[lldb/telemetry] Report exit status only once

[labath]

SetExitStatus can be called the second time when we reap the debug server process. This shouldn't be interesting as at that point, we've already told everyone that the process has exited.

I believe/hope this will also help with sporadic shutdown crashes that have cropped up recently. They happen because the debug server is monitored from a detached thread, so this code can be called after main returns (and starts destroying everything). This isn't a real fix for that though, as the situation can still happen (it's just that it usually happens after the exit status has already been set). I think the real fix for that is to make sure these threads terminate before we start shutting everything down.

[llvmbot]

@llvm/pr-subscribers-lldb

Author: Pavel Labath (labath)

Changes

SetExitStatus can be called the second time when we reap the debug server process. This shouldn't be interesting as at that point, we've already told everyone that the process has exited.

I believe/hope this will also help with sporadic shutdown crashes that have cropped up recently. They happen because the debug server is monitored from a detached thread, so this code can be called after main returns (and starts destroying everything). This isn't a real fix for that though, as the situation can still happen (it's just that it usually happens after the exit status has already been set). I think the real fix for that is to make sure these threads terminate before we start shutting everything down.

---

Full diff: https://github.com/llvm/llvm-project/pull/134078.diff

1 Files Affected:

• (modified) lldb/source/Target/Process.cpp (+14-14)

diff --git a/lldb/source/Target/Process.cpp b/lldb/source/Target/Process.cpp
index 369933234ccca..7936cf28467b2 100644
--- a/lldb/source/Target/Process.cpp
+++ b/lldb/source/Target/Process.cpp
@@ -1067,6 +1067,20 @@ const char *Process::GetExitDescription() {
 bool Process::SetExitStatus(int status, llvm::StringRef exit_string) {
   // Use a mutex to protect setting the exit status.
   std::lock_guard<std::mutex> guard(m_exit_status_mutex);
+  Log *log(GetLog(LLDBLog::State | LLDBLog::Process));
+  LLDB_LOG(log, "(plugin = {0} status = {1} ({1:x8}), description=\"{2}\")",
+           GetPluginName(), status, exit_string);
+
+  // We were already in the exited state
+  if (m_private_state.GetValue() == eStateExited) {
+    LLDB_LOG(
+        log,
+        "(plugin = {0}) ignoring exit status because state was already set "
+        "to eStateExited",
+        GetPluginName());
+    return false;
+  }
+
   telemetry::ScopedDispatcher<telemetry::ProcessExitInfo> helper;
 
   UUID module_uuid;
@@ -1089,20 +1103,6 @@ bool Process::SetExitStatus(int status, llvm::StringRef exit_string) {
     info->pid = m_pid;
   });
 
-  Log *log(GetLog(LLDBLog::State | LLDBLog::Process));
-  LLDB_LOG(log, "(plugin = {0} status = {1} ({1:x8}), description=\"{2}\")",
-           GetPluginName(), status, exit_string);
-
-  // We were already in the exited state
-  if (m_private_state.GetValue() == eStateExited) {
-    LLDB_LOG(
-        log,
-        "(plugin = {0}) ignoring exit status because state was already set "
-        "to eStateExited",
-        GetPluginName());
-    return false;
-  }
-
   m_exit_status = status;
   if (!exit_string.empty())
     m_exit_string = exit_string.str();