- 28 1月, 2015 4 次提交
-
-
由 sdong 提交于
Summary: I'm not sure the expected results of std::atomic::fetch_sub() when using memory_order_relaxed, and I suspect TSAN complains. Test Plan: make all check Reviewers: rven, yhchiang, igor Reviewed By: igor Subscribers: leveldb, dhruba Differential Revision: https://reviews.facebook.net/D32259
-
由 sdong 提交于
Summary: ThreadSanitizer complains data race of super version and version's destructor with Get(). This patch will fix those warning. The warning is likely from ColumnFamilyData::ReturnThreadLocalSuperVersion(). With relaxed consistency of CAS, reading the data of the super version can technically happen after swapping it in, enabling the background thread to clean it up. Test Plan: make all check Reviewers: rven, igor, yhchiang Reviewed By: yhchiang Subscribers: leveldb, dhruba Differential Revision: https://reviews.facebook.net/D32265
-
由 sdong 提交于
Summary: We see failure of the test in travis but I can't repro it. Add more logging in failure cases to help us figure out which failure it is. Also makes synchronization slightly stronger, though there isn't seem to be a problem without it Test Plan: Run the test Reviewers: rven, yhchiang, igor Reviewed By: igor Subscribers: leveldb, dhruba Differential Revision: https://reviews.facebook.net/D32319
-
由 Yueh-Hsuan Chiang 提交于
Summary: Update the comment for the removal of mac-install-gflags.sh Test Plan: n/a Reviewers: igor Reviewed By: igor Subscribers: dhruba Differential Revision: https://reviews.facebook.net/D32295
-
- 27 1月, 2015 5 次提交
-
-
由 sdong 提交于
Summary: log_dir_unsynced_ is a confusing name. Rename it to log_dir_synced_ and flip the value. Test Plan: Run ./fault_injection_test Reviewers: rven, yhchiang, igor Reviewed By: igor Subscribers: leveldb, dhruba Differential Revision: https://reviews.facebook.net/D32235
-
由 sdong 提交于
Summary: Currently fault_injection_test has a test case to drop all the unsynced data. Add one more case to take a randomized bytes from it. Test Plan: Run the test Reviewers: rven, yhchiang, igor Reviewed By: igor Subscribers: leveldb, dhruba Differential Revision: https://reviews.facebook.net/D32229
-
由 sdong 提交于
Summary: 1. If WAL directory is different from db directory. Sync the directory after creating a log file under it. 2. After creating an SST file, sync its parent directory instead of DB directory. 3. change the check of kResetDeleteUnsyncedFiles in fault_injection_test. Since we changed the behavior to sync log files' parent directory after first WAL sync, instead of creating, kResetDeleteUnsyncedFiles will not guarantee to show post sync updates. Test Plan: make all check Reviewers: yhchiang, rven, igor Reviewed By: igor Subscribers: leveldb, dhruba Differential Revision: https://reviews.facebook.net/D32067
-
由 Igor Canadi 提交于
-
由 Igor Canadi 提交于
Summary: This is first in a series of diffs that fixes data races detected by thread sanitizer. Here the problem is that we call Ref() on a column family during a single-threaded write, without holding a mutex. Test Plan: TSAN is no longer complaining about LevelLimitReopen. Reviewers: yhchiang, rven, sdong Reviewed By: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D32121
-
- 26 1月, 2015 1 次提交
-
-
由 Yueh-Hsuan Chiang 提交于
[RocksJava] Fix native library loader
-
- 24 1月, 2015 11 次提交
-
-
由 fyrz 提交于
Summary: Prior to this the native library loader instance didn`t care about a state. So if library loading was called multiple times, multiple copies of the shared object were put into the tmp folder and loaded into the JVM. This changed within this commit to the following behavior: - library loading is now synchronized - library is loaded within the first call - if loading was successful the library loaded sets a flag - every subsequent call checks for a boolean flag indicating if there was already a successful attempt Test Plan: - Execute example and watch tmp folder while the example is running - After this patch only one shared object will be in the tmp folder Usual tests: - make rocksdbjava jtest - mvn -f rocksjni.pom package Reviewers: adamretter, ankgup87, yhchiang Subscribers: dhruba Differential Revision: https://reviews.facebook.net/D32133
-
由 Igor Canadi 提交于
-
由 Igor Canadi 提交于
Summary: We should not be calling InternalStats methods outside of the mutex. Test Plan: COMPILE_WITH_TSAN=1 m db_test && ROCKSDB_TESTS=CompactionTrigger ./db_test failing before the diff, works now Reviewers: yhchiang, rven, sdong Reviewed By: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D32127
-
由 Igor Canadi 提交于
Summary: More race condition bugs with our archive WAL files. I do believe this caused t5988326, but can't reproduce the failure unfortunately. Test Plan: make check Reviewers: yhchiang, rven, sdong Reviewed By: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D32103
-
由 sdong 提交于
