Diagnose extensions in 'offsetof'

https://www.open-std.org/jtc1/sc22/wg14/www/docs/n2350.htm made very
clear that it is an UB having type definitions with in offsetof.
Clang supports defining a type as the first argument as a conforming
extension due to how many projects use the construct in C99 and earlier
to calculate the alignment of a type. GCC also supports defining a type
as the first argument.

This adds extension warnings and documentation for the functionality
Clang explicitly supports.

Fixes #57065
Reverts the revert of 39da55e8f548a11f7dadefa73ea73d809a5f1729

Co-authored-by: Yingchi Long <i@lyc.dev>
Co-authored-by: Aaron Ballman <aaron@aaronballman.com>

Differential Revision: https://reviews.llvm.org/D133574

1 files changed, 23 insertions, 5 deletions

diff --git a/clang/lib/Parse/ParseExpr.cpp b/clang/lib/Parse/ParseExpr.cpp
index b1bf988..392ed29 100644
--- a/clang/lib/Parse/ParseExpr.cpp
+++ b/clang/lib/Parse/ParseExpr.cpp
@@ -2592,10 +2592,21 @@ ExprResult Parser::ParseBuiltinPrimaryExpression() {
   }
   case tok::kw___builtin_offsetof: {
     SourceLocation TypeLoc = Tok.getLocation();
-    TypeResult Ty = ParseTypeName();
-    if (Ty.isInvalid()) {
-      SkipUntil(tok::r_paren, StopAtSemi);
-      return ExprError();
+    auto OOK = Sema::OffsetOfKind::OOK_Builtin;
+    if (Tok.getLocation().isMacroID()) {
+      StringRef MacroName = Lexer::getImmediateMacroNameForDiagnostics(
+          Tok.getLocation(), PP.getSourceManager(), getLangOpts());
+      if (MacroName == "offsetof")
+        OOK = Sema::OffsetOfKind::OOK_Macro;
+    }
+    TypeResult Ty;
+    {
+      OffsetOfStateRAIIObject InOffsetof(*this, OOK);
+      Ty = ParseTypeName();
+      if (Ty.isInvalid()) {
+        SkipUntil(tok::r_paren, StopAtSemi);
+        return ExprError();
+      }
     }
 
     if (ExpectAndConsume(tok::comma)) {
@@ -2618,6 +2629,12 @@ ExprResult Parser::ParseBuiltinPrimaryExpression() {
     Comps.back().U.IdentInfo = Tok.getIdentifierInfo();
     Comps.back().LocStart = Comps.back().LocEnd = ConsumeToken();
 
+    enum class Kind { MemberAccess, ArraySubscript };
+    auto DiagExt = [&](SourceLocation Loc, Kind K) {
+      Diag(Loc, diag::ext_offsetof_member_designator)
+          << (K == Kind::ArraySubscript) << (OOK == Sema::OOK_Macro);
+    };
+
     // FIXME: This loop leaks the index expressions on error.
     while (true) {
       if (Tok.is(tok::period)) {
@@ -2631,9 +2648,9 @@ ExprResult Parser::ParseBuiltinPrimaryExpression() {
           SkipUntil(tok::r_paren, StopAtSemi);
           return ExprError();
         }
+        DiagExt(Comps.back().LocStart, Kind::MemberAccess);
         Comps.back().U.IdentInfo = Tok.getIdentifierInfo();
         Comps.back().LocEnd = ConsumeToken();
-
       } else if (Tok.is(tok::l_square)) {
         if (CheckProhibitedCXX11Attribute())
           return ExprError();
@@ -2649,6 +2666,7 @@ ExprResult Parser::ParseBuiltinPrimaryExpression() {
           SkipUntil(tok::r_paren, StopAtSemi);
           return Res;
         }
+        DiagExt(Comps.back().LocStart, Kind::ArraySubscript);
         Comps.back().U.E = Res.get();
 
         ST.consumeClose();