hanchenye-llvm-project/lldb/source/Target/ThreadPlanStepRange.cpp

559 lines
20 KiB
C++
Raw Normal View History

//===-- ThreadPlanStepRange.cpp ---------------------------------*- C++ -*-===//
//
// The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
#include "lldb/Target/ThreadPlanStepRange.h"
// C Includes
// C++ Includes
// Other libraries and framework includes
// Project includes
#include "lldb/lldb-private-log.h"
#include "lldb/Breakpoint/BreakpointLocation.h"
#include "lldb/Breakpoint/BreakpointSite.h"
#include "lldb/Core/Disassembler.h"
#include "lldb/Core/Log.h"
#include "lldb/Core/Stream.h"
#include "lldb/Symbol/Function.h"
#include "lldb/Symbol/Symbol.h"
#include "lldb/Target/ExecutionContext.h"
Abtracted the old "lldb_private::Thread::StopInfo" into an abtract class. This will allow debugger plug-ins to make any instance of "lldb_private::StopInfo" that can completely describe any stop reason. It also provides a framework for doing intelligent things with the stop info at important times in the lifetime of the inferior. Examples include the signal stop info in StopInfoUnixSignal. It will check with the process to see that the current action is for the signal. These actions include wether to stop for the signal, wether the notify that the signal was hit, and wether to pass the signal along to the inferior process. The StopInfoUnixSignal class overrides the "ShouldStop()" method of StopInfo and this allows the stop info to determine if it should stop at the signal or continue the process. StopInfo subclasses must override the following functions: virtual lldb::StopReason GetStopReason () const = 0; virtual const char * GetDescription () = 0; StopInfo subclasses can override the following functions: // If the subclass returns "false", the inferior will resume. The default // version of this function returns "true" which means the default stop // info will stop the process. The breakpoint subclass will check if // the breakpoint wants us to stop by calling any installed callback on // the breakpoint, and also checking if the breakpoint is for the current // thread. Signals will check if they should stop based off of the // UnixSignal settings in the process. virtual bool ShouldStop (Event *event_ptr); // Sublasses can state if they want to notify the debugger when "ShouldStop" // returns false. This would be handy for breakpoints where you want to // log information and continue and is also used by the signal stop info // to notify that a signal was received (after it checks with the process // signal settings). virtual bool ShouldNotify (Event *event_ptr) { return false; } // Allow subclasses to do something intelligent right before we resume. // The signal class will figure out if the signal should be propagated // to the inferior process and pass that along to the debugger plug-ins. virtual void WillResume (lldb::StateType resume_state) { // By default, don't do anything } The support the Mach exceptions was moved into the lldb/source/Plugins/Process/Utility folder and now doesn't polute the lldb_private::Thread class with platform specific code. llvm-svn: 110184
2010-08-04 09:40:35 +08:00
#include "lldb/Target/Process.h"
#include "lldb/Target/RegisterContext.h"
#include "lldb/Target/StopInfo.h"
#include "lldb/Target/Target.h"
Abtracted the old "lldb_private::Thread::StopInfo" into an abtract class. This will allow debugger plug-ins to make any instance of "lldb_private::StopInfo" that can completely describe any stop reason. It also provides a framework for doing intelligent things with the stop info at important times in the lifetime of the inferior. Examples include the signal stop info in StopInfoUnixSignal. It will check with the process to see that the current action is for the signal. These actions include wether to stop for the signal, wether the notify that the signal was hit, and wether to pass the signal along to the inferior process. The StopInfoUnixSignal class overrides the "ShouldStop()" method of StopInfo and this allows the stop info to determine if it should stop at the signal or continue the process. StopInfo subclasses must override the following functions: virtual lldb::StopReason GetStopReason () const = 0; virtual const char * GetDescription () = 0; StopInfo subclasses can override the following functions: // If the subclass returns "false", the inferior will resume. The default // version of this function returns "true" which means the default stop // info will stop the process. The breakpoint subclass will check if // the breakpoint wants us to stop by calling any installed callback on // the breakpoint, and also checking if the breakpoint is for the current // thread. Signals will check if they should stop based off of the // UnixSignal settings in the process. virtual bool ShouldStop (Event *event_ptr); // Sublasses can state if they want to notify the debugger when "ShouldStop" // returns false. This would be handy for breakpoints where you want to // log information and continue and is also used by the signal stop info // to notify that a signal was received (after it checks with the process // signal settings). virtual bool ShouldNotify (Event *event_ptr) { return false; } // Allow subclasses to do something intelligent right before we resume. // The signal class will figure out if the signal should be propagated // to the inferior process and pass that along to the debugger plug-ins. virtual void WillResume (lldb::StateType resume_state) { // By default, don't do anything } The support the Mach exceptions was moved into the lldb/source/Plugins/Process/Utility folder and now doesn't polute the lldb_private::Thread class with platform specific code. llvm-svn: 110184
2010-08-04 09:40:35 +08:00
#include "lldb/Target/Thread.h"
#include "lldb/Target/ThreadPlanRunToAddress.h"
using namespace lldb;
using namespace lldb_private;
//----------------------------------------------------------------------
// ThreadPlanStepRange: Step through a stack range, either stepping over or into
// based on the value of \a type.
