As noted by Richard in the post:
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20130624/082605.html, the following code should not add an entry
into PendingLocalImplicitInstantiations, since local instantiations
should only occur within the context of other instantiations:
int foo(double y) {
struct Lambda {
template<class T> T operator()(T t) const { return t; };
} lambda;
return lambda(y);
}
Hence the attached code does the following:
1) In MarkFunctionReferenced, check if ActiveInstantiations.size()
is non-zero before adding to PendingLocalImplicitInstantiations.
2) In InstantiateFunctionDefinition, we swap out/in
PendingLocalImplicitInstantiations so that only those
pending local instantiations that are added during the instantiation
of the current function are instantiated recursively.
llvm-svn: 184903
Also, print the cache hits statistics if the log is in debugging mode vs. LLDB being a debug build - this should make it easier to gather useful metrics on cache success rate for real users
llvm-svn: 184900
The argument to -w (--category) in type * list is a regular expression
This caused unhappiness with the gnu-libstdc++ category because of the double ++
Now we check for exact textual match as-well-as regexp matching
llvm-svn: 184898
debug statements to add a missing newline. Also canonicalize to '\n' instead of
"\n"; the latter calls a function with a loop the former does not.
llvm-svn: 184897
(thanks!) by deferring the free of the filename until we finish writing
the coverage data to that file.
Bill, let me know if you'd prefer a different approach!
llvm-svn: 184895
The semi-unofficial way of returning a status from a Python command was to return a string (e.g. return "no such variable was found") that LLDB would pick as a clue of an error having happened
This checkin changes that:
- SBCommandReturnObject now exports a SetError() call, which can take an SBError or a plain C-string
- script commands now drop any return value and expect the SBCommandReturnObject ("return object") to be filled in appropriately - if you do nothing, a success will be assumed
If your commands were relying on returning a value and having LLDB pick that up as an error, please change your commands to SetError() through the return object or expect changes in behavior
llvm-svn: 184893
When the decl that we're getting alignment for is a FieldDecl, and the field's
parent record is invalid, skip the actual field alignment calculation (and
return 1-byte alignment in the general case).
Also, assert in in getASTRecordLayout that the decl is valid. This was
inspired by PR16292; see also r184581 and r184751.
llvm-svn: 184883
A default template-argument shall not be specified in a friend template
declaration.
Interestingly, we properly handled default template arguments on friend
class members but not on just friend classes.
llvm-svn: 184882
This is easier to read than the internal fixed-point representation.
If anybody knows the correct algorithm for converting fixed-point
numbers to base 10, feel free to fix it.
llvm-svn: 184881
This patch modifies TableGen to generate a function in
${TARGET}GenInstrInfo.inc called getNamedOperandIdx(), which can be used
to look up indices for operands based on their names.
In order to activate this feature for an instruction, you must set the
UseNamedOperandTable bit.
For example, if you have an instruction like:
def ADD : TargetInstr <(outs GPR:$dst), (ins GPR:$src0, GPR:$src1)>;
You can look up the operand indices using the new function, like this:
Target::getNamedOperandIdx(Target::ADD, Target::OpName::dst) => 0
Target::getNamedOperandIdx(Target::ADD, Target::OpName::src0) => 1
Target::getNamedOperandIdx(Target::ADD, Target::OpName::src1) => 2
The operand names are case sensitive, so $dst and $DST are considered
different operands.
This change is useful for R600 which has instructions with a large number
of operands, many of which model single bit instruction configuration
values. These configuration bits are common across most instructions,
but may have a different operand index depending on the instruction type.
It is useful to have a convenient way to look up the operand indices,
so these bits can be generically set on any instruction.
llvm-svn: 184879
When a 1-element vector alloca is promoted, a store instruction can often be
rewritten without converting the value to a scalar and using an insertelement
instruction to stuff it into the new alloca. This patch just adds a check
to skip that conversion when it is unnecessary. This turns out to be really
important for some ARM Neon operations where <1 x i64> is used to get around
the fact that i64 is not a legal type.
llvm-svn: 184870
It is defined on recent FreeBSD versions, so must not be mutually
exclusive with an #elif FreeBSD block.
Patch submitted by Robert Millan.
Fixes PR#16447.
llvm-svn: 184867
Richard Smith