llvm-svn: 146798

valobj.AddressOf() returns None when an address is expected in a SyntheticChildrenProvider

Patch from Enrico Granata:

The problem was that the frozen object created by the expression parser was a copy of the contents of the StgClosure, rather than a pointer to it. Thus, the expression parser was correctly computing the result of the arithmetic&cast operation along with its address, but only saving it in the live object. This meant that the frozen copy acted as an address-less variable, hence the problem.

The fix attached to this email lets the expression parser store the "live address" in the frozen copy of the address when the object is built without a valid address of its own.
Doing so, along with delegating ValueObjectConstResult to calculate its own address when necessary, solves the issue. I have also added a new test case to check for regressions in this area, and checked that existing test cases pass correctly.

llvm-svn: 146768

we handle Objective-C method calls.  Currently,
LLDB treats the result of an Objective-C method
as unknown if the type information doesn't have
the method's signature.  Now Clang can cast the
result to id if it isn't explicitly cast.

I also added a test case for this, as well as a
fix for a type import problem that this feature
exposed.

llvm-svn: 146756

handles.

llvm-svn: 146746

1 -- an address pointing off into non-executable memory -- don't
abort the unwind.  We'll use the ABI's default UnwindPlan to try
to get out of frame 1 and on many platforms with a standard frame
chain stack layout we can get back on track and get a valid frame
2.  This preserves the lldb behavior to-date; the change last week
to require the memory region to be executable broke it.

I'd like to mark this frame specially when displayed to the user;
I tried to override the places where the frame's pc value is returned
to change it to a sentinel value (e.g. LLDB_INVALID_ADDRESS) but
couldn't get that to work cleanly so I backed that part out for
now.  When this happens we'll often miss one of the user's actual
frames, the one that's of most interest to the user, so I'd like
to make this visually distinctive.

Note that a frame in non-executable memory region is only allowed
for frame 1.  After that we should be solid on the unwind and any
pc address in non-executable memory indicates a failure and we
should stop unwinding.

llvm-svn: 146723

llvm-svn: 146717

Fix a bug where when debugging with .o files, we end up with two symbols for each real OBJC_CLASS_$_whatever, one of which is correctly classified as an ObjCClass symbol, and the other is just a data symbol.  This was messing up the ObjC dynamic type detection.
<rdar://problem/10589527>

llvm-svn: 146712

Added a static memory pressure function in SBDebugger:

    void SBDebugger::MemoryPressureDetected ()

This can be called by applications that detect memory pressure to cause LLDB to release cached information.

llvm-svn: 146640

    size_t
    SBProcess::ReadCStringFromMemory (addr_t addr, void *buf, size_t size, lldb::SBError &error);

    uint64_t
    SBProcess::ReadUnsignedFromMemory (addr_t addr, uint32_t byte_size, lldb::SBError &error);

    lldb::addr_t
    SBProcess::ReadPointerFromMemory (addr_t addr, lldb::SBError &error);

These ReadCStringFromMemory() has some SWIG type magic that makes it return the
python string directly and the "buf" is not needed:

error = SBError()
max_cstr_len = 256
cstr = lldb.process.ReadCStringFromMemory (0x1000, max_cstr_len, error)
if error.Success():
    ....

The other two functions behave as expteced. This will make it easier to get integer values
from the inferior process that are correctly byte swapped. Also for pointers, the correct
pointer byte size will be used.

Also cleaned up a few printf style warnings for the 32 bit lldb build on darwin.

llvm-svn: 146636

lldb::SBValue::CreateValueFromAddress does not verify SBType::GetPointerType succeeds

SBValue::CreateValueFromAddress() should check the validity of type and its derived pointer type
before using it.  Add a test case.

llvm-svn: 146629

clients to disassemble a series of raw bytes as
demonstrated by a new testcase.

In the future, this API will also allow clients
to provide a callback that adds comments for
addresses in the disassembly.

I also modified the SWIG harness to ensure that
Python ByteArrays work as well as strings as
sources of raw data.

llvm-svn: 146611

dispatch functions that are implemented in hand-written assembly.
There is also hand-written eh_frame instructions for unwinding
from these functions.

Normally we don't use eh_frame instructions for the currently
executing function, prefering the assembly instruction profiling
method.  But in these hand-written dispatch functions, the
profiling is doomed and we should use the eh_frame instructions.

