Index: lib/sanitizer_common/sanitizer_allocator.cc
===================================================================
--- lib/sanitizer_common/sanitizer_allocator.cc
+++ lib/sanitizer_common/sanitizer_allocator.cc
@@ -180,6 +180,25 @@
 // LowLevelAllocator
 static LowLevelAllocateCallback low_level_alloc_callback;
 
+int InternalReallocArr(void *addr, uptr count, uptr size,
+                       InternalAllocatorCache *cache) {
+  void *oaddr;
+  void *naddr = 0;
+
+  internal_memcpy(&oaddr, addr, sizeof(addr));
+  if (!count || !size) {
+    InternalFree(oaddr, cache);
+    internal_memcpy(addr, &naddr, sizeof(addr));
+    return 0;
+  }
+
+  naddr = InternalRealloc(oaddr, count * size, cache);
+  if (naddr == 0) return -1;
+
+  internal_memcpy(addr, &naddr, sizeof(addr));
+  return 0;
+}
+
 void *LowLevelAllocator::Allocate(uptr size) {
   // Align allocation size.
   size = RoundUpTo(size, 8);
Index: lib/sanitizer_common/sanitizer_allocator_internal.h
===================================================================
--- lib/sanitizer_common/sanitizer_allocator_internal.h
+++ lib/sanitizer_common/sanitizer_allocator_internal.h
@@ -57,6 +57,8 @@
 void *InternalCalloc(uptr countr, uptr size,
                      InternalAllocatorCache *cache = nullptr);
 void InternalFree(void *p, InternalAllocatorCache *cache = nullptr);
+int InternalReallocArr(void *p, uptr count, uptr size,
+                       InternalAllocatorCache *cache = nullptr);
 InternalAllocator *internal_allocator();
 
 enum InternalAllocEnum {
Index: lib/tsan/rtl/tsan_interceptors.cc
===================================================================
--- lib/tsan/rtl/tsan_interceptors.cc
+++ lib/tsan/rtl/tsan_interceptors.cc
@@ -265,7 +265,7 @@
   return ctx;
 }
 
-#if !SANITIZER_MAC
+#if !SANITIZER_MAC && !SANITIZER_NETBSD
 static unsigned g_thread_finalize_key;
 #endif
 
@@ -391,7 +391,28 @@
   void *arg;
 };
 
-static void at_exit_wrapper(void *arg) {
+static struct AtExitCtx **AtExitStack;
+static int AtExitStackSize;
+
+static void at_exit_wrapper() {
+  ThreadState *thr = cur_thread();
+  uptr pc = 0;
+
+  // Pop AtExitCtx from the top of the stack of callback functions
+  AtExitCtx *ctx = AtExitStack[AtExitStackSize - 1];
+  --AtExitStackSize;
+  if (InternalReallocArr(&AtExitStack, AtExitStackSize,
+                         sizeof(struct AtExitCtx *))) {
+    Printf("InternalReallocArr error");
+    Die();
+  }
+
+  Acquire(thr, pc, (uptr)ctx);
+  ((void(*)())ctx->f)();
+  InternalFree(ctx);
+}
+
+static void cxa_at_exit_wrapper(void *arg) {
   ThreadState *thr = cur_thread();
   uptr pc = 0;
   Acquire(thr, pc, (uptr)arg);
@@ -430,12 +451,29 @@
   // Memory allocation in __cxa_atexit will race with free during exit,
   // because we do not see synchronization around atexit callback list.
   ThreadIgnoreBegin(thr, pc);
-  int res = REAL(__cxa_atexit)(at_exit_wrapper, ctx, dso);
+  int res;
+  if (dso == 0) {
+    // NetBSD does not preserve the 2nd argument if dso is equal to 0
+    // Store ctx in a local stack-like structure
+    res = REAL(__cxa_atexit)((void (*)(void *a))at_exit_wrapper, 0, 0);
+    // Push AtExitCtx on the top of the stack of callback functions
+    if (res == 0) {
+      ++AtExitStackSize;
+      if (InternalReallocArr(&AtExitStack, AtExitStackSize,
+                             sizeof(struct AtExitCtx *))) {
+        Printf("InternalReallocArr error");
+        Die();
+      }
+      AtExitStack[AtExitStackSize - 1] = ctx;
+    }
+  } else {
+    res = REAL(__cxa_atexit)(cxa_at_exit_wrapper, ctx, dso);
+  }
   ThreadIgnoreEnd(thr, pc);
   return res;
 }
 
-#if !SANITIZER_MAC
+#if !SANITIZER_MAC && !SANITIZER_NETBSD
 static void on_exit_wrapper(int status, void *arg) {
   ThreadState *thr = cur_thread();
   uptr pc = 0;
@@ -460,6 +498,9 @@
   ThreadIgnoreEnd(thr, pc);
   return res;
 }
+#define TSAN_MAYBE_INTERCEPT_ON_EXIT TSAN_INTERCEPT(on_exit)
+#else
+#define TSAN_MAYBE_INTERCEPT_ON_EXIT
 #endif
 
 // Cleanup old bufs.
@@ -866,7 +907,7 @@
 }
 }  // namespace __tsan
 
-#if !SANITIZER_MAC
+#if !SANITIZER_MAC && !SANITIZER_NETBSD
 static void thread_finalize(void *v) {
   uptr iter = (uptr)v;
   if (iter > 1) {
@@ -896,7 +937,7 @@
     ThreadState *thr = cur_thread();
     // Thread-local state is not initialized yet.
     ScopedIgnoreInterceptors ignore;
-#if !SANITIZER_MAC
+#if !SANITIZER_MAC && !SANITIZER_NETBSD
     ThreadIgnoreBegin(thr, 0);
     if (pthread_setspecific(g_thread_finalize_key,
                             (void *)GetPthreadDestructorIterations())) {
@@ -2445,6 +2486,17 @@
 }
 #endif
 
+#if SANITIZER_NETBSD
+TSAN_INTERCEPTOR(void, _lwp_exit) {
+  SCOPED_TSAN_INTERCEPTOR(_lwp_exit);
+  REAL(_lwp_exit)();
+  DestroyThreadState();
+}
+#define TSAN_MAYBE_INTERCEPT__LWP_EXIT TSAN_INTERCEPT(_lwp_exit)
+#else
+#define TSAN_MAYBE_INTERCEPT__LWP_EXIT
+#endif
+
 namespace __tsan {
 
 static void finalize(void *arg) {
@@ -2605,7 +2657,7 @@
 #if !SANITIZER_ANDROID
   TSAN_INTERCEPT(dl_iterate_phdr);
 #endif
-  TSAN_INTERCEPT(on_exit);
+  TSAN_MAYBE_INTERCEPT_ON_EXIT;
   TSAN_INTERCEPT(__cxa_atexit);
   TSAN_INTERCEPT(_exit);
 
@@ -2613,6 +2665,8 @@
   TSAN_INTERCEPT(__tls_get_addr);
 #endif
 
+  TSAN_MAYBE_INTERCEPT__LWP_EXIT;
+
 #if !SANITIZER_MAC && !SANITIZER_ANDROID
   // Need to setup it, because interceptors check that the function is resolved.
   // But atexit is emitted directly into the module, so can't be resolved.
@@ -2624,7 +2678,7 @@
     Die();
   }
 
-#if !SANITIZER_MAC
+#if !SANITIZER_MAC && !SANITIZER_NETBSD
   if (pthread_key_create(&g_thread_finalize_key, &thread_finalize)) {
     Printf("ThreadSanitizer: failed to create thread key\n");
     Die();