jdk8-mips64-public/hotspot: src/share/vm/oops/cpCacheOop.cpp@1d7922586cf6 (annotated)

src/share/vm/oops/cpCacheOop.cpp@1d7922586cf6 (annotated)

src/share/vm/oops/cpCacheOop.cpp

Tue, 24 Jul 2012 10:51:00 -0700

author: twisti
date: Tue, 24 Jul 2012 10:51:00 -0700
changeset 3969: 1d7922586cf6
parent 3601: f096e1b74d85
permissions: -rw-r--r--

7023639: JSR 292 method handle invocation needs a fast path for compiled code
6984705: JSR 292 method handle creation should not go through JNI
Summary: remove assembly code for JDK 7 chained method handles
Reviewed-by: jrose, twisti, kvn, mhaupt
Contributed-by: John Rose <john.r.rose@oracle.com>, Christian Thalinger <christian.thalinger@oracle.com>, Michael Haupt <michael.haupt@oracle.com>

 /*
  * Copyright (c) 1998, 2012, 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 "gc_implementation/shared/markSweep.inline.hpp"
 #include "interpreter/interpreter.hpp"
 #include "interpreter/rewriter.hpp"
 #include "memory/universe.inline.hpp"
 #include "oops/cpCacheOop.hpp"
 #include "oops/objArrayOop.hpp"
 #include "oops/oop.inline.hpp"
 #include "prims/jvmtiRedefineClassesTrace.hpp"
 #include "prims/methodHandles.hpp"
 #include "runtime/handles.inline.hpp"
 // Implememtation of ConstantPoolCacheEntry
 void ConstantPoolCacheEntry::initialize_entry(int index) {
   assert(0 < index && index < 0x10000, "sanity check");
   _indices = index;
   assert(constant_pool_index() == index, "");
 }
 void ConstantPoolCacheEntry::initialize_secondary_entry(int main_index) {
   assert(0 <= main_index && main_index < 0x10000, "sanity check");
   _indices = (main_index << main_cp_index_bits);
   assert(main_entry_index() == main_index, "");
 }
 int ConstantPoolCacheEntry::make_flags(TosState state,
                                        int option_bits,
                                        int field_index_or_method_params) {
   assert(state < number_of_states, "Invalid state in make_flags");
   int f = ((int)state << tos_state_shift) | option_bits | field_index_or_method_params;
   // Preserve existing flag bit values
   // The low bits are a field offset, or else the method parameter size.
 #ifdef ASSERT
   TosState old_state = flag_state();
   assert(old_state == (TosState)0 || old_state == state,
          "inconsistent cpCache flags state");
 #endif
   return (_flags | f) ;
 }
 void ConstantPoolCacheEntry::set_bytecode_1(Bytecodes::Code code) {
   assert(!is_secondary_entry(), "must not overwrite main_entry_index");
 #ifdef ASSERT
   // Read once.
   volatile Bytecodes::Code c = bytecode_1();
   assert(c == 0 || c == code || code == 0, "update must be consistent");
 #endif
   // Need to flush pending stores here before bytecode is written.
   OrderAccess::release_store_ptr(&_indices, _indices | ((u_char)code << bytecode_1_shift));
 }
 void ConstantPoolCacheEntry::set_bytecode_2(Bytecodes::Code code) {
   assert(!is_secondary_entry(), "must not overwrite main_entry_index");
 #ifdef ASSERT
   // Read once.
   volatile Bytecodes::Code c = bytecode_2();
   assert(c == 0 || c == code || code == 0, "update must be consistent");
 #endif
   // Need to flush pending stores here before bytecode is written.
   OrderAccess::release_store_ptr(&_indices, _indices | ((u_char)code << bytecode_2_shift));
 }
 // Sets f1, ordering with previous writes.
 void ConstantPoolCacheEntry::release_set_f1(oop f1) {
   // Use barriers as in oop_store
   assert(f1 != NULL, "");
   oop* f1_addr = (oop*) &_f1;
   update_barrier_set_pre(f1_addr, f1);
   OrderAccess::release_store_ptr((intptr_t*)f1_addr, f1);
   update_barrier_set((void*) f1_addr, f1);
 }
 // Sets flags, but only if the value was previously zero.
