInverseBranching_((Random::kMaxNext + 1) / kBranching_),
-      compare_(cmp),
-      allocator_(allocator),
-      head_(AllocateNode(0, max_height)),
-      max_height_(1),
-      prev_height_(1) {
-  assert(max_height > 0 && kMaxHeight_ == static_cast<uint32_t>(max_height));
-  assert(branching_factor > 1 &&
-         kBranching_ == static_cast<uint32_t>(branching_factor));
-  assert(kScaledInverseBranching_ > 0);
-  // Allocate the prev_ Node* array, directly from the passed-in allocator.
-  // prev_ does not need to be freed, as its life cycle is tied up with
-  // the allocator as a whole.
-  prev_ = reinterpret_cast<Node**>(
-      allocator_->AllocateAligned(sizeof(Node*) * kMaxHeight_));
-  for (int i = 0; i < kMaxHeight_; i++) {
-    head_->SetNext(i, nullptr);
-    prev_[i] = head_;
-  }
-}
-
-template <class Comparator>
-char* InlineSkipList<Comparator>::AllocateKey(size_t key_size) {
-  return const_cast<char*>(AllocateNode(key_size, RandomHeight())->Key());
-}
-
-template <class Comparator>
-typename InlineSkipList<Comparator>::Node*
-InlineSkipList<Comparator>::AllocateNode(size_t key_size, int height) {
-  auto prefix = sizeof(std::atomic<Node*>) * (height - 1);
-
-  // prefix is space for the height - 1 pointers that we store before
-  // the Node instance (next_[-(height - 1) .. -1]).  Node starts at
-  // raw + prefix, and holds the bottom-mode (level 0) skip list pointer
-  // next_[0].  key_size is the bytes for the key, which comes just after
-  // the Node.
-  char* raw = allocator_->AllocateAligned(prefix + sizeof(Node) + key_size);
-  Node* x = reinterpret_cast<Node*>(raw + prefix);
-
-  // Once we've linked the node into the skip list we don't actually need
-  // to know its height, because we can implicitly use the fact that we
-  // traversed into a node at level h to known that h is a valid level
-  // for that node.  We need to convey the height to the Insert step,
-  // however, so that it can perform the proper links.  Since we're not
-  // using the pointers at the moment, StashHeight temporarily borrow
-  // storage from next_[0] for that purpose.
-  x->StashHeight(height);
-  return x;
-}
-
-template <class Comparator>
-void InlineSkipList<Comparator>::Insert(const char* key) {
-  // InsertConcurrently often can't maintain the prev_ invariants, so
-  // it just sets prev_height_ to zero, letting us know that we should
-  // ignore it.  A relaxed load suffices here because write thread
-  // synchronization separates Insert calls from InsertConcurrently calls.
-  auto prev_height = prev_height_.load(std::memory_order_relaxed);
-
-  // fast path for sequential insertion
-  if (prev_height > 0 && !KeyIsAfterNode(key, prev_[0]->NoBarrier_Next(0)) &&
-      (prev_[0] == head_ || KeyIsAfterNode(key, prev_[0]))) {
-    assert(prev_[0] != head_ || (prev_height == 1 && GetMaxHeight() == 1));
-
-    // Outside of this method prev_[1..max_height_] is the predecessor
-    // of prev_[0], and prev_height_ refers to prev_[0].  Inside Insert
-    // prev_[0..max_height - 1] is the predecessor of key.  Switch from
-    // the external state to the internal
-    for (int i = 1; i < prev_height; i++) {
-      prev_[i] = prev_[0];
-    }
-  } else {
-    // TODO(opt): we could use a NoBarrier predecessor search as an
-    // optimization for architectures where memory_order_acquire needs
-    // a synchronization instruction.  Doesn't matter on x86
-    FindLessThan(key, prev_);
-  }
-
-  // Our data structure does not allow duplicate insertion
-  assert(prev_[0]->Next(0) == nullptr || !Equal(key, prev_[0]->Next(0)->Key()));
-
-  // Find the Node that we placed before the key in AllocateKey
-  Node* x = reinterpret_cast<Node*>(const_cast<char*>(key)) - 1;
-  int height = x->UnstashHeight();
-  assert(height >= 1 && height <= kMaxHeight_);
-
-  if (height > GetMaxHeight()) {
-    for (int i = GetMaxHeight(); i < height; i++) {
-      prev_[i] = head_;
-    }
-
-    // It is ok to mutate max_height_ without any synchronization
-    // with concurrent readers.  A concurrent reader that observes
-    // the new value of max_height_ will see either the old value of
-    // new level pointers from head_ (nullptr), or a new value set in
-    // the loop below.  In the former case the reader will
-    // immediately drop to the next level since nullptr sorts after all
-    // keys.  In the latter case the reader will use the new node.
-    max_height_.store(height, std::memory_order_relaxed);
-  }
-
-  for (int i = 0; i < height; i++) {
-    // NoBarrier_SetNext() suffices since we will add a barrier when
-    // we publish a pointer to "x" in prev[i].
-    x->NoBarrier_SetNext(i, prev_[i]->NoBarrier_Next(i));
-    prev_[i]->SetNext(i, x);
-  }
-  prev_[0] = x;
-  prev_height_.store(height, std::memory_order_relaxed);
-}
-
-template <class Comparator>
-void InlineSkipList<Comparator>::FindLevelSplice(const char* key, Node* before,
-                                                 Node* after, int level,
-                                                 Node** out_prev,
-                                                 Node** out_next) {
-  while (true) {
-    Node* next = before->Next(level);
-    assert(before == head_ || next == nullptr ||
-           KeyIsAfterNode(next->Key(), before));
-    assert(before == head_ || KeyIsAfterNode(key, before));
-    if (next == after || !KeyIsAfterNode(key, next)) {
-      // found it
-      *out_prev = before;
-      *out_next = next;
-      return;
-    }
-    before = next;
-  }
-}
-
-template <class Comparator>
-void InlineSkipList<Comparator>::InsertConcurrently(const char* key) {
-  Node* x = reinterpret_cast<Node*>(const_cast<char*>(key)) - 1;
-  int height = x->UnstashHeight();
-  assert(height >= 1 && height <= kMaxHeight_);
-
-  // We don't have a lock-free algorithm for updating prev_, but we do have
-  // the option of invalidating the entire sequential-insertion cache.
-  // prev_'s invariant is that prev_[i] (i > 0) is the predecessor of
-  // prev_[0] at that level.  We're only going to violate that if height
-  // > 1 and key lands after prev_[height - 1] but before prev_[0].
-  // Comparisons are pretty expensive, so an easier version is to just
-  // clear the cache if height > 1.  We only write to prev_height_ if the
-  // nobody else has, to avoid invalidating the root of the skip list in
-  // all of the other CPU caches.
-  if (height > 1 && prev_height_.load(std::memory_order_relaxed) != 0) {
-    prev_height_.store(0, std::memory_order_relaxed);
-  }
-
-  int max_height = max_height_.load(std::memory_order_relaxed);
-  while (height > max_height) {
-    if (max_height_.compare_exchange_strong(max_height, height)) {
-      // successfully updated it
-      max_height = height;
-      break;
-    }
-    // else retry, possibly exiting the loop because somebody else
-    // increased it
-  }
-  assert(max_height <= kMaxPossibleHeight);
-
-  Node* prev[kMaxPossibleHeight + 1];
-  Node* next[kMaxPossibleHeight + 1];
-  prev[max_height] = head_;
-  next[max_height] = nullptr;
-  for (int i = max_height - 1; i >= 0; --i) {
-    FindLevelSplice(key, prev[i + 1], next[i + 1], i, &prev[i], &next[i]);
-  }
-  for (int i = 0; i < height; ++i) {
-    while (true) {
-      x->NoBarrier_SetNext(i, next[i]);
-      if (prev[i]->CASNext(i, next[i], x)) {
-        // success
-        break;
-      }
-      // CAS failed, we need to recompute prev and next. It is unlikely
-      // to be helpful to try to use a different level as we redo the
-      // search, because it should be unlikely that lots of nodes have
-      // been inserted between prev[i] and next[i]. No point in using
-      // next[i] as the after hint, because we know it is stale.
-      FindLevelSplice(key, prev[i], nullptr, i, &prev[i], &next[i]);
-    }
-  }
-}
-
-template <class Comparator>
-bool InlineSkipList<Comparator>::Contains(const char* key) const {
-  Node* x = FindGreaterOrEqual(key);
-  if (x != nullptr && Equal(key, x->Key())) {
-    return true;
-  } else {
-    return false;
-  }
-}
-
-}  // namespace rocksdb
@@ -1,475 +0,0 @@
-//  Copyright (c) 2013, Facebook, Inc.  All rights reserved.
-//  This source code is licensed under the BSD-style license found in the
-//  LICENSE file in the root directory of this source tree. An additional grant
-//  of patent rights can be found in the PATENTS file in the same 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 "db/inlineskiplist.h"
-#include <set>
-#include "rocksdb/env.h"
-#include "util/concurrent_arena.h"
-#include "util/hash.h"
-#include "util/random.h"
-#include "util/testharness.h"
-
-namespace rocksdb {
-
-// Our test skip list stores 8-byte unsigned integers
-typedef uint64_t Key;
-
-static const char* Encode(const uint64_t* key) {
-  return reinterpret_cast<const char*>(key);
-}
-
-static Key Decode(const char* key) {
-  Key rv;
-  memcpy(&rv, key, sizeof(Key));
-  return rv;
-}
-
-struct TestComparator {
-  int operator()(const char* a, const char* b) const {
-    if (Decode(a) < Decode(b)) {
-      return -1;
-    } else if (Decode(a) > Decode(b)) {
-      return +1;
-    } else {
-      return 0;
-    }
-  }
-};
-
-class InlineSkipTest : public testing::Test {};
-
-TEST_F(InlineSkipTest, Empty) {
-  Arena arena;
-  TestComparator cmp;
-  InlineSkipList<TestComparator> list(cmp, &arena);
-  Key key = 10;
-  ASSERT_TRUE(!list.Contains(Encode(&key)));
-
-  InlineSkipList<TestComparator>::Iterator iter(&list);
-  ASSERT_TRUE(!iter.Valid());
-  iter.SeekToFirst();
-  ASSERT_TRUE(!iter.Valid());
-  key = 100;
-  iter.Seek(Encode(&key));
-  ASSERT_TRUE(!iter.Valid());
-  iter.SeekToLast();
-  ASSERT_TRUE(!iter.Valid());
-}
-
-TEST_F(InlineSkipTest, InsertAndLookup) {
-  const int N = 2000;
-  const int R = 5000;
-  Random rnd(1000);
-  std::set<Key> keys;
-  ConcurrentArena arena;
-  TestComparator cmp;
-  InlineSkipList<TestComparator> list(cmp, &arena);
-  for (int i = 0; i < N; i++) {
-    Key key = rnd.Next() % R;
-    if (keys.insert(key).second) {
-      char* buf = list.AllocateKey(sizeof(Key));
-      memcpy(buf, &key, sizeof(Key));
-      list.Insert(buf);
-    }
-  }
-
-  for (Key i = 0; i < R; i++) {
-    if (list.Contains(Encode(&i))) {
-      ASSERT_EQ(keys.count(i), 1U);
-    } else {
-      ASSERT_EQ(keys.count(i), 0U);
-    }
-  }
-
-  // Simple iterator tests
-  {
-    InlineSkipList<TestComparator>::Iterator iter(&list);
-    ASSERT_TRUE(!iter.Valid());
-
-    uint64_t zero = 0;
-    iter.Seek(Encode(&zero));
-    ASSERT_TRUE(iter.Valid());
-    ASSERT_EQ(*(keys.begin()), Decode(iter.key()));
-
-    iter.SeekToFirst();
-    ASSERT_TRUE(iter.Valid());
-    ASSERT_EQ(*(keys.begin()), Decode(iter.key()));
-
-    iter.SeekToLast();
-    ASSERT_TRUE(iter.Valid());
-    ASSERT_EQ(*(keys.rbegin()), Decode(iter.key()));
-  }
-
-  // Forward iteration test
-  for (Key i = 0; i < R; i++) {
-    InlineSkipList<TestComparator>::Iterator iter(&list);
-    iter.Seek(Encode(&i));
-
-    // Compare against model iterator
-    std::set<Key>::iterator model_iter = keys.lower_bound(i);
-    for (int j = 0; j < 3; j++) {
-      if (model_iter == keys.end()) {
-        ASSERT_TRUE(!iter.Valid());
-        break;
-      } else {
-        ASSERT_TRUE(iter.Valid());
-        ASSERT_EQ(*model_iter, Decode(iter.key()));
-        ++model_iter;
-        iter.Next();
-      }
-    }
-  }
-
-  // Backward iteration test
-  {
-    InlineSkipList<TestComparator>::Iterator iter(&list);
-    iter.SeekToLast();
-
-    // Compare against model iterator
-    for (std::set<Key>::reverse_iterator model_iter = keys.rbegin();
-         model_iter != keys.rend(); ++model_iter) {
-      ASSERT_TRUE(iter.Valid());
-      ASSERT_EQ(*model_iter, Decode(iter.key()));
-      iter.Prev();
-    }
-    ASSERT_TRUE(!iter.Valid());
-  }
-}
-
-// We want to make sure that with a single writer and multiple
-// concurrent readers (with no synchronization other than when a
-// reader's iterator is created), the reader always observes all the
-// data that was present in the skip list when the iterator was
-// constructor.  Because insertions are happening concurrently, we may
-// also observe new values that were inserted since the iterator was
-// constructed, but we should never miss any values that were present
-// at iterator construction time.
-//
-// We generate multi-part keys:
-//     <key,gen,hash>
-// where:
-//     key is in range [0..K-1]
-//     gen is a generation number for key
-//     hash is hash(key,gen)
-//
-// The insertion code picks a random key, sets gen to be 1 + the last
-// generation number inserted for that key, and sets hash to Hash(key,gen).
-//
-// At the beginning of a read, we snapshot the last inserted
-// generation number for each key.  We then iterate, including random
-// calls to Next() and Seek().  For every key we encounter, we
-// check that it is either expected given the initial snapshot or has
-// been concurrently added since the iterator started.
-class ConcurrentTest {
- public:
-  static const uint32_t K = 8;
-
- private:
-  static uint64_t key(Key key) { return (key >> 40); }
-  static uint64_t gen(Key key) { return (key >> 8) & 0xffffffffu; }
-  static uint64_t hash(Key key) { return key & 0xff; }
-
-  static uint64_t HashNumbers(uint64_t k, uint64_t g) {
-    uint64_t data[2] = {k, g};
-    return Hash(reinterpret_cast<char*>(data), sizeof(data), 0);
-  }
-
-  static Key MakeKey(uint64_t k, uint64_t g) {
-    assert(sizeof(Key) == sizeof(uint64_t));
-    assert(k <= K);  // We sometimes pass K to seek to the end of the skiplist
-    assert(g <= 0xffffffffu);
-    return ((k << 40) | (g << 8) | (HashNumbers(k, g) & 0xff));
-  }
-
-  static bool IsValidKey(Key k) {
-    return hash(k) == (HashNumbers(key(k), gen(k)) & 0xff);
-  }
-
-  static Key RandomTarget(Random* rnd) {
-    switch (rnd->Next() % 10) {
-      case 0:
-        // Seek to beginning
-        return MakeKey(0, 0);
-      case 1:
-        // Seek to end
-        return MakeKey(K, 0);
-      default:
-        // Seek to middle
-        return MakeKey(rnd->Next() % K, 0);
-    }
-  }
-
-  // Per-key generation
-  struct State {
-    std::atomic<int> generation[K];
-    void Set(int k, int v) {
-      generation[k].store(v, std::memory_order_release);
-    }
-    int Get(int k) { return generation[k].load(std::memory_order_acquire); }
-
-    State() {
-      for (unsigned int k = 0; k < K; k++) {
-        Set(k, 0);
-      }
-    }
-  };
-
-  // Current state of the test
-  State current_;
-
-  ConcurrentArena arena_;
-
-  // InlineSkipList is not protected by mu_.  We just use a single writer
-  // thread to modify it.
-  InlineSkipList<TestComparator> list_;
-
- public:
-  ConcurrentTest() : list_(TestComparator(), &arena_) {}
-
-  // REQUIRES: No concurrent calls to WriteStep or ConcurrentWriteStep
-  void WriteStep(Random* rnd) {
-    const uint32_t k = rnd->Next() % K;
-    const int g = current_.Get(k) + 1;
-    const Key new_key = MakeKey(k, g);
-    char* buf = list_.AllocateKey(sizeof(Key));
-    memcpy(buf, &new_key, sizeof(Key));
-    list_.Insert(buf);
-    current_.Set(k, g);
-  }
-
-  // REQUIRES: No concurrent calls for the same k
-  void ConcurrentWriteStep(uint32_t k) {
-    const int g = current_.Get(k) + 1;
-    const Key new_key = MakeKey(k, g);
-    char* buf = list_.AllocateKey(sizeof(Key));
-    memcpy(buf, &new_key, sizeof(Key));
-    list_.InsertConcurrently(buf);
-    ASSERT_EQ(g, current_.Get(k) + 1);
-    current_.Set(k, g);
-  }
-
-  void ReadStep(Random* rnd) {
-    // Remember the initial committed state of the skiplist.
-    State initial_state;
-    for (unsigned int k = 0; k < K; k++) {
-      initial_state.Set(k, current_.Get(k));
-    }
-
-    Key pos = RandomTarget(rnd);
-    InlineSkipList<TestComparator>::Iterator iter(&list_);
-    iter.Seek(Encode(&pos));
-    while (true) {
-      Key current;
-      if (!iter.Valid()) {
-        current = MakeKey(K, 0);
-      } else {
-        current = Decode(iter.key());
-        ASSERT_TRUE(IsValidKey(current)) << current;
-      }
-      ASSERT_LE(pos, current) << "should not go backwards";
-
-      // Verify that everything in [pos,current) was not present in
-      // initial_state.
-      while (pos < current) {
-        ASSERT_LT(key(pos), K) << pos;
-
-        // Note that generation 0 is never inserted, so it is ok if
-        // <*,0,*> is missing.
-        ASSERT_TRUE((gen(pos) == 0U) ||
-                    (gen(pos) > static_cast<uint64_t>(initial_state.Get(
-                                    static_cast<int>(key(pos))))))
-            << "key: " << key(pos) << "; gen: " << gen(pos)
-            << "; initgen: " << initial_state.Get(static_cast<int>(key(pos)));
-
-        // Advance to next key in the valid key space
-        if (key(pos) < key(current)) {
-          pos = MakeKey(key(pos) + 1, 0);
-        } else {
-          pos = MakeKey(key(pos), gen(pos) + 1);
-        }
-      }
-
-      if (!iter.Valid()) {
-        break;
-      }
-
-      if (rnd->Next() % 2) {
-        iter.Next();
-        pos = MakeKey(key(pos), gen(pos) + 1);
-      } else {
-        Key new_target = RandomTarget(rnd);
-        if (new_target > pos) {
-          pos = new_target;
-          iter.Seek(Encode(&new_target));
-        }
-      }
-    }
-  }
-};
-const uint32_t ConcurrentTest::K;
-
-// Simple test that does single-threaded testing of the ConcurrentTest
-// scaffolding.
-TEST_F(InlineSkipTest, ConcurrentReadWithoutThreads) {
-  ConcurrentTest test;
-  Random rnd(test::RandomSeed());
-  for (int i = 0; i < 10000; i++) {
-    test.ReadStep(&rnd);
-    test.WriteStep(&rnd);
-  }
-}
-
-TEST_F(InlineSkipTest, ConcurrentInsertWithoutThreads) {
-  ConcurrentTest test;
-  Random rnd(test::RandomSeed());
-  for (int i = 0; i < 10000; i++) {
-    test.ReadStep(&rnd);
-    uint32_t base = rnd.Next();
-    for (int j = 0; j < 4; ++j) {
-      test.ConcurrentWriteStep((base + j) % ConcurrentTest::K);
-    }
-  }
-}
-
-class TestState {
- public:
-  ConcurrentTest t_;
-  int seed_;
-  std::atomic<bool> quit_flag_;
-  std::atomic<uint32_t> next_writer_;
-
-  enum ReaderState { STARTING, RUNNING, DONE };
-
-  explicit TestState(int s)
-      : seed_(s),
-        quit_flag_(false),
-        state_(STARTING),
-        pending_writers_(0),
-        state_cv_(&mu_) {}
-
-  void Wait(ReaderState s) {
-    mu_.Lock();
-    while (state_ != s) {
-      state_cv_.Wait();
-    }
-    mu_.Unlock();
-  }
-
-  void Change(ReaderState s) {
-    mu_.Lock();
-    state_ = s;
-    state_cv_.Signal();
-    mu_.Unlock();
-  }
-
-  void AdjustPendingWriters(int delta) {
-    mu_.Lock();
-    pending_writers_ += delta;
-    if (pending_writers_ == 0) {
-      state_cv_.Signal();
-    }
-    mu_.Unlock();
-  }
-
-  void WaitForPendingWriters() {
-    mu_.Lock();
-    while (pending_writers_ != 0) {
-      state_cv_.Wait();
-    }
-    mu_.Unlock();
-  }
-
- private:
-  port::Mutex mu_;
-  ReaderState state_;
-  int pending_writers_;
-  port::CondVar state_cv_;
-};
-
-static void ConcurrentReader(void* arg) {
-  TestState* state = reinterpret_cast<TestState*>(arg);
-  Random rnd(state->seed_);
-  int64_t reads = 0;
-  state->Change(TestState::RUNNING);
-  while (!state->quit_flag_.load(std::memory_order_acquire)) {
-    state->t_.ReadStep(&rnd);
-    ++reads;
-  }
-  state->Change(TestState::DONE);
-}
-
-static void ConcurrentWriter(void* arg) {
-  TestState* state = reinterpret_cast<TestState*>(arg);
-  uint32_t k = state->next_writer_++ % ConcurrentTest::K;
-  state->t_.ConcurrentWriteStep(k);
-  state->AdjustPendingWriters(-1);
-}
-
-static void RunConcurrentRead(int run) {
-  const int seed = test::RandomSeed() + (run * 100);
-  Random rnd(seed);
-  const int N = 1000;
-  const int kSize = 1000;
-  for (int i = 0; i < N; i++) {
-    if ((i % 100) == 0) {
-      fprintf(stderr, "Run %d of %d\n", i, N);
-    }
-    TestState state(seed + 1);
-    Env::Default()->Schedule(ConcurrentReader, &state);
-    state.Wait(TestState::RUNNING);
-    for (int k = 0; k < kSize; ++k) {
-      state.t_.WriteStep(&rnd);
-    }
-    state.quit_flag_.store(true, std::memory_order_release);
-    state.Wait(TestState::DONE);
-  }
-}
-
-static void RunConcurrentInsert(int run, int write_parallelism = 4) {
-  Env::Default()->SetBackgroundThreads(1 + write_parallelism,
-                                       Env::Priority::LOW);
-  const int seed = test::RandomSeed() + (run * 100);
-  Random rnd(seed);
-  const int N = 1000;
-  const int kSize = 1000;
-  for (int i = 0; i < N; i++) {
-    if ((i % 100) == 0) {
-      fprintf(stderr, "Run %d of %d\n", i, N);
-    }
-    TestState state(seed + 1);
-    Env::Default()->Schedule(ConcurrentReader, &state);
-    state.Wait(TestState::RUNNING);
-    for (int k = 0; k < kSize; k += write_parallelism) {
-      state.next_writer_ = rnd.Next();
-      state.AdjustPendingWriters(write_parallelism);
-      for (int p = 0; p < write_parallelism; ++p) {
-        Env::Default()->Schedule(ConcurrentWriter, &state);
-      }
-      state.WaitForPendingWriters();
-    }
-    state.quit_flag_.store(true, std::memory_order_release);
-    state.Wait(TestState::DONE);
-  }
-}
-
-TEST_F(InlineSkipTest, ConcurrentRead1) { RunConcurrentRead(1); }
-TEST_F(InlineSkipTest, ConcurrentRead2) { RunConcurrentRead(2); }
-TEST_F(InlineSkipTest, ConcurrentRead3) { RunConcurrentRead(3); }
-TEST_F(InlineSkipTest, ConcurrentRead4) { RunConcurrentRead(4); }
-TEST_F(InlineSkipTest, ConcurrentRead5) { RunConcurrentRead(5); }
-TEST_F(InlineSkipTest, ConcurrentInsert1) { RunConcurrentInsert(1); }
-TEST_F(InlineSkipTest, ConcurrentInsert2) { RunConcurrentInsert(2); }
-TEST_F(InlineSkipTest, ConcurrentInsert3) { RunConcurrentInsert(3); }
-
-}  // namespace rocksdb
-
-int main(int argc, char** argv) {
-  ::testing::InitGoogleTest(&argc, argv);
-  return RUN_ALL_TESTS();
-}
@@ -35,7 +35,8 @@ void PrintLevelStatsHeader(char* buf, size_t len, const std::string& cf_name) {
      "\n** Compaction Stats [%s] **\n"
      "Level    Files   Size(MB) Score Read(GB)  Rn(GB) Rnp1(GB) "
      "Write(GB) Wnew(GB) Moved(GB) W-Amp Rd(MB/s) Wr(MB/s) "
-      "Comp(sec) Comp(cnt) Avg(sec) KeyIn KeyDrop\n"
+      "Comp(sec) Comp(cnt) Avg(sec) "
+      "Stall(cnt)  KeyIn KeyDrop\n"
      "--------------------------------------------------------------------"
      "-----------------------------------------------------------"
      "--------------------------------------\n",
@@ -43,9 +44,9 @@ void PrintLevelStatsHeader(char* buf, size_t len, const std::string& cf_name) {
 }

 void PrintLevelStats(char* buf, size_t len, const std::string& name,
-  int num_files, int being_compacted, double total_file_size,
-  double score, double w_amp,
-  const InternalStats::CompactionStats& stats) {
+    int num_files, int being_compacted, double total_file_size, double score,
+    double w_amp, uint64_t stalls,
+    const InternalStats::CompactionStats& stats) {
  uint64_t bytes_read =
      stats.bytes_read_non_output_levels + stats.bytes_read_output_level;
  int64_t bytes_new =
@@ -56,7 +57,7 @@ void PrintLevelStats(char* buf, size_t len, const std::string& name,
      NumberToHumanString(stats.num_dropped_records);

  snprintf(buf, len,
-           "%4s %6d/%-3d %8.2f %5.1f " /* Level, Files, Size(MB), Score */
+           "%4s %6d/%-3d %8.0f %5.1f " /* Level, Files, Size(MB), Score */
           "%8.1f "                    /* Read(GB) */
           "%7.1f "                    /* Rn(GB) */
           "%8.1f "                    /* Rnp1(GB) */
@@ -69,6 +70,8 @@ void PrintLevelStats(char* buf, size_t len, const std::string& name,
           "%9.0f "                    /* Comp(sec) */
           "%9d "                      /* Comp(cnt) */
           "%8.3f "                    /* Avg(sec) */
+           "%10" PRIu64
+           " "      /* Stall(cnt) */
           "%7s "   /* KeyIn */
           "%6s\n", /* KeyDrop */
           name.c_str(),
@@ -79,7 +82,7 @@ void PrintLevelStats(char* buf, size_t len, const std::string& name,
           bytes_read / kMB / elapsed, stats.bytes_written / kMB / elapsed,
           stats.micros / kMicrosInSec, stats.count,
           stats.count == 0 ? 0 : stats.micros / kMicrosInSec / stats.count,
-           num_input_records.c_str(), num_dropped_records.c_str());
+           stalls, num_input_records.c_str(), num_dropped_records.c_str());
 }
 }

