The problem this change addresses is that we treat __is_pod and
__is_empty as keywords in C++, because they are built-in type traits
in GCC >= 4.3. However, GNU libstdc++ 4.2 (and possibly earlier
versions) define implementation-detail struct templates named __is_pod
and __is_empty.
This commit solves the problem by recognizing
struct __is_pod
and
struct __is_empty
as special token sequences. When one of these token sequences is
encountered, the keyword (__is_pod or __is_empty) is implicitly
downgraded to an identifier so that parsing can continue. This is an
egregious hack, but it has the virtue of "just working" whether
someone is using libstdc++ 4.2 or not, without the need for special
flags.
llvm-svn: 80988
x->Base::f
We no longer try to "enter" the context of the type that "x" points
to. Instead, we drag that object type through the parser and pass it
into the Sema routines that need to know how to perform lookup within
member access expressions.
We now implement most of the crazy name lookup rules in C++
[basic.lookup.classref] for non-templated code, including performing
lookup both in the context of the type referred to by the member
access and in the scope of the member access itself and then detecting
ambiguities when the two lookups collide (p1 and p4; p3 and p7 are
still TODO). This change also corrects our handling of name lookup
within template arguments of template-ids inside the
nested-name-specifier (p6; we used to look into the scope of the
object expression for them) and fixes PR4703.
I have disabled some tests that involve member access expressions
where the object expression has dependent type, because we don't yet
have the ability to describe dependent nested-name-specifiers starting
with an identifier.
llvm-svn: 80843
TypenameType if getTypeName is looking at a member of an unknown
specialization. This allows us to properly parse class templates that
derived from type that could only otherwise be described by a typename type,
e.g.,
template<class T> struct X {};
template<typename T> struct Y : public X<T>::X { };
Fixes PR4381.
llvm-svn: 80123
their members, including member class template, member function
templates, and member classes and functions of member templates.
To actually parse the nested-name-specifiers that qualify the name of
an out-of-line definition of a member template, e.g.,
template<typename X> template<typename Y>
X Outer<X>::Inner1<Y>::foo(Y) {
return X();
}
we need to look for the template names (e.g., "Inner1") as a member of
the current instantiation (Outer<X>), even before we have entered the
scope of the current instantiation. Since we can't do this in general
(i.e., we should not be looking into all dependent
nested-name-specifiers as if they were the current instantiation), we
rely on the parser to tell us when it is parsing a declaration
specifier sequence, and, therefore, when we should consider the
current scope specifier to be a current instantiation.
Printing of complicated, dependent nested-name-specifiers may be
somewhat broken by this commit; I'll add tests for this issue and fix
the problem (if it still exists) in a subsequent commit.
llvm-svn: 80044
member templates declared inside other templates. This allows us to
match out-of-line definitions of member function templates within
class templates to the declarations within the class template. We
still can't handle out-of-line definitions for member class templates,
however.
llvm-svn: 79955
and will participate in overload resolution. Unify the instantiation
of CXXMethodDecls and CXXConstructorDecls, which had already gotten
out-of-sync.
llvm-svn: 79658
FriendFunctionDecl, and create instances as appropriate.
The design of FriendFunctionDecl is still somewhat up in the air; you can
befriend arbitrary types of functions --- methods, constructors, etc. ---
and it's not clear that this representation captures that very well.
We'll have a better picture when we start consuming this data in access
control.
llvm-svn: 78653
Fixes PR4704 problems
Addresses Eli's patch feedback re: ugly cast code
Updates all postfix operators to remove ParenListExprs. While this is awful,
no better solution (say, in the parser) is obvious to me. Better solutions
welcome.
llvm-svn: 78621
--- Reverse-merging r78535 into '.':
D test/Sema/altivec-init.c
U include/clang/Basic/DiagnosticSemaKinds.td
U include/clang/AST/Expr.h
U include/clang/AST/StmtNodes.def
U include/clang/Parse/Parser.h
U include/clang/Parse/Action.h
U tools/clang-cc/clang-cc.cpp
U lib/Frontend/PrintParserCallbacks.cpp
U lib/CodeGen/CGExprScalar.cpp
U lib/Sema/SemaInit.cpp
U lib/Sema/Sema.h
U lib/Sema/SemaExpr.cpp
U lib/Sema/SemaTemplateInstantiateExpr.cpp
U lib/AST/StmtProfile.cpp
U lib/AST/Expr.cpp
U lib/AST/StmtPrinter.cpp
U lib/Parse/ParseExpr.cpp
U lib/Parse/ParseExprCXX.cpp
llvm-svn: 78551
In addition to being defined by the AltiVec PIM, this is also the vector
initializer syntax used by OpenCL, so that vector literals are compatible
with macro arguments.
llvm-svn: 78535
implementation of '#pragma unused' by not constructing intermediate
DeclRefExprs, but instead do the name lookup directly. The
implementation is greatly simplified.
Along the way, degrade '#pragma unused(undeclaredvariable)' to a
warning instead of being a hard error. This implements:
<rdar://problem/6761874> [sema] allow #pragma unused to reference undefined variable (with warning)
llvm-svn: 78019
elsewhere. Very slightly decouples DeclSpec users from knowing the exact
diagnostics to report, and makes it easier to provide different diagnostics in
some places.
llvm-svn: 77990
and __has_trivial_constructor builtin pseudo-functions and
additionally implements __has_trivial_copy and __has_trivial_assign,
from John McCall!
llvm-svn: 76916
point that covers templates and non-templates. This should eliminate
the flood of warnings I introduced yesterday.
Removed the ActOnClassTemplate action, which is no longer used.
llvm-svn: 76881
templates, e.g.,
template<typename T>
struct Outer {
struct Inner;
};
template<typename T>
struct Outer<T>::Inner {
// ...
};
Implementing this feature required some extensions to ActOnTag, which
now takes a set of template parameter lists, and is the precursor to
removing the ActOnClassTemplate function from the parser Action
interface. The reason for this approach is simple: the parser cannot
tell the difference between a class template definition and the
definition of a member of a class template; both have template
parameter lists, and semantic analysis determines what that template
parameter list means.
There is still some cleanup to do with ActOnTag and
ActOnClassTemplate. This commit provides the basic functionality we
need, however.
llvm-svn: 76820
declaration in the AST.
The new ASTContext::getCommentForDecl function searches for a comment
that is attached to the given declaration, and returns that comment,
which may be composed of several comment blocks.
Comments are always available in an AST. However, to avoid harming
performance, we don't actually parse the comments. Rather, we keep the
source ranges of all of the comments within a large, sorted vector,
then lazily extract comments via a binary search in that vector only
when needed (which never occurs in a "normal" compile).
Comments are written to a precompiled header/AST file as a blob of
source ranges. That blob is only lazily loaded when one requests a
comment for a declaration (this never occurs in a "normal" compile).
The indexer testbed now supports comment extraction. When the
-point-at location points to a declaration with a Doxygen-style
comment, the indexer testbed prints the associated comment
block(s). See test/Index/comments.c for an example.
Some notes:
- We don't actually attempt to parse the comment blocks themselves,
beyond identifying them as Doxygen comment blocks to associate them
with a declaration.
- We won't find comment blocks that aren't adjacent to the
declaration, because we start our search based on the location of
the declaration.
- We don't go through the necessary hops to find, for example,
whether some redeclaration of a declaration has comments when our
current declaration does not. Similarly, we don't attempt to
associate a \param Foo marker in a function body comment with the
parameter named Foo (although that is certainly possible).
- Verification of my "no performance impact" claims is still "to be
done".
llvm-svn: 74704
Another case where we should use SmallVector::data() instead of taking the
address of element 0 of a SmallVector when the SmallVector has no elements.
llvm-svn: 74556
making sure we return true when annotating a function template with
explicit template arguments, but not when we don't annotate anything.)
llvm-svn: 74383
compilation, and (hopefully) introduce RAII objects for changing the
"potentially evaluated" state at all of the necessary places within
Sema and Parser. Other changes:
- Set the unevaluated/potentially-evaluated context appropriately
during template instantiation.
- We now recognize three different states while parsing or
instantiating expressions: unevaluated, potentially evaluated, and
potentially potentially evaluated (for C++'s typeid).