 bool ConstantPoolCacheEntry::init_flags_atomic(intptr_t flags) {
   intptr_t result = Atomic::cmpxchg_ptr(flags, &_flags, 0);
   return (result == 0);
 }
 #ifdef ASSERT
 // It is possible to have two different dummy methodOops created
 // when the resolve code for invoke interface executes concurrently
 // Hence the assertion below is weakened a bit for the invokeinterface
 // case.
 bool ConstantPoolCacheEntry::same_methodOop(oop cur_f1, oop f1) {
   return (cur_f1 == f1 || ((methodOop)cur_f1)->name() ==
          ((methodOop)f1)->name() || ((methodOop)cur_f1)->signature() ==
          ((methodOop)f1)->signature());
 }
 #endif
 // Note that concurrent update of both bytecodes can leave one of them
 // reset to zero.  This is harmless; the interpreter will simply re-resolve
 // the damaged entry.  More seriously, the memory synchronization is needed
 // to flush other fields (f1, f2) completely to memory before the bytecodes
 // are updated, lest other processors see a non-zero bytecode but zero f1/f2.
 void ConstantPoolCacheEntry::set_field(Bytecodes::Code get_code,
                                        Bytecodes::Code put_code,
                                        KlassHandle field_holder,
                                        int field_index,
                                        int field_offset,
                                        TosState field_type,
                                        bool is_final,
                                        bool is_volatile) {
   set_f1(field_holder()->java_mirror());
   set_f2(field_offset);
   assert((field_index & field_index_mask) == field_index,
          "field index does not fit in low flag bits");
   set_field_flags(field_type,
                   ((is_volatile ? 1 : 0) << is_volatile_shift) |
                   ((is_final    ? 1 : 0) << is_final_shift),
                   field_index);
   set_bytecode_1(get_code);
   set_bytecode_2(put_code);
   NOT_PRODUCT(verify(tty));
 }
 void ConstantPoolCacheEntry::set_parameter_size(int value) {
   // This routine is called only in corner cases where the CPCE is not yet initialized.
   // See AbstractInterpreter::deopt_continue_after_entry.
   assert(_flags == 0 || parameter_size() == 0 || parameter_size() == value,
          err_msg("size must not change: parameter_size=%d, value=%d", parameter_size(), value));
   // Setting the parameter size by itself is only safe if the
   // current value of _flags is 0, otherwise another thread may have
   // updated it and we don't want to overwrite that value.  Don't
   // bother trying to update it once it's nonzero but always make
   // sure that the final parameter size agrees with what was passed.
   if (_flags == 0) {
     Atomic::cmpxchg_ptr((value & parameter_size_mask), &_flags, 0);
   }
   guarantee(parameter_size() == value,
             err_msg("size must not change: parameter_size=%d, value=%d", parameter_size(), value));
 }
 void ConstantPoolCacheEntry::set_method(Bytecodes::Code invoke_code,
                                         methodHandle method,
                                         int vtable_index) {
   assert(!is_secondary_entry(), "");
   assert(method->interpreter_entry() != NULL, "should have been set at this point");
   assert(!method->is_obsolete(),  "attempt to write obsolete method to cpCache");
   int byte_no = -1;
   bool change_to_virtual = false;
   switch (invoke_code) {
     case Bytecodes::_invokeinterface:
       // We get here from InterpreterRuntime::resolve_invoke when an invokeinterface
       // instruction somehow links to a non-interface method (in Object).
       // In that case, the method has no itable index and must be invoked as a virtual.
       // Set a flag to keep track of this corner case.
       change_to_virtual = true;
       // ...and fall through as if we were handling invokevirtual:
     case Bytecodes::_invokevirtual:
       {
         if (method->can_be_statically_bound()) {
           // set_f2_as_vfinal_method checks if is_vfinal flag is true.
           set_method_flags(as_TosState(method->result_type()),
                            (                             1      << is_vfinal_shift) |
                            ((method->is_final_method() ? 1 : 0) << is_final_shift)  |
                            ((change_to_virtual         ? 1 : 0) << is_forced_virtual_shift),
                            method()->size_of_parameters());
           set_f2_as_vfinal_method(method());
         } else {
           assert(vtable_index >= 0, "valid index");
           assert(!method->is_final_method(), "sanity");
           set_method_flags(as_TosState(method->result_type()),
                            ((change_to_virtual ? 1 : 0) << is_forced_virtual_shift),
                            method()->size_of_parameters());
           set_f2(vtable_index);
         }
         byte_no = 2;
         break;
       }
     case Bytecodes::_invokespecial:
     case Bytecodes::_invokestatic:
       // Note:  Read and preserve the value of the is_vfinal flag on any
       // invokevirtual bytecode shared with this constant pool cache entry.
