8076289: Move the StrongRootsScope out of SharedHeap

Reviewed-by: stefank, sjohanss, david

hotspot/src/share/vm/gc_implementation/concurrentMarkSweep/concurrentMarkSweepGeneration.cpp

@@ -53,6 +53,7 @@
 #include "memory/padded.hpp"
 #include "memory/referencePolicy.hpp"
 #include "memory/resourceArea.hpp"
+#include "memory/strongRootsScope.hpp"
 #include "memory/tenuredGeneration.hpp"
 #include "oops/oop.inline.hpp"
 #include "prims/jvmtiExport.hpp"
@@ -3014,10 +3015,10 @@ void CMSCollector::checkpointRootsInitialWork() {
       gch->set_par_threads(n_workers);
       initialize_sequential_subtasks_for_young_gen_rescan(n_workers);
       if (n_workers > 1) {
-        GenCollectedHeap::StrongRootsScope srs(gch);
+        StrongRootsScope srs;
         workers->run_task(&tsk);
       } else {
-        GenCollectedHeap::StrongRootsScope srs(gch);
+        StrongRootsScope srs;
         tsk.work(0);
       }
       gch->set_par_threads(0);
@@ -5112,11 +5113,11 @@ void CMSCollector::do_remark_parallel() {
     // necessarily be so, since it's possible that we are doing
     // ST marking.
     ReferenceProcessorMTDiscoveryMutator mt(ref_processor(), true);
-    GenCollectedHeap::StrongRootsScope srs(gch);
+    StrongRootsScope srs;
     workers->run_task(&tsk);
   } else {
     ReferenceProcessorMTDiscoveryMutator mt(ref_processor(), false);
-    GenCollectedHeap::StrongRootsScope srs(gch);
+    StrongRootsScope srs;
     tsk.work(0);
   }
 
@@ -5184,7 +5185,7 @@ void CMSCollector::do_remark_non_parallel() {
     verify_work_stacks_empty();
 
     gch->rem_set()->prepare_for_younger_refs_iterate(false); // Not parallel.
-    GenCollectedHeap::StrongRootsScope srs(gch);
+    StrongRootsScope srs;
 
     gch->gen_process_roots(_cmsGen->level(),
                            true,  // younger gens as roots

hotspot/src/share/vm/gc_implementation/g1/concurrentMark.cpp

@@ -46,6 +46,7 @@
 #include "memory/genOopClosures.inline.hpp"
 #include "memory/referencePolicy.hpp"
 #include "memory/resourceArea.hpp"
+#include "memory/strongRootsScope.hpp"
 #include "oops/oop.inline.hpp"
 #include "runtime/handles.inline.hpp"
 #include "runtime/java.hpp"
@@ -2650,7 +2651,7 @@ void ConcurrentMark::checkpointRootsFinalWork() {
 
   g1h->ensure_parsability(false);
 
-  G1CollectedHeap::StrongRootsScope srs(g1h);
+  StrongRootsScope srs;
   // this is remark, so we'll use up all active threads
   uint active_workers = g1h->workers()->active_workers();
   if (active_workers == 0) {

hotspot/src/share/vm/gc_implementation/g1/g1RootProcessor.cpp

@@ -116,7 +116,7 @@ void G1RootProcessor::wait_until_all_strong_classes_discovered() {
 G1RootProcessor::G1RootProcessor(G1CollectedHeap* g1h) :
     _g1h(g1h),
     _process_strong_tasks(new SubTasksDone(G1RP_PS_NumElements)),
-    _srs(g1h),
+    _srs(),
     _lock(Mutex::leaf, "G1 Root Scanning barrier lock", false, Monitor::_safepoint_check_never),
     _n_workers_discovered_strong_classes(0) {}
 

hotspot/src/share/vm/gc_implementation/g1/g1RootProcessor.hpp

@@ -26,7 +26,7 @@
 #define SHARE_VM_GC_IMPLEMENTATION_G1_ROOTPROCESSOR_HPP
 
 #include "memory/allocation.hpp"
-#include "memory/sharedHeap.hpp"
+#include "memory/strongRootsScope.hpp"
 #include "runtime/mutex.hpp"
 
 class CLDClosure;
@@ -46,7 +46,7 @@ class SubTasksDone;
 class G1RootProcessor : public StackObj {
   G1CollectedHeap* _g1h;
   SubTasksDone* _process_strong_tasks;
-  SharedHeap::StrongRootsScope _srs;
+  StrongRootsScope _srs;
 
