an architecture into ArchSpec:
uint32_t
ArchSpec::GetMinimumOpcodeByteSize() const;
uint32_t
ArchSpec::GetMaximumOpcodeByteSize() const;
Added an AddressClass to the Instruction class in Disassembler.h.
This allows decoded instructions to know know if they are code,
code with alternate ISA (thumb), or even data which can be mixed
into code. The instruction does have an address, but it is a good
idea to cache this value so we don't have to look it up more than
once.
Fixed an issue in Opcode::SetOpcodeBytes() where the length wasn't
getting set.
Changed:
bool
SymbolContextList::AppendIfUnique (const SymbolContext& sc);
To:
bool
SymbolContextList::AppendIfUnique (const SymbolContext& sc,
bool merge_symbol_into_function);
This function was typically being used when looking up functions
and symbols. Now if you lookup a function, then find the symbol,
they can be merged into the same symbol context and not cause
multiple symbol contexts to appear in a symbol context list that
describes the same function.
Fixed the SymbolContext not equal operator which was causing mixed
mode disassembly to not work ("disassembler --mixed --name main").
Modified the disassembler classes to know about the fact we know,
for a given architecture, what the min and max opcode byte sizes
are. The InstructionList class was modified to return the max
opcode byte size for all of the instructions in its list.
These two fixes means when disassemble a list of instructions and dump
them and show the opcode bytes, we can format the output more
intelligently when showing opcode bytes. This affects any architectures
that have varying opcode byte sizes (x86_64 and i386). Knowing the max
opcode byte size also helps us to be able to disassemble N instructions
without having to re-read data if we didn't read enough bytes.
Added the ability to set the architecture for the disassemble command.
This means you can easily cross disassemble data for any supported
architecture. I also added the ability to specify "thumb" as an
architecture so that we can force disassembly into thumb mode when
needed. In GDB this was done using a hack of specifying an odd
address when disassembling. I don't want to repeat this hack in LLDB,
so the auto detection between ARM and thumb is failing, just specify
thumb when disassembling:
(lldb) disassemble --arch thumb --name main
You can also have data in say an x86_64 file executable and disassemble
data as any other supported architecture:
% lldb a.out
Current executable set to 'a.out' (x86_64).
(lldb) b main
(lldb) run
(lldb) disassemble --arch thumb --count 2 --start-address 0x0000000100001080 --bytes
0x100001080: 0xb580 push {r7, lr}
0x100001082: 0xaf00 add r7, sp, #0
Fixed Target::ReadMemory(...) to be able to deal with Address argument object
that isn't section offset. When an address object was supplied that was
out on the heap or stack, target read memory would fail. Disassembly uses
Target::ReadMemory(...), and the example above where we disassembler thumb
opcodes in an x86 binary was failing do to this bug.
llvm-svn: 128347
rdar://problem/9173060 lldb hangs while running unique-types
disappears if running with clang version >= 3. Modify the TestUniqueTypes.py
to detect if we are running with clang version < 3 and, if true, skip the test.
Update the lldbtest.system() function to return a tuple of (stdoutdata, stderrdata)
since we need the stderr data from "clang -v" command. Modify existing clients of
lldbtest.system() to now use, for example:
# First, capture the golden output emitted by the oracle, i.e., the
# series of printf statements.
- go = system("./a.out", sender=self)
+ go = system("./a.out", sender=self)[0]
# This golden list contains a list of (variable, value) pairs extracted
# from the golden output.
gl = []
And add two utility functions to lldbutil.py.
llvm-svn: 128162
overlap in the SWIG integration which has now been fixed by introducing
callbacks for initializing SWIG for each language (python only right now).
There was also a breakpoint command callback that called into SWIG which has
been abtracted into a callback to avoid cross over as well.
Added a new binary: lldb-platform
This will be the start of the remote platform that will use as much of the
Host functionality to do its job so it should just work on all platforms.
It is pretty hollowed out for now, but soon it will implement a platform
using the GDB remote packets as the transport.