       // It is cheap and safe to consult is_vfinal() at all times.
       // Once is_vfinal is set, it must stay that way, lest we get a dangling oop.
       set_method_flags(as_TosState(method->result_type()),
                        ((is_vfinal()               ? 1 : 0) << is_vfinal_shift) |
                        ((method->is_final_method() ? 1 : 0) << is_final_shift),
                        method()->size_of_parameters());
       set_f1(method());
       byte_no = 1;
       break;
     default:
       ShouldNotReachHere();
       break;
   }
   // Note:  byte_no also appears in TemplateTable::resolve.
   if (byte_no == 1) {
     assert(invoke_code != Bytecodes::_invokevirtual &&
            invoke_code != Bytecodes::_invokeinterface, "");
     set_bytecode_1(invoke_code);
   } else if (byte_no == 2)  {
     if (change_to_virtual) {
       assert(invoke_code == Bytecodes::_invokeinterface, "");
       // NOTE: THIS IS A HACK - BE VERY CAREFUL!!!
       //
       // Workaround for the case where we encounter an invokeinterface, but we
       // should really have an _invokevirtual since the resolved method is a
       // virtual method in java.lang.Object. This is a corner case in the spec
       // but is presumably legal. javac does not generate this code.
       //
       // We set bytecode_1() to _invokeinterface, because that is the
       // bytecode # used by the interpreter to see if it is resolved.
       // We set bytecode_2() to _invokevirtual.
       // See also interpreterRuntime.cpp. (8/25/2000)
       // Only set resolved for the invokeinterface case if method is public.
       // Otherwise, the method needs to be reresolved with caller for each
       // interface call.
       if (method->is_public()) set_bytecode_1(invoke_code);
     } else {
       assert(invoke_code == Bytecodes::_invokevirtual, "");
     }
     // set up for invokevirtual, even if linking for invokeinterface also:
     set_bytecode_2(Bytecodes::_invokevirtual);
   } else {
     ShouldNotReachHere();
   }
   NOT_PRODUCT(verify(tty));
 }
 void ConstantPoolCacheEntry::set_interface_call(methodHandle method, int index) {
   assert(!is_secondary_entry(), "");
   klassOop interf = method->method_holder();
   assert(instanceKlass::cast(interf)->is_interface(), "must be an interface");
   assert(!method->is_final_method(), "interfaces do not have final methods; cannot link to one here");
   set_f1(interf);
   set_f2(index);
   set_method_flags(as_TosState(method->result_type()),
 ,  // no option bits
                    method()->size_of_parameters());
   set_bytecode_1(Bytecodes::_invokeinterface);
 }
 void ConstantPoolCacheEntry::set_method_handle(methodHandle adapter, Handle appendix) {
   assert(!is_secondary_entry(), "");
   set_method_handle_common(Bytecodes::_invokehandle, adapter, appendix);
 }
 void ConstantPoolCacheEntry::set_dynamic_call(methodHandle adapter, Handle appendix) {
   assert(is_secondary_entry(), "");
   set_method_handle_common(Bytecodes::_invokedynamic, adapter, appendix);
 }
 void ConstantPoolCacheEntry::set_method_handle_common(Bytecodes::Code invoke_code, methodHandle adapter, Handle appendix) {
   // NOTE: This CPCE can be the subject of data races.
   // There are three words to update: flags, f2, f1 (in that order).
   // Writers must store all other values before f1.
   // Readers must test f1 first for non-null before reading other fields.
   // Competing writers must acquire exclusive access on the first
   // write, to flags, using a compare/exchange.
   // A losing writer must spin until the winner writes f1,
   // so that when he returns, he can use the linked cache entry.
   bool has_appendix = appendix.not_null();
   if (!has_appendix) {
     // The extra argument is not used, but we need a non-null value to signify linkage state.
     // Set it to something benign that will never leak memory.
     appendix = Universe::void_mirror();
   }
   bool owner =
     init_method_flags_atomic(as_TosState(adapter->result_type()),
                    ((has_appendix ?  1 : 0) << has_appendix_shift) |
                    (                 1      << is_vfinal_shift)    |
                    (                 1      << is_final_shift),
                    adapter->size_of_parameters());
   if (!owner) {
     while (is_f1_null()) {
       // Pause momentarily on a low-level lock, to allow racing thread to win.
