hanchenye-llvm-project/lldb/source/Commands/CommandObjectBreakpoint.cpp

1636 lines
61 KiB
C++
Raw Normal View History

//===-- CommandObjectBreakpoint.cpp -----------------------------*- C++ -*-===//
//
// The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
#include "CommandObjectBreakpoint.h"
#include "CommandObjectBreakpointCommand.h"
// C Includes
// C++ Includes
// Other libraries and framework includes
// Project includes
#include "lldb/Breakpoint/Breakpoint.h"
#include "lldb/Breakpoint/BreakpointIDList.h"
#include "lldb/Breakpoint/BreakpointLocation.h"
#include "lldb/Interpreter/Options.h"
#include "lldb/Core/RegularExpression.h"
#include "lldb/Core/StreamString.h"
#include "lldb/Interpreter/CommandInterpreter.h"
#include "lldb/Interpreter/CommandReturnObject.h"
#include "lldb/Target/Target.h"
#include "lldb/Interpreter/CommandCompletions.h"
#include "lldb/Target/StackFrame.h"
#include "lldb/Target/Thread.h"
#include "lldb/Target/ThreadSpec.h"
#include <vector>
using namespace lldb;
using namespace lldb_private;
static void
AddBreakpointDescription (Stream *s, Breakpoint *bp, lldb::DescriptionLevel level)
{
s->IndentMore();
bp->GetDescription (s, level, true);
s->IndentLess();
s->EOL();
}
//-------------------------------------------------------------------------
// CommandObjectBreakpointSet::CommandOptions
//-------------------------------------------------------------------------
#pragma mark Set::CommandOptions
CommandObjectBreakpointSet::CommandOptions::CommandOptions(CommandInterpreter &interpreter) :
Options (interpreter),
m_filename (),
m_line_num (0),
m_column (0),
m_check_inlines (true),
m_func_name (),
Added function name types to allow us to set breakpoints by name more intelligently. The four name types we currently have are: eFunctionNameTypeFull = (1 << 1), // The function name. // For C this is the same as just the name of the function // For C++ this is the demangled version of the mangled name. // For ObjC this is the full function signature with the + or // - and the square brackets and the class and selector eFunctionNameTypeBase = (1 << 2), // The function name only, no namespaces or arguments and no class // methods or selectors will be searched. eFunctionNameTypeMethod = (1 << 3), // Find function by method name (C++) with no namespace or arguments eFunctionNameTypeSelector = (1 << 4) // Find function by selector name (ObjC) names this allows much more flexibility when setting breakoints: (lldb) breakpoint set --name main --basename (lldb) breakpoint set --name main --fullname (lldb) breakpoint set --name main --method (lldb) breakpoint set --name main --selector The default: (lldb) breakpoint set --name main will inspect the name "main" and look for any parens, or if the name starts with "-[" or "+[" and if any are found then a full name search will happen. Else a basename search will be the default. Fixed some command option structures so not all options are required when they shouldn't be. Cleaned up the breakpoint output summary. Made the "image lookup --address <addr>" output much more verbose so it shows all the important symbol context results. Added a GetDescription method to many of the SymbolContext objects for the more verbose output. llvm-svn: 107075
2010-06-29 05:30:43 +08:00
m_func_name_type_mask (0),
m_func_regexp (),
m_modules (),
m_load_addr(),
m_ignore_count (0),
m_thread_id(LLDB_INVALID_THREAD_ID),
m_thread_index (UINT32_MAX),
m_thread_name(),
m_queue_name()
{
}
CommandObjectBreakpointSet::CommandOptions::~CommandOptions ()
{
}
OptionDefinition
CommandObjectBreakpointSet::CommandOptions::g_option_table[] =
{
{ LLDB_OPT_SET_ALL, false, "shlib", 's', required_argument, NULL, CommandCompletions::eModuleCompletion, eArgTypeShlibName,
"Set the breakpoint only in this shared library (can use this option multiple times for multiple shlibs)."},
{ LLDB_OPT_SET_ALL, false, "ignore-count", 'i', required_argument, NULL, 0, eArgTypeCount,
"Set the number of times this breakpoint is skipped before stopping." },
{ LLDB_OPT_SET_ALL, false, "thread-index", 'x', required_argument, NULL, NULL, eArgTypeThreadIndex,
"The breakpoint stops only for the thread whose index matches this argument."},
{ LLDB_OPT_SET_ALL, false, "thread-id", 't', required_argument, NULL, NULL, eArgTypeThreadID,
"The breakpoint stops only for the thread whose TID matches this argument."},
{ LLDB_OPT_SET_ALL, false, "thread-name", 'T', required_argument, NULL, NULL, eArgTypeThreadName,
"The breakpoint stops only for the thread whose thread name matches this argument."},
{ LLDB_OPT_SET_ALL, false, "queue-name", 'q', required_argument, NULL, NULL, eArgTypeQueueName,
"The breakpoint stops only for threads in the queue whose name is given by this argument."},
{ LLDB_OPT_SET_1, false, "file", 'f', required_argument, NULL, CommandCompletions::eSourceFileCompletion, eArgTypeFilename,
"Set the breakpoint by source location in this particular file."},
{ LLDB_OPT_SET_1, true, "line", 'l', required_argument, NULL, 0, eArgTypeLineNum,
"Set the breakpoint by source location at this particular line."},
// Comment out this option for the moment, as we don't actually use it, but will in the future.
// This way users won't see it, but the infrastructure is left in place.
// { 0, false, "column", 'c', required_argument, NULL, "<column>",
// "Set the breakpoint by source location at this particular column."},
{ LLDB_OPT_SET_2, true, "address", 'a', required_argument, NULL, 0, eArgTypeAddress,
"Set the breakpoint by address, at the specified address."},
{ LLDB_OPT_SET_3, true, "name", 'n', required_argument, NULL, CommandCompletions::eSymbolCompletion, eArgTypeFunctionName,
"Set the breakpoint by function name." },
{ LLDB_OPT_SET_4, true, "fullname", 'F', required_argument, NULL, CommandCompletions::eSymbolCompletion, eArgTypeFullName,
"Set the breakpoint by fully qualified function names. For C++ this means namespaces and all arguemnts, and "
"for Objective C this means a full function prototype with class and selector." },
Added function name types to allow us to set breakpoints by name more intelligently. The four name types we currently have are: eFunctionNameTypeFull = (1 << 1), // The function name. // For C this is the same as just the name of the function // For C++ this is the demangled version of the mangled name. // For ObjC this is the full function signature with the + or // - and the square brackets and the class and selector eFunctionNameTypeBase = (1 << 2), // The function name only, no namespaces or arguments and no class // methods or selectors will be searched. eFunctionNameTypeMethod = (1 << 3), // Find function by method name (C++) with no namespace or arguments eFunctionNameTypeSelector = (1 << 4) // Find function by selector name (ObjC) names this allows much more flexibility when setting breakoints: (lldb) breakpoint set --name main --basename (lldb) breakpoint set --name main --fullname (lldb) breakpoint set --name main --method (lldb) breakpoint set --name main --selector The default: (lldb) breakpoint set --name main will inspect the name "main" and look for any parens, or if the name starts with "-[" or "+[" and if any are found then a full name search will happen. Else a basename search will be the default. Fixed some command option structures so not all options are required when they shouldn't be. Cleaned up the breakpoint output summary. Made the "image lookup --address <addr>" output much more verbose so it shows all the important symbol context results. Added a GetDescription method to many of the SymbolContext objects for the more verbose output. llvm-svn: 107075
2010-06-29 05:30:43 +08:00
{ LLDB_OPT_SET_5, true, "selector", 'S', required_argument, NULL, 0, eArgTypeSelector,
"Set the breakpoint by ObjC selector name." },
Added function name types to allow us to set breakpoints by name more intelligently. The four name types we currently have are: eFunctionNameTypeFull = (1 << 1), // The function name. // For C this is the same as just the name of the function // For C++ this is the demangled version of the mangled name. // For ObjC this is the full function signature with the + or // - and the square brackets and the class and selector eFunctionNameTypeBase = (1 << 2), // The function name only, no namespaces or arguments and no class // methods or selectors will be searched. eFunctionNameTypeMethod = (1 << 3), // Find function by method name (C++) with no namespace or arguments eFunctionNameTypeSelector = (1 << 4) // Find function by selector name (ObjC) names this allows much more flexibility when setting breakoints: (lldb) breakpoint set --name main --basename (lldb) breakpoint set --name main --fullname (lldb) breakpoint set --name main --method (lldb) breakpoint set --name main --selector The default: (lldb) breakpoint set --name main will inspect the name "main" and look for any parens, or if the name starts with "-[" or "+[" and if any are found then a full name search will happen. Else a basename search will be the default. Fixed some command option structures so not all options are required when they shouldn't be. Cleaned up the breakpoint output summary. Made the "image lookup --address <addr>" output much more verbose so it shows all the important symbol context results. Added a GetDescription method to many of the SymbolContext objects for the more verbose output. llvm-svn: 107075
2010-06-29 05:30:43 +08:00
{ LLDB_OPT_SET_6, true, "method", 'M', required_argument, NULL, 0, eArgTypeMethod,
"Set the breakpoint by C++ method names." },
Added function name types to allow us to set breakpoints by name more intelligently. The four name types we currently have are: eFunctionNameTypeFull = (1 << 1), // The function name. // For C this is the same as just the name of the function // For C++ this is the demangled version of the mangled name. // For ObjC this is the full function signature with the + or // - and the square brackets and the class and selector eFunctionNameTypeBase = (1 << 2), // The function name only, no namespaces or arguments and no class // methods or selectors will be searched. eFunctionNameTypeMethod = (1 << 3), // Find function by method name (C++) with no namespace or arguments eFunctionNameTypeSelector = (1 << 4) // Find function by selector name (ObjC) names this allows much more flexibility when setting breakoints: (lldb) breakpoint set --name main --basename (lldb) breakpoint set --name main --fullname (lldb) breakpoint set --name main --method (lldb) breakpoint set --name main --selector The default: (lldb) breakpoint set --name main will inspect the name "main" and look for any parens, or if the name starts with "-[" or "+[" and if any are found then a full name search will happen. Else a basename search will be the default. Fixed some command option structures so not all options are required when they shouldn't be. Cleaned up the breakpoint output summary. Made the "image lookup --address <addr>" output much more verbose so it shows all the important symbol context results. Added a GetDescription method to many of the SymbolContext objects for the more verbose output. llvm-svn: 107075
2010-06-29 05:30:43 +08:00
{ LLDB_OPT_SET_7, true, "func-regex", 'r', required_argument, NULL, 0, eArgTypeRegularExpression,
"Set the breakpoint by function name, evaluating a regular-expression to find the function name(s)." },
{ LLDB_OPT_SET_8, true, "basename", 'b', required_argument, NULL, CommandCompletions::eSymbolCompletion, eArgTypeFunctionName,
"Set the breakpoint by function basename (C++ namespaces and arguments will be ignored)." },
{ 0, false, NULL, 0, 0, NULL, 0, eArgTypeNone, NULL }
};
const OptionDefinition*
CommandObjectBreakpointSet::CommandOptions::GetDefinitions ()
{
return g_option_table;
}
Error
Added two new classes for command options: lldb_private::OptionGroup lldb_private::OptionGroupOptions OptionGroup lets you define a class that encapsulates settings that you want to reuse in multiple commands. It contains only the option definitions and the ability to set the option values, but it doesn't directly interface with the lldb_private::Options class that is the front end to all of the CommandObject option parsing. For that the OptionGroupOptions class can be used. It aggregates one or more OptionGroup objects and directs the option setting to the appropriate OptionGroup class. For an example of this, take a look at the CommandObjectFile and how it uses its "m_option_group" object shown below to be able to set values in both the FileOptionGroup and PlatformOptionGroup classes. The members used in CommandObjectFile are: OptionGroupOptions m_option_group; FileOptionGroup m_file_options; PlatformOptionGroup m_platform_options; Then in the constructor for CommandObjectFile you can combine the option settings. The code below shows a simplified version of the constructor: CommandObjectFile::CommandObjectFile(CommandInterpreter &interpreter) : CommandObject (...), m_option_group (interpreter), m_file_options (), m_platform_options(true) { m_option_group.Append (&m_file_options); m_option_group.Append (&m_platform_options); m_option_group.Finalize(); } We append the m_file_options and then the m_platform_options and then tell the option group the finalize the results. This allows the m_option_group to become the organizer of our prefs and after option parsing we end up with valid preference settings in both the m_file_options and m_platform_options objects. This also allows any other commands to use the FileOptionGroup and PlatformOptionGroup classes to implement options for their commands. Renamed: virtual void Options::ResetOptionValues(); to: virtual void Options::OptionParsingStarting(); And implemented a new callback named: virtual Error Options::OptionParsingFinished(); This allows Options subclasses to verify that the options all go together after all of the options have been specified and gives the chance for the command object to return an error. It also gives a chance to take all of the option values and produce or initialize objects after all options have completed parsing. Modfied: virtual Error SetOptionValue (int option_idx, const char *option_arg) = 0; to be: virtual Error SetOptionValue (uint32_t option_idx, const char *option_arg) = 0; (option_idx is now unsigned). llvm-svn: 129415
