From e98dd95446e009ace1722498effbf32250d623e4 Mon Sep 17 00:00:00 2001
From: Yiming Zhou <yimingzhou@meta.com>
Date: Sat, 21 Jun 2025 01:32:02 +0000
Subject: [PATCH] [nativert] Move SerialGraphExecutor to PyTorch core (#156459)

Summary: `SerialGraphExecutor` inherits from `GraphExecutorBase` and executes all nodes in the graph in a serial manner

Test Plan:
CI

Rollback Plan:

Differential Revision: D76917966

Pull Request resolved: https://github.com/pytorch/pytorch/pull/156459
Approved by: https://github.com/zhxchen17, https://github.com/jingsh
---
 build_variables.bzl                           |  1 +
 .../nativert/executor/SerialGraphExecutor.cpp | 33 +++++++++++++++++++
 torch/nativert/executor/SerialGraphExecutor.h | 23 +++++++++++++
 3 files changed, 57 insertions(+)
 create mode 100644 torch/nativert/executor/SerialGraphExecutor.cpp
 create mode 100644 torch/nativert/executor/SerialGraphExecutor.h

diff --git a/build_variables.bzl b/build_variables.bzl
index 0c7442d36262..77e6ab46837b 100644
--- a/build_variables.bzl
+++ b/build_variables.bzl
@@ -603,6 +603,7 @@ libtorch_nativert_sources = [
     "torch/nativert/executor/GraphExecutorBase.cpp",
     "torch/nativert/executor/OpKernel.cpp",
     "torch/nativert/executor/PlacementUtils.cpp",
+    "torch/nativert/executor/SerialGraphExecutor.cpp",
     "torch/nativert/executor/Weights.cpp",
     "torch/nativert/executor/memory/FunctionSchema.cpp",
     "torch/nativert/common/FileUtil.cpp",
diff --git a/torch/nativert/executor/SerialGraphExecutor.cpp b/torch/nativert/executor/SerialGraphExecutor.cpp
new file mode 100644
index 000000000000..f1ef0491eda1
--- /dev/null
+++ b/torch/nativert/executor/SerialGraphExecutor.cpp
@@ -0,0 +1,33 @@
+#include <torch/nativert/executor/ExecutionPlanner.h>
+#include <torch/nativert/executor/ExecutorConfig.h>
+#include <torch/nativert/executor/SerialGraphExecutor.h>
+
+namespace torch::nativert {
+
+std::vector<c10::IValue> SerialGraphExecutor::execute(
+    ExecutionFrame& executionFrame,
+    std::vector<c10::IValue> inputs) {
+  fillUserInputs(executionFrame, std::move(inputs));
+
+  return executeWithPrefilledFrame(executionFrame);
+}
+
+std::vector<c10::IValue> SerialGraphExecutor::executeWithPrefilledFrame(
+    ExecutionFrame& executionFrame) {
+  // Execute kernels for all nodes except prim.Input and prim.Output
+  for (NodeIndex nodeIdx = 1; nodeIdx < nodeKernels_.size() - 1; ++nodeIdx) {
+    nodeKernels_[nodeIdx]->compute(executionFrame);
+
+    // don't free intermediate values when static memory planning is enabled
+    if (!executorConfig_.enableStaticMemoryPlanning) {
+      // Free the intermediate values that are no used anymore
+      for (const auto& valueKey : execPlan_->valuesToFree[nodeIdx]) {
+        executionFrame.releaseValue(valueKey);
+      }
+    }
+  }
+
+  return executionFrame.tryMoveUserOutputs();
+}
+
+} // namespace torch::nativert
diff --git a/torch/nativert/executor/SerialGraphExecutor.h b/torch/nativert/executor/SerialGraphExecutor.h
new file mode 100644
index 000000000000..cae3313e61e8
--- /dev/null
+++ b/torch/nativert/executor/SerialGraphExecutor.h
@@ -0,0 +1,23 @@
+#pragma once
+
+#include <torch/nativert/executor/GraphExecutorBase.h>
+
+namespace torch::nativert {
+
+class SerialGraphExecutor : public GraphExecutorBase {
+ public:
+  SerialGraphExecutor(
+      const Graph& graph,
+      std::vector<std::unique_ptr<OpKernel>> nodeKernels,
+      const ExecutorConfig& executorConfig)
+      : GraphExecutorBase(graph, std::move(nodeKernels), executorConfig) {}
+
+  std::vector<c10::IValue> execute(
+      ExecutionFrame& frame,
+      std::vector<c10::IValue> inputs) override;
+
+  std::vector<c10::IValue> executeWithPrefilledFrame(
+      ExecutionFrame& frame) override;
+};
+
+} // namespace torch::nativert