       MutexLockerEx mu(Patching_lock, Mutex::_no_safepoint_check_flag);
       os::yield();
     }
     return;
   }
   if (TraceInvokeDynamic) {
     tty->print_cr("set_method_handle bc=%d appendix="PTR_FORMAT"%s method="PTR_FORMAT" ",
                   invoke_code,
                   (intptr_t)appendix(), (has_appendix ? "" : " (unused)"),
                   (intptr_t)adapter());
     adapter->print();
     if (has_appendix)  appendix()->print();
   }
   // Method handle invokes and invokedynamic sites use both cp cache words.
   // f1, if not null, contains a value passed as a trailing argument to the adapter.
   // In the general case, this could be the call site's MethodType,
   // for use with java.lang.Invokers.checkExactType, or else a CallSite object.
   // f2 contains the adapter method which manages the actual call.
   // In the general case, this is a compiled LambdaForm.
   // (The Java code is free to optimize these calls by binding other
   // sorts of methods and appendices to call sites.)
   // JVM-level linking is via f2, as if for invokevfinal, and signatures are erased.
   // The appendix argument (if any) is added to the signature, and is counted in the parameter_size bits.
   // In principle this means that the method (with appendix) could take up to 256 parameter slots.
   //
   // This means that given a call site like (List)mh.invoke("foo"),
   // the f2 method has signature '(Ljl/Object;Ljl/invoke/MethodType;)Ljl/Object;',
   // not '(Ljava/lang/String;)Ljava/util/List;'.
   // The fact that String and List are involved is encoded in the MethodType in f1.
   // This allows us to create fewer method oops, while keeping type safety.
   //
   set_f2_as_vfinal_method(adapter());
   assert(appendix.not_null(), "needed for linkage state");
   release_set_f1(appendix());  // This must be the last one to set (see NOTE above)!
   if (!is_secondary_entry()) {
     // The interpreter assembly code does not check byte_2,
     // but it is used by is_resolved, method_if_resolved, etc.
     set_bytecode_2(invoke_code);
   }
   NOT_PRODUCT(verify(tty));
   if (TraceInvokeDynamic) {
     this->print(tty, 0);
   }
 }
 methodOop ConstantPoolCacheEntry::method_if_resolved(constantPoolHandle cpool) {
   if (is_secondary_entry()) {
     if (!is_f1_null())
       return f2_as_vfinal_method();
     return NULL;
   }
   // Decode the action of set_method and set_interface_call
   Bytecodes::Code invoke_code = bytecode_1();
   if (invoke_code != (Bytecodes::Code)0) {
     oop f1 = _f1;
     if (f1 != NULL) {
       switch (invoke_code) {
       case Bytecodes::_invokeinterface:
         assert(f1->is_klass(), "");
         return klassItable::method_for_itable_index(klassOop(f1), f2_as_index());
       case Bytecodes::_invokestatic:
       case Bytecodes::_invokespecial:
         assert(!has_appendix(), "");
         assert(f1->is_method(), "");
         return methodOop(f1);
       }
     }
   }
   invoke_code = bytecode_2();
   if (invoke_code != (Bytecodes::Code)0) {
     switch (invoke_code) {
     case Bytecodes::_invokevirtual:
       if (is_vfinal()) {
         // invokevirtual
         methodOop m = f2_as_vfinal_method();
         assert(m->is_method(), "");
         return m;
       } else {
         int holder_index = cpool->uncached_klass_ref_index_at(constant_pool_index());
         if (cpool->tag_at(holder_index).is_klass()) {
           klassOop klass = cpool->resolved_klass_at(holder_index);
           if (!Klass::cast(klass)->oop_is_instance())
             klass = SystemDictionary::Object_klass();
           return instanceKlass::cast(klass)->method_at_vtable(f2_as_index());
         }
       }
       break;
     case Bytecodes::_invokehandle:
