[ExecuTorch][WebGPU] Add per-pass dispatch ordering + scratch buffer tests

Pull Request resolved: #20080

Native unit tests for two runtime enablers: per-pass compute-dispatch ordering (D107543258) and graph-owned scratch buffers (D107543259). `test/native/test_dispatch_order.cpp` exercises multi-dispatch read-after-write ordering through a single `execute()` using dependency chains -- a single-input `add` self-chain and a heterogeneous `rms_norm` -> `add` chain, both lowered via `VulkanPartitioner` -- comparing GPU output to a torch-computed golden per element. `test/native/test_scratch_buffer.cpp` is a white-box test of `WebGPUGraph::create_scratch_buffer` (no black-box consumer exists below the SDPA op): allocation + zero-size guard, copy round-trip, a compute Storage round-trip (its actual use), and a create/destroy lifecycle stress. Authored with assistance from Claude.
ghstack-source-id: 390672000
@exported-using-ghexport

Differential Revision: [D107576199](https://our.internmc.facebook.com/intern/diff/D107576199/)

Author: JCNTH

backends/webgpu/CMakeLists.txt

@@ -118,4 +118,10 @@ endfunction()
 if(EXECUTORCH_BUILD_WEBGPU_TEST)
   add_webgpu_native_test(webgpu_native_test test/test_webgpu_native.cpp)
   add_webgpu_native_test(webgpu_rms_norm_test test/native/test_rms_norm.cpp)
+  add_webgpu_native_test(
+    webgpu_dispatch_order_test test/native/test_dispatch_order.cpp
+  )
+  add_webgpu_native_test(
+    webgpu_scratch_buffer_test test/native/test_scratch_buffer.cpp
+  )
 endif()

backends/webgpu/test/native/test_dispatch_order.cpp

@@ -0,0 +1,167 @@
+/*
+ * Copyright (c) Meta Platforms, Inc. and affiliates.
+ * All rights reserved.
+ *
+ * This source code is licensed under the BSD-style license found in the
+ * LICENSE file in the root directory of this source tree.
+ */
+
+#include <executorch/backends/webgpu/runtime/WebGPUDevice.h>
+#include <executorch/extension/module/module.h>
+#include <executorch/extension/tensor/tensor.h>
+
+#include <algorithm>
+#include <cmath>
+#include <cstdio>
+#include <cstdlib>
+#include <fstream>
+#include <string>
+#include <vector>
+
+using namespace executorch::backends::webgpu;
+using namespace executorch::extension;
+using namespace executorch::runtime;
+
+namespace {
+
+struct Case {
+  const char* name;
+  std::vector<int32_t> sizes;
+};
+
+// Mirrors _CASES in test_dispatch_order.py (add-chain or rms_norm+add chain).
+const std::vector<Case> kCases = {
+    {"single", {16, 16}},
+    {"chain3", {64, 64}},
+    {"chain5_tiny", {1, 1}},
+    {"chain5_wide", {7, 896}},
+    {"chain8", {256, 256}},
+    {"deep32", {128, 128}},
+    {"large_chain", {1024, 1024}},
+    {"het_small", {1, 1, 7, 896}},
+    {"het_deep", {1, 1, 5, 256}},
+};
+
+std::vector<float> read_f32_bin(const std::string& path) {
+  std::ifstream f(path, std::ios::binary | std::ios::ate);
+  if (!f) {
+    return {};
+  }
+  const auto file_size = static_cast<size_t>(f.tellg());
+  if (file_size % sizeof(float) != 0) {
+    return {}; // truncated/corrupt golden; caller treats empty as failure
+  }
+  f.seekg(0);
+  std::vector<float> data(file_size / sizeof(float));
+  f.read(
+      reinterpret_cast<char*>(data.data()),
+      static_cast<std::streamsize>(file_size));
+  return data;
+}
+
+bool run_case(const std::string& dir, const Case& tc) {
+  printf("\n--- dispatch_order[%s] ---\n", tc.name);
+  const std::string base = dir + "/" + tc.name;
+  std::vector<float> input = read_f32_bin(base + ".input.bin");
+  std::vector<float> golden = read_f32_bin(base + ".golden.bin");
+  if (input.empty() || golden.empty()) {
+    printf("FAIL: could not read input/golden for %s\n", tc.name);
+    return false;
+  }
+
+  Module module(base + ".pte");
+  if (module.load_forward() != Error::Ok) {
+    printf("FAIL: could not load %s.pte\n", tc.name);
+    return false;
+  }
+
+  size_t expected = 1;
+  for (int32_t d : tc.sizes) {
+    expected *= static_cast<size_t>(d);
+  }
+  if (input.size() != expected) {
+    printf(
+        "FAIL: input numel %zu != expected %zu for %s\n",
+        input.size(),
+        expected,
+        tc.name);
+    return false;
+  }
+  auto x = make_tensor_ptr(tc.sizes, std::vector<float>(input));
+  auto result = module.forward({EValue(x)});
+  if (!result.ok()) {
+    printf("FAIL: forward failed (error %d)\n", (int)result.error());
+    return false;
+  }
+  const auto& outputs = result.get();
+  if (outputs.empty() || !outputs[0].isTensor()) {
+    printf("FAIL: no tensor output\n");
+    return false;
+  }
+  const auto& out_tensor = outputs[0].toTensor();
+  if (static_cast<size_t>(out_tensor.numel()) != golden.size()) {
+    printf(
+        "FAIL: output numel %zu != golden %zu\n",
+        (size_t)out_tensor.numel(),
+        golden.size());
+    return false;
+  }
+  const float* out_data = out_tensor.const_data_ptr<float>();
+
+  float max_abs_err = 0.0f;
+  float max_rel_err = 0.0f;
+  for (size_t i = 0; i < golden.size(); i++) {
+    const float abs_err = std::abs(out_data[i] - golden[i]);
+    max_abs_err = std::max(max_abs_err, abs_err);
+    const float denom = std::max(std::abs(golden[i]), 1e-6f);
+    max_rel_err = std::max(max_rel_err, abs_err / denom);
+  }
+  printf(
+      "Max abs error: %e   Max rel error: %e (%zu elements)\n",
+      max_abs_err,
+      max_rel_err,
+      golden.size());
+  // Lenient gate: pass iff abs<=tol OR rel<=tol (near-zero goldens).
+  if (max_abs_err > 1e-3f && max_rel_err > 1e-3f) {
+    printf("FAIL: dispatch_order[%s] exceeds tolerance 1e-3\n", tc.name);
+    return false;
+  }
+  printf("PASS: dispatch_order[%s]\n", tc.name);
+  return true;
+}
+
+} // namespace
+
+int main(int argc, char** argv) {
+  std::string dir = "/tmp/dispatch_order";
+  if (argc > 1) {
+    dir = argv[1];
+  }
+  if (const char* env = std::getenv("WEBGPU_DISPATCH_ORDER_DIR")) {
+    dir = env;
+  }
+
+  WebGPUContext ctx;
+  try {
+    ctx = create_webgpu_context();
+  } catch (const std::exception& e) {
+    printf("SKIP: %s\n", e.what());
+    return 0;
+  }
+  set_default_webgpu_context(&ctx);
+  printf("WebGPU device acquired (native); case dir: %s\n", dir.c_str());
+
+  bool ok = true;
+  for (const auto& tc : kCases) {
+    ok = run_case(dir, tc) && ok;
+  }
+
+  set_default_webgpu_context(nullptr);
+  destroy_webgpu_context(ctx);
+
+  if (!ok) {
+    return 1;
+  }
+  printf("\nAll dispatch_order tests passed\n");
+  return 0;
+}