Unfortunately there's no easy way to flag/extend the eh_frame/debug_frame
sections to annotate if the unwind instructions are accurate at
all addresses ("asynchronous") or if they are only accurate at locations
that can throw an exception ("synchronous" and the normal case for 
gcc/clang generated eh_frame/debug_frame CFI).

<rdar://problem/10508134>

llvm-svn: 146551

Add null checks to several functions.  Plus add test scenario for passing None to SBTarget.FindFirstType(None) and friends.

llvm-svn: 146540

the expression parser to locate instances where
dyn_cast<>() and isa<>() are used on types, and
replace them with getAs<>() as appropriate.

The difference is that dyn_cast<>() and isa<>()
are essentially LLVM/Clang's equivalent of RTTI
-- that is, they try to downcast the object and
return NULL if they cannot -- but getAs<>() can
traverse typedefs to perform a semantic cast.

llvm-svn: 146537

Check that the pc value for frames up the stack is in a
mapped+executable region of memory.

Check that the stack pointer for frames up the stack is
in a mapped+readable region of memory.

If the unwinder ever makes a mistake walking the stack,
these checks will help to keep it from going too far into
the weeds.

These aren't fixing any bugs that I know of, but they
add extra robustness to a complicated task.

llvm-svn: 146478

if this is a mapped/executable region of memory.  If it isn't, we've jumped
through a bad pointer and we know how to unwind the stack correctly based
on the ABI.  

Previously I had 0x0 special cased but if you jumped to 0x2 on x86_64 one
frame would be skipped because the unwinder would try using the x86_64 
ArchDefaultUnwindPlan which relied on the rbp.

Fixes <rdar://problem/10508291>

llvm-svn: 146477

parse.

llvm-svn: 146473

validates the "self," "this," and "_cmd" pointers
that get passed into expressions.  It used to check
them aggressively for validity before allowing the
expression to run as an object method; now, this
functionality is gated by a bool and off by default.

Now the default is that when LLDB is stopped in a
method of a class, code entered using "expr" will
always masquerade as an instance method.  If for
some reason "self," "this," or "_cmd" is unavailable
it will be reported as NULL.  This may cause the
expression to crash if it relies on those pointers,
but for example getting the addresses of ivars will
now work as the user would expect.

llvm-svn: 146465

There were two problems associated with this radar:
1. "settings show target.source-map" failed to show the source-map after, for example,
   "settings set target.source-map /Volumes/data/lldb/svn/trunk/test/source-manager /Volumes/data/lldb/svn/trunk/test/source-manager/hidden"
   has been executed to set the source-map.
2. "list -n main" failed to display the source of the main() function after we properly set the source-map.

The first was fixed by adding the missing functionality to TargetInstanceSettings::GetInstanceSettingsValue (Target.cpp)
and updating the support files PathMappingList.h/.cpp; the second by modifying SourceManager.cpp to fix several places
with incorrect logic.

Also added a test case test_move_and_then_display_source() to TestSourceManager.py, which moves main.c to hidden/main.c,
sets target.source-map to perform the directory mapping, and then verifies that "list -n main" can still show the main()
function.

llvm-svn: 146422

it is not required.

llvm-svn: 146418

<rdar://problem/10561406>

Stopped the SymbolFileDWARF::FindFunctions (...) from always calculating
the line table entry for all functions that were found. This can slow down
the expression parser if it ends up finding a bunch of matches. Fixed the 
places that were relying on the line table entry being filled in.

Discovered a recursive stack blowout that happened when "main" didn't have
line info for it and there was no line information for "main"

llvm-svn: 146330

expression parser would never try getting typed
variables from the target.

llvm-svn: 146317

- Even if a frame isn't present, we always try
  to use FindGlobalVariable to find variables.
  Instead of using frame->TrackGlobalVariable()
  to promote the VariableSP into a ValueObject,
  we now simply use ValueObjectVariable.

- When requesting the value of a variable, we
  allow returning of the "live version" of the
  variable -- that is, the variable in the
  target instead of a pointer to its freeze
  dried version in LLDB -- even if there is no
  process present.

llvm-svn: 146315

An assertion was firing when parsing types due to trying to complete parent
class decl contenxt types too often.

