> On Oct. 16, 2018, 1:47 p.m., Sahil Takiar wrote: > > ql/src/java/org/apache/hadoop/hive/ql/exec/spark/LocalHiveSparkClient.java > > Line 73 (original), 73 (patched) > > <https://reviews.apache.org/r/69022/diff/2/?file=2097990#file2097990line75> > > > > if we expect multiple sessions to access this, should we make this > > `volatile`? > > denys kuzmenko wrote: > it's being accesed only inside of the critical section (within the lock > boundaries) > > Sahil Takiar wrote: > does java guarantee that non-volatile variables accessed inside a > critical section are not cached locally by a CPU?
In short - yes. JSR 133 (Java Memory Model) Synchronization ensures that memory writes by a thread before or during a synchronized block are made visible in a predictable manner to other threads which synchronize on the same monitor. After we exit a synchronized block, we release the monitor, which has the effect of flushing the cache to main memory, so that writes made by this thread can be visible to other threads. Before we can enter a synchronized block, we acquire the monitor, which has the effect of invalidating the local processor cache so that variables will be reloaded from main memory. We will then be able to see all of the writes made visible by the previous release. - denys ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/69022/#review209617 ----------------------------------------------------------- On Oct. 16, 2018, 4:49 p.m., denys kuzmenko wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/69022/ > ----------------------------------------------------------- > > (Updated Oct. 16, 2018, 4:49 p.m.) > > > Review request for hive, Sahil Takiar and Adam Szita. > > > Bugs: HIVE-20737 > https://issues.apache.org/jira/browse/HIVE-20737 > > > Repository: hive-git > > > Description > ------- > > 1. Local SparkContext is shared between user sessions and should be closed > only when there is no active. > 2. Possible race condition in SparkSession.open() in case when user queries > run in parallel within the same session. > > > Diffs > ----- > > ql/src/java/org/apache/hadoop/hive/ql/exec/spark/LocalHiveSparkClient.java > 72ff53e3bd > > ql/src/java/org/apache/hadoop/hive/ql/exec/spark/session/SparkSessionImpl.java > bb50129518 > > ql/src/test/org/apache/hadoop/hive/ql/exec/spark/TestLocalHiveSparkClient.java > PRE-CREATION > > > Diff: https://reviews.apache.org/r/69022/diff/3/ > > > Testing > ------- > > Added TestLocalHiveSparkClient test > > > File Attachments > ---------------- > > HIVE-20737.7.patch > > https://reviews.apache.org/media/uploaded/files/2018/10/15/9cf8a2b3-9ec1-4316-81d0-3cd124b1a9fd__HIVE-20737.7.patch > > > Thanks, > > denys kuzmenko > >
