Commit Graph

3417 Commits

Author SHA1 Message Date
Douglas Gregor 721fb2b6e4 Diagnose the use of incomplete types in C++ typeid expressions
llvm-svn: 92045
2009-12-23 21:06:06 +00:00
Douglas Gregor f45f6828c6 Remove cv-qualifiers from the argument to typeid
llvm-svn: 92041
2009-12-23 20:51:04 +00:00
Eli Friedman 38b9ad88e2 Fix the overflow calculation in Sema::CheckTemplateArgument to be a bit more
accurate.

llvm-svn: 92018
2009-12-23 18:44:58 +00:00
Douglas Gregor 25ab25f39d When using a default function argument for a function template (or
member function thereof), perform the template instantiation each time
the default argument is needed. This ensures that
  (1) We get different CXXTemporary objects for each instantiation, and
  (2) Any other instantiations or definitions triggered by the
  instantiation of the default argument expression are guaranteed to
  happen; previously, they might have been suppressed, e.g., because
  they happened in an unevaluated context.

This fixes the majority of PR5810. However, it does not address the
problem where we may have multiple uses of the same CXXTemporary
within an expression when the temporary came from a non-instantiated
default argument expression.

llvm-svn: 92015
2009-12-23 18:19:08 +00:00
Nuno Lopes cfca1f0dc1 move a few more symbols to .rodata/.data.rel.ro
llvm-svn: 92012
2009-12-23 17:49:57 +00:00
Douglas Gregor e6600379b1 Fix DISABLE_SMART_POINTERS build
llvm-svn: 92008
2009-12-23 17:40:29 +00:00
Eric Christopher c87915629b Update for the intrinsic changes in llvm: the object size intrinsic
only takes a boolean second argument now. Update tests accordingly.
Currently the builtin still accepts the full range for compatibility.

llvm-svn: 91983
2009-12-23 03:49:37 +00:00
John McCall 0a4bb26ed0 Set a member's access specifier even if it doesn't match the previous specifier.
Prevents an assert on successive redeclarations.

Fixed PR5573.

llvm-svn: 91956
2009-12-23 00:37:40 +00:00
Douglas Gregor 04c5f973ee Objective-C methods can be variadic, too. Who knew.
llvm-svn: 91951
2009-12-23 00:21:46 +00:00
Douglas Gregor 8d48e9a26e Switch Sema::CreateOverloadedUnaryOp over to InitializationSequence.
llvm-svn: 91948
2009-12-23 00:02:00 +00:00
Nuno Lopes 56abcbdb47 warn when attribute warn_unused_result is applied to void functions.
while at it, remove an outdated FIXME

llvm-svn: 91946
2009-12-22 23:59:52 +00:00
Douglas Gregor 684d7bdc43 Allow the first parameter of operator new to be a cv-qualified
size_t. Also, fix an issue with initialization of parameters in calls,
where we weren't removing the cv-qualifiers on the parameter type
itself. Fixes PR5823.

llvm-svn: 91941
2009-12-22 23:42:49 +00:00
Douglas Gregor 19fc0b753a Make sure that reinterpret_cast gets a CastKind on all successful
paths. Fixes "cannot compile this unexpected cast lvalue yet" error in
llvm/lib/Analysis/IPA/GlobalsModRef.cpp.

llvm-svn: 91932
2009-12-22 22:47:22 +00:00
John McCall a1d8550778 Add using shadow decls to the "instantiated locals" map, fixing PR5847.
llvm-svn: 91928
2009-12-22 22:26:37 +00:00
Douglas Gregor b6ea60872d Switch Sema::AddCXXDirectInitializerToDecl over to InitializationSequence
llvm-svn: 91927
2009-12-22 22:17:25 +00:00
Douglas Gregor 0a70c4d9a2 Switch parameter passing for overloaded binary operators to
InitializationSequence. Fixes the -fsyntax-only failure in
llvm/lib/Transforms/Scalar/InstructionCombining.cpp.

llvm-svn: 91921
2009-12-22 21:44:34 +00:00
Douglas Gregor 2e9c795df0 When transforming a C++ "new" expression that was not explicitly given
a size, check whether the transformed type is itself an array type. If
so, take the major array bound as the size to allocate. Fixes PR5833.