- When we're in a potentially potentially-evaluated context, queue
up MarkDeclarationReferenced calls in a stack. For C++ typeid
expressions that are potentially evaluated, we will play back
these MarkDeclarationReferenced calls when we exit the
corresponding potentially potentially-evaluated context.
- Non-type template arguments are now parsed as constant
expressions, so they are not potentially-evaluated.
llvm-svn: 73899
C++. This logic is required to trigger implicit instantiation of
function templates and member functions of class templates, which will
be implemented separately.
This commit includes support for -Wunused-parameter, printing warnings
for named parameters that are not used within a function/Objective-C
method/block. Fixes <rdar://problem/6505209>.
llvm-svn: 73797
ExpectAndConsume instead of custom diag logic. This gets us an
insertion hint and positions the ; at the end of the line
instead of on the next token. Before:
t.c:5:1: error: expected ';' after return statement
}
^
after:
t.c:4:11: error: expected ';' after return statement
return 4
^
;
llvm-svn: 73315
specifier resulted in the creation of a new TagDecl node, which
happens either when the tag specifier was a definition or when the tag
specifier was the first declaration of that tag type. This information
has several uses, the first of which is implemented in this commit:
1) In C++, one is not allowed to define tag types within a type
specifier (e.g., static_cast<struct S { int x; } *>(0) is
ill-formed) or within the result or parameter types of a
function. We now diagnose this.
2) We can extend DeclGroups to contain information about any tags
that are declared/defined within the declaration specifiers of a
variable, e.g.,
struct Point { int x, y, z; } p;
This will help improve AST printing and template instantiation,
among other things.
3) For C99, we can keep track of whether a tag type is defined
within the type of a parameter, to properly cope with cases like,
e.g.,
int bar(struct T2 { int x; } y) {
struct T2 z;
}
We can also do similar things wherever there is a type specifier,
e.g., to keep track of where the definition of S occurs in this
legal C99 code:
(struct S { int x, y; } *)0
llvm-svn: 72555
parser. Rather than placing all of the delayed member function
declarations and inline definitions into a single bucket corresponding
to the top-level class, we instead mirror the nesting structure of the
nested classes and place the delayed member functions into their
appropriate place. Then, when we actually parse the delayed member
function declarations, set up the scope stack the same way as it was
when we originally saw the declaration, so that we can find, e.g.,
template parameters that are in scope.
llvm-svn: 72502
references. There are several smallish fixes here:
- Make sure we look through template parameter scope when
determining whether we're parsing a nested class (or nested class
*template*). This makes sure that we delay parsing the bodies of
inline member functions until after we're out of the outermost
class (template) scope.
- Since the bodies of member functions are always parsed
"out-of-line", even when they were declared in-line, teach
unqualified name lookup to look into the (semantic) parents.
- Use the new InstantiateDeclRef to handle the instantiation of a
reference to a declaration (in DeclRefExpr), which drastically
simplifies template instantiation for DeclRefExprs.
- When we're instantiating a ParmVarDecl, it must be in the current
instantiation scope, so only look there.
Also, remove the #if 0's and FIXME's from the dynarray example, which
now compiles and executes thanks to Anders and Eli.
llvm-svn: 72481
(T(*)(int[x+y]));
is an (invalid) paren expression, but "x+y" will be parsed as part of the (rejected) type-id,
so unnecessary Action calls are made for an unused (and possibly leaked) "x+y".
Use a different scheme, similar to parsing inline methods. The parenthesized tokens are cached,
the context that follows is determined (possibly by parsing a cast-expression),
and then we re-introduce the cached tokens into the token stream and parse them appropriately.
llvm-svn: 72279
a paren expression without considering the context past the parentheses.
Behold:
(T())x; - type-id
(T())*x; - type-id
(T())/x; - expression
(T()); - expression
llvm-svn: 72260
Embed its functionality into it's only user, ParseCXXCasts.
CXXCasts now get the "actual" expression directly, they no longer always receive a ParenExpr. This is better since the
parentheses are always part of the C++ casts syntax.
llvm-svn: 72257
-Makes typeof consistent with sizeof/alignof
-Fixes a bug when '>' is in a typeof expression, inside a template type param:
A<typeof(x>1)> a;
llvm-svn: 72255
redundant functionality. The result (ASTOwningVector) lives in
clang/Parse/Ownership.h and is used by both the parser and semantic
analysis. No intended functionality change.
llvm-svn: 72214
template<typename T>
struct X {
struct Inner;
};
template struct X<int>::Inner;
This change is larger than it looks because it also fixes some
a problem with nested-name-specifiers and tags. We weren't requiring
the DeclContext associated with the scope specifier of a tag to be
complete. Therefore, when looking for something like "struct
X<int>::Inner", we weren't instantiating X<int>.
This, naturally, uncovered a problem with member pointers, where we
were requiring the left-hand side of a member pointer access
expression (e.g., x->*) to be a complete type. However, this is wrong:
the semantics of this expression does not require a complete type (EDG
agrees).
Stuart vouched for me. Blame him.
llvm-svn: 71756
template class X<int>;
This also cleans up the propagation of template information through
declaration parsing, which is used to improve some diagnostics.
llvm-svn: 71608
parse just a single declaration and provide a reasonable diagnostic
when the "only one declarator per template declaration" rule is
violated. This eliminates some ugly, ugly hackery where we used to
require thatn the layout of a DeclGroup of a single element be the
same as the layout of a single declaration.
llvm-svn: 71596
'objc_ownership_cfretain' -> 'cf_ownership_retain'
'objc_ownership_cfrelease' -> 'cf_ownership_release'
Motivation: Core Foundation objects can be used in isolation from Objective-C,
and this forces users to reason about the separate semantics of CF objects. More
Sema support pending.
llvm-svn: 70884
return type and the selector. This is inconsistent with C functions
(where such attributes would be placed on the return type, not the the
FunctionDecl), and is inconsistent with what people are use to seeing.
llvm-svn: 70878
appear between the return type and the selector. This is a separate code path
from regular attribute processing, as we only want to (a) accept only a specific
set of attributes in this place and (b) want to distinguish to clients the
context in which an attribute was added to an ObjCMethodDecl.
Currently, the attribute 'objc_ownership_returns' is the only attribute that
uses this new feature. Shortly I will add a warning for 'objc_ownership_returns'
to be placed at the end of a method declaration.
llvm-svn: 70504
type and argument types are missing, and let return type deduction
happen before we give errors for returning from a noreturn block.
Radar 6441502
llvm-svn: 70413
that if we're going to print an extension warning anyway,
there's no point to changing behavior based on NoExtensions: it will
only make error recovery worse.
Note that this doesn't cause any behavior change because NoExtensions
isn't used by the current front-end. I'm still considering what to do about
the remaining use of NoExtensions in IdentifierTable.cpp.
llvm-svn: 70273
as 'objc_ownership_cfretain' except that the method acts like a CFRetain instead
of a [... retain] (important in GC modes). Checker support is wired up, but
currently only for Objective-C message expressions (not function calls).
llvm-svn: 70218
This gets rid of a bunch of random InvalidDecl bools in sema, changing
us to use the following approach:
1. When analyzing a declspec or declarator, if an error is found, we
set a bit in Declarator saying that it is invalid.
2. Once the Decl is created by sema, we immediately set the isInvalid
bit on it from what is in the declarator. From this point on, sema
consistently looks at and sets the bit on the decl.
This gives a very clear separation of concerns and simplifies a bunch
of code. In addition to this, this patch makes these changes:
1. it renames DeclSpec::getInvalidType() -> isInvalidType().
2. various "merge" functions no longer return bools: they just set the
invalid bit on the dest decl if invalid.
3. The ActOnTypedefDeclarator/ActOnFunctionDeclarator/ActOnVariableDeclarator
methods now set invalid on the decl returned instead of returning an
invalid bit byref.
4. In SemaType, refering to a typedef that was invalid now propagates the
bit into the resultant type. Stuff declared with the invalid typedef
will now be marked invalid.
5. Various methods like CheckVariableDeclaration now return void and set the
invalid bit on the decl they check.
