Merge commit 'f2c339ca8b76' from apple/stable/20200714 into swift/main

Author: git apple-llvm automerger

lldb/include/lldb/Utility/ReproducerInstrumentation.h

@@ -17,6 +17,7 @@
 #include "llvm/Support/ErrorHandling.h"
 
 #include <map>
+#include <thread>
 #include <type_traits>
 
 template <typename T,
@@ -78,6 +79,13 @@ template <typename... Ts> inline std::string stringify_args(const Ts &... ts) {
 // is initialized or enabled.
 // #define LLDB_REPRO_INSTR_TRACE
 
+#ifdef LLDB_REPRO_INSTR_TRACE
+inline llvm::raw_ostream &this_thread_id() {
+  size_t tid = std::hash<std::thread::id>{}(std::this_thread::get_id());
+  return llvm::errs().write_hex(tid) << " :: ";
+}
+#endif
+
 #define LLDB_REGISTER_CONSTRUCTOR(Class, Signature)                            \
   R.Register<Class * Signature>(&construct<Class Signature>::record, "",       \
                                 #Class, #Class, #Signature)
@@ -325,6 +333,7 @@ class Deserializer {
   }
 
   template <typename T> const T &HandleReplayResult(const T &t) {
+    CheckSequence(Deserialize<unsigned>());
     unsigned result = Deserialize<unsigned>();
     if (is_trivially_serializable<T>::value)
       return t;
@@ -334,6 +343,7 @@ class Deserializer {
 
   /// Store the returned value in the index-to-object mapping.
   template <typename T> T &HandleReplayResult(T &t) {
+    CheckSequence(Deserialize<unsigned>());
     unsigned result = Deserialize<unsigned>();
     if (is_trivially_serializable<T>::value)
       return t;
@@ -343,6 +353,7 @@ class Deserializer {
 
   /// Store the returned value in the index-to-object mapping.
   template <typename T> T *HandleReplayResult(T *t) {
+    CheckSequence(Deserialize<unsigned>());
     unsigned result = Deserialize<unsigned>();
     if (is_trivially_serializable<T>::value)
       return t;
@@ -352,6 +363,7 @@ class Deserializer {
   /// All returned types are recorded, even when the function returns a void.
   /// The latter requires special handling.
   void HandleReplayResultVoid() {
+    CheckSequence(Deserialize<unsigned>());
     unsigned result = Deserialize<unsigned>();
     assert(result == 0);
     (void)result;
@@ -361,6 +373,10 @@ class Deserializer {
     return m_index_to_object.GetAllObjects();
   }
 
+  void SetExpectedSequence(unsigned sequence) {
+    m_expected_sequence = sequence;
+  }
+
 private:
   template <typename T> T Read(ValueTag) {
     assert(HasData(sizeof(T)));
@@ -402,11 +418,17 @@ class Deserializer {
     return *(new UnderlyingT(Deserialize<UnderlyingT>()));
   }
 
+  /// Verify that the given sequence number matches what we expect.
+  void CheckSequence(unsigned sequence);
+
   /// Mapping of indices to objects.
   IndexToObject m_index_to_object;
 
   /// Buffer containing the serialized data.
   llvm::StringRef m_buffer;
+
+  /// The result's expected sequence number.
+  llvm::Optional<unsigned> m_expected_sequence;
 };
 
