- Add static method to test if the current lexer is a non-macro/non-pragma
lexer.
- Refactor some code in PPLexerChange to use this static method.
- No performance change.
llvm-svn: 59486
This is because the PTHLexer will not support this method. Performance testing
on preprocessing Cocoa.h shows that this results in a negligible performance
difference (less than 1%).
I tried making Lexer::SetCommentRetentionState() an out-of-line function (a
precursor to making it a virtual function in PreprocessorLexer) and noticed a 1%
decrease in speed (it is called in a hot part of the Preprocessor).
llvm-svn: 59477
alias for the current PreprocessorLexer. No functionality change. Performance
testing shows this results in no performance degradation when preprocessing
Cocoa.h.
llvm-svn: 59474
PreprocessorLexer, which will either be a 'Lexer' or 'PTHLexer'.
- Added stub field 'CurPTHLexer' to keep track of the current PTHLexer.
- Modified IncludeStackInfo to track both the current PTHLexer and
current PreprocessorLexer.
llvm-svn: 59472
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
where the control reaches the end of a non-void function and also allows the
compiler to generate better code. When this assertion is false we can easily
add more else cases.
llvm-svn: 59468
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
representing the names of declarations in the C family of
languages. DeclarationName is used in NamedDecl to store the name of
the declaration (naturally), and ObjCMethodDecl is now a NamedDecl.
llvm-svn: 59441
- RegionView and RegionViewMap is introduced to assist back-mapping from
super region to subregions.
- GDM is used to carry RegionView information.
- AnonTypedRegion is added to represent a typed region introduced by pointer
casting. Later AnonTypedRegion can be used in other similar cases, e.g.,
malloc()'ed region.
- The specific conversion is delegated to store manager.
llvm-svn: 59382
In that patch I added a bogus type promotion for unary '!'.
The real bug was more fallout from edges cases with compound assignments and conjured symbolic values. Now the conjured value has the type of the LHS expression, and we do a promotion to the computation type. We also now correctly do a conversion from the computation type back to the LHS type.
llvm-svn: 59349
- Block-expression for 'ObjCForCollectionStmt' is not alive before it occurs
- Recursively visit 'element' expression for ObjCForCollectionStmt to get liveness for referenced block-level expressions and variables.
llvm-svn: 59316
function call created in response to the use of operator syntax that
resolves to an overloaded operator in C++, e.g., "str1 +
str2" that resolves to std::operator+(str1, str2)". We now build a
CXXOperatorCallExpr in C++ when we pick an overloaded operator. (But
only for binary operators, where we actually implement overloading)
I decided *not* to refactor the current CallExpr to make it abstract
(with FunctionCallExpr and CXXOperatorCallExpr as derived
classes). Doing so would allow us to make CXXOperatorCallExpr a little
bit smaller, at the cost of making the argument and callee accessors
virtual. We won't know if this is going to be a win until we can parse
lots of C++ code to determine how much memory we'll save by making
this change vs. the performance penalty due to the extra virtual
calls.
llvm-svn: 59306
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
some more bullet-proofing/enhancements for tryEvaluate. This shouldn't
cause any behavior changes except for handling cases where we were
crashing before and being able to evaluate a few more cases in tryEvaluate.
This should settle the minor mess surrounding r59196.
llvm-svn: 59224
little rude; I figure it's cleaner to just back this out now so
it doesn't get forgotten or mixed up with other checkins.
The modification to isICE is simply wrong; I've added a test that the
change to isICE breaks.
I'm pretty sure the modification to tryEvaluate is also wrong.
At the very least, there's some serious miscommunication going on here,
as this is going in exactly the opposite direction of r59105. My
understanding is that tryEvaluate is not supposed to care about side
effects. That said, a lot of the clients to tryEvaluate are
expecting it to enforce a no-side-effects policy, so we probably need
another method that provides that guarantee.
llvm-svn: 59212
landing pads.