There are a few minor changes to tests with this, but the only major bad
result is test/SemaCXX/constructor-recovery.cpp. I'll take a look at this
next.
llvm-svn: 70020
to the checker yet, but essentially it allows a user to specify that an
Objective-C method or C function increments the reference count of a passed
object.
llvm-svn: 70005
up to the checker yet, but essentially it allows a user to specify that an
Objective-C method or C function returns an owned an Objective-C object.
llvm-svn: 70001
by correctly propagating the fact that the type was invalid up to the
attributeRuns decl, then returning an ExprError when attributeRuns is
formed (like we do for normal declrefexprs).
llvm-svn: 69998
for scoping to match C99 even when in C89 mode. This patch fixes this
(eliminating a "redefinition of thisKey" error), and also prevents non-sensical
diagnostics in -pedantic mode like this:
t.m:7:8: warning: variable declaration in for loop is a C99-specific feature
for (id thisKey in keys) ;
^
llvm-svn: 69760
Remove an atrocious amount of trailing whitespace in the overloaded operator mangler. Sorry, couldn't help myself.
Change the DeclType parameter of Sema::CheckReferenceInit to be passed by value instead of reference. It wasn't changed anywhere.
Let the parser handle C++'s irregular grammar around assignment-expression and conditional-expression.
And finally, the reason for all this stuff: implement C++ semantics for the conditional operator. The implementation is complete except for determining lvalueness.
llvm-svn: 69299
with other diagnostic mapping. In the new scheme, -Wfoo or -Wno-foo or
-Werror=foo all override the -pedantic options, and __extension__
robustly silences all extension diagnostics in their scope.
An added bonus of this change is that MAP_DEFAULT goes away, meaning that
per-diagnostic mapping information can now be stored in 2 bits, doubling
the density of the Diagnostic::DiagMapping array. This also
substantially simplifies Diagnostic::getDiagnosticLevel.
OTOH, this temporarily introduces some "macro intensive" code in
Diagnostic.cpp. This will be addressed in a later patch.
llvm-svn: 69154
nested name specifiers. Now we emit stuff like:
t.cpp:8:13: error: unknown type name 'X'
static foo::X P;
~~~~ ^
instead of:
t.cpp:8:16: error: invalid token after top level declarator
static foo::X P;
^
This is inspired by a really awful error message I got from
g++ when I misspelt diag::kind as diag::Kind.
llvm-svn: 69086
that I noticed working on other things.
Instead of emitting:
t2.cc:1:8: error: use of undeclared identifier 'g'
int x(*g);
^
t2.cc:1:10: error: expected ')'
int x(*g);
^
t2.cc:1:6: note: to match this '('
int x(*g);
^
We now only emit:
t2.cc:1:7: warning: type specifier missing, defaults to 'int'
int x(*g);
^
Note that the example in SemaCXX/nested-name-spec.cpp:f4 is still
not great, we now produce both of:
void f4(undef::C); // expected-error {{use of undeclared identifier 'undef'}} \
expected-error {{variable has incomplete type 'void'}}
The second diagnostic should be silenced by something getting marked invalid.
I don't plan to fix this though.
llvm-svn: 68919
struct xyz { int y; };
enum abc { ZZZ };
static xyz b;
abc c;
we used to produce:
t2.c:4:8: error: unknown type name 'xyz'
static xyz b;
^
t2.c:5:1: error: unknown type name 'abc'
abc c;
^
we now produce:
t2.c:4:8: error: use of tagged type 'xyz' without 'struct' tag
static xyz b;
^
struct
t2.c:5:1: error: use of tagged type 'abc' without 'enum' tag
abc c;
^
enum
GCC produces the normal:
t2.c:4: error: expected ‘=’, ‘,’, ‘;’, ‘asm’ or ‘__attribute__’ before ‘b’
t2.c:5: error: expected ‘=’, ‘,’, ‘;’, ‘asm’ or ‘__attribute__’ before ‘c’
rdar://6783347
llvm-svn: 68914
which tries to do better error recovery when it is "obvious" that an
identifier is a mis-typed typename. In this case, we try to parse
it as a typename instead of as the identifier in a declarator, which
gives us several options for better error recovery and immediately
makes diagnostics more useful. For example, we now produce:
t.c:4:8: error: unknown type name 'foo_t'
static foo_t a = 4;
^
instead of:
t.c:4:14: error: invalid token after top level declarator
static foo_t a = 4;
^
Also, since we now parse "a" correctly, we make a decl for it,
preventing later uses of 'a' from emitting things like:
t.c:12:20: error: use of undeclared identifier 'a'
int bar() { return a + b; }
^
I'd really appreciate any scrutiny possible on this, it
is a tricky area.
llvm-svn: 68911
Implement the rvalue reference overload dance for returning local objects. Returning a local object first tries to find a move constructor now.
The error message when no move constructor is defined (or is not applicable) and the copy constructor is deleted is quite ugly, though.
llvm-svn: 68902
down to the ActionBase class. This eliminates dependencies of (e.g.)
DeclSpec.h on Action.h, meaning that action.h can now include these
headers and use their types directly in the actions interfaces.
This is a refactoring to support a future change, no functionality
change.
llvm-svn: 68869
clients of the analyzer to designate custom assertion routines as "noreturn"
functions from the analyzer's perspective but not the compiler's.
llvm-svn: 68746
of the range is now the ';' location. For something like this:
$ cat t2.c
#define bool int
void f(int x, int y) {
bool b = !x && y;
}
We used to produce:
$ clang-cc t2.c -ast-dump
typedef char *__builtin_va_list;
void f(int x, int y)
(CompoundStmt 0x2201f10 <t2.c:3:22, line:5:1>
(DeclStmt 0x2201ef0 <line:2:14> <----
0x2201a20 "int b =
(BinaryOperator 0x2201ed0 <line:4:10, col:16> 'int' '&&'
(UnaryOperator 0x2201e90 <col:10, col:11> 'int' prefix '!'
(DeclRefExpr 0x2201c90 <col:11> 'int' ParmVar='x' 0x2201a50))
(DeclRefExpr 0x2201eb0 <col:16> 'int' ParmVar='y' 0x2201e10))")
Now we produce:
$ clang-cc t2.c -ast-dump
typedef char *__builtin_va_list;
void f(int x, int y)
(CompoundStmt 0x2201f10 <t2.c:3:22, line:5:1>
(DeclStmt 0x2201ef0 <line:2:14, line:4:17> <------
0x2201a20 "int b =
(BinaryOperator 0x2201ed0 <col:10, col:16> 'int' '&&'
(UnaryOperator 0x2201e90 <col:10, col:11> 'int' prefix '!'
(DeclRefExpr 0x2201c90 <col:11> 'int' ParmVar='x' 0x2201a50))
(DeclRefExpr 0x2201eb0 <col:16> 'int' ParmVar='y' 0x2201e10))")
llvm-svn: 68288
failures that involve malformed types, e.g., "typename X::foo" where
"foo" isn't a type, or "std::vector<void>" that doens't instantiate
properly.
Similarly, be a bit smarter in our handling of ambiguities that occur
in Sema::getTypeName, to eliminate duplicate error messages about
ambiguous name lookup.
This eliminates two XFAILs in test/SemaCXX, one of which was crying
out to us, trying to tell us that we were producing repeated error
messages.
llvm-svn: 68251
within nested-name-specifiers, e.g., for the "apply" in
typename MetaFun::template apply<T1, T2>::type
At present, we can't instantiate these nested-name-specifiers, so our
testing is sketchy.
llvm-svn: 68081
representation handles the various ways in which one can name a
template, including unqualified references ("vector"), qualified
references ("std::vector"), and dependent template names
("MetaFun::template apply").
One immediate effect of this change is that the representation of
nested-name-specifiers in type names for class template
specializations (e.g., std::vector<int>) is more accurate. Rather than
representing std::vector<int> as
std::(vector<int>)
we represent it as
(std::vector)<int>
which more closely follows the C++ grammar.
Additionally, templates are no longer represented as declarations
(DeclPtrTy) in Parse-Sema interactions. Instead, I've introduced a new
OpaquePtr type (TemplateTy) that holds the representation of a
TemplateName. This will simplify the handling of dependent
template-names, once we get there.
llvm-svn: 68074
productions (except the already broken ObjC cases like @class X,Y;) in
the parser that can produce more than one Decl return a DeclGroup instead
of a Decl, etc.
This allows elimination of the Decl::NextDeclarator field, and exposes
various clients that should look at all decls in a group, but which were
only looking at one (such as the dumper, printer, etc). These have been
fixed.
