2009-12-16 04:14:24 +08:00
|
|
|
// RUN: %clang_cc1 -fsyntax-only -verify %s
|
2009-12-07 16:29:39 +08:00
|
|
|
|
|
|
|
namespace PR5557 {
|
|
|
|
template <class T> struct A {
|
|
|
|
A();
|
Rework when and how vtables are emitted, by tracking where vtables are
"used" (e.g., we will refer to the vtable in the generated code) and
when they are defined (i.e., because we've seen the key function
definition). Previously, we were effectively tracking "potential
definitions" rather than uses, so we were a bit too eager about emitting
vtables for classes without key functions.
The new scheme:
- For every use of a vtable, Sema calls MarkVTableUsed() to indicate
the use. For example, this occurs when calling a virtual member
function of the class, defining a constructor of that class type,
dynamic_cast'ing from that type to a derived class, casting
to/through a virtual base class, etc.
- For every definition of a vtable, Sema calls MarkVTableUsed() to
indicate the definition. This happens at the end of the translation
unit for classes whose key function has been defined (so we can
delay computation of the key function; see PR6564), and will also
occur with explicit template instantiation definitions.
- For every vtable defined/used, we mark all of the virtual member
functions of that vtable as defined/used, unless we know that the key
function is in another translation unit. This instantiates virtual
member functions when needed.
- At the end of the translation unit, Sema tells CodeGen (via the
ASTConsumer) which vtables must be defined (CodeGen will define
them) and which may be used (for which CodeGen will define the
vtables lazily).
From a language perspective, both the old and the new schemes are
permissible: we're allowed to instantiate virtual member functions
whenever we want per the standard. However, all other C++ compilers
were more lazy than we were, and our eagerness was both a performance
issue (we instantiated too much) and a portability problem (we broke
Boost test cases, which now pass).
Notes:
(1) There's a ton of churn in the tests, because the order in which
vtables get emitted to IR has changed. I've tried to isolate some of
the larger tests from these issues.
(2) Some diagnostics related to
implicitly-instantiated/implicitly-defined virtual member functions
have moved to the point of first use/definition. It's better this
way.
(3) I could use a review of the places where we MarkVTableUsed, to
see if I missed any place where the language effectively requires a
vtable.
Fixes PR7114 and PR6564.
llvm-svn: 103718
2010-05-14 00:44:06 +08:00
|
|
|
virtual void anchor();
|
2009-12-07 16:29:39 +08:00
|
|
|
virtual int a(T x);
|
|
|
|
};
|
|
|
|
template<class T> A<T>::A() {}
|
2010-01-06 12:44:19 +08:00
|
|
|
template<class T> void A<T>::anchor() { }
|
|
|
|
|
2009-12-07 16:29:39 +08:00
|
|
|
template<class T> int A<T>::a(T x) {
|
|
|
|
return *x; // expected-error{{requires pointer operand}}
|
|
|
|
}
|
|
|
|
|
2010-01-06 12:44:19 +08:00
|
|
|
void f(A<int> x) {
|
Rework when and how vtables are emitted, by tracking where vtables are
"used" (e.g., we will refer to the vtable in the generated code) and
when they are defined (i.e., because we've seen the key function
definition). Previously, we were effectively tracking "potential
definitions" rather than uses, so we were a bit too eager about emitting
vtables for classes without key functions.
The new scheme:
- For every use of a vtable, Sema calls MarkVTableUsed() to indicate
the use. For example, this occurs when calling a virtual member
function of the class, defining a constructor of that class type,
dynamic_cast'ing from that type to a derived class, casting
to/through a virtual base class, etc.
- For every definition of a vtable, Sema calls MarkVTableUsed() to
indicate the definition. This happens at the end of the translation
unit for classes whose key function has been defined (so we can
delay computation of the key function; see PR6564), and will also
occur with explicit template instantiation definitions.
- For every vtable defined/used, we mark all of the virtual member
functions of that vtable as defined/used, unless we know that the key
function is in another translation unit. This instantiates virtual
member functions when needed.
- At the end of the translation unit, Sema tells CodeGen (via the
ASTConsumer) which vtables must be defined (CodeGen will define
them) and which may be used (for which CodeGen will define the
vtables lazily).
