Fix the thread wait case in error_handler (#7700)

Summary:
In error_handler auto recovery case, if recovery_in_prog_ is false, the recover is finished or failed. In this case, the auto recovery thread should finish its execution so recovery_thread_ should be null. However, in some cases, it is not null, the caller should not directly returned. Instead, it should wait for a while and create a new thread to execute the new recovery.

Pull Request resolved: #7700

Test Plan: make check, error_handler_fs_test

Reviewed By: anand1976

Differential Revision: D25098233

Pulled By: zhichao-cao

fbshipit-source-id: 5a1cba234ca18f6dd5d1be88e02d66e1d5ce931b

db/error_handler.cc

@@ -584,12 +584,20 @@ Status ErrorHandler::StartRecoverFromRetryableBGIOError(IOStatus io_error) {
   if (bg_error_.ok() || io_error.ok()) {
     return Status::OK();
   }
-  if (db_options_.max_bgerror_resume_count <= 0 || recovery_in_prog_ ||
-      recovery_thread_) {
+  if (db_options_.max_bgerror_resume_count <= 0 || recovery_in_prog_) {
     // Auto resume BG error is not enabled, directly return bg_error_.
     return bg_error_;
   }
 
+  if (recovery_thread_) {
+    // In this case, if recovery_in_prog_ is false, current thread should
+    // wait the previous recover thread to finish and create a new thread
+    // to recover from the bg error.
+    db_mutex_->Unlock();
+    recovery_thread_->join();
+    db_mutex_->Lock();
+  }
+
   recovery_in_prog_ = true;
   recovery_thread_.reset(
       new port::Thread(&ErrorHandler::RecoverFromRetryableBGIOError, this));