2011-04-13 08:18:08 +08:00
CommandObjectBreakpointSet::CommandOptions::SetOptionValue (uint32_t option_idx, const char *option_arg)
{
Error error;
char short_option = (char) m_getopt_table[option_idx].val;
switch (short_option)
{
case 'a':
m_load_addr = Args::StringToUInt64(option_arg, LLDB_INVALID_ADDRESS, 0);
if (m_load_addr == LLDB_INVALID_ADDRESS)
m_load_addr = Args::StringToUInt64(option_arg, LLDB_INVALID_ADDRESS, 16);
if (m_load_addr == LLDB_INVALID_ADDRESS)
error.SetErrorStringWithFormat ("Invalid address string '%s'.\n", option_arg);
break;
case 'c':
m_column = Args::StringToUInt32 (option_arg, 0);
break;
Added function name types to allow us to set breakpoints by name more intelligently. The four name types we currently have are: eFunctionNameTypeFull = (1 << 1), // The function name. // For C this is the same as just the name of the function // For C++ this is the demangled version of the mangled name. // For ObjC this is the full function signature with the + or // - and the square brackets and the class and selector eFunctionNameTypeBase = (1 << 2), // The function name only, no namespaces or arguments and no class // methods or selectors will be searched. eFunctionNameTypeMethod = (1 << 3), // Find function by method name (C++) with no namespace or arguments eFunctionNameTypeSelector = (1 << 4) // Find function by selector name (ObjC) names this allows much more flexibility when setting breakoints: (lldb) breakpoint set --name main --basename (lldb) breakpoint set --name main --fullname (lldb) breakpoint set --name main --method (lldb) breakpoint set --name main --selector The default: (lldb) breakpoint set --name main will inspect the name "main" and look for any parens, or if the name starts with "-[" or "+[" and if any are found then a full name search will happen. Else a basename search will be the default. Fixed some command option structures so not all options are required when they shouldn't be. Cleaned up the breakpoint output summary. Made the "image lookup --address <addr>" output much more verbose so it shows all the important symbol context results. Added a GetDescription method to many of the SymbolContext objects for the more verbose output. llvm-svn: 107075
2010-06-29 05:30:43 +08:00
case 'f':
Added the ability to get the min and max instruction byte size for 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
2011-03-27 03:14:58 +08:00
m_filename.assign (option_arg);
break;
Added function name types to allow us to set breakpoints by name more intelligently. The four name types we currently have are: eFunctionNameTypeFull = (1 << 1), // The function name. // For C this is the same as just the name of the function // For C++ this is the demangled version of the mangled name. // For ObjC this is the full function signature with the + or // - and the square brackets and the class and selector eFunctionNameTypeBase = (1 << 2), // The function name only, no namespaces or arguments and no class // methods or selectors will be searched. eFunctionNameTypeMethod = (1 << 3), // Find function by method name (C++) with no namespace or arguments eFunctionNameTypeSelector = (1 << 4) // Find function by selector name (ObjC) names this allows much more flexibility when setting breakoints: (lldb) breakpoint set --name main --basename (lldb) breakpoint set --name main --fullname (lldb) breakpoint set --name main --method (lldb) breakpoint set --name main --selector The default: (lldb) breakpoint set --name main will inspect the name "main" and look for any parens, or if the name starts with "-[" or "+[" and if any are found then a full name search will happen. Else a basename search will be the default. Fixed some command option structures so not all options are required when they shouldn't be. Cleaned up the breakpoint output summary. Made the "image lookup --address <addr>" output much more verbose so it shows all the important symbol context results. Added a GetDescription method to many of the SymbolContext objects for the more verbose output. llvm-svn: 107075
2010-06-29 05:30:43 +08:00
case 'l':
m_line_num = Args::StringToUInt32 (option_arg, 0);
break;
Added function name types to allow us to set breakpoints by name more intelligently. The four name types we currently have are: eFunctionNameTypeFull = (1 << 1), // The function name. // For C this is the same as just the name of the function // For C++ this is the demangled version of the mangled name. // For ObjC this is the full function signature with the + or // - and the square brackets and the class and selector eFunctionNameTypeBase = (1 << 2), // The function name only, no namespaces or arguments and no class // methods or selectors will be searched. eFunctionNameTypeMethod = (1 << 3), // Find function by method name (C++) with no namespace or arguments eFunctionNameTypeSelector = (1 << 4) // Find function by selector name (ObjC) names this allows much more flexibility when setting breakoints: (lldb) breakpoint set --name main --basename (lldb) breakpoint set --name main --fullname (lldb) breakpoint set --name main --method (lldb) breakpoint set --name main --selector The default: (lldb) breakpoint set --name main will inspect the name "main" and look for any parens, or if the name starts with "-[" or "+[" and if any are found then a full name search will happen. Else a basename search will be the default. Fixed some command option structures so not all options are required when they shouldn't be. Cleaned up the breakpoint output summary. Made the "image lookup --address <addr>" output much more verbose so it shows all the important symbol context results. Added a GetDescription method to many of the SymbolContext objects for the more verbose output. llvm-svn: 107075
2010-06-29 05:30:43 +08:00
case 'b':
Added the ability to get the min and max instruction byte size for 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
2011-03-27 03:14:58 +08:00
m_func_name.assign (option_arg);
Added function name types to allow us to set breakpoints by name more intelligently. The four name types we currently have are: eFunctionNameTypeFull = (1 << 1), // The function name. // For C this is the same as just the name of the function // For C++ this is the demangled version of the mangled name. // For ObjC this is the full function signature with the + or // - and the square brackets and the class and selector eFunctionNameTypeBase = (1 << 2), // The function name only, no namespaces or arguments and no class // methods or selectors will be searched. eFunctionNameTypeMethod = (1 << 3), // Find function by method name (C++) with no namespace or arguments eFunctionNameTypeSelector = (1 << 4) // Find function by selector name (ObjC) names this allows much more flexibility when setting breakoints: (lldb) breakpoint set --name main --basename (lldb) breakpoint set --name main --fullname (lldb) breakpoint set --name main --method (lldb) breakpoint set --name main --selector The default: (lldb) breakpoint set --name main will inspect the name "main" and look for any parens, or if the name starts with "-[" or "+[" and if any are found then a full name search will happen. Else a basename search will be the default. Fixed some command option structures so not all options are required when they shouldn't be. Cleaned up the breakpoint output summary. Made the "image lookup --address <addr>" output much more verbose so it shows all the important symbol context results. Added a GetDescription method to many of the SymbolContext objects for the more verbose output. llvm-svn: 107075
2010-06-29 05:30:43 +08:00
m_func_name_type_mask |= eFunctionNameTypeBase;
break;
case 'n':
Added the ability to get the min and max instruction byte size for 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
2011-03-27 03:14:58 +08:00
m_func_name.assign (option_arg);
m_func_name_type_mask |= eFunctionNameTypeAuto;
break;
Added function name types to allow us to set breakpoints by name more intelligently. The four name types we currently have are: eFunctionNameTypeFull = (1 << 1), // The function name. // For C this is the same as just the name of the function // For C++ this is the demangled version of the mangled name. // For ObjC this is the full function signature with the + or // - and the square brackets and the class and selector eFunctionNameTypeBase = (1 << 2), // The function name only, no namespaces or arguments and no class // methods or selectors will be searched. eFunctionNameTypeMethod = (1 << 3), // Find function by method name (C++) with no namespace or arguments eFunctionNameTypeSelector = (1 << 4) // Find function by selector name (ObjC) names this allows much more flexibility when setting breakoints: (lldb) breakpoint set --name main --basename (lldb) breakpoint set --name main --fullname (lldb) breakpoint set --name main --method (lldb) breakpoint set --name main --selector The default: (lldb) breakpoint set --name main will inspect the name "main" and look for any parens, or if the name starts with "-[" or "+[" and if any are found then a full name search will happen. Else a basename search will be the default. Fixed some command option structures so not all options are required when they shouldn't be. Cleaned up the breakpoint output summary. Made the "image lookup --address <addr>" output much more verbose so it shows all the important symbol context results. Added a GetDescription method to many of the SymbolContext objects for the more verbose output. llvm-svn: 107075
2010-06-29 05:30:43 +08:00
case 'F':
Added the ability to get the min and max instruction byte size for 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
2011-03-27 03:14:58 +08:00
m_func_name.assign (option_arg);
Added function name types to allow us to set breakpoints by name more intelligently. The four name types we currently have are: eFunctionNameTypeFull = (1 << 1), // The function name. // For C this is the same as just the name of the function // For C++ this is the demangled version of the mangled name. // For ObjC this is the full function signature with the + or // - and the square brackets and the class and selector eFunctionNameTypeBase = (1 << 2), // The function name only, no namespaces or arguments and no class // methods or selectors will be searched. eFunctionNameTypeMethod = (1 << 3), // Find function by method name (C++) with no namespace or arguments eFunctionNameTypeSelector = (1 << 4) // Find function by selector name (ObjC) names this allows much more flexibility when setting breakoints: (lldb) breakpoint set --name main --basename (lldb) breakpoint set --name main --fullname (lldb) breakpoint set --name main --method (lldb) breakpoint set --name main --selector The default: (lldb) breakpoint set --name main will inspect the name "main" and look for any parens, or if the name starts with "-[" or "+[" and if any are found then a full name search will happen. Else a basename search will be the default. Fixed some command option structures so not all options are required when they shouldn't be. Cleaned up the breakpoint output summary. Made the "image lookup --address <addr>" output much more verbose so it shows all the important symbol context results. Added a GetDescription method to many of the SymbolContext objects for the more verbose output. llvm-svn: 107075
2010-06-29 05:30:43 +08:00
m_func_name_type_mask |= eFunctionNameTypeFull;
break;
case 'S':
Added the ability to get the min and max instruction byte size for 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
2011-03-27 03:14:58 +08:00
m_func_name.assign (option_arg);