@@ -127,8 +130,6 @@ static const std::string aggregated_table_properties =
    "aggregated-table-properties";
 static const std::string aggregated_table_properties_at_level =
    aggregated_table_properties + "-at-level";
-static const std::string num_running_compactions = "num-running-compactions";
-static const std::string num_running_flushes = "num-running-flushes";

 const std::string DB::Properties::kNumFilesAtLevelPrefix =
                      rocksdb_prefix + num_files_at_level_prefix;
@@ -142,10 +143,6 @@ const std::string DB::Properties::kMemTableFlushPending =
                      rocksdb_prefix + mem_table_flush_pending;
 const std::string DB::Properties::kCompactionPending =
                      rocksdb_prefix + compaction_pending;
-const std::string DB::Properties::kNumRunningCompactions =
-    rocksdb_prefix + num_running_compactions;
-const std::string DB::Properties::kNumRunningFlushes =
-    rocksdb_prefix + num_running_flushes;
 const std::string DB::Properties::kBackgroundErrors =
                      rocksdb_prefix + background_errors;
 const std::string DB::Properties::kCurSizeActiveMemTable =
@@ -263,10 +260,6 @@ DBPropertyType GetPropertyType(const Slice& property, bool* is_int_property,
    return kTotalSstFilesSize;
  } else if (in == estimate_pending_comp_bytes) {
    return kEstimatePendingCompactionBytes;
-  } else if (in == num_running_flushes) {
-    return kNumRunningFlushes;
-  } else if (in == num_running_compactions) {
-    return kNumRunningCompactions;
  }
  return kUnknown;
 }
@@ -395,17 +388,11 @@ bool InternalStats::GetIntProperty(DBPropertyType property_type,
      // Return number of mem tables that are ready to flush (made immutable)
      *value = (cfd_->imm()->IsFlushPending() ? 1 : 0);
      return true;
-    case kNumRunningFlushes:
-      *value = db->num_running_flushes();
-      return true;
    case kCompactionPending:
      // 1 if the system already determines at least one compaction is needed.
      // 0 otherwise,
      *value = (cfd_->compaction_picker()->NeedsCompaction(vstorage) ? 1 : 0);
      return true;
-    case kNumRunningCompactions:
-      *value = db->num_running_compactions_;
-      return true;
    case kBackgroundErrors:
      // Accumulated number of  errors in background flushes or compactions.
      *value = GetBackgroundErrorCount();
@@ -485,14 +472,14 @@ void InternalStats::DumpDBStats(std::string* value) {
           seconds_up, interval_seconds_up);
  value->append(buf);
  // Cumulative
-  uint64_t user_bytes_written = GetDBStats(InternalStats::BYTES_WRITTEN);
-  uint64_t num_keys_written = GetDBStats(InternalStats::NUMBER_KEYS_WRITTEN);
-  uint64_t write_other = GetDBStats(InternalStats::WRITE_DONE_BY_OTHER);
-  uint64_t write_self = GetDBStats(InternalStats::WRITE_DONE_BY_SELF);
-  uint64_t wal_bytes = GetDBStats(InternalStats::WAL_FILE_BYTES);
-  uint64_t wal_synced = GetDBStats(InternalStats::WAL_FILE_SYNCED);
-  uint64_t write_with_wal = GetDBStats(InternalStats::WRITE_WITH_WAL);
-  uint64_t write_stall_micros = GetDBStats(InternalStats::WRITE_STALL_MICROS);
+  uint64_t user_bytes_written = db_stats_[InternalStats::BYTES_WRITTEN];
+  uint64_t num_keys_written = db_stats_[InternalStats::NUMBER_KEYS_WRITTEN];
+  uint64_t write_other = db_stats_[InternalStats::WRITE_DONE_BY_OTHER];
+  uint64_t write_self = db_stats_[InternalStats::WRITE_DONE_BY_SELF];
+  uint64_t wal_bytes = db_stats_[InternalStats::WAL_FILE_BYTES];
+  uint64_t wal_synced = db_stats_[InternalStats::WAL_FILE_SYNCED];
+  uint64_t write_with_wal = db_stats_[InternalStats::WRITE_WITH_WAL];
+  uint64_t write_stall_micros = db_stats_[InternalStats::WRITE_STALL_MICROS];
  uint64_t compact_bytes_read = 0;
  uint64_t compact_bytes_write = 0;
  uint64_t compact_micros = 0;
@@ -674,38 +661,44 @@ void InternalStats::DumpCFStats(std::string* value) {
  int total_files = 0;
  int total_files_being_compacted = 0;
  double total_file_size = 0;
+  uint64_t total_slowdown_count_soft = 0;
+  uint64_t total_slowdown_count_hard = 0;
+  uint64_t total_stall_count = 0;
  for (int level = 0; level < number_levels_; level++) {
    int files = vstorage->NumLevelFiles(level);
    total_files += files;
    total_files_being_compacted += files_being_compacted[level];
    if (comp_stats_[level].micros > 0 || files > 0) {
+      uint64_t stalls = level == 0 ?
+        (cf_stats_count_[LEVEL0_SLOWDOWN] +
+         cf_stats_count_[LEVEL0_NUM_FILES] +
+         cf_stats_count_[MEMTABLE_COMPACTION])
+        : (stall_leveln_slowdown_count_soft_[level] +
+           stall_leveln_slowdown_count_hard_[level]);
+
      stats_sum.Add(comp_stats_[level]);
      total_file_size += vstorage->NumLevelBytes(level);
+      total_stall_count += stalls;
+      total_slowdown_count_soft += stall_leveln_slowdown_count_soft_[level];
+      total_slowdown_count_hard += stall_leveln_slowdown_count_hard_[level];
      double w_amp =
          (comp_stats_[level].bytes_read_non_output_levels == 0) ? 0.0
          : static_cast<double>(comp_stats_[level].bytes_written) /
            comp_stats_[level].bytes_read_non_output_levels;
      PrintLevelStats(buf, sizeof(buf), "L" + ToString(level), files,
                      files_being_compacted[level],
-                      static_cast<double>(vstorage->NumLevelBytes(level)),
-                      compaction_score[level],
-                      w_amp, comp_stats_[level]);
+                      vstorage->NumLevelBytes(level), compaction_score[level],
+                      w_amp, stalls, comp_stats_[level]);
      value->append(buf);
    }
  }
  uint64_t curr_ingest = cf_stats_value_[BYTES_FLUSHED];
  // Cumulative summary
  double w_amp = stats_sum.bytes_written / static_cast<double>(curr_ingest + 1);
-  uint64_t total_stall_count =
-      cf_stats_count_[LEVEL0_SLOWDOWN_TOTAL] +
-      cf_stats_count_[LEVEL0_NUM_FILES_TOTAL] +
-      cf_stats_count_[SOFT_PENDING_COMPACTION_BYTES_LIMIT] +
-      cf_stats_count_[HARD_PENDING_COMPACTION_BYTES_LIMIT] +
-      cf_stats_count_[MEMTABLE_COMPACTION] + cf_stats_count_[MEMTABLE_SLOWDOWN];
  // Stats summary across levels
  PrintLevelStats(buf, sizeof(buf), "Sum", total_files,
-                  total_files_being_compacted, total_file_size, 0, w_amp,
-                  stats_sum);
+      total_files_being_compacted, total_file_size, 0, w_amp,
+      total_stall_count, stats_sum);
  value->append(buf);
  // Interval summary
  uint64_t interval_ingest =
@@ -713,7 +706,9 @@ void InternalStats::DumpCFStats(std::string* value) {
  CompactionStats interval_stats(stats_sum);
  interval_stats.Subtract(cf_stats_snapshot_.comp_stats);
  w_amp = interval_stats.bytes_written / static_cast<double>(interval_ingest);
-  PrintLevelStats(buf, sizeof(buf), "Int", 0, 0, 0, 0, w_amp, interval_stats);
+  PrintLevelStats(buf, sizeof(buf), "Int", 0, 0, 0, 0,
+      w_amp, total_stall_count - cf_stats_snapshot_.stall_count,
+      interval_stats);
  value->append(buf);

  snprintf(buf, sizeof(buf),
@@ -721,32 +716,15 @@ void InternalStats::DumpCFStats(std::string* value) {
           curr_ingest / kGB, interval_ingest / kGB);
  value->append(buf);

-  snprintf(buf, sizeof(buf), "Stalls(count): %" PRIu64
-                             " level0_slowdown, "
-                             "%" PRIu64
-                             " level0_slowdown_with_compaction, "
-                             "%" PRIu64
-                             " level0_numfiles, "
-                             "%" PRIu64
-                             " level0_numfiles_with_compaction, "
-                             "%" PRIu64
-                             " stop for pending_compaction_bytes, "
-                             "%" PRIu64
-                             " slowdown for pending_compaction_bytes, "
-                             "%" PRIu64
-                             " memtable_compaction, "
-                             "%" PRIu64
-                             " memtable_slowdown, "
-                             "interval %" PRIu64 " total count\n",
-           cf_stats_count_[LEVEL0_SLOWDOWN_TOTAL],
-           cf_stats_count_[LEVEL0_SLOWDOWN_WITH_COMPACTION],
-           cf_stats_count_[LEVEL0_NUM_FILES_TOTAL],
-           cf_stats_count_[LEVEL0_NUM_FILES_WITH_COMPACTION],
-           cf_stats_count_[HARD_PENDING_COMPACTION_BYTES_LIMIT],
-           cf_stats_count_[SOFT_PENDING_COMPACTION_BYTES_LIMIT],
+  snprintf(buf, sizeof(buf),
+           "Stalls(count): %" PRIu64 " level0_slowdown, "
+           "%" PRIu64 " level0_numfiles, %" PRIu64 " memtable_compaction, "
+           "%" PRIu64 " leveln_slowdown_soft, "
+           "%" PRIu64 " leveln_slowdown_hard\n",
+           cf_stats_count_[LEVEL0_SLOWDOWN],
+           cf_stats_count_[LEVEL0_NUM_FILES],
           cf_stats_count_[MEMTABLE_COMPACTION],
-           cf_stats_count_[MEMTABLE_SLOWDOWN],
-           total_stall_count - cf_stats_snapshot_.stall_count);
+           total_slowdown_count_soft, total_slowdown_count_hard);
  value->append(buf);

  cf_stats_snapshot_.ingest_bytes = curr_ingest;
@@ -38,10 +38,7 @@ enum DBPropertyType : uint32_t {
                                 // in memory that have already been flushed
  kMemtableFlushPending,         // Return 1 if mem table flushing is pending,
                                 // otherwise 0.
-  kNumRunningFlushes,      // Return the number of currently running flushes.
  kCompactionPending,      // Return 1 if a compaction is pending. Otherwise 0.
-  kNumRunningCompactions,  // Return the number of currently running
-                           // compactions.
  kBackgroundErrors,       // Return accumulated background errors encountered.
  kCurSizeActiveMemTable,  // Return current size of the active memtable
  kCurSizeAllMemTables,    // Return current size of unflushed
@@ -52,10 +49,10 @@ enum DBPropertyType : uint32_t {
                                   // memtable.
  kNumEntriesInImmutableMemtable,  // Return sum of number of entries in all
                                   // the immutable mem tables.
-  kNumDeletesInMutableMemtable,    // Return number of deletion entries in the
-                                   // mutable memtable.
-  kNumDeletesInImmutableMemtable,  // Return the total number of deletion
-                                   // entries in all the immutable mem tables.
+  kNumDeletesInMutableMemtable,    // Return number of entries in the mutable
+                                   // memtable.
+  kNumDeletesInImmutableMemtable,  // Return sum of number of deletes in all
+                                   // the immutable mem tables.
  kEstimatedNumKeys,  // Estimated total number of keys in the database.
  kEstimatedUsageByTableReaders,  // Estimated memory by table readers.
  kIsFileDeletionEnabled,         // Equals disable_delete_obsolete_files_,
@@ -83,14 +80,9 @@ extern DBPropertyType GetPropertyType(const Slice& property,
 class InternalStats {
 public:
  enum InternalCFStatsType {
-    LEVEL0_SLOWDOWN_TOTAL,
-    LEVEL0_SLOWDOWN_WITH_COMPACTION,
+    LEVEL0_SLOWDOWN,
    MEMTABLE_COMPACTION,
-    MEMTABLE_SLOWDOWN,
-    LEVEL0_NUM_FILES_TOTAL,
-    LEVEL0_NUM_FILES_WITH_COMPACTION,
-    SOFT_PENDING_COMPACTION_BYTES_LIMIT,
-    HARD_PENDING_COMPACTION_BYTES_LIMIT,
+    LEVEL0_NUM_FILES,
    WRITE_STALLS_ENUM_MAX,
    BYTES_FLUSHED,
    INTERNAL_CF_STATS_ENUM_MAX,
@@ -109,16 +101,30 @@ class InternalStats {
  };

  InternalStats(int num_levels, Env* env, ColumnFamilyData* cfd)
-      : db_stats_{},
-        cf_stats_value_{},
-        cf_stats_count_{},
+      : db_stats_(INTERNAL_DB_STATS_ENUM_MAX),
+        cf_stats_value_(INTERNAL_CF_STATS_ENUM_MAX),
+        cf_stats_count_(INTERNAL_CF_STATS_ENUM_MAX),
        comp_stats_(num_levels),
+        stall_leveln_slowdown_count_hard_(num_levels),
+        stall_leveln_slowdown_count_soft_(num_levels),
        file_read_latency_(num_levels),
        bg_error_count_(0),
        number_levels_(num_levels),
        env_(env),
        cfd_(cfd),
-        started_at_(env->NowMicros()) {}
+        started_at_(env->NowMicros()) {
+    for (int i = 0; i< INTERNAL_DB_STATS_ENUM_MAX; ++i) {
+      db_stats_[i] = 0;
+    }
+    for (int i = 0; i< INTERNAL_CF_STATS_ENUM_MAX; ++i) {
+      cf_stats_value_[i] = 0;
+      cf_stats_count_[i] = 0;
+    }
+    for (int i = 0; i < num_levels; ++i) {
+      stall_leveln_slowdown_count_hard_[i] = 0;
+      stall_leveln_slowdown_count_soft_[i] = 0;
+    }
+  }

  // Per level compaction stats.  comp_stats_[level] stores the stats for
  // compactions that produced data for the specified "level".
@@ -225,19 +231,21 @@ class InternalStats {
    comp_stats_[level].bytes_moved += amount;
  }

+  void RecordLevelNSlowdown(int level, bool soft) {
+    if (soft) {
+      ++stall_leveln_slowdown_count_soft_[level];
+    } else {
+      ++stall_leveln_slowdown_count_hard_[level];
+    }
+  }
+
  void AddCFStats(InternalCFStatsType type, uint64_t value) {
    cf_stats_value_[type] += value;
    ++cf_stats_count_[type];
  }

  void AddDBStats(InternalDBStatsType type, uint64_t value) {
-    auto& v = db_stats_[type];
-    v.store(v.load(std::memory_order_relaxed) + value,
-            std::memory_order_relaxed);
-  }
-
-  uint64_t GetDBStats(InternalDBStatsType type) {
-    return db_stats_[type].load(std::memory_order_relaxed);
+    db_stats_[type] += value;
  }

  HistogramImpl* GetFileReadHist(int level) {
@@ -262,12 +270,15 @@ class InternalStats {
  void DumpCFStats(std::string* value);

  // Per-DB stats
-  std::atomic<uint64_t> db_stats_[INTERNAL_DB_STATS_ENUM_MAX];
+  std::vector<uint64_t> db_stats_;
  // Per-ColumnFamily stats
-  uint64_t cf_stats_value_[INTERNAL_CF_STATS_ENUM_MAX];
-  uint64_t cf_stats_count_[INTERNAL_CF_STATS_ENUM_MAX];
+  std::vector<uint64_t> cf_stats_value_;
+  std::vector<uint64_t> cf_stats_count_;
  // Per-ColumnFamily/level compaction stats
  std::vector<CompactionStats> comp_stats_;
+  // These count the number of microseconds for which MakeRoomForWrite stalls.
+  std::vector<uint64_t> stall_leveln_slowdown_count_hard_;
+  std::vector<uint64_t> stall_leveln_slowdown_count_soft_;
  std::vector<HistogramImpl> file_read_latency_;

  // Used to compute per-interval statistics
@@ -339,14 +350,9 @@ class InternalStats {
 class InternalStats {
 public:
  enum InternalCFStatsType {
-    LEVEL0_SLOWDOWN_TOTAL,
-    LEVEL0_SLOWDOWN_WITH_COMPACTION,
+    LEVEL0_SLOWDOWN,
    MEMTABLE_COMPACTION,
-    MEMTABLE_SLOWDOWN,
-    LEVEL0_NUM_FILES_TOTAL,
-    LEVEL0_NUM_FILES_WITH_COMPACTION,
-    SOFT_PENDING_COMPACTION_BYTES_LIMIT,
-    HARD_PENDING_COMPACTION_BYTES_LIMIT,
+    LEVEL0_NUM_FILES,
    WRITE_STALLS_ENUM_MAX,
    BYTES_FLUSHED,
    INTERNAL_CF_STATS_ENUM_MAX,
@@ -392,6 +398,8 @@ class InternalStats {

  void IncBytesMoved(int level, uint64_t amount) {}

+  void RecordLevelNSlowdown(int level, bool soft) {}
+
  void AddCFStats(InternalCFStatsType type, uint64_t value) {}

  void AddDBStats(InternalDBStatsType type, uint64_t value) {}
@@ -4,12 +4,10 @@
 //  of patent rights can be found in the PATENTS file in the same directory.

 #include "db/db_impl.h"
-#include "db/db_test_util.h"
 #include "db/dbformat.h"
 #include "db/filename.h"
 #include "db/version_set.h"
 #include "db/write_batch_internal.h"
-#include "memtable/hash_linklist_rep.h"
 #include "rocksdb/cache.h"
 #include "rocksdb/compaction_filter.h"
 #include "rocksdb/db.h"
@@ -24,6 +22,7 @@
 #include "table/block_based_table_factory.h"
 #include "table/plain_table_factory.h"
 #include "util/hash.h"
+#include "util/hash_linklist_rep.h"
 #include "util/logging.h"
 #include "util/mutexlock.h"
 #include "util/rate_limiter.h"
@@ -38,45 +37,119 @@

 namespace rocksdb {

-class EventListenerTest : public DBTestBase {
+class EventListenerTest : public testing::Test {
 public:
-  EventListenerTest() : DBTestBase("/listener_test") {}
+  EventListenerTest() {
+    dbname_ = test::TmpDir() + "/listener_test";
+    EXPECT_OK(DestroyDB(dbname_, Options()));
+    db_ = nullptr;
+    Reopen();
+  }
+
+  ~EventListenerTest() {
+    Close();
+    Options options;
+    options.db_paths.emplace_back(dbname_, 0);
+    options.db_paths.emplace_back(dbname_ + "_2", 0);
+    options.db_paths.emplace_back(dbname_ + "_3", 0);
+    options.db_paths.emplace_back(dbname_ + "_4", 0);
+    EXPECT_OK(DestroyDB(dbname_, options));
+  }
+
+  void CreateColumnFamilies(const std::vector<std::string>& cfs,
+                            const ColumnFamilyOptions* options = nullptr) {
+    ColumnFamilyOptions cf_opts;
+    cf_opts = ColumnFamilyOptions(Options());
+    size_t cfi = handles_.size();
+    handles_.resize(cfi + cfs.size());
+    for (auto cf : cfs) {
+      ASSERT_OK(db_->CreateColumnFamily(cf_opts, cf, &handles_[cfi++]));
+    }
+  }
+
+  void Close() {
+    for (auto h : handles_) {
+      delete h;
+    }
+    handles_.clear();
+    delete db_;
+    db_ = nullptr;
+  }
+
+  void ReopenWithColumnFamilies(const std::vector<std::string>& cfs,
+                                const Options* options = nullptr) {
+    ASSERT_OK(TryReopenWithColumnFamilies(cfs, options));
+  }
+
+  Status TryReopenWithColumnFamilies(const std::vector<std::string>& cfs,
+                                     const Options* options = nullptr) {
+    Close();
+    Options opts = (options == nullptr) ? Options() : *options;
+    std::vector<const Options*> v_opts(cfs.size(), &opts);
+    return TryReopenWithColumnFamilies(cfs, v_opts);
+  }
+
+  Status TryReopenWithColumnFamilies(
+      const std::vector<std::string>& cfs,
+      const std::vector<const Options*>& options) {
+    Close();
+    EXPECT_EQ(cfs.size(), options.size());
+    std::vector<ColumnFamilyDescriptor> column_families;
+    for (size_t i = 0; i < cfs.size(); ++i) {
+      column_families.push_back(ColumnFamilyDescriptor(cfs[i], *options[i]));
+    }
+    DBOptions db_opts = DBOptions(*options[0]);
+    return DB::Open(db_opts, dbname_, column_families, &handles_, &db_);
+  }
+
+  Status TryReopen(Options* options = nullptr) {
+    Close();
+    Options opts;
+    if (options != nullptr) {
+      opts = *options;
+    } else {
+      opts.create_if_missing = true;
+    }
+
+    return DB::Open(opts, dbname_, &db_);
+  }
+
+  void Reopen(Options* options = nullptr) {
+    ASSERT_OK(TryReopen(options));
+  }
+
+  void CreateAndReopenWithCF(const std::vector<std::string>& cfs,
+                             const Options* options = nullptr) {
+    CreateColumnFamilies(cfs, options);
+    std::vector<std::string> cfs_plus_default = cfs;
+    cfs_plus_default.insert(cfs_plus_default.begin(), kDefaultColumnFamilyName);
+    ReopenWithColumnFamilies(cfs_plus_default, options);
+  }
+
+  DBImpl* dbfull() {
+    return reinterpret_cast<DBImpl*>(db_);
+  }
+
+  Status Put(int cf, const Slice& k, const Slice& v,
+             WriteOptions wo = WriteOptions()) {
+    return db_->Put(wo, handles_[cf], k, v);
+  }
+
+  Status Flush(size_t cf = 0) {
+    FlushOptions opt = FlushOptions();
+    opt.wait = true;
+    if (cf == 0) {
+      return db_->Flush(opt);
+    } else {
+      return db_->Flush(opt, handles_[cf]);
+    }
+  }

  const size_t k110KB = 110 << 10;
-};