llvm-svn: 91907
2009-12-22 17:13:37 +00:00
Douglas Gregor bbeb5c391c Switch parameter-passing for calls via function pointers (where we
don't have a FunctionDecl) over to InitializationSequence.

llvm-svn: 91906
2009-12-22 16:09:06 +00:00
Douglas Gregor 1b3039344b Switch InitializedEntity from TypeLoc down to just QualTypes, since we don't use the location information but we did spend a bunch of time building faked-up TypeLocs
llvm-svn: 91905
2009-12-22 15:35:07 +00:00
Douglas Gregor 96596c98fb Switch initialization of parameters in a call over to
InitializationSequence (when a FunctionDecl is present). This required
a few small fixes to initialization sequences:

  - Make sure to use the adjusted parameter type for initialization of
  function parameters.
  - Implement transparent union calling semantics in C

llvm-svn: 91902
2009-12-22 07:24:36 +00:00
Chris Lattner d806cbc98d fix a bug handling the gnu ?: extension. Patch by Storlek on IRC,
who prefers to be stealthy and mysterious.

llvm-svn: 91888
2009-12-22 06:00:13 +00:00
Anders Carlsson 9c45ad7a1b When we simply return a retained member expression when instantiating, we must also mark the member decl as referenced.
llvm-svn: 91887
2009-12-22 05:24:09 +00:00
Eli Friedman e2808379b8 Zap CheckInitializerTypes.
llvm-svn: 91884
2009-12-22 02:50:30 +00:00
Eli Friedman 5f101b95c6 Switch default arguments over to InitializationSequence.
llvm-svn: 91883
2009-12-22 02:46:13 +00:00
Eli Friedman a553d4ac95 Switch compound literals over to InitializationSequence.
llvm-svn: 91882
2009-12-22 02:35:53 +00:00
Eli Friedman 463e523ad8 Switch file-scope assignment initialization over to InitializationSequence.
llvm-svn: 91881
2009-12-22 02:10:53 +00:00
Douglas Gregor da8cdbce35 Centralize the emission/suppression/delay of diagnostics describing runtime before in the new function Sema::DiagRuntimeBehavior, addressing one of Chris' comments.
llvm-svn: 91870
2009-12-22 01:01:55 +00:00
John McCall 463e10cbd5 Stop diagnosing the use of inner classes as friends. ddunbar asked whether
this was useful, and on review Doug and I decided it was probably on the level
of a bug in the standard and therefore not worth a warning even in -pedantic.
If someone disagrees and urgently wants clang++ to warn about this in strict
c++98 mode, we can talk about it.

llvm-svn: 91868
2009-12-22 00:59:39 +00:00
Douglas Gregor 507eb87f05 Eliminate the ASTContext argument to CXXConstructorDecl::isCopyConstructor, since the context is available in the Decl
llvm-svn: 91862
2009-12-22 00:34:07 +00:00
Douglas Gregor 4141d5bea2 When converting from a type to itself or one of its base classes via a
constructor call, the conversion is only a standard conversion
sequence if that constructor is a copy constructor. This fixes PR5834
in a semi-lame way, because the "real" fix will be to move over to
InitializationSequence. That will happen "soonish", but not now.

llvm-svn: 91861
2009-12-22 00:21:20 +00:00
Douglas Gregor 2bb0765bb4 When filling in value initializations within an initializer list, be
sure to fill in the initialized member of a union when a member was
explicitly designated. Fixes PR5843.

llvm-svn: 91858
2009-12-22 00:05:34 +00:00
Douglas Gregor 8364e6b568 When a template-id refers to a single function template, and the
explicitly-specified template arguments are enough to determine the
instantiation, and either template argument deduction fails or is not
performed in that context, we can resolve the template-id down to a
function template specialization (so sayeth C++0x
[temp.arg.explicit]p3). Fixes PR5811.

llvm-svn: 91852
2009-12-21 23:17:24 +00:00
Douglas Gregor 3985791d62 Introduce an assertion to ensure that template argument deduction does
not deduce an "overload" type. Such a deduction indicates a failure in
semantic analysis  (e.g., PR5811) that currently isn't caught until
code-generation time. This assertions makes it clearer that this
particular issue is a semantic-analysis problem, not a code-gen problem.

llvm-svn: 91844
2009-12-21 21:27:38 +00:00
Fariborz Jahanian ffc420cb51 Allow comparison of 'void *' with function pointer
as a g++ extension (fixes radar 7481987).

llvm-svn: 91827
2009-12-21 18:19:17 +00:00
John McCall b5a0d316dc Reorganize the base-lookup bits of ActOnMemInitializer in order to better
support diagnostics and error recovery.

llvm-svn: 91825
2009-12-21 10:41:20 +00:00
Eli Friedman 9025ec2ee5 Fix for PR5840: fix the kind of name lookup used for classes in
Sema::getTypeName.