     case Bytecodes::_invokedynamic:
       return f2_as_vfinal_method();
     }
   }
   return NULL;
 }
 class LocalOopClosure: public OopClosure {
  private:
   void (*_f)(oop*);
  public:
   LocalOopClosure(void f(oop*))        { _f = f; }
   virtual void do_oop(oop* o)          { _f(o); }
   virtual void do_oop(narrowOop *o)    { ShouldNotReachHere(); }
 };
 void ConstantPoolCacheEntry::oops_do(void f(oop*)) {
   LocalOopClosure blk(f);
   oop_iterate(&blk);
 }
 void ConstantPoolCacheEntry::oop_iterate(OopClosure* blk) {
   assert(in_words(size()) == 4, "check code below - may need adjustment");
   // field[1] is always oop or NULL
   blk->do_oop((oop*)&_f1);
   if (is_vfinal()) {
     blk->do_oop((oop*)&_f2);
   }
 }
 void ConstantPoolCacheEntry::oop_iterate_m(OopClosure* blk, MemRegion mr) {
   assert(in_words(size()) == 4, "check code below - may need adjustment");
   // field[1] is always oop or NULL
   if (mr.contains((oop *)&_f1)) blk->do_oop((oop*)&_f1);
   if (is_vfinal()) {
     if (mr.contains((oop *)&_f2)) blk->do_oop((oop*)&_f2);
   }
 }
 void ConstantPoolCacheEntry::follow_contents() {
   assert(in_words(size()) == 4, "check code below - may need adjustment");
   // field[1] is always oop or NULL
   MarkSweep::mark_and_push((oop*)&_f1);
   if (is_vfinal()) {
     MarkSweep::mark_and_push((oop*)&_f2);
   }
 }
 #ifndef SERIALGC
 void ConstantPoolCacheEntry::follow_contents(ParCompactionManager* cm) {
   assert(in_words(size()) == 4, "check code below - may need adjustment");
   // field[1] is always oop or NULL
   PSParallelCompact::mark_and_push(cm, (oop*)&_f1);
   if (is_vfinal()) {
     PSParallelCompact::mark_and_push(cm, (oop*)&_f2);
   }
 }
 #endif // SERIALGC
 void ConstantPoolCacheEntry::adjust_pointers() {
   assert(in_words(size()) == 4, "check code below - may need adjustment");
   // field[1] is always oop or NULL
   MarkSweep::adjust_pointer((oop*)&_f1);
   if (is_vfinal()) {
     MarkSweep::adjust_pointer((oop*)&_f2);
   }
 }
 #ifndef SERIALGC
 void ConstantPoolCacheEntry::update_pointers() {
   assert(in_words(size()) == 4, "check code below - may need adjustment");
   // field[1] is always oop or NULL
   PSParallelCompact::adjust_pointer((oop*)&_f1);
   if (is_vfinal()) {
     PSParallelCompact::adjust_pointer((oop*)&_f2);
   }
 }
 #endif // SERIALGC
 // RedefineClasses() API support:
 // If this constantPoolCacheEntry refers to old_method then update it
 // to refer to new_method.
 bool ConstantPoolCacheEntry::adjust_method_entry(methodOop old_method,
        methodOop new_method, bool * trace_name_printed) {
   if (is_vfinal()) {
     // virtual and final so _f2 contains method ptr instead of vtable index
     if (f2_as_vfinal_method() == old_method) {
       // match old_method so need an update
       // NOTE: can't use set_f2_as_vfinal_method as it asserts on different values
       _f2 = (intptr_t)new_method;
       if (RC_TRACE_IN_RANGE(0x00100000, 0x00400000)) {
         if (!(*trace_name_printed)) {
           // RC_TRACE_MESG macro has an embedded ResourceMark
           RC_TRACE_MESG(("adjust: name=%s",
             Klass::cast(old_method->method_holder())->external_name()));
           *trace_name_printed = true;
         }
         // RC_TRACE macro has an embedded ResourceMark
         RC_TRACE(0x00400000, ("cpc vf-entry update: %s(%s)",
           new_method->name()->as_C_string(),
           new_method->signature()->as_C_string()));
       }
       return true;
     }
     // f1() is not used with virtual entries so bail out
     return false;
   }
   if ((oop)_f1 == NULL) {
     // NULL f1() means this is a virtual entry so bail out
     // We are assuming that the vtable index does not need change.
