[ https://issues.apache.org/jira/browse/THRIFT-3768?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15251965#comment-15251965 ]
ASF GitHub Bot commented on THRIFT-3768: ---------------------------------------- Github user jeking3 commented on a diff in the pull request: https://github.com/apache/thrift/pull/980#discussion_r60590542 --- Diff: lib/cpp/src/thrift/concurrency/ThreadManager.cpp --- @@ -421,7 +416,7 @@ void ThreadManager::Impl::removeWorker(size_t value) { { Synchronized s(workerMonitor_); - while (workerCount_ != workerMaxCount_) { + while (workerCount_ > goalCount) { --- End diff -- This routine blocks until there are fewer actual workers than the number of workers when it was called less the value passed in. On a busy system where workers are being added this means it might be in here for a while. I did not change this behavior; simply I ensured that the effects of multiple threads calling removeWorker() at the same time don't clobber each-other's desired goal of how many to remove. It would be better to have two calls; one to trim the maximum number of workers but it doesn't block; another being a barrier you can use to get to that number. Callers may want to trim without blocking, for example. > TThreadedServer may crash if it is destroyed immediately after it returns > from serve(); TThreadedServer disconnects clients when they connec > -------------------------------------------------------------------------------------------------------------------------------------------- > > Key: THRIFT-3768 > URL: https://issues.apache.org/jira/browse/THRIFT-3768 > Project: Thrift > Issue Type: Bug > Components: C++ - Library > Affects Versions: 0.9.3 > Reporter: Ted Wang > Assignee: James E. King, III > Priority: Minor > > Here's a sequence that shows the race: > Thread-1 (Users of TThreadedServer): Calls TThreadedServer::stop(), which > calls interruptChildren and initiates the tearing down of client connections. > Thread-2: In TServerFramework::serve(), broke out of accept, and now blocks > in TThreadedServer::serve() waiting to drain all the clients. > Thread-3 (The connected client thread created by TThreadedServer): In > disposeConnectedClient, running because the server is shutting down and the > shared_ptr specified this function to be the cleanup function for the client. > This thread just returned from onClientDisconnected and now context switches. > Thread-2: TThreadedServer::serve() is notified that all of the clients have > disconnected and completes. > Thread-1: Joins on Thread-2 and destroys the server object because it is done. > Thread-3: Finally gets a chance to run, but now encounters undefined behavior > because it is still executing a member function of an object that has been > destroyed. > You can force this race in action if you put sleep(1) before > onClientDisconnected() in disposeConnectedClient -- This message was sent by Atlassian JIRA (v6.3.4#6332)