   // Used to implement the Thread work barrier.
   Monitor _lock;

hotspot/src/share/vm/gc_implementation/parNew/parNewGeneration.cpp

@@ -42,7 +42,7 @@
 #include "memory/generation.hpp"
 #include "memory/referencePolicy.hpp"
 #include "memory/resourceArea.hpp"
-#include "memory/sharedHeap.hpp"
+#include "memory/strongRootsScope.hpp"
 #include "memory/space.hpp"
 #include "oops/objArrayOop.hpp"
 #include "oops/oop.inline.hpp"
@@ -974,10 +974,10 @@ void ParNewGeneration::collect(bool   full,
   // in the multi-threaded case, but we special-case n=1 here to get
   // repeatable measurements of the 1-thread overhead of the parallel code.
   if (n_workers > 1) {
-    GenCollectedHeap::StrongRootsScope srs(gch);
+    StrongRootsScope srs;
     workers->run_task(&tsk);
   } else {
-    GenCollectedHeap::StrongRootsScope srs(gch);
+    StrongRootsScope srs;
     tsk.work(0);
   }
   thread_state_set.reset(0 /* Bad value in debug if not reset */,

hotspot/src/share/vm/gc_implementation/parallelScavenge/parallelScavengeHeap.hpp

@@ -34,6 +34,7 @@
 #include "gc_implementation/shared/gcWhen.hpp"
 #include "gc_interface/collectedHeap.inline.hpp"
 #include "memory/collectorPolicy.hpp"
+#include "memory/strongRootsScope.hpp"
 #include "utilities/ostream.hpp"
 
 class AdjoiningGenerations;
@@ -237,7 +238,7 @@ class ParallelScavengeHeap : public CollectedHeap {
   void gen_mangle_unused_area() PRODUCT_RETURN;
 
   // Call these in sequential code around the processing of strong roots.
-  class ParStrongRootsScope : public MarkingCodeBlobClosure::MarkScope {
+  class ParStrongRootsScope : public MarkScope {
    public:
     ParStrongRootsScope();
     ~ParStrongRootsScope();

hotspot/src/share/vm/memory/genCollectedHeap.cpp

@@ -39,7 +39,7 @@
 #include "memory/genOopClosures.inline.hpp"
 #include "memory/generationSpec.hpp"
 #include "memory/resourceArea.hpp"
-#include "memory/sharedHeap.hpp"
+#include "memory/strongRootsScope.hpp"
 #include "memory/space.hpp"
 #include "oops/oop.inline.hpp"
 #include "runtime/biasedLocking.hpp"
@@ -596,7 +596,7 @@ void GenCollectedHeap::process_roots(bool activate_scope,
                                      CLDClosure* strong_cld_closure,
                                      CLDClosure* weak_cld_closure,
                                      CodeBlobClosure* code_roots) {
-  StrongRootsScope srs(this, activate_scope);
+  StrongRootsScope srs(activate_scope);
 
   // General roots.
   assert(Threads::thread_claim_parity() != 0, "must have called prologue code");

hotspot/src/share/vm/memory/iterator.cpp

@@ -64,16 +64,6 @@ void CodeBlobToOopClosure::do_code_blob(CodeBlob* cb) {
   }
 }
 