- Primarily a cleanliness issue instead of a performance issue (this
eliminates all blocks w/o predecessors on 176.gcc/expr.c), but this
also allows subsequent code to recognize it is unreachable and
potentially avoid IRgen.
llvm-svn: 59211
- Use dotted notation for blocks related to a particular statement
type.
- Use .end for landing pads.
No functionality change in NDEBUG mode. :)
llvm-svn: 59210
- Indicates that caller is done with the block and it can be dropped
if it has no predecessors. Useful for callers who need to make
landing pads but which may not be reached.
No functionality change.
llvm-svn: 59207
- Evaluation of , operator used bogus assumption that LHS could be
evaluated as an integral expression even though its type is
unspecified.
This change is making isICE very permissive of the LHS in non-evaluated
contexts because it is not clear what predicate we would use to reject
code here. The standard didn't offer me any guidance; opinions?
llvm-svn: 59196
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
PreprocessorLexer now has a virtual method "IndirectLex" which allows it to call the lex method of its subclasses. This is not for performance intensive operations.
llvm-svn: 59185
- Rename SetSVal to BindLoc
- Add BindDecl
- Add BindExpr
GRState:
- Environment now binds to Stmt* instead of Expr*. This is needed for processing ObjCForCollectionStmt (essentially the declaration of the the 'element' variable can have an SVal attached to it).
- BindDecl no longer accepts Expr* for the initialization value; use SVal* instead.
llvm-svn: 59152
functions for built-in operators, e.g., the builtin
bool operator==(int const*, int const*)
can be used for the expression "x1 == x2" given:
struct X {
operator int const*();
} x1, x2;
The scheme for handling these built-in operators is relatively simple:
for each candidate required by the standard, create a special kind of
candidate function for the built-in. If overload resolution picks the
built-in operator, we perform the appropriate conversions on the
arguments and then let the normal built-in operator take care of it.
There may be some optimization opportunity left: if we can reduce the
number of built-in operator overloads we generate, overload resolution
for these cases will go faster. However, one must be careful when
doing this: GCC generates too few operator overloads in our little
test program, and fails to compile it because none of the overloads it
generates match.
Note that we only support operator overload for non-member binary
operators at the moment. The other operators will follow.
As part of this change, ImplicitCastExpr can now be an lvalue.
llvm-svn: 59148
This pushes it a lot closer to being able to deal with most of the stuff
CodeGen's constant expression evaluator knows how to deal with. This
also fixes PR3003.
The test could possibly use some improvement, but this'll work for now.
Test 6 is inspired by PR3003; the other tests are mostly just designed
to exercise the new code. The reason for the funny structure of the
tests is that type fixing for arrays inside of structs is the only place
in Sema that calls tryEvaluate, at least for the moment.
llvm-svn: 59125
- Split out "simple" statements which can easily handle IR generation
when there is no insert point. These are generally statements which
start by emitting a new block or are only containers for other
statements.
- This fixes a regression in emitting dummy blocks, notably for case
statements.
- This also fixes spurious emission of a number of debug stoppoint
intrinsic instructions.
Remove unneeded sw.body block, just clear the insertion point.
Lift out CodeGenFunction::EmitStopPoint which calls into the
CGDebugInfo class when generating debug info.
Normalize definitions of Emit{Break,Continue}Stmt and usage of
ErrorUnsupported.
llvm-svn: 59118
CodeGenFunction.cpp. Change VisitConditionalOperator to use
constant fold instead of codegen'ing a constant conditional.
Change ForStmt to use EmitBranchOnBoolExpr, this shrinks
expr.c very slightly to 40239 lines.
llvm-svn: 59113
crashing because we errors are ignored in subexpressions that are not evaluated,
but we still evaluate the result of parents. This would cause an assertion
because the erroneous subexpr didn't have its result value set to the right type.
llvm-svn: 59110
- EmitStmt is no longer required to finish with a current insertion
point defined (i.e. it does not need to make dummy
blocks). Instead, it can clear the insertion point in the builder
which indicates that the current insertion point is unreachable.
