"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.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@103718 91177308-0d34-0410-b5e6-96231b3b80d8
potentially-evaluated expression context, to ensure that used
declarations get properly marked. Fixes PR7123.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@103624 91177308-0d34-0410-b5e6-96231b3b80d8
referenced unless we see one of them defined (or the key function
defined, if it as one) or if we need the vtable for something. Fixes
PR7114.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@103497 91177308-0d34-0410-b5e6-96231b3b80d8
explicit instantiations of template. C++0x clarifies the intent
(they're ill-formed in some cases; see [temp.explicit] for
details). However, one could squint at the C++98/03 standard and
conclude they are permitted, so reduce the error to a warning
(controlled by -Wc++0x-compat) in C++98/03 mode.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@103482 91177308-0d34-0410-b5e6-96231b3b80d8
value-dependent if their initializers are value-dependent; my recent
tweak to these dependent rules overstepped by taking away this
value-dependents. Fixes a Boost.GIL regression.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@103476 91177308-0d34-0410-b5e6-96231b3b80d8
of the current instantiation is value-dependent. The C++ standard
fails to enumerate this case and, therefore, we missed it. Chandler
did all of the hard work of reducing the last remaining
Boost.PtrContainer failure (which had to do with static initialization
in the Serialization library) down to this simple little test.
While I'm at it, clean up the dependence rules for template arguments
that are declarations, and implement the dependence rules for template
argument packs.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@103464 91177308-0d34-0410-b5e6-96231b3b80d8
particular, don't complain about unused variables that have dependent
type until instantiation time, so that we can look at the type of the
variable. Moreover, only complain about unused variables that have
neither a user-declared constructor nor a non-trivial destructor.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@103362 91177308-0d34-0410-b5e6-96231b3b80d8
for, and switch), be careful to construct the full expressions as soon
as we perform template instantation, so we don't either forget to call
temporary destructors or destroy temporaries at the wrong time. This
is the template-instantiation analogue to r103187, during which I
hadn't realized that the issue would affect the handling of these
constructs differently inside and outside of templates.
Fixes a regression in Boost.Function.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@103357 91177308-0d34-0410-b5e6-96231b3b80d8
conflicting deduced template argument values, give a more specific
reason along with those values, e.g.,
test/SemaTemplate/overload-candidates.cpp:4:10: note: candidate template
ignored: deduced conflicting types for parameter 'T' ('int' vs. 'long')
const T& min(const T&, const T&);
^
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@103339 91177308-0d34-0410-b5e6-96231b3b80d8
ensure that we complete the type when we need to look at constructors
during reference binding.
When determining whether the two types involved in reference binding
are reference-compatible, reference-related, etc., do not complete the
type of the reference itself because it is not necessary to determine
well-formedness of the program. Complete the type that we are binding
to, since that can affect whether we know about a derived-to-base
conversion.
Re-fixes PR7080.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@103283 91177308-0d34-0410-b5e6-96231b3b80d8
are reference-compatible, reference-related, etc., do not complete the
type of the reference itself because it is not necessary to determine
well-formedness of the program. Complete the type that we are binding
to, since that can affect whether we know about a derived-to-base
conversion.
Fixes PR7080.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@103220 91177308-0d34-0410-b5e6-96231b3b80d8
different tag kind ("struct" vs. "class") than the primary template,
which has an affect on access control.
Should fix the last remaining Boost.Accumulors failure.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@103144 91177308-0d34-0410-b5e6-96231b3b80d8
ParseOptionalCXXScopeSpecifier() only annotates the subset of
template-ids which are not subject to lexical ambiguity. Add support
for the more general case in ParseUnqualifiedId() to handle cases
such as A::template B().
Also improve some diagnostic locations.
Fixes PR7030, from Alp Toker!
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@103081 91177308-0d34-0410-b5e6-96231b3b80d8
typedef int functype(int, int);
functype func;
also instantiate the synthesized function parameters for the resulting
function declaration.
With this change, Boost.Wave builds and passes all of its regression
tests.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@103025 91177308-0d34-0410-b5e6-96231b3b80d8
friend function template, be sure to adjust the computed template
argument lists based on the location of the definition of the function
template: it's possible that the definition we're instantiating with
and the template declaration that we found when creating the
specialization are in different contexts, which meant that we would
end up using the wrong template arguments for instantiation.