-MarkingCodeBlobClosure::MarkScope::MarkScope(bool activate)
-  : _active(activate)
-{
-  if (_active)  nmethod::oops_do_marking_prologue();
-}
-
-MarkingCodeBlobClosure::MarkScope::~MarkScope() {
-  if (_active)  nmethod::oops_do_marking_epilogue();
-}
-
 void MarkingCodeBlobClosure::do_code_blob(CodeBlob* cb) {
   nmethod* nm = cb->as_nmethod_or_null();
   if (nm != NULL && !nm->test_set_oops_do_mark()) {

hotspot/src/share/vm/memory/iterator.hpp

@@ -288,16 +288,6 @@ class MarkingCodeBlobClosure : public CodeBlobToOopClosure {
   // Called for each code blob, but at most once per unique blob.
 
   virtual void do_code_blob(CodeBlob* cb);
-
-  class MarkScope : public StackObj {
-  protected:
-    bool _active;
-  public:
-    MarkScope(bool activate = true);
-      // = { if (active) nmethod::oops_do_marking_prologue(); }
-    ~MarkScope();
-      // = { if (active) nmethod::oops_do_marking_epilogue(); }
-  };
 };
 
 // MonitorClosure is used for iterating over monitors in the monitors cache

hotspot/src/share/vm/memory/sharedHeap.cpp

@@ -23,32 +23,9 @@
  */
 
 #include "precompiled.hpp"
-#include "classfile/stringTable.hpp"
+#include "gc_interface/collectedHeap.hpp"
-#include "classfile/systemDictionary.hpp"
-#include "code/codeCache.hpp"
-#include "gc_interface/collectedHeap.inline.hpp"
 #include "memory/sharedHeap.hpp"
-#include "oops/oop.inline.hpp"
-#include "runtime/atomic.inline.hpp"
-#include "runtime/fprofiler.hpp"
-#include "runtime/java.hpp"
-#include "utilities/copy.hpp"
-#include "utilities/workgroup.hpp"
 
 SharedHeap::SharedHeap() :
   CollectedHeap()
 {}
-
-SharedHeap::StrongRootsScope::StrongRootsScope(SharedHeap* heap, bool activate)
-  : MarkScope(activate), _sh(heap)
-{
-  if (_active) {
-    Threads::change_thread_claim_parity();
-    // Zero the claimed high water mark in the StringTable
-    StringTable::clear_parallel_claimed_index();
-  }
-}
-
-SharedHeap::StrongRootsScope::~StrongRootsScope() {
-  Threads::assert_all_threads_claimed();
-}

hotspot/src/share/vm/memory/sharedHeap.hpp

@@ -27,62 +27,6 @@
 
 #include "gc_interface/collectedHeap.hpp"
 