-struct TestPropertiesCollector : public rocksdb::TablePropertiesCollector {
-  virtual rocksdb::Status AddUserKey(const rocksdb::Slice& key,
-                                     const rocksdb::Slice& value,
-                                     rocksdb::EntryType type,
-                                     rocksdb::SequenceNumber seq,
-                                     uint64_t file_size) override {
-    return Status::OK();
-  }
-  virtual rocksdb::Status Finish(
-      rocksdb::UserCollectedProperties* properties) override {
-    properties->insert({"0", "1"});
-    return Status::OK();
-  }
-
-  virtual const char* Name() const override {
-    return "TestTablePropertiesCollector";
-  }
-
-  rocksdb::UserCollectedProperties GetReadableProperties() const override {
-    rocksdb::UserCollectedProperties ret;
-    ret["2"] = "3";
-    return ret;
-  }
-};
-
-class TestPropertiesCollectorFactory : public TablePropertiesCollectorFactory {
- public:
-  virtual TablePropertiesCollector* CreateTablePropertiesCollector(
-      TablePropertiesCollectorFactory::Context context) override {
-    return new TestPropertiesCollector;
-  }
-  const char* Name() const override { return "TestTablePropertiesCollector"; }
+  DB* db_;
+  std::string dbname_;
+  std::vector<ColumnFamilyHandle*> handles_;
 };

 class TestCompactionListener : public EventListener {
@@ -88,16 +161,6 @@ class TestCompactionListener : public EventListener {
    ASSERT_GT(ci.output_files.size(), 0U);
    ASSERT_EQ(db->GetEnv()->GetThreadID(), ci.thread_id);
    ASSERT_GT(ci.thread_id, 0U);
-
-    for (auto fl : {ci.input_files, ci.output_files}) {
-      for (auto fn : fl) {
-        auto it = ci.table_properties.find(fn);
-        ASSERT_NE(it, ci.table_properties.end());
-        auto tp = it->second;
-        ASSERT_TRUE(tp != nullptr);
-        ASSERT_EQ(tp->user_collected_properties.find("0")->second, "1");
-      }
-    }
  }

  std::vector<DB*> compacted_dbs_;
@@ -123,15 +186,13 @@ TEST_F(EventListenerTest, OnSingleDBCompactionTest) {
  options.enable_thread_tracking = true;
 #endif  // ROCKSDB_USING_THREAD_STATUS
  options.level0_file_num_compaction_trigger = kNumL0Files;
-  options.table_properties_collector_factories.push_back(
-      std::make_shared<TestPropertiesCollectorFactory>());

  TestCompactionListener* listener = new TestCompactionListener();
  options.listeners.emplace_back(listener);
  std::vector<std::string> cf_names = {
      "pikachu", "ilya", "muromec", "dobrynia",
      "nikitich", "alyosha", "popovich"};
-  CreateAndReopenWithCF(cf_names, options);
+  CreateAndReopenWithCF(cf_names, &options);
  ASSERT_OK(Put(1, "pikachu", std::string(90000, 'p')));
  ASSERT_OK(Put(2, "ilya", std::string(90000, 'i')));
  ASSERT_OK(Put(3, "muromec", std::string(90000, 'm')));
@@ -139,12 +200,12 @@ TEST_F(EventListenerTest, OnSingleDBCompactionTest) {
  ASSERT_OK(Put(5, "nikitich", std::string(90000, 'n')));
  ASSERT_OK(Put(6, "alyosha", std::string(90000, 'a')));
  ASSERT_OK(Put(7, "popovich", std::string(90000, 'p')));
-  for (int i = 1; i < 8; ++i) {
+  for (size_t i = 1; i < 8; ++i) {
    ASSERT_OK(Flush(i));
-    const Slice kRangeStart = "a";
-    const Slice kRangeEnd = "z";
+    const Slice kStart = "a";
+    const Slice kEnd = "z";
    ASSERT_OK(dbfull()->CompactRange(CompactRangeOptions(), handles_[i],
-                                     &kRangeStart, &kRangeEnd));
+                                     &kStart, &kEnd));
    dbfull()->TEST_WaitForFlushMemTable();
    dbfull()->TEST_WaitForCompact();
  }
@@ -213,8 +274,6 @@ class TestFlushListener : public EventListener {
    ASSERT_EQ(prev_fc_info_.file_path, info.file_path);
    ASSERT_EQ(db->GetEnv()->GetThreadID(), info.thread_id);
    ASSERT_GT(info.thread_id, 0U);
-    ASSERT_EQ(info.table_properties.user_collected_properties.find("0")->second,
-              "1");
  }

  std::vector<std::string> flushed_column_family_names_;
@@ -240,9 +299,7 @@ TEST_F(EventListenerTest, OnSingleDBFlushTest) {
  std::vector<std::string> cf_names = {
      "pikachu", "ilya", "muromec", "dobrynia",
      "nikitich", "alyosha", "popovich"};
-  options.table_properties_collector_factories.push_back(
-      std::make_shared<TestPropertiesCollectorFactory>());
-  CreateAndReopenWithCF(cf_names, options);
+  CreateAndReopenWithCF(cf_names, &options);

  ASSERT_OK(Put(1, "pikachu", std::string(90000, 'p')));
  ASSERT_OK(Put(2, "ilya", std::string(90000, 'i')));
@@ -251,7 +308,7 @@ TEST_F(EventListenerTest, OnSingleDBFlushTest) {
  ASSERT_OK(Put(5, "nikitich", std::string(90000, 'n')));
  ASSERT_OK(Put(6, "alyosha", std::string(90000, 'a')));
  ASSERT_OK(Put(7, "popovich", std::string(90000, 'p')));
-  for (int i = 1; i < 8; ++i) {
+  for (size_t i = 1; i < 8; ++i) {
    ASSERT_OK(Flush(i));
    dbfull()->TEST_WaitForFlushMemTable();
    ASSERT_EQ(listener->flushed_dbs_.size(), i);
@@ -273,12 +330,10 @@ TEST_F(EventListenerTest, MultiCF) {
 #endif  // ROCKSDB_USING_THREAD_STATUS
  TestFlushListener* listener = new TestFlushListener(options.env);
  options.listeners.emplace_back(listener);
-  options.table_properties_collector_factories.push_back(
-      std::make_shared<TestPropertiesCollectorFactory>());
  std::vector<std::string> cf_names = {
      "pikachu", "ilya", "muromec", "dobrynia",
      "nikitich", "alyosha", "popovich"};
-  CreateAndReopenWithCF(cf_names, options);
+  CreateAndReopenWithCF(cf_names, &options);

  ASSERT_OK(Put(1, "pikachu", std::string(90000, 'p')));
  ASSERT_OK(Put(2, "ilya", std::string(90000, 'i')));
@@ -287,7 +342,7 @@ TEST_F(EventListenerTest, MultiCF) {
  ASSERT_OK(Put(5, "nikitich", std::string(90000, 'n')));
  ASSERT_OK(Put(6, "alyosha", std::string(90000, 'a')));
  ASSERT_OK(Put(7, "popovich", std::string(90000, 'p')));
-  for (int i = 1; i < 8; ++i) {
+  for (size_t i = 1; i < 8; ++i) {
    ASSERT_OK(Flush(i));
    ASSERT_EQ(listener->flushed_dbs_.size(), i);
    ASSERT_EQ(listener->flushed_column_family_names_.size(), i);
@@ -305,8 +360,6 @@ TEST_F(EventListenerTest, MultiDBMultiListeners) {
 #if ROCKSDB_USING_THREAD_STATUS
  options.enable_thread_tracking = true;
 #endif  // ROCKSDB_USING_THREAD_STATUS
-  options.table_properties_collector_factories.push_back(
-      std::make_shared<TestPropertiesCollectorFactory>());
  std::vector<TestFlushListener*> listeners;
  const int kNumDBs = 5;
  const int kNumListeners = 10;
@@ -401,10 +454,8 @@ TEST_F(EventListenerTest, DisableBGCompaction) {
  options.compaction_style = kCompactionStyleNone;
  options.compression = kNoCompression;
  options.write_buffer_size = 100000;  // Small write buffer
-  options.table_properties_collector_factories.push_back(
-      std::make_shared<TestPropertiesCollectorFactory>());

-  CreateAndReopenWithCF({"pikachu"}, options);
+  CreateAndReopenWithCF({"pikachu"}, &options);
  ColumnFamilyMetaData cf_meta;
  db_->GetColumnFamilyMetaData(handles_[1], &cf_meta);

@@ -418,167 +469,6 @@ TEST_F(EventListenerTest, DisableBGCompaction) {
  ASSERT_GE(listener->slowdown_count, kSlowdownTrigger * 9);
 }

-class TestCompactionReasonListener : public EventListener {
- public:
-  void OnCompactionCompleted(DB* db, const CompactionJobInfo& ci) override {
-    std::lock_guard<std::mutex> lock(mutex_);
-    compaction_reasons_.push_back(ci.compaction_reason);
-  }
-
-  std::vector<CompactionReason> compaction_reasons_;
-  std::mutex mutex_;
-};
-
-TEST_F(EventListenerTest, CompactionReasonLevel) {
-  Options options;
-  options.create_if_missing = true;
-  options.memtable_factory.reset(
-      new SpecialSkipListFactory(DBTestBase::kNumKeysByGenerateNewRandomFile));
-
-  TestCompactionReasonListener* listener = new TestCompactionReasonListener();
-  options.listeners.emplace_back(listener);
-
-  options.level0_file_num_compaction_trigger = 4;
-  options.compaction_style = kCompactionStyleLevel;
-
-  DestroyAndReopen(options);
-  Random rnd(301);
-
-  // Write 4 files in L0
-  for (int i = 0; i < 4; i++) {
-    GenerateNewRandomFile(&rnd);
-  }
-  dbfull()->TEST_WaitForCompact();
-
-  ASSERT_EQ(listener->compaction_reasons_.size(), 1);
-  ASSERT_EQ(listener->compaction_reasons_[0],
-            CompactionReason::kLevelL0FilesNum);
-
-  DestroyAndReopen(options);
-
-  // Write 3 non-overlapping files in L0
-  for (int k = 1; k <= 30; k++) {
-    ASSERT_OK(Put(Key(k), Key(k)));
-    if (k % 10 == 0) {
-      Flush();
-    }
-  }
-
-  // Do a trivial move from L0 -> L1
-  db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
-
-  options.max_bytes_for_level_base = 1;
-  Close();
-  listener->compaction_reasons_.clear();
-  Reopen(options);
-
-  dbfull()->TEST_WaitForCompact();
-  ASSERT_GT(listener->compaction_reasons_.size(), 1);
-
-  for (auto compaction_reason : listener->compaction_reasons_) {
-    ASSERT_EQ(compaction_reason, CompactionReason::kLevelMaxLevelSize);
-  }
-
-  options.disable_auto_compactions = true;
-  Close();
-  listener->compaction_reasons_.clear();
-  Reopen(options);
-
-  db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
-  ASSERT_GT(listener->compaction_reasons_.size(), 0);
-  for (auto compaction_reason : listener->compaction_reasons_) {
-    ASSERT_EQ(compaction_reason, CompactionReason::kManualCompaction);
-  }
-}
-
-TEST_F(EventListenerTest, CompactionReasonUniversal) {
-  Options options;
-  options.create_if_missing = true;
-  options.memtable_factory.reset(
-      new SpecialSkipListFactory(DBTestBase::kNumKeysByGenerateNewRandomFile));
-
-  TestCompactionReasonListener* listener = new TestCompactionReasonListener();
-  options.listeners.emplace_back(listener);
-
-  options.compaction_style = kCompactionStyleUniversal;
-
-  Random rnd(301);
-
-  options.level0_file_num_compaction_trigger = 8;
-  options.compaction_options_universal.max_size_amplification_percent = 100000;
-  options.compaction_options_universal.size_ratio = 100000;
-  DestroyAndReopen(options);
-  listener->compaction_reasons_.clear();
-
-  // Write 8 files in L0
-  for (int i = 0; i < 8; i++) {
-    GenerateNewRandomFile(&rnd);
-  }
-  dbfull()->TEST_WaitForCompact();
-
-  ASSERT_GT(listener->compaction_reasons_.size(), 0);
-  for (auto compaction_reason : listener->compaction_reasons_) {
-    ASSERT_EQ(compaction_reason, CompactionReason::kUniversalSortedRunNum);
-  }
-
-  options.level0_file_num_compaction_trigger = 8;
-  options.compaction_options_universal.max_size_amplification_percent = 1;
-  options.compaction_options_universal.size_ratio = 100000;
-
-  DestroyAndReopen(options);
-  listener->compaction_reasons_.clear();
-
-  // Write 8 files in L0
-  for (int i = 0; i < 8; i++) {
-    GenerateNewRandomFile(&rnd);
-  }
-  dbfull()->TEST_WaitForCompact();
-
-  ASSERT_GT(listener->compaction_reasons_.size(), 0);
-  for (auto compaction_reason : listener->compaction_reasons_) {
-    ASSERT_EQ(compaction_reason, CompactionReason::kUniversalSizeAmplification);
-  }
-
-  options.disable_auto_compactions = true;
-  Close();
-  listener->compaction_reasons_.clear();
-  Reopen(options);
-
-  db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
-
-  ASSERT_GT(listener->compaction_reasons_.size(), 0);
-  for (auto compaction_reason : listener->compaction_reasons_) {
-    ASSERT_EQ(compaction_reason, CompactionReason::kManualCompaction);
-  }
-}
-
-TEST_F(EventListenerTest, CompactionReasonFIFO) {
-  Options options;
-  options.create_if_missing = true;
-  options.memtable_factory.reset(
-      new SpecialSkipListFactory(DBTestBase::kNumKeysByGenerateNewRandomFile));
-
-  TestCompactionReasonListener* listener = new TestCompactionReasonListener();
-  options.listeners.emplace_back(listener);
-
-  options.level0_file_num_compaction_trigger = 4;
-  options.compaction_style = kCompactionStyleFIFO;
-  options.compaction_options_fifo.max_table_files_size = 1;
-
-  DestroyAndReopen(options);
-  Random rnd(301);
-
-  // Write 4 files in L0
-  for (int i = 0; i < 4; i++) {
-    GenerateNewRandomFile(&rnd);
-  }
-  dbfull()->TEST_WaitForCompact();
-
-  ASSERT_GT(listener->compaction_reasons_.size(), 0);
-  for (auto compaction_reason : listener->compaction_reasons_) {
-    ASSERT_EQ(compaction_reason, CompactionReason::kFIFOMaxSize);
-  }
-}
 }  // namespace rocksdb

 #endif  // ROCKSDB_LITE
@@ -587,3 +477,4 @@ int main(int argc, char** argv) {
  ::testing::InitGoogleTest(&argc, argv);
  return RUN_ALL_TESTS();
 }
+
@@ -22,24 +22,14 @@ enum RecordType {
  // For fragments
  kFirstType = 2,
  kMiddleType = 3,
-  kLastType = 4,
-
-  // For recycled log files
-  kRecyclableFullType = 5,
-  kRecyclableFirstType = 6,
-  kRecyclableMiddleType = 7,
-  kRecyclableLastType = 8,
+  kLastType = 4
 };
-static const int kMaxRecordType = kRecyclableLastType;
+static const int kMaxRecordType = kLastType;

 static const unsigned int kBlockSize = 32768;

 // Header is checksum (4 bytes), type (1 byte), length (2 bytes).
 static const int kHeaderSize = 4 + 1 + 2;

-// Recyclable header is checksum (4 bytes), type (1 byte), log number
-// (4 bytes), length (2 bytes).
-static const int kRecyclableHeaderSize = 4 + 1 + 4 + 2;
-
 }  // namespace log
 }  // namespace rocksdb
@@ -21,12 +21,9 @@ namespace log {
 Reader::Reporter::~Reporter() {
 }

-Reader::Reader(std::shared_ptr<Logger> info_log,
-	       unique_ptr<SequentialFileReader>&& _file,
-               Reporter* reporter, bool checksum, uint64_t initial_offset,
-               uint64_t log_num)
-    : info_log_(info_log),
-      file_(std::move(_file)),
+Reader::Reader(unique_ptr<SequentialFileReader>&& _file, Reporter* reporter,
+               bool checksum, uint64_t initial_offset)
+    : file_(std::move(_file)),
      reporter_(reporter),
      checksum_(checksum),
      backing_store_(new char[kBlockSize]),
@@ -36,8 +33,7 @@ Reader::Reader(std::shared_ptr<Logger> info_log,
      eof_offset_(0),
      last_record_offset_(0),
      end_of_buffer_offset_(0),
-      initial_offset_(initial_offset),
-      log_number_(log_num) {}
+      initial_offset_(initial_offset) {}

 Reader::~Reader() {
  delete[] backing_store_;
@@ -66,15 +62,8 @@ bool Reader::SkipToInitialBlock() {
  return true;
 }

-// For kAbsoluteConsistency, on clean shutdown we don't expect any error
-// in the log files.  For other modes, we can ignore only incomplete records
-// in the last log file, which are presumably due to a write in progress
-// during restart (or from log recycling).
-//
-// TODO krad: Evaluate if we need to move to a more strict mode where we
-// restrict the inconsistency to only the last log
 bool Reader::ReadRecord(Slice* record, std::string* scratch,
-                        WALRecoveryMode wal_recovery_mode) {
+                        const bool report_eof_inconsistency) {
  if (last_record_offset_ < initial_offset_) {
    if (!SkipToInitialBlock()) {
      return false;
@@ -91,11 +80,10 @@ bool Reader::ReadRecord(Slice* record, std::string* scratch,
  Slice fragment;
  while (true) {
    uint64_t physical_record_offset = end_of_buffer_offset_ - buffer_.size();
-    size_t drop_size;
-    const unsigned int record_type = ReadPhysicalRecord(&fragment, &drop_size);
+    const unsigned int record_type =
+        ReadPhysicalRecord(&fragment, report_eof_inconsistency);
    switch (record_type) {
      case kFullType:
-      case kRecyclableFullType:
        if (in_fragmented_record && !scratch->empty()) {
          // Handle bug in earlier versions of log::Writer where
          // it could emit an empty kFirstType record at the tail end
@@ -110,7 +98,6 @@ bool Reader::ReadRecord(Slice* record, std::string* scratch,
        return true;

      case kFirstType:
-      case kRecyclableFirstType:
        if (in_fragmented_record && !scratch->empty()) {
          // Handle bug in earlier versions of log::Writer where
          // it could emit an empty kFirstType record at the tail end
@@ -124,7 +111,6 @@ bool Reader::ReadRecord(Slice* record, std::string* scratch,
        break;

      case kMiddleType:
-      case kRecyclableMiddleType:
        if (!in_fragmented_record) {
          ReportCorruption(fragment.size(),
                           "missing start of fragmented record(1)");
@@ -134,7 +120,6 @@ bool Reader::ReadRecord(Slice* record, std::string* scratch,
        break;

      case kLastType:
-      case kRecyclableLastType:
        if (!in_fragmented_record) {
          ReportCorruption(fragment.size(),
                           "missing start of fragmented record(2)");
@@ -146,17 +131,9 @@ bool Reader::ReadRecord(Slice* record, std::string* scratch,
        }
        break;

-      case kBadHeader:
-        if (wal_recovery_mode == WALRecoveryMode::kAbsoluteConsistency) {
-          // in clean shutdown we don't expect any error in the log files
-          ReportCorruption(drop_size, "truncated header");
-        }
-      // fall-thru
-
      case kEof:
        if (in_fragmented_record) {
-          if (wal_recovery_mode == WALRecoveryMode::kAbsoluteConsistency) {
-            // in clean shutdown we don't expect any error in the log files
+          if (report_eof_inconsistency) {
            ReportCorruption(scratch->size(), "error reading trailing data");
          }
          // This can be caused by the writer dying immediately after
@@ -166,23 +143,6 @@ bool Reader::ReadRecord(Slice* record, std::string* scratch,
        }
        return false;

-      case kOldRecord:
-        if (wal_recovery_mode != WALRecoveryMode::kSkipAnyCorruptedRecords) {
-          // Treat a record from a previous instance of the log as EOF.
-          if (in_fragmented_record) {
-            if (wal_recovery_mode == WALRecoveryMode::kAbsoluteConsistency) {
-              // in clean shutdown we don't expect any error in the log files
-              ReportCorruption(scratch->size(), "error reading trailing data");
-            }
-            // This can be caused by the writer dying immediately after
-            //  writing a physical record but before completing the next; don't
-            //  treat it as a corruption, just ignore the entire logical record.
-            scratch->clear();
-          }
-          return false;
-        }
-      // fall-thru
-
      case kBadRecord:
        if (in_fragmented_record) {
          ReportCorruption(scratch->size(), "error in middle of record");
@@ -284,49 +244,36 @@ void Reader::ReportDrop(size_t bytes, const Status& reason) {
  }
 }

-bool Reader::ReadMore(size_t* drop_size, int *error) {
-  if (!eof_ && !read_error_) {
-    // Last read was a full read, so this is a trailer to skip
-    buffer_.clear();
-    Status status = file_->Read(kBlockSize, &buffer_, backing_store_);
-    end_of_buffer_offset_ += buffer_.size();
-    if (!status.ok()) {
-      buffer_.clear();
-      ReportDrop(kBlockSize, status);
-      read_error_ = true;
-      *error = kEof;
-      return false;
-    } else if (buffer_.size() < (size_t)kBlockSize) {
-      eof_ = true;
-      eof_offset_ = buffer_.size();
-    }
-    return true;
-  } else {
-    // Note that if buffer_ is non-empty, we have a truncated header at the
-    //  end of the file, which can be caused by the writer crashing in the
-    //  middle of writing the header. Unless explicitly requested we don't
-    //  considering this an error, just report EOF.
-    if (buffer_.size()) {
-      *drop_size = buffer_.size();
-      buffer_.clear();
-      *error = kBadHeader;
-      return false;
-    }
-    buffer_.clear();
-    *error = kEof;
-    return false;
-  }
-}
-
-unsigned int Reader::ReadPhysicalRecord(Slice* result, size_t* drop_size) {
+unsigned int Reader::ReadPhysicalRecord(Slice* result,
+                                        const bool report_eof_inconsistency) {
  while (true) {
-    // We need at least the minimum header size
    if (buffer_.size() < (size_t)kHeaderSize) {
-      int r;
-      if (!ReadMore(drop_size, &r)) {
-	return r;
+      if (!eof_ && !read_error_) {
+        // Last read was a full read, so this is a trailer to skip
+        buffer_.clear();
+        Status status = file_->Read(kBlockSize, &buffer_, backing_store_);
+        end_of_buffer_offset_ += buffer_.size();
+        if (!status.ok()) {
+          buffer_.clear();
+          ReportDrop(kBlockSize, status);
+          read_error_ = true;
+          return kEof;
+        } else if (buffer_.size() < (size_t)kBlockSize) {
+          eof_ = true;
+          eof_offset_ = buffer_.size();
+        }
+        continue;
+      } else {
+        // Note that if buffer_ is non-empty, we have a truncated header at the
+        //  end of the file, which can be caused by the writer crashing in the
+        //  middle of writing the header. Unless explicitly requested we don't
+        //  considering this an error, just report EOF.
+        if (buffer_.size() && report_eof_inconsistency) {
+          ReportCorruption(buffer_.size(), "truncated header");
+        }
+        buffer_.clear();
+        return kEof;
      }
-      continue;
    }

