Fix optimized code caching in FastNewClosureStub.

src/arm/code-stubs-arm.cc

 // Copyright 2012 the V8 project authors. All rights reserved.
 // Redistribution and use in source and binary forms, with or without
 // modification, are permitted provided that the following conditions are
 // met:
 //
 //     * Redistributions of source code must retain the above copyright
 //       notice, this list of conditions and the following disclaimer.
 //     * Redistributions in binary form must reproduce the above
 //       copyright notice, this list of conditions and the following
 //       disclaimer in the documentation and/or other materials provided
@@ skipping 140 matching lines @@
 
   // r2 holds global context, r1 points to fixed array of 3-element entries
   // (global context, optimized code, literals).
   // The optimized code map must never be empty, so check the first elements.
   Label install_optimized;
   // Speculatively move code object into r4.
   __ ldr(r4, FieldMemOperand(r1, FixedArray::kHeaderSize + kPointerSize));
   __ ldr(r5, FieldMemOperand(r1, FixedArray::kHeaderSize));
   __ cmp(r2, r5);
   __ b(eq, &install_optimized);
-  __ b(&install_unoptimized);
 
   // Iterate through the rest of map backwards.  r4 holds an index as a Smi.
   Label loop;
   __ ldr(r4, FieldMemOperand(r1, FixedArray::kLengthOffset));
   __ bind(&loop);
   // Do not double check first entry.
 
   __ cmp(r4, Operand(Smi::FromInt(SharedFunctionInfo::kEntryLength)));
   __ b(eq, &install_unoptimized);
   __ sub(r4, r4, Operand(
@@ skipping 7340 matching lines @@
   __ StoreNumberToDoubleElements(r0, r3, r1, r5, r6, r7, r9, r2,
                                  &slow_elements);
   __ Ret();
 }
 
 #undef __
 
 } }  // namespace v8::internal
 
 #endif  // V8_TARGET_ARCH_ARM