-// A "SharedHeap" is an implementation of a java heap for HotSpot.  This
-// is an abstract class: there may be many different kinds of heaps.  This
-// class defines the functions that a heap must implement, and contains
-// infrastructure common to all heaps.
-
-// Note on use of FlexibleWorkGang's for GC.
-// There are three places where task completion is determined.
-// In
-//    1) ParallelTaskTerminator::offer_termination() where _n_threads
-//    must be set to the correct value so that count of workers that
-//    have offered termination will exactly match the number
-//    working on the task.  Tasks such as those derived from GCTask
-//    use ParallelTaskTerminator's.  Tasks that want load balancing
-//    by work stealing use this method to gauge completion.
-//    2) SubTasksDone has a variable _n_threads that is used in
-//    all_tasks_completed() to determine completion.  all_tasks_complete()
-//    counts the number of tasks that have been done and then reset
-//    the SubTasksDone so that it can be used again.  When the number of
-//    tasks is set to the number of GC workers, then _n_threads must
-//    be set to the number of active GC workers. G1RootProcessor and
-//    GenCollectedHeap have SubTasksDone.
-//    3) SequentialSubTasksDone has an _n_threads that is used in
-//    a way similar to SubTasksDone and has the same dependency on the
-//    number of active GC workers.  CompactibleFreeListSpace and Space
-//    have SequentialSubTasksDone's.
-//
-// Examples of using SubTasksDone and SequentialSubTasksDone:
-//  G1RootProcessor and GenCollectedHeap::process_roots() use
-//  SubTasksDone* _process_strong_tasks to claim tasks for workers
-//
-//  GenCollectedHeap::gen_process_roots() calls
-//      rem_set()->younger_refs_iterate()
-//  to scan the card table and which eventually calls down into
-//  CardTableModRefBS::par_non_clean_card_iterate_work().  This method
-//  uses SequentialSubTasksDone* _pst to claim tasks.
-//  Both SubTasksDone and SequentialSubTasksDone call their method
-//  all_tasks_completed() to count the number of GC workers that have
-//  finished their work.  That logic is "when all the workers are
-//  finished the tasks are finished".
-//
-//  The pattern that appears  in the code is to set _n_threads
-//  to a value > 1 before a task that you would like executed in parallel
-//  and then to set it to 0 after that task has completed.  A value of
-//  0 is a "special" value in set_n_threads() which translates to
-//  setting _n_threads to 1.
-//
-//  Some code uses _n_termination to decide if work should be done in
-//  parallel.  The notorious possibly_parallel_oops_do() in threads.cpp
-//  is an example of such code.  Look for variable "is_par" for other
-//  examples.
-//
-//  The active_workers is not reset to 0 after a parallel phase.  It's
-//  value may be used in later phases and in one instance at least
-//  (the parallel remark) it has to be used (the parallel remark depends
-//  on the partitioning done in the previous parallel scavenge).
-
 class SharedHeap : public CollectedHeap {
   friend class VMStructs;
 
@@ -90,48 +34,6 @@ protected:
   // Full initialization is done in a concrete subtype's "initialize"
   // function.
   SharedHeap();
-
-public:
-  // Note, the below comment needs to be updated to reflect the changes
-  // introduced by JDK-8076225. This should be done as part of JDK-8076289.
-  //
-  //Some collectors will perform "process_strong_roots" in parallel.
-  // Such a call will involve claiming some fine-grained tasks, such as
-  // scanning of threads.  To make this process simpler, we provide the
-  // "strong_roots_parity()" method.  Collectors that start parallel tasks
-  // whose threads invoke "process_strong_roots" must
-  // call "change_strong_roots_parity" in sequential code starting such a
-  // task.  (This also means that a parallel thread may only call
-  // process_strong_roots once.)
-  //
-  // For calls to process_roots by sequential code, the parity is
-  // updated automatically.
-  //
-  // The idea is that objects representing fine-grained tasks, such as
-  // threads, will contain a "parity" field.  A task will is claimed in the
-  // current "process_roots" call only if its parity field is the
-  // same as the "strong_roots_parity"; task claiming is accomplished by
-  // updating the parity field to the strong_roots_parity with a CAS.
-  //
-  // If the client meats this spec, then strong_roots_parity() will have
-  // the following properties:
-  //   a) to return a different value than was returned before the last
-  //      call to change_strong_roots_parity, and
-  //   c) to never return a distinguished value (zero) with which such
-  //      task-claiming variables may be initialized, to indicate "never
-  //      claimed".
- public:
-
-  // Call these in sequential code around process_roots.
-  // strong_roots_prologue calls change_strong_roots_parity, if
-  // parallel tasks are enabled.
-  class StrongRootsScope : public MarkingCodeBlobClosure::MarkScope {
-    SharedHeap*   _sh;
-
-   public:
-    StrongRootsScope(SharedHeap* heap, bool activate = true);
-    ~StrongRootsScope();
-  };
  };
 