Still TODO:
1) there are some FIXME's in the code about potentially using
DeclGroup for better location info.
2) ParseObjCAtDirectives should return a DeclGroup due to @class etc.
3) I'm not sure what is going on with StmtIterator.cpp, or if it can
be radically simplified now.
4) I put a truly horrible hack in ParseTemplate.cpp.
I plan to bring up #3/4 on the mailing list, but don't plan to tackle
#1/2 in the short term.
llvm-svn: 68002
pointer. Its purpose in life is to be a glorified void*, but which does not
implicitly convert to void* or other OpaquePtr's with a different UID.
Introduce Action::DeclPtrTy which is a typedef for OpaquePtr<0>. Change the
entire parser/sema interface to use DeclPtrTy instead of DeclTy*. This
makes the C++ compiler enforce that these aren't convertible to other opaque
types.
We should also convert ExprTy, StmtTy, TypeTy, AttrTy, BaseTy, etc,
but I don't plan to do that in the short term.
The one outstanding known problem with this patch is that we lose the
bitmangling optimization where ActionResult<DeclPtrTy> doesn't know how to
bitmangle the success bit into the low bit of DeclPtrTy. I will rectify
this with a subsequent patch.
llvm-svn: 67952
instantiation for C++ typename-specifiers such as
typename T::type
The parsing of typename-specifiers is relatively easy thanks to
annotation tokens. When we see the "typename", we parse the
typename-specifier and produce a typename annotation token. There are
only a few places where we need to handle this. We currently parse the
typename-specifier form that terminates in an identifier, but not the
simple-template-id form, e.g.,
typename T::template apply<U, V>
Parsing of nested-name-specifiers has a similar problem, since at this
point we don't have any representation of a class template
specialization whose template-name is unknown.
Semantic analysis is only partially complete, with some support for
template instantiation that works for simple examples.
llvm-svn: 67875
class C {
C() { }
int a;
};
C::C() : a(10) { }
We also diagnose when initializers are used on declarations that aren't constructors:
t.cpp:1:10: error: only constructors take base initializers
void f() : a(10) { }
^
Doug and/or Sebastian: I'd appreciate a review, especially the nested-name-spec test results (from the looks of it we now match gcc in that test.)
llvm-svn: 67672
failure to perform a declaration. Instead, explicitly note semantic
failures that occur during template parsing with a DeclResult. Fixes
PR3872.
llvm-svn: 67659
qualified name, e.g.,
foo::x
so that we retain the nested-name-specifier as written in the source
code and can reproduce that qualified name when printing the types
back (e.g., in diagnostics). This is PR3493, which won't be complete
until finished the other tasks mentioned near the end of this commit.
The parser's representation of nested-name-specifiers, CXXScopeSpec,
is now a bit fatter, because it needs to contain the scopes that
precede each '::' and keep track of whether the global scoping
operator '::' was at the beginning. For example, we need to keep track
of the leading '::', 'foo', and 'bar' in
::foo::bar::x
The Action's CXXScopeTy * is no longer a DeclContext *. It's now the
opaque version of the new NestedNameSpecifier, which contains a single
component of a nested-name-specifier (either a DeclContext * or a Type
*, bitmangled).
The new sugar type QualifiedNameType composes a sequence of
NestedNameSpecifiers with a representation of the type we're actually
referring to. At present, we only build QualifiedNameType nodes within
Sema::getTypeName. This will be extended to other type-constructing
actions (e.g., ActOnClassTemplateId).
Also on the way: QualifiedDeclRefExprs will also store a sequence of
NestedNameSpecifiers, so that we can print out the property
nested-name-specifier. I expect to also use this for handling
dependent names like Fibonacci<I - 1>::value.
llvm-svn: 67265
diagnostics. This builds on the patch that Sebastian committed and
then revert. Major differences are:
- We don't remove or use the current ".def" files. Instead, for now,
we just make sure that we're building the ".inc" files.
- Fixed CMake makefiles to run TableGen and build the ".inc" files
when needed. Tested with both the Xcode and Makefile generators
provided by CMake, so it should be solid.
- Fixed normal makefiles to handle out-of-source builds that involve
the ".inc" files.
I'll send a separate patch to the list with Sebastian's changes that
eliminate the use of the .def files.
llvm-svn: 67058
Introduce a new PrettyStackTraceDecl.
Use it to add the top level LLVM IR generation stuff in
Backend.cpp to stack traces. We now get crashes like:
Stack dump:
0. Program arguments: clang t.c -emit-llvm
1. <eof> parser at end of file
2. t.c:1:5: LLVM IR generation of declaration 'a'
Abort
for IR generation crashes.
llvm-svn: 66153
For example:
Stack dump:
0. Program arguments: clang t.cpp
1. t.cpp:4:8: current parser token: ';'
2. t.cpp:3:1: parsing struct/union/class body 'x'
Abort
It is weird that the parser is always "underneath" any parse context
actions, but the parser is created first.
llvm-svn: 66148
like this:
Stack dump:
0. using-directive.cpp:9:14: in compound statement ('{}')
1. using-directive.cpp:9:14: parsing function body 'A::B::f'
2. using-directive.cpp:7:3: parsing namespace 'A::B'
3. using-directive.cpp:5:1: parsing namespace 'A'
4. clang using-directive.cpp
Abort
for testcase like like:
namespace A {
short i;
namespace B {
long i;
void f() { <<crash>>
llvm-svn: 66123
parser. For example, we now print out:
0. t.c:5:10: in compound statement {}
1. t.c:3:12: in compound statement {}
2. clang t.c -fsyntax-only
llvm-svn: 66108
multiple sequential case statements instead of doing it with recursion. This
fixes a problem where we run out of stack space parsing 100K directly nested
cases.
There are a couple other problems that prevent this from being useful in
practice (right now the example only parses correctly with -disable-free and
doesn't work with -emit-llvm), but this is a start.
I'm not including a testcase because it is large and uninteresting for
regtesting.
Sebastian, I would appreciate it if you could scrutinize the smart pointer
gymnastics I do.
llvm-svn: 66011
Also necessary to fix:
<rdar://problem/6632061> [sema] non object types should not be allowed in @catch statements
<rdar://problem/6252237> [sema] qualified id should be disallowed in @catch statements
llvm-svn: 65964
know how to recover from an error, we can attach a hint to the
diagnostic that states how to modify the code, which can be one of:
- Insert some new code (a text string) at a particular source
location
- Remove the code within a given range
- Replace the code within a given range with some new code (a text
string)
Right now, we use these hints to annotate diagnostic information. For
example, if one uses the '>>' in a template argument in C++98, as in
this code:
template<int I> class B { };
B<1000 >> 2> *b1;
we'll warn that the behavior will change in C++0x. The fix is to
insert parenthese, so we use code insertion annotations to illustrate
where the parentheses go:
test.cpp:10:10: warning: use of right-shift operator ('>>') in template
argument will require parentheses in C++0x
B<1000 >> 2> *b1;
^
( )
Use of these annotations is partially implemented for HTML
diagnostics, but it's not (yet) producing valid HTML, which may be
related to PR2386, so it has been #if 0'd out.
In this future, we could consider hooking this mechanism up to the
rewriter to actually try to fix these problems during compilation (or,
after a compilation whose only errors have fixes). For now, however, I
suggest that we use these code modification hints whenever we can, so
that we get better diagnostics now and will have better coverage when
we find better ways to use this information.
This also fixes PR3410 by placing the complaint about missing tokens
just after the previous token (rather than at the location of the next
token).
llvm-svn: 65570
vector<vector<double>> Matrix;
In C++98/03, this token always means "right shift". However, if we're in
a context where we know that it can't mean "right shift", provide a
friendly reminder to put a space between the two >'s and then treat it
as two >'s as part of recovery.
