"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
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
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
(-Wunused-exception-parameter) than normal variables, since it's more
common to name and then ignore an exception parameter. This warning is
neither enabled by default nor by -Wall. Fixes <rdar://problem/7931045>.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@102931 91177308-0d34-0410-b5e6-96231b3b80d8
already knows what context it's looking in. Just pass that context in
instead of (questionably) recalculating it.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@102818 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
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
(e.g., no typename, enum, class, etc.), e.g., because the context is
one that is known to refer to a type. Patch from Enea Zaffanella!
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@102243 91177308-0d34-0410-b5e6-96231b3b80d8
arguments. Rather than having the parser call ActOnParamDeclarator
(which is a bit of a hack), call a new ActOnObjCExceptionDecl
action. We'll be moving more functionality into this handler to
perform earlier checking of @catch.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@102222 91177308-0d34-0410-b5e6-96231b3b80d8
way that C does. Among other differences, elaborated type specifiers
are defined to skip "non-types", which, as you might imagine, does not
include typedefs. Rework our use of IDNS masks to capture the semantics
of different kinds of declarations better, and remove most current lookup
filters. Removing the last remaining filter is more complicated and will
happen in a separate patch.
Fixes PR 6885 as well some spectrum of unfiled bugs.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@102164 91177308-0d34-0410-b5e6-96231b3b80d8
function declaration, since it may end up being changed (e.g.,
"extern" can become "static" if a prior declaration was static). Patch
by Enea Zaffanella and Paolo Bolzoni.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@101826 91177308-0d34-0410-b5e6-96231b3b80d8
in case it ends up doing something that might trigger diagnostics
(template instantiation, ambiguity reporting, access
reporting). Noticed while working on PR6831.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@101412 91177308-0d34-0410-b5e6-96231b3b80d8
ASTContext::getTypeSize() rather than ASTContext::getIntWidth() for
the width of an integral type. The former includes padding for bools
(to the target's size) while the latter does not, so we woud end up
zero-extending bools to the target width when we shouldn't. Fixes a
crash-on-valid in the included test.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@101372 91177308-0d34-0410-b5e6-96231b3b80d8
generally recover from typos in keywords (since we would effectively
have to mangle the token stream). However, there are still benefits to
typo-correcting with keywords:
- We don't make stupid suggestions when the user typed something
that is similar to a keyword.
- We can suggest the keyword in a diagnostic (did you mean
"static_cast"?), even if we can't recover and therefore don't have
a fix-it.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@101274 91177308-0d34-0410-b5e6-96231b3b80d8
are invalid. Prevents a crash-on-invalid during template instantiation.
I... really don't understand how this wasn't already present.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@101203 91177308-0d34-0410-b5e6-96231b3b80d8
friend declaration; this used to be important but is now just a waste of time
plus an unreasonable assertion. Fixes PR6174.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@101112 91177308-0d34-0410-b5e6-96231b3b80d8
ended up finding a function template that we didn't expect. Recover
more gracefully, and fix a similar issue for class templates.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@101040 91177308-0d34-0410-b5e6-96231b3b80d8
function's type is (strictly speaking) non-dependent. This ensures
that, e.g., default function arguments get instantiated properly.
And, since I couldn't resist, collapse the two implementations of
function-parameter instantiation into calls to a single, new function
(Sema::SubstParmVarDecl), since the two had nearly identical code (and
each had bugs the other didn't!). More importantly, factored out the
semantic analysis of a parameter declaration into
Sema::CheckParameter, which is called both by
Sema::ActOnParamDeclarator (when parameters are parsed) and when a
parameter is instantiated. Previously, we were missing some
Objective-C and address-space checks on instantiated function
parameters.
Fixes PR6733.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@101029 91177308-0d34-0410-b5e6-96231b3b80d8
warning. It's not harmful to have such pointless declarations, and GCC
does not diagnose this issue consistently.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@100814 91177308-0d34-0410-b5e6-96231b3b80d8
nested-name-specifier (e.g., "class T::foo") fails to find a tag
member in the scope nominated by the
nested-name-specifier. Previously, we gave a bland
error: 'Nested' does not name a tag member in the specified scope
which didn't actually say where we were looking, which was rather
horrible when the nested-name-specifier was instantiated. Now, we give
something a bit better:
error: no class named 'Nested' in 'NoDepBase<T>'
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@100060 91177308-0d34-0410-b5e6-96231b3b80d8
this was parsed as a typename-specifier, elaborated-type-specifier
(including the kind), or just a dependent qualified type name.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@100039 91177308-0d34-0410-b5e6-96231b3b80d8
This introduces FunctionType::ExtInfo to hold the calling convention and the
noreturn attribute. The next patch will extend it to include the regparm
attribute and fix the bug.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@99920 91177308-0d34-0410-b5e6-96231b3b80d8
doing all the cleanup tasks and checks. This gives us the proper context for
checking access to base and member destructors.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@99559 91177308-0d34-0410-b5e6-96231b3b80d8
since we have absolutely no way to match them when they are declared
nor do we have a way to represent these parsed-but-not-checked friend
declarations.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@99407 91177308-0d34-0410-b5e6-96231b3b80d8
template <> friend void foo(int);
we need to change it to
friend void foo<>(int);
or else the user won't get the template specialization they obviously want.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@99390 91177308-0d34-0410-b5e6-96231b3b80d8
entering a function or block definition, not on every single declaration.
