PR15067: Don't assert when a UCN appears in a C90 file.

Unfortunately, we can't accept the UCN as an extension because we're
required to treat it as two tokens for preprocessing purposes.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@173622 91177308-0d34-0410-b5e6-96231b3b80d8

4 files changed, 16 insertions, 8 deletions

diff --git a/include/clang/Basic/DiagnosticLexKinds.td b/include/clang/Basic/DiagnosticLexKinds.td
index e3bff51b8e..81e8e9cf84 100644
--- a/include/clang/Basic/DiagnosticLexKinds.td
+++ b/include/clang/Basic/DiagnosticLexKinds.td
@@ -127,6 +127,11 @@ def warn_cxx98_compat_literal_ucn_escape_basic_scs : Warning<
 def warn_cxx98_compat_literal_ucn_control_character : Warning<
   "universal character name referring to a control character "
   "is incompatible with C++98">, InGroup<CXX98Compat>, DefaultIgnore;
+def warn_ucn_not_valid_in_c89 : Warning<
+  "universal character names are only valid in C99 or C++; "
+  "treating as '\\' followed by identifier">, InGroup<Unicode>;
+def warn_ucn_not_valid_in_c89_literal : ExtWarn<
+  "universal character names are only valid in C99 or C++">, InGroup<Unicode>;
 
 
 // Literal
@@ -165,8 +170,6 @@ def ext_string_too_long : Extension<"string literal of length %0 exceeds "
   "support">, InGroup<OverlengthStrings>;
 def err_character_too_large : Error<
   "character too large for enclosing character literal type">;
-def warn_ucn_not_valid_in_c89 : ExtWarn<
-  "unicode escape sequences are only valid in C99 or C++">, InGroup<Unicode>;
 def warn_cxx98_compat_unicode_literal : Warning<
   "unicode literals are incompatible with C++98">,
   InGroup<CXX98Compat>, DefaultIgnore;
diff --git a/lib/Lex/Lexer.cpp b/lib/Lex/Lexer.cpp
index 429780eaea..91c4f66403 100644
--- a/lib/Lex/Lexer.cpp
+++ b/lib/Lex/Lexer.cpp
@@ -2698,8 +2698,6 @@ bool Lexer::isCodeCompletionPoint(const char *CurPtr) const {
 
 uint32_t Lexer::tryReadUCN(const char *&StartPtr, const char *SlashLoc,
                            Token *Result) {
-  assert(LangOpts.CPlusPlus || LangOpts.C99);
-
   unsigned CharSize;
   char Kind = getCharAndSize(StartPtr, CharSize);
 
@@ -2711,6 +2709,11 @@ uint32_t Lexer::tryReadUCN(const char *&StartPtr, const char *SlashLoc,
   else
     return 0;
 
+  if (!LangOpts.CPlusPlus && !LangOpts.C99) {
+    Diag(SlashLoc, diag::warn_ucn_not_valid_in_c89);
+    return 0;
+  }
+
   const char *CurPtr = StartPtr + CharSize;
   const char *KindLoc = &CurPtr[-1];
 
@@ -2737,7 +2740,7 @@ uint32_t Lexer::tryReadUCN(const char *&StartPtr, const char *SlashLoc,
           }
         }
       }
-      
+
       return 0;
     }
 
diff --git a/lib/Lex/LiteralSupport.cpp b/lib/Lex/LiteralSupport.cpp
index 1647aa5381..09a94a1143 100644
--- a/lib/Lex/LiteralSupport.cpp
+++ b/lib/Lex/LiteralSupport.cpp
@@ -277,7 +277,7 @@ static bool ProcessUCNEscape(const char *ThisTokBegin, const char *&ThisTokBuf,
 
   if (!Features.CPlusPlus && !Features.C99 && Diags)
     Diag(Diags, Features, Loc, ThisTokBegin, UcnBegin, ThisTokBuf,
-         diag::warn_ucn_not_valid_in_c89);
+         diag::warn_ucn_not_valid_in_c89_literal);
 
   return true;
 }
diff --git a/test/Lexer/c90.c b/test/Lexer/c90.c
index 7142c09ac4..826d910b29 100644
--- a/test/Lexer/c90.c
+++ b/test/Lexer/c90.c
@@ -29,8 +29,8 @@ void test2() {
 }
 
 void test3() {
-  (void)L"\u1234";  // expected-error {{unicode escape sequences are only valid in C99 or C++}}
-  (void)L'\u1234';  // expected-error {{unicode escape sequences are only valid in C99 or C++}}
+  (void)L"\u1234";  // expected-error {{universal character names are only valid in C99 or C++}}
+  (void)L'\u1234';  // expected-error {{universal character names are only valid in C99 or C++}}
 }
 
 #define PREFIX(x) foo ## x
@@ -39,3 +39,5 @@ int test4() {
   int *p = &PREFIX(0p+1);
   return p[-1];
 }
+
+#define MY_UCN \u00FC // expected-warning {{universal character names are only valid in C99 or C++; treating as '\' followed by identifier}}