    // Parse the header
@@ -335,34 +282,18 @@ unsigned int Reader::ReadPhysicalRecord(Slice* result, size_t* drop_size) {
    const uint32_t b = static_cast<uint32_t>(header[5]) & 0xff;
    const unsigned int type = header[6];
    const uint32_t length = a | (b << 8);
-    int header_size = kHeaderSize;
-    if (type >= kRecyclableFullType && type <= kRecyclableLastType) {
-      header_size = kRecyclableHeaderSize;
-      // We need enough for the larger header
-      if (buffer_.size() < (size_t)kRecyclableHeaderSize) {
-	int r;
-	if (!ReadMore(drop_size, &r)) {
-          return r;
-        }
-	continue;
-      }
-      const uint32_t log_num = DecodeFixed32(header + 7);
-      if (log_num != log_number_) {
-        return kOldRecord;
-      }
-    }
-    if (header_size + length > buffer_.size()) {
-      *drop_size = buffer_.size();
+    if (kHeaderSize + length > buffer_.size()) {
+      size_t drop_size = buffer_.size();
      buffer_.clear();
      if (!eof_) {
-        ReportCorruption(*drop_size, "bad record length");
+        ReportCorruption(drop_size, "bad record length");
        return kBadRecord;
      }
      // If the end of the file has been reached without reading |length| bytes
      // of payload, assume the writer died in the middle of writing the record.
      // Don't report a corruption unless requested.
-      if (*drop_size) {
-        return kBadHeader;
+      if (drop_size && report_eof_inconsistency) {
+        ReportCorruption(drop_size, "truncated header");
      }
      return kEof;
    }
@@ -380,29 +311,29 @@ unsigned int Reader::ReadPhysicalRecord(Slice* result, size_t* drop_size) {
    // Check crc
    if (checksum_) {
      uint32_t expected_crc = crc32c::Unmask(DecodeFixed32(header));
-      uint32_t actual_crc = crc32c::Value(header + 6, length + header_size - 6);
+      uint32_t actual_crc = crc32c::Value(header + 6, 1 + length);
      if (actual_crc != expected_crc) {
        // Drop the rest of the buffer since "length" itself may have
        // been corrupted and if we trust it, we could find some
        // fragment of a real log record that just happens to look
        // like a valid log record.
-        *drop_size = buffer_.size();
+        size_t drop_size = buffer_.size();
        buffer_.clear();
-        ReportCorruption(*drop_size, "checksum mismatch");
+        ReportCorruption(drop_size, "checksum mismatch");
        return kBadRecord;
      }
    }

-    buffer_.remove_prefix(header_size + length);
+    buffer_.remove_prefix(kHeaderSize + length);

    // Skip physical record that started before initial_offset_
-    if (end_of_buffer_offset_ - buffer_.size() - header_size - length <
+    if (end_of_buffer_offset_ - buffer_.size() - kHeaderSize - length <
        initial_offset_) {
      result->clear();
      return kBadRecord;
    }

-    *result = Slice(header + header_size, length);
+    *result = Slice(header + kHeaderSize, length);
    return type;
  }
 }
@@ -14,12 +14,10 @@
 #include "db/log_format.h"
 #include "rocksdb/slice.h"
 #include "rocksdb/status.h"
-#include "rocksdb/options.h"

 namespace rocksdb {

 class SequentialFileReader;
-class Logger;
 using std::unique_ptr;

 namespace log {
@@ -53,10 +51,8 @@ class Reader {
  //
  // The Reader will start reading at the first record located at physical
  // position >= initial_offset within the file.
-  Reader(std::shared_ptr<Logger> info_log,
-	 unique_ptr<SequentialFileReader>&& file,
-         Reporter* reporter, bool checksum, uint64_t initial_offset,
-         uint64_t log_num);
+  Reader(unique_ptr<SequentialFileReader>&& file, Reporter* reporter,
+         bool checksum, uint64_t initial_offset);

  ~Reader();

@@ -66,8 +62,7 @@ class Reader {
  // will only be valid until the next mutating operation on this
  // reader or the next mutation to *scratch.
  bool ReadRecord(Slice* record, std::string* scratch,
-                  WALRecoveryMode wal_recovery_mode =
-                      WALRecoveryMode::kTolerateCorruptedTailRecords);
+                  bool report_eof_inconsistency = false);

  // Returns the physical offset of the last record returned by ReadRecord.
  //
@@ -89,7 +84,6 @@ class Reader {
  SequentialFileReader* file() { return file_.get(); }

 private:
-  std::shared_ptr<Logger> info_log_;
  const unique_ptr<SequentialFileReader> file_;
  Reporter* const reporter_;
  bool const checksum_;
@@ -110,9 +104,6 @@ class Reader {
  // Offset at which to start looking for the first record to return
  uint64_t const initial_offset_;

-  // which log number this is
-  uint64_t const log_number_;
-
  // Extend record types with the following special values
  enum {
    kEof = kMaxRecordType + 1,
@@ -121,11 +112,7 @@ class Reader {
    // * The record has an invalid CRC (ReadPhysicalRecord reports a drop)
    // * The record is a 0-length record (No drop is reported)
    // * The record is below constructor's initial_offset (No drop is reported)
-    kBadRecord = kMaxRecordType + 2,
-    // Returned when we fail to read a valid header.
-    kBadHeader = kMaxRecordType + 3,
-    // Returned when we read an old record from a previous user of the log.
-    kOldRecord = kMaxRecordType + 4,
+    kBadRecord = kMaxRecordType + 2
  };

  // Skips all blocks that are completely before "initial_offset_".
@@ -134,10 +121,8 @@ class Reader {
  bool SkipToInitialBlock();

  // Return type, or one of the preceding special values
-  unsigned int ReadPhysicalRecord(Slice* result, size_t* drop_size);
-
-  // Read some more
-  bool ReadMore(size_t* drop_size, int *error);
+  unsigned int ReadPhysicalRecord(Slice* result,
+                                  bool report_eof_inconsistency = false);

  // Reports dropped bytes to the reporter.
  // buffer_ must be updated to remove the dropped bytes prior to invocation.
@@ -43,7 +43,7 @@ static std::string RandomSkewedString(int i, Random* rnd) {
  return BigString(NumberString(i), rnd->Skewed(17));
 }

-class LogTest : public ::testing::TestWithParam<int> {
+class LogTest : public testing::Test {
 private:
  class StringSource : public SequentialFile {
   public:
@@ -153,26 +153,19 @@ class LogTest : public ::testing::TestWithParam<int> {

  // Record metadata for testing initial offset functionality
  static size_t initial_offset_record_sizes_[];
-  uint64_t initial_offset_last_record_offsets_[4];
+  static uint64_t initial_offset_last_record_offsets_[];

 public:
  LogTest()
      : reader_contents_(),
-        dest_holder_(test::GetWritableFileWriter(
-            new test::StringSink(&reader_contents_))),
+        dest_holder_(
+            test::GetWritableFileWriter(
+              new test::StringSink(&reader_contents_))),
        source_holder_(
            test::GetSequentialFileReader(new StringSource(reader_contents_))),
-        writer_(std::move(dest_holder_), 123, GetParam()),
-        reader_(NULL, std::move(source_holder_), &report_, true /*checksum*/,
-                0 /*initial_offset*/, 123) {
-    int header_size = GetParam() ? kRecyclableHeaderSize : kHeaderSize;
-    initial_offset_last_record_offsets_[0] = 0;
-    initial_offset_last_record_offsets_[1] = header_size + 10000;
-    initial_offset_last_record_offsets_[2] = 2 * (header_size + 10000);
-    initial_offset_last_record_offsets_[3] = 2 * (header_size + 10000) +
-                                             (2 * log::kBlockSize - 1000) +
-                                             3 * header_size;
-  }
+        writer_(std::move(dest_holder_)),
+        reader_(std::move(source_holder_), &report_, true /*checksum*/,
+                0 /*initial_offset*/) {}

  void Write(const std::string& msg) {
    writer_.AddRecord(Slice(msg));
@@ -182,11 +175,10 @@ class LogTest : public ::testing::TestWithParam<int> {
    return dest_contents().size();
  }

-  std::string Read(const WALRecoveryMode wal_recovery_mode =
-                       WALRecoveryMode::kTolerateCorruptedTailRecords) {
+  std::string Read(const bool report_eof_inconsistency = false) {
    std::string scratch;
    Slice record;
-    if (reader_.ReadRecord(&record, &scratch, wal_recovery_mode)) {
+    if (reader_.ReadRecord(&record, &scratch, report_eof_inconsistency)) {
      return record.ToString();
    } else {
      return "EOF";
@@ -208,11 +200,9 @@ class LogTest : public ::testing::TestWithParam<int> {
    dest->Drop(bytes);
  }

-  void FixChecksum(int header_offset, int len, bool recyclable) {
+  void FixChecksum(int header_offset, int len) {
    // Compute crc of type/len/data
-    int header_size = recyclable ? kRecyclableHeaderSize : kHeaderSize;
-    uint32_t crc = crc32c::Value(&dest_contents()[header_offset + 6],
-                                 header_size - 6 + len);
+    uint32_t crc = crc32c::Value(&dest_contents()[header_offset+6], 1 + len);
    crc = crc32c::Mask(crc);
    EncodeFixed32(&dest_contents()[header_offset], crc);
  }
@@ -269,8 +259,8 @@ class LogTest : public ::testing::TestWithParam<int> {
    unique_ptr<SequentialFileReader> file_reader(
        test::GetSequentialFileReader(new StringSource(reader_contents_)));
    unique_ptr<Reader> offset_reader(
-        new Reader(NULL, std::move(file_reader), &report_,
-                   true /*checksum*/, WrittenBytes() + offset_past_end, 123));
+        new Reader(std::move(file_reader), &report_, true /*checksum*/,
+                   WrittenBytes() + offset_past_end));
    Slice record;
    std::string scratch;
    ASSERT_TRUE(!offset_reader->ReadRecord(&record, &scratch));
@@ -281,9 +271,8 @@ class LogTest : public ::testing::TestWithParam<int> {
    WriteInitialOffsetLog();
    unique_ptr<SequentialFileReader> file_reader(
        test::GetSequentialFileReader(new StringSource(reader_contents_)));
-    unique_ptr<Reader> offset_reader(
-        new Reader(NULL, std::move(file_reader), &report_,
-                   true /*checksum*/, initial_offset, 123));
+    unique_ptr<Reader> offset_reader(new Reader(
+        std::move(file_reader), &report_, true /*checksum*/, initial_offset));
    Slice record;
    std::string scratch;
    ASSERT_TRUE(offset_reader->ReadRecord(&record, &scratch));
@@ -302,9 +291,16 @@ size_t LogTest::initial_offset_record_sizes_[] =
     2 * log::kBlockSize - 1000,  // Span three blocks
     1};

-TEST_P(LogTest, Empty) { ASSERT_EQ("EOF", Read()); }
+uint64_t LogTest::initial_offset_last_record_offsets_[] =
+    {0,
+     kHeaderSize + 10000,
+     2 * (kHeaderSize + 10000),
+     2 * (kHeaderSize + 10000) +
+         (2 * log::kBlockSize - 1000) + 3 * kHeaderSize};

-TEST_P(LogTest, ReadWrite) {
+TEST_F(LogTest, Empty) { ASSERT_EQ("EOF", Read()); }
+
+TEST_F(LogTest, ReadWrite) {
  Write("foo");
  Write("bar");
  Write("");
@@ -317,7 +313,7 @@ TEST_P(LogTest, ReadWrite) {
  ASSERT_EQ("EOF", Read());  // Make sure reads at eof work
 }

-TEST_P(LogTest, ManyBlocks) {
+TEST_F(LogTest, ManyBlocks) {
  for (int i = 0; i < 100000; i++) {
    Write(NumberString(i));
  }
@@ -327,7 +323,7 @@ TEST_P(LogTest, ManyBlocks) {
  ASSERT_EQ("EOF", Read());
 }

-TEST_P(LogTest, Fragmentation) {
+TEST_F(LogTest, Fragmentation) {
  Write("small");
  Write(BigString("medium", 50000));
  Write(BigString("large", 100000));
@@ -337,12 +333,11 @@ TEST_P(LogTest, Fragmentation) {
  ASSERT_EQ("EOF", Read());
 }

-TEST_P(LogTest, MarginalTrailer) {
+TEST_F(LogTest, MarginalTrailer) {
  // Make a trailer that is exactly the same length as an empty record.
-  int header_size = GetParam() ? kRecyclableHeaderSize : kHeaderSize;
-  const int n = kBlockSize - 2 * header_size;
+  const int n = kBlockSize - 2*kHeaderSize;
  Write(BigString("foo", n));
-  ASSERT_EQ((unsigned int)(kBlockSize - header_size), WrittenBytes());
+  ASSERT_EQ((unsigned int)(kBlockSize - kHeaderSize), WrittenBytes());
  Write("");
  Write("bar");
  ASSERT_EQ(BigString("foo", n), Read());
@@ -351,12 +346,11 @@ TEST_P(LogTest, MarginalTrailer) {
  ASSERT_EQ("EOF", Read());
 }

-TEST_P(LogTest, MarginalTrailer2) {
+TEST_F(LogTest, MarginalTrailer2) {
  // Make a trailer that is exactly the same length as an empty record.
-  int header_size = GetParam() ? kRecyclableHeaderSize : kHeaderSize;
-  const int n = kBlockSize - 2 * header_size;
+  const int n = kBlockSize - 2*kHeaderSize;
  Write(BigString("foo", n));
-  ASSERT_EQ((unsigned int)(kBlockSize - header_size), WrittenBytes());
+  ASSERT_EQ((unsigned int)(kBlockSize - kHeaderSize), WrittenBytes());
  Write("bar");
  ASSERT_EQ(BigString("foo", n), Read());
  ASSERT_EQ("bar", Read());
@@ -365,11 +359,10 @@ TEST_P(LogTest, MarginalTrailer2) {
  ASSERT_EQ("", ReportMessage());
 }

-TEST_P(LogTest, ShortTrailer) {
-  int header_size = GetParam() ? kRecyclableHeaderSize : kHeaderSize;
-  const int n = kBlockSize - 2 * header_size + 4;
+TEST_F(LogTest, ShortTrailer) {
+  const int n = kBlockSize - 2*kHeaderSize + 4;
  Write(BigString("foo", n));
-  ASSERT_EQ((unsigned int)(kBlockSize - header_size + 4), WrittenBytes());
+  ASSERT_EQ((unsigned int)(kBlockSize - kHeaderSize + 4), WrittenBytes());
  Write("");
  Write("bar");
  ASSERT_EQ(BigString("foo", n), Read());
@@ -378,16 +371,15 @@ TEST_P(LogTest, ShortTrailer) {
  ASSERT_EQ("EOF", Read());
 }

-TEST_P(LogTest, AlignedEof) {
-  int header_size = GetParam() ? kRecyclableHeaderSize : kHeaderSize;
-  const int n = kBlockSize - 2 * header_size + 4;
+TEST_F(LogTest, AlignedEof) {
+  const int n = kBlockSize - 2*kHeaderSize + 4;
  Write(BigString("foo", n));
-  ASSERT_EQ((unsigned int)(kBlockSize - header_size + 4), WrittenBytes());
+  ASSERT_EQ((unsigned int)(kBlockSize - kHeaderSize + 4), WrittenBytes());
  ASSERT_EQ(BigString("foo", n), Read());
  ASSERT_EQ("EOF", Read());
 }

-TEST_P(LogTest, RandomRead) {
+TEST_F(LogTest, RandomRead) {
  const int N = 500;
  Random write_rnd(301);
  for (int i = 0; i < N; i++) {
@@ -402,7 +394,7 @@ TEST_P(LogTest, RandomRead) {

 // Tests of all the error paths in log_reader.cc follow:

-TEST_P(LogTest, ReadError) {
+TEST_F(LogTest, ReadError) {
  Write("foo");
  ForceError();
  ASSERT_EQ("EOF", Read());
@@ -410,17 +402,17 @@ TEST_P(LogTest, ReadError) {
  ASSERT_EQ("OK", MatchError("read error"));
 }

-TEST_P(LogTest, BadRecordType) {
+TEST_F(LogTest, BadRecordType) {
  Write("foo");
  // Type is stored in header[6]
  IncrementByte(6, 100);
-  FixChecksum(0, 3, false);
+  FixChecksum(0, 3);
  ASSERT_EQ("EOF", Read());
  ASSERT_EQ(3U, DroppedBytes());
  ASSERT_EQ("OK", MatchError("unknown record type"));
 }

-TEST_P(LogTest, TruncatedTrailingRecordIsIgnored) {
+TEST_F(LogTest, TruncatedTrailingRecordIsIgnored) {
  Write("foo");
  ShrinkSize(4);   // Drop all payload as well as a header byte
  ASSERT_EQ("EOF", Read());
@@ -429,18 +421,17 @@ TEST_P(LogTest, TruncatedTrailingRecordIsIgnored) {
  ASSERT_EQ("", ReportMessage());
 }

-TEST_P(LogTest, TruncatedTrailingRecordIsNotIgnored) {
+TEST_F(LogTest, TruncatedTrailingRecordIsNotIgnored) {
  Write("foo");
  ShrinkSize(4);  // Drop all payload as well as a header byte
-  ASSERT_EQ("EOF", Read(WALRecoveryMode::kAbsoluteConsistency));
+  ASSERT_EQ("EOF", Read(/*report_eof_inconsistency*/ true));
  // Truncated last record is ignored, not treated as an error
  ASSERT_GT(DroppedBytes(), 0U);
  ASSERT_EQ("OK", MatchError("Corruption: truncated header"));
 }

-TEST_P(LogTest, BadLength) {
-  int header_size = GetParam() ? kRecyclableHeaderSize : kHeaderSize;
-  const int kPayloadSize = kBlockSize - header_size;
+TEST_F(LogTest, BadLength) {
+  const int kPayloadSize = kBlockSize - kHeaderSize;
  Write(BigString("bar", kPayloadSize));
  Write("foo");
  // Least significant size byte is stored in header[4].
@@ -450,7 +441,7 @@ TEST_P(LogTest, BadLength) {
  ASSERT_EQ("OK", MatchError("bad record length"));
 }

-TEST_P(LogTest, BadLengthAtEndIsIgnored) {
+TEST_F(LogTest, BadLengthAtEndIsIgnored) {
  Write("foo");
  ShrinkSize(1);
  ASSERT_EQ("EOF", Read());
@@ -458,63 +449,63 @@ TEST_P(LogTest, BadLengthAtEndIsIgnored) {
  ASSERT_EQ("", ReportMessage());
 }

-TEST_P(LogTest, BadLengthAtEndIsNotIgnored) {
+TEST_F(LogTest, BadLengthAtEndIsNotIgnored) {
  Write("foo");
  ShrinkSize(1);
-  ASSERT_EQ("EOF", Read(WALRecoveryMode::kAbsoluteConsistency));
+  ASSERT_EQ("EOF", Read(/*report_eof_inconsistency=*/true));
  ASSERT_GT(DroppedBytes(), 0U);
  ASSERT_EQ("OK", MatchError("Corruption: truncated header"));
 }

-TEST_P(LogTest, ChecksumMismatch) {
-  Write("foooooo");
-  IncrementByte(0, 14);
+TEST_F(LogTest, ChecksumMismatch) {
+  Write("foo");
+  IncrementByte(0, 10);
  ASSERT_EQ("EOF", Read());
-  ASSERT_EQ(14U + 4 * !!GetParam(), DroppedBytes());
+  ASSERT_EQ(10U, DroppedBytes());
  ASSERT_EQ("OK", MatchError("checksum mismatch"));
 }

-TEST_P(LogTest, UnexpectedMiddleType) {
+TEST_F(LogTest, UnexpectedMiddleType) {
  Write("foo");
-  SetByte(6, GetParam() ? kRecyclableMiddleType : kMiddleType);
-  FixChecksum(0, 3, !!GetParam());
+  SetByte(6, kMiddleType);
+  FixChecksum(0, 3);
  ASSERT_EQ("EOF", Read());
  ASSERT_EQ(3U, DroppedBytes());
  ASSERT_EQ("OK", MatchError("missing start"));
 }

-TEST_P(LogTest, UnexpectedLastType) {
+TEST_F(LogTest, UnexpectedLastType) {
  Write("foo");
-  SetByte(6, GetParam() ? kRecyclableLastType : kLastType);
-  FixChecksum(0, 3, !!GetParam());
+  SetByte(6, kLastType);
+  FixChecksum(0, 3);
  ASSERT_EQ("EOF", Read());
  ASSERT_EQ(3U, DroppedBytes());
  ASSERT_EQ("OK", MatchError("missing start"));
 }

-TEST_P(LogTest, UnexpectedFullType) {
+TEST_F(LogTest, UnexpectedFullType) {
  Write("foo");
  Write("bar");
-  SetByte(6, GetParam() ? kRecyclableFirstType : kFirstType);
-  FixChecksum(0, 3, !!GetParam());
+  SetByte(6, kFirstType);
+  FixChecksum(0, 3);
  ASSERT_EQ("bar", Read());
  ASSERT_EQ("EOF", Read());
  ASSERT_EQ(3U, DroppedBytes());
  ASSERT_EQ("OK", MatchError("partial record without end"));
 }

-TEST_P(LogTest, UnexpectedFirstType) {
+TEST_F(LogTest, UnexpectedFirstType) {
  Write("foo");
  Write(BigString("bar", 100000));
-  SetByte(6, GetParam() ? kRecyclableFirstType : kFirstType);
-  FixChecksum(0, 3, !!GetParam());
+  SetByte(6, kFirstType);
+  FixChecksum(0, 3);
  ASSERT_EQ(BigString("bar", 100000), Read());
  ASSERT_EQ("EOF", Read());
  ASSERT_EQ(3U, DroppedBytes());
  ASSERT_EQ("OK", MatchError("partial record without end"));
 }

-TEST_P(LogTest, MissingLastIsIgnored) {
+TEST_F(LogTest, MissingLastIsIgnored) {
  Write(BigString("bar", kBlockSize));
  // Remove the LAST block, including header.
  ShrinkSize(14);
@@ -523,16 +514,16 @@ TEST_P(LogTest, MissingLastIsIgnored) {
  ASSERT_EQ(0U, DroppedBytes());
 }

-TEST_P(LogTest, MissingLastIsNotIgnored) {
+TEST_F(LogTest, MissingLastIsNotIgnored) {
  Write(BigString("bar", kBlockSize));
  // Remove the LAST block, including header.
  ShrinkSize(14);
-  ASSERT_EQ("EOF", Read(WALRecoveryMode::kAbsoluteConsistency));
+  ASSERT_EQ("EOF", Read(/*report_eof_inconsistency=*/true));
  ASSERT_GT(DroppedBytes(), 0U);
  ASSERT_EQ("OK", MatchError("Corruption: error reading trailing data"));
 }

-TEST_P(LogTest, PartialLastIsIgnored) {
+TEST_F(LogTest, PartialLastIsIgnored) {
  Write(BigString("bar", kBlockSize));
  // Cause a bad record length in the LAST block.
  ShrinkSize(1);
@@ -541,18 +532,18 @@ TEST_P(LogTest, PartialLastIsIgnored) {
  ASSERT_EQ(0U, DroppedBytes());
 }

-TEST_P(LogTest, PartialLastIsNotIgnored) {
+TEST_F(LogTest, PartialLastIsNotIgnored) {
  Write(BigString("bar", kBlockSize));
  // Cause a bad record length in the LAST block.
  ShrinkSize(1);
-  ASSERT_EQ("EOF", Read(WALRecoveryMode::kAbsoluteConsistency));
+  ASSERT_EQ("EOF", Read(/*report_eof_inconsistency=*/true));
  ASSERT_GT(DroppedBytes(), 0U);
  ASSERT_EQ("OK", MatchError(
                      "Corruption: truncated headerCorruption: "
                      "error reading trailing data"));
 }

-TEST_P(LogTest, ErrorJoinsRecords) {
+TEST_F(LogTest, ErrorJoinsRecords) {
  // Consider two fragmented records:
  //    first(R1) last(R1) first(R2) last(R2)
  // where the middle two fragments disappear.  We do not want
@@ -575,60 +566,46 @@ TEST_P(LogTest, ErrorJoinsRecords) {
  ASSERT_GE(dropped, 2 * kBlockSize);
 }

-TEST_P(LogTest, ReadStart) { CheckInitialOffsetRecord(0, 0); }
+TEST_F(LogTest, ReadStart) { CheckInitialOffsetRecord(0, 0); }

-TEST_P(LogTest, ReadSecondOneOff) { CheckInitialOffsetRecord(1, 1); }
+TEST_F(LogTest, ReadSecondOneOff) { CheckInitialOffsetRecord(1, 1); }

-TEST_P(LogTest, ReadSecondTenThousand) { CheckInitialOffsetRecord(10000, 1); }
+TEST_F(LogTest, ReadSecondTenThousand) { CheckInitialOffsetRecord(10000, 1); }

-TEST_P(LogTest, ReadSecondStart) {
-  int header_size = GetParam() ? kRecyclableHeaderSize : kHeaderSize;
-  CheckInitialOffsetRecord(10000 + header_size, 1);
-}
+TEST_F(LogTest, ReadSecondStart) { CheckInitialOffsetRecord(10007, 1); }

-TEST_P(LogTest, ReadThirdOneOff) {
-  int header_size = GetParam() ? kRecyclableHeaderSize : kHeaderSize;
-  CheckInitialOffsetRecord(10000 + header_size + 1, 2);
-}
+TEST_F(LogTest, ReadThirdOneOff) { CheckInitialOffsetRecord(10008, 2); }

-TEST_P(LogTest, ReadThirdStart) {
-  int header_size = GetParam() ? kRecyclableHeaderSize : kHeaderSize;
-  CheckInitialOffsetRecord(20000 + 2 * header_size, 2);
-}
+TEST_F(LogTest, ReadThirdStart) { CheckInitialOffsetRecord(20014, 2); }