Added function name types to allow us to set breakpoints by name more intelligently. The four name types we currently have are: eFunctionNameTypeFull = (1 << 1), // The function name. // For C this is the same as just the name of the function // For C++ this is the demangled version of the mangled name. // For ObjC this is the full function signature with the + or // - and the square brackets and the class and selector eFunctionNameTypeBase = (1 << 2), // The function name only, no namespaces or arguments and no class // methods or selectors will be searched. eFunctionNameTypeMethod = (1 << 3), // Find function by method name (C++) with no namespace or arguments eFunctionNameTypeSelector = (1 << 4) // Find function by selector name (ObjC) names this allows much more flexibility when setting breakoints: (lldb) breakpoint set --name main --basename (lldb) breakpoint set --name main --fullname (lldb) breakpoint set --name main --method (lldb) breakpoint set --name main --selector The default: (lldb) breakpoint set --name main will inspect the name "main" and look for any parens, or if the name starts with "-[" or "+[" and if any are found then a full name search will happen. Else a basename search will be the default. Fixed some command option structures so not all options are required when they shouldn't be. Cleaned up the breakpoint output summary. Made the "image lookup --address <addr>" output much more verbose so it shows all the important symbol context results. Added a GetDescription method to many of the SymbolContext objects for the more verbose output. llvm-svn: 107075
2010-06-29 05:30:43 +08:00
m_func_name_type_mask |= eFunctionNameTypeSelector;
break;
case 'M':
Added the ability to get the min and max instruction byte size for 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
2011-03-27 03:14:58 +08:00
m_func_name.assign (option_arg);
Added function name types to allow us to set breakpoints by name more intelligently. The four name types we currently have are: eFunctionNameTypeFull = (1 << 1), // The function name. // For C this is the same as just the name of the function // For C++ this is the demangled version of the mangled name. // For ObjC this is the full function signature with the + or // - and the square brackets and the class and selector eFunctionNameTypeBase = (1 << 2), // The function name only, no namespaces or arguments and no class // methods or selectors will be searched. eFunctionNameTypeMethod = (1 << 3), // Find function by method name (C++) with no namespace or arguments eFunctionNameTypeSelector = (1 << 4) // Find function by selector name (ObjC) names this allows much more flexibility when setting breakoints: (lldb) breakpoint set --name main --basename (lldb) breakpoint set --name main --fullname (lldb) breakpoint set --name main --method (lldb) breakpoint set --name main --selector The default: (lldb) breakpoint set --name main will inspect the name "main" and look for any parens, or if the name starts with "-[" or "+[" and if any are found then a full name search will happen. Else a basename search will be the default. Fixed some command option structures so not all options are required when they shouldn't be. Cleaned up the breakpoint output summary. Made the "image lookup --address <addr>" output much more verbose so it shows all the important symbol context results. Added a GetDescription method to many of the SymbolContext objects for the more verbose output. llvm-svn: 107075
2010-06-29 05:30:43 +08:00
m_func_name_type_mask |= eFunctionNameTypeMethod;
break;
case 'r':
Added the ability to get the min and max instruction byte size for 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
2011-03-27 03:14:58 +08:00
m_func_regexp.assign (option_arg);
break;
Added function name types to allow us to set breakpoints by name more intelligently. The four name types we currently have are: eFunctionNameTypeFull = (1 << 1), // The function name. // For C this is the same as just the name of the function // For C++ this is the demangled version of the mangled name. // For ObjC this is the full function signature with the + or // - and the square brackets and the class and selector eFunctionNameTypeBase = (1 << 2), // The function name only, no namespaces or arguments and no class // methods or selectors will be searched. eFunctionNameTypeMethod = (1 << 3), // Find function by method name (C++) with no namespace or arguments eFunctionNameTypeSelector = (1 << 4) // Find function by selector name (ObjC) names this allows much more flexibility when setting breakoints: (lldb) breakpoint set --name main --basename (lldb) breakpoint set --name main --fullname (lldb) breakpoint set --name main --method (lldb) breakpoint set --name main --selector The default: (lldb) breakpoint set --name main will inspect the name "main" and look for any parens, or if the name starts with "-[" or "+[" and if any are found then a full name search will happen. Else a basename search will be the default. Fixed some command option structures so not all options are required when they shouldn't be. Cleaned up the breakpoint output summary. Made the "image lookup --address <addr>" output much more verbose so it shows all the important symbol context results. Added a GetDescription method to many of the SymbolContext objects for the more verbose output. llvm-svn: 107075
2010-06-29 05:30:43 +08:00
case 's':
{
m_modules.push_back (std::string (option_arg));
break;
}
case 'i':
{
m_ignore_count = Args::StringToUInt32(option_arg, UINT32_MAX, 0);
if (m_ignore_count == UINT32_MAX)
error.SetErrorStringWithFormat ("Invalid ignore count '%s'.\n", option_arg);
}
break;
case 't' :
{
m_thread_id = Args::StringToUInt64(option_arg, LLDB_INVALID_THREAD_ID, 0);
if (m_thread_id == LLDB_INVALID_THREAD_ID)
error.SetErrorStringWithFormat ("Invalid thread id string '%s'.\n", option_arg);
}
break;
case 'T':
Added the ability to get the min and max instruction byte size for 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
2011-03-27 03:14:58 +08:00
m_thread_name.assign (option_arg);
break;
case 'q':
Added the ability to get the min and max instruction byte size for 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
2011-03-27 03:14:58 +08:00
m_queue_name.assign (option_arg);
break;
case 'x':
{
m_thread_index = Args::StringToUInt32(option_arg, UINT32_MAX, 0);
if (m_thread_id == UINT32_MAX)
error.SetErrorStringWithFormat ("Invalid thread index string '%s'.\n", option_arg);
}
break;
default:
error.SetErrorStringWithFormat ("Unrecognized option '%c'.\n", short_option);
break;
}
return error;
}
void
Added two new classes for command options: lldb_private::OptionGroup lldb_private::OptionGroupOptions OptionGroup lets you define a class that encapsulates settings that you want to reuse in multiple commands. It contains only the option definitions and the ability to set the option values, but it doesn't directly interface with the lldb_private::Options class that is the front end to all of the CommandObject option parsing. For that the OptionGroupOptions class can be used. It aggregates one or more OptionGroup objects and directs the option setting to the appropriate OptionGroup class. For an example of this, take a look at the CommandObjectFile and how it uses its "m_option_group" object shown below to be able to set values in both the FileOptionGroup and PlatformOptionGroup classes. The members used in CommandObjectFile are: OptionGroupOptions m_option_group; FileOptionGroup m_file_options; PlatformOptionGroup m_platform_options; Then in the constructor for CommandObjectFile you can combine the option settings. The code below shows a simplified version of the constructor: CommandObjectFile::CommandObjectFile(CommandInterpreter &interpreter) : CommandObject (...), m_option_group (interpreter), m_file_options (), m_platform_options(true) { m_option_group.Append (&m_file_options); m_option_group.Append (&m_platform_options); m_option_group.Finalize(); } We append the m_file_options and then the m_platform_options and then tell the option group the finalize the results. This allows the m_option_group to become the organizer of our prefs and after option parsing we end up with valid preference settings in both the m_file_options and m_platform_options objects. This also allows any other commands to use the FileOptionGroup and PlatformOptionGroup classes to implement options for their commands. Renamed: virtual void Options::ResetOptionValues(); to: virtual void Options::OptionParsingStarting(); And implemented a new callback named: virtual Error Options::OptionParsingFinished(); This allows Options subclasses to verify that the options all go together after all of the options have been specified and gives the chance for the command object to return an error. It also gives a chance to take all of the option values and produce or initialize objects after all options have completed parsing. Modfied: virtual Error SetOptionValue (int option_idx, const char *option_arg) = 0; to be: virtual Error SetOptionValue (uint32_t option_idx, const char *option_arg) = 0; (option_idx is now unsigned). llvm-svn: 129415
2011-04-13 08:18:08 +08:00
CommandObjectBreakpointSet::CommandOptions::OptionParsingStarting ()
{
m_filename.clear();
m_line_num = 0;
m_column = 0;
m_func_name.clear();
Added function name types to allow us to set breakpoints by name more intelligently. The four name types we currently have are: eFunctionNameTypeFull = (1 << 1), // The function name. // For C this is the same as just the name of the function // For C++ this is the demangled version of the mangled name. // For ObjC this is the full function signature with the + or // - and the square brackets and the class and selector eFunctionNameTypeBase = (1 << 2), // The function name only, no namespaces or arguments and no class // methods or selectors will be searched. eFunctionNameTypeMethod = (1 << 3), // Find function by method name (C++) with no namespace or arguments eFunctionNameTypeSelector = (1 << 4) // Find function by selector name (ObjC) names this allows much more flexibility when setting breakoints: (lldb) breakpoint set --name main --basename (lldb) breakpoint set --name main --fullname (lldb) breakpoint set --name main --method (lldb) breakpoint set --name main --selector The default: (lldb) breakpoint set --name main will inspect the name "main" and look for any parens, or if the name starts with "-[" or "+[" and if any are found then a full name search will happen. Else a basename search will be the default. Fixed some command option structures so not all options are required when they shouldn't be. Cleaned up the breakpoint output summary. Made the "image lookup --address <addr>" output much more verbose so it shows all the important symbol context results. Added a GetDescription method to many of the SymbolContext objects for the more verbose output. llvm-svn: 107075
2010-06-29 05:30:43 +08:00
m_func_name_type_mask = 0;
m_func_regexp.clear();
m_load_addr = LLDB_INVALID_ADDRESS;
m_modules.clear();
m_ignore_count = 0;
m_thread_id = LLDB_INVALID_THREAD_ID;
m_thread_index = UINT32_MAX;
m_thread_name.clear();
m_queue_name.clear();
}
//-------------------------------------------------------------------------
// CommandObjectBreakpointSet
//-------------------------------------------------------------------------
#pragma mark Set
CommandObjectBreakpointSet::CommandObjectBreakpointSet (CommandInterpreter &interpreter) :
CommandObject (interpreter,
"breakpoint set",
"Sets a breakpoint or set of breakpoints in the executable.",
"breakpoint set <cmd-options>"),
m_options (interpreter)
{
}
CommandObjectBreakpointSet::~CommandObjectBreakpointSet ()
{
}
Options *
CommandObjectBreakpointSet::GetOptions ()
{
return &m_options;
}
bool
CommandObjectBreakpointSet::Execute
(
Args& command,
CommandReturnObject &result
)
{
Target *target = m_interpreter.GetDebugger().GetSelectedTarget().get();
if (target == NULL)
{
result.AppendError ("Invalid target. Must set target before setting breakpoints (see 'target create' command).");
result.SetStatus (eReturnStatusFailed);
return false;
}
// The following are the various types of breakpoints that could be set:
// 1). -f -l -p [-s -g] (setting breakpoint by source location)
// 2). -a [-s -g] (setting breakpoint by address)
// 3). -n [-s -g] (setting breakpoint by function name)
// 4). -r [-s -g] (setting breakpoint by function name regular expression)
BreakpointSetType break_type = eSetTypeInvalid;
if (m_options.m_line_num != 0)
break_type = eSetTypeFileAndLine;
else if (m_options.m_load_addr != LLDB_INVALID_ADDRESS)
break_type = eSetTypeAddress;
else if (!m_options.m_func_name.empty())
break_type = eSetTypeFunctionName;
else if (!m_options.m_func_regexp.empty())
break_type = eSetTypeFunctionRegexp;
Breakpoint *bp = NULL;
FileSpec module_spec;
bool use_module = false;
int num_modules = m_options.m_modules.size();
if ((num_modules > 0) && (break_type != eSetTypeAddress))
use_module = true;
switch (break_type)
{
case eSetTypeFileAndLine: // Breakpoint by source position
{
FileSpec file;
if (m_options.m_filename.empty())
{
uint32_t default_line;
// First use the Source Manager's default file.
// Then use the current stack frame's file.
