Do not emit diagnostics for invalid unicode characters in preprocessing mode

cor3ntin · AaronBallman · commit 2715c4da5038 · 2021-08-18T09:12:36.000-04:00
This amends 4e80636 with a fix for https://lab.llvm.org/buildbot/#/builders/139/builds/8943
diff --git a/clang/lib/Lex/Lexer.cpp b/clang/lib/Lex/Lexer.cpp
@@ -1652,7 +1652,8 @@ bool Lexer::tryConsumeIdentifierUCN(const char *&CurPtr, unsigned Size,
   if (!isAllowedIDChar(CodePoint, LangOpts)) {
     if (isASCII(CodePoint) || isUnicodeWhitespace(CodePoint))
       return false;
-    if (!isLexingRawMode())
+    if (!isLexingRawMode() && !ParsingPreprocessorDirective &&
+        !PP->isPreprocessedOutput())
       diagnoseInvalidUnicodeCodepointInIdentifier(
           PP->getDiagnostics(), LangOpts, CodePoint,
           makeCharRange(*this, CurPtr, UCNPtr),
@@ -1691,7 +1692,8 @@ bool Lexer::tryConsumeIdentifierUTF8Char(const char *&CurPtr) {
     if (isASCII(CodePoint) || isUnicodeWhitespace(CodePoint))
       return false;
 
-    if (!isLexingRawMode())
+    if (!isLexingRawMode() && !ParsingPreprocessorDirective &&
+        !PP->isPreprocessedOutput())
       diagnoseInvalidUnicodeCodepointInIdentifier(
           PP->getDiagnostics(), LangOpts, CodePoint,
           makeCharRange(*this, CurPtr, UnicodePtr), /*IsFirst=*/false);
diff --git a/clang/test/Lexer/unicode.c b/clang/test/Lexer/unicode.c
@@ -1,7 +1,7 @@
 // RUN: %clang_cc1 -fsyntax-only -verify -x c -std=c11 %s
 // RUN: %clang_cc1 -fsyntax-only -verify -x c++ -std=c++11 %s
-// RUN: %clang_cc1 -E -DPP_ONLY=1 %s -o %t
-// RUN: FileCheck --strict-whitespace --input-file=%t %s
+// RUN: %clang_cc1 -triple x86_64-scei-ps4 -E -DPP_ONLY=1 %s | FileCheck %s --strict-whitespace
+// RUN: %clang_cc1 -E -DPP_ONLY=1 %s | FileCheck %s --strict-whitespace
 
 // This file contains Unicode characters; please do not "fix" them!
 