Summary: 3 iterations were disabled by mistake by one recent commit, causing CLANG build error. Fix it Test Plan: USE_CLANG=1 make fault_injection_test and run the test Reviewers: igor Reviewed By: igor Subscribers: leveldb, dhruba Differential Revision: https://reviews.facebook.net/D32109
-
由 sdong 提交于
Summary: fault_injection_test.cc has two variable names not following the convention fix it. Test Plan: run the test Reviewers: rven, yhchiang, igor Reviewed By: igor Subscribers: leveldb, dhruba Differential Revision: https://reviews.facebook.net/D32097
-
由 sdong 提交于
Summary: Wrapper classes in fault_injection_test doesn't simulate RocksDB Env behavior close enough. Improve it by: (1) when fsync, don't sync parent (2) support directory fsync (3) support multiple directories Add test cases of (1) persisting by WAL fsync, not just compact range (2) different WAL dir (3) combination of (1) and (2) (4) data directory is not the same as db name. Test Plan: Run the test and make sure it passes. Reviewers: rven, yhchiang, igor Subscribers: leveldb, dhruba Differential Revision: https://reviews.facebook.net/D32031
-
由 Igor Canadi 提交于
Summary: It looks like ASAN with gcc 4.9 works better than 4.8.1. It detected this possibility of heap buffer overflow. This was in our codebase for a year :) Test Plan: COMPILE_WITH_ASAN=1 make backupable_db && ./backupable_db Reviewers: yhchiang, rven, sdong Reviewed By: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D32085
-
由 Igor Canadi 提交于
Summary: We need this because we build MySQL with 4.8.1. Test Plan: ROCKSDB_FBCODE_BUILD_WITH_481=1 make check Reviewers: sdong, yhchiang, rven, yoshinorim Reviewed By: yoshinorim Subscribers: jonahcohen, yoshinorim, dhruba, leveldb Differential Revision: https://reviews.facebook.net/D32073
-
由 Igor Canadi 提交于
Summary: When you compile with COMPILE_WITH_TSAN=1, we will compile the code with -fsanitize=thread. This will resolve bunch of data race issues we might have. Test Plan: COMPILE_WITH_TSAN=1 m db_test Reviewers: yhchiang, rven, sdong Reviewed By: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D32019
-
由 Igor Canadi 提交于
Summary: Upgrade gcc to 4.9.1 and clang to dev. With new compilers I succeeded to run thread sanitizer, too. I'll post output (doesn't look good) and fix some things in separate diffs. Test Plan: compiles with both g++ and clang Reviewers: yhchiang, rven, sdong Reviewed By: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D32007
-
- 23 1月, 2015 14 次提交
-
-
由 Yueh-Hsuan Chiang 提交于
[RocksJava] Allow using ReadOptions with iterators
-
由 fyrz 提交于
The methods: - newIterator - iterators support now also ReadOptions. That allows a user of the Java API to retrieve RocksIterator instances on a snapshot.
-
由 Yueh-Hsuan Chiang 提交于
Summary: Make options_test runnable on ROCKSDB_LITE by blocking those tests that require non-ROCKSDB_LITE feature. Test Plan: make options_test OPT=-DROCKSDB_LITE -j32 ./options_test make clean make options_test -j32 ./options_test Reviewers: sdong, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D32025
-
由 Yueh-Hsuan Chiang 提交于
[RocksJava] Format version support in BlockBasedTableConfig
-
由 Yueh-Hsuan Chiang 提交于
-
由 Yueh-Hsuan Chiang 提交于
-
由 Yueh-Hsuan Chiang 提交于
[RocksJava] ColumnFamily name JNI correction
-
由 Igor Canadi 提交于
-
由 fyrz 提交于
- added spaces between operators - removed obsolete text in JavaDoc
-
由 fyrz 提交于
- Added support for format version in BlockBasedTableConfig
-
由 fyrz 提交于
-
由 sdong 提交于
Summary: Now we don't sync manifest file when initializing it, so DB cannot be safely reopened before the first mem table flush. Fix it by syncing it. This fixes fault_injection_test. Test Plan: make all check Reviewers: rven, yhchiang, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D32001
-
由 fyrz 提交于
Summary: Previous to this commit there was a problem with unterminated String usage as jByteArrays are not zero terminated. Test Plan: make rocksdbjava make jtest mvn -f rocksjni.pom package Reviewers: yhchiang, adamretter, ankgup87 Subscribers: dhruba Differential Revision: https://reviews.facebook.net/D31809
-
由 fyrz 提交于
Previous to this commit there was a problem with unterminated String usage as jByteArrays are not zero terminated.
-
- 22 1月, 2015 3 次提交
-
-
由 Igor Canadi 提交于
-
由 Igor Canadi 提交于
Summary: I saw this when running readrandom benchmark with corrupted database -- benchmark worked! If a Get() returns corruption we should probably abort. Test Plan: compiles Reviewers: yhchiang, rven, sdong Reviewed By: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D31701
-
由 sdong 提交于
Summary: GetLiveFilesMetaData() already adds a leading "/" in file name. No need to add one extra "/" in DBImpl::CheckConsistency() Test Plan: make all check Reviewers: yhchiang, rven, igor Reviewed By: igor Subscribers: leveldb, dhruba Differential Revision: https://reviews.facebook.net/D31779
-
- 20 1月, 2015 2 次提交
-
-
由 Yueh-Hsuan Chiang 提交于
[RocksJava] TTL DB support
-
由 fyrz 提交于
-