if (!target->GetSourceManager().GetDefaultFileAndLine(file, default_line))
{
StackFrame *cur_frame = m_interpreter.GetExecutionContext().frame;
if (cur_frame == NULL)
{
result.AppendError ("Attempting to set breakpoint by line number alone with no selected frame.");
result.SetStatus (eReturnStatusFailed);
break;
}
else if (!cur_frame->HasDebugInformation())
{
result.AppendError ("Attempting to set breakpoint by line number alone but selected frame has no debug info.");
result.SetStatus (eReturnStatusFailed);
break;
}
else
{
const SymbolContext &sc = cur_frame->GetSymbolContext (eSymbolContextLineEntry);
if (sc.line_entry.file)
{
file = sc.line_entry.file;
}
else
{
result.AppendError ("Attempting to set breakpoint by line number alone but can't find the file for the selected frame.");
result.SetStatus (eReturnStatusFailed);
break;
}
}
}
}
else
{
file.SetFile(m_options.m_filename.c_str(), false);
}
if (use_module)
{
for (int i = 0; i < num_modules; ++i)
{
module_spec.SetFile(m_options.m_modules[i].c_str(), false);
bp = target->CreateBreakpoint (&module_spec,
file,
m_options.m_line_num,
m_options.m_check_inlines).get();
if (bp)
{
Stream &output_stream = result.GetOutputStream();
result.AppendMessage ("Breakpoint created: ");
bp->GetDescription(&output_stream, lldb::eDescriptionLevelBrief);
output_stream.EOL();
if (bp->GetNumLocations() == 0)
output_stream.Printf ("WARNING: Unable to resolve breakpoint to any actual"
" locations.\n");
result.SetStatus (eReturnStatusSuccessFinishResult);
}
else
{
result.AppendErrorWithFormat("Breakpoint creation failed: No breakpoint created in module '%s'.\n",
m_options.m_modules[i].c_str());
result.SetStatus (eReturnStatusFailed);
}
}
}
else
bp = target->CreateBreakpoint (NULL,
file,
m_options.m_line_num,
m_options.m_check_inlines).get();
}
break;
case eSetTypeAddress: // Breakpoint by address
bp = target->CreateBreakpoint (m_options.m_load_addr, false).get();
break;
Added function name types to allow us to set breakpoints by name more intelligently. The four name types we currently have are: eFunctionNameTypeFull = (1 << 1), // The function name. // For C this is the same as just the name of the function // For C++ this is the demangled version of the mangled name. // For ObjC this is the full function signature with the + or // - and the square brackets and the class and selector eFunctionNameTypeBase = (1 << 2), // The function name only, no namespaces or arguments and no class // methods or selectors will be searched. eFunctionNameTypeMethod = (1 << 3), // Find function by method name (C++) with no namespace or arguments eFunctionNameTypeSelector = (1 << 4) // Find function by selector name (ObjC) names this allows much more flexibility when setting breakoints: (lldb) breakpoint set --name main --basename (lldb) breakpoint set --name main --fullname (lldb) breakpoint set --name main --method (lldb) breakpoint set --name main --selector The default: (lldb) breakpoint set --name main will inspect the name "main" and look for any parens, or if the name starts with "-[" or "+[" and if any are found then a full name search will happen. Else a basename search will be the default. Fixed some command option structures so not all options are required when they shouldn't be. Cleaned up the breakpoint output summary. Made the "image lookup --address <addr>" output much more verbose so it shows all the important symbol context results. Added a GetDescription method to many of the SymbolContext objects for the more verbose output. llvm-svn: 107075
2010-06-29 05:30:43 +08:00
case eSetTypeFunctionName: // Breakpoint by function name
{
Added function name types to allow us to set breakpoints by name more intelligently. The four name types we currently have are: eFunctionNameTypeFull = (1 << 1), // The function name. // For C this is the same as just the name of the function // For C++ this is the demangled version of the mangled name. // For ObjC this is the full function signature with the + or // - and the square brackets and the class and selector eFunctionNameTypeBase = (1 << 2), // The function name only, no namespaces or arguments and no class // methods or selectors will be searched. eFunctionNameTypeMethod = (1 << 3), // Find function by method name (C++) with no namespace or arguments eFunctionNameTypeSelector = (1 << 4) // Find function by selector name (ObjC) names this allows much more flexibility when setting breakoints: (lldb) breakpoint set --name main --basename (lldb) breakpoint set --name main --fullname (lldb) breakpoint set --name main --method (lldb) breakpoint set --name main --selector The default: (lldb) breakpoint set --name main will inspect the name "main" and look for any parens, or if the name starts with "-[" or "+[" and if any are found then a full name search will happen. Else a basename search will be the default. Fixed some command option structures so not all options are required when they shouldn't be. Cleaned up the breakpoint output summary. Made the "image lookup --address <addr>" output much more verbose so it shows all the important symbol context results. Added a GetDescription method to many of the SymbolContext objects for the more verbose output. llvm-svn: 107075
2010-06-29 05:30:43 +08:00
uint32_t name_type_mask = m_options.m_func_name_type_mask;
if (name_type_mask == 0)
name_type_mask = eFunctionNameTypeAuto;
Added function name types to allow us to set breakpoints by name more intelligently. The four name types we currently have are: eFunctionNameTypeFull = (1 << 1), // The function name. // For C this is the same as just the name of the function // For C++ this is the demangled version of the mangled name. // For ObjC this is the full function signature with the + or // - and the square brackets and the class and selector eFunctionNameTypeBase = (1 << 2), // The function name only, no namespaces or arguments and no class // methods or selectors will be searched. eFunctionNameTypeMethod = (1 << 3), // Find function by method name (C++) with no namespace or arguments eFunctionNameTypeSelector = (1 << 4) // Find function by selector name (ObjC) names this allows much more flexibility when setting breakoints: (lldb) breakpoint set --name main --basename (lldb) breakpoint set --name main --fullname (lldb) breakpoint set --name main --method (lldb) breakpoint set --name main --selector The default: (lldb) breakpoint set --name main will inspect the name "main" and look for any parens, or if the name starts with "-[" or "+[" and if any are found then a full name search will happen. Else a basename search will be the default. Fixed some command option structures so not all options are required when they shouldn't be. Cleaned up the breakpoint output summary. Made the "image lookup --address <addr>" output much more verbose so it shows all the important symbol context results. Added a GetDescription method to many of the SymbolContext objects for the more verbose output. llvm-svn: 107075
2010-06-29 05:30:43 +08:00
if (use_module)
{
for (int i = 0; i < num_modules; ++i)
{
module_spec.SetFile(m_options.m_modules[i].c_str(), false);
bp = target->CreateBreakpoint (&module_spec,
m_options.m_func_name.c_str(),
name_type_mask,
Breakpoint::Exact).get();
Added function name types to allow us to set breakpoints by name more intelligently. The four name types we currently have are: eFunctionNameTypeFull = (1 << 1), // The function name. // For C this is the same as just the name of the function // For C++ this is the demangled version of the mangled name. // For ObjC this is the full function signature with the + or // - and the square brackets and the class and selector eFunctionNameTypeBase = (1 << 2), // The function name only, no namespaces or arguments and no class // methods or selectors will be searched. eFunctionNameTypeMethod = (1 << 3), // Find function by method name (C++) with no namespace or arguments eFunctionNameTypeSelector = (1 << 4) // Find function by selector name (ObjC) names this allows much more flexibility when setting breakoints: (lldb) breakpoint set --name main --basename (lldb) breakpoint set --name main --fullname (lldb) breakpoint set --name main --method (lldb) breakpoint set --name main --selector The default: (lldb) breakpoint set --name main will inspect the name "main" and look for any parens, or if the name starts with "-[" or "+[" and if any are found then a full name search will happen. Else a basename search will be the default. Fixed some command option structures so not all options are required when they shouldn't be. Cleaned up the breakpoint output summary. Made the "image lookup --address <addr>" output much more verbose so it shows all the important symbol context results. Added a GetDescription method to many of the SymbolContext objects for the more verbose output. llvm-svn: 107075
2010-06-29 05:30:43 +08:00
if (bp)
{
Stream &output_stream = result.GetOutputStream();
Added function name types to allow us to set breakpoints by name more intelligently. The four name types we currently have are: eFunctionNameTypeFull = (1 << 1), // The function name. // For C this is the same as just the name of the function // For C++ this is the demangled version of the mangled name. // For ObjC this is the full function signature with the + or // - and the square brackets and the class and selector eFunctionNameTypeBase = (1 << 2), // The function name only, no namespaces or arguments and no class // methods or selectors will be searched. eFunctionNameTypeMethod = (1 << 3), // Find function by method name (C++) with no namespace or arguments eFunctionNameTypeSelector = (1 << 4) // Find function by selector name (ObjC) names this allows much more flexibility when setting breakoints: (lldb) breakpoint set --name main --basename (lldb) breakpoint set --name main --fullname (lldb) breakpoint set --name main --method (lldb) breakpoint set --name main --selector The default: (lldb) breakpoint set --name main will inspect the name "main" and look for any parens, or if the name starts with "-[" or "+[" and if any are found then a full name search will happen. Else a basename search will be the default. Fixed some command option structures so not all options are required when they shouldn't be. Cleaned up the breakpoint output summary. Made the "image lookup --address <addr>" output much more verbose so it shows all the important symbol context results. Added a GetDescription method to many of the SymbolContext objects for the more verbose output. llvm-svn: 107075
2010-06-29 05:30:43 +08:00
output_stream.Printf ("Breakpoint created: ");
bp->GetDescription(&output_stream, lldb::eDescriptionLevelBrief);
output_stream.EOL();
if (bp->GetNumLocations() == 0)
output_stream.Printf ("WARNING: Unable to resolve breakpoint to any actual"
" locations.\n");
Added function name types to allow us to set breakpoints by name more intelligently. The four name types we currently have are: eFunctionNameTypeFull = (1 << 1), // The function name. // For C this is the same as just the name of the function // For C++ this is the demangled version of the mangled name. // For ObjC this is the full function signature with the + or // - and the square brackets and the class and selector eFunctionNameTypeBase = (1 << 2), // The function name only, no namespaces or arguments and no class // methods or selectors will be searched. eFunctionNameTypeMethod = (1 << 3), // Find function by method name (C++) with no namespace or arguments eFunctionNameTypeSelector = (1 << 4) // Find function by selector name (ObjC) names this allows much more flexibility when setting breakoints: (lldb) breakpoint set --name main --basename (lldb) breakpoint set --name main --fullname (lldb) breakpoint set --name main --method (lldb) breakpoint set --name main --selector The default: (lldb) breakpoint set --name main will inspect the name "main" and look for any parens, or if the name starts with "-[" or "+[" and if any are found then a full name search will happen. Else a basename search will be the default. Fixed some command option structures so not all options are required when they shouldn't be. Cleaned up the breakpoint output summary. Made the "image lookup --address <addr>" output much more verbose so it shows all the important symbol context results. Added a GetDescription method to many of the SymbolContext objects for the more verbose output. llvm-svn: 107075
2010-06-29 05:30:43 +08:00
result.SetStatus (eReturnStatusSuccessFinishResult);
}
else
{
result.AppendErrorWithFormat("Breakpoint creation failed: No breakpoint created in module '%s'.\n",
m_options.m_modules[i].c_str());
result.SetStatus (eReturnStatusFailed);
}
}
}
Added function name types to allow us to set breakpoints by name more intelligently. The four name types we currently have are: eFunctionNameTypeFull = (1 << 1), // The function name. // For C this is the same as just the name of the function // For C++ this is the demangled version of the mangled name. // For ObjC this is the full function signature with the + or // - and the square brackets and the class and selector eFunctionNameTypeBase = (1 << 2), // The function name only, no namespaces or arguments and no class // methods or selectors will be searched. eFunctionNameTypeMethod = (1 << 3), // Find function by method name (C++) with no namespace or arguments eFunctionNameTypeSelector = (1 << 4) // Find function by selector name (ObjC) names this allows much more flexibility when setting breakoints: (lldb) breakpoint set --name main --basename (lldb) breakpoint set --name main --fullname (lldb) breakpoint set --name main --method (lldb) breakpoint set --name main --selector The default: (lldb) breakpoint set --name main will inspect the name "main" and look for any parens, or if the name starts with "-[" or "+[" and if any are found then a full name search will happen. Else a basename search will be the default. Fixed some command option structures so not all options are required when they shouldn't be. Cleaned up the breakpoint output summary. Made the "image lookup --address <addr>" output much more verbose so it shows all the important symbol context results. Added a GetDescription method to many of the SymbolContext objects for the more verbose output. llvm-svn: 107075
2010-06-29 05:30:43 +08:00
else
bp = target->CreateBreakpoint (NULL, m_options.m_func_name.c_str(), name_type_mask, Breakpoint::Exact).get();
}
break;
Added function name types to allow us to set breakpoints by name more intelligently. The four name types we currently have are: eFunctionNameTypeFull = (1 << 1), // The function name. // For C this is the same as just the name of the function // For C++ this is the demangled version of the mangled name. // For ObjC this is the full function signature with the + or // - and the square brackets and the class and selector eFunctionNameTypeBase = (1 << 2), // The function name only, no namespaces or arguments and no class // methods or selectors will be searched. eFunctionNameTypeMethod = (1 << 3), // Find function by method name (C++) with no namespace or arguments eFunctionNameTypeSelector = (1 << 4) // Find function by selector name (ObjC) names this allows much more flexibility when setting breakoints: (lldb) breakpoint set --name main --basename (lldb) breakpoint set --name main --fullname (lldb) breakpoint set --name main --method (lldb) breakpoint set --name main --selector The default: (lldb) breakpoint set --name main will inspect the name "main" and look for any parens, or if the name starts with "-[" or "+[" and if any are found then a full name search will happen. Else a basename search will be the default. Fixed some command option structures so not all options are required when they shouldn't be. Cleaned up the breakpoint output summary. Made the "image lookup --address <addr>" output much more verbose so it shows all the important symbol context results. Added a GetDescription method to many of the SymbolContext objects for the more verbose output. llvm-svn: 107075
2010-06-29 05:30:43 +08:00
case eSetTypeFunctionRegexp: // Breakpoint by regular expression function name
{
RegularExpression regexp(m_options.m_func_regexp.c_str());
if (use_module)
{
for (int i = 0; i < num_modules; ++i)
{
