[V8] Implement SIMD Boolean vector types to allow mask registers.

src/compiler/js-native-context-specialization.cc

 // Copyright 2015 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/js-native-context-specialization.h"
 
 #include "src/accessors.h"
 #include "src/code-factory.h"
 #include "src/compilation-dependencies.h"
 #include "src/compiler/access-builder.h"
@@ skipping 1507 matching lines @@
         case MachineRepresentation::kTagged:
           break;
         case MachineRepresentation::kNone:
         case MachineRepresentation::kBit:
         case MachineRepresentation::kWord8:
         case MachineRepresentation::kWord16:
         case MachineRepresentation::kWord32:
         case MachineRepresentation::kWord64:
         case MachineRepresentation::kFloat32:
         case MachineRepresentation::kSimd128:
+        case MachineRepresentation::kSimdBool4:
+        case MachineRepresentation::kSimdBool8:
+        case MachineRepresentation::kSimdBool16:
           UNREACHABLE();
           break;
       }
       Handle<Map> transition_map;
       if (access_info.transition_map().ToHandle(&transition_map)) {
         effect = graph()->NewNode(
             common()->BeginRegion(RegionObservability::kObservable), effect);
         effect = graph()->NewNode(
             simplified()->StoreField(AccessBuilder::ForMap()), receiver,
             jsgraph()->Constant(transition_map), effect, control);
@@ skipping 709 matching lines @@
   return jsgraph()->javascript();
 }
 
 SimplifiedOperatorBuilder* JSNativeContextSpecialization::simplified() const {
   return jsgraph()->simplified();
 }
 
 }  // namespace compiler
 }  // namespace internal
 }  // namespace v8