In C++0x, this token is always broken into two '>' tokens.
llvm-svn: 65484
std::vector<int>::allocator_type
When we parse a template-id that names a type, it will become either a
template-id annotation (which is a parsed representation of a
template-id that has not yet been through semantic analysis) or a
typename annotation (where semantic analysis has resolved the
template-id to an actual type), depending on the context. We only
produce a type in contexts where we know that we only need type
information, e.g., in a type specifier. Otherwise, we create a
template-id annotation that can later be "upgraded" by transforming it
into a typename annotation when the parser needs a type. This occurs,
for example, when we've parsed "std::vector<int>" above and then see
the '::' after it. However, it means that when writing something like
this:
template<> class Outer::Inner<int> { ... };
We have two tokens to represent Outer::Inner<int>: one token for the
nested name specifier Outer::, and one template-id annotation token
for Inner<int>, which will be passed to semantic analysis to define
the class template specialization.
Most of the churn in the template tests in this patch come from an
improvement in our error recovery from ill-formed template-ids.
llvm-svn: 65467
us whether there was an error in trying to parse a type-name (type-id
in C++). This allows propagation of errors further in the compiler,
suppressing more bogus error messages.
llvm-svn: 64922
any named parameters, e.g., this is accepted in C:
void f(...) __attribute__((overloadable));
although this would be rejected:
void f(...);
To do this, moved the checking of the "ellipsis without any named
arguments" condition from the parser into Sema (where it belongs anyway).
llvm-svn: 64902
specialization of class templates, e.g.,
template<typename T> class X;
template<> class X<int> { /* blah */ };
Each specialization is a different *Decl node (naturally), and can
have different members. We keep track of forward declarations and
definitions as for other class/struct/union types.
This is only the basic framework: we still have to deal with checking
the template headers properly, improving recovery when there are
failures, handling nested name specifiers, etc.
llvm-svn: 64848
1) implement parser and sema support for reading and verifying attribute(warnunusedresult).
2) rename hasLocalSideEffect to isUnusedResultAWarning, inverting the sense
of its result.
3) extend isUnusedResultAWarning to directly return the loc and range
info that should be reported to the user. Make it substantially more
precise in some cases than what was previously reported.
4) teach isUnusedResultAWarning about CallExpr to decls that are
pure/const/warnunusedresult, fixing a fixme.
5) change warn_attribute_wrong_decl_type to not pass in english strings, instead,
pass in integers and use %select.
llvm-svn: 64543
This commit adds a new attribute, "overloadable", that enables C++
function overloading in C. The attribute can only be added to function
declarations, e.g.,
int *f(int) __attribute__((overloadable));
If the "overloadable" attribute exists on a function with a given
name, *all* functions with that name (and in that scope) must have the
"overloadable" attribute. Sets of overloaded functions with the
"overloadable" attribute then follow the normal C++ rules for
overloaded functions, e.g., overloads must have different
parameter-type-lists from each other.
When calling an overloaded function in C, we follow the same
overloading rules as C++, with three extensions to the set of standard
conversions:
- A value of a given struct or union type T can be converted to the
type T. This is just the identity conversion. (In C++, this would
go through a copy constructor).
- A value of pointer type T* can be converted to a value of type U*
if T and U are compatible types. This conversion has Conversion
rank (it's considered a pointer conversion in C).
- A value of type T can be converted to a value of type U if T and U
are compatible (and are not both pointer types). This conversion
has Conversion rank (it's considered to be a new kind of
conversion unique to C, a "compatible" conversion).
Known defects (and, therefore, next steps):
1) The standard-conversion handling does not understand conversions
involving _Complex or vector extensions, so it is likely to get
these wrong. We need to add these conversions.
2) All overloadable functions with the same name will have the same
linkage name, which means we'll get a collision in the linker (if
not sooner). We'll need to mangle the names of these functions.
llvm-svn: 64336
arguments. This commit covers checking and merging default template
arguments from previous declarations, but it does not cover the actual
use of default template arguments when naming class template
specializations.
llvm-svn: 64229
disambiguation contexts, so that we properly parse template arguments
such as
A<int()>
as type-ids rather than as expressions. Since this can be confusing
(especially when the template parameter is a non-type template
parameter), we try to give a friendly error message.
Almost, eliminate a redundant error message (that should have been a
note) and add some ultra-basic checks for non-type template
arguments.
llvm-svn: 64189
representation for template arguments. Also simplifies the interface
for ActOnClassTemplateSpecialization and eliminates some annoying
allocations of TemplateArgs.
My attempt at smart pointers for template arguments lists is
relatively lame. We can improve it once we're sure that we have the
right representation for template arguments.
llvm-svn: 64154
to a class template. For example, the template-id 'vector<int>' now
has a nice, sugary type in the type system. What we can do now:
- Parse template-ids like 'vector<int>' (where 'vector' names a
class template) and form proper types for them in the type system.
- Parse icky template-ids like 'A<5>' and 'A<(5 > 0)>' properly,
using (sadly) a bool in the parser to tell it whether '>' should
be treated as an operator or not.
This is a baby-step, with major problems and limitations:
- There are currently two ways that we handle template arguments
(whether they are types or expressions). These will be merged, and,
most likely, TemplateArg will disappear.
- We don't have any notion of the declaration of class template
specializations or of template instantiations, so all template-ids
are fancy names for 'int' :)
llvm-svn: 64153
than a Decl, which gives us some more flexibility to express the
results with the type system. There are no clients using this
flexibility yet, but it's meant to be able to describe qualified names
as written in the source (e.g., "foo::type") or template-ids that name
a class template specialization (e.g., "std::vector<INT>").
DeclSpec's TST_typedef has become TST_typename, to reflect its use to
describe types found by name (that may or may not be typedefs). The
type representation of a DeclSpec with TST_typename is an opaque
QualType pointer. All users of TST_typedef, both direct and indirect,
have been updated for these changes.
llvm-svn: 64141
redeclarations. For example, checks that a class template
redeclaration has the same template parameters as previous
declarations.
Detangled class-template checking from ActOnTag, whose logic was
getting rather convoluted because it tried to handle C, C++, and C++
template semantics in one shot.
Made some inroads toward eliminating extraneous "declaration does not
declare anything" errors by adding an "error" type specifier.
llvm-svn: 63973
This shrinks OwningResult by one pointer. Since it is no longer larger than OwningPtr, merge the two.
This leads to simpler client code and speeds up my benchmark by 2.7%.
For some reason, this exposes a previously hidden bug, causing a regression in SemaCXX/condition.cpp.
llvm-svn: 63867
- Support initialization of reference members; complain if any
reference members are left uninitialized.
- Use C++ copy-initialization for initializing each element (falls
back to constraint checking in C)
- Make sure we diagnose when one tries to provide an initializer
list for a non-aggregate.
- Don't complain about empty initializers in C++ (they are permitted)
- Unrelated but necessary: don't bother trying to convert the
decl-specifier-seq to a type when we're dealing with a C++
constructor, destructor, or conversion operator; it results in
spurious warnings.
llvm-svn: 63431
.def file for each library. This means that adding a diagnostic
to sema doesn't require all the other libraries to be rebuilt.
Patch by Anders Johnsen!
llvm-svn: 63111
- When it's safe, ActionResult uses the low bit of the pointer for
the "invalid" flag rather than a separate "bool" value. This keeps
GCC from generating some truly awful code, for a > 3x speedup in the
result-passing microbenchmark.
- When DISABLE_SMART_POINTERS is defined, store an ActionResult
within ASTOwningResult rather than an ASTOwningPtr. Brings the
performance benefits of the above to smart pointers with
DISABLE_SMART_POINTERS defined.
Sadly, these micro-benchmark performance improvements don't seem to
make much of a difference on Cocoa.h right now. However, they're
harmless and might help with future optimizations.
llvm-svn: 63061
special action, inside function prototype scope. This avoids confusion
when we try to inject these parameters into the scope of the function
body before the function itself has been added to the surrounding
scope. Fixes <rdar://problem/6097326>.
llvm-svn: 62849
Uncomment the define in Ownership.h to disable the smart pointers.
Disabled, the smart pointers no longer contain a pointer
to the action, and no longer have special destruction or
copying semantics. They are, compiler willing, raw
pointers or ActionResult equivalents.
llvm-svn: 62767
designated initializers. This implementation should cover all of the
constraints in C99 6.7.8, including long, complex designations and
computing the size of incomplete array types initialized with a
designated initializer. Please see the new test-case and holler if you
find cases where this doesn't work.
There are still some wrinkles with GNU's anonymous structs and
anonymous unions (it isn't clear how these should work; we'll just
follow GCC's lead) and with designated initializers for the members of a
union. I'll tackle those very soon.