- CodeGenFunction provides HaveInsertPoint and EnsureInsertPoint
which respectively test if there is an insert point and ensure an
insertion point exists (by making a dummy block).
- Clearly mark functions in CodeGenFunction which can be called with
no insertion point defined. Currently this is a limited set, and
EmitStmt simply EnsureInsertPoint()s before emitting subsequent IR.
Remove EmitDummyBlock, which is no longer needed. Clients who haven't
already cleared the insertion point (typically via EmitBranch) can do
so by hand.
Remove isDummyBlock, which has effectively been renamed to
HaveInsertPoint.
The main thrust of this change is that we no longer have create dummy
blocks just to destroy them a short time later in EmitBlock in the
common case that there is no unreachable code following something like
a goto.
Additionally, this means that we are not using the hokey condition in
isDummyBlock that a block without a name is a dummy block. Guess how
well that works when we never emit block names!
llvm-svn: 59089
- Added support for ObjCForCollectionStmt
- Fixed bug where expression values would be always set to uninitialized when loops were involved
llvm-svn: 59076
the size of the -O0 output on some cases. For example, on expr.c from
176.gcc, it shrinks the .ll file from 43164 to 42835 lines, and removed
references to two external symbols.
llvm-svn: 59034
1. emit proper debug info for forward decls of structs.
2. emit DW_TAG_member nodes around members of a field like llvm-gcc does.
This slows down debug info generation, but is required for correctness.
llvm-svn: 58973
-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
When allocating an array for ParamInfo, the "decl->getNumParams()" call was used, but this will return 0 since it checks ParamInfo (which isn't yet defined and is null).
The result was that ParamInfo got an array of zero length to hold the ParmVarDecls.
llvm-svn: 58850
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
duplication in the handling of copy-initialization by constructor,
which occurs both for initialization of a declaration and for
overloading. The initialization code is due for some refactoring.
llvm-svn: 58756
cope with the case where a user-defined conversion is actually a copy
construction, and therefore can be compared against other standard
conversion sequences. While I called this a hack before, now I'm
convinced that it's the right way to go.
Compare overloads based on derived-to-base conversions that invoke
copy constructors.
Suppress user-defined conversions when attempting to call a
user-defined conversion.
llvm-svn: 58629
when appropriate.
Conversions for class types now make use of copy constructors. I've
replaced the egregious hack allowing class-to-class conversions with a
slightly less egregious hack calling these conversions standard
conversions (for overloading reasons).
llvm-svn: 58622
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
sets the whole struct to Unknown. Then we cannot assume the V passed to
BindStruct() is always a CompoundVal. When it is an UnknownVal, we call
BindStructToVal(UnknownVal).
2. Change the signature of InitializeStructToUndefined() to BindStructToVal()
to reuse the code.
llvm-svn: 58564
the types for size_t and ptrdiff_t more accurate. I think all of these
are correct, but please compare the defines for __PTRDIFF_TYPE__ and
__SIZE_TYPE__ to gcc to double-check; this particularly applies to
those on BSD variants, since I'm not sure what they do here; I assume
here that they're the same as on Linux.
Fixes wchar_t to be "int", not "unsigned int" (which I think is
correct on everything but Windows).
Fixes ptrdiff_t to be "int" rather than "short" on PIC16; "short" is an
somewhat strange choice because it normally gets promoted, and it's not
consistent with the choice for size_t.
llvm-svn: 58556
conversions.
Notes:
- Overload resolution for converting constructors need to prohibit
user-defined conversions (hence, the test isn't -verify safe yet).
- We still use hacks for conversions from a class type to itself.
This will be the case until we start implicitly declaring the appropriate
special member functions. (That's next on my list)
llvm-svn: 58513
Chris Lattner