-TEST_P(LogTest, ReadFourthOneOff) {
-  int header_size = GetParam() ? kRecyclableHeaderSize : kHeaderSize;
-  CheckInitialOffsetRecord(20000 + 2 * header_size + 1, 3);
-}
+TEST_F(LogTest, ReadFourthOneOff) { CheckInitialOffsetRecord(20015, 3); }

-TEST_P(LogTest, ReadFourthFirstBlockTrailer) {
+TEST_F(LogTest, ReadFourthFirstBlockTrailer) {
  CheckInitialOffsetRecord(log::kBlockSize - 4, 3);
 }

-TEST_P(LogTest, ReadFourthMiddleBlock) {
+TEST_F(LogTest, ReadFourthMiddleBlock) {
  CheckInitialOffsetRecord(log::kBlockSize + 1, 3);
 }

-TEST_P(LogTest, ReadFourthLastBlock) {
+TEST_F(LogTest, ReadFourthLastBlock) {
  CheckInitialOffsetRecord(2 * log::kBlockSize + 1, 3);
 }

-TEST_P(LogTest, ReadFourthStart) {
-  int header_size = GetParam() ? kRecyclableHeaderSize : kHeaderSize;
+TEST_F(LogTest, ReadFourthStart) {
  CheckInitialOffsetRecord(
-      2 * (header_size + 1000) + (2 * log::kBlockSize - 1000) + 3 * header_size,
+      2 * (kHeaderSize + 1000) + (2 * log::kBlockSize - 1000) + 3 * kHeaderSize,
      3);
 }

-TEST_P(LogTest, ReadEnd) { CheckOffsetPastEndReturnsNoRecords(0); }
+TEST_F(LogTest, ReadEnd) { CheckOffsetPastEndReturnsNoRecords(0); }

-TEST_P(LogTest, ReadPastEnd) { CheckOffsetPastEndReturnsNoRecords(5); }
+TEST_F(LogTest, ReadPastEnd) { CheckOffsetPastEndReturnsNoRecords(5); }

-TEST_P(LogTest, ClearEofSingleBlock) {
+TEST_F(LogTest, ClearEofSingleBlock) {
  Write("foo");
  Write("bar");
-  int header_size = GetParam() ? kRecyclableHeaderSize : kHeaderSize;
-  ForceEOF(3 + header_size + 2);
+  ForceEOF(3 + kHeaderSize + 2);
  ASSERT_EQ("foo", Read());
  UnmarkEOF();
  ASSERT_EQ("bar", Read());
@@ -640,13 +617,12 @@ TEST_P(LogTest, ClearEofSingleBlock) {
  ASSERT_TRUE(IsEOF());
 }

-TEST_P(LogTest, ClearEofMultiBlock) {
+TEST_F(LogTest, ClearEofMultiBlock) {
  size_t num_full_blocks = 5;
-  int header_size = GetParam() ? kRecyclableHeaderSize : kHeaderSize;
-  size_t n = (kBlockSize - header_size) * num_full_blocks + 25;
+  size_t n = (kBlockSize - kHeaderSize) * num_full_blocks + 25;
  Write(BigString("foo", n));
  Write(BigString("bar", n));
-  ForceEOF(n + num_full_blocks * header_size + header_size + 3);
+  ForceEOF(n + num_full_blocks * kHeaderSize + 10);
  ASSERT_EQ(BigString("foo", n), Read());
  ASSERT_TRUE(IsEOF());
  UnmarkEOF();
@@ -658,7 +634,7 @@ TEST_P(LogTest, ClearEofMultiBlock) {
  ASSERT_TRUE(IsEOF());
 }

-TEST_P(LogTest, ClearEofError) {
+TEST_F(LogTest, ClearEofError) {
  // If an error occurs during Read() in UnmarkEOF(), the records contained
  // in the buffer should be returned on subsequent calls of ReadRecord()
  // until no more full records are left, whereafter ReadRecord() should return
@@ -676,7 +652,7 @@ TEST_P(LogTest, ClearEofError) {
  ASSERT_EQ("EOF", Read());
 }

-TEST_P(LogTest, ClearEofError2) {
+TEST_F(LogTest, ClearEofError2) {
  Write("foo");
  Write("bar");
  UnmarkEOF();
@@ -690,8 +666,6 @@ TEST_P(LogTest, ClearEofError2) {
  ASSERT_EQ("OK", MatchError("read error"));
 }

-INSTANTIATE_TEST_CASE_P(bool, LogTest, ::testing::Values(0, 2));
-
 }  // namespace log
 }  // namespace rocksdb

@@ -18,12 +18,8 @@
 namespace rocksdb {
 namespace log {

-Writer::Writer(unique_ptr<WritableFileWriter>&& dest,
-               uint64_t log_number, bool recycle_log_files)
-    : dest_(std::move(dest)),
-      block_offset_(0),
-      log_number_(log_number),
-      recycle_log_files_(recycle_log_files) {
+Writer::Writer(unique_ptr<WritableFileWriter>&& dest)
+    : dest_(std::move(dest)), block_offset_(0) {
  for (int i = 0; i <= kMaxRecordType; i++) {
    char t = static_cast<char>(i);
    type_crc_[i] = crc32c::Value(&t, 1);
@@ -37,46 +33,40 @@ Status Writer::AddRecord(const Slice& slice) {
  const char* ptr = slice.data();
  size_t left = slice.size();

-  // Header size varies depending on whether we are recycling or not.
-  const int header_size =
-      recycle_log_files_ ? kRecyclableHeaderSize : kHeaderSize;
-
  // Fragment the record if necessary and emit it.  Note that if slice
  // is empty, we still want to iterate once to emit a single
  // zero-length record
  Status s;
  bool begin = true;
  do {
-    const int64_t leftover = kBlockSize - block_offset_;
+    const int leftover = kBlockSize - block_offset_;
    assert(leftover >= 0);
-    if (leftover < header_size) {
+    if (leftover < kHeaderSize) {
      // Switch to a new block
      if (leftover > 0) {
-        // Fill the trailer (literal below relies on kHeaderSize and
-        // kRecyclableHeaderSize being <= 11)
-        assert(header_size <= 11);
-        dest_->Append(
-            Slice("\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00", leftover));
+        // Fill the trailer (literal below relies on kHeaderSize being 7)
+        assert(kHeaderSize == 7);
+        dest_->Append(Slice("\x00\x00\x00\x00\x00\x00", leftover));
      }
      block_offset_ = 0;
    }

-    // Invariant: we never leave < header_size bytes in a block.
-    assert(static_cast<int64_t>(kBlockSize - block_offset_) >= header_size);
+    // Invariant: we never leave < kHeaderSize bytes in a block.
+    assert(static_cast<int>(kBlockSize) - block_offset_ >= kHeaderSize);

-    const size_t avail = kBlockSize - block_offset_ - header_size;
+    const size_t avail = kBlockSize - block_offset_ - kHeaderSize;
    const size_t fragment_length = (left < avail) ? left : avail;

    RecordType type;
    const bool end = (left == fragment_length);
    if (begin && end) {
-      type = recycle_log_files_ ? kRecyclableFullType : kFullType;
+      type = kFullType;
    } else if (begin) {
-      type = recycle_log_files_ ? kRecyclableFirstType : kFirstType;
+      type = kFirstType;
    } else if (end) {
-      type = recycle_log_files_ ? kRecyclableLastType : kLastType;
+      type = kLastType;
    } else {
-      type = recycle_log_files_ ? kRecyclableMiddleType : kMiddleType;
+      type = kMiddleType;
    }

    s = EmitPhysicalRecord(type, ptr, fragment_length);
@@ -89,48 +79,28 @@ Status Writer::AddRecord(const Slice& slice) {

 Status Writer::EmitPhysicalRecord(RecordType t, const char* ptr, size_t n) {
  assert(n <= 0xffff);  // Must fit in two bytes
-
-  size_t header_size;
-  char buf[kRecyclableHeaderSize];
+  assert(block_offset_ + kHeaderSize + n <= kBlockSize);

  // Format the header
+  char buf[kHeaderSize];
  buf[4] = static_cast<char>(n & 0xff);
  buf[5] = static_cast<char>(n >> 8);
  buf[6] = static_cast<char>(t);

-  uint32_t crc = type_crc_[t];
-  if (t < kRecyclableFullType) {
-    // Legacy record format
-    assert(block_offset_ + kHeaderSize + n <= kBlockSize);
-    header_size = kHeaderSize;
-  } else {
-    // Recyclable record format
-    assert(block_offset_ + kRecyclableHeaderSize + n <= kBlockSize);
-    header_size = kRecyclableHeaderSize;
-
-    // Only encode low 32-bits of the 64-bit log number.  This means
-    // we will fail to detect an old record if we recycled a log from
-    // ~4 billion logs ago, but that is effectively impossible, and
-    // even if it were we'dbe far more likely to see a false positive
-    // on the 32-bit CRC.
-    EncodeFixed32(buf + 7, static_cast<uint32_t>(log_number_));
-    crc = crc32c::Extend(crc, buf + 7, 4);
-  }
-
  // Compute the crc of the record type and the payload.
-  crc = crc32c::Extend(crc, ptr, n);
-  crc = crc32c::Mask(crc);  // Adjust for storage
+  uint32_t crc = crc32c::Extend(type_crc_[t], ptr, n);
+  crc = crc32c::Mask(crc);                 // Adjust for storage
  EncodeFixed32(buf, crc);

  // Write the header and the payload
-  Status s = dest_->Append(Slice(buf, header_size));
+  Status s = dest_->Append(Slice(buf, kHeaderSize));
  if (s.ok()) {
    s = dest_->Append(Slice(ptr, n));
    if (s.ok()) {
      s = dest_->Flush();
    }
  }
-  block_offset_ += header_size + n;
+  block_offset_ += kHeaderSize + n;
  return s;
 }

@@ -43,7 +43,7 @@ namespace log {
 * Data is written out in kBlockSize chunks. If next record does not fit
 * into the space left, the leftover space will be padded with \0.
 *
- * Legacy record format:
+ * Record format:
 *
 * +---------+-----------+-----------+--- ... ---+
 * |CRC (4B) | Size (2B) | Type (1B) | Payload   |
@@ -57,23 +57,13 @@ namespace log {
 *        blocks that are larger than kBlockSize
 * Payload = Byte stream as long as specified by the payload size
 *
- * Recyclable record format:
- *
- * +---------+-----------+-----------+----------------+--- ... ---+
- * |CRC (4B) | Size (2B) | Type (1B) | Log number (4B)| Payload   |
- * +---------+-----------+-----------+----------------+--- ... ---+
- *
- * Same as above, with the addition of
- * Log number = 32bit log file number, so that we can distinguish between
- * records written by the most recent log writer vs a previous one.
 */
 class Writer {
 public:
  // Create a writer that will append data to "*dest".
  // "*dest" must be initially empty.
  // "*dest" must remain live while this Writer is in use.
-  explicit Writer(unique_ptr<WritableFileWriter>&& dest,
-                  uint64_t log_number, bool recycle_log_files);
+  explicit Writer(unique_ptr<WritableFileWriter>&& dest);
  ~Writer();

  Status AddRecord(const Slice& slice);
@@ -83,9 +73,7 @@ class Writer {

 private:
  unique_ptr<WritableFileWriter> dest_;
-  size_t block_offset_;       // Current offset in block
-  uint64_t log_number_;
-  bool recycle_log_files_;
+  int block_offset_;       // Current offset in block

  // crc32c values for all supported record types.  These are
  // pre-computed to reduce the overhead of computing the crc of the
@@ -21,7 +21,6 @@
 #include "rocksdb/iterator.h"
 #include "rocksdb/merge_operator.h"
 #include "rocksdb/slice_transform.h"
-#include "table/internal_iterator.h"
 #include "table/merger.h"
 #include "util/arena.h"
 #include "util/coding.h"
@@ -60,7 +59,7 @@ MemTable::MemTable(const InternalKeyComparator& cmp,
      moptions_(ioptions, mutable_cf_options),
      refs_(0),
      kArenaBlockSize(OptimizeBlockSize(moptions_.arena_block_size)),
-      arena_(moptions_.arena_block_size, 0),
+      arena_(moptions_.arena_block_size),
      allocator_(&arena_, write_buffer),
      table_(ioptions.memtable_factory->CreateMemTableRep(
          comparator_, &allocator_, ioptions.prefix_extractor,
@@ -78,12 +77,12 @@ MemTable::MemTable(const InternalKeyComparator& cmp,
                 ? moptions_.inplace_update_num_locks
                 : 0),
      prefix_extractor_(ioptions.prefix_extractor),
-      flush_state_(FLUSH_NOT_REQUESTED),
+      should_flush_(ShouldFlushNow()),
+      flush_scheduled_(false),
      env_(ioptions.env) {
-  UpdateFlushState();
-  // something went wrong if we need to flush before inserting anything
-  assert(!ShouldScheduleFlush());
-
+  // if should_flush_ == true without an entry inserted, something must have
+  // gone wrong already.
+  assert(!should_flush_);
  if (prefix_extractor_ && moptions_.memtable_prefix_bloom_bits > 0) {
    prefix_bloom_.reset(new DynamicBloom(
        &allocator_,
@@ -167,17 +166,6 @@ bool MemTable::ShouldFlushNow() const {
  return arena_.AllocatedAndUnused() < kArenaBlockSize / 4;
 }

-void MemTable::UpdateFlushState() {
-  auto state = flush_state_.load(std::memory_order_relaxed);
-  if (state == FLUSH_NOT_REQUESTED && ShouldFlushNow()) {
-    // ignore CAS failure, because that means somebody else requested
-    // a flush
-    flush_state_.compare_exchange_strong(state, FLUSH_REQUESTED,
-                                         std::memory_order_relaxed,
-                                         std::memory_order_relaxed);
-  }
-}
-
 int MemTable::KeyComparator::operator()(const char* prefix_len_key1,
                                        const char* prefix_len_key2) const {
  // Internal keys are encoded as length-prefixed strings.
@@ -214,7 +202,7 @@ const char* EncodeKey(std::string* scratch, const Slice& target) {
  return scratch->data();
 }

-class MemTableIterator : public InternalIterator {
+class MemTableIterator: public Iterator {
 public:
  MemTableIterator(
      const MemTable& mem, const ReadOptions& read_options, Arena* arena)
@@ -242,15 +230,10 @@ class MemTableIterator : public InternalIterator {
  virtual void Seek(const Slice& k) override {
    PERF_TIMER_GUARD(seek_on_memtable_time);
    PERF_COUNTER_ADD(seek_on_memtable_count, 1);
-    if (bloom_ != nullptr) {
-      if (!bloom_->MayContain(
-              prefix_extractor_->Transform(ExtractUserKey(k)))) {
-        PERF_COUNTER_ADD(bloom_memtable_miss_count, 1);
-        valid_ = false;
-        return;
-      } else {
-        PERF_COUNTER_ADD(bloom_memtable_hit_count, 1);
-      }
+    if (bloom_ != nullptr &&
+        !bloom_->MayContain(prefix_extractor_->Transform(ExtractUserKey(k)))) {
+      valid_ = false;
+      return;
    }
    iter_->Seek(k, nullptr);
    valid_ = iter_->Valid();
@@ -285,21 +268,6 @@ class MemTableIterator : public InternalIterator {

  virtual Status status() const override { return Status::OK(); }

-  virtual Status PinData() override {
-    // memtable data is always pinned
-    return Status::OK();
-  }
-
-  virtual Status ReleasePinnedData() override {
-    // memtable data is always pinned
-    return Status::OK();
-  }
-
-  virtual bool IsKeyPinned() const override {
-    // memtable data is always pinned
-    return true;
-  }
-
 private:
  DynamicBloom* bloom_;
  const SliceTransform* const prefix_extractor_;
@@ -312,8 +280,7 @@ class MemTableIterator : public InternalIterator {
  void operator=(const MemTableIterator&);
 };

-InternalIterator* MemTable::NewIterator(const ReadOptions& read_options,
-                                        Arena* arena) {
+Iterator* MemTable::NewIterator(const ReadOptions& read_options, Arena* arena) {
  assert(arena != nullptr);
  auto mem = arena->AllocateAligned(sizeof(MemTableIterator));
  return new (mem) MemTableIterator(*this, read_options, arena);
@@ -346,7 +313,7 @@ uint64_t MemTable::ApproximateSize(const Slice& start_ikey,

 void MemTable::Add(SequenceNumber s, ValueType type,
                   const Slice& key, /* user key */
-                   const Slice& value, bool allow_concurrent) {
+                   const Slice& value) {
  // Format of an entry is concatenation of:
  //  key_size     : varint32 of internal_key.size()
  //  key bytes    : char[internal_key.size()]
@@ -360,7 +327,7 @@ void MemTable::Add(SequenceNumber s, ValueType type,
                               val_size;
  char* buf = nullptr;
  KeyHandle handle = table_->Allocate(encoded_len, &buf);
-
+  assert(buf != nullptr);
  char* p = EncodeVarint32(buf, internal_key_size);
  memcpy(p, key.data(), key_size);
  p += key_size;
@@ -370,64 +337,32 @@ void MemTable::Add(SequenceNumber s, ValueType type,
  p = EncodeVarint32(p, val_size);
  memcpy(p, value.data(), val_size);
  assert((unsigned)(p + val_size - buf) == (unsigned)encoded_len);
-  if (!allow_concurrent) {
-    table_->Insert(handle);
-
-    // this is a bit ugly, but is the way to avoid locked instructions
-    // when incrementing an atomic
-    num_entries_.store(num_entries_.load(std::memory_order_relaxed) + 1,
-                       std::memory_order_relaxed);
-    data_size_.store(data_size_.load(std::memory_order_relaxed) + encoded_len,
+  table_->Insert(handle);
+  num_entries_.store(num_entries_.load(std::memory_order_relaxed) + 1,
                     std::memory_order_relaxed);
-    if (type == kTypeDeletion) {
-      num_deletes_.store(num_deletes_.load(std::memory_order_relaxed) + 1,
-                         std::memory_order_relaxed);
-    }
-
-    if (prefix_bloom_) {
-      assert(prefix_extractor_);
-      prefix_bloom_->Add(prefix_extractor_->Transform(key));
-    }
-
-    // The first sequence number inserted into the memtable
-    assert(first_seqno_ == 0 || s > first_seqno_);
-    if (first_seqno_ == 0) {
-      first_seqno_.store(s, std::memory_order_relaxed);
-
-      if (earliest_seqno_ == kMaxSequenceNumber) {
-        earliest_seqno_.store(GetFirstSequenceNumber(),
-                              std::memory_order_relaxed);
-      }
-      assert(first_seqno_.load() >= earliest_seqno_.load());
-    }
-  } else {
-    table_->InsertConcurrently(handle);
-
-    num_entries_.fetch_add(1, std::memory_order_relaxed);
-    data_size_.fetch_add(encoded_len, std::memory_order_relaxed);
-    if (type == kTypeDeletion) {
-      num_deletes_.fetch_add(1, std::memory_order_relaxed);
-    }
-
-    if (prefix_bloom_) {
-      assert(prefix_extractor_);
-      prefix_bloom_->AddConcurrently(prefix_extractor_->Transform(key));
-    }
-
-    // atomically update first_seqno_ and earliest_seqno_.
-    uint64_t cur_seq_num = first_seqno_.load(std::memory_order_relaxed);
-    while ((cur_seq_num == 0 || s < cur_seq_num) &&
-           !first_seqno_.compare_exchange_weak(cur_seq_num, s)) {
-    }
-    uint64_t cur_earliest_seqno =
-        earliest_seqno_.load(std::memory_order_relaxed);
-    while (
-        (cur_earliest_seqno == kMaxSequenceNumber || s < cur_earliest_seqno) &&
-        !first_seqno_.compare_exchange_weak(cur_earliest_seqno, s)) {
-    }
+  data_size_.store(data_size_.load(std::memory_order_relaxed) + encoded_len,
+                   std::memory_order_relaxed);
+  if (type == kTypeDeletion) {
+    num_deletes_++;
  }

-  UpdateFlushState();
+  if (prefix_bloom_) {
+    assert(prefix_extractor_);
+    prefix_bloom_->Add(prefix_extractor_->Transform(key));
+  }
+
+  // The first sequence number inserted into the memtable
+  assert(first_seqno_ == 0 || s > first_seqno_);
+  if (first_seqno_ == 0) {
+    first_seqno_ = s;
+
+    if (earliest_seqno_ == kMaxSequenceNumber) {
+      earliest_seqno_ = first_seqno_;
+    }
+    assert(first_seqno_ >= earliest_seqno_);
+  }
+
+  should_flush_ = ShouldFlushNow();
 }

 // Callback from MemTable::Get()
@@ -500,7 +435,7 @@ static bool SaveValue(void* arg, const char* entry) {
            *(s->status) =
                Status::Corruption("Error: Could not perform merge.");
          }
-        } else if (s->value != nullptr) {
+        } else {
          s->value->assign(v.data(), v.size());
        }
        if (s->inplace_update_support) {
@@ -509,10 +444,9 @@ static bool SaveValue(void* arg, const char* entry) {
        *(s->found_final_value) = true;
        return false;
      }
-      case kTypeDeletion:
-      case kTypeSingleDeletion: {
+      case kTypeDeletion: {
        if (*(s->merge_in_progress)) {
-          assert(merge_operator != nullptr);
+          assert(merge_operator);
          *(s->status) = Status::OK();
          bool merge_success = false;
          {
@@ -573,18 +507,12 @@ bool MemTable::Get(const LookupKey& key, std::string* value, Status* s,
  Slice user_key = key.user_key();
  bool found_final_value = false;
  bool merge_in_progress = s->IsMergeInProgress();
-  bool const may_contain =
-      nullptr == prefix_bloom_
-          ? false
-          : prefix_bloom_->MayContain(prefix_extractor_->Transform(user_key));
-  if (prefix_bloom_ && !may_contain) {
+
+  if (prefix_bloom_ &&
+      !prefix_bloom_->MayContain(prefix_extractor_->Transform(user_key))) {
    // iter is null if prefix bloom says the key does not exist
-    PERF_COUNTER_ADD(bloom_memtable_miss_count, 1);
    *seq = kMaxSequenceNumber;
  } else {
-    if (prefix_bloom_) {
-      PERF_COUNTER_ADD(bloom_memtable_hit_count, 1);
-    }
    Saver saver;
    saver.status = s;
    saver.found_final_value = &found_final_value;
@@ -728,16 +656,16 @@ bool MemTable::UpdateCallback(SequenceNumber seq,
              }
            }
            RecordTick(moptions_.statistics, NUMBER_KEYS_UPDATED);
-            UpdateFlushState();
+            should_flush_ = ShouldFlushNow();
            return true;
          } else if (status == UpdateStatus::UPDATED) {
            Add(seq, kTypeValue, key, Slice(str_value));
            RecordTick(moptions_.statistics, NUMBER_KEYS_WRITTEN);
-            UpdateFlushState();
+            should_flush_ = ShouldFlushNow();
            return true;
          } else if (status == UpdateStatus::UPDATE_FAILED) {
            // No action required. Return.
-            UpdateFlushState();
+            should_flush_ = ShouldFlushNow();
            return true;
          }
        }
@@ -8,11 +8,10 @@
 // found in the LICENSE file. See the AUTHORS file for names of contributors.

