Move index closer to start of Node{3,16}

This should slightly improve cache hit rate

ConflictSet.cpp

@@ -338,11 +338,12 @@ struct Node3 : Node {
   constexpr static auto kMaxNodes = 3;
   constexpr static auto kType = Type_Node3;
 
-  TaggedNodePointer children[kMaxNodes];
-  InternalVersionT childMaxVersion[kMaxNodes];
   // Sorted
   uint8_t index[kMaxNodes];
 
+  TaggedNodePointer children[kMaxNodes];
+  InternalVersionT childMaxVersion[kMaxNodes];
+
   uint8_t *partialKey() {
     assert(!releaseDeferred);
     return (uint8_t *)(this + 1);
@@ -357,11 +358,12 @@ struct Node16 : Node {
   constexpr static auto kType = Type_Node16;
   constexpr static auto kMaxNodes = 16;
 
-  TaggedNodePointer children[kMaxNodes];
-  InternalVersionT childMaxVersion[kMaxNodes];
   // Sorted
   uint8_t index[kMaxNodes];
 
+  TaggedNodePointer children[kMaxNodes];
+  InternalVersionT childMaxVersion[kMaxNodes];
+
   uint8_t *partialKey() {
     assert(!releaseDeferred);
     return (uint8_t *)(this + 1);
@@ -440,7 +442,10 @@ inline void Node3::copyChildrenAndKeyFrom(const Node0 &other) {
 inline void Node3::copyChildrenAndKeyFrom(const Node3 &other) {
   memcpy((char *)this + kNodeCopyBegin, (char *)&other + kNodeCopyBegin,
          kNodeCopySize);
-  memcpy(children, other.children, sizeof(*this) - sizeof(Node));
+  memcpy(index, other.index, kMaxNodes);
+  memcpy(children, other.children, kMaxNodes * sizeof(children[0])); // NOLINT
+  memcpy(childMaxVersion, other.childMaxVersion,
+         kMaxNodes * sizeof(childMaxVersion[0]));
   memcpy(partialKey(), &other + 1, partialKeyLen);
   for (int i = 0; i < numChildren; ++i) {
     assert(children[i]->parent == &other);