 /// Partial specialization for C-style strings. We read the string value
@@ -600,8 +622,8 @@ class Serializer {
   /// objects (in which case we serialize their index).
   template <typename T> void Serialize(T *t) {
 #ifdef LLDB_REPRO_INSTR_TRACE
-    llvm::errs() << "Serializing with " << LLVM_PRETTY_FUNCTION << " -> "
-                 << stringify_args(t) << "\n";
+    this_thread_id() << "Serializing with " << LLVM_PRETTY_FUNCTION << " -> "
+                     << stringify_args(t) << "\n";
 #endif
     if (std::is_fundamental<T>::value) {
       Serialize(*t);
@@ -616,8 +638,8 @@ class Serializer {
   /// to objects (in which case we serialize their index).
   template <typename T> void Serialize(T &t) {
 #ifdef LLDB_REPRO_INSTR_TRACE
-    llvm::errs() << "Serializing with " << LLVM_PRETTY_FUNCTION << " -> "
-                 << stringify_args(t) << "\n";
+    this_thread_id() << "Serializing with " << LLVM_PRETTY_FUNCTION << " -> "
+                     << stringify_args(t) << "\n";
 #endif
     if (is_trivially_serializable<T>::value) {
       m_stream.write(reinterpret_cast<const char *>(&t), sizeof(T));
@@ -637,8 +659,8 @@ class Serializer {
 
   void Serialize(const char *t) {
 #ifdef LLDB_REPRO_INSTR_TRACE
-    llvm::errs() << "Serializing with " << LLVM_PRETTY_FUNCTION << " -> "
-                 << stringify_args(t) << "\n";
+    this_thread_id() << "Serializing with " << LLVM_PRETTY_FUNCTION << " -> "
+                     << stringify_args(t) << "\n";
 #endif
     const size_t size = t ? strlen(t) : std::numeric_limits<size_t>::max();
     Serialize(size);
@@ -737,19 +759,23 @@ class Recorder {
     if (!ShouldCapture())
       return;
 
+    std::lock_guard<std::mutex> lock(g_mutex);
+    unsigned sequence = GetSequenceNumber();
     unsigned id = registry.GetID(uintptr_t(f));
 
 #ifdef LLDB_REPRO_INSTR_TRACE
     Log(id);
 #endif
 
+    serializer.SerializeAll(sequence);
     serializer.SerializeAll(id);
     serializer.SerializeAll(args...);
 
     if (std::is_class<typename std::remove_pointer<
             typename std::remove_reference<Result>::type>::type>::value) {
       m_result_recorded = false;
     } else {
+      serializer.SerializeAll(sequence);
       serializer.SerializeAll(0);
       m_result_recorded = true;
     }
@@ -763,16 +789,20 @@ class Recorder {
     if (!ShouldCapture())
       return;
 
+    std::lock_guard<std::mutex> lock(g_mutex);
+    unsigned sequence = GetSequenceNumber();
     unsigned id = registry.GetID(uintptr_t(f));
 
 #ifdef LLDB_REPRO_INSTR_TRACE
     Log(id);
 #endif
 
+    serializer.SerializeAll(sequence);
     serializer.SerializeAll(id);
     serializer.SerializeAll(args...);
 
     // Record result.
+    serializer.SerializeAll(sequence);
     serializer.SerializeAll(0);
     m_result_recorded = true;
   }
@@ -798,7 +828,9 @@ class Recorder {
     if (update_boundary)
       UpdateBoundary();
     if (m_serializer && ShouldCapture()) {
+      std::lock_guard<std::mutex> lock(g_mutex);
       assert(!m_result_recorded);
+      m_serializer->SerializeAll(GetSequenceNumber());
       m_serializer->SerializeAll(r);
       m_result_recorded = true;
     }
@@ -808,6 +840,7 @@ class Recorder {
   template <typename Result, typename T>
   Result Replay(Deserializer &deserializer, Registry &registry, uintptr_t addr,
                 bool update_boundary) {
+    deserializer.SetExpectedSequence(deserializer.Deserialize<unsigned>());
     unsigned actual_id = registry.GetID(addr);
     unsigned id = deserializer.Deserialize<unsigned>();
     registry.CheckID(id, actual_id);
@@ -818,6 +851,7 @@ class Recorder {
   }
 
   void Replay(Deserializer &deserializer, Registry &registry, uintptr_t addr) {
+    deserializer.SetExpectedSequence(deserializer.Deserialize<unsigned>());
     unsigned actual_id = registry.GetID(addr);
     unsigned id = deserializer.Deserialize<unsigned>();
     registry.CheckID(id, actual_id);
@@ -833,7 +867,14 @@ class Recorder {
 
   bool ShouldCapture() { return m_local_boundary; }
 
+  /// Mark the current thread as a private thread and pretend that everything
+  /// on this thread is behind happening behind the API boundary.
+  static void PrivateThread() { g_global_boundary = true; }
+
 private:
+  static unsigned GetNextSequenceNumber() { return g_sequence++; }
+  unsigned GetSequenceNumber() const;
+
   template <typename T> friend struct replay;
   void UpdateBoundary() {
     if (m_local_boundary)
@@ -842,8 +883,8 @@ class Recorder {
 