 #pragma once
-#include <atomic>
-#include <deque>
-#include <functional>
-#include <memory>
 #include <string>
+#include <memory>
+#include <functional>
+#include <deque>
 #include <vector>
 #include "db/dbformat.h"
 #include "db/skiplist.h"
@@ -22,9 +21,8 @@
 #include "rocksdb/memtablerep.h"
 #include "rocksdb/immutable_options.h"
 #include "db/memtable_allocator.h"
-#include "util/concurrent_arena.h"
+#include "util/arena.h"
 #include "util/dynamic_bloom.h"
-#include "util/instrumented_mutex.h"
 #include "util/mutable_cf_options.h"

 namespace rocksdb {
@@ -33,7 +31,6 @@ class Mutex;
 class MemTableIterator;
 class MergeContext;
 class WriteBuffer;
-class InternalIterator;

 struct MemTableOptions {
  explicit MemTableOptions(
@@ -126,17 +123,10 @@ class MemTable {
  // This method heuristically determines if the memtable should continue to
  // host more data.
  bool ShouldScheduleFlush() const {
-    return flush_state_.load(std::memory_order_relaxed) == FLUSH_REQUESTED;
+    return flush_scheduled_ == false && should_flush_;
  }

-  // Returns true if a flush should be scheduled and the caller should
-  // be the one to schedule it
-  bool MarkFlushScheduled() {
-    auto before = FLUSH_REQUESTED;
-    return flush_state_.compare_exchange_strong(before, FLUSH_SCHEDULED,
-                                                std::memory_order_relaxed,
-                                                std::memory_order_relaxed);
-  }
+  void MarkFlushScheduled() { flush_scheduled_ = true; }

  // Return an iterator that yields the contents of the memtable.
  //
@@ -150,16 +140,17 @@ class MemTable {
  // arena: If not null, the arena needs to be used to allocate the Iterator.
  //        Calling ~Iterator of the iterator will destroy all the states but
  //        those allocated in arena.
-  InternalIterator* NewIterator(const ReadOptions& read_options, Arena* arena);
+  Iterator* NewIterator(const ReadOptions& read_options, Arena* arena);

  // Add an entry into memtable that maps key to value at the
  // specified sequence number and with the specified type.
  // Typically value will be empty if type==kTypeDeletion.
  //
-  // REQUIRES: if allow_concurrent = false, external synchronization to prevent
-  // simultaneous operations on the same MemTable.
-  void Add(SequenceNumber seq, ValueType type, const Slice& key,
-           const Slice& value, bool allow_concurrent = false);
+  // REQUIRES: external synchronization to prevent simultaneous
+  // operations on the same MemTable.
+  void Add(SequenceNumber seq, ValueType type,
+           const Slice& key,
+           const Slice& value);

  // If memtable contains a value for key, store it in *value and return true.
  // If memtable contains a deletion for key, store a NotFound() error
@@ -197,7 +188,7 @@ class MemTable {
              const Slice& key,
              const Slice& value);

-  // If prev_value for key exists, attempts to update it inplace.
+  // If prev_value for key exits, attempts to update it inplace.
  // else returns false
  // Pseudocode
  //   if key exists in current memtable && prev_value is of type kTypeValue
@@ -228,9 +219,7 @@ class MemTable {
  // Get total number of deletes in the mem table.
  // REQUIRES: external synchronization to prevent simultaneous
  // operations on the same MemTable (unless this Memtable is immutable).
-  uint64_t num_deletes() const {
-    return num_deletes_.load(std::memory_order_relaxed);
-  }
+  uint64_t num_deletes() const { return num_deletes_; }

  // Returns the edits area that is needed for flushing the memtable
  VersionEdit* GetEdits() { return &edit_; }
@@ -244,9 +233,7 @@ class MemTable {
  // into the memtable.
  // REQUIRES: external synchronization to prevent simultaneous
  // operations on the same MemTable (unless this Memtable is immutable).
-  SequenceNumber GetFirstSequenceNumber() {
-    return first_seqno_.load(std::memory_order_relaxed);
-  }
+  SequenceNumber GetFirstSequenceNumber() { return first_seqno_; }

  // Returns the sequence number that is guaranteed to be smaller than or equal
  // to the sequence number of any key that could be inserted into this
@@ -255,9 +242,7 @@ class MemTable {
  //
  // If the earliest sequence number could not be determined,
  // kMaxSequenceNumber will be returned.
-  SequenceNumber GetEarliestSequenceNumber() {
-    return earliest_seqno_.load(std::memory_order_relaxed);
-  }
+  SequenceNumber GetEarliestSequenceNumber() { return earliest_seqno_; }

  // Returns the next active logfile number when this memtable is about to
  // be flushed to storage
@@ -304,7 +289,8 @@ class MemTable {
  const MemTableOptions* GetMemTableOptions() const { return &moptions_; }

 private:
-  enum FlushStateEnum { FLUSH_NOT_REQUESTED, FLUSH_REQUESTED, FLUSH_SCHEDULED };
+  // Dynamically check if we can add more incoming entries
+  bool ShouldFlushNow() const;

  friend class MemTableIterator;
  friend class MemTableBackwardIterator;
@@ -314,14 +300,14 @@ class MemTable {
  const MemTableOptions moptions_;
  int refs_;
  const size_t kArenaBlockSize;
-  ConcurrentArena arena_;
+  Arena arena_;
  MemTableAllocator allocator_;
  unique_ptr<MemTableRep> table_;

  // Total data size of all data inserted
  std::atomic<uint64_t> data_size_;
  std::atomic<uint64_t> num_entries_;
-  std::atomic<uint64_t> num_deletes_;
+  uint64_t num_deletes_;

  // These are used to manage memtable flushes to storage
  bool flush_in_progress_; // started the flush
@@ -333,11 +319,11 @@ class MemTable {
  VersionEdit edit_;

  // The sequence number of the kv that was inserted first
-  std::atomic<SequenceNumber> first_seqno_;
+  SequenceNumber first_seqno_;

  // The db sequence number at the time of creation or kMaxSequenceNumber
  // if not set.
-  std::atomic<SequenceNumber> earliest_seqno_;
+  SequenceNumber earliest_seqno_;

  // The log files earlier than this number can be deleted.
  uint64_t mem_next_logfile_number_;
@@ -345,22 +331,19 @@ class MemTable {
  // rw locks for inplace updates
  std::vector<port::RWMutex> locks_;

+  // No copying allowed
+  MemTable(const MemTable&);
+  void operator=(const MemTable&);
+
  const SliceTransform* const prefix_extractor_;
  std::unique_ptr<DynamicBloom> prefix_bloom_;

-  std::atomic<FlushStateEnum> flush_state_;
+  // a flag indicating if a memtable has met the criteria to flush
+  bool should_flush_;

+  // a flag indicating if flush has been scheduled
+  bool flush_scheduled_;
  Env* env_;
-
-  // Returns a heuristic flush decision
-  bool ShouldFlushNow() const;
-
-  // Updates flush_state_ using ShouldFlushNow()
-  void UpdateFlushState();
-
-  // No copying allowed
-  MemTable(const MemTable&);
-  MemTable& operator=(const MemTable&);
 };

 extern const char* EncodeKey(std::string* scratch, const Slice& target);
@@ -7,42 +7,46 @@
 // 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 "db/memtable_allocator.h"
-
 #include <assert.h>
+
+#include "db/memtable_allocator.h"
 #include "db/writebuffer.h"
 #include "util/arena.h"

 namespace rocksdb {

-MemTableAllocator::MemTableAllocator(Allocator* allocator,
-                                     WriteBuffer* write_buffer)
-    : allocator_(allocator), write_buffer_(write_buffer), bytes_allocated_(0) {}
+MemTableAllocator::MemTableAllocator(Arena* arena, WriteBuffer* write_buffer)
+    : arena_(arena), write_buffer_(write_buffer), bytes_allocated_(0) {
+}

-MemTableAllocator::~MemTableAllocator() { DoneAllocating(); }
+MemTableAllocator::~MemTableAllocator() {
+  DoneAllocating();
+}

 char* MemTableAllocator::Allocate(size_t bytes) {
  assert(write_buffer_ != nullptr);
-  bytes_allocated_.fetch_add(bytes, std::memory_order_relaxed);
+  bytes_allocated_ += bytes;
  write_buffer_->ReserveMem(bytes);
-  return allocator_->Allocate(bytes);
+  return arena_->Allocate(bytes);
 }

 char* MemTableAllocator::AllocateAligned(size_t bytes, size_t huge_page_size,
                                         Logger* logger) {
  assert(write_buffer_ != nullptr);
-  bytes_allocated_.fetch_add(bytes, std::memory_order_relaxed);
+  bytes_allocated_ += bytes;
  write_buffer_->ReserveMem(bytes);
-  return allocator_->AllocateAligned(bytes, huge_page_size, logger);
+  return arena_->AllocateAligned(bytes, huge_page_size, logger);
 }

 void MemTableAllocator::DoneAllocating() {
  if (write_buffer_ != nullptr) {
-    write_buffer_->FreeMem(bytes_allocated_.load(std::memory_order_relaxed));
+    write_buffer_->FreeMem(bytes_allocated_);
    write_buffer_ = nullptr;
  }
 }

-size_t MemTableAllocator::BlockSize() const { return allocator_->BlockSize(); }
+size_t MemTableAllocator::BlockSize() const {
+  return arena_->BlockSize();
+}

 }  // namespace rocksdb
@@ -11,18 +11,17 @@
 // to WriteBuffer so we can track and enforce overall write buffer limits.

 #pragma once
-
-#include <atomic>
 #include "util/allocator.h"

 namespace rocksdb {

+class Arena;
 class Logger;
 class WriteBuffer;

 class MemTableAllocator : public Allocator {
 public:
-  explicit MemTableAllocator(Allocator* allocator, WriteBuffer* write_buffer);
+  explicit MemTableAllocator(Arena* arena, WriteBuffer* write_buffer);
  ~MemTableAllocator();

  // Allocator interface
@@ -36,9 +35,9 @@ class MemTableAllocator : public Allocator {
  void DoneAllocating();

 private:
-  Allocator* allocator_;
+  Arena* arena_;
  WriteBuffer* write_buffer_;
-  std::atomic<size_t> bytes_allocated_;
+  size_t bytes_allocated_;

  // No copying allowed
  MemTableAllocator(const MemTableAllocator&);
@@ -138,9 +138,9 @@ bool MemTableListVersion::GetFromList(std::list<MemTable*>* list,
  return false;
 }

-void MemTableListVersion::AddIterators(
-    const ReadOptions& options, std::vector<InternalIterator*>* iterator_list,
-    Arena* arena) {
+void MemTableListVersion::AddIterators(const ReadOptions& options,
+                                       std::vector<Iterator*>* iterator_list,
+                                       Arena* arena) {
  for (auto& m : memlist_) {
    iterator_list->push_back(m->NewIterator(options, arena));
  }
@@ -74,8 +74,7 @@ class MemTableListVersion {
  }

  void AddIterators(const ReadOptions& options,
-                    std::vector<InternalIterator*>* iterator_list,
-                    Arena* arena);
+                    std::vector<Iterator*>* iterator_list, Arena* arena);

  void AddIterators(const ReadOptions& options,
                    MergeIteratorBuilder* merge_iter_builder);
@@ -132,6 +132,8 @@ DEFINE_int64(seed, 0,
             "Seed base for random number generators. "
             "When 0 it is deterministic.");

+static rocksdb::Env* FLAGS_env = rocksdb::Env::Default();
+
 namespace rocksdb {

 namespace {
@@ -590,7 +592,6 @@ int main(int argc, char** argv) {
  std::unique_ptr<rocksdb::MemTableRepFactory> factory;
  if (FLAGS_memtablerep == "skiplist") {
    factory.reset(new rocksdb::SkipListFactory);
-#ifndef ROCKSDB_LITE
  } else if (FLAGS_memtablerep == "vector") {
    factory.reset(new rocksdb::VectorRepFactory);
  } else if (FLAGS_memtablerep == "hashskiplist") {
@@ -612,7 +613,6 @@ int main(int argc, char** argv) {
        static_cast<uint32_t>(FLAGS_hash_function_count)));
    options.prefix_extractor.reset(
        rocksdb::NewFixedPrefixTransform(FLAGS_prefix_length));
-#endif  // ROCKSDB_LITE
  } else {
    fprintf(stdout, "Unknown memtablerep: %s\n", FLAGS_memtablerep.c_str());
    exit(1);
@@ -12,9 +12,9 @@
 #include "rocksdb/comparator.h"
 #include "rocksdb/db.h"
 #include "rocksdb/merge_operator.h"
-#include "table/internal_iterator.h"
 #include "util/perf_context_imp.h"
 #include "util/statistics.h"
+#include "util/stop_watch.h"

 namespace rocksdb {

@@ -41,7 +41,8 @@ Status MergeHelper::TimedFullMerge(const Slice& key, const Slice* value,
  bool success =
      merge_operator->FullMerge(key, value, operands, result, logger);

-  RecordTick(statistics, MERGE_OPERATION_TOTAL_TIME, timer.ElapsedNanosSafe());
+  RecordTick(statistics, MERGE_OPERATION_TOTAL_TIME,
+             env != nullptr ? timer.ElapsedNanos() : 0);

  if (!success) {
    RecordTick(statistics, NUMBER_MERGE_FAILURES);
@@ -57,47 +58,39 @@ Status MergeHelper::TimedFullMerge(const Slice& key, const Slice* value,
 //       keys_ stores the list of keys encountered while merging.
 //       operands_ stores the list of merge operands encountered while merging.
 //       keys_[i] corresponds to operands_[i] for each i.
-Status MergeHelper::MergeUntil(InternalIterator* iter,
-                               const SequenceNumber stop_before,
-                               const bool at_bottom) {
+Status MergeHelper::MergeUntil(Iterator* iter, const SequenceNumber stop_before,
+                               const bool at_bottom, Statistics* stats,
+                               Env* env_) {
  // Get a copy of the internal key, before it's invalidated by iter->Next()
  // Also maintain the list of merge operands seen.
  assert(HasOperator());
  keys_.clear();
  operands_.clear();
+  keys_.push_front(iter->key().ToString());
+  operands_.push_front(iter->value().ToString());
  assert(user_merge_operator_);
-  bool first_key = true;

  // We need to parse the internal key again as the parsed key is
  // backed by the internal key!
  // Assume no internal key corruption as it has been successfully parsed
  // by the caller.
-  // original_key_is_iter variable is just caching the information:
-  // original_key_is_iter == (iter->key().ToString() == original_key)
-  bool original_key_is_iter = true;
-  std::string original_key = iter->key().ToString();
-  // Important:
-  // orig_ikey is backed by original_key if keys_.empty()
-  // orig_ikey is backed by keys_.back() if !keys_.empty()
+  // Invariant: keys_.back() will not change. Hence, orig_ikey is always valid.
  ParsedInternalKey orig_ikey;
-  ParseInternalKey(original_key, &orig_ikey);
+  ParseInternalKey(keys_.back(), &orig_ikey);

  Status s;
  bool hit_the_next_user_key = false;
-  for (; iter->Valid(); iter->Next(), original_key_is_iter = false) {
+  for (iter->Next(); iter->Valid(); iter->Next()) {
    ParsedInternalKey ikey;
+    assert(operands_.size() >= 1);        // Should be invariants!
    assert(keys_.size() == operands_.size());

    if (!ParseInternalKey(iter->key(), &ikey)) {
      // stop at corrupted key
      if (assert_valid_internal_key_) {
-        assert(!"Corrupted internal key not expected.");
-        return Status::Corruption("Corrupted internal key not expected.");
+        assert(!"corrupted internal key is not expected");
      }
      break;
-    } else if (first_key) {
-      assert(user_comparator_->Equal(ikey.user_key, orig_ikey.user_key));
-      first_key = false;
    } else if (!user_comparator_->Equal(ikey.user_key, orig_ikey.user_key)) {
      // hit a different user key, stop right here
      hit_the_next_user_key = true;
@@ -109,31 +102,14 @@ Status MergeHelper::MergeUntil(InternalIterator* iter,

    // At this point we are guaranteed that we need to process this key.

-    assert(IsValueType(ikey.type));
+    assert(ikey.type <= kValueTypeForSeek);
    if (ikey.type != kTypeMerge) {
-      if (ikey.type != kTypeValue && ikey.type != kTypeDeletion) {
-        // Merges operands can only be used with puts and deletions, single
-        // deletions are not supported.
-        assert(false);
-        // release build doesn't have asserts, so we return error status
-        return Status::InvalidArgument(
-            " Merges operands can only be used with puts and deletions, single "
-            "deletions are not supported.");
-      }
-
      // hit a put/delete
      //   => merge the put value or a nullptr with operands_
      //   => store result in operands_.back() (and update keys_.back())
      //   => change the entry type to kTypeValue for keys_.back()
      // We are done! Success!
-
-      // If there are no operands, just return the Status::OK(). That will cause
-      // the compaction iterator to write out the key we're currently at, which
-      // is the put/delete we just encountered.
-      if (keys_.empty()) {
-        return Status::OK();
-      }
-
+      //
      // TODO(noetzli) If the merge operator returns false, we are currently
      // (almost) silently dropping the put/delete. That's probably not what we
      // want.
@@ -141,14 +117,14 @@ Status MergeHelper::MergeUntil(InternalIterator* iter,
      const Slice* val_ptr = (kTypeValue == ikey.type) ? &val : nullptr;
      std::string merge_result;
      s = TimedFullMerge(ikey.user_key, val_ptr, operands_,
-                         user_merge_operator_, stats_, env_, logger_,
+                         user_merge_operator_, stats, env_, logger_,
                         &merge_result);

      // We store the result in keys_.back() and operands_.back()
      // if nothing went wrong (i.e.: no operand corruption on disk)
      if (s.ok()) {
        // The original key encountered
-        original_key = std::move(keys_.back());
+        std::string original_key = std::move(keys_.back());
        orig_ikey.type = kTypeValue;
        UpdateInternalKey(&original_key, orig_ikey.sequence, orig_ikey.type);
        keys_.clear();
@@ -162,42 +138,18 @@ Status MergeHelper::MergeUntil(InternalIterator* iter,
      return s;
    } else {
      // hit a merge
-      //   => if there is a compaction filter, apply it.
      //   => merge the operand into the front of the operands_ list
-      //      if not filtered
+      //   => use the user's associative merge function to determine how.
      //   => then continue because we haven't yet seen a Put/Delete.
-      //
-      // Keep queuing keys and operands until we either meet a put / delete
+      assert(!operands_.empty()); // Should have at least one element in it
+
+      // keep queuing keys and operands until we either meet a put / delete
      // request or later did a partial merge.
-
-      Slice value_slice = iter->value();
-      // add an operand to the list if:
-      // 1) it's included in one of the snapshots. in that case we *must* write
-      // it out, no matter what compaction filter says
-      // 2) it's not filtered by a compaction filter
-      if (ikey.sequence <= latest_snapshot_ ||
-          !FilterMerge(orig_ikey.user_key, value_slice)) {
-        if (original_key_is_iter) {
-          // this is just an optimization that saves us one memcpy
-          keys_.push_front(std::move(original_key));
-        } else {
-          keys_.push_front(iter->key().ToString());
-        }
-        if (keys_.size() == 1) {
-          // we need to re-anchor the orig_ikey because it was anchored by
-          // original_key before
-          ParseInternalKey(keys_.back(), &orig_ikey);
-        }
-        operands_.push_front(value_slice.ToString());
-      }
+      keys_.push_front(iter->key().ToString());
+      operands_.push_front(iter->value().ToString());
    }
  }

-  if (operands_.size() == 0) {
-    // we filtered out all the merge operands
-    return Status::OK();
-  }
-
  // We are sure we have seen this key's entire history if we are at the
  // last level and exhausted all internal keys of this user key.
  // NOTE: !iter->Valid() does not necessarily mean we hit the
@@ -222,13 +174,11 @@ Status MergeHelper::MergeUntil(InternalIterator* iter,
    assert(operands_.size() == keys_.size());
    std::string merge_result;
    s = TimedFullMerge(orig_ikey.user_key, nullptr, operands_,
-                       user_merge_operator_, stats_, env_, logger_,
+                       user_merge_operator_, stats, env_, logger_,
                       &merge_result);
    if (s.ok()) {
      // The original key encountered
-      // We are certain that keys_ is not empty here (see assertions couple of
-      // lines before).
-      original_key = std::move(keys_.back());
+      std::string original_key = std::move(keys_.back());
      orig_ikey.type = kTypeValue;
      UpdateInternalKey(&original_key, orig_ikey.sequence, orig_ikey.type);
      keys_.clear();
@@ -250,14 +200,14 @@ Status MergeHelper::MergeUntil(InternalIterator* iter,
      bool merge_success = false;
      std::string merge_result;
      {
-        StopWatchNano timer(env_, stats_ != nullptr);
+        StopWatchNano timer(env_, stats != nullptr);
        PERF_TIMER_GUARD(merge_operator_time_nanos);
        merge_success = user_merge_operator_->PartialMergeMulti(
            orig_ikey.user_key,
            std::deque<Slice>(operands_.begin(), operands_.end()),
            &merge_result, logger_);
-        RecordTick(stats_, MERGE_OPERATION_TOTAL_TIME,
-                   timer.ElapsedNanosSafe());
+        RecordTick(stats, MERGE_OPERATION_TOTAL_TIME,
+                   env_ != nullptr ? timer.ElapsedNanos() : 0);
      }
      if (merge_success) {
        // Merging of operands (associative merge) was successful.
@@ -281,6 +231,7 @@ MergeOutputIterator::MergeOutputIterator(const MergeHelper* merge_helper)
 void MergeOutputIterator::SeekToFirst() {
  const auto& keys = merge_helper_->keys();
  const auto& values = merge_helper_->values();
+  assert(keys.size() > 0);
  assert(keys.size() == values.size());
  it_keys_ = keys.rbegin();
  it_values_ = values.rbegin();
@@ -291,17 +242,4 @@ void MergeOutputIterator::Next() {
  ++it_values_;
 }

-bool MergeHelper::FilterMerge(const Slice& user_key, const Slice& value_slice) {
-  if (compaction_filter_ == nullptr) {
-    return false;
-  }
-  if (stats_ != nullptr) {
-    filter_timer_.Start();
-  }
-  bool to_delete =
-      compaction_filter_->FilterMergeOperand(level_, user_key, value_slice);
-  total_filter_time_ += filter_timer_.ElapsedNanosSafe();
-  return to_delete;
-}
-
 } // namespace rocksdb
@@ -10,10 +10,8 @@
 #include <string>

 #include "db/dbformat.h"
-#include "rocksdb/compaction_filter.h"
 #include "rocksdb/env.h"
 #include "rocksdb/slice.h"
-#include "util/stop_watch.h"

 namespace rocksdb {

@@ -22,30 +20,20 @@ class Iterator;
 class Logger;
 class MergeOperator;
 class Statistics;
-class InternalIterator;

 class MergeHelper {
 public:
-  MergeHelper(Env* env, const Comparator* user_comparator,
-              const MergeOperator* user_merge_operator,
-              const CompactionFilter* compaction_filter, Logger* logger,
+  MergeHelper(const Comparator* user_comparator,
+              const MergeOperator* user_merge_operator, Logger* logger,
              unsigned min_partial_merge_operands,
-              bool assert_valid_internal_key, SequenceNumber latest_snapshot,
-              int level = 0, Statistics* stats = nullptr)
-      : env_(env),
-        user_comparator_(user_comparator),
+              bool assert_valid_internal_key)
+      : user_comparator_(user_comparator),
        user_merge_operator_(user_merge_operator),
-        compaction_filter_(compaction_filter),
        logger_(logger),
        min_partial_merge_operands_(min_partial_merge_operands),
        assert_valid_internal_key_(assert_valid_internal_key),
-        latest_snapshot_(latest_snapshot),
-        level_(level),
        keys_(),
-        operands_(),
-        filter_timer_(env_),
-        total_filter_time_(0U),
-        stats_(stats) {
+        operands_() {
    assert(user_comparator_ != nullptr);
  }

@@ -74,22 +62,15 @@ class MergeHelper {
  //                   0 means no restriction
  // at_bottom:   (IN) true if the iterator covers the bottem level, which means
  //                   we could reach the start of the history of this user key.
-  //
  // Returns one of the following statuses:
  // - OK: Entries were successfully merged.
  // - MergeInProgress: Put/Delete not encountered and unable to merge operands.
-  // - Corruption: Merge operator reported unsuccessful merge or a corrupted
-  //   key has been encountered and not expected (applies only when compiling
-  //   with asserts removed).
+  // - Corruption: Merge operator reported unsuccessful merge.
  //
  // REQUIRED: The first key in the input is not corrupted.
-  Status MergeUntil(InternalIterator* iter,
-                    const SequenceNumber stop_before = 0,
-                    const bool at_bottom = false);
-
-  // Filters a merge operand using the compaction filter specified
-  // in the constructor. Returns true if the operand should be filtered out.
-  bool FilterMerge(const Slice& user_key, const Slice& value_slice);
+  Status MergeUntil(Iterator* iter, const SequenceNumber stop_before = 0,
+                    const bool at_bottom = false, Statistics* stats = nullptr,
+                    Env* env_ = nullptr);

  // Query the merge result
  // These are valid until the next MergeUntil call
@@ -118,28 +99,19 @@ class MergeHelper {
  // TODO: Re-style this comment to be like the first one
  const std::deque<std::string>& keys() const { return keys_; }
  const std::deque<std::string>& values() const { return operands_; }
-  uint64_t TotalFilterTime() const { return total_filter_time_; }
  bool HasOperator() const { return user_merge_operator_ != nullptr; }

 private:
-  Env* env_;
  const Comparator* user_comparator_;
  const MergeOperator* user_merge_operator_;
-  const CompactionFilter* compaction_filter_;
  Logger* logger_;
  unsigned min_partial_merge_operands_;
  bool assert_valid_internal_key_; // enforce no internal key corruption?
-  SequenceNumber latest_snapshot_;
-  int level_;