Unfortunately we don't have previous-lookup results around when it's time
to make this decision, so we have to redo the lookup. The alternative is
to use delayed diagnostics.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@99172 91177308-0d34-0410-b5e6-96231b3b80d8
If a struct has an invalid field, mark it as invalid. Also avoid producing
errors about incomplete types that are invalid.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@99150 91177308-0d34-0410-b5e6-96231b3b80d8
This object controls when the warnings are executed, allowing the client code
in Sema to selectively disable warnings as needed.
Centralizing the logic for analysis-based warnings allows us to optimize
when and how they are run.
Along the way, remove the redundant logic for the 'check fall-through' warning
for blocks; now the same logic is used for both blocks and functions.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@99085 91177308-0d34-0410-b5e6-96231b3b80d8
ActOnStartCXXMemberDeclaration. We haven't started the field collector on this
class yet, so don't stop it. Fixes a crash in the VS buildbot and a memory error
on all the others.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@98760 91177308-0d34-0410-b5e6-96231b3b80d8
considered to be a part of the translation unit unless they're named
in a way that brings them into existence.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@98729 91177308-0d34-0410-b5e6-96231b3b80d8
on unqualified declarations.
Patch by Enea Zaffanella! Minimal adjustments: allocate the ExtInfo nodes
with the ASTContext and delete them during Destroy(). I audited a bunch of
Destroy methods at the same time, to ensure that the correct teardown was
being done.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@98540 91177308-0d34-0410-b5e6-96231b3b80d8
instantiation. Based on a patch by Enea Zaffanella! I found a way to
reduce some of the redundancy between TreeTransform's "standard"
FunctionProtoType transformation and TemplateInstantiator's override,
and I killed off the old SubstFunctionType by adding type source info
for the last cases where we were creating FunctionDecls without TSI
(at least that get passed through template instantiation).
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@98252 91177308-0d34-0410-b5e6-96231b3b80d8
injected class name of a class template or class template partial specialization.
This is a non-canonical type; the canonical type is still a template
specialization type. This becomes the TypeForDecl of the pattern declaration,
which cleans up some amount of code (and complicates some other parts, but
whatever).
Fixes PR6326 and probably a few others, primarily by re-establishing a few
invariants about TypeLoc sizes.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@98134 91177308-0d34-0410-b5e6-96231b3b80d8
uninitialized. This seems not to be the case in C++0x, where we still
call the (trivial) default constructor for a POD class
(!). Previously, we had implemented only the C++0x rules; now we
implement both. Fixes PR6536.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@97928 91177308-0d34-0410-b5e6-96231b3b80d8
which has the label map, switch statement stack, etc. Previously, we
had a single set of maps in Sema (for the function) along with a stack
of block scopes. However, this lead to funky behavior with nested
functions, e.g., in the member functions of local classes.
The explicit-stack approach is far cleaner, and we retain a 1-element
cache so that we're not malloc/free'ing every time we enter a
function. Fixes PR6382.
Also, tweaked the unused-variable warning suppression logic to look at
errors within a given Scope rather than within a given function. The
prior code wasn't looking at the right number-of-errors count when
dealing with blocks, since the block's count would be deallocated
before we got to ActOnPopScope. This approach works with nested
blocks/functions, and gives tighter error recovery.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@97518 91177308-0d34-0410-b5e6-96231b3b80d8
re-declare them. This fixes PR6317. Also add the beginnings of an interesting
test case for p1 of [class.friend] which also covers PR6317.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@97499 91177308-0d34-0410-b5e6-96231b3b80d8
a fixme and PR6451.
Only perform jump checking if the containing function has no errors,
and add the infrastructure needed to do this.
On the testcase in the PR, we produce:
t.cc:6:3: error: illegal goto into protected scope
goto later;
^
t.cc:7:5: note: jump bypasses variable initialization
X x;
^
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@97497 91177308-0d34-0410-b5e6-96231b3b80d8
errors, e.g.:
t.c:1:21: error: redefinition of parameter 'x'
int test(int x, int x);
^
t.c:1:14: note: previous declaration is here
int test(int x, int x);
^
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@96769 91177308-0d34-0410-b5e6-96231b3b80d8
fixing up a few callers that thought they were propagating NoReturn
information but were in fact saying something about exception
specifications.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@96766 91177308-0d34-0410-b5e6-96231b3b80d8
are for out of line declarations more easily. This simplifies the logic and
handles the case of out-of-line class definitions correctly. Fixes PR6107.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@96729 91177308-0d34-0410-b5e6-96231b3b80d8
Fariborz Jahanian
Douglas Gregor
Abramo Bagnara
John McCall
Anders Carlsson
Devang Patel
Daniel Dunbar
Chris Lattner
Jeffrey Yasskin
Rafael Espindola
Ted Kremenek
Eli Friedman
Chandler Carruth