//----------------------------------------------------------------------
2011-02-08 12:27:50 +08:00
ThreadPlanStepRange::ThreadPlanStepRange (ThreadPlanKind kind,
const char *name,
Thread &thread,
const AddressRange &range,
const SymbolContext &addr_context,
lldb::RunMode stop_others,
bool given_ranges_only) :
ThreadPlan (kind, name, thread, eVoteNoOpinion, eVoteNoOpinion),
m_addr_context (addr_context),
m_address_ranges (),
m_stop_others (stop_others),
m_stack_id (),
m_parent_stack_id(),
m_no_more_plans (false),
m_first_run_event (true),
m_use_fast_step(false),
m_given_ranges_only (given_ranges_only)
{
m_use_fast_step = GetTarget().GetUseFastStepping();
AddRange(range);
m_stack_id = m_thread.GetStackFrameAtIndex(0)->GetStackID();
StackFrameSP parent_stack = m_thread.GetStackFrameAtIndex(1);
if (parent_stack)
m_parent_stack_id = parent_stack->GetStackID();
}
ThreadPlanStepRange::~ThreadPlanStepRange ()
{
ClearNextBranchBreakpoint();
size_t num_instruction_ranges = m_instruction_ranges.size();
// FIXME: The DisassemblerLLVMC has a reference cycle and won't go away if it has any active instructions.
// I'll fix that but for now, just clear the list and it will go away nicely.
for (size_t i = 0; i < num_instruction_ranges; i++)
{
if (m_instruction_ranges[i])
m_instruction_ranges[i]->GetInstructionList().Clear();
}
}
void
ThreadPlanStepRange::DidPush ()
{
// See if we can find a "next range" breakpoint:
SetNextBranchBreakpoint();
}
bool
ThreadPlanStepRange::ValidatePlan (Stream *error)
{
return true;
}
Vote
ThreadPlanStepRange::ShouldReportStop (Event *event_ptr)
{
Log *log(lldb_private::GetLogIfAllCategoriesSet (LIBLLDB_LOG_STEP));
const Vote vote = IsPlanComplete() ? eVoteYes : eVoteNo;
if (log)
log->Printf ("ThreadPlanStepRange::ShouldReportStop() returning vote %i\n", vote);
return vote;
}
void
ThreadPlanStepRange::AddRange(const AddressRange &new_range)
{
// For now I'm just adding the ranges. At some point we may want to
// condense the ranges if they overlap, though I don't think it is likely
// to be very important.
m_address_ranges.push_back (new_range);
// Fill the slot for this address range with an empty DisassemblerSP in the instruction ranges. I want the
// indices to match, but I don't want to do the work to disassemble this range if I don't step into it.
m_instruction_ranges.push_back (DisassemblerSP());
}
void
ThreadPlanStepRange::DumpRanges(Stream *s)
{
size_t num_ranges = m_address_ranges.size();
if (num_ranges == 1)
{
m_address_ranges[0].Dump (s, m_thread.CalculateTarget().get(), Address::DumpStyleLoadAddress);
}
else
{
for (size_t i = 0; i < num_ranges; i++)
{
s->PutCString("%d: ");
m_address_ranges[i].Dump (s, m_thread.CalculateTarget().get(), Address::DumpStyleLoadAddress);
}
}
}
bool
ThreadPlanStepRange::InRange ()
{
Log *log(lldb_private::GetLogIfAllCategoriesSet (LIBLLDB_LOG_STEP));
bool ret_value = false;
lldb::addr_t pc_load_addr = m_thread.GetRegisterContext()->GetPC();
size_t num_ranges = m_address_ranges.size();
for (size_t i = 0; i < num_ranges; i++)
{
ret_value = m_address_ranges[i].ContainsLoadAddress(pc_load_addr, m_thread.CalculateTarget().get());
if (ret_value)
break;
}
if (!ret_value && !m_given_ranges_only)
{
// See if we've just stepped to another part of the same line number...