module_spec.SetFile(m_options.m_modules[i].c_str(), false);
bp = target->CreateBreakpoint (&module_spec, regexp).get();
if (bp)
{
Stream &output_stream = result.GetOutputStream();
output_stream.Printf ("Breakpoint created: ");
bp->GetDescription(&output_stream, lldb::eDescriptionLevelBrief);
output_stream.EOL();
if (bp->GetNumLocations() == 0)
output_stream.Printf ("WARNING: Unable to resolve breakpoint to any actual"
" locations.\n");
result.SetStatus (eReturnStatusSuccessFinishResult);
}
else
{
result.AppendErrorWithFormat("Breakpoint creation failed: No breakpoint created in module '%s'.\n",
m_options.m_modules[i].c_str());
result.SetStatus (eReturnStatusFailed);
}
}
}
else
bp = target->CreateBreakpoint (NULL, regexp).get();
}
break;
Added function name types to allow us to set breakpoints by name more intelligently. The four name types we currently have are: eFunctionNameTypeFull = (1 << 1), // The function name. // For C this is the same as just the name of the function // For C++ this is the demangled version of the mangled name. // For ObjC this is the full function signature with the + or // - and the square brackets and the class and selector eFunctionNameTypeBase = (1 << 2), // The function name only, no namespaces or arguments and no class // methods or selectors will be searched. eFunctionNameTypeMethod = (1 << 3), // Find function by method name (C++) with no namespace or arguments eFunctionNameTypeSelector = (1 << 4) // Find function by selector name (ObjC) names this allows much more flexibility when setting breakoints: (lldb) breakpoint set --name main --basename (lldb) breakpoint set --name main --fullname (lldb) breakpoint set --name main --method (lldb) breakpoint set --name main --selector The default: (lldb) breakpoint set --name main will inspect the name "main" and look for any parens, or if the name starts with "-[" or "+[" and if any are found then a full name search will happen. Else a basename search will be the default. Fixed some command option structures so not all options are required when they shouldn't be. Cleaned up the breakpoint output summary. Made the "image lookup --address <addr>" output much more verbose so it shows all the important symbol context results. Added a GetDescription method to many of the SymbolContext objects for the more verbose output. llvm-svn: 107075
2010-06-29 05:30:43 +08:00
default:
break;
}
// Now set the various options that were passed in:
if (bp)
{
if (m_options.m_thread_id != LLDB_INVALID_THREAD_ID)
bp->SetThreadID (m_options.m_thread_id);
if (m_options.m_thread_index != UINT32_MAX)
bp->GetOptions()->GetThreadSpec()->SetIndex(m_options.m_thread_index);
if (!m_options.m_thread_name.empty())
bp->GetOptions()->GetThreadSpec()->SetName(m_options.m_thread_name.c_str());
if (!m_options.m_queue_name.empty())
bp->GetOptions()->GetThreadSpec()->SetQueueName(m_options.m_queue_name.c_str());
if (m_options.m_ignore_count != 0)
bp->GetOptions()->SetIgnoreCount(m_options.m_ignore_count);
}
if (bp && !use_module)
{
Stream &output_stream = result.GetOutputStream();
output_stream.Printf ("Breakpoint created: ");
bp->GetDescription(&output_stream, lldb::eDescriptionLevelBrief);
output_stream.EOL();
if (bp->GetNumLocations() == 0)
output_stream.Printf ("WARNING: Unable to resolve breakpoint to any actual locations.\n");
result.SetStatus (eReturnStatusSuccessFinishResult);
}
else if (!bp)
{
result.AppendError ("Breakpoint creation failed: No breakpoint created.");
result.SetStatus (eReturnStatusFailed);
}
return result.Succeeded();
}
//-------------------------------------------------------------------------
// CommandObjectMultiwordBreakpoint
//-------------------------------------------------------------------------
#pragma mark MultiwordBreakpoint
CommandObjectMultiwordBreakpoint::CommandObjectMultiwordBreakpoint (CommandInterpreter &interpreter) :
CommandObjectMultiword (interpreter,
"breakpoint",
"A set of commands for operating on breakpoints. Also see _regexp-break.",
"breakpoint <command> [<command-options>]")
{
bool status;
CommandObjectSP list_command_object (new CommandObjectBreakpointList (interpreter));
CommandObjectSP enable_command_object (new CommandObjectBreakpointEnable (interpreter));
CommandObjectSP disable_command_object (new CommandObjectBreakpointDisable (interpreter));
CommandObjectSP clear_command_object (new CommandObjectBreakpointClear (interpreter));
CommandObjectSP delete_command_object (new CommandObjectBreakpointDelete (interpreter));
CommandObjectSP set_command_object (new CommandObjectBreakpointSet (interpreter));
CommandObjectSP command_command_object (new CommandObjectBreakpointCommand (interpreter));
CommandObjectSP modify_command_object (new CommandObjectBreakpointModify(interpreter));
list_command_object->SetCommandName ("breakpoint list");
enable_command_object->SetCommandName("breakpoint enable");
disable_command_object->SetCommandName("breakpoint disable");
clear_command_object->SetCommandName("breakpoint clear");
delete_command_object->SetCommandName("breakpoint delete");
set_command_object->SetCommandName("breakpoint set");
command_command_object->SetCommandName ("breakpoint command");
modify_command_object->SetCommandName ("breakpoint modify");
status = LoadSubCommand ("list", list_command_object);
status = LoadSubCommand ("enable", enable_command_object);
status = LoadSubCommand ("disable", disable_command_object);
status = LoadSubCommand ("clear", clear_command_object);
status = LoadSubCommand ("delete", delete_command_object);
status = LoadSubCommand ("set", set_command_object);
status = LoadSubCommand ("command", command_command_object);
status = LoadSubCommand ("modify", modify_command_object);
}
CommandObjectMultiwordBreakpoint::~CommandObjectMultiwordBreakpoint ()
{
}
void
CommandObjectMultiwordBreakpoint::VerifyBreakpointIDs (Args &args, Target *target, CommandReturnObject &result,
BreakpointIDList *valid_ids)
{
// args can be strings representing 1). integers (for breakpoint ids)
// 2). the full breakpoint & location canonical representation
// 3). the word "to" or a hyphen, representing a range (in which case there
// had *better* be an entry both before & after of one of the first two types.
// If args is empty, we will use the last created breakpoint (if there is one.)
Args temp_args;
if (args.GetArgumentCount() == 0)
{
if (target->GetLastCreatedBreakpoint())
{
valid_ids->AddBreakpointID (BreakpointID(target->GetLastCreatedBreakpoint()->GetID(), LLDB_INVALID_BREAK_ID));
result.SetStatus (eReturnStatusSuccessFinishNoResult);
}
else
{
result.AppendError("No breakpoint specified and no last created breakpoint.");
result.SetStatus (eReturnStatusFailed);
}
return;
}
// Create a new Args variable to use; copy any non-breakpoint-id-ranges stuff directly from the old ARGS to
// the new TEMP_ARGS. Do not copy breakpoint id range strings over; instead generate a list of strings for
// all the breakpoint ids in the range, and shove all of those breakpoint id strings into TEMP_ARGS.
BreakpointIDList::FindAndReplaceIDRanges (args, target, result, temp_args);
// NOW, convert the list of breakpoint id strings in TEMP_ARGS into an actual BreakpointIDList:
valid_ids->InsertStringArray (temp_args.GetConstArgumentVector(), temp_args.GetArgumentCount(), result);
// At this point, all of the breakpoint ids that the user passed in have been converted to breakpoint IDs
// and put into valid_ids.
if (result.Succeeded())
{
// Now that we've converted everything from args into a list of breakpoint ids, go through our tentative list
// of breakpoint id's and verify that they correspond to valid/currently set breakpoints.
const size_t count = valid_ids->GetSize();
for (size_t i = 0; i < count; ++i)
{
BreakpointID cur_bp_id = valid_ids->GetBreakpointIDAtIndex (i);
Breakpoint *breakpoint = target->GetBreakpointByID (cur_bp_id.GetBreakpointID()).get();
if (breakpoint != NULL)
{
int num_locations = breakpoint->GetNumLocations();
if (cur_bp_id.GetLocationID() > num_locations)
{
StreamString id_str;
BreakpointID::GetCanonicalReference (&id_str,
cur_bp_id.GetBreakpointID(),
cur_bp_id.GetLocationID());
i = valid_ids->GetSize() + 1;
result.AppendErrorWithFormat ("'%s' is not a currently valid breakpoint/location id.\n",
id_str.GetData());
result.SetStatus (eReturnStatusFailed);
}
}
else
{
i = valid_ids->GetSize() + 1;
result.AppendErrorWithFormat ("'%d' is not a currently valid breakpoint id.\n", cur_bp_id.GetBreakpointID());
result.SetStatus (eReturnStatusFailed);
}
}
}
}
//-------------------------------------------------------------------------
// CommandObjectBreakpointList::Options
//-------------------------------------------------------------------------
#pragma mark List::CommandOptions
CommandObjectBreakpointList::CommandOptions::CommandOptions(CommandInterpreter &interpreter) :
Options (interpreter),
m_level (lldb::eDescriptionLevelBrief) // Breakpoint List defaults to brief descriptions
{
}
CommandObjectBreakpointList::CommandOptions::~CommandOptions ()
{
}
OptionDefinition
CommandObjectBreakpointList::CommandOptions::g_option_table[] =
{
{ LLDB_OPT_SET_ALL, false, "internal", 'i', no_argument, NULL, 0, eArgTypeNone,
"Show debugger internal breakpoints" },
{ LLDB_OPT_SET_1, false, "brief", 'b', no_argument, NULL, 0, eArgTypeNone,
"Give a brief description of the breakpoint (no location info)."},
// FIXME: We need to add an "internal" command, and then add this sort of thing to it.
// But I need to see it for now, and don't want to wait.
{ LLDB_OPT_SET_2, false, "full", 'f', no_argument, NULL, 0, eArgTypeNone,
"Give a full description of the breakpoint and its locations."},
{ LLDB_OPT_SET_3, false, "verbose", 'v', no_argument, NULL, 0, eArgTypeNone,
"Explain everything we know about the breakpoint (for debugging debugger bugs)." },
{ 0, false, NULL, 0, 0, NULL, 0, eArgTypeNone, NULL }
};
const OptionDefinition*
CommandObjectBreakpointList::CommandOptions::GetDefinitions ()
{
return g_option_table;
}
Error
Added two new classes for command options: lldb_private::OptionGroup lldb_private::OptionGroupOptions OptionGroup lets you define a class that encapsulates settings that you want to reuse in multiple commands. It contains only the option definitions and the ability to set the option values, but it doesn't directly interface with the lldb_private::Options class that is the front end to all of the CommandObject option parsing. For that the OptionGroupOptions class can be used. It aggregates one or more OptionGroup objects and directs the option setting to the appropriate OptionGroup class. For an example of this, take a look at the CommandObjectFile and how it uses its "m_option_group" object shown below to be able to set values in both the FileOptionGroup and PlatformOptionGroup classes. The members used in CommandObjectFile are: OptionGroupOptions m_option_group; FileOptionGroup m_file_options; PlatformOptionGroup m_platform_options; Then in the constructor for CommandObjectFile you can combine the option settings. The code below shows a simplified version of the constructor: CommandObjectFile::CommandObjectFile(CommandInterpreter &interpreter) : CommandObject (...), m_option_group (interpreter), m_file_options (), m_platform_options(true) { m_option_group.Append (&m_file_options); m_option_group.Append (&m_platform_options); m_option_group.Finalize(); } We append the m_file_options and then the m_platform_options and then tell the option group the finalize the results. This allows the m_option_group to become the organizer of our prefs and after option parsing we end up with valid preference settings in both the m_file_options and m_platform_options objects. This also allows any other commands to use the FileOptionGroup and PlatformOptionGroup classes to implement options for their commands. Renamed: virtual void Options::ResetOptionValues(); to: virtual void Options::OptionParsingStarting(); And implemented a new callback named: virtual Error Options::OptionParsingFinished(); This allows Options subclasses to verify that the options all go together after all of the options have been specified and gives the chance for the command object to return an error. It also gives a chance to take all of the option values and produce or initialize objects after all options have completed parsing. Modfied: virtual Error SetOptionValue (int option_idx, const char *option_arg) = 0; to be: virtual Error SetOptionValue (uint32_t option_idx, const char *option_arg) = 0; (option_idx is now unsigned). llvm-svn: 129415
2011-04-13 08:18:08 +08:00
CommandObjectBreakpointList::CommandOptions::SetOptionValue (uint32_t option_idx, const char *option_arg)
{
Error error;
char short_option = (char) m_getopt_table[option_idx].val;
switch (short_option)
{
case 'b':
m_level = lldb::eDescriptionLevelBrief;
break;
case 'f':
m_level = lldb::eDescriptionLevelFull;
break;
case 'v':
m_level = lldb::eDescriptionLevelVerbose;
break;
case 'i':
m_internal = true;
break;
default:
error.SetErrorStringWithFormat ("Unrecognized option '%c'.\n", short_option);
break;
}
return error;
}
void
Added two new classes for command options: lldb_private::OptionGroup lldb_private::OptionGroupOptions OptionGroup lets you define a class that encapsulates settings that you want to reuse in multiple commands. It contains only the option definitions and the ability to set the option values, but it doesn't directly interface with the lldb_private::Options class that is the front end to all of the CommandObject option parsing. For that the OptionGroupOptions class can be used. It aggregates one or more OptionGroup objects and directs the option setting to the appropriate OptionGroup class. For an example of this, take a look at the CommandObjectFile and how it uses its "m_option_group" object shown below to be able to set values in both the FileOptionGroup and PlatformOptionGroup classes. The members used in CommandObjectFile are: OptionGroupOptions m_option_group; FileOptionGroup m_file_options; PlatformOptionGroup m_platform_options; Then in the constructor for CommandObjectFile you can combine the option settings. The code below shows a simplified version of the constructor: CommandObjectFile::CommandObjectFile(CommandInterpreter &interpreter) : CommandObject (...), m_option_group (interpreter), m_file_options (), m_platform_options(true) { m_option_group.Append (&m_file_options); m_option_group.Append (&m_platform_options); m_option_group.Finalize(); } We append the m_file_options and then the m_platform_options and then tell the option group the finalize the results. This allows the m_option_group to become the organizer of our prefs and after option parsing we end up with valid preference settings in both the m_file_options and m_platform_options objects. This also allows any other commands to use the FileOptionGroup and PlatformOptionGroup classes to implement options for their commands. Renamed: virtual void Options::ResetOptionValues(); to: virtual void Options::OptionParsingStarting(); And implemented a new callback named: virtual Error Options::OptionParsingFinished(); This allows Options subclasses to verify that the options all go together after all of the options have been specified and gives the chance for the command object to return an error. It also gives a chance to take all of the option values and produce or initialize objects after all options have completed parsing. Modfied: virtual Error SetOptionValue (int option_idx, const char *option_arg) = 0; to be: virtual Error SetOptionValue (uint32_t option_idx, const char *option_arg) = 0; (option_idx is now unsigned). llvm-svn: 129415
2011-04-13 08:18:08 +08:00
CommandObjectBreakpointList::CommandOptions::OptionParsingStarting ()
{
m_level = lldb::eDescriptionLevelFull;
m_internal = false;
}
//-------------------------------------------------------------------------
// CommandObjectBreakpointList
//-------------------------------------------------------------------------
#pragma mark List
CommandObjectBreakpointList::CommandObjectBreakpointList (CommandInterpreter &interpreter) :
CommandObject (interpreter,
"breakpoint list",
"List some or all breakpoints at configurable levels of detail.",
NULL),
m_options (interpreter)
{
CommandArgumentEntry arg;
CommandArgumentData bp_id_arg;
// Define the first (and only) variant of this arg.