Fixes PR7013; all Boost.DynamicBitset tests now pass.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@102974 91177308-0d34-0410-b5e6-96231b3b80d8
mapping from the declaration in the template to the instantiated
declaration before transforming the initializer, in case some crazy
lunatic decides to use a variable in its own initializer. Fixes PR7016.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@102945 91177308-0d34-0410-b5e6-96231b3b80d8
to enter the instantiated parameter declarations into the local
instantiation scope; they can't be referenced anyway. Fixes PR7022.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@102914 91177308-0d34-0410-b5e6-96231b3b80d8
assignment operators.
Previously, Sema provided type-checking and template instantiation for
copy assignment operators, then CodeGen would synthesize the actual
body of the copy constructor. Unfortunately, the two were not in sync,
and CodeGen might pick a copy-assignment operator that is different
from what Sema chose, leading to strange failures, e.g., link-time
failures when CodeGen called a copy-assignment operator that was not
instantiation, run-time failures when copy-assignment operators were
overloaded for const/non-const references and the wrong one was
picked, and run-time failures when by-value copy-assignment operators
did not have their arguments properly copy-initialized.
This implementation synthesizes the implicitly-defined copy assignment
operator bodies in Sema, so that the resulting ASTs encode exactly
what CodeGen needs to do; there is no longer any special code in
CodeGen to synthesize copy-assignment operators. The synthesis of the
body is relatively simple, and we generate one of three different
kinds of copy statements for each base or member:
- For a class subobject, call the appropriate copy-assignment
operator, after overload resolution has determined what that is.
- For an array of scalar types or an array of class types that have
trivial copy assignment operators, construct a call to
__builtin_memcpy.
- For an array of class types with non-trivial copy assignment
operators, synthesize a (possibly nested!) for loop whose inner
statement calls the copy constructor.
- For a scalar type, use built-in assignment.
This patch fixes at least a few tests cases in Boost.Spirit that were
failing because CodeGen picked the wrong copy-assignment operator
(leading to link-time failures), and I suspect a number of undiagnosed
problems will also go away with this change.
Some of the diagnostics we had previously have gotten worse with this
change, since we're going through generic code for our
type-checking. I will improve this in a subsequent patch.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@102853 91177308-0d34-0410-b5e6-96231b3b80d8
parameter with pointer-to-member type, we may have to perform a
qualification conversion, since the pointee type of the parameter
might be more qualified than the pointee type of the argument we form
from the declaration. Fixes PR6986.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@102777 91177308-0d34-0410-b5e6-96231b3b80d8
of the mapping from local declarations to their instantiated
counterparts during template instantiation. Previously, we tried to do
some unholy merging of local instantiation scopes that involved
storing a single hash table along with an "undo" list on the
side... which was ugly, and never handled function parameters
properly.
Now, we just keep separate hash tables for each local instantiation
scope, and "combining" two scopes means that we'll look in each of the
combined hash tables. The combined scope stack is rarely deep, and
this makes it easy to avoid the "undo" issues we were hitting. Also,
I've simplified the logic for function parameters: if we're declaring
a function and we need the function parameters to live longer, we just
push them back into the local instantiation scope where we need them.
Fixes PR6990.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@102732 91177308-0d34-0410-b5e6-96231b3b80d8
InjectedClassNameType's Decl to point at the definition. It's a little
messy, but we do the same thing with classes and their record types,
since much of Clang expects that the TagDecl* one gets out of a type
is the definition. Fixes several Boost.Proto failures.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@102691 91177308-0d34-0410-b5e6-96231b3b80d8
entering the current instantiation. Set up a little to preserve type location
information for typename types while we're in there.
Fixes a Boost failure.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@102673 91177308-0d34-0410-b5e6-96231b3b80d8
bindings when the template argument is still an expression; it happens
while checking the template arguments of a class template partial
specializations. Fixes PR6964.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@102595 91177308-0d34-0410-b5e6-96231b3b80d8
Amadini.
This change introduces a new expression node type, OffsetOfExpr, that
describes __builtin_offsetof. Previously, __builtin_offsetof was
implemented using a unary operator whose subexpression involved
various synthesized array-subscript and member-reference expressions,
which was ugly and made it very hard to instantiate as a
template. OffsetOfExpr represents the AST more faithfully, with proper
type source information and a more compact representation.