StackFrame *frame = m_thread.GetStackFrameAtIndex(0).get();
SymbolContext new_context(frame->GetSymbolContext(eSymbolContextEverything));
if (m_addr_context.line_entry.IsValid() && new_context.line_entry.IsValid())
{
if (m_addr_context.line_entry.file == new_context.line_entry.file)
{
if (m_addr_context.line_entry.line == new_context.line_entry.line)
{
m_addr_context = new_context;
AddRange(m_addr_context.line_entry.range);
ret_value = true;
if (log)
{
StreamString s;
m_addr_context.line_entry.Dump (&s,
m_thread.CalculateTarget().get(),
true,
Address::DumpStyleLoadAddress,
Address::DumpStyleLoadAddress,
true);
log->Printf ("Step range plan stepped to another range of same line: %s", s.GetData());
}
}
else if (new_context.line_entry.line == 0)
{
new_context.line_entry.line = m_addr_context.line_entry.line;
m_addr_context = new_context;
AddRange(m_addr_context.line_entry.range);
ret_value = true;
if (log)
{
StreamString s;
m_addr_context.line_entry.Dump (&s,
m_thread.CalculateTarget().get(),
true,
Address::DumpStyleLoadAddress,
Address::DumpStyleLoadAddress,
true);
log->Printf ("Step range plan stepped to a range at linenumber 0 stepping through that range: %s", s.GetData());
}
}
else if (new_context.line_entry.range.GetBaseAddress().GetLoadAddress(m_thread.CalculateTarget().get())
!= pc_load_addr)
{
// Another thing that sometimes happens here is that we step out of one line into the MIDDLE of another
// line. So far I mostly see this due to bugs in the debug information.
// But we probably don't want to be in the middle of a line range, so in that case reset the stepping
// range to the line we've stepped into the middle of and continue.
m_addr_context = new_context;
m_address_ranges.clear();
AddRange(m_addr_context.line_entry.range);
ret_value = true;
if (log)
{
StreamString s;
m_addr_context.line_entry.Dump (&s,
m_thread.CalculateTarget().get(),
true,
Address::DumpStyleLoadAddress,
Address::DumpStyleLoadAddress,
true);
log->Printf ("Step range plan stepped to the middle of new line(%d): %s, continuing to clear this line.",
new_context.line_entry.line,
s.GetData());
}
}
}
}
}
if (!ret_value && log)
log->Printf ("Step range plan out of range to 0x%" PRIx64, pc_load_addr);
return ret_value;
}
bool
ThreadPlanStepRange::InSymbol()
{
lldb::addr_t cur_pc = m_thread.GetRegisterContext()->GetPC();
if (m_addr_context.function != NULL)
{
return m_addr_context.function->GetAddressRange().ContainsLoadAddress (cur_pc, m_thread.CalculateTarget().get());
}
else if (m_addr_context.symbol)
{
AddressRange range(m_addr_context.symbol->GetAddress(), m_addr_context.symbol->GetByteSize());
return range.ContainsLoadAddress (cur_pc, m_thread.CalculateTarget().get());
}
return false;
}
// FIXME: This should also handle inlining if we aren't going to do inlining in the
// main stack.