     return false;
   }
   if ((oop)_f1 == old_method) {
     _f1 = new_method;
     if (RC_TRACE_IN_RANGE(0x00100000, 0x00400000)) {
       if (!(*trace_name_printed)) {
         // RC_TRACE_MESG macro has an embedded ResourceMark
         RC_TRACE_MESG(("adjust: name=%s",
           Klass::cast(old_method->method_holder())->external_name()));
         *trace_name_printed = true;
       }
       // RC_TRACE macro has an embedded ResourceMark
       RC_TRACE(0x00400000, ("cpc entry update: %s(%s)",
         new_method->name()->as_C_string(),
         new_method->signature()->as_C_string()));
     }
     return true;
   }
   return false;
 }
 bool ConstantPoolCacheEntry::is_interesting_method_entry(klassOop k) {
   if (!is_method_entry()) {
     // not a method entry so not interesting by default
     return false;
   }
   methodOop m = NULL;
   if (is_vfinal()) {
     // virtual and final so _f2 contains method ptr instead of vtable index
     m = f2_as_vfinal_method();
   } else if (is_f1_null()) {
     // NULL _f1 means this is a virtual entry so also not interesting
     return false;
   } else {
     oop f1 = _f1;  // _f1 is volatile
     if (!f1->is_method()) {
       // _f1 can also contain a klassOop for an interface
       return false;
     }
     m = f1_as_method();
   }
   assert(m != NULL && m->is_method(), "sanity check");
   if (m == NULL || !m->is_method() || m->method_holder() != k) {
     // robustness for above sanity checks or method is not in
     // the interesting class
     return false;
   }
   // the method is in the interesting class so the entry is interesting
   return true;
 }
 void ConstantPoolCacheEntry::print(outputStream* st, int index) const {
   // print separator
   if (index == 0) st->print_cr("                 -------------");
   // print entry
   st->print("%3d  ("PTR_FORMAT")  ", index, (intptr_t)this);
   if (is_secondary_entry())
     st->print_cr("[%5d|secondary]", main_entry_index());
   else
     st->print_cr("[%02x|%02x|%5d]", bytecode_2(), bytecode_1(), constant_pool_index());
   st->print_cr("                 [   "PTR_FORMAT"]", (intptr_t)(oop)_f1);
   st->print_cr("                 [   "PTR_FORMAT"]", (intptr_t)_f2);
   st->print_cr("                 [   "PTR_FORMAT"]", (intptr_t)_flags);
   st->print_cr("                 -------------");
 }
 void ConstantPoolCacheEntry::verify(outputStream* st) const {
   // not implemented yet
 }
 // Implementation of ConstantPoolCache
 void constantPoolCacheOopDesc::initialize(intArray& inverse_index_map) {
   assert(inverse_index_map.length() == length(), "inverse index map must have same length as cache");
   for (int i = 0; i < length(); i++) {
     ConstantPoolCacheEntry* e = entry_at(i);
     int original_index = inverse_index_map[i];
     if ((original_index & Rewriter::_secondary_entry_tag) != 0) {
       int main_index = (original_index - Rewriter::_secondary_entry_tag);
       assert(!entry_at(main_index)->is_secondary_entry(), "valid main index");
       e->initialize_secondary_entry(main_index);
     } else {
       e->initialize_entry(original_index);
     }
     assert(entry_at(i) == e, "sanity");
   }
 }
 // RedefineClasses() API support:
 // If any entry of this constantPoolCache points to any of
 // old_methods, replace it with the corresponding new_method.
 void constantPoolCacheOopDesc::adjust_method_entries(methodOop* old_methods, methodOop* new_methods,
                                                      int methods_length, bool * trace_name_printed) {
   if (methods_length == 0) {
     // nothing to do if there are no methods
     return;
   }
   // get shorthand for the interesting class
   klassOop old_holder = old_methods[0]->method_holder();
   for (int i = 0; i < length(); i++) {
     if (!entry_at(i)->is_interesting_method_entry(old_holder)) {
       // skip uninteresting methods
       continue;
     }
     // The constantPoolCache contains entries for several different
     // things, but we only care about methods. In fact, we only care
     // about methods in the same class as the one that contains the
     // old_methods. At this point, we have an interesting entry.
     for (int j = 0; j < methods_length; j++) {
       methodOop old_method = old_methods[j];
       methodOop new_method = new_methods[j];
       if (entry_at(i)->adjust_method_entry(old_method, new_method,
           trace_name_printed)) {
         // current old_method matched this entry and we updated it so
         // break out and get to the next interesting entry if there one
         break;
       }
     }
   }
 }

Mercurial > jdk8-mips64-public > hotspot / annotate

src/share/vm/oops/cpCacheOop.cpp@1d7922586cf6 (annotated)

src/share/vm/oops/cpCacheOop.cpp