From a language perspective, both the old and the new schemes are
permissible: we're allowed to instantiate virtual member functions
whenever we want per the standard. However, all other C++ compilers
were more lazy than we were, and our eagerness was both a performance
issue (we instantiated too much) and a portability problem (we broke
Boost test cases, which now pass).
Notes:
(1) There's a ton of churn in the tests, because the order in which
vtables get emitted to IR has changed. I've tried to isolate some of
the larger tests from these issues.
(2) Some diagnostics related to
implicitly-instantiated/implicitly-defined virtual member functions
have moved to the point of first use/definition. It's better this
way.
(3) I could use a review of the places where we MarkVTableUsed, to
see if I missed any place where the language effectively requires a
vtable.
Fixes PR7114 and PR6564.
llvm-svn: 103718
2010-05-14 00:44:06 +08:00
|
|
|
x.anchor(); // expected-note{{instantiation}}
|
2010-01-06 12:44:19 +08:00
|
|
|
}
|
2009-12-07 16:29:39 +08:00
|
|
|
|
|
|
|
template<typename T>
|
|
|
|
struct X {
|
|
|
|
virtual void f();
|
|
|
|
};
|
|
|
|
|
|
|
|
template<>
|
|
|
|
void X<int>::f() { }
|
|
|
|
}
|
2010-01-06 12:44:19 +08:00
|
|
|
|
|
|
|
template<typename T>
|
|
|
|
struct Base {
|
|
|
|
virtual ~Base() {
|
|
|
|
int *ptr = 0;
|
|
|
|
T t = ptr; // expected-error{{cannot initialize}}
|
|
|
|
}
|
|
|
|
};
|
|
|
|
|
|
|
|
template<typename T>
|
|
|
|
struct Derived : Base<T> {
|
2010-03-16 13:36:30 +08:00
|
|
|
virtual void foo() { }
|
2010-01-06 12:44:19 +08:00
|
|
|
};
|
|
|
|
|
2010-03-16 13:36:30 +08:00
|
|
|
template struct Derived<int>; // expected-note {{in instantiation of member function 'Base<int>::~Base' requested here}}
|
2010-01-06 12:44:19 +08:00
|
|
|
|
2010-01-07 04:27:16 +08:00
|
|
|
template<typename T>
|
|
|
|
struct HasOutOfLineKey {
|
Rework when and how vtables are emitted, by tracking where vtables are
"used" (e.g., we will refer to the vtable in the generated code) and
when they are defined (i.e., because we've seen the key function
definition). Previously, we were effectively tracking "potential
definitions" rather than uses, so we were a bit too eager about emitting
vtables for classes without key functions.
The new scheme:
- For every use of a vtable, Sema calls MarkVTableUsed() to indicate
the use. For example, this occurs when calling a virtual member
function of the class, defining a constructor of that class type,
dynamic_cast'ing from that type to a derived class, casting
to/through a virtual base class, etc.
- For every definition of a vtable, Sema calls MarkVTableUsed() to
indicate the definition. This happens at the end of the translation
unit for classes whose key function has been defined (so we can
delay computation of the key function; see PR6564), and will also
occur with explicit template instantiation definitions.
- For every vtable defined/used, we mark all of the virtual member
functions of that vtable as defined/used, unless we know that the key
function is in another translation unit. This instantiates virtual
member functions when needed.
- At the end of the translation unit, Sema tells CodeGen (via the
ASTConsumer) which vtables must be defined (CodeGen will define
them) and which may be used (for which CodeGen will define the
vtables lazily).
From a language perspective, both the old and the new schemes are
permissible: we're allowed to instantiate virtual member functions
whenever we want per the standard. However, all other C++ compilers
were more lazy than we were, and our eagerness was both a performance
issue (we instantiated too much) and a portability problem (we broke
Boost test cases, which now pass).
Notes:
(1) There's a ton of churn in the tests, because the order in which
vtables get emitted to IR has changed. I've tried to isolate some of
the larger tests from these issues.
(2) Some diagnostics related to
implicitly-instantiated/implicitly-defined virtual member functions
have moved to the point of first use/definition. It's better this
way.