CodeGen is still nonexistent, and there's some leftover code in the
parser's representation of designators that I'll also need to clean up.
llvm-svn: 62737
function DeclaratorChunk in common cases. This uses a fixed array in
Declarator when it is small enough for the first function declarator chunk
in a declarator.
This eliminates all malloc/free traffic from DeclaratorChunk::getFunction
when running on Cocoa.h except for five functions: signal/bsd_signal/sigset,
which have multiple Function DeclChunk's, and
CFUUIDCreateWithBytes/CFUUIDGetConstantUUIDWithBytes, which take more than
16 arguments.
This patch was pair programmed with Steve.
llvm-svn: 62599
allocating them from a recycling bump pointer allocator. This
reduces malloc/free traffic of parse-noop (but no other mode),
which makes sharking -parse-noop more meaningful.
llvm-svn: 62460
C++ handle anonymous structs/unions in the same way. Addresses several
bugs:
<rdar://problem/6259534>
<rdar://problem/6481130>
<rdar://problem/6483159>
The test case in PR clang/1750 now passes with -fsyntax-only, but
CodeGen for inline assembler still fails.
llvm-svn: 62112
or enum to be outside that struct, union, or enum. Fixes several
regressions:
<rdar://problem/6487662>
<rdar://problem/6487669>
<rdar://problem/6487684>
<rdar://problem/6487702>
PR clang/3305
PR clang/3312
There is still some work to do in Objective-C++, but this requires
that each of the Objective-C entities (interfaces, implementations,
etc.) to be introduced into the context stack with
PushDeclContext/PopDeclContext. This will be a separate fix, later.
llvm-svn: 62091
that is neither a definition nor a forward declaration and where X has
not yet been declared as a tag, introduce a declaration
into the appropriate scope (which is likely *not* to be the current
scope). The rules for the placement of the declaration differ slightly
in C and C++, so we implement both and test the various corner
cases. This implementation isn't 100% correct due to some lingering
issues with the function prototype scope (for a function parameter
list) not being the same scope as the scope of the function
definition. Testcase is FIXME'd; this probably isn't an important issue.
Addresses <rdar://problem/6484805>.
llvm-svn: 62014
introduce a Scope for the body of a tag. This reduces the number of
semantic differences between C and C++ structs and unions, and will
help with other features (e.g., anonymous unions) in C. Some important
points:
- Fields are now in the "member" namespace (IDNS_Member), to keep
them separate from tags and ordinary names in C. See the new test
in Sema/member-reference.c for an example of why this matters. In
C++, ordinary and member name lookup will find members in both the
ordinary and member namespace, so the difference between
IDNS_Member and IDNS_Ordinary is erased by Sema::LookupDecl (but
only in C++!).
- We always introduce a Scope and push a DeclContext when we're
defining a tag, in both C and C++. Previously, we had different
actions and different Scope/CurContext behavior for enums, C
structs/unions, and C++ structs/unions/classes. Now, it's one pair
of actions. (Yay!)
There's still some fuzziness in the handling of struct/union/enum
definitions within other struct/union/enum definitions in C. We'll
need to do some more cleanup to eliminate some reliance on CurContext
before we can solve this issue for real. What we want is for something
like this:
struct X {
struct T { int x; } t;
};
to introduce T into translation unit scope (placing it at the
appropriate point in the IdentifierResolver chain, too), but it should
still have struct X as its lexical declaration
context. PushOnScopeChains isn't smart enough to do that yet, though,
so there's a FIXME test in nested-redef.c
llvm-svn: 61940
- Simplify ParseDeclCXX to use early exit on error instead of nesting.
- Change ParseDeclCXX to using the 'skip on error' form of ExpectAndConsume.
- If we don't see the ; in a using directive, still call the action, for
hopefully better error recovery.
llvm-svn: 61801
Make C++ classes track the POD property (C++ [class]p4)
Track the existence of a copy assignment operator.
Implicitly declare the copy assignment operator if none is provided.
Implement most of the parsing job for the G++ type traits extension.
Fully implement the low-hanging fruit of the type traits:
__is_pod: Whether a type is a POD.
__is_class: Whether a type is a (non-union) class.
__is_union: Whether a type is a union.
__is_enum: Whether a type is an enum.
__is_polymorphic: Whether a type is polymorphic (C++ [class.virtual]p1).
llvm-svn: 61746
DeclContexts whose members are visible from enclosing DeclContexts up
to (and including) the innermost enclosing non-transparent
DeclContexts. Transparent DeclContexts unify the mechanism to be used
for various language features, including C enumerations, anonymous
unions, C++0x inline namespaces, and C++ linkage
specifications. Please refer to the documentation in the Clang
internals manual for more information.
Only enumerations and linkage specifications currently use transparent
DeclContexts.
Still to do: use transparent DeclContexts to implement anonymous
unions and GCC's anonymous structs extension, and, later, the C++0x
features. We also need to tighten up the DeclContext/ScopedDecl link
to ensure that every ScopedDecl is in a single DeclContext, which
will ensure that we can then enforce ownership and reduce the memory
footprint of DeclContext.
llvm-svn: 61735
verified to be simple type specifiers, so there is no need for it
to call TryAnnotateTypeOrScopeToken.
Make MaybeParseCXXScopeSpecifier reject ::new and ::delete with a
hard error now that it may never be transitively called in a
context where these are legal. This allows me to start
disentangling things more.
llvm-svn: 61659
ParseCastExpression into the switch. This gets it out of the hot
path through ParseCastExpression for all the non-identifier and
non-:: tokens.
llvm-svn: 61643
semantics and improve our handling of default arguments. Specifically,
we follow this order:
- As soon as the see the '}' in the class definition, the class is
complete and we add any implicit declarations (default constructor,
copy constructor, etc.) to the class.
- If there are any default function arguments, parse them
- If there were any inline member function definitions, parse them
As part of this change, we now keep track of the the fact that we've
seen unparsed default function arguments within the AST. See the new
ParmVarDecl::hasUnparsedDefaultArg member. This allows us to properly
cope with calls inside default function arguments to other functions
where we're making use of the default arguments.
Made some C++ error messages regarding failed initializations more
specific.
llvm-svn: 61406
DeclContext. Instead, just keep the list of currently-active
declarations and only build the OverloadedFunctionDecl when we
absolutely need it.
This is a half-step toward eliminating the need to explicitly build
OverloadedFunctionDecls that store sets of overloaded
functions. This was suggested by Argiris a while back, and it's a good
thing for several reasons: first, it eliminates the messy logic that
currently tries to keep the OverloadedFunctionDecl in sync with the
declarations that are being added. Second, it will (eventually)
eliminate the need to allocate memory for overload sets, which could
help performance. Finally, it helps set us up for when name lookup can
return multiple (possibly ambiguous) results, as can happen with
lookup of class members in C++.
Next steps: make the IdentifierResolver store overloads as separate
entries in its list rather than replacing them with an
OverloadedFunctionDecl now, then see how far we can go toward
eliminating OverloadedFunctionDecl entirely.
llvm-svn: 61357
become useful or correct until we (1) parse template arguments
correctly, (2) have some way to turn template-ids into types,
declarators, etc., and (3) have a real representation of templates.
llvm-svn: 61208
array size declarators. No need to go through all the trouble
of parsing crazy things like [static const 4] when most code
doesn't need it.
llvm-svn: 61200
warning: statement was disambiguated as declaration
because it is currently firing in cases where the declaration would
not actually parse as a statement. We'd love to bring this warning
back if we can make it more accurate.
llvm-svn: 61137
is completely defined (C++ [class.mem]p2).
Reverse the order in which we process the definitions of member
functions specified inline. This way, we'll get diagnostics in the
order in which the member functions were declared in the class.
llvm-svn: 61103
Substantially improve error recovery after broken if conditions by
parsing the full if when we have a semantic error instead of using
parser recovery techniques to recover from a semantic error.
This fixes rdar://6094870 - spurious error after invalid 'if' condition
llvm-svn: 60929
and separates lexical name lookup from qualified name lookup. In
particular:
* Make DeclContext the central data structure for storing and
looking up declarations within existing declarations, e.g., members
of structs/unions/classes, enumerators in C++0x enums, members of
C++ namespaces, and (later) members of Objective-C
interfaces/implementations. DeclContext uses a lazily-constructed
data structure optimized for fast lookup (array for small contexts,
hash table for larger contexts).