Also, relax where "dsymutil" binary can come from in the Makefile.rules.

llvm-svn: 146310

hard to ensure it doesn't get invalidated out from under us.  Instead look it up from the ThreadID
and StackID when asked for it.
<rdar://problem/10554409>

llvm-svn: 146309

creating appropriate setter/getter methods for
property definitions.

llvm-svn: 146295

that if we prefer the current compile unit, followed by any compile units that
already had their DIEs parsed, followed by the rest of the matches, that we
might save some memory. This turned out not to help much. The code is commented
out, but I want to check it in so I don't lose the code in case it could help
later.

Added the ability to efficiently find the objective C class implementation
when using the new .apple_types acclerator tables with the type flags. If the
type flags are not available, we default back to what we were doing before.

llvm-svn: 146250

PlatformDarwin.cpp -- call it from both PlatformRemoteiOS.cpp
and the native process PlatformDarwin.cpp when running on an arm
system.

Bump lldb version number to 94.

llvm-svn: 146249

Rework how the breakpoint conditions & callbacks are handled.  We now iterate over all the locations at the site
that got hit, and first check the condition, and if that location's condition says we should stop, then we
run the callback.  In the end if any location's condition and callback say we should stop, then we stop. 

llvm-svn: 146242

in the context in which it was originally found, the
expression parser now goes hunting for it in all modules
(in the appropriate namespace, if applicable).  This means
that forward-declared types that exist in another shared
library will now be resolved correctly.

Added a test case to cover this.  The test case also tests
"frame variable," which does not have this functionality
yet.

llvm-svn: 146204

<rdar://problem/10545069>

llvm-svn: 146173

pointer to make the result of an expression.  LLDB now
dumps the ivars of the Objective-C object and all of
its parents.  This just required fixing a bug where we
didn't distinguish between Objective-C object pointers
and regular C-style pointers.

Also added a testcase to verify that this continues to
work.

llvm-svn: 146164

that is in a class from the expression parser, and it was causing an
assertion. There is now a function that will correctly resolve a type
even if it is in a class.

llvm-svn: 146141

<rdar://problem/10535460> lldb expression evaluation doesn't handle bit fields in ObjC classes properly

llvm-svn: 146134

take a SymbolFile reference and a lldb::user_id_t and be used in objects
which represent things in debug symbols that have types where we don't need
to know the true type yet, such as in lldb_private::Variable objects. This
allows us to defer resolving the type until something is used. More specifically
this allows us to get 1000 local variables from the current function, and if
the user types "frame variable argc", we end up _only_ resolving the type for
"argc" and not for the 999 other local variables. We can expand the use of this
as needed in the future.

Modified the DWARFMappedHash class to be able to read the HashData that has
more than just the DIE offset. It currently will read the atoms in the header
definition and read the data correctly. Currently only the DIE offset and 
type flags are supported. This is needed for adding type flags to the 
.apple_types hash accelerator tables.

Fixed a assertion crash that would happen if we have a variable that had a
DW_AT_const_value instead of a location where "location.LocationContains_DW_OP_addr()"
would end up asserting when it tried to parse the variable location as a
DWARF opcode list.

Decreased the amount of memory that LLDB would use when evaluating an expression
by 3x - 4x for clang. There was a place in the namespace lookup code that was
parsing all namespaces with a certain name in a DWARF file instead of stopping
when it found the first match. This was causing all of the compile units with
a matching namespace to get parsed into memory and causing unnecessary memory
bloat. 

Improved "Target::EvaluateExpression(...)" to not try and find a variable
when the expression contains characters that would certainly cause an expression
to need to be evaluated by the debugger. 

llvm-svn: 146130

for now to fix testcases.  Once we have a valid use
for the function information (i.e., once properties
returning UnknownAnyTy are allowed, once we read
return type information from the runtime, among
other uses) I will re-enable this.

llvm-svn: 146129

llvm-svn: 146126

symbols.  Now we find the correct method.

Unfortunately we don't get the superclass from the
runtime yet so the method doesn't import correctly
(and I added a check to make sure that doesn't hurt
us) but once we get that information right we will
report methods correctly to the parser as well.

Getting superclass information requires a common AST
context for all Objective-C runtime information,
meaning that the superclass and the subclass are in
the same AST context in all cases.  That is the next
thing that needs to be done here.

llvm-svn: 146089

avalable when a global variable is looked up.
In ClangExpressionDeclMap, a frame should usually
be available.

llvm-svn: 146066

llvm-svn: 146061