Merge bastille into main

clang/include/clang/Basic/Attr.td

@@ -3594,3 +3594,11 @@ def ReleaseHandle : InheritableParamAttr {
   let Subjects = SubjectList<[ParmVar]>;
   let Documentation = [ReleaseHandleDocs];
 }
+
+def Builtin : InheritableAttr {
+  let Spellings = [];
+  let Args = [UnsignedArgument<"ID">];
+  let Subjects = SubjectList<[Function]>;
+  let SemaHandler = 0;
+  let Documentation = [Undocumented];
+}

clang/include/clang/Basic/Builtins.def

@@ -75,6 +75,9 @@
 //  U -> pure
 //  c -> const
 //  t -> signature is meaningless, use custom typechecking
+//  T -> type is not important to semantic analysis and codegen; recognize as
+//       builtin even if type doesn't match signature, and don't warn if we
+//       can't be sure the type is right
 //  F -> this is a libc/libm function with a '__builtin_' prefix added.
 //  f -> this is a libc/libm function without the '__builtin_' prefix. It can
 //       be followed by ':headername:' to state which header this function
@@ -893,7 +896,7 @@ LANGBUILTIN(__va_start,       "vc**.", "nt", ALL_MS_LANGUAGES)
 LANGBUILTIN(__fastfail, "vUi",    "nr", ALL_MS_LANGUAGES)
 
 // Microsoft library builtins.
-LIBBUILTIN(_setjmpex, "iJ", "fj",   "setjmpex.h", ALL_MS_LANGUAGES)
+LIBBUILTIN(_setjmpex, "iJ", "fjT", "setjmpex.h", ALL_MS_LANGUAGES)
 
 // C99 library functions
 // C99 stdarg.h
@@ -987,8 +990,8 @@ LIBBUILTIN(wmemmove,"w*w*wC*z", "f", "wchar.h", ALL_LANGUAGES)
 // In some systems setjmp is a macro that expands to _setjmp. We undefine
 // it here to avoid having two identical LIBBUILTIN entries.
 #undef setjmp
-LIBBUILTIN(setjmp, "iJ",          "fj",    "setjmp.h", ALL_LANGUAGES)
-LIBBUILTIN(longjmp, "vJi",        "fr",    "setjmp.h", ALL_LANGUAGES)
+LIBBUILTIN(setjmp, "iJ",          "fjT",   "setjmp.h", ALL_LANGUAGES)
+LIBBUILTIN(longjmp, "vJi",        "frT",   "setjmp.h", ALL_LANGUAGES)
 
 // Non-C library functions, active in GNU mode only.
 // Functions with (returns_twice) attribute (marked as "j") are still active in
@@ -1015,20 +1018,22 @@ LIBBUILTIN(strcasecmp, "icC*cC*", "f",     "strings.h", ALL_GNU_LANGUAGES)
 LIBBUILTIN(strncasecmp, "icC*cC*z", "f",   "strings.h", ALL_GNU_LANGUAGES)
 // POSIX unistd.h
 LIBBUILTIN(_exit, "vi",           "fr",    "unistd.h", ALL_GNU_LANGUAGES)
-LIBBUILTIN(vfork, "p",            "fj",    "unistd.h", ALL_LANGUAGES)
+LIBBUILTIN(vfork, "p",            "fjT",   "unistd.h", ALL_LANGUAGES)
 // POSIX pthread.h
+// FIXME: Should specify argument types.
 LIBBUILTIN(pthread_create, "",  "fC<2,3>", "pthread.h", ALL_GNU_LANGUAGES)
 
 // POSIX setjmp.h
 
-LIBBUILTIN(_setjmp, "iJ",         "fj",   "setjmp.h", ALL_LANGUAGES)
-LIBBUILTIN(__sigsetjmp, "iSJi",   "fj",   "setjmp.h", ALL_LANGUAGES)
-LIBBUILTIN(sigsetjmp, "iSJi",     "fj",   "setjmp.h", ALL_LANGUAGES)
-LIBBUILTIN(savectx, "iJ",         "fj",   "setjmp.h", ALL_LANGUAGES)
-LIBBUILTIN(getcontext, "iK*",     "fj",   "setjmp.h", ALL_LANGUAGES)
+// FIXME: MinGW _setjmp has an additional void* parameter.
+LIBBUILTIN(_setjmp, "iJ",         "fjT",   "setjmp.h", ALL_LANGUAGES)
+LIBBUILTIN(__sigsetjmp, "iSJi",   "fjT",   "setjmp.h", ALL_LANGUAGES)
+LIBBUILTIN(sigsetjmp, "iSJi",     "fjT",   "setjmp.h", ALL_LANGUAGES)
+LIBBUILTIN(savectx, "iJ",         "fjT",   "setjmp.h", ALL_LANGUAGES)
+LIBBUILTIN(getcontext, "iK*",     "fjT",   "setjmp.h", ALL_LANGUAGES)
 
-LIBBUILTIN(_longjmp, "vJi",       "fr",    "setjmp.h", ALL_GNU_LANGUAGES)
-LIBBUILTIN(siglongjmp, "vSJi",    "fr",    "setjmp.h", ALL_GNU_LANGUAGES)
+LIBBUILTIN(_longjmp, "vJi",       "frT",   "setjmp.h", ALL_GNU_LANGUAGES)
+LIBBUILTIN(siglongjmp, "vSJi",    "frT",   "setjmp.h", ALL_GNU_LANGUAGES)
 // non-standard but very common
 LIBBUILTIN(strlcpy, "zc*cC*z",    "f",     "string.h", ALL_GNU_LANGUAGES)
 LIBBUILTIN(strlcat, "zc*cC*z",    "f",     "string.h", ALL_GNU_LANGUAGES)

clang/include/clang/Basic/Builtins.h

@@ -158,6 +158,13 @@ class Context {
     return strchr(getRecord(ID).Attributes, 't') != nullptr;
   }
 