* Implement C++ qualified name lookup in terms of lookup into
DeclContext.
* Implement C++ unqualified name lookup in terms of
qualified+unqualified name lookup (since unqualified lookup is not
purely lexical in C++!)
* Limit the use of the chains of declarations stored in
IdentifierInfo to those names declared lexically.
* Eliminate CXXFieldDecl, collapsing its behavior into
FieldDecl. (FieldDecl is now a ScopedDecl).
* Make RecordDecl into a DeclContext and eliminates its
Members/NumMembers fields (since one can just iterate through the
DeclContext to get the fields).
llvm-svn: 60878
"else" clause, e.g.,
if (int X = foo()) {
} else {
if (X) { // warning: X is always zero in this context
}
}
Fixes rdar://6425550 and lets me think about something other than
DeclContext.
llvm-svn: 60858
template<typename T> void f(T x) {
g(x); // g is a dependent name, so don't even bother to look it up
g(); // error: g is not a dependent name
}
Note that when we see "g(", we build a CXXDependentNameExpr. However,
if none of the call arguments are type-dependent, we will force the
resolution of the name "g" and replace the CXXDependentNameExpr with
its result.
GCC actually produces a nice error message when you make this
mistake, and even offers to compile your code with -fpermissive. I'll
do the former next, but I don't plan to do the latter.
llvm-svn: 60618
parameters, with some semantic analysis:
- Template parameters are introduced into template parameter scope
- Complain about template parameter shadowing (except in Microsoft mode)
Note that we leak template parameter declarations like crazy, a
problem we'll remedy once we actually create proper declarations for
templates.
Next up: dependent types and value-dependent/type-dependent
expressions.
llvm-svn: 60597
- Template parameter scope to hold the template parameters
- Template parameter context for parsing declarators
- Actions for template type parameters and non-type template
parameters
llvm-svn: 60387
-Change Parser::ParseCXXScopeSpecifier to MaybeParseCXXScopeSpecifier
-Remove Parser::isTokenCXXScopeSpecifier and fold it into MaybeParseCXXScopeSpecifier
-Rename Parser::TryAnnotateScopeToken to TryAnnotateCXXScopeToken and only allow it to be called when in C++
llvm-svn: 60117
Implemented anonymous category (also know as continuation class)
used to override main class's property attribute. This is work in
propgress.
llvm-svn: 60114
with implicit quotes around them. This has a bunch of follow-on
effects and requires tweaking to a whole lot of code. This causes
a regression in two tests (xfailed) by causing it to emit things like:
Line 10: duplicate interface declaration for category 'MyClass1' ('Category1')
instead of:
Line 10: duplicate interface declaration for category 'MyClass1(Category1)'
I will fix this in a follow-up commit.
As part of this, I had to start switching stuff to use ->getDeclName() instead
of Decl::getName() for consistency. This is good, but I was planning to do this
as an independent patch. There will be several follow-on patches
to clean up some of the mess, but this patch is already too big.
llvm-svn: 59917
one for building up the diagnostic that is in flight (DiagnosticBuilder)
and one for pulling structured information out of the diagnostic when
formatting and presenting it.
There is no functionality change with this patch.
llvm-svn: 59849
built-in operator candidates. Test overloading of '&' and ','.
In C++, a comma expression is an lvalue if its right-hand
subexpression is an lvalue. Update Expr::isLvalue accordingly.
llvm-svn: 59643
The core fix in Sema::ActOnClassMessage(). All the other changes have to do with passing down the SourceLocation for the receiver (to properly position the cursor when producing an error diagnostic).
llvm-svn: 59639
post-decrement, including support for generating all of the built-in
operator candidates for these operators.
C++ and C have different rules for the arguments to the builtin unary
'+' and '-'. Implemented both variants in Sema::ActOnUnaryOp.
In C++, pre-increment and pre-decrement return lvalues. Update
Expr::isLvalue accordingly.
llvm-svn: 59638
With this snippet:
void f(a::b);
An assert is hit:
Assertion failed: CachedTokens[CachedLexPos-1].getLocation() == Tok.getAnnotationEndLoc() && "The annotation should be until the most recent cached token", file ..\..\lib\Lex\PPCaching.cpp, line 98
Introduce Preprocessor::RevertCachedTokens that reverts a specific number of tokens when backtracking is enabled.
llvm-svn: 59636
operator+, directly, using the same mechanism as all other special
names.
Removed the "special" identifiers for the overloaded operators from
the identifier table and IdentifierInfo data structure. IdentifierInfo
is back to representing only real identifiers.
Added a new Action, ActOnOperatorFunctionIdExpr, that builds an
expression from an parsed operator-function-id (e.g., "operator
+"). ActOnIdentifierExpr used to do this job, but
operator-function-ids are no longer represented by IdentifierInfo's.
Extended Declarator to store overloaded operator names.
Sema::GetNameForDeclarator now knows how to turn the operator
name into a DeclarationName for the overloaded operator.
Except for (perhaps) consolidating the functionality of
ActOnIdentifier, ActOnOperatorFunctionIdExpr, and
ActOnConversionFunctionExpr into a common routine that builds an
appropriate DeclRefExpr by looking up a DeclarationName, all of the
work on normalizing declaration names should be complete with this
commit.
llvm-svn: 59526
and let the clients push whatever they want into the DiagnosticInfo
instead of hard coding a few forms. Also switch various clients to
use Diag(Tok, ...) instead of Diag(Tok.getLocation(), ...) as the
canonical form to simplify the code a bit.
llvm-svn: 59509
are formed. In particular, a diagnostic with all its strings and ranges is now
packaged up and sent to DiagnosticClients as a DiagnosticInfo instead of as a
ton of random stuff. This has the benefit of simplifying the interface, making
it more extensible, and allowing us to do more checking for things like access
past the end of the various arrays passed in.
In addition to introducing DiagnosticInfo, this also substantially changes how
Diagnostic::Report works. Instead of being passed in all of the info required
to issue a diagnostic, Report now takes only the required info (a location and
ID) and returns a fresh DiagnosticInfo *by value*. The caller is then free to
stuff strings and ranges into the DiagnosticInfo with the << operator. When
the dtor runs on the DiagnosticInfo object (which should happen at the end of
the statement), the diagnostic is actually emitted with all of the accumulated
information. This is a somewhat tricky dance, but it means that the
accumulated DiagnosticInfo is allowed to keep pointers to other expression
temporaries without those pointers getting invalidated.
This is just the minimal change to get this stuff working, but this will allow
us to eliminate the zillions of variant "Diag" methods scattered throughout
(e.g.) sema. For example, instead of calling:
Diag(BuiltinLoc, diag::err_overload_no_match, typeNames,
SourceRange(BuiltinLoc, RParenLoc));
We will soon be able to just do:
Diag(BuiltinLoc, diag::err_overload_no_match)
<< typeNames << SourceRange(BuiltinLoc, RParenLoc));
This scales better to support arbitrary types being passed in (not just
strings) in a type-safe way. Go operator overloading?!
llvm-svn: 59502
strings instead of array of strings. This reduces string copying
in some not-very-important cases, but paves the way for future
improvements.
llvm-svn: 59494
destructors, and conversion functions. The placeholders were used to
work around the fact that the parser and some of Sema really wanted
declarators to have simple identifiers; now, the code that deals with
declarators will use DeclarationNames.
llvm-svn: 59469
C++ constructors, destructors, and conversion functions now have a
FETokenInfo field that IdentifierResolver can access, so that these
special names are handled just like ordinary identifiers. A few other
Sema routines now use DeclarationNames instead of IdentifierInfo*'s.
To validate this design, this code also implements parsing and
semantic analysis for id-expressions that name conversion functions,
e.g.,
return operator bool();
The new parser action ActOnConversionFunctionExpr takes the result of
parsing "operator type-id" and turning it into an expression, using
the IdentifierResolver with the DeclarationName of the conversion
function. ActOnDeclarator pushes those conversion function names into
scope so that the IdentifierResolver can find them, of course.
llvm-svn: 59462
1. In the top level of ParseStatementOrDeclaration, don't eat a } if we
just parsed a statement if it list there. Also, don't even bother
emitting an error about a missing semicolon if the statement had a
bug (an rbrace is fine).