 #ifdef LLDB_REPRO_INSTR_TRACE
   void Log(unsigned id) {
-    llvm::errs() << "Recording " << id << ": " << m_pretty_func << " ("
-                 << m_pretty_args << ")\n";
+    this_thread_id() << "Recording " << id << ": " << m_pretty_func << " ("
+                     << m_pretty_args << ")\n";
   }
 #endif
 
@@ -859,8 +900,17 @@ class Recorder {
   /// Whether the return value was recorded explicitly.
   bool m_result_recorded;
 
+  /// The sequence number for this pair of function and result.
+  unsigned m_sequence;
+
   /// Whether we're currently across the API boundary.
-  static bool g_global_boundary;
+  static thread_local bool g_global_boundary;
+
+  /// Global mutex to protect concurrent access.
+  static std::mutex g_mutex;
+
+  /// Unique, monotonically increasing sequence number.
+  static std::atomic<unsigned> g_sequence;
 };
 
 /// To be used as the "Runtime ID" of a constructor. It also invokes the
@@ -1002,6 +1052,7 @@ struct invoke_char_ptr<Result (Class::*)(Args...) const> {
 
     static Result replay(Recorder &recorder, Deserializer &deserializer,
                          Registry &registry, char *str) {
+      deserializer.SetExpectedSequence(deserializer.Deserialize<unsigned>());
       deserializer.Deserialize<unsigned>();
       Class *c = deserializer.Deserialize<Class *>();
       deserializer.Deserialize<const char *>();
@@ -1023,6 +1074,7 @@ struct invoke_char_ptr<Result (Class::*)(Args...)> {
 
     static Result replay(Recorder &recorder, Deserializer &deserializer,
                          Registry &registry, char *str) {
+      deserializer.SetExpectedSequence(deserializer.Deserialize<unsigned>());
       deserializer.Deserialize<unsigned>();
       Class *c = deserializer.Deserialize<Class *>();
       deserializer.Deserialize<const char *>();
@@ -1043,6 +1095,7 @@ struct invoke_char_ptr<Result (*)(Args...)> {
 
     static Result replay(Recorder &recorder, Deserializer &deserializer,
                          Registry &registry, char *str) {
+      deserializer.SetExpectedSequence(deserializer.Deserialize<unsigned>());
       deserializer.Deserialize<unsigned>();
       deserializer.Deserialize<const char *>();
       size_t l = deserializer.Deserialize<size_t>();

lldb/source/Plugins/ScriptInterpreter/Python/ScriptInterpreterPython.cpp

@@ -32,6 +32,7 @@
 #include "lldb/Interpreter/CommandReturnObject.h"
 #include "lldb/Target/Thread.h"
 #include "lldb/Target/ThreadPlan.h"
+#include "lldb/Utility/ReproducerInstrumentation.h"
 #include "lldb/Utility/Timer.h"
 #include "llvm/ADT/STLExtras.h"
 #include "llvm/ADT/StringRef.h"
@@ -430,6 +431,7 @@ ScriptInterpreterPythonImpl::Locker::Locker(
     : ScriptInterpreterLocker(),
       m_teardown_session((on_leave & TearDownSession) == TearDownSession),
       m_python_interpreter(py_interpreter) {
+  repro::Recorder::PrivateThread();
   DoAcquireLock();
   if ((on_entry & InitSession) == InitSession) {
     if (!DoInitSession(on_entry, in, out, err)) {

lldb/source/Utility/ReproducerInstrumentation.cpp

@@ -8,6 +8,7 @@
 
 #include "lldb/Utility/ReproducerInstrumentation.h"
 #include "lldb/Utility/Reproducer.h"
+#include <limits>
 #include <stdio.h>
 #include <stdlib.h>
 #include <thread>
@@ -84,6 +85,16 @@ template <> const char **Deserializer::Deserialize<const char **>() {
   return r;
 }
 