bp_id_arg.arg_type = eArgTypeBreakpointID;
bp_id_arg.arg_repetition = eArgRepeatOptional;
// There is only one variant this argument could be; put it into the argument entry.
arg.push_back (bp_id_arg);
// Push the data for the first argument into the m_arguments vector.
m_arguments.push_back (arg);
}
CommandObjectBreakpointList::~CommandObjectBreakpointList ()
{
}
Options *
CommandObjectBreakpointList::GetOptions ()
{
return &m_options;
}
bool
CommandObjectBreakpointList::Execute
(
Args& args,
CommandReturnObject &result
)
{
Target *target = m_interpreter.GetDebugger().GetSelectedTarget().get();
if (target == NULL)
{
result.AppendError ("Invalid target. No current target or breakpoints.");
result.SetStatus (eReturnStatusSuccessFinishNoResult);
return true;
}
const BreakpointList &breakpoints = target->GetBreakpointList(m_options.m_internal);
Mutex::Locker locker;
target->GetBreakpointList(m_options.m_internal).GetListMutex(locker);
size_t num_breakpoints = breakpoints.GetSize();
if (num_breakpoints == 0)
{
result.AppendMessage ("No breakpoints currently set.");
result.SetStatus (eReturnStatusSuccessFinishNoResult);
return true;
}
Stream &output_stream = result.GetOutputStream();
if (args.GetArgumentCount() == 0)
{
// No breakpoint selected; show info about all currently set breakpoints.
result.AppendMessage ("Current breakpoints:");
for (size_t i = 0; i < num_breakpoints; ++i)
{
Breakpoint *breakpoint = breakpoints.GetBreakpointAtIndex (i).get();
AddBreakpointDescription (&output_stream, breakpoint, m_options.m_level);
}
result.SetStatus (eReturnStatusSuccessFinishNoResult);
}
else
{
// Particular breakpoints selected; show info about that breakpoint.
BreakpointIDList valid_bp_ids;
CommandObjectMultiwordBreakpoint::VerifyBreakpointIDs (args, target, result, &valid_bp_ids);
if (result.Succeeded())
{
for (size_t i = 0; i < valid_bp_ids.GetSize(); ++i)
{
BreakpointID cur_bp_id = valid_bp_ids.GetBreakpointIDAtIndex (i);
Breakpoint *breakpoint = target->GetBreakpointByID (cur_bp_id.GetBreakpointID()).get();
AddBreakpointDescription (&output_stream, breakpoint, m_options.m_level);
}
result.SetStatus (eReturnStatusSuccessFinishNoResult);
}
else
{
result.AppendError ("Invalid breakpoint id.");
result.SetStatus (eReturnStatusFailed);
}
}
return result.Succeeded();
}
//-------------------------------------------------------------------------
// CommandObjectBreakpointEnable
//-------------------------------------------------------------------------
#pragma mark Enable
CommandObjectBreakpointEnable::CommandObjectBreakpointEnable (CommandInterpreter &interpreter) :
CommandObject (interpreter,
"enable",
"Enable the specified disabled breakpoint(s). If no breakpoints are specified, enable all of them.",
NULL)
{
CommandArgumentEntry arg;
CommandArgumentData bp_id_arg;
CommandArgumentData bp_id_range_arg;
// Create the first variant for the first (and only) argument for this command.
bp_id_arg.arg_type = eArgTypeBreakpointID;
bp_id_arg.arg_repetition = eArgRepeatOptional;
// Create the second variant for the first (and only) argument for this command.
bp_id_range_arg.arg_type = eArgTypeBreakpointIDRange;
bp_id_range_arg.arg_repetition = eArgRepeatOptional;
// The first (and only) argument for this command could be either a bp_id or a bp_id_range.
// Push both variants into the entry for the first argument for this command.
arg.push_back (bp_id_arg);
arg.push_back (bp_id_range_arg);
// Add the entry for the first argument for this command to the object's arguments vector.
m_arguments.push_back (arg);
}
CommandObjectBreakpointEnable::~CommandObjectBreakpointEnable ()
{
}
bool
CommandObjectBreakpointEnable::Execute
(
Args& args,
CommandReturnObject &result
)
{
Target *target = m_interpreter.GetDebugger().GetSelectedTarget().get();
if (target == NULL)
{
result.AppendError ("Invalid target. No existing target or breakpoints.");
result.SetStatus (eReturnStatusFailed);
return false;
}
Mutex::Locker locker;
target->GetBreakpointList().GetListMutex(locker);
const BreakpointList &breakpoints = target->GetBreakpointList();
size_t num_breakpoints = breakpoints.GetSize();
if (num_breakpoints == 0)
{
result.AppendError ("No breakpoints exist to be enabled.");
result.SetStatus (eReturnStatusFailed);
return false;
}
if (args.GetArgumentCount() == 0)
{
// No breakpoint selected; enable all currently set breakpoints.
target->EnableAllBreakpoints ();
result.AppendMessageWithFormat ("All breakpoints enabled. (%d breakpoints)\n", num_breakpoints);
result.SetStatus (eReturnStatusSuccessFinishNoResult);
}
else
{
// Particular breakpoint selected; enable that breakpoint.
BreakpointIDList valid_bp_ids;
CommandObjectMultiwordBreakpoint::VerifyBreakpointIDs (args, target, result, &valid_bp_ids);
if (result.Succeeded())
{
int enable_count = 0;
int loc_count = 0;
const size_t count = valid_bp_ids.GetSize();
for (size_t i = 0; i < count; ++i)
{
BreakpointID cur_bp_id = valid_bp_ids.GetBreakpointIDAtIndex (i);
if (cur_bp_id.GetBreakpointID() != LLDB_INVALID_BREAK_ID)
{
Breakpoint *breakpoint = target->GetBreakpointByID (cur_bp_id.GetBreakpointID()).get();
if (cur_bp_id.GetLocationID() != LLDB_INVALID_BREAK_ID)
{
BreakpointLocation *location = breakpoint->FindLocationByID (cur_bp_id.GetLocationID()).get();
if (location)
{
location->SetEnabled (true);
++loc_count;
}
}
else
{
breakpoint->SetEnabled (true);
++enable_count;
}
}
}
result.AppendMessageWithFormat ("%d breakpoints enabled.\n", enable_count + loc_count);
result.SetStatus (eReturnStatusSuccessFinishNoResult);
}
}
return result.Succeeded();
}
//-------------------------------------------------------------------------
// CommandObjectBreakpointDisable
//-------------------------------------------------------------------------
#pragma mark Disable
CommandObjectBreakpointDisable::CommandObjectBreakpointDisable (CommandInterpreter &interpreter) :
CommandObject (interpreter,
"breakpoint disable",
"Disable the specified breakpoint(s) without removing it/them. If no breakpoints are specified, disable them all.",
NULL)
{
CommandArgumentEntry arg;
CommandArgumentData bp_id_arg;
CommandArgumentData bp_id_range_arg;
// Create the first variant for the first (and only) argument for this command.
bp_id_arg.arg_type = eArgTypeBreakpointID;
bp_id_arg.arg_repetition = eArgRepeatOptional;
// Create the second variant for the first (and only) argument for this command.
bp_id_range_arg.arg_type = eArgTypeBreakpointIDRange;
bp_id_range_arg.arg_repetition = eArgRepeatOptional;
// The first (and only) argument for this command could be either a bp_id or a bp_id_range.
// Push both variants into the entry for the first argument for this command.
arg.push_back (bp_id_arg);
arg.push_back (bp_id_range_arg);
// Add the entry for the first argument for this command to the object's arguments vector.
m_arguments.push_back (arg);
}
CommandObjectBreakpointDisable::~CommandObjectBreakpointDisable ()
{
}
bool
CommandObjectBreakpointDisable::Execute
(
Args& args,
CommandReturnObject &result
)
{
Target *target = m_interpreter.GetDebugger().GetSelectedTarget().get();
if (target == NULL)
{
result.AppendError ("Invalid target. No existing target or breakpoints.");
result.SetStatus (eReturnStatusFailed);
return false;
}
Mutex::Locker locker;
target->GetBreakpointList().GetListMutex(locker);
const BreakpointList &breakpoints = target->GetBreakpointList();
size_t num_breakpoints = breakpoints.GetSize();
if (num_breakpoints == 0)
{
result.AppendError ("No breakpoints exist to be disabled.");
result.SetStatus (eReturnStatusFailed);
return false;
}
if (args.GetArgumentCount() == 0)
{
// No breakpoint selected; disable all currently set breakpoints.
target->DisableAllBreakpoints ();
result.AppendMessageWithFormat ("All breakpoints disabled. (%d breakpoints)\n", num_breakpoints);
result.SetStatus (eReturnStatusSuccessFinishNoResult);
}
else
{
// Particular breakpoint selected; disable that breakpoint.