llvm-svn: 128053
types that have different contents. Currently LLDB is incorrectly uniquing,
on MacOSX, the std::vector _VectorImpl class from the two different vector
templates. The DWARF looks like:
0x0000008e: DW_TAG_structure_type [7] *
DW_AT_name( "_Vector_base<int,std::allocator<int> >" )
DW_AT_declaration( 0x01 )
DW_AT_sibling( {0x00000103} )
0x00000098: DW_TAG_structure_type [8] *
DW_AT_name( "_Vector_impl" )
DW_AT_byte_size( 0x18 )
DW_AT_decl_file( "/usr/include/c++/4.2.1/bits/stl_vector.h" )
DW_AT_decl_line( 83 )
0x000000a0: DW_TAG_inheritance [9]
DW_AT_type( {0x000006fa} ( allocator<int> ) )
DW_AT_data_member_location( +0 )
DW_AT_accessibility( DW_ACCESS_public )
0x0000011b: DW_TAG_structure_type [7] *
DW_AT_name( "_Vector_base<short int,std::allocator<short int> >" )
DW_AT_declaration( 0x01 )
DW_AT_sibling( {0x00000190} )
0x00000125: DW_TAG_structure_type [8] *
DW_AT_name( "_Vector_impl" )
DW_AT_byte_size( 0x18 )
DW_AT_decl_file( "/usr/include/c++/4.2.1/bits/stl_vector.h" )
DW_AT_decl_line( 83 )
0x0000012d: DW_TAG_inheritance [9]
DW_AT_type( {0x00000f75} ( allocator<short int> ) )
DW_AT_data_member_location( +0 )
DW_AT_accessibility( DW_ACCESS_public )
In this case it using DIE 0x00000098 for both 0x00000098 and 0x00000125.
This test will help detect this issue once I have a fix for it. I have a fix
that I am testing.
llvm-svn: 127660
This uses pexpect module to spawn a 'lldb' program and uses pseudo-TTY to talk to
the child application.
The test cases test setting breakpoints, adding a stop-hook with line range, and
verifies that when the inferior stops, the stop-hook will fire off when it is
within range and will not fire off when it is out of range.
llvm-svn: 127519
SBTarget.Launch() API, stop at a breakpoint, get the stopped thread, and verify that the
pid of the stopped thread's process is equal to the pid of the process returned by
SBTarget.Launch().
llvm-svn: 127444
Currently it has only test cases for SBThread.GetStopDescription() API.
Also modified lldb.swig to add typemap for (char *dst, size_t dst_len)
which occurs for SBThread::GetStopDescription() C++ API. For Python
scripting:
# Due to the typemap magic (see lldb.swig), we pass in an (int)length to GetStopDescription
# and expect to get a Python string as the result object!
# The 100 is just an arbitrary number specifying the buffer size.
stop_description = thread.GetStopDescription(100)
llvm-svn: 127173
API with a process not in eStateConnected, and checks that the remote launch failed.
Modify SBProcess::RemoteLaunch()/RemoteAttachToProcessWithID()'s log statements to fix a
crasher when logging is turned on.
llvm-svn: 127055
We start a fake debugserver listening on localhost:12345 and issue the command
'process connect connect://localhost:12345' to connect to it.
llvm-svn: 127048
This allows us to override CFLAGS on the command line:
$ CFLAGS='-arch $(ARCH) -gdwarf-2 -O0' ./dotest.py -C clang -A i386 -v objc-optimized
Session logs for test failures/errors will go into directory '2011-03-04-10_33_57'
Command invoked: python ./dotest.py -C clang -A i386 -v objc-optimized
----------------------------------------------------------------------
Collected 2 tests
1: test_break_with_dsym (TestObjcOptimized.ObjcOptimizedTestCase)
Test 'expr member' continues to work for optimized build. ... ok
2: test_break_with_dwarf (TestObjcOptimized.ObjcOptimizedTestCase)
Test 'expr member' continues to work for optimized build. ... ok
----------------------------------------------------------------------
Ran 2 tests in 1.902s
OK
$
llvm-svn: 127011
test that objective-c expression parser continues to work for optimized build.
Radar filed:
# rdar://problem/9087739
# test failure: objc_optimized does not work for "-C clang -A i386"
llvm-svn: 127009
on the command line. For example, use '-A x86_64^i386' to launch the inferior use both x86_64
and i386.
This is an example of building the debuggee using both clang and gcc compiers:
[17:30:46] johnny:/Volumes/data/lldb/svn/trunk/test $ ./dotest.py -C clang^gcc -v -f SourceManagerTestCase.test_modify_source_file_while_debugging
Session logs for test failures/errors will go into directory '2011-03-03-17_31_39'
Command invoked: python ./dotest.py -C clang^gcc -v -f SourceManagerTestCase.test_modify_source_file_while_debugging
Configuration: compiler=clang
----------------------------------------------------------------------
Collected 1 test
1: test_modify_source_file_while_debugging (TestSourceManager.SourceManagerTestCase)
Modify a source file while debugging the executable. ... Command 'run' failed!
original content: #include <stdio.h>
int main(int argc, char const *argv[]) {
printf("Hello world.\n"); // Set break point at this line.
return 0;
}
new content: #include <stdio.h>
int main(int argc, char const *argv[]) {
printf("Hello lldb.\n"); // Set break point at this line.
return 0;
}
os.path.getmtime() after writing new content: 1299202305.0
content restored to: #include <stdio.h>
int main(int argc, char const *argv[]) {
printf("Hello world.\n"); // Set break point at this line.
