[turbofan] Added the RoundInt64ToFloat64 instruction to TurboFan.

src/compiler/arm64/code-generator-arm64.cc

 // Copyright 2014 the V8 project authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file.
 
 #include "src/compiler/code-generator.h"
 
 #include "src/arm64/frames-arm64.h"
 #include "src/arm64/macro-assembler-arm64.h"
 #include "src/compiler/code-generator-impl.h"
 #include "src/compiler/gap-resolver.h"
@@ skipping 894 matching lines @@
       break;
     case kArm64Float64ToInt32:
       __ Fcvtzs(i.OutputRegister32(), i.InputDoubleRegister(0));
       break;
     case kArm64Float64ToUint32:
       __ Fcvtzu(i.OutputRegister32(), i.InputDoubleRegister(0));
       break;
     case kArm64Int32ToFloat64:
       __ Scvtf(i.OutputDoubleRegister(), i.InputRegister32(0));
       break;
+    case kArm64Int64ToFloat64:
+      __ Scvtf(i.OutputDoubleRegister(), i.InputRegister64(0));
+      break;
     case kArm64Uint32ToFloat64:
       __ Ucvtf(i.OutputDoubleRegister(), i.InputRegister32(0));
       break;
     case kArm64Float64ExtractLowWord32:
       __ Fmov(i.OutputRegister32(), i.InputFloat32Register(0));
       break;
     case kArm64Float64ExtractHighWord32:
       // TODO(arm64): This should use MOV (to general) when NEON is supported.
       __ Fmov(i.OutputRegister(), i.InputFloat64Register(0));
       __ Lsr(i.OutputRegister(), i.OutputRegister(), 32);
@@ skipping 532 matching lines @@
       padding_size -= kInstructionSize;
     }
   }
 }
 
 #undef __
 
 }  // namespace compiler
 }  // namespace internal
 }  // namespace v8