OffsetOfExpr also has support for dependent __builtin_offsetof
expressions; it can be value-dependent, but will never be
type-dependent (like sizeof or alignof). This commit introduces
template instantiation for __builtin_offsetof as well.
There are two major caveats to this patch:
1) CodeGen cannot handle the case where __builtin_offsetof is not a
constant expression, so it produces an error. So, to avoid
regressing in C, we retain the old UnaryOperator-based
__builtin_offsetof implementation in C while using the shiny new
OffsetOfExpr implementation in C++. The old implementation can go
away once we have proper CodeGen support for this case, which we
expect won't cause much trouble in C++.
2) __builtin_offsetof doesn't work well with non-POD class types,
particularly when the designated field is found within a base
class. I will address this in a subsequent patch.
Fixes PR5880 and a bunch of assertions when building Boost.Python
tests.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@102542 91177308-0d34-0410-b5e6-96231b3b80d8
complete, return an error rather than falling back to building a
dependent declaration reference, since we might not be in a dependent
context. Fixes a fiendish crash-on-invalid in Boost.FunctionTypes that
I wasn't able to reduce to anything useful.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@102491 91177308-0d34-0410-b5e6-96231b3b80d8
template argument deduction, use the lexical declaration context as
the owner for friend function templates. Fixes 2 failures in
Boost.Graph.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@102489 91177308-0d34-0410-b5e6-96231b3b80d8
of a class template or class template partial specialization. That is to
say, in
template <class T> class A { ... };
or
template <class T> class B<const T*> { ... };
make 'A<T>' and 'B<const T*>' sugar for the corresponding InjectedClassNameType
when written inside the appropriate context. This allows us to track the
current instantiation appropriately even inside AST routines. It also allows
us to compute a DeclContext for a type much more efficiently, at some extra
cost every time we write a template specialization (which can be optimized,
but I've left it simple in this patch).
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@102407 91177308-0d34-0410-b5e6-96231b3b80d8
that the type we're copying is complete.
Boost.Regex now builds, although it's failing its regression tests
with our favorite "Sema doesn't consider destructor as used."
assertion.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@102271 91177308-0d34-0410-b5e6-96231b3b80d8
template parameter, by sure to mark that declaration as
"referenced". The Boost.Iterator library now passes all tests.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@102256 91177308-0d34-0410-b5e6-96231b3b80d8
when they are not complete (since we could not match them up to
anything) and ensuring that enum parsing can cope with dependent
elaborated-type-specifiers. Fixes PR6915 and PR6649.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@102247 91177308-0d34-0410-b5e6-96231b3b80d8
declaration that this typedef gives the tag a name. Fixes a problem
uncovered by Boost.GIL (Generic Image Library).
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@102180 91177308-0d34-0410-b5e6-96231b3b80d8
method parameter, provide a note pointing at the parameter itself so
the user does not have to manually look for the function/method being
called and match up parameters to arguments. For example, we now get:
t.c:4:5: warning: incompatible pointer types passing 'long *' to
parameter of
type 'int *' [-pedantic]
f(long_ptr);
^~~~~~~~
t.c:1:13: note: passing argument to parameter 'x' here
void f(int *x);
^
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@102038 91177308-0d34-0410-b5e6-96231b3b80d8
we will print with each error that occurs during template
instantiation. When the backtrace is longer than that, we will print
N/2 of the innermost backtrace entries and N/2 of the outermost
backtrace entries, then skip the middle entries with a note such as:
note: suppressed 2 template instantiation contexts; use
-ftemplate-backtrace-limit=N to change the number of template
instantiation entries shown
This should eliminate some excessively long backtraces that aren't
providing any value.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@101882 91177308-0d34-0410-b5e6-96231b3b80d8
resolution. There are two sources of problems involving user-defined
conversions that this change eliminates, along with providing simpler
interfaces for checking implicit conversions:
- It eliminates a case of infinite recursion found in Boost.
- It eliminates the search for the constructor needed to copy a temporary
generated by an implicit conversion from overload
resolution. Overload resolution assumes that, if it gets a value
of the parameter's class type (or a derived class thereof), there
is a way to copy if... even if there isn't. We now model this
properly.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@101680 91177308-0d34-0410-b5e6-96231b3b80d8
Douglas Gregor
Daniel Dunbar
Chandler Carruth
John McCall