[libSyntax] Optimise layout of RawSyntax to be more space efficient

This decreases the size of RawSyntax nodes from 88 to 64 bytes by
- Avoiding some padding by moving RefCount further up
- Limiting the length of tokens and their trivia to 32 bits. We would
  hit this limit with files >4GB but we also hit this limit at other
  places like the TextLength property in the Common bits.

Author: ahoppen

include/swift/Syntax/RawSyntax.h

@@ -160,15 +160,17 @@ class RawSyntax final
   /// An ID of this node that is stable across incremental parses
   SyntaxNodeId NodeId;
 
+  mutable std::atomic<int> RefCount;
+
   /// If this node was allocated using a \c SyntaxArena's bump allocator, a
   /// reference to the arena to keep the underlying memory buffer of this node
   /// alive. If this is a \c nullptr, the node owns its own memory buffer.
   RC<SyntaxArena> Arena;
 
   union {
     struct {
-      // FIXME: Reduce TextLength to 30 bits so that common fits in 4 bytes?
-      /// Number of bytes this node takes up spelled out in the source code
+      /// Number of bytes this node takes up spelled out in the source code.
+      /// Always 0 if the node is missing.
       unsigned TextLength : 32;
       /// Whether this piece of syntax was actually present in the source.
       unsigned Presence : 1;
@@ -196,10 +198,13 @@ class RawSyntax final
                     "Only 64 bits reserved for standard syntax bits");
       uint64_t : bitmax(NumRawSyntaxBits, 64); // align to 16 bits
       /// The kind of token this "token" node represents.
+      const char *LeadingTrivia;
+      const char *TokenText;
+      const char *TrailingTrivia;
+      unsigned LeadingTriviaLength : 32;
+      unsigned TokenLength : 32;
+      unsigned TrailingTriviaLength : 32;
       unsigned TokenKind : 16;
-      StringRef LeadingTrivia;
-      StringRef TokenText;
-      StringRef TrailingTrivia;
     } Token;
   } Bits;
 
@@ -238,8 +243,6 @@ class RawSyntax final
     return TextLength;
   }
 
-  mutable std::atomic<int> RefCount;
-
 public:
   ~RawSyntax();
 
@@ -392,19 +395,20 @@ class RawSyntax final
   /// disappear when the syntax node gets freed.
   StringRef getTokenText() const {
     assert(isToken());
-    return Bits.Token.TokenText;
+    return StringRef(Bits.Token.TokenText, Bits.Token.TokenLength);
   }
 
   /// Return the unparsed leading trivia of the token.
   StringRef getLeadingTrivia() const {
     assert(isToken());
-    return Bits.Token.LeadingTrivia;
+    return StringRef(Bits.Token.LeadingTrivia, Bits.Token.LeadingTriviaLength);
   }
 
   /// Return the unparsed trailing trivia of the token.
   StringRef getTrailingTrivia() const {
     assert(isToken());
-    return Bits.Token.TrailingTrivia;
+    return StringRef(Bits.Token.TrailingTrivia,
+                     Bits.Token.TrailingTriviaLength);
   }
 
   /// Return pieces that make up the leading trivia of the token.

lib/Syntax/RawSyntax.cpp

@@ -111,8 +111,8 @@ RawSyntax::RawSyntax(SyntaxKind Kind, ArrayRef<RC<RawSyntax>> Layout,
                      size_t TextLength, SourcePresence Presence,
                      const RC<SyntaxArena> &Arena,
                      llvm::Optional<unsigned> NodeId)
-    : Arena(Arena), Bits({{unsigned(TextLength), unsigned(Presence), false}}),
-      RefCount(0) {
+    : RefCount(0), Arena(Arena),
+      Bits({{unsigned(TextLength), unsigned(Presence), false}}) {
   assert(Arena && "RawSyntax nodes must always be allocated in an arena");
   assert(Kind != SyntaxKind::Token &&
          "'token' syntax node must be constructed with dedicated constructor");
@@ -143,23 +143,33 @@ RawSyntax::RawSyntax(tok TokKind, StringRef Text, size_t TextLength,
                      StringRef LeadingTrivia, StringRef TrailingTrivia,
                      SourcePresence Presence, const RC<SyntaxArena> &Arena,
                      llvm::Optional<unsigned> NodeId)
-    : Arena(Arena), Bits({{unsigned(TextLength), unsigned(Presence), true}}),
-      RefCount(0) {
+    : RefCount(0), Arena(Arena),
+      Bits({{unsigned(TextLength), unsigned(Presence), true}}) {
   assert(Arena && "RawSyntax nodes must always be allocated in an arena");
   copyToArenaIfNecessary(LeadingTrivia, Arena);
   copyToArenaIfNecessary(Text, Arena);
   copyToArenaIfNecessary(TrailingTrivia, Arena);
 
+  if (Presence == SourcePresence::Missing) {
+    assert(TextLength == 0);
+  } else {
+    assert(TextLength ==
+           LeadingTrivia.size() + Text.size() + TrailingTrivia.size());
+  }
+
   if (NodeId.hasValue()) {
     this->NodeId = NodeId.getValue();
     NextFreeNodeId = std::max(this->NodeId + 1, NextFreeNodeId);
   } else {
     this->NodeId = NextFreeNodeId++;
   }
+  Bits.Token.LeadingTrivia = LeadingTrivia.data();
+  Bits.Token.TokenText = Text.data();
+  Bits.Token.TrailingTrivia = TrailingTrivia.data();
+  Bits.Token.LeadingTriviaLength = LeadingTrivia.size();
+  Bits.Token.TokenLength = Text.size();
+  Bits.Token.TrailingTriviaLength = TrailingTrivia.size();
   Bits.Token.TokenKind = unsigned(TokKind);
-  Bits.Token.LeadingTrivia = LeadingTrivia;
-  Bits.Token.TokenText = Text;
-  Bits.Token.TrailingTrivia = TrailingTrivia;
 }
 
 RawSyntax::~RawSyntax() {