  // the scratch area that holds the result of MergeUntil
  // valid up to the next MergeUntil call
  std::deque<std::string> keys_;    // Keeps track of the sequence of keys seen
  std::deque<std::string> operands_;  // Parallel with keys_; stores the values
-
-  StopWatchNano filter_timer_;
-  uint64_t total_filter_time_;
-  Statistics* stats_;
 };

 // MergeOutputIterator can be used to iterate over the result of a merge.
@@ -18,24 +18,37 @@ namespace rocksdb {

 class MergeHelperTest : public testing::Test {
 public:
-  MergeHelperTest() { env_ = Env::Default(); }
-
+  MergeHelperTest() = default;
  ~MergeHelperTest() = default;

-  Status Run(SequenceNumber stop_before, bool at_bottom,
-             SequenceNumber latest_snapshot = 0) {
-    iter_.reset(new test::VectorIterator(ks_, vs_));
-    iter_->SeekToFirst();
-    merge_helper_.reset(new MergeHelper(env_, BytewiseComparator(),
-                                        merge_op_.get(), filter_.get(), nullptr,
-                                        2U, false, latest_snapshot));
-    return merge_helper_->MergeUntil(iter_.get(), stop_before, at_bottom);
+  Status RunUInt64MergeHelper(SequenceNumber stop_before, bool at_bottom) {
+    InitIterator();
+    merge_op_ = MergeOperators::CreateUInt64AddOperator();
+    merge_helper_.reset(new MergeHelper(BytewiseComparator(), merge_op_.get(),
+                                        nullptr, 2U, false));
+    return merge_helper_->MergeUntil(iter_.get(), stop_before, at_bottom,
+                                     nullptr, Env::Default());
+  }
+
+  Status RunStringAppendMergeHelper(SequenceNumber stop_before,
+                                    bool at_bottom) {
+    InitIterator();
+    merge_op_ = MergeOperators::CreateStringAppendTESTOperator();
+    merge_helper_.reset(new MergeHelper(BytewiseComparator(), merge_op_.get(),
+                                        nullptr, 2U, false));
+    return merge_helper_->MergeUntil(iter_.get(), stop_before, at_bottom,
+                                     nullptr, Env::Default());
+  }
+
+  std::string Key(const std::string& user_key, const SequenceNumber& seq,
+      const ValueType& t) {
+    return InternalKey(user_key, seq, t).Encode().ToString();
  }

  void AddKeyVal(const std::string& user_key, const SequenceNumber& seq,
                 const ValueType& t, const std::string& val,
                 bool corrupt = false) {
-    InternalKey ikey(user_key, seq, t);
+    InternalKey ikey = InternalKey(user_key, seq, t);
    if (corrupt) {
      test::CorruptKeyType(&ikey);
    }
@@ -43,63 +56,66 @@ class MergeHelperTest : public testing::Test {
    vs_.push_back(val);
  }

-  Env* env_;
+  void InitIterator() {
+    iter_.reset(new test::VectorIterator(ks_, vs_));
+    iter_->SeekToFirst();
+  }
+
+  std::string EncodeInt(uint64_t x) {
+    std::string result;
+    PutFixed64(&result, x);
+    return result;
+  }
+
  std::unique_ptr<test::VectorIterator> iter_;
  std::shared_ptr<MergeOperator> merge_op_;
  std::unique_ptr<MergeHelper> merge_helper_;
  std::vector<std::string> ks_;
  std::vector<std::string> vs_;
-  std::unique_ptr<test::FilterNumber> filter_;
 };

 // If MergeHelper encounters a new key on the last level, we know that
 // the key has no more history and it can merge keys.
 TEST_F(MergeHelperTest, MergeAtBottomSuccess) {
-  merge_op_ = MergeOperators::CreateUInt64AddOperator();
+  AddKeyVal("a", 20, kTypeMerge, EncodeInt(1U));
+  AddKeyVal("a", 10, kTypeMerge, EncodeInt(3U));
+  AddKeyVal("b", 10, kTypeMerge, EncodeInt(4U));  // <- Iterator after merge

-  AddKeyVal("a", 20, kTypeMerge, test::EncodeInt(1U));
-  AddKeyVal("a", 10, kTypeMerge, test::EncodeInt(3U));
-  AddKeyVal("b", 10, kTypeMerge, test::EncodeInt(4U));  // <- iter_ after merge
-
-  ASSERT_TRUE(Run(0, true).ok());
+  ASSERT_TRUE(RunUInt64MergeHelper(0, true).ok());
  ASSERT_EQ(ks_[2], iter_->key());
-  ASSERT_EQ(test::KeyStr("a", 20, kTypeValue), merge_helper_->keys()[0]);
-  ASSERT_EQ(test::EncodeInt(4U), merge_helper_->values()[0]);
+  ASSERT_EQ(Key("a", 20, kTypeValue), merge_helper_->keys()[0]);
+  ASSERT_EQ(EncodeInt(4U), merge_helper_->values()[0]);
  ASSERT_EQ(1U, merge_helper_->keys().size());
  ASSERT_EQ(1U, merge_helper_->values().size());
 }

 // Merging with a value results in a successful merge.
 TEST_F(MergeHelperTest, MergeValue) {
-  merge_op_ = MergeOperators::CreateUInt64AddOperator();
+  AddKeyVal("a", 40, kTypeMerge, EncodeInt(1U));
+  AddKeyVal("a", 30, kTypeMerge, EncodeInt(3U));
+  AddKeyVal("a", 20, kTypeValue, EncodeInt(4U));  // <- Iterator after merge
+  AddKeyVal("a", 10, kTypeMerge, EncodeInt(1U));

-  AddKeyVal("a", 40, kTypeMerge, test::EncodeInt(1U));
-  AddKeyVal("a", 30, kTypeMerge, test::EncodeInt(3U));
-  AddKeyVal("a", 20, kTypeValue, test::EncodeInt(4U));  // <- iter_ after merge
-  AddKeyVal("a", 10, kTypeMerge, test::EncodeInt(1U));
-
-  ASSERT_TRUE(Run(0, false).ok());
+  ASSERT_TRUE(RunUInt64MergeHelper(0, false).ok());
  ASSERT_EQ(ks_[3], iter_->key());
-  ASSERT_EQ(test::KeyStr("a", 40, kTypeValue), merge_helper_->keys()[0]);
-  ASSERT_EQ(test::EncodeInt(8U), merge_helper_->values()[0]);
+  ASSERT_EQ(Key("a", 40, kTypeValue), merge_helper_->keys()[0]);
+  ASSERT_EQ(EncodeInt(8U), merge_helper_->values()[0]);
  ASSERT_EQ(1U, merge_helper_->keys().size());
  ASSERT_EQ(1U, merge_helper_->values().size());
 }

 // Merging stops before a snapshot.
 TEST_F(MergeHelperTest, SnapshotBeforeValue) {
-  merge_op_ = MergeOperators::CreateUInt64AddOperator();
+  AddKeyVal("a", 50, kTypeMerge, EncodeInt(1U));
+  AddKeyVal("a", 40, kTypeMerge, EncodeInt(3U));  // <- Iterator after merge
+  AddKeyVal("a", 30, kTypeMerge, EncodeInt(1U));
+  AddKeyVal("a", 20, kTypeValue, EncodeInt(4U));
+  AddKeyVal("a", 10, kTypeMerge, EncodeInt(1U));

-  AddKeyVal("a", 50, kTypeMerge, test::EncodeInt(1U));
-  AddKeyVal("a", 40, kTypeMerge, test::EncodeInt(3U));  // <- iter_ after merge
-  AddKeyVal("a", 30, kTypeMerge, test::EncodeInt(1U));
-  AddKeyVal("a", 20, kTypeValue, test::EncodeInt(4U));
-  AddKeyVal("a", 10, kTypeMerge, test::EncodeInt(1U));
-
-  ASSERT_TRUE(Run(31, true).IsMergeInProgress());
+  ASSERT_TRUE(RunUInt64MergeHelper(31, true).IsMergeInProgress());
  ASSERT_EQ(ks_[2], iter_->key());
-  ASSERT_EQ(test::KeyStr("a", 50, kTypeMerge), merge_helper_->keys()[0]);
-  ASSERT_EQ(test::EncodeInt(4U), merge_helper_->values()[0]);
+  ASSERT_EQ(Key("a", 50, kTypeMerge), merge_helper_->keys()[0]);
+  ASSERT_EQ(EncodeInt(4U), merge_helper_->values()[0]);
  ASSERT_EQ(1U, merge_helper_->keys().size());
  ASSERT_EQ(1U, merge_helper_->values().size());
 }
@@ -107,17 +123,15 @@ TEST_F(MergeHelperTest, SnapshotBeforeValue) {
 // MergeHelper preserves the operand stack for merge operators that
 // cannot do a partial merge.
 TEST_F(MergeHelperTest, NoPartialMerge) {
-  merge_op_ = MergeOperators::CreateStringAppendTESTOperator();
-
  AddKeyVal("a", 50, kTypeMerge, "v2");
-  AddKeyVal("a", 40, kTypeMerge, "v");  // <- iter_ after merge
+  AddKeyVal("a", 40, kTypeMerge, "v");  // <- Iterator after merge
  AddKeyVal("a", 30, kTypeMerge, "v");

-  ASSERT_TRUE(Run(31, true).IsMergeInProgress());
+  ASSERT_TRUE(RunStringAppendMergeHelper(31, true).IsMergeInProgress());
  ASSERT_EQ(ks_[2], iter_->key());
-  ASSERT_EQ(test::KeyStr("a", 40, kTypeMerge), merge_helper_->keys()[0]);
+  ASSERT_EQ(Key("a", 40, kTypeMerge), merge_helper_->keys()[0]);
  ASSERT_EQ("v", merge_helper_->values()[0]);
-  ASSERT_EQ(test::KeyStr("a", 50, kTypeMerge), merge_helper_->keys()[1]);
+  ASSERT_EQ(Key("a", 50, kTypeMerge), merge_helper_->keys()[1]);
  ASSERT_EQ("v2", merge_helper_->values()[1]);
  ASSERT_EQ(2U, merge_helper_->keys().size());
  ASSERT_EQ(2U, merge_helper_->values().size());
@@ -125,162 +139,44 @@ TEST_F(MergeHelperTest, NoPartialMerge) {

 // A single operand can not be merged.
 TEST_F(MergeHelperTest, SingleOperand) {
-  merge_op_ = MergeOperators::CreateUInt64AddOperator();
+  AddKeyVal("a", 50, kTypeMerge, EncodeInt(1U));

-  AddKeyVal("a", 50, kTypeMerge, test::EncodeInt(1U));
-
-  ASSERT_TRUE(Run(31, true).IsMergeInProgress());
+  ASSERT_TRUE(RunUInt64MergeHelper(31, true).IsMergeInProgress());
  ASSERT_FALSE(iter_->Valid());
-  ASSERT_EQ(test::KeyStr("a", 50, kTypeMerge), merge_helper_->keys()[0]);
-  ASSERT_EQ(test::EncodeInt(1U), merge_helper_->values()[0]);
+  ASSERT_EQ(Key("a", 50, kTypeMerge), merge_helper_->keys()[0]);
+  ASSERT_EQ(EncodeInt(1U), merge_helper_->values()[0]);
  ASSERT_EQ(1U, merge_helper_->keys().size());
  ASSERT_EQ(1U, merge_helper_->values().size());
 }

 // Merging with a deletion turns the deletion into a value
 TEST_F(MergeHelperTest, MergeDeletion) {
-  merge_op_ = MergeOperators::CreateUInt64AddOperator();
-
-  AddKeyVal("a", 30, kTypeMerge, test::EncodeInt(3U));
+  AddKeyVal("a", 30, kTypeMerge, EncodeInt(3U));
  AddKeyVal("a", 20, kTypeDeletion, "");

-  ASSERT_TRUE(Run(15, false).ok());
+  ASSERT_TRUE(RunUInt64MergeHelper(15, false).ok());
  ASSERT_FALSE(iter_->Valid());
-  ASSERT_EQ(test::KeyStr("a", 30, kTypeValue), merge_helper_->keys()[0]);
-  ASSERT_EQ(test::EncodeInt(3U), merge_helper_->values()[0]);
+  ASSERT_EQ(Key("a", 30, kTypeValue), merge_helper_->keys()[0]);
+  ASSERT_EQ(EncodeInt(3U), merge_helper_->values()[0]);
  ASSERT_EQ(1U, merge_helper_->keys().size());
  ASSERT_EQ(1U, merge_helper_->values().size());
 }

 // The merge helper stops upon encountering a corrupt key
 TEST_F(MergeHelperTest, CorruptKey) {
-  merge_op_ = MergeOperators::CreateUInt64AddOperator();
-
-  AddKeyVal("a", 30, kTypeMerge, test::EncodeInt(3U));
-  AddKeyVal("a", 25, kTypeMerge, test::EncodeInt(1U));
+  AddKeyVal("a", 30, kTypeMerge, EncodeInt(3U));
+  AddKeyVal("a", 25, kTypeMerge, EncodeInt(1U));
  // Corrupt key
-  AddKeyVal("a", 20, kTypeDeletion, "", true);  // <- iter_ after merge
+  AddKeyVal("a", 20, kTypeDeletion, "", true);  // <- Iterator after merge

-  ASSERT_TRUE(Run(15, false).IsMergeInProgress());
+  ASSERT_TRUE(RunUInt64MergeHelper(15, false).IsMergeInProgress());
  ASSERT_EQ(ks_[2], iter_->key());
-  ASSERT_EQ(test::KeyStr("a", 30, kTypeMerge), merge_helper_->keys()[0]);
-  ASSERT_EQ(test::EncodeInt(4U), merge_helper_->values()[0]);
+  ASSERT_EQ(Key("a", 30, kTypeMerge), merge_helper_->keys()[0]);
+  ASSERT_EQ(EncodeInt(4U), merge_helper_->values()[0]);
  ASSERT_EQ(1U, merge_helper_->keys().size());
  ASSERT_EQ(1U, merge_helper_->values().size());
 }

-// The compaction filter is called on every merge operand
-TEST_F(MergeHelperTest, FilterMergeOperands) {
-  merge_op_ = MergeOperators::CreateUInt64AddOperator();
-  filter_.reset(new test::FilterNumber(5U));
-
-  AddKeyVal("a", 30, kTypeMerge, test::EncodeInt(3U));
-  AddKeyVal("a", 29, kTypeMerge, test::EncodeInt(5U));  // Filtered
-  AddKeyVal("a", 28, kTypeMerge, test::EncodeInt(3U));
-  AddKeyVal("a", 27, kTypeMerge, test::EncodeInt(1U));
-  AddKeyVal("a", 26, kTypeMerge, test::EncodeInt(5U));  // Filtered
-  AddKeyVal("a", 25, kTypeValue, test::EncodeInt(1U));
-
-  ASSERT_TRUE(Run(15, false).ok());
-  ASSERT_FALSE(iter_->Valid());
-  MergeOutputIterator merge_output_iter(merge_helper_.get());
-  merge_output_iter.SeekToFirst();
-  ASSERT_EQ(test::KeyStr("a", 30, kTypeValue),
-            merge_output_iter.key().ToString());
-  ASSERT_EQ(test::EncodeInt(8U), merge_output_iter.value().ToString());
-  merge_output_iter.Next();
-  ASSERT_FALSE(merge_output_iter.Valid());
-}
-
-TEST_F(MergeHelperTest, FilterAllMergeOperands) {
-  merge_op_ = MergeOperators::CreateUInt64AddOperator();
-  filter_.reset(new test::FilterNumber(5U));
-
-  AddKeyVal("a", 30, kTypeMerge, test::EncodeInt(5U));
-  AddKeyVal("a", 29, kTypeMerge, test::EncodeInt(5U));
-  AddKeyVal("a", 28, kTypeMerge, test::EncodeInt(5U));
-  AddKeyVal("a", 27, kTypeMerge, test::EncodeInt(5U));
-  AddKeyVal("a", 26, kTypeMerge, test::EncodeInt(5U));
-  AddKeyVal("a", 25, kTypeMerge, test::EncodeInt(5U));
-
-  // filtered out all
-  ASSERT_TRUE(Run(15, false).ok());
-  ASSERT_FALSE(iter_->Valid());
-  MergeOutputIterator merge_output_iter(merge_helper_.get());
-  merge_output_iter.SeekToFirst();
-  ASSERT_FALSE(merge_output_iter.Valid());
-
-  // we have one operand that will survive because it's a delete
-  AddKeyVal("a", 24, kTypeDeletion, test::EncodeInt(5U));
-  AddKeyVal("b", 23, kTypeValue, test::EncodeInt(5U));
-  ASSERT_TRUE(Run(15, true).ok());
-  merge_output_iter = MergeOutputIterator(merge_helper_.get());
-  ASSERT_TRUE(iter_->Valid());
-  merge_output_iter.SeekToFirst();
-  ASSERT_FALSE(merge_output_iter.Valid());
-
-  // when all merge operands are filtered out, we leave the iterator pointing to
-  // the Put/Delete that survived
-  ASSERT_EQ(test::KeyStr("a", 24, kTypeDeletion), iter_->key().ToString());
-  ASSERT_EQ(test::EncodeInt(5U), iter_->value().ToString());
-}
-
-// Make sure that merge operands are filtered at the beginning
-TEST_F(MergeHelperTest, FilterFirstMergeOperand) {
-  merge_op_ = MergeOperators::CreateUInt64AddOperator();
-  filter_.reset(new test::FilterNumber(5U));
-
-  AddKeyVal("a", 31, kTypeMerge, test::EncodeInt(5U));  // Filtered
-  AddKeyVal("a", 30, kTypeMerge, test::EncodeInt(5U));  // Filtered
-  AddKeyVal("a", 29, kTypeMerge, test::EncodeInt(2U));
-  AddKeyVal("a", 28, kTypeMerge, test::EncodeInt(1U));
-  AddKeyVal("a", 27, kTypeMerge, test::EncodeInt(3U));
-  AddKeyVal("a", 26, kTypeMerge, test::EncodeInt(5U));  // Filtered
-  AddKeyVal("a", 25, kTypeMerge, test::EncodeInt(5U));  // Filtered
-  AddKeyVal("b", 24, kTypeValue, test::EncodeInt(5U));  // next user key
-
-  ASSERT_OK(Run(15, true));
-  ASSERT_TRUE(iter_->Valid());
-  MergeOutputIterator merge_output_iter(merge_helper_.get());
-  merge_output_iter.SeekToFirst();
-  // sequence number is 29 here, because the first merge operand got filtered
-  // out
-  ASSERT_EQ(test::KeyStr("a", 29, kTypeValue),
-            merge_output_iter.key().ToString());
-  ASSERT_EQ(test::EncodeInt(6U), merge_output_iter.value().ToString());
-  merge_output_iter.Next();
-  ASSERT_FALSE(merge_output_iter.Valid());
-
-  // make sure that we're passing user keys into the filter
-  ASSERT_EQ("a", filter_->last_merge_operand_key());
-}
-
-// Make sure that merge operands are not filtered out if there's a snapshot
-// pointing at them
-TEST_F(MergeHelperTest, DontFilterMergeOperandsBeforeSnapshotTest) {
-  merge_op_ = MergeOperators::CreateUInt64AddOperator();
-  filter_.reset(new test::FilterNumber(5U));
-
-  AddKeyVal("a", 31, kTypeMerge, test::EncodeInt(5U));
-  AddKeyVal("a", 30, kTypeMerge, test::EncodeInt(5U));
-  AddKeyVal("a", 29, kTypeMerge, test::EncodeInt(2U));
-  AddKeyVal("a", 28, kTypeMerge, test::EncodeInt(1U));
-  AddKeyVal("a", 27, kTypeMerge, test::EncodeInt(3U));
-  AddKeyVal("a", 26, kTypeMerge, test::EncodeInt(5U));
-  AddKeyVal("a", 25, kTypeMerge, test::EncodeInt(5U));
-  AddKeyVal("b", 24, kTypeValue, test::EncodeInt(5U));
-
-  ASSERT_OK(Run(15, true, 32));
-  ASSERT_TRUE(iter_->Valid());
-  MergeOutputIterator merge_output_iter(merge_helper_.get());
-  merge_output_iter.SeekToFirst();
-  ASSERT_EQ(test::KeyStr("a", 31, kTypeValue),
-            merge_output_iter.key().ToString());
-  ASSERT_EQ(test::EncodeInt(26U), merge_output_iter.value().ToString());
-  merge_output_iter.Next();
-  ASSERT_FALSE(merge_output_iter.Valid());
-}
-
 }  // namespace rocksdb

 int main(int argc, char** argv) {
@@ -20,6 +20,7 @@
 #include "utilities/merge_operators.h"
 #include "util/testharness.h"

+using namespace std;
 using namespace rocksdb;

 namespace {
@@ -75,7 +76,7 @@ class CountMergeOperator : public AssociativeMergeOperator {
 };

 namespace {
-std::shared_ptr<DB> OpenDb(const std::string& dbname, const bool ttl = false,
+std::shared_ptr<DB> OpenDb(const string& dbname, const bool ttl = false,
                           const size_t max_successive_merges = 0,
                           const uint32_t min_partial_merge_operands = 2) {
  DB* db;
@@ -89,7 +90,7 @@ std::shared_ptr<DB> OpenDb(const std::string& dbname, const bool ttl = false,
 // DBWithTTL is not supported in ROCKSDB_LITE
 #ifndef ROCKSDB_LITE
  if (ttl) {
-    std::cout << "Opening database with TTL\n";
+    cout << "Opening database with TTL\n";
    DBWithTTL* db_with_ttl;
    s = DBWithTTL::Open(options, dbname, &db_with_ttl);
    db = db_with_ttl;
@@ -101,7 +102,7 @@ std::shared_ptr<DB> OpenDb(const std::string& dbname, const bool ttl = false,
  s = DB::Open(options, dbname, &db);
 #endif  // !ROCKSDB_LITE
  if (!s.ok()) {
-    std::cerr << s.ToString() << std::endl;
+    cerr << s.ToString() << endl;
    assert(false);
  }
  return std::shared_ptr<DB>(db);
@@ -141,7 +142,7 @@ class Counters {
  // if the underlying level db operation failed.