//
// Ideally we should remember the whole stack frame list, and then compare that
// to the current list.
lldb::FrameComparison
ThreadPlanStepRange::CompareCurrentFrameToStartFrame()
{
FrameComparison frame_order;
StackID cur_frame_id = m_thread.GetStackFrameAtIndex(0)->GetStackID();
if (cur_frame_id == m_stack_id)
{
frame_order = eFrameCompareEqual;
}
else if (cur_frame_id < m_stack_id)
{
frame_order = eFrameCompareYounger;
}
else
{
StackFrameSP cur_parent_frame = m_thread.GetStackFrameAtIndex(1);
StackID cur_parent_id;
if (cur_parent_frame)
cur_parent_id = cur_parent_frame->GetStackID();
if (m_parent_stack_id.IsValid()
&& cur_parent_id.IsValid()
&& m_parent_stack_id == cur_parent_id)
frame_order = eFrameCompareSameParent;
else
frame_order = eFrameCompareOlder;
}
return frame_order;
}
bool
ThreadPlanStepRange::StopOthers ()
{
if (m_stop_others == lldb::eOnlyThisThread
|| m_stop_others == lldb::eOnlyDuringStepping)
return true;
else
return false;
}
InstructionList *
ThreadPlanStepRange::GetInstructionsForAddress(lldb::addr_t addr, size_t &range_index, size_t &insn_offset)
{
size_t num_ranges = m_address_ranges.size();
for (size_t i = 0; i < num_ranges; i++)
{
if (m_address_ranges[i].ContainsLoadAddress(addr, &GetTarget()))
{
// Some joker added a zero size range to the stepping range...
if (m_address_ranges[i].GetByteSize() == 0)
return NULL;
if (!m_instruction_ranges[i])
{
//Disassemble the address range given:
ExecutionContext exe_ctx (m_thread.GetProcess());
const char *plugin_name = NULL;
const char *flavor = NULL;
const bool prefer_file_cache = true;
m_instruction_ranges[i] = Disassembler::DisassembleRange(GetTarget().GetArchitecture(),
plugin_name,
flavor,
exe_ctx,
m_address_ranges[i],
prefer_file_cache);
}
if (!m_instruction_ranges[i])
return NULL;
else
{
// Find where we are in the instruction list as well. If we aren't at an instruction,
// return NULL. In this case, we're probably lost, and shouldn't try to do anything fancy.
insn_offset = m_instruction_ranges[i]->GetInstructionList().GetIndexOfInstructionAtLoadAddress(addr, GetTarget());
if (insn_offset == UINT32_MAX)
return NULL;
else
{
range_index = i;
return &m_instruction_ranges[i]->GetInstructionList();
}
}
}
}
return NULL;
}
void
ThreadPlanStepRange::ClearNextBranchBreakpoint()
{
if (m_next_branch_bp_sp)
{
Log *log(lldb_private::GetLogIfAllCategoriesSet (LIBLLDB_LOG_STEP));
if (log)
log->Printf ("Removing next branch breakpoint: %d.", m_next_branch_bp_sp->GetID());
GetTarget().RemoveBreakpointByID (m_next_branch_bp_sp->GetID());
m_next_branch_bp_sp.reset();
}
}
bool
ThreadPlanStepRange::SetNextBranchBreakpoint ()
{
if (m_next_branch_bp_sp)
return true;
Log *log(lldb_private::GetLogIfAllCategoriesSet (LIBLLDB_LOG_STEP));
// Stepping through ranges using breakpoints doesn't work yet, but with this off we fall back to instruction
// single stepping.
if (!m_use_fast_step)
return false;
lldb::addr_t cur_addr = GetThread().GetRegisterContext()->GetPC();
// Find the current address in our address ranges, and fetch the disassembly if we haven't already:
size_t pc_index;
size_t range_index;
InstructionList *instructions = GetInstructionsForAddress (cur_addr, range_index, pc_index);
if (instructions == NULL)
return false;
else
{
uint32_t branch_index;
branch_index = instructions->GetIndexOfNextBranchInstruction (pc_index);
Address run_to_address;
// If we didn't find a branch, run to the end of the range.