BreakpointIDList valid_bp_ids;
CommandObjectMultiwordBreakpoint::VerifyBreakpointIDs (args, target, result, &valid_bp_ids);
if (result.Succeeded())
{
int disable_count = 0;
int loc_count = 0;
const size_t count = valid_bp_ids.GetSize();
for (size_t i = 0; i < count; ++i)
{
BreakpointID cur_bp_id = valid_bp_ids.GetBreakpointIDAtIndex (i);
if (cur_bp_id.GetBreakpointID() != LLDB_INVALID_BREAK_ID)
{
Breakpoint *breakpoint = target->GetBreakpointByID (cur_bp_id.GetBreakpointID()).get();
if (cur_bp_id.GetLocationID() != LLDB_INVALID_BREAK_ID)
{
BreakpointLocation *location = breakpoint->FindLocationByID (cur_bp_id.GetLocationID()).get();
if (location)
{
location->SetEnabled (false);
++loc_count;
}
}
else
{
breakpoint->SetEnabled (false);
++disable_count;
}
}
}
result.AppendMessageWithFormat ("%d breakpoints disabled.\n", disable_count + loc_count);
result.SetStatus (eReturnStatusSuccessFinishNoResult);
}
}
return result.Succeeded();
}
//-------------------------------------------------------------------------
// CommandObjectBreakpointClear::CommandOptions
//-------------------------------------------------------------------------
#pragma mark Clear::CommandOptions
CommandObjectBreakpointClear::CommandOptions::CommandOptions(CommandInterpreter &interpreter) :
Options (interpreter),
m_filename (),
m_line_num (0)
{
}
CommandObjectBreakpointClear::CommandOptions::~CommandOptions ()
{
}
OptionDefinition
CommandObjectBreakpointClear::CommandOptions::g_option_table[] =
{
{ LLDB_OPT_SET_1, false, "file", 'f', required_argument, NULL, CommandCompletions::eSourceFileCompletion, eArgTypeFilename,
"Specify the breakpoint by source location in this particular file."},
{ LLDB_OPT_SET_1, true, "line", 'l', required_argument, NULL, 0, eArgTypeLineNum,
"Specify the breakpoint by source location at this particular line."},
{ 0, false, NULL, 0, 0, NULL, 0, eArgTypeNone, NULL }
};
const OptionDefinition*
CommandObjectBreakpointClear::CommandOptions::GetDefinitions ()
{
return g_option_table;
}
Error
Added two new classes for command options: lldb_private::OptionGroup lldb_private::OptionGroupOptions OptionGroup lets you define a class that encapsulates settings that you want to reuse in multiple commands. It contains only the option definitions and the ability to set the option values, but it doesn't directly interface with the lldb_private::Options class that is the front end to all of the CommandObject option parsing. For that the OptionGroupOptions class can be used. It aggregates one or more OptionGroup objects and directs the option setting to the appropriate OptionGroup class. For an example of this, take a look at the CommandObjectFile and how it uses its "m_option_group" object shown below to be able to set values in both the FileOptionGroup and PlatformOptionGroup classes. The members used in CommandObjectFile are: OptionGroupOptions m_option_group; FileOptionGroup m_file_options; PlatformOptionGroup m_platform_options; Then in the constructor for CommandObjectFile you can combine the option settings. The code below shows a simplified version of the constructor: CommandObjectFile::CommandObjectFile(CommandInterpreter &interpreter) : CommandObject (...), m_option_group (interpreter), m_file_options (), m_platform_options(true) { m_option_group.Append (&m_file_options); m_option_group.Append (&m_platform_options); m_option_group.Finalize(); } We append the m_file_options and then the m_platform_options and then tell the option group the finalize the results. This allows the m_option_group to become the organizer of our prefs and after option parsing we end up with valid preference settings in both the m_file_options and m_platform_options objects. This also allows any other commands to use the FileOptionGroup and PlatformOptionGroup classes to implement options for their commands. Renamed: virtual void Options::ResetOptionValues(); to: virtual void Options::OptionParsingStarting(); And implemented a new callback named: virtual Error Options::OptionParsingFinished(); This allows Options subclasses to verify that the options all go together after all of the options have been specified and gives the chance for the command object to return an error. It also gives a chance to take all of the option values and produce or initialize objects after all options have completed parsing. Modfied: virtual Error SetOptionValue (int option_idx, const char *option_arg) = 0; to be: virtual Error SetOptionValue (uint32_t option_idx, const char *option_arg) = 0; (option_idx is now unsigned). llvm-svn: 129415
2011-04-13 08:18:08 +08:00
CommandObjectBreakpointClear::CommandOptions::SetOptionValue (uint32_t option_idx, const char *option_arg)
{
Error error;
char short_option = (char) m_getopt_table[option_idx].val;
switch (short_option)
{
case 'f':
Added the ability to get the min and max instruction byte size for 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
2011-03-27 03:14:58 +08:00
m_filename.assign (option_arg);
break;
case 'l':
m_line_num = Args::StringToUInt32 (option_arg, 0);
break;
default:
error.SetErrorStringWithFormat ("Unrecognized option '%c'.\n", short_option);
break;
}
return error;
}
void
Added two new classes for command options: lldb_private::OptionGroup lldb_private::OptionGroupOptions OptionGroup lets you define a class that encapsulates settings that you want to reuse in multiple commands. It contains only the option definitions and the ability to set the option values, but it doesn't directly interface with the lldb_private::Options class that is the front end to all of the CommandObject option parsing. For that the OptionGroupOptions class can be used. It aggregates one or more OptionGroup objects and directs the option setting to the appropriate OptionGroup class. For an example of this, take a look at the CommandObjectFile and how it uses its "m_option_group" object shown below to be able to set values in both the FileOptionGroup and PlatformOptionGroup classes. The members used in CommandObjectFile are: OptionGroupOptions m_option_group; FileOptionGroup m_file_options; PlatformOptionGroup m_platform_options; Then in the constructor for CommandObjectFile you can combine the option settings. The code below shows a simplified version of the constructor: CommandObjectFile::CommandObjectFile(CommandInterpreter &interpreter) : CommandObject (...), m_option_group (interpreter), m_file_options (), m_platform_options(true) { m_option_group.Append (&m_file_options); m_option_group.Append (&m_platform_options); m_option_group.Finalize(); } We append the m_file_options and then the m_platform_options and then tell the option group the finalize the results. This allows the m_option_group to become the organizer of our prefs and after option parsing we end up with valid preference settings in both the m_file_options and m_platform_options objects. This also allows any other commands to use the FileOptionGroup and PlatformOptionGroup classes to implement options for their commands. Renamed: virtual void Options::ResetOptionValues(); to: virtual void Options::OptionParsingStarting(); And implemented a new callback named: virtual Error Options::OptionParsingFinished(); This allows Options subclasses to verify that the options all go together after all of the options have been specified and gives the chance for the command object to return an error. It also gives a chance to take all of the option values and produce or initialize objects after all options have completed parsing. Modfied: virtual Error SetOptionValue (int option_idx, const char *option_arg) = 0; to be: virtual Error SetOptionValue (uint32_t option_idx, const char *option_arg) = 0; (option_idx is now unsigned). llvm-svn: 129415
2011-04-13 08:18:08 +08:00
CommandObjectBreakpointClear::CommandOptions::OptionParsingStarting ()
{
m_filename.clear();
m_line_num = 0;
}
//-------------------------------------------------------------------------
// CommandObjectBreakpointClear
//-------------------------------------------------------------------------
#pragma mark Clear
CommandObjectBreakpointClear::CommandObjectBreakpointClear (CommandInterpreter &interpreter) :
CommandObject (interpreter,
"breakpoint clear",
"Clears a breakpoint or set of breakpoints in the executable.",
"breakpoint clear <cmd-options>"),
m_options (interpreter)
{
}
CommandObjectBreakpointClear::~CommandObjectBreakpointClear ()
{
}
Options *
CommandObjectBreakpointClear::GetOptions ()
{
return &m_options;
}
bool
CommandObjectBreakpointClear::Execute
(
Args& command,
CommandReturnObject &result
)
{
Target *target = m_interpreter.GetDebugger().GetSelectedTarget().get();
if (target == NULL)
{
result.AppendError ("Invalid target. No existing target or breakpoints.");
result.SetStatus (eReturnStatusFailed);
return false;
}
// The following are the various types of breakpoints that could be cleared:
// 1). -f -l (clearing breakpoint by source location)
BreakpointClearType break_type = eClearTypeInvalid;
if (m_options.m_line_num != 0)
break_type = eClearTypeFileAndLine;
Mutex::Locker locker;
target->GetBreakpointList().GetListMutex(locker);
BreakpointList &breakpoints = target->GetBreakpointList();
size_t num_breakpoints = breakpoints.GetSize();
// Early return if there's no breakpoint at all.
if (num_breakpoints == 0)
{
result.AppendError ("Breakpoint clear: No breakpoint cleared.");
result.SetStatus (eReturnStatusFailed);
return result.Succeeded();
}
// Find matching breakpoints and delete them.
// First create a copy of all the IDs.
std::vector<break_id_t> BreakIDs;
for (size_t i = 0; i < num_breakpoints; ++i)
BreakIDs.push_back(breakpoints.GetBreakpointAtIndex(i).get()->GetID());
int num_cleared = 0;
StreamString ss;
switch (break_type)
{
case eClearTypeFileAndLine: // Breakpoint by source position
{
const ConstString filename(m_options.m_filename.c_str());
BreakpointLocationCollection loc_coll;
for (size_t i = 0; i < num_breakpoints; ++i)
{
Breakpoint *bp = breakpoints.FindBreakpointByID(BreakIDs[i]).get();
if (bp->GetMatchingFileLine(filename, m_options.m_line_num, loc_coll))
{
// If the collection size is 0, it's a full match and we can just remove the breakpoint.
if (loc_coll.GetSize() == 0)
{
bp->GetDescription(&ss, lldb::eDescriptionLevelBrief);
ss.EOL();
target->RemoveBreakpointByID (bp->GetID());
++num_cleared;
}
}
}
}
break;
default:
break;
}
if (num_cleared > 0)
{
Stream &output_stream = result.GetOutputStream();
output_stream.Printf ("%d breakpoints cleared:\n", num_cleared);
output_stream << ss.GetData();
output_stream.EOL();
result.SetStatus (eReturnStatusSuccessFinishNoResult);
}
else
{
result.AppendError ("Breakpoint clear: No breakpoint cleared.");
result.SetStatus (eReturnStatusFailed);
}
return result.Succeeded();
}
//-------------------------------------------------------------------------
// CommandObjectBreakpointDelete
//-------------------------------------------------------------------------
#pragma mark Delete
CommandObjectBreakpointDelete::CommandObjectBreakpointDelete(CommandInterpreter &interpreter) :
CommandObject (interpreter,
"breakpoint delete",
"Delete the specified breakpoint(s). If no breakpoints are specified, delete them all.",
NULL)
{
CommandArgumentEntry arg;
CommandArgumentData bp_id_arg;
CommandArgumentData bp_id_range_arg;
// Create the first variant for the first (and only) argument for this command.
bp_id_arg.arg_type = eArgTypeBreakpointID;
bp_id_arg.arg_repetition = eArgRepeatOptional;
// Create the second variant for the first (and only) argument for this command.
bp_id_range_arg.arg_type = eArgTypeBreakpointIDRange;
bp_id_range_arg.arg_repetition = eArgRepeatOptional;
// The first (and only) argument for this command could be either a bp_id or a bp_id_range.
// Push both variants into the entry for the first argument for this command.
arg.push_back (bp_id_arg);
arg.push_back (bp_id_range_arg);
// Add the entry for the first argument for this command to the object's arguments vector.
m_arguments.push_back (arg);
}
CommandObjectBreakpointDelete::~CommandObjectBreakpointDelete ()
{
}
bool
CommandObjectBreakpointDelete::Execute
(
Args& args,
CommandReturnObject &result
)
{
Target *target = m_interpreter.GetDebugger().GetSelectedTarget().get();
if (target == NULL)
{
result.AppendError ("Invalid target. No existing target or breakpoints.");
result.SetStatus (eReturnStatusFailed);
return false;
}
Mutex::Locker locker;
target->GetBreakpointList().GetListMutex(locker);
const BreakpointList &breakpoints = target->GetBreakpointList();
size_t num_breakpoints = breakpoints.GetSize();
if (num_breakpoints == 0)
{
result.AppendError ("No breakpoints exist to be deleted.");
result.SetStatus (eReturnStatusFailed);
return false;
}
if (args.GetArgumentCount() == 0)
{
if (!m_interpreter.Confirm ("About to delete all breakpoints, do you want to do that?", true))
{
result.AppendMessage("Operation cancelled...");
}
else
{
target->RemoveAllBreakpoints ();
result.AppendMessageWithFormat ("All breakpoints removed. (%d breakpoints)\n", num_breakpoints);
}
result.SetStatus (eReturnStatusSuccessFinishNoResult);
}
else
{
// Particular breakpoint selected; disable that breakpoint.
BreakpointIDList valid_bp_ids;
CommandObjectMultiwordBreakpoint::VerifyBreakpointIDs (args, target, result, &valid_bp_ids);
if (result.Succeeded())
{
int delete_count = 0;
int disable_count = 0;
const size_t count = valid_bp_ids.GetSize();
for (size_t i = 0; i < count; ++i)
{
BreakpointID cur_bp_id = valid_bp_ids.GetBreakpointIDAtIndex (i);
if (cur_bp_id.GetBreakpointID() != LLDB_INVALID_BREAK_ID)
{
if (cur_bp_id.GetLocationID() != LLDB_INVALID_BREAK_ID)
{
Breakpoint *breakpoint = target->GetBreakpointByID (cur_bp_id.GetBreakpointID()).get();
BreakpointLocation *location = breakpoint->FindLocationByID (cur_bp_id.GetLocationID()).get();
// It makes no sense to try to delete individual locations, so we disable them instead.