 #endif // SHARE_VM_MEMORY_SHAREDHEAP_HPP

hotspot/src/share/vm/memory/strongRootsScope.cpp

@@ -0,0 +1,53 @@
+/*
+ * Copyright (c) 2015, 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
+ * under the terms of the GNU General Public License version 2 only, as
+ * published by the Free Software Foundation.
+ *
+ * This code is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
+ * version 2 for more details (a copy is included in the LICENSE file that
+ * accompanied this code).
+ *
+ * You should have received a copy of the GNU General Public License version
+ * 2 along with this work; if not, write to the Free Software Foundation,
+ * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
+ *
+ * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
+ * or visit www.oracle.com if you need additional information or have any
+ * questions.
+ *
+ */
+
+#include "precompiled.hpp"
+#include "classfile/stringTable.hpp"
+#include "code/nmethod.hpp"
+#include "memory/strongRootsScope.hpp"
+#include "runtime/thread.hpp"
+
+MarkScope::MarkScope(bool activate) : _active(activate) {
+  if (_active) {
+    nmethod::oops_do_marking_prologue();
+  }
+}
+
+MarkScope::~MarkScope() {
+  if (_active) {
+    nmethod::oops_do_marking_epilogue();
+  }
+}
+
+StrongRootsScope::StrongRootsScope(bool activate) : MarkScope(activate) {
+  if (_active) {
+    Threads::change_thread_claim_parity();
+    // Zero the claimed high water mark in the StringTable
+    StringTable::clear_parallel_claimed_index();
+  }
+}
+
+StrongRootsScope::~StrongRootsScope() {
+  Threads::assert_all_threads_claimed();
+}

hotspot/src/share/vm/memory/strongRootsScope.hpp

@@ -0,0 +1,46 @@
+/*
+ * Copyright (c) 2015, 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
+ * under the terms of the GNU General Public License version 2 only, as
+ * published by the Free Software Foundation.
+ *
+ * This code is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
+ * version 2 for more details (a copy is included in the LICENSE file that
+ * accompanied this code).
+ *
+ * You should have received a copy of the GNU General Public License version
+ * 2 along with this work; if not, write to the Free Software Foundation,
+ * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
+ *
+ * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
+ * or visit www.oracle.com if you need additional information or have any
+ * questions.
+ *
+ */
+
+#ifndef SHARE_VM_MEMORY_STRONGROOTSSCOPE_HPP
+#define SHARE_VM_MEMORY_STRONGROOTSSCOPE_HPP
+
+#include "memory/allocation.hpp"
+
+class MarkScope : public StackObj {
+ protected:
+  bool _active;
+ public:
+  MarkScope(bool activate = true);
+  ~MarkScope();
+};
+
+// Sets up and tears down the required state for parallel root processing.
+
+class StrongRootsScope : public MarkScope {
+ public:
+  StrongRootsScope(bool activate = true);
+  ~StrongRootsScope();
+};
+
+#endif // SHARE_VM_MEMORY_STRONGROOTSSCOPE_HPP

hotspot/src/share/vm/runtime/thread.hpp

@@ -1886,10 +1886,23 @@ class Threads: AllStatic {
   // Does not include JNI_VERSION_1_1
   static jboolean is_supported_jni_version(jint version);
 
+  // The "thread claim parity" provides a way for threads to be claimed
+  // by parallel worker tasks.
+  //
+  // Each thread contains a a "parity" field. A task will claim the
+  // thread only if its parity field is the same as the global parity,
+  // which is updated by calling change_thread_claim_parity().
+  //
+  // For this to work change_thread_claim_parity() needs to be called
+  // exactly once in sequential code before starting parallel tasks
+  // that should claim threads.
+  //
+  // New threads get their parity set to 0 and change_thread_claim_parity()
+  // never set the global parity to 0.
   static int thread_claim_parity() { return _thread_claim_parity; }
   static void change_thread_claim_parity();
-
   static void assert_all_threads_claimed() PRODUCT_RETURN;
+
   // Apply "f->do_oop" to all root oops in all threads.
   // This version may only be called by sequential code.
   static void oops_do(OopClosure* f, CLDClosure* cld_f, CodeBlobClosure* cf);