if (branch_index == UINT32_MAX)
{
branch_index = instructions->GetSize() - 1;
}
if (branch_index - pc_index > 1)
{
const bool is_internal = true;
run_to_address = instructions->GetInstructionAtIndex(branch_index)->GetAddress();
m_next_branch_bp_sp = GetTarget().CreateBreakpoint(run_to_address, is_internal, false);
if (m_next_branch_bp_sp)
{
if (log)
{
lldb::break_id_t bp_site_id = LLDB_INVALID_BREAK_ID;
BreakpointLocationSP bp_loc = m_next_branch_bp_sp->GetLocationAtIndex(0);
if (bp_loc)
{
BreakpointSiteSP bp_site = bp_loc->GetBreakpointSite();
if (bp_site)
{
bp_site_id = bp_site->GetID();
}
}
log->Printf ("ThreadPlanStepRange::SetNextBranchBreakpoint - Setting breakpoint %d (site %d) to run to address 0x%" PRIx64,
m_next_branch_bp_sp->GetID(),
bp_site_id,
run_to_address.GetLoadAddress(&m_thread.GetProcess()->GetTarget()));
}
m_next_branch_bp_sp->SetThreadID(m_thread.GetID());
m_next_branch_bp_sp->SetBreakpointKind ("next-branch-location");
return true;
}
else
return false;
}
}
return false;
}
bool
ThreadPlanStepRange::NextRangeBreakpointExplainsStop (lldb::StopInfoSP stop_info_sp)
{
Log *log(lldb_private::GetLogIfAllCategoriesSet (LIBLLDB_LOG_STEP));
if (!m_next_branch_bp_sp)
return false;
break_id_t bp_site_id = stop_info_sp->GetValue();
BreakpointSiteSP bp_site_sp = m_thread.GetProcess()->GetBreakpointSiteList().FindByID(bp_site_id);
if (!bp_site_sp)
return false;
else if (!bp_site_sp->IsBreakpointAtThisSite (m_next_branch_bp_sp->GetID()))
return false;
else
{
// If we've hit the next branch breakpoint, then clear it.
size_t num_owners = bp_site_sp->GetNumberOfOwners();
bool explains_stop = true;
// If all the owners are internal, then we are probably just stepping over this range from multiple threads,
// or multiple frames, so we want to continue. If one is not internal, then we should not explain the stop,
// and let the user breakpoint handle the stop.
for (size_t i = 0; i < num_owners; i++)
{
if (!bp_site_sp->GetOwnerAtIndex(i)->GetBreakpoint().IsInternal())
{
explains_stop = false;
break;
}
}
if (log)
log->Printf ("ThreadPlanStepRange::NextRangeBreakpointExplainsStop - Hit next range breakpoint which has %" PRIu64 " owners - explains stop: %u.",
(uint64_t)num_owners,
explains_stop);
ClearNextBranchBreakpoint();
return explains_stop;
}
}
bool
ThreadPlanStepRange::WillStop ()
{
return true;
}
StateType
ThreadPlanStepRange::GetPlanRunState ()
{
if (m_next_branch_bp_sp)
return eStateRunning;
else
return eStateStepping;
}
bool
ThreadPlanStepRange::MischiefManaged ()
{
// If we have pushed some plans between ShouldStop & MischiefManaged, then we're not done...
// I do this check first because we might have stepped somewhere that will fool InRange into
// thinking it needs to step past the end of that line. This happens, for instance, when stepping
// over inlined code that is in the middle of the current line.
if (!m_no_more_plans)
return false;
bool done = true;
if (!IsPlanComplete())
{
if (InRange())
{
done = false;
}
else
{
FrameComparison frame_order = CompareCurrentFrameToStartFrame();
if (frame_order != eFrameCompareOlder)
{
if (m_no_more_plans)
done = true;
else
done = false;
}
else
done = true;
}
}
if (done)
{
Log *log(lldb_private::GetLogIfAllCategoriesSet (LIBLLDB_LOG_STEP));
if (log)
log->Printf("Completed step through range plan.");
ClearNextBranchBreakpoint();
ThreadPlan::MischiefManaged ();
return true;
}
else
{
return false;
}
}
bool
ThreadPlanStepRange::IsPlanStale ()
{
Log *log(lldb_private::GetLogIfAllCategoriesSet (LIBLLDB_LOG_STEP));
FrameComparison frame_order = CompareCurrentFrameToStartFrame();
if (frame_order == eFrameCompareOlder)
{
if (log)
{
log->Printf("ThreadPlanStepRange::IsPlanStale returning true, we've stepped out.");
}
return true;
}
else if (frame_order == eFrameCompareEqual && InSymbol())
{
// If we are not in a place we should step through, we've gotten stale.
// One tricky bit here is that some stubs don't push a frame, so we should.
// check that we are in the same symbol.
if (!InRange())
{
return true;
}
}
return false;
}