if (location)
{
location->SetEnabled (false);
++disable_count;
}
}
else
{
target->RemoveBreakpointByID (cur_bp_id.GetBreakpointID());
++delete_count;
}
}
}
result.AppendMessageWithFormat ("%d breakpoints deleted; %d breakpoint locations disabled.\n",
delete_count, disable_count);
result.SetStatus (eReturnStatusSuccessFinishNoResult);
}
}
return result.Succeeded();
}
//-------------------------------------------------------------------------
// CommandObjectBreakpointModify::CommandOptions
//-------------------------------------------------------------------------
#pragma mark Modify::CommandOptions
CommandObjectBreakpointModify::CommandOptions::CommandOptions(CommandInterpreter &interpreter) :
Options (interpreter),
m_ignore_count (0),
m_thread_id(LLDB_INVALID_THREAD_ID),
m_thread_id_passed(false),
m_thread_index (UINT32_MAX),
m_thread_index_passed(false),
m_thread_name(),
m_queue_name(),
m_condition (),
m_enable_passed (false),
m_enable_value (false),
m_name_passed (false),
m_queue_passed (false),
m_condition_passed (false)
{
}
CommandObjectBreakpointModify::CommandOptions::~CommandOptions ()
{
}
OptionDefinition
CommandObjectBreakpointModify::CommandOptions::g_option_table[] =
{
{ LLDB_OPT_SET_ALL, false, "ignore-count", 'i', required_argument, NULL, NULL, eArgTypeCount, "Set the number of times this breakpoint is skipped before stopping." },
{ LLDB_OPT_SET_ALL, false, "thread-index", 'x', required_argument, NULL, NULL, eArgTypeThreadIndex, "The breakpoint stops only for the thread whose indeX matches this argument."},
{ LLDB_OPT_SET_ALL, false, "thread-id", 't', required_argument, NULL, NULL, eArgTypeThreadID, "The breakpoint stops only for the thread whose TID matches this argument."},
{ LLDB_OPT_SET_ALL, false, "thread-name", 'T', required_argument, NULL, NULL, eArgTypeThreadName, "The breakpoint stops only for the thread whose thread name matches this argument."},
{ LLDB_OPT_SET_ALL, false, "queue-name", 'q', required_argument, NULL, NULL, eArgTypeQueueName, "The breakpoint stops only for threads in the queue whose name is given by this argument."},
{ LLDB_OPT_SET_ALL, false, "condition", 'c', required_argument, NULL, NULL, eArgTypeExpression, "The breakpoint stops only if this condition expression evaluates to true."},
{ LLDB_OPT_SET_1, false, "enable", 'e', no_argument, NULL, NULL, eArgTypeNone, "Enable the breakpoint."},
{ LLDB_OPT_SET_2, false, "disable", 'd', no_argument, NULL, NULL, eArgTypeNone, "Disable the breakpoint."},
{ 0, false, NULL, 0 , 0, NULL, 0, eArgTypeNone, NULL }
};
const OptionDefinition*
CommandObjectBreakpointModify::CommandOptions::GetDefinitions ()
{
return g_option_table;
}
Error
Added two new classes for command options: lldb_private::OptionGroup lldb_private::OptionGroupOptions OptionGroup lets you define a class that encapsulates settings that you want to reuse in multiple commands. It contains only the option definitions and the ability to set the option values, but it doesn't directly interface with the lldb_private::Options class that is the front end to all of the CommandObject option parsing. For that the OptionGroupOptions class can be used. It aggregates one or more OptionGroup objects and directs the option setting to the appropriate OptionGroup class. For an example of this, take a look at the CommandObjectFile and how it uses its "m_option_group" object shown below to be able to set values in both the FileOptionGroup and PlatformOptionGroup classes. The members used in CommandObjectFile are: OptionGroupOptions m_option_group; FileOptionGroup m_file_options; PlatformOptionGroup m_platform_options; Then in the constructor for CommandObjectFile you can combine the option settings. The code below shows a simplified version of the constructor: CommandObjectFile::CommandObjectFile(CommandInterpreter &interpreter) : CommandObject (...), m_option_group (interpreter), m_file_options (), m_platform_options(true) { m_option_group.Append (&m_file_options); m_option_group.Append (&m_platform_options); m_option_group.Finalize(); } We append the m_file_options and then the m_platform_options and then tell the option group the finalize the results. This allows the m_option_group to become the organizer of our prefs and after option parsing we end up with valid preference settings in both the m_file_options and m_platform_options objects. This also allows any other commands to use the FileOptionGroup and PlatformOptionGroup classes to implement options for their commands. Renamed: virtual void Options::ResetOptionValues(); to: virtual void Options::OptionParsingStarting(); And implemented a new callback named: virtual Error Options::OptionParsingFinished(); This allows Options subclasses to verify that the options all go together after all of the options have been specified and gives the chance for the command object to return an error. It also gives a chance to take all of the option values and produce or initialize objects after all options have completed parsing. Modfied: virtual Error SetOptionValue (int option_idx, const char *option_arg) = 0; to be: virtual Error SetOptionValue (uint32_t option_idx, const char *option_arg) = 0; (option_idx is now unsigned). llvm-svn: 129415
2011-04-13 08:18:08 +08:00
CommandObjectBreakpointModify::CommandOptions::SetOptionValue (uint32_t option_idx, const char *option_arg)
{
Error error;
char short_option = (char) m_getopt_table[option_idx].val;
switch (short_option)
{
case 'c':
if (option_arg != NULL)
Added the ability to get the min and max instruction byte size for 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
2011-03-27 03:14:58 +08:00
m_condition.assign (option_arg);
else
m_condition.clear();
m_condition_passed = true;
break;
case 'd':
m_enable_passed = true;
m_enable_value = false;
break;
case 'e':
m_enable_passed = true;
m_enable_value = true;
break;
case 'i':
{
m_ignore_count = Args::StringToUInt32(option_arg, UINT32_MAX, 0);
if (m_ignore_count == UINT32_MAX)
error.SetErrorStringWithFormat ("Invalid ignore count '%s'.\n", option_arg);
}
break;
case 't' :
{
if (option_arg[0] == '\0')
{
m_thread_id = LLDB_INVALID_THREAD_ID;
m_thread_id_passed = true;
}
else
{
m_thread_id = Args::StringToUInt64(option_arg, LLDB_INVALID_THREAD_ID, 0);
if (m_thread_id == LLDB_INVALID_THREAD_ID)
error.SetErrorStringWithFormat ("Invalid thread id string '%s'.\n", option_arg);
else
m_thread_id_passed = true;
}
}
break;
case 'T':
if (option_arg != NULL)
Added the ability to get the min and max instruction byte size for 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
2011-03-27 03:14:58 +08:00
m_thread_name.assign (option_arg);
else
m_thread_name.clear();
m_name_passed = true;
break;
case 'q':
if (option_arg != NULL)
Added the ability to get the min and max instruction byte size for 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
2011-03-27 03:14:58 +08:00
m_queue_name.assign (option_arg);
else
m_queue_name.clear();
m_queue_passed = true;
break;
case 'x':
{
if (option_arg[0] == '\n')
{
m_thread_index = UINT32_MAX;
m_thread_index_passed = true;
}
else
{
m_thread_index = Args::StringToUInt32 (option_arg, UINT32_MAX, 0);
if (m_thread_id == UINT32_MAX)
error.SetErrorStringWithFormat ("Invalid thread index string '%s'.\n", option_arg);
else
m_thread_index_passed = true;
}
}
break;
default:
error.SetErrorStringWithFormat ("Unrecognized option '%c'.\n", short_option);
break;
}
return error;
}
void
Added two new classes for command options: lldb_private::OptionGroup lldb_private::OptionGroupOptions OptionGroup lets you define a class that encapsulates settings that you want to reuse in multiple commands. It contains only the option definitions and the ability to set the option values, but it doesn't directly interface with the lldb_private::Options class that is the front end to all of the CommandObject option parsing. For that the OptionGroupOptions class can be used. It aggregates one or more OptionGroup objects and directs the option setting to the appropriate OptionGroup class. For an example of this, take a look at the CommandObjectFile and how it uses its "m_option_group" object shown below to be able to set values in both the FileOptionGroup and PlatformOptionGroup classes. The members used in CommandObjectFile are: OptionGroupOptions m_option_group; FileOptionGroup m_file_options; PlatformOptionGroup m_platform_options; Then in the constructor for CommandObjectFile you can combine the option settings. The code below shows a simplified version of the constructor: CommandObjectFile::CommandObjectFile(CommandInterpreter &interpreter) : CommandObject (...), m_option_group (interpreter), m_file_options (), m_platform_options(true) { m_option_group.Append (&m_file_options); m_option_group.Append (&m_platform_options); m_option_group.Finalize(); } We append the m_file_options and then the m_platform_options and then tell the option group the finalize the results. This allows the m_option_group to become the organizer of our prefs and after option parsing we end up with valid preference settings in both the m_file_options and m_platform_options objects. This also allows any other commands to use the FileOptionGroup and PlatformOptionGroup classes to implement options for their commands. Renamed: virtual void Options::ResetOptionValues(); to: virtual void Options::OptionParsingStarting(); And implemented a new callback named: virtual Error Options::OptionParsingFinished(); This allows Options subclasses to verify that the options all go together after all of the options have been specified and gives the chance for the command object to return an error. It also gives a chance to take all of the option values and produce or initialize objects after all options have completed parsing. Modfied: virtual Error SetOptionValue (int option_idx, const char *option_arg) = 0; to be: virtual Error SetOptionValue (uint32_t option_idx, const char *option_arg) = 0; (option_idx is now unsigned). llvm-svn: 129415
2011-04-13 08:18:08 +08:00
CommandObjectBreakpointModify::CommandOptions::OptionParsingStarting ()
{
m_ignore_count = 0;
m_thread_id = LLDB_INVALID_THREAD_ID;
m_thread_id_passed = false;
m_thread_index = UINT32_MAX;
m_thread_index_passed = false;
m_thread_name.clear();
m_queue_name.clear();
m_condition.clear();
m_enable_passed = false;
m_queue_passed = false;
m_name_passed = false;
m_condition_passed = false;
}
//-------------------------------------------------------------------------
// CommandObjectBreakpointModify
//-------------------------------------------------------------------------
#pragma mark Modify
CommandObjectBreakpointModify::CommandObjectBreakpointModify (CommandInterpreter &interpreter) :
CommandObject (interpreter,
"breakpoint modify",
"Modify the options on a breakpoint or set of breakpoints in the executable. "
"If no breakpoint is specified, acts on the last created breakpoint. "
"With the exception of -e, -d and -i, passing an empty argument clears the modification.",
NULL),
m_options (interpreter)
{
CommandArgumentEntry arg;
CommandArgumentData bp_id_arg;
CommandArgumentData bp_id_range_arg;
// Create the first variant for the first (and only) argument for this command.
bp_id_arg.arg_type = eArgTypeBreakpointID;
bp_id_arg.arg_repetition = eArgRepeatPlain;
// Create the second variant for the first (and only) argument for this command.
bp_id_range_arg.arg_type = eArgTypeBreakpointIDRange;
bp_id_range_arg.arg_repetition = eArgRepeatPlain;
// The first (and only) argument for this command could be either a bp_id or a bp_id_range.
// Push both variants into the entry for the first argument for this command.
arg.push_back (bp_id_arg);
arg.push_back (bp_id_range_arg);
// Add the entry for the first argument for this command to the object's arguments vector.
m_arguments.push_back (arg);
}
CommandObjectBreakpointModify::~CommandObjectBreakpointModify ()
{
}
Options *
CommandObjectBreakpointModify::GetOptions ()
{
return &m_options;
}
bool
CommandObjectBreakpointModify::Execute
(
Args& command,
CommandReturnObject &result
)
{
Target *target = m_interpreter.GetDebugger().GetSelectedTarget().get();
if (target == NULL)
{
result.AppendError ("Invalid target. No existing target or breakpoints.");
result.SetStatus (eReturnStatusFailed);
return false;
}
Mutex::Locker locker;
target->GetBreakpointList().GetListMutex(locker);
BreakpointIDList valid_bp_ids;
CommandObjectMultiwordBreakpoint::VerifyBreakpointIDs (command, target, result, &valid_bp_ids);
if (result.Succeeded())
{
const size_t count = valid_bp_ids.GetSize();
for (size_t i = 0; i < count; ++i)
{
BreakpointID cur_bp_id = valid_bp_ids.GetBreakpointIDAtIndex (i);
if (cur_bp_id.GetBreakpointID() != LLDB_INVALID_BREAK_ID)
{
Breakpoint *bp = target->GetBreakpointByID (cur_bp_id.GetBreakpointID()).get();
if (cur_bp_id.GetLocationID() != LLDB_INVALID_BREAK_ID)
{
BreakpointLocation *location = bp->FindLocationByID (cur_bp_id.GetLocationID()).get();
if (location)
{
if (m_options.m_thread_id_passed)
location->SetThreadID (m_options.m_thread_id);
if (m_options.m_thread_index_passed)
location->GetLocationOptions()->GetThreadSpec()->SetIndex(m_options.m_thread_index);
if (m_options.m_name_passed)
location->GetLocationOptions()->GetThreadSpec()->SetName(m_options.m_thread_name.c_str());
if (m_options.m_queue_passed)
location->GetLocationOptions()->GetThreadSpec()->SetQueueName(m_options.m_queue_name.c_str());
if (m_options.m_ignore_count != 0)
location->GetLocationOptions()->SetIgnoreCount(m_options.m_ignore_count);
if (m_options.m_enable_passed)
location->SetEnabled (m_options.m_enable_value);
if (m_options.m_condition_passed)
location->SetCondition (m_options.m_condition.c_str());
}
}
else
{
if (m_options.m_thread_id_passed)
bp->SetThreadID (m_options.m_thread_id);
if (m_options.m_thread_index_passed)
bp->GetOptions()->GetThreadSpec()->SetIndex(m_options.m_thread_index);
if (m_options.m_name_passed)
bp->GetOptions()->GetThreadSpec()->SetName(m_options.m_thread_name.c_str());
if (m_options.m_queue_passed)
bp->GetOptions()->GetThreadSpec()->SetQueueName(m_options.m_queue_name.c_str());
if (m_options.m_ignore_count != 0)
bp->GetOptions()->SetIgnoreCount(m_options.m_ignore_count);
if (m_options.m_enable_passed)
bp->SetEnabled (m_options.m_enable_value);
if (m_options.m_condition_passed)
bp->SetCondition (m_options.m_condition.c_str());
}
}
}
}
return result.Succeeded();
}