Check for @end in ParseObjCInterfaceDeclList instead of in each caller
Handle @required and @optional with the same code

Add some fixmes about some apparently objc2 code that is being accepted
in objc1.

llvm-svn: 57803

where it would reject @required in non-protocols, but then go
ahead and tag methods with required anyway.  Instead, if we see
this in something other than a protocol, just ignore the request.

Also, improve error recovery a bit when we see something bogus 
inside an interface.

llvm-svn: 57798

to make it easier to understand.  No functionality change.

llvm-svn: 57797

move it to its own predicate to make it more clear.

llvm-svn: 57796

can't stick an attributes?

llvm-svn: 57795

of whether a '(' was a grouping paren or the start of a function declarator.
This is PR2796.

Now we eat the attribute before deciding whether the paren is grouping or
not, then apply it to the resultant decl or to the first argument as needed.

One somewhat surprising aspect of this is that attributes interact with
implicit int in cases like this:

void a(x, y) // k&r style function
void b(__attribute__(()) x, y); // function with two implicit int arguments
void c(x, __attribute__(()) y); // error, can't have attr in identifier list.

Fun stuff.

llvm-svn: 57790

llvm-svn: 57789

llvm-svn: 57788

when ObjC is turned on.

llvm-svn: 57787

be either deleted or referenced afterwards.

llvm-svn: 57786

llvm-svn: 57785

this everywhere in LegalizeTypes.

llvm-svn: 57783

elements.  Otherwise LegalizeTypes will, reasonably
enough, legalize the mask, which may result in it
no longer being a BUILD_VECTOR node (LegalizeDAG
simply ignores the legality or not of vector masks).

llvm-svn: 57782

The casts<> require that "Decl.h" is included before "Type.h", use reinterpret_cast<> to remove that kind of dependency.

llvm-svn: 57781

llvm-svn: 57780

llvm-svn: 57779

 - Support noreturn on function-typed variables.

 - Extend isFunctionOrMethod to return true for K&R functions and
   provide hasFunctionProto to check if a decl has information about
   its arguments. This code needs some serious cleaning, but works.

 - Add/improve test cases for noreturn and unused.

llvm-svn: 57778

llvm-svn: 57777

llvm-svn: 57775

Patch by Timo Sirainen!

llvm-svn: 57772

the previous patch this one actually passes make check.

"Fix PR2356 on PowerPC: if we have an input and output that are tied together
that have different sizes (e.g. i32 and i64) make sure to reserve registers for
the bigger operand."

llvm-svn: 57771

llvm-svn: 57770

for Obj-C methods.

llvm-svn: 57769

llvm-svn: 57766

llvm-svn: 57765

llvm-svn: 57764

Generalize VisitLValue: not only can CallExprs and ObjCMessageExprs return aggregate temporaries, so can many other expressions.

llvm-svn: 57761

Function calls and ObjC message expressions can be used in a lvalue context if they return a structure.  E.g  foo().x == 1.  We don't really support, however, such temporaries yet in the environment or the store.

llvm-svn: 57760

retain/release checker: Check if a tracked value escapes if we also try binding it to the store and the store doesn't support that binding (i.e., it cannot track it).  This has the nice feature that the checker will automatically get more powerful if we use a more powerful store model.

llvm-svn: 57755

Use "VisitLValue" when processing the base for "x.f" field accesses, and "Visit" when processing the base for "x->f" field accesses.

llvm-svn: 57754

llvm-svn: 57751

llvm-svn: 57750

llvm-svn: 57749

and add a TargetLowering hook for it to use to determine when this
is legal (i.e. not in PIC mode, etc.)

This allows instruction selection to emit folded constant offsets
in more cases, such as the included testcase, eliminating the need
for explicit arithmetic instructions.

This eliminates the need for the C++ code in X86ISelDAGToDAG.cpp
that attempted to achieve the same effect, but wasn't as effective.

Also, fix handling of offsets in GlobalAddressSDNodes in several
places, including changing GlobalAddressSDNode's offset from
int to int64_t.

The Mips, Alpha, Sparc, and CellSPU targets appear to be
unaware of GlobalAddress offsets currently, so set the hook to
false on those targets.

llvm-svn: 57748

test/CodeGen/X86/2008-09-17-inline-asm-1.ll
and a few others, and it breaks the llvm-gcc build.

llvm-svn: 57747

No functionality change.

llvm-svn: 57746

llvm-svn: 57745

llvm-svn: 57744

llvm-svn: 57743

llvm-svn: 57742