  // mapped to a levedb Put
-  bool set(const std::string& key, uint64_t value) {
+  bool set(const string& key, uint64_t value) {
    // just treat the internal rep of int64 as the string
    Slice slice((char *)&value, sizeof(value));
    auto s = db_->Put(put_option_, key, slice);
@@ -149,26 +150,26 @@ class Counters {
    if (s.ok()) {
      return true;
    } else {
-      std::cerr << s.ToString() << std::endl;
+      cerr << s.ToString() << endl;
      return false;
    }
  }

  // mapped to a rocksdb Delete
-  bool remove(const std::string& key) {
+  bool remove(const string& key) {
    auto s = db_->Delete(delete_option_, key);

    if (s.ok()) {
      return true;
    } else {
-      std::cerr << s.ToString() << std::endl;
+      cerr << s.ToString() << std::endl;
      return false;
    }
  }

  // mapped to a rocksdb Get
-  bool get(const std::string& key, uint64_t* value) {
-    std::string str;
+  bool get(const string& key, uint64_t *value) {
+    string str;
    auto s = db_->Get(get_option_, key, &str);

    if (s.IsNotFound()) {
@@ -178,33 +179,35 @@ class Counters {
    } else if (s.ok()) {
      // deserialization
      if (str.size() != sizeof(uint64_t)) {
-        std::cerr << "value corruption\n";
+        cerr << "value corruption\n";
        return false;
      }
      *value = DecodeFixed64(&str[0]);
      return true;
    } else {
-      std::cerr << s.ToString() << std::endl;
+      cerr << s.ToString() << std::endl;
      return false;
    }
  }

  // 'add' is implemented as get -> modify -> set
  // An alternative is a single merge operation, see MergeBasedCounters
-  virtual bool add(const std::string& key, uint64_t value) {
+  virtual bool add(const string& key, uint64_t value) {
    uint64_t base = default_;
    return get(key, &base) && set(key, base + value);
  }


  // convenience functions for testing
-  void assert_set(const std::string& key, uint64_t value) {
+  void assert_set(const string& key, uint64_t value) {
    assert(set(key, value));
  }

-  void assert_remove(const std::string& key) { assert(remove(key)); }
+  void assert_remove(const string& key) {
+    assert(remove(key));
+  }

-  uint64_t assert_get(const std::string& key) {
+  uint64_t assert_get(const string& key) {
    uint64_t value = default_;
    int result = get(key, &value);
    assert(result);
@@ -212,7 +215,7 @@ class Counters {
    return value;
  }

-  void assert_add(const std::string& key, uint64_t value) {
+  void assert_add(const string& key, uint64_t value) {
    int result = add(key, value);
    assert(result);
    if (result == 0) exit(1); // Disable unused variable warning.
@@ -231,7 +234,7 @@ class MergeBasedCounters : public Counters {
  }

  // mapped to a rocksdb Merge operation
-  virtual bool add(const std::string& key, uint64_t value) override {
+  virtual bool add(const string& key, uint64_t value) override {
    char encoded[sizeof(uint64_t)];
    EncodeFixed64(encoded, value);
    Slice slice(encoded, sizeof(uint64_t));
@@ -240,7 +243,7 @@ class MergeBasedCounters : public Counters {
    if (s.ok()) {
      return true;
    } else {
-      std::cerr << s.ToString() << std::endl;
+      cerr << s.ToString() << endl;
      return false;
    }
  }
@@ -251,7 +254,7 @@ void dumpDb(DB* db) {
  auto it = unique_ptr<Iterator>(db->NewIterator(ReadOptions()));
  for (it->SeekToFirst(); it->Valid(); it->Next()) {
    uint64_t value = DecodeFixed64(it->value().data());
-    std::cout << it->key().ToString() << ": " << value << std::endl;
+    cout << it->key().ToString() << ": "  << value << endl;
  }
  assert(it->status().ok());  // Check for any errors found during the scan
 }
@@ -299,9 +302,9 @@ void testCounters(Counters& counters, DB* db, bool test_compaction) {
  if (test_compaction) {
    db->Flush(o);

-    std::cout << "Compaction started ...\n";
+    cout << "Compaction started ...\n";
    db->CompactRange(CompactRangeOptions(), nullptr, nullptr);
-    std::cout << "Compaction ended\n";
+    cout << "Compaction ended\n";

    dumpDb(db);

@@ -397,7 +400,7 @@ void testSingleBatchSuccessiveMerge(DB* db, size_t max_num_merges,

  // Get the value
  resetNumMergeOperatorCalls();
-  std::string get_value_str;
+  string get_value_str;
  {
    Status s = db->Get(ReadOptions(), key, &get_value_str);
    assert(s.ok());
@@ -409,24 +412,24 @@ void testSingleBatchSuccessiveMerge(DB* db, size_t max_num_merges,
            static_cast<size_t>((num_merges % (max_num_merges + 1))));
 }

-void runTest(int argc, const std::string& dbname, const bool use_ttl = false) {
+void runTest(int argc, const string& dbname, const bool use_ttl = false) {
  bool compact = false;
  if (argc > 1) {
    compact = true;
-    std::cout << "Turn on Compaction\n";
+    cout << "Turn on Compaction\n";
  }

  {
    auto db = OpenDb(dbname, use_ttl);

    {
-      std::cout << "Test read-modify-write counters... \n";
+      cout << "Test read-modify-write counters... \n";
      Counters counters(db, 0);
      testCounters(counters, db.get(), true);
    }

    {
-      std::cout << "Test merge-based counters... \n";
+      cout << "Test merge-based counters... \n";
      MergeBasedCounters counters(db, 0);
      testCounters(counters, db.get(), compact);
    }
@@ -435,7 +438,7 @@ void runTest(int argc, const std::string& dbname, const bool use_ttl = false) {
  DestroyDB(dbname, Options());

  {
-    std::cout << "Test merge in memtable... \n";
+    cout << "Test merge in memtable... \n";
    size_t max_merge = 5;
    auto db = OpenDb(dbname, use_ttl, max_merge);
    MergeBasedCounters counters(db, 0);
@@ -446,7 +449,7 @@ void runTest(int argc, const std::string& dbname, const bool use_ttl = false) {
  }

  {
-    std::cout << "Test Partial-Merge\n";
+    cout << "Test Partial-Merge\n";
    size_t max_merge = 100;
    for (uint32_t min_merge = 5; min_merge < 25; min_merge += 5) {
      for (uint32_t count = min_merge - 1; count <= min_merge + 1; count++) {
@@ -466,7 +469,7 @@ void runTest(int argc, const std::string& dbname, const bool use_ttl = false) {
  }

  {
-    std::cout << "Test merge-operator not set after reopen\n";
+    cout << "Test merge-operator not set after reopen\n";
    {
      auto db = OpenDb(dbname);
      MergeBasedCounters counters(db, 0);
@@ -486,7 +489,7 @@ void runTest(int argc, const std::string& dbname, const bool use_ttl = false) {

  /* Temporary remove this test
  {
-    std::cout << "Test merge-operator not set after reopen (recovery case)\n";
+    cout << "Test merge-operator not set after reopen (recovery case)\n";
    {
      auto db = OpenDb(dbname);
      MergeBasedCounters counters(db, 0);
@@ -1,119 +0,0 @@
-//  Copyright (c) 2015, Facebook, Inc.  All rights reserved.
-//  This source code is licensed under the BSD-style license found in the
-//  LICENSE file in the root directory of this source tree. An additional grant
-//  of patent rights can be found in the PATENTS file in the same directory.
-
-#ifndef ROCKSDB_LITE
-#include <string>
-
-#include "db/db_impl.h"
-#include "db/db_test_util.h"
-#include "rocksdb/options.h"
-#include "rocksdb/table.h"
-#include "util/testharness.h"
-
-namespace rocksdb {
-class OptionsFileTest : public testing::Test {
- public:
-  OptionsFileTest() : dbname_(test::TmpDir() + "/options_file_test") {}
-
-  std::string dbname_;
-};
-
-namespace {
-void UpdateOptionsFiles(DB* db,
-                        std::unordered_set<std::string>* filename_history,
-                        int* options_files_count) {
-  std::vector<std::string> filenames;
-  db->GetEnv()->GetChildren(db->GetName(), &filenames);
-  uint64_t number;
-  FileType type;
-  *options_files_count = 0;
-  for (auto filename : filenames) {
-    if (ParseFileName(filename, &number, &type) && type == kOptionsFile) {
-      filename_history->insert(filename);
-      (*options_files_count)++;
-    }
-  }
-}
-
-// Verify whether the current Options Files are the latest ones.
-void VerifyOptionsFileName(
-    DB* db, const std::unordered_set<std::string>& past_filenames) {
-  std::vector<std::string> filenames;
-  std::unordered_set<std::string> current_filenames;
-  db->GetEnv()->GetChildren(db->GetName(), &filenames);
-  uint64_t number;
-  FileType type;
-  for (auto filename : filenames) {
-    if (ParseFileName(filename, &number, &type) && type == kOptionsFile) {
-      current_filenames.insert(filename);
-    }
-  }
-  for (auto past_filename : past_filenames) {
-    if (current_filenames.find(past_filename) != current_filenames.end()) {
-      continue;
-    }
-    for (auto filename : current_filenames) {
-      ASSERT_GT(filename, past_filename);
-    }
-  }
-}
-}  // namespace
-
-TEST_F(OptionsFileTest, NumberOfOptionsFiles) {
-  const int kReopenCount = 20;
-  Options opt;
-  opt.create_if_missing = true;
-  DestroyDB(dbname_, opt);
-  std::unordered_set<std::string> filename_history;
-  DB* db;
-  for (int i = 0; i < kReopenCount; ++i) {
-    ASSERT_OK(DB::Open(opt, dbname_, &db));
-    int num_options_files = 0;
-    UpdateOptionsFiles(db, &filename_history, &num_options_files);
-    ASSERT_GT(num_options_files, 0);
-    ASSERT_LE(num_options_files, 2);
-    // Make sure we always keep the latest option files.
-    VerifyOptionsFileName(db, filename_history);
-    delete db;
-  }
-}
-
-TEST_F(OptionsFileTest, OptionsFileName) {
-  const uint64_t kOptionsFileNum = 12345;
-  uint64_t number;
-  FileType type;
-
-  auto options_file_name = OptionsFileName("", kOptionsFileNum);
-  ASSERT_TRUE(ParseFileName(options_file_name, &number, &type, nullptr));
-  ASSERT_EQ(type, kOptionsFile);
-  ASSERT_EQ(number, kOptionsFileNum);
-
-  const uint64_t kTempOptionsFileNum = 54352;
-  auto temp_options_file_name = TempOptionsFileName("", kTempOptionsFileNum);
-  ASSERT_TRUE(ParseFileName(temp_options_file_name, &number, &type, nullptr));
-  ASSERT_NE(temp_options_file_name.find(kTempFileNameSuffix),
-            std::string::npos);
-  ASSERT_EQ(type, kTempFile);
-  ASSERT_EQ(number, kTempOptionsFileNum);
-}
-}  // namespace rocksdb
-
-int main(int argc, char** argv) {
-#if !(defined NDEBUG) || !defined(OS_WIN)
-  ::testing::InitGoogleTest(&argc, argv);
-  return RUN_ALL_TESTS();
-#else
-  return 0;
-#endif  // !(defined NDEBUG) || !defined(OS_WIN)
-}
-#else
-
-#include <cstdio>
-
-int main(int argc, char** argv) {
-  printf("Skipped as Options file is not supported in RocksDBLite.\n");
-  return 0;
-}
-#endif  // !ROCKSDB_LITE
@@ -5,7 +5,6 @@
 //
 #include <algorithm>
 #include <iostream>
-#include <thread>
 #include <vector>

 #include "rocksdb/db.h"
@@ -445,7 +444,6 @@ void ProfileQueries(bool enabled_time = false) {
  }
 }

-#ifndef ROCKSDB_LITE
 TEST_F(PerfContextTest, KeyComparisonCount) {
  SetPerfLevel(kEnableCount);
  ProfileQueries();
@@ -456,7 +454,6 @@ TEST_F(PerfContextTest, KeyComparisonCount) {
  SetPerfLevel(kEnableTime);
  ProfileQueries(true);
 }
-#endif  // ROCKSDB_LITE

 // make perf_context_test
 // export ROCKSDB_TESTS=PerfContextTest.SeekKeyComparison
@@ -542,49 +539,6 @@ TEST_F(PerfContextTest, SeekKeyComparison) {
  }
 }

-TEST_F(PerfContextTest, DBMutexLockCounter) {
-  SetPerfLevel(kEnableTime);
-  int stats_code[] = {0, static_cast<int>(DB_MUTEX_WAIT_MICROS)};
-  for (int c = 0; c < 2; ++c) {
-    InstrumentedMutex mutex(nullptr, Env::Default(), stats_code[c]);
-    mutex.Lock();
-    std::thread child_thread([&] {
-      SetPerfLevel(kEnableTime);
-      perf_context.Reset();
-      ASSERT_EQ(perf_context.db_mutex_lock_nanos, 0);
-      mutex.Lock();
-      mutex.Unlock();
-      if (stats_code[c] == DB_MUTEX_WAIT_MICROS) {
-        // increment the counter only when it's a DB Mutex
-        ASSERT_GT(perf_context.db_mutex_lock_nanos, 0);
-      } else {
-        ASSERT_EQ(perf_context.db_mutex_lock_nanos, 0);
-      }
-    });
-    Env::Default()->SleepForMicroseconds(100);
-    mutex.Unlock();
-    child_thread.join();
-  }
-}
-
-TEST_F(PerfContextTest, FalseDBMutexWait) {
-  SetPerfLevel(kEnableTime);
-  int stats_code[] = {0, static_cast<int>(DB_MUTEX_WAIT_MICROS)};
-  for (int c = 0; c < 2; ++c) {
-    InstrumentedMutex mutex(nullptr, Env::Default(), stats_code[c]);
-    InstrumentedCondVar lock(&mutex);
-    perf_context.Reset();
-    mutex.Lock();
-    lock.TimedWait(100);
-    mutex.Unlock();
-    if (stats_code[c] == static_cast<int>(DB_MUTEX_WAIT_MICROS)) {
-      // increment the counter only when it's a DB Mutex
-      ASSERT_GT(perf_context.db_condition_wait_nanos, 0);
-    } else {
-      ASSERT_EQ(perf_context.db_condition_wait_nanos, 0);
-    }
-  }
-}
 }

 int main(int argc, char** argv) {
Author	SHA1	Message	Date
Yueh-Hsuan Chiang	a9a6a2f434	Merge pull request #764 from dmittendorf/fix-java-static-packaging Fix crossbuild jar packaging Conflicts: Makefile java/crossbuild/build-linux-centos.sh	2015-10-15 15:05:20 -07:00
Yueh-Hsuan Chiang	be5cb800ef	Merge pull request #759 from jwlent55/statically-load-compression-libraries Ensure that the compression libraries are statically linked into dyna…	2015-10-15 15:04:53 -07:00
agiardullo	0fb70a800e	Return MergeInProgress when fetching from transactions or WBWI with overwrite_key Summary: WriteBatchWithIndex::GetFromBatchAndDB only works correctly for overwrite_key=false. Transactions use overwrite_key=true (since WriteBatchWithIndex::GetIteratorWithBase only works when overwrite_key=true). So currently, Transactions could return incorrectly merged results when calling Get/GetForUpdate(). Until a permanent fix can be put in place, Transaction::Get[ForUpdate] and WriteBatchWithIndex::GetFromBatch[AndDB] will now return MergeInProgress if the most recent write to a key in the batch is a Merge. Test Plan: more tests Reviewers: sdong, yhchiang, rven, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D47817	2015-09-30 11:45:32 -07:00
agiardullo	8dbf5a39bd	Fix accidental object copy in transactions Summary: Should have used auto& instead of auto. Also needed to change the code a bit due to const correctness. Test Plan: unit tests Reviewers: sdong, igor, yoshinorim, yhchiang Reviewed By: yhchiang Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D47787	2015-09-30 11:37:27 -07:00
Yueh-Hsuan Chiang	0798a49a28	Make CompactionJobStatsTest.UniversalCompactionTest more robust Summary: CompactionJobStatsTest.UniversalCompactionTest assumes compaction kicks in when the number of L0 files equals to the compaction trigger. However, in some case, the compaction might not catch up the write speed and thus compaction might not kick in until the number of L0 files is GREATER than the compaction trigger. This patch tries to fix this corner case by making the Put thread wait for a potential compaction whenever it flushes. Test Plan: ./compaction_job_stats_test Reviewers: sdong, anthony, IslamAbdelRahman, igor Subscribers: dhruba Differential Revision: https://reviews.facebook.net/D47589	2015-09-30 11:15:35 -07:00
Islam AbdelRahman	555f3be85b	Clear SyncPoint Trace in DeleteSchedulerTests Summary: DeleteSchedulerTests is running the same test with different rates, After the first iteraton sync points become useless because ClearTrace was not being called Test Plan: Run the test Reviewers: sdong, yhchiang, igor Reviewed By: igor Subscribers: dhruba Differential Revision: https://reviews.facebook.net/D47709	2015-09-29 22:29:09 -07:00
Yueh-Hsuan Chiang	e29a73b89e	Fixed a tsan warning in db_stress.cc Summary: Fixed the following tsan warning in db_stress.cc WARNING: ThreadSanitizer: data race (pid=3163194) Read of size 8 at 0x7fd1797cb518 by thread T32: #0 VerifyDb tools/db_stress.cc:1731 (db_stress+0x000000040674) #1 rocksdb::StressTest::ThreadBody(void) tools/db_stress.cc:1191 (db_stress+0x0000000625a9) #2 StartThreadWrapper util/env_posix.cc:1648 (db_stress+0x00000028bbbd) Previous write of size 8 at 0x7fd1797cb518 by thread T31: #0 VerifyDb tools/db_stress.cc:1726 (db_stress+0x00000004072a) #1 rocksdb::StressTest::ThreadBody(void) tools/db_stress.cc:1191 (db_stress+0x0000000625a9) #2 StartThreadWrapper util/env_posix.cc:1648 (db_stress+0x00000028bbbd) The cause is that in VerifyDb(), the static local const variable long max_key can be read and written at the same time. This patch fixed it by making it non-static. Test Plan: db_stress Reviewers: igor, sdong, IslamAbdelRahman, anthony Reviewed By: anthony Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D47703	2015-09-29 22:28:45 -07:00
Yueh-Hsuan Chiang	0b7dd9131f	Fixed a bug which causes rocksdb.flush.write.bytes stat is always zero Summary: Fixed a bug which causes rocksdb.flush.write.bytes stat is always zero Test Plan: augment existing db_test Reviewers: sdong, anthony, IslamAbdelRahman, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D47595	2015-09-28 18:39:35 -07:00
sdong	8dcaa16b54	RandomAccessFileReader should not inherit RandomAccessFile Summary: RandomAccessFileReader unnecessarily inherited RandomAccessFile, which can introduce unnecessarily extra costs. Remove it. Test Plan: Run all existing tests Reviewers: yhchiang, anthony, igor, kradhakrishnan, rven, IslamAbdelRahman Reviewed By: IslamAbdelRahman Subscribers: leveldb, dhruba Differential Revision: https://reviews.facebook.net/D47409	2015-09-28 18:38:54 -07:00
Andres Noetzli	1d2d22e738	Fix clang-format on Travis Summary: Commit `1b598213aa` does not quite work as intended. This patch fixes Travis by adding clang-format to the packages list and excluding clang-format from the OSX build. In addition the build does not stop when bad formatting is detected but the build still fails. Test Plan: See https://travis-ci.org/facebook/rocksdb/builds/81037125 Reviewers: igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D47223	2015-09-28 18:37:54 -07:00
Andres Noetzli	fbd60cf3c4	Check formatting in Travis Summary: Check code formatting before running tests. Goal is to make it easier to deal with external pull requests. Test Plan: See https://travis-ci.org/facebook/rocksdb/builds/80952963 Reviewers: igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D47211	2015-09-28 18:37:38 -07:00
Andres Noetzli	c5024c03b6	Fix gflags build in Travis script Summary: There were two minor issues with `9566342d28`: - The install command was doing `cmake` instead of `cmake .` - `LIBRARY_PATH` was not set Fix both issues. Test Plan: Run tests on Travis: https://travis-ci.org/facebook/rocksdb/jobs/80699627 (bloom_test works now). Reviewers: igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D47103	2015-09-28 18:37:08 -07:00
Andres Noetzli	074154582a	Fix wrong constants in db_test_util Summary: Fix two constants in WaitFor() that multiply a value with 000 instead of 1000. Test Plan: make clean all check Reviewers: rven, anthony, yhchiang, aekmekji, sdong, MarkCallaghan, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D47091	2015-09-28 18:36:54 -07:00
Andres Noetzli	3967a4f0d1	Build gflags from source for Travis Summary: Unfortunately it is not easily possible to install gflags as a package on Travis. Change the Travis script to download and build gflags. This patch avoids "sudo make install" so that we can continue to use Travis' container-based infrastructure. Test Plan: Run tests on Travis. Reviewers: sdong, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D47037	2015-09-28 18:36:02 -07:00
Andres Noetzli	9a4d86b3da	Initialize variable to avoid warning Summary: RocksDB debug version failed to build under gcc-4.8.1 on sandcastle with the following error: ``` db/db_compaction_filter_test.cc:570:33: error: â€˜snapshotâ€™ may be used uninitialized in this function [-Werror=maybe-uninitialized] ``` Test Plan: make db_compaction_filter_test && ./db_compaction_filter_test Reviewers: rven, anthony, yhchiang, aekmekji, igor, sdong Reviewed By: igor, sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D46725	2015-09-16 15:55:30 -07:00
Igor Canadi	0833b75104	LogAndApply() should fail if the column family has been dropped Summary: This patch finally fixes the ColumnFamilyTest.ReadDroppedColumnFamily test. The test has been failing very sporadically and it was hard to repro. However, I managed to write a new tests that reproes the failure deterministically. Here's what happens: 1. We start the flush for the column family 2. We check if the column family was dropped here: https://github.com/facebook/rocksdb/blob/a3fc49bfddcdb1ff29409aacd06c04df56c7a1d7/db/flush_job.cc#L149 3. This check goes through, ends up in InstallMemtableFlushResults() and it goes into LogAndApply() 4. At about this time, we start dropping the column family. Dropping the column family process gets to LogAndApply() at about the same time as LogAndApply() from flush process 5. Drop column family goes through LogAndApply() first, marking the column family as dropped. 6. Flush process gets woken up and gets a chance to write to the MANIFEST. However, this is where it gets stuck: https://github.com/facebook/rocksdb/blob/a3fc49bfddcdb1ff29409aacd06c04df56c7a1d7/db/version_set.cc#L1975 7. We see that the column family was dropped, so there is no need to write to the MANIFEST. We return OK. 8. Flush gets OK back from LogAndApply() and it deletes the memtable, thinking that the data is now safely persisted to sst file. The fix is pretty simple. Instead of OK, we return ShutdownInProgress. This is not really true, but we have been using this status code to also mean "this operation was canceled because the column family has been dropped". The fix is only one LOC. All other code is related to tests. I added a new test that reproes the failure. I also moved SleepingBackgroundTask to util/testutil.h (because I needed it in column_family_test for my new test). There's plenty of other places where we reimplement SleepingBackgroundTask, but I'll address that in a separate commit. Test Plan: 1. new test 2. make check 3. Make sure the ColumnFamilyTest.ReadDroppedColumnFamily doesn't fail on Travis: https://travis-ci.org/facebook/rocksdb/jobs/79952386 Reviewers: yhchiang, anthony, IslamAbdelRahman, kradhakrishnan, rven, sdong Reviewed By: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D46773	2015-09-15 11:30:44 -07:00
Dmitry Marakasov	664acfa087	Fix printf format for size_t	2015-09-15 11:00:52 -07:00
Dmitry Marakasov	66a007f832	Fix `integer overflow in expression' error	2015-09-15 11:00:45 -07:00
sdong	5559488e92	DBImpl::FindObsoleteFiles() shouldn't release mutex between getting min_pending_output and scanning files Summary: Releasing mutex between getting min_pending_output and scanning files may cause min_pending_output to be max but some non-final files are found in file scanning, ending up with deleting wrong files. As a recent regression, mutex can be released while waiting for log sync. We move it to after file scanning. Test Plan: Run all existing tests. Don't think it is easy to write a unit test. Maybe we should find a way to assert lock not released so that we can have some test verification for similar cases. Reviewers: igor, anthony, IslamAbdelRahman, kradhakrishnan, yhchiang, kolmike, rven Reviewed By: rven Subscribers: leveldb, dhruba Differential Revision: https://reviews.facebook.net/D46899	2015-09-15 10:57:00 -07:00
Islam AbdelRahman	d0ce28bb27	Fix compaction_job_stats under ROCKSDB_LITE Summary: Fix compaction_job_stats under ROCKSDB_LITE Test Plan: compile using ROCKSDB_LITE Reviewers: yhchiang, igor, sdong Reviewed By: sdong Subscribers: dhruba Differential Revision: https://reviews.facebook.net/D46887	2015-09-15 10:50:14 -07:00
Andres Noetzli	248802038b	Relax asserts in arena_test Summary: Commit `c67d206898` did not fix all test conditions which use Arena::MemoryAllocatedBytes() (see Travis failure https://travis-ci.org/facebook/rocksdb/jobs/79957700). The assumption of that commit was that aligned allocations do not call Arena::AllocateNewBlock(), so malloc_usable_block_size() would not be used for Arena::MemoryAllocatedBytes(). However, there is a code path where Arena::AllocateAligned() calls AllocateFallback() which in turn calls Arena::AllocateNewBlock(), so Arena::MemoryAllocatedBytes() may return a greater value than expected even for aligned requests. Test Plan: make arena_test && ./arena_test Reviewers: rven, anthony, yhchiang, aekmekji, igor, sdong Reviewed By: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D46869	2015-09-15 10:50:09 -07:00
agiardullo	af06d1a871	Transactions: Release Locks when rolling back to a savepoint Summary: Transaction::RollbackToSavePoint() will now release any locks that were taken since the previous SavePoint. To do this cleanly, I moved tracked_keys_ management into TransactionBase. Test Plan: New Transaction test. Reviewers: igor, rven, sdong Reviewed By: sdong Subscribers: dhruba, spetrunia, leveldb Differential Revision: https://reviews.facebook.net/D46761	2015-09-15 10:50:01 -07:00
Andres Noetzli	5266ed3cc5	Fixed arena_test failure due to malloc_usable_size() Summary: ArenaTest.MemoryAllocatedBytes on Travis failed: https://travis-ci.org/facebook/rocksdb/jobs/79887849 . This is probably due to malloc_usable_size() returning a value greater than the requested size. From the man page: The value returned by malloc_usable_size() may be greater than the requested size of the allocation because of alignment and minimum size constraints. Although the excess bytes can be overwritten by the application without ill effects, this is not good programming practice: the number of excess bytes in an allocation depends on the underlying implementation. Test Plan: make arena_test && ./arena_test Reviewers: rven, anthony, yhchiang, aekmekji, sdong, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D46743	2015-09-15 10:49:53 -07:00
Andres Noetzli	7584091c7f	Fixed bug in compaction iterator Summary: During the refactoring, the condition that makes sure that compaction filters are only applied to records newer than the latest snapshot got butchered. This patch fixes the condition and adds a test case. Test Plan: make db_compaction_filter_test && ./db_compaction_filter_test Reviewers: rven, anthony, yhchiang, sdong, aekmekji, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D46707	2015-09-15 10:49:37 -07:00