return 0;
}
os.path.getmtime() after restore: 1299202307.0
ok
----------------------------------------------------------------------
Ran 1 test in 8.259s
OK
Configuration: compiler=gcc
----------------------------------------------------------------------
Collected 1 test
1: test_modify_source_file_while_debugging (TestSourceManager.SourceManagerTestCase)
Modify a source file while debugging the executable. ... original content: #include <stdio.h>
int main(int argc, char const *argv[]) {
printf("Hello world.\n"); // Set break point at this line.
return 0;
}
new content: #include <stdio.h>
int main(int argc, char const *argv[]) {
printf("Hello lldb.\n"); // Set break point at this line.
return 0;
}
os.path.getmtime() after writing new content: 1299202307.0
content restored to: #include <stdio.h>
int main(int argc, char const *argv[]) {
printf("Hello world.\n"); // Set break point at this line.
return 0;
}
os.path.getmtime() after restore: 1299202309.0
ok
----------------------------------------------------------------------
Ran 1 test in 2.301s
OK
[17:31:49] johnny:/Volumes/data/lldb/svn/trunk/test $
llvm-svn: 126979
among other things:
// When stopped on breakppint 1, we can get the line entry using SBFrame API
// SBFrame.GetLineEntry(). We'll get the start address for the the line entry
// with the SBAddress type, resolve the symbol context using the SBTarget API
// SBTarget.ResolveSymbolContextForAddress() in order to get the SBSymbol.
//
// We then stop at breakpoint 2, get the SBFrame, and the the SBFunction object.
//
// The address from calling GetStartAddress() on the symbol and the function
// should point to the same address, and we also verify that.
And add one utility function disassemble(target, function_or_symbol) to lldbutil.py:
"""Disassemble the function or symbol given a target.
It returns the disassembly content in a string object.
"""
TestDisasm.py uses the disassemble() function to do disassembly on the SBSymbol, and
then the SBFunction object.
llvm-svn: 126955
// When stopped on breakppint 1, and then 2, we can get the line entries using
// SBFrame API SBFrame.GetLineEntry(). We'll get the start addresses for the
// two line entries; with the start address (of SBAddress type), we can then
// resolve the symbol context using the SBTarget API
// SBTarget.ResolveSymbolContextForAddress().
//
// The two symbol context should point to the same symbol, i.e., 'a' function.
Add two utility functions to lldbutil.py:
o get_stopped_threads(process, reason):
return the list of threads with the specified stop reason or an empty list if not found
o get_stopped_thread(process, reason):
return the first thread with the given stop reason or None if not found
llvm-svn: 126916
o int_to_bytearray()
o bytearray_to_int()
They return/interpret the bytearray in the little endian format.
For big endian, simply perform ba.reverse() on the bytearray object.
And modify TestProcessAPI.py to take advantage of the functions.
llvm-svn: 126813
among other SBProcess APIs, to write (int)256 into a memory location of a global variable
(int)my_int and reads/checks the variable afterwards.
llvm-svn: 126792
the SBProcess.ReadMemory() API, which, due to SWIG typemap'ing, expects 3 arguments (the location
to read from, the size in bytes to read, and an SBError object), and returns the result as a
Python string object.
On SnowLeopard where this has been tested, the SWIG script needs to be pampered (use the exact
same parameter names as in SBProcess.h) in order for this to work.
llvm-svn: 126736
current working directory when running the inferior. Radar filed:
# rdar://problem/9056462
# The process launch flag '-w' for setting the current working directory not working?
llvm-svn: 126529
of Stephen Wilson's idea (thanks for the input Stephen!). What I ended up
doing was:
- Got rid of ArchSpec::CPU (which was a generic CPU enumeration that mimics
the contents of llvm::Triple::ArchType). We now rely upon the llvm::Triple
to give us the machine type from llvm::Triple::ArchType.
- There is a new ArchSpec::Core definition which further qualifies the CPU
core we are dealing with into a single enumeration. If you need support for
a new Core and want to debug it in LLDB, it must be added to this list. In
the future we can allow for dynamic core registration, but for now it is
hard coded.
- The ArchSpec can now be initialized with a llvm::Triple or with a C string
that represents the triple (it can just be an arch still like "i386").
- The ArchSpec can still initialize itself with a architecture type -- mach-o
with cpu type and subtype, or ELF with e_machine + e_flags -- and this will
then get translated into the internal llvm::Triple::ArchSpec + ArchSpec::Core.
The mach-o cpu type and subtype can be accessed using the getter functions:
uint32_t
ArchSpec::GetMachOCPUType () const;
uint32_t
ArchSpec::GetMachOCPUSubType () const;
But these functions are just converting out internal llvm::Triple::ArchSpec
+ ArchSpec::Core back into mach-o. Same goes for ELF.
All code has been updated to deal with the changes.
This should abstract us until later when the llvm::TargetSpec stuff gets
finalized and we can then adopt it.
llvm-svn: 126278
Jim Ingham