"LookupNestedNameSpecifierName" isn't quite the right kind of lookup, though; 
it doesn't ignore namespaces.  Someone more familiar with the lookup code
should fix this properly.

llvm-svn: 91809
2009-12-21 01:42:38 +00:00
Eli Friedman a26001bf96 Small comment fix.
llvm-svn: 91808
2009-12-21 01:26:45 +00:00
Nuno Lopes 518e370719 fix PR4010: add support for the warn_unused_result for function pointers
llvm-svn: 91803
2009-12-20 23:11:08 +00:00
Eli Friedman 7dac3712a2 Make sure we instantiate the destructor for variables initialized by
assignment.

llvm-svn: 91798
2009-12-20 22:29:11 +00:00
Eli Friedman 4afe9a3518 Fix review comment; no visible change.
llvm-svn: 91797
2009-12-20 22:12:03 +00:00
Douglas Gregor 39c778b443 Switch default-initialization of variables of class type (or array thereof) over to InitializationSequence. I could swear that this fixes a PR somewhere, but I couldn't figure out which one
llvm-svn: 91796
2009-12-20 22:01:25 +00:00
Douglas Gregor 1b6d10555f Revert accidental commit
llvm-svn: 91795
2009-12-20 21:10:58 +00:00
Douglas Gregor b752f1e25c Fix CMake build on windows, from Cedric Venet
llvm-svn: 91794
2009-12-20 20:56:30 +00:00
John McCall 1c7e6ec27b Don't inject the class name until that magical lbrace.
Because of the rules of base-class lookup* and the restrictions on typedefs, it
was actually impossible for this to cause any problems more serious than the
spurious acceptance of
  template <class T> class A : B<A> { ... };
instead of
  template <class T> class A : B<A<T> > { ... };
but I'm sure we can all agree that that is a very important restriction which
is well worth making another Parser->Sema call for.

(*) n.b. clang++ does not implement these rules correctly;  we are not ignoring
    non-type names

llvm-svn: 91792
2009-12-20 07:58:13 +00:00
John McCall b878801046 Kill off PreDeclaratorDC.
llvm-svn: 91772
2009-12-19 10:53:49 +00:00
John McCall 6df5fef637 Refactor to remove more dependencies on PreDeclaratorDC. I seem to have made
the redeclaration problems in the [temp.explicit]p3 testcase worse, but I can
live with that;  they'll need to be fixed more holistically anyhow.

llvm-svn: 91771
2009-12-19 10:49:29 +00:00
John McCall 4d6d6137a0 Don't use EnterDeclaratorContext when rebuilding a type in the current
instantiation, since we're not using a Scope object for that anyway.

llvm-svn: 91770
2009-12-19 09:35:56 +00:00
John McCall 1f4ee7bd2f Just push a new scope when parsing an out-of-line variable definition.
Magically fixes all the terrible lookup problems associated with not pushing
a new scope.  Resolves an ancient xfail and an LLVM misparse.

llvm-svn: 91769
2009-12-19 09:28:58 +00:00
Eli Friedman 7827520ce8 Initialization improvements: addition of string initialization and a few
small bug fixes in SemaInit, switch over SemaDecl to use it more often, and
change a bunch of diagnostics which are different with the new initialization
code.

llvm-svn: 91767
2009-12-19 08:11:05 +00:00
Douglas Gregor a4b592a7d5 Switch more of Sema::CheckInitializerTypes over to
InitializationSequence. Specially, switch initialization of a C++
class type (either copy- or direct-initialization). 

Also, make sure that we create an elidable copy-construction when
performing copy initialization of a C++ class variable. Fixes PR5826.

llvm-svn: 91750
2009-12-19 03:01:41 +00:00