(3) I could use a review of the places where we MarkVTableUsed, to
see if I missed any place where the language effectively requires a
vtable.
Fixes PR7114 and PR6564.
llvm-svn: 103718
2010-05-14 00:44:06 +08:00
|
|
|
HasOutOfLineKey() { }
|
2010-01-07 04:27:16 +08:00
|
|
|
virtual T *f(float *fp);
|
|
|
|
};
|
|
|
|
|
|
|
|
template<typename T>
|
|
|
|
T *HasOutOfLineKey<T>::f(float *fp) {
|
|
|
|
return fp; // expected-error{{cannot initialize return object of type 'int *' with an lvalue of type 'float *'}}
|
|
|
|
}
|
|
|
|
|
Rework when and how vtables are emitted, by tracking where vtables are
"used" (e.g., we will refer to the vtable in the generated code) and
when they are defined (i.e., because we've seen the key function
definition). Previously, we were effectively tracking "potential
definitions" rather than uses, so we were a bit too eager about emitting
vtables for classes without key functions.
The new scheme:
- For every use of a vtable, Sema calls MarkVTableUsed() to indicate
the use. For example, this occurs when calling a virtual member
function of the class, defining a constructor of that class type,
dynamic_cast'ing from that type to a derived class, casting
to/through a virtual base class, etc.
- For every definition of a vtable, Sema calls MarkVTableUsed() to
indicate the definition. This happens at the end of the translation
unit for classes whose key function has been defined (so we can
delay computation of the key function; see PR6564), and will also
occur with explicit template instantiation definitions.
- For every vtable defined/used, we mark all of the virtual member
functions of that vtable as defined/used, unless we know that the key
function is in another translation unit. This instantiates virtual
member functions when needed.
- At the end of the translation unit, Sema tells CodeGen (via the
ASTConsumer) which vtables must be defined (CodeGen will define
them) and which may be used (for which CodeGen will define the
vtables lazily).
From a language perspective, both the old and the new schemes are
permissible: we're allowed to instantiate virtual member functions
whenever we want per the standard. However, all other C++ compilers
were more lazy than we were, and our eagerness was both a performance
issue (we instantiated too much) and a portability problem (we broke
Boost test cases, which now pass).
Notes:
(1) There's a ton of churn in the tests, because the order in which
vtables get emitted to IR has changed. I've tried to isolate some of
the larger tests from these issues.
(2) Some diagnostics related to
implicitly-instantiated/implicitly-defined virtual member functions
have moved to the point of first use/definition. It's better this
way.
(3) I could use a review of the places where we MarkVTableUsed, to
see if I missed any place where the language effectively requires a
vtable.
Fixes PR7114 and PR6564.
llvm-svn: 103718
2010-05-14 00:44:06 +08:00
|
|
|
HasOutOfLineKey<int> out_of_line; // expected-note{{in instantiation of member function 'HasOutOfLineKey<int>::f' requested here}}
|
|
|
|
|
|
|
|
namespace std {
|
|
|
|
class type_info;
|
|
|
|
}
|
|
|
|
|
|
|
|
namespace PR7114 {
|
|
|
|
class A { virtual ~A(); }; // expected-note{{declared private here}}
|
|
|
|
|
|
|
|
template<typename T>
|
|
|
|
class B {
|
|
|
|
public:
|
|
|
|
class Inner : public A { }; // expected-error{{base class 'PR7114::A' has private destructor}}
|
|
|
|
static Inner i;
|
|
|
|
static const unsigned value = sizeof(i) == 4;
|
|
|
|
};
|
|
|
|
|
|
|
|
int f() { return B<int>::value; }
|
|
|
|
|
|
|
|
void test_typeid(B<float>::Inner bfi) {
|
|
|
|
(void)typeid(bfi); // expected-note{{implicit default destructor}}
|
|
|
|
}
|
|
|
|
|
|
|
|
template<typename T>
|
|
|
|
struct X : A {
|
|
|
|
void f() { }
|
|
|
|
};
|
|
|
|
|
|
|
|
void test_X(X<int> xi, X<float> xf) {
|
|
|
|
xi.f();
|
|
|
|
}
|
|
|
|
}
|