From 50e916f38bfb45846159716a3831919dc8bb6c1d Mon Sep 17 00:00:00 2001
From: TFLM-bot <tflm-github-bot@google.com>
Date: Sat, 21 Dec 2024 14:02:32 +0000
Subject: [PATCH] Sync from upstream TF.

---
 tensorflow/lite/core/api/flatbuffer_conversions.h | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/tensorflow/lite/core/api/flatbuffer_conversions.h b/tensorflow/lite/core/api/flatbuffer_conversions.h
index 376b0eeb630..5bc70cc6dea 100644
--- a/tensorflow/lite/core/api/flatbuffer_conversions.h
+++ b/tensorflow/lite/core/api/flatbuffer_conversions.h
@@ -42,9 +42,8 @@ class BuiltinDataAllocator {
   // deallocation.
   template <typename T>
   T* AllocatePOD() {
-    // TODO(b/154346074): Change this to is_trivially_destructible when all
-    // platform targets support that properly.
-    static_assert(std::is_pod<T>::value, "Builtin data structure must be POD.");
+    static_assert(std::is_trivially_destructible<T>::value,
+                  "Builtin data structure must be POD.");
     void* allocated_memory = this->Allocate(sizeof(T), alignof(T));
     return new (allocated_memory) T();
   }