[trace][intel-pt] Scaffold the 'thread trace start | stop' commands

Depends on D90490.

The stop command is simple and invokes the new method Trace::StopTracingThread(thread).

On the other hand, the start command works by delegating its implementation to a CommandObject provided by the Trace plugin. This is necessary because each trace plugin needs different options for this command. There's even the chance that a Trace plugin can't support live tracing, but instead supports offline decoding and analysis, which means that "thread trace dump instructions" works but "thread trace start" doest. Because of this and a few other reasons, it's better to have each plugin provide this implementation.

Besides, I'm using the GetSupportedTraceType method introduced in D90490 to quickly infer what's the trace plug-in that works for the current process.

As an implementation note, I moved CommandObjectIterateOverThreads to its header so that I can use it from the IntelPT plugin. Besides, the actual start and stop logic for intel-pt is not part of this diff.

Reviewed By: clayborg

Differential Revision: https://reviews.llvm.org/D90729

1 files changed, 158 insertions, 0 deletions

diff --git a/lldb/source/Commands/CommandObjectThreadUtil.cpp b/lldb/source/Commands/CommandObjectThreadUtil.cpp
new file mode 100644
index 0000000..b93698c
--- /dev/null
+++ b/lldb/source/Commands/CommandObjectThreadUtil.cpp
@@ -0,0 +1,158 @@
+//===-- CommandObjectThreadUtil.cpp -----------------------------*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#include "CommandObjectThreadUtil.h"
+
+#include "lldb/Interpreter/CommandReturnObject.h"
+#include "lldb/Target/Process.h"
+#include "lldb/Target/Thread.h"
+
+using namespace lldb;
+using namespace lldb_private;
+using namespace llvm;
+
+CommandObjectIterateOverThreads::CommandObjectIterateOverThreads(
+    CommandInterpreter &interpreter, const char *name, const char *help,
+    const char *syntax, uint32_t flags)
+    : CommandObjectParsed(interpreter, name, help, syntax, flags) {}
+
+bool CommandObjectIterateOverThreads::DoExecute(Args &command,
+                                                CommandReturnObject &result) {
+  result.SetStatus(m_success_return);
+
+  bool all_threads = false;
+  if (command.GetArgumentCount() == 0) {
+    Thread *thread = m_exe_ctx.GetThreadPtr();
+    if (!thread || !HandleOneThread(thread->GetID(), result))
+      return false;
+    return result.Succeeded();
+  } else if (command.GetArgumentCount() == 1) {
+    all_threads = ::strcmp(command.GetArgumentAtIndex(0), "all") == 0;
+    m_unique_stacks = ::strcmp(command.GetArgumentAtIndex(0), "unique") == 0;
+  }
+
+  // Use tids instead of ThreadSPs to prevent deadlocking problems which
+  // result from JIT-ing code while iterating over the (locked) ThreadSP
+  // list.
+  std::vector<lldb::tid_t> tids;
+
+  if (all_threads || m_unique_stacks) {
+    Process *process = m_exe_ctx.GetProcessPtr();
+
+    for (ThreadSP thread_sp : process->Threads())
+      tids.push_back(thread_sp->GetID());
+  } else {
+    const size_t num_args = command.GetArgumentCount();
+    Process *process = m_exe_ctx.GetProcessPtr();
+
+    std::lock_guard<std::recursive_mutex> guard(
+        process->GetThreadList().GetMutex());
+
+    for (size_t i = 0; i < num_args; i++) {
+      uint32_t thread_idx;
+      if (!llvm::to_integer(command.GetArgumentAtIndex(i), thread_idx)) {
+        result.AppendErrorWithFormat("invalid thread specification: \"%s\"\n",
+                                     command.GetArgumentAtIndex(i));
+        result.SetStatus(eReturnStatusFailed);
+        return false;
+      }
+
+      ThreadSP thread =
+          process->GetThreadList().FindThreadByIndexID(thread_idx);
+
+      if (!thread) {
+        result.AppendErrorWithFormat("no thread with index: \"%s\"\n",
+                                     command.GetArgumentAtIndex(i));
+        result.SetStatus(eReturnStatusFailed);
+        return false;
+      }
+
+      tids.push_back(thread->GetID());
+    }
+  }
+
+  if (m_unique_stacks) {
+    // Iterate over threads, finding unique stack buckets.
+    std::set<UniqueStack> unique_stacks;
+    for (const lldb::tid_t &tid : tids) {
+      if (!BucketThread(tid, unique_stacks, result)) {
+        return false;
+      }
+    }
+
+    // Write the thread id's and unique call stacks to the output stream
+    Stream &strm = result.GetOutputStream();
+    Process *process = m_exe_ctx.GetProcessPtr();
+    for (const UniqueStack &stack : unique_stacks) {
+      // List the common thread ID's
+      const std::vector<uint32_t> &thread_index_ids =
+          stack.GetUniqueThreadIndexIDs();
+      strm.Format("{0} thread(s) ", thread_index_ids.size());
+      for (const uint32_t &thread_index_id : thread_index_ids) {
+        strm.Format("#{0} ", thread_index_id);
+      }
+      strm.EOL();
+
+      // List the shared call stack for this set of threads
+      uint32_t representative_thread_id = stack.GetRepresentativeThread();
+      ThreadSP thread = process->GetThreadList().FindThreadByIndexID(
+          representative_thread_id);
+      if (!HandleOneThread(thread->GetID(), result)) {
+        return false;
+      }
+    }
+  } else {
+    uint32_t idx = 0;
+    for (const lldb::tid_t &tid : tids) {
+      if (idx != 0 && m_add_return)
+        result.AppendMessage("");
+
+      if (!HandleOneThread(tid, result))
+        return false;
+
+      ++idx;
+    }
+  }
+  return result.Succeeded();
+}
+
+bool CommandObjectIterateOverThreads::BucketThread(
+    lldb::tid_t tid, std::set<UniqueStack> &unique_stacks,
+    CommandReturnObject &result) {
+  // Grab the corresponding thread for the given thread id.
+  Process *process = m_exe_ctx.GetProcessPtr();
+  Thread *thread = process->GetThreadList().FindThreadByID(tid).get();
+  if (thread == nullptr) {
+    result.AppendErrorWithFormatv("Failed to process thread #{0}.\n", tid);
+    result.SetStatus(eReturnStatusFailed);
+    return false;
+  }
+
+  // Collect the each frame's address for this call-stack
+  std::stack<lldb::addr_t> stack_frames;
+  const uint32_t frame_count = thread->GetStackFrameCount();
+  for (uint32_t frame_index = 0; frame_index < frame_count; frame_index++) {
+    const lldb::StackFrameSP frame_sp =
+        thread->GetStackFrameAtIndex(frame_index);
+    const lldb::addr_t pc = frame_sp->GetStackID().GetPC();
+    stack_frames.push(pc);
+  }
+
+  uint32_t thread_index_id = thread->GetIndexID();
+  UniqueStack new_unique_stack(stack_frames, thread_index_id);
+
+  // Try to match the threads stack to and existing entry.
+  std::set<UniqueStack>::iterator matching_stack =
+      unique_stacks.find(new_unique_stack);
+  if (matching_stack != unique_stacks.end()) {
+    matching_stack->AddThread(thread_index_id);
+  } else {
+    unique_stacks.insert(new_unique_stack);
+  }
+  return true;
+}