2. In do/while parsing, don't require a 'while' to be present if the do
body didn't parse.
This allows us to generate a clean diagnostic for this code:
t.c:1:22: error: expected expression
void foo (void) { do . while (0); }
^
Thanks to Neil for pointing this out.
llvm-svn: 59256
conversion functions. Instead, we just use a placeholder identifier
for these (e.g., "<constructor>") and override NamedDecl::getName() to
provide a human-readable name.
This is one potential solution to the problem; another solution would
be to replace the use of IdentifierInfo* in NamedDecl with a different
class that deals with identifiers better. I'm also prototyping that to
see how it compares, but this commit is better than what we had
previously.
llvm-svn: 59193
-When parsing declarators, don't depend on "CurScope->isCXXClassScope() == true" for constructors/destructors
-For C++ member declarations, don't depend on "Declarator.getContext() == Declarator::MemberContext"
llvm-svn: 58866
functions in C++, e.g.,
struct X {
operator bool() const;
};
Note that these conversions don't actually do anything, since we don't
yet have the ability to use them for implicit or explicit conversions.
llvm-svn: 58860
operators. For example, one can now write "x + y" where x or y is a
class or enumeration type, and Clang will perform overload resolution
for "+" based on the overloaded operators it finds.
The other kinds of overloadable operators in C++ will follow this same
approach.
Three major issues remain:
1) We don't find member operators
2) Since we don't have user-defined conversion operators, we can't
call any of the built-in overloaded operators in C++ [over.built].
3) Once we've done the semantic checks, we drop the overloaded
operator on the floor; it doesn't get into the AST at all.
llvm-svn: 58821
operators in C++. Overloaded operators can be called directly via
their operator-function-ids, e.g., "operator+(foo, bar)", but we don't
yet implement the semantics of operator overloading to handle, e.g.,
"foo + bar".
llvm-svn: 58817
Implicit declaration of destructors (when necessary).
Extended Declarator to store information about parsed constructors
and destructors; this will be extended to deal with declarators that
name overloaded operators (e.g., "operator +") and user-defined
conversion operators (e.g., "operator int").
llvm-svn: 58767
reference-collapsing.
Implement diagnostic for formation of a reference to cv void.
Drop cv-qualifiers added to a reference type when the reference type
comes from a typedef.
llvm-svn: 58612
etc more generic. For some targets, long may not be equal to pointer size. For
example: PIC16 has int as i16, ptr as i16 but long as i32.
Also fixed a few build warnings in assert() functions in CFRefCount.cpp,
CGDecl.cpp, SemaDeclCXX.cpp and ParseDeclCXX.cpp.
llvm-svn: 58501
Notes:
- Constructors are never found by name lookup, so they'll never get
pushed into any scope. Instead, they are stored as an
OverloadedFunctionDecl in CXXRecordDecl for easy overloading.
- There's a new action isCurrentClassName that determines whether an
identifier is the name of the innermost class currently being defined;
we use this to identify the declarator-id grammar rule that refers to
a type-name.
- MinimalAction does *not* support parsing constructors.
- We now handle virtual and explicit function specifiers.
llvm-svn: 58499
were being treated as type names for non-Objective-C files.
- Other lines are just because MinimalAction didn't have access to
the LangOptions.
llvm-svn: 58498
of copy initialization. Other pieces of the puzzle:
- Try/Perform-ImplicitConversion now handles implicit conversions
that don't involve references.
- Try/Perform-CopyInitialization uses
CheckSingleAssignmentConstraints for C. PerformCopyInitialization
is now used for all argument passing and returning values from a
function.
- Diagnose errors with declaring references and const values without
an initializer. (Uses a new Action callback, ActOnUninitializedDecl).
We do not yet have implicit conversion sequences for reference
binding, which means that we don't have any overloading support for
reference parameters yet.
llvm-svn: 58353
- CastExpr is the root of all casts
- ImplicitCastExpr is (still) used for all explicit casts
- ExplicitCastExpr is now the root of all *explicit* casts
- ExplicitCCastExpr (new name needed!?) is a C-style cast in C or C++
- CXXFunctionalCastExpr inherits from ExplicitCastExpr
- CXXNamedCastExpr inherits from ExplicitCastExpr and is the root of all
of the C++ named cast expression types (static_cast, dynamic_cast, etc.)
- Added classes CXXStaticCastExpr, CXXDynamicCastExpr,
CXXReinterpretCastExpr, and CXXConstCastExpr to
Also, fixed returned-stack-addr.cpp, which broke once when we fixed
reinterpret_cast to diagnose double->int* conversions and again when
we eliminated implicit conversions to reference types. The fix is in
both testcase and SemaChecking.cpp.
Most of this patch is simply support for the renaming. There's very
little actual change in semantics.
llvm-svn: 58264
struct foo Y[10] = {
[4] .arr [2] 4 // expected-error {{expected '=' or another designator}}
};
because the "missing equals" extension only is valid if there
is exactly one array designator.
llvm-svn: 58215
1. It introduces new parser level abstractions for designators
that are used to communicate between parser and sema.
2. This fixes a FIXME where "identifier ':'" was considered to be
a designator even if it wasn't the first in a designator list.
3. In the "identifier ':'" case, it actually builds the designator
representation.
llvm-svn: 58205
1. Remove a bogus assertion, clients other than sema can return a
null pointer from actions that result in ParseTypeName returning null.
2. Remove dead RParenLoc variable.
3. Simplify control flow handling error conditions.
4. On a major failure, we should skip until ')' not until '}'.
llvm-svn: 57949
crash.m:8:12: error: type name requires a specifier or qualifier
@property (readonlyx, getter=isAwesome) int _awesome;
^
crash.m:8:29: error: expected ';' at end of declaration list
@property (readonlyx, getter=isAwesome) int _awesome;
^
crash.m:8:39: error: expected identifier or '('
@property (readonlyx, getter=isAwesome) int _awesome;
^
we now get:
crash.m:8:12: error: unknown property attribute 'readonlyx'
@property (readonlyx, getter=isAwesome) int _awesome;
^
Also, we can eliminate isObjCPropertyAttribute now.
llvm-svn: 57811
ParseObjCPropertyAttribute. Before, on this code (where
a comma was forgotten):
@property (readonly getter=isAwesome) int _awesome;
we emitted:
crash.m:9:11: error: expected ')'
@property (readonly getter=isAwesome) int _awesome;
^
crash.m:9:37: error: type name requires a specifier or qualifier
@property (readonly getter=isAwesome) int _awesome;
^
crash.m:9:37: error: expected identifier or '('
crash.m:9:37: error: expected ';' at end of declaration list
crash.m:9:1: error: @property requires fields to be named
@property (readonly getter=isAwesome) int _awesome;
^
now we emit:
crash.m:9:21: error: expected ')'
@property (readonly getter=isAwesome) int _awesome;
^
crash.m:9:11: error: to match this '('
@property (readonly getter=isAwesome) int _awesome;
^
llvm-svn: 57809
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
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
This is how this kind of initializers appear in the AST:
-The Init expression of the VarDecl is a functional type construction (of the VarDecl's type).
-The new VarDecl::hasCXXDirectInitializer() returns true.
e.g, for "int x(1);":
-VarDecl 'x' has Init with expression "int(1)" (CXXFunctionalCastExpr).
-hasCXXDirectInitializer() of VarDecl 'x' returns true.
A major benefit is that clients that don't particularly care about which exactly form was the initializer can handle both cases without special case code.
Note that codegening works now for "int x(1);" without any changes to CodeGen.
llvm-svn: 57178
This was the motivation of the following changes:
-'TentativeParsingResult' enum is replaced by a 'TPResult' class that basically encapsulates the enum.
-TPR_true, TPR_false, TPR_ambiguous, and TPR_error enum constants are replaced by TPResult::True(), TPResult::False(), etc. calls that return a TPResult object.
-Also fixed the subtle bug in Parser::isCXXFunctionDeclarator (caught by the above changes as a compilation error).
llvm-svn: 57125
'ParseTentative.cpp' implements the functionality needed to resolve ambiguous C++ statements, to either a declaration or an expression, by "tentatively parsing" them.
llvm-svn: 57084
Douglas Gregor