diff src/share/vm/gc_implementation/g1/sparsePRT.hpp @ 2173:97ba643ea3ed

7014261: G1: RSet-related failures Summary: A race between the concurrent cleanup thread and the VM thread while it is processing the "expanded sparse table list" causes both threads to try to free the same sparse table entry and either causes one of the threads to fail or leaves the entry in an inconsistent state. The solution is purge all entries on the expanded list that correspond go regions that are being cleaned up. Reviewed-by: brutisso, johnc
author tonyp
date Tue, 25 Jan 2011 17:58:19 -0500
parents f95d63e2154a
children d2a62e0f25eb
line wrap: on
line diff
--- a/src/share/vm/gc_implementation/g1/sparsePRT.hpp	Fri Jan 21 11:30:22 2011 -0500
+++ b/src/share/vm/gc_implementation/g1/sparsePRT.hpp	Tue Jan 25 17:58:19 2011 -0500
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2001, 2010, Oracle and/or its affiliates. All rights reserved.
+ * Copyright (c) 2001, 2011, Oracle and/or its affiliates. All rights reserved.
  * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
  *
  * This code is free software; you can redistribute it and/or modify it
@@ -212,8 +212,11 @@
 // mutex.
 
 class SparsePRTIter;
+class SparsePRTCleanupTask;
 
 class SparsePRT VALUE_OBJ_CLASS_SPEC {
+  friend class SparsePRTCleanupTask;
+
   //  Iterations are done on the _cur hash table, since they only need to
   //  see entries visible at the start of a collection pause.
   //  All other operations are done using the _next hash table.
@@ -238,6 +241,8 @@
   SparsePRT* next_expanded() { return _next_expanded; }
   void set_next_expanded(SparsePRT* nxt) { _next_expanded = nxt; }
 
+  bool should_be_on_expanded_list();
+
   static SparsePRT* _head_expanded_list;
 
 public:
@@ -284,12 +289,36 @@
   static void add_to_expanded_list(SparsePRT* sprt);
   static SparsePRT* get_from_expanded_list();
 
+  // The purpose of these three methods is to help the GC workers
+  // during the cleanup pause to recreate the expanded list, purging
+  // any tables from it that belong to regions that are freed during
+  // cleanup (if we don't purge those tables, there is a race that
+  // causes various crashes; see CR 7014261).
+  //
+  // We chose to recreate the expanded list, instead of purging
+  // entries from it by iterating over it, to avoid this serial phase
+  // at the end of the cleanup pause.
+  //
+  // The three methods below work as follows:
+  // * reset_for_cleanup_tasks() : Nulls the expanded list head at the
+  //   start of the cleanup pause.
+  // * do_cleanup_work() : Called by the cleanup workers for every
+  //   region that is not free / is being freed by the cleanup
+  //   pause. It creates a list of expanded tables whose head / tail
+  //   are on the thread-local SparsePRTCleanupTask object.
+  // * finish_cleanup_task() : Called by the cleanup workers after
+  //   they complete their cleanup task. It adds the local list into
+  //   the global expanded list. It assumes that the
+  //   ParGCRareEvent_lock is being held to ensure MT-safety.
+  static void reset_for_cleanup_tasks();
+  void do_cleanup_work(SparsePRTCleanupTask* sprt_cleanup_task);
+  static void finish_cleanup_task(SparsePRTCleanupTask* sprt_cleanup_task);
+
   bool contains_card(RegionIdx_t region_id, CardIdx_t card_index) const {
     return _next->contains_card(region_id, card_index);
   }
 };
 
-
 class SparsePRTIter: public RSHashTableIter {
 public:
   void init(const SparsePRT* sprt) {
@@ -300,4 +329,22 @@
   }
 };
 
+// This allows each worker during a cleanup pause to create a
+// thread-local list of sparse tables that have been expanded and need
+// to be processed at the beginning of the next GC pause. This lists
+// are concatenated into the single expanded list at the end of the
+// cleanup pause.
+class SparsePRTCleanupTask VALUE_OBJ_CLASS_SPEC {
+private:
+  SparsePRT* _head;
+  SparsePRT* _tail;
+
+public:
+  SparsePRTCleanupTask() : _head(NULL), _tail(NULL) { }
+
+  void add(SparsePRT* sprt);
+  SparsePRT* head() { return _head; }
+  SparsePRT* tail() { return _tail; }
+};
+
 #endif // SHARE_VM_GC_IMPLEMENTATION_G1_SPARSEPRT_HPP