Note that incompatible-protocol-qualified-types.m is currently failing. This is
unrelated to this patch and Steve is looking at the general problem of not reporting
incompitible pointer types in return stetement..

llvm-svn: 44897

"@class Protocol;"

llvm-svn: 44670

standard says that we should adjust the "reference to T" type to "T"
before analysis.

llvm-svn: 44530

http://llvm.org/viewvc/llvm-project?view=rev&revision=44089

"Decl" once again can no longer be NULL, so the NULL checks are not needed.

llvm-svn: 44336

llvm-svn: 44222

in TagDecl*. This allows the deserializer to use ASTContext to create the
TagTypes. Deserialize TagTypes then rely on pointer-backpatching to resolve
the decls.

This may not be the interface that we want, but as the implementation of
TagTypes will potentially change significantly in the future, I'm leaving this
for now. An appropriate FIXME is in place.

llvm-svn: 44089

llvm-svn: 44087

Removed tons of dead code in ASTContext concerning how types use to be
serialized.
Removed serialization methods from QualType that are no longer used.

llvm-svn: 44070

the new serialization API.

llvm-svn: 44035

Teach ASTContext::tagTypesAreCompatible() about the built-in ObjC types (Class and id), removing a bogus warning.

llvm-svn: 43809

serialize Type objects in the order they are serialized in the Types
vector.  We also now rely on the methods within ASTContext to unique
Type objects and handle the actual creation of Type objects (these are
now called by the deserialization code).  This approach solves some
hairy issues with ownership of objects and allows us to naturally
handle recursive types.

llvm-svn: 43787

for ASTContext is still rapidly evolving.

llvm-svn: 43774

of type sets when emitting complex types and pointer types that are
also considered builtins.  These types are automatically created in
the ctor of ASTContext, and thus should not be serialized (was
producing an error during deserialization).

llvm-svn: 43733

Implement rewrite rules for ObjC string constants.

llvm-svn: 43665

Now, at AST level record info is maintained by ASTRecordLayout class.
Now, at code gen  level record info is maintained by CGRecordLayout class.

llvm-svn: 43619

ownership model of some type pointers.  Added FIXMEs to serialization.

Added comments to ASTContext indicating which variables we are intentionally
*not* serializing.

llvm-svn: 43618

method types.

llvm-svn: 43617

each type.  This ensures that the order in which the types are serialized
is clear and remains persistent.

llvm-svn: 43615

llvm-svn: 43604

(hypothetically) read in/write out most of the types.  Bugs likely exist.

llvm-svn: 43584

We now serialize ivar references inside of ASTContext.

llvm-svn: 43571

llvm-svn: 43561

Added skeleton code for serialization of ASTContext.

llvm-svn: 43558

llvm-svn: 43540

llvm-svn: 43512

llvm-svn: 43504

(VLAs with a specified size expresssion).  This vector owns the
references to these type objects.

llvm-svn: 43502

llvm-svn: 43485

and inserted into a FoldingSet owned by ASTContext.

llvm-svn: 43482

llvm-svn: 43481

llvm-svn: 43445

llvm-svn: 43443

llvm-svn: 43439

llvm-svn: 43075

llvm-svn: 43038

Change the type of ObjCStringLiteral from "struct __builtin_CFString *" to "NSConstantString *".

This makes the typecheck much happier. Without this change, the type checker would have to special case "struct __builtin_CFString *". This change does assume the interface for NSConstantString is declared in the translation unit.

I left ASTContext::getCFConstantStringType() around for now (with a comment that says it is currently unused).

llvm-svn: 43021

Move type compatibility predicates from Type to ASTContext. In addition, the predicates are now instance methods (they were previously static class methods on Type).

This allowed me to fix the following hack from this weekend...

// FIXME: Devise a way to do this without using strcmp.
// Would like to say..."return getAsStructureType() == IdStructType;", but
// we don't have a pointer to ASTContext.
bool Type::isObjcIdType() const {
  if (const RecordType *RT = getAsStructureType())
    return !strcmp(RT->getDecl()->getName(), "objc_object");
  return false;
}

...which is now...

bool isObjcIdType(QualType T) const {
  return T->getAsStructureType() == IdStructType;
}

Side notes:

- I had to remove a convenience function from the TypesCompatibleExpr class.

int typesAreCompatible() const {return Type::typesAreCompatible(Type1,Type2);}

Which required a couple clients get a little more verbose...

-    Result = TCE->typesAreCompatible();
+    Result = Ctx.typesAreCompatible(TCE->getArgType1(), TCE->getArgType2());

Overall, I think this change also makes sense for a couple reasons...

1) Since ASTContext vends types, it makes sense for the type compatibility API to be there.
2) This allows the type compatibility predeciates to refer to data not strictly present in the AST (which I have found problematic on several occasions).

llvm-svn: 43009

Added ASTContext::setObjcIdType/getObjcIdType(), set by Sema.

Also noticed ASTContext::BuiltinVaListType wasn't being initialized to the null type (so I set it).

llvm-svn: 42983

llvm-svn: 42858

Add __builtin_va_start to the list of builtins, make __builtin_va_list available to builtin functions.

llvm-svn: 42857