+void Deserializer::CheckSequence(unsigned sequence) {
+  if (m_expected_sequence && *m_expected_sequence != sequence)
+    llvm::report_fatal_error(
+        "The result does not match the preceding "
+        "function. This is probably the result of concurrent "
+        "use of the SB API during capture, which is currently not "
+        "supported.");
+  m_expected_sequence.reset();
+}
+
 bool Registry::Replay(const FileSpec &file) {
   auto error_or_file = llvm::MemoryBuffer::getFile(file.GetPath());
   if (auto err = error_or_file.getError())
@@ -107,6 +118,7 @@ bool Registry::Replay(Deserializer &deserializer) {
   setvbuf(stdout, nullptr, _IONBF, 0);
 
   while (deserializer.HasData(1)) {
+    unsigned sequence = deserializer.Deserialize<unsigned>();
     unsigned id = deserializer.Deserialize<unsigned>();
 
 #ifndef LLDB_REPRO_INSTR_TRACE
@@ -115,6 +127,7 @@ bool Registry::Replay(Deserializer &deserializer) {
     llvm::errs() << "Replaying " << id << ": " << GetSignature(id) << "\n";
 #endif
 
+    deserializer.SetExpectedSequence(sequence);
     GetReplayer(id)->operator()(deserializer);
   }
 
@@ -181,21 +194,24 @@ unsigned ObjectToIndex::GetIndexForObjectImpl(const void *object) {
 
 Recorder::Recorder()
     : m_serializer(nullptr), m_pretty_func(), m_pretty_args(),
-      m_local_boundary(false), m_result_recorded(true) {
+      m_local_boundary(false), m_result_recorded(true),
+      m_sequence(std::numeric_limits<unsigned>::max()) {
   if (!g_global_boundary) {
     g_global_boundary = true;
     m_local_boundary = true;
+    m_sequence = GetNextSequenceNumber();
   }
 }
 
 Recorder::Recorder(llvm::StringRef pretty_func, std::string &&pretty_args)
     : m_serializer(nullptr), m_pretty_func(pretty_func),
       m_pretty_args(pretty_args), m_local_boundary(false),
-      m_result_recorded(true) {
+      m_result_recorded(true),
+      m_sequence(std::numeric_limits<unsigned>::max()) {
   if (!g_global_boundary) {
     g_global_boundary = true;
     m_local_boundary = true;
-
+    m_sequence = GetNextSequenceNumber();
     LLDB_LOG(GetLogIfAllCategoriesSet(LIBLLDB_LOG_API), "{0} ({1})",
              m_pretty_func, m_pretty_args);
   }
@@ -206,6 +222,11 @@ Recorder::~Recorder() {
   UpdateBoundary();
 }
 
+unsigned Recorder::GetSequenceNumber() const {
+  assert(m_sequence != std::numeric_limits<unsigned>::max());
+  return m_sequence;
+}
+
 void InstrumentationData::Initialize(Serializer &serializer,
                                      Registry &registry) {
   InstanceImpl().emplace(serializer, registry);
@@ -227,4 +248,6 @@ llvm::Optional<InstrumentationData> &InstrumentationData::InstanceImpl() {
   return g_instrumentation_data;
 }
 
-bool lldb_private::repro::Recorder::g_global_boundary;
+thread_local bool lldb_private::repro::Recorder::g_global_boundary = false;
+std::atomic<unsigned> lldb_private::repro::Recorder::g_sequence;
+std::mutex lldb_private::repro::Recorder::g_mutex;