+  /// Determines whether a declaration of this builtin should be recognized
+  /// even if the type doesn't match the specified signature.
+  bool allowTypeMismatch(unsigned ID) const {
+    return strchr(getRecord(ID).Attributes, 'T') != nullptr ||
+           hasCustomTypechecking(ID);
+  }
+
   /// Determines whether this builtin has a result or any arguments which
   /// are pointer types.
   bool hasPtrArgsOrResult(unsigned ID) const {

clang/include/clang/Basic/IdentifierTable.h

@@ -221,18 +221,6 @@ class alignas(IdentifierInfoAlignment) IdentifierInfo {
   }
   void setObjCKeywordID(tok::ObjCKeywordKind ID) { ObjCOrBuiltinID = ID; }
 
-  /// True if setNotBuiltin() was called.
-  bool hasRevertedBuiltin() const {
-    return ObjCOrBuiltinID == tok::NUM_OBJC_KEYWORDS;
-  }
-
-  /// Revert the identifier to a non-builtin identifier. We do this if
-  /// the name of a known builtin library function is used to declare that
-  /// function, but an unexpected type is specified.
-  void revertBuiltin() {
-    setBuiltinID(0);
-  }
-
   /// Return a value indicating whether this is a builtin function.
   ///
   /// 0 is not-built-in. 1+ are specific builtin functions.

clang/include/clang/Sema/Sema.h

@@ -3990,6 +3990,8 @@ class Sema final {
   ObjCInterfaceDecl *getObjCInterfaceDecl(IdentifierInfo *&Id,
                                           SourceLocation IdLoc,
                                           bool TypoCorrection = false);
+  FunctionDecl *CreateBuiltin(IdentifierInfo *II, QualType Type, unsigned ID,
+                              SourceLocation Loc);
   NamedDecl *LazilyCreateBuiltin(IdentifierInfo *II, unsigned ID,
                                  Scope *S, bool ForRedeclaration,
                                  SourceLocation Loc);

clang/lib/AST/Decl.cpp

@@ -3163,44 +3163,24 @@ FunctionDecl *FunctionDecl::getCanonicalDecl() { return getFirstDecl(); }
 /// functions as their wrapped builtins. This shouldn't be done in general, but
 /// it's useful in Sema to diagnose calls to wrappers based on their semantics.
 unsigned FunctionDecl::getBuiltinID(bool ConsiderWrapperFunctions) const {
-  unsigned BuiltinID;
+  unsigned BuiltinID = 0;
 
   if (const auto *ABAA = getAttr<ArmBuiltinAliasAttr>()) {
     BuiltinID = ABAA->getBuiltinName()->getBuiltinID();
-  } else {
-    if (!getIdentifier())
-      return 0;
-
-    BuiltinID = getIdentifier()->getBuiltinID();
+  } else if (const auto *A = getAttr<BuiltinAttr>()) {
+    BuiltinID = A->getID();
   }
 
   if (!BuiltinID)
     return 0;
 
-  ASTContext &Context = getASTContext();
-  if (Context.getLangOpts().CPlusPlus) {
-    const auto *LinkageDecl =
-        dyn_cast<LinkageSpecDecl>(getFirstDecl()->getDeclContext());
-    // In C++, the first declaration of a builtin is always inside an implicit
-    // extern "C".
-    // FIXME: A recognised library function may not be directly in an extern "C"
-    // declaration, for instance "extern "C" { namespace std { decl } }".
-    if (!LinkageDecl) {
-      if (BuiltinID == Builtin::BI__GetExceptionInfo &&
-          Context.getTargetInfo().getCXXABI().isMicrosoft())
-        return Builtin::BI__GetExceptionInfo;
-      return 0;
-    }
-    if (LinkageDecl->getLanguage() != LinkageSpecDecl::lang_c)
-      return 0;
-  }
-
   // If the function is marked "overloadable", it has a different mangled name
   // and is not the C library function.
   if (!ConsiderWrapperFunctions && hasAttr<OverloadableAttr>() &&
       !hasAttr<ArmBuiltinAliasAttr>())
     return 0;
 
+  ASTContext &Context = getASTContext();
   if (!Context.BuiltinInfo.isPredefinedLibFunction(BuiltinID))
     return BuiltinID;
 

clang/lib/CodeGen/CGBuiltin.cpp

@@ -3827,11 +3827,13 @@ RValue CodeGenFunction::EmitBuiltinExpr(const GlobalDecl GD, unsigned BuiltinID,
   case Builtin::BI_abnormal_termination:
     return RValue::get(EmitSEHAbnormalTermination());
   case Builtin::BI_setjmpex:
-    if (getTarget().getTriple().isOSMSVCRT())
+    if (getTarget().getTriple().isOSMSVCRT() && E->getNumArgs() == 1 &&
+        E->getArg(0)->getType()->isPointerType())
       return EmitMSVCRTSetJmp(*this, MSVCSetJmpKind::_setjmpex, E);
     break;
   case Builtin::BI_setjmp:
-    if (getTarget().getTriple().isOSMSVCRT()) {
+    if (getTarget().getTriple().isOSMSVCRT() && E->getNumArgs() == 1 &&
+        E->getArg(0)->getType()->isPointerType()) {
       if (getTarget().getTriple().getArch() == llvm::Triple::x86)
         return EmitMSVCRTSetJmp(*this, MSVCSetJmpKind::_setjmp3, E);
       else if (getTarget().getTriple().getArch() == llvm::Triple::aarch64)