src: allow blobs in addition to FILE*s in embedder snapshot API

This is a shared follow-up to 06bb6b4 and a466fea now that both have been merged.
nodejs · nodejs-github-bot · Feb 6, 2023 · Feb 3, 2023 · Feb 6, 2023 · Feb 6, 2023
commit 56170f5e9031f5962bd101f77d1b62aa7e7bf096
diff --git a/src/api/embed_helpers.cc b/src/api/embed_helpers.cc
@@ -291,7 +291,8 @@ EmbedderSnapshotData::Pointer EmbedderSnapshotData::BuiltinSnapshotData() {
       SnapshotBuilder::GetEmbeddedSnapshotData(), false)};
 }
 
-EmbedderSnapshotData::Pointer EmbedderSnapshotData::FromFile(FILE* in) {
+EmbedderSnapshotData::Pointer EmbedderSnapshotData::FromBlob(
+    const std::vector<char>& in) {
   SnapshotData* snapshot_data = new SnapshotData();
   CHECK_EQ(snapshot_data->data_ownership, SnapshotData::DataOwnership::kOwned);
   EmbedderSnapshotData::Pointer result{
@@ -302,8 +303,16 @@ EmbedderSnapshotData::Pointer EmbedderSnapshotData::FromFile(FILE* in) {
   return result;
 }
 
+EmbedderSnapshotData::Pointer EmbedderSnapshotData::FromFile(FILE* in) {
+  return FromBlob(ReadFileSync(in));
+}
+
+std::vector<char> EmbedderSnapshotData::ToBlob() const {
+  return impl_->ToBlob();
+}
+
 void EmbedderSnapshotData::ToFile(FILE* out) const {
-  impl_->ToBlob(out);
+  impl_->ToFile(out);
 }
 
 EmbedderSnapshotData::EmbedderSnapshotData(const SnapshotData* impl,

diff --git a/src/env.h b/src/env.h
@@ -517,11 +517,13 @@ struct SnapshotData {
   // v8::ScriptCompiler::CachedData is not copyable.
   std::vector<builtins::CodeCacheInfo> code_cache;
 
-  void ToBlob(FILE* out) const;
+  void ToFile(FILE* out) const;
+  std::vector<char> ToBlob() const;
   // If returns false, the metadata doesn't match the current Node.js binary,
   // and the caller should not consume the snapshot data.
   bool Check() const;
-  static bool FromBlob(SnapshotData* out, FILE* in);
+  static bool FromFile(SnapshotData* out, FILE* in);
+  static bool FromBlob(SnapshotData* out, const std::vector<char>& in);
   static const SnapshotData* FromEmbedderWrapper(
       const EmbedderSnapshotData* data);
   EmbedderSnapshotData::Pointer AsEmbedderWrapper() const;

diff --git a/src/node.cc b/src/node.cc
@@ -1146,7 +1146,7 @@ ExitCode GenerateAndWriteSnapshotData(const SnapshotData** snapshot_data_ptr,
 
   FILE* fp = fopen(snapshot_blob_path.c_str(), "wb");
   if (fp != nullptr) {
-    (*snapshot_data_ptr)->ToBlob(fp);
+    (*snapshot_data_ptr)->ToFile(fp);
     fclose(fp);
   } else {
     fprintf(stderr,
@@ -1174,7 +1174,7 @@ ExitCode LoadSnapshotDataAndRun(const SnapshotData** snapshot_data_ptr,
       return exit_code;
     }
     std::unique_ptr<SnapshotData> read_data = std::make_unique<SnapshotData>();
-    bool ok = SnapshotData::FromBlob(read_data.get(), fp);
+    bool ok = SnapshotData::FromFile(read_data.get(), fp);
     fclose(fp);
     if (!ok) {
       // If we fail to read the customized snapshot, simply exit with 1.

diff --git a/src/node.h b/src/node.h
@@ -515,11 +515,13 @@ class EmbedderSnapshotData {
   // The FILE* handle can be closed immediately following this call.
   // If the snapshot is invalid, this returns an empty pointer.
   static Pointer FromFile(FILE* in);
+  static Pointer FromBlob(const std::vector<char>& in);
 
   // Write this EmbedderSnapshotData object to an output file.
   // Calling this method will not close the FILE* handle.
   // The FILE* handle can be closed immediately following this call.
   void ToFile(FILE* out) const;
+  std::vector<char> ToBlob() const;
 
   // Returns whether custom snapshots can be used. Currently, this means
   // that V8 was configured without the shared-readonly-heap feature.

diff --git a/src/node_snapshotable.cc b/src/node_snapshotable.cc
@@ -840,7 +840,7 @@ size_t SnapshotSerializer::Write(const SnapshotMetadata& data) {
 // [    ...       ]  env_info
 // [    ...       ]  code_cache
 
-void SnapshotData::ToBlob(FILE* out) const {
+std::vector<char> SnapshotData::ToBlob() const {
   SnapshotSerializer w;
   w.Debug("SnapshotData::ToBlob()\n");
 
@@ -858,9 +858,14 @@ void SnapshotData::ToBlob(FILE* out) const {
   written_total += w.Write<EnvSerializeInfo>(env_info);
   w.Debug("Write code_cache\n");
   written_total += w.WriteVector<builtins::CodeCacheInfo>(code_cache);
-  size_t num_written = fwrite(w.sink.data(), w.sink.size(), 1, out);
-  CHECK_EQ(num_written, 1);
   w.Debug("SnapshotData::ToBlob() Wrote %d bytes\n", written_total);
+  return w.sink;
+}
+
+void SnapshotData::ToFile(FILE* out) const {
+  const std::vector<char> sink = ToBlob();
+  size_t num_written = fwrite(sink.data(), sink.size(), 1, out);
+  CHECK_EQ(num_written, 1);
 }
 
 const SnapshotData* SnapshotData::FromEmbedderWrapper(
@@ -872,20 +877,12 @@ EmbedderSnapshotData::Pointer SnapshotData::AsEmbedderWrapper() const {
   return EmbedderSnapshotData::Pointer{new EmbedderSnapshotData(this, false)};
 }
 
-bool SnapshotData::FromBlob(SnapshotData* out, FILE* in) {
-  CHECK_EQ(ftell(in), 0);
-  int err = fseek(in, 0, SEEK_END);
-  CHECK_EQ(err, 0);
-  size_t size = ftell(in);
-  CHECK_NE(size, static_cast<size_t>(-1L));
-  err = fseek(in, 0, SEEK_SET);
-  CHECK_EQ(err, 0);
-
-  std::vector<char> sink(size);
-  size_t num_read = fread(sink.data(), size, 1, in);
-  CHECK_EQ(num_read, 1);
+bool SnapshotData::FromFile(SnapshotData* out, FILE* in) {
+  return FromBlob(out, ReadFileSync(in));
+}
 
-  SnapshotDeserializer r(sink);
+bool SnapshotData::FromBlob(SnapshotData* out, const std::vector<char>& in) {
+  SnapshotDeserializer r(in);
   r.Debug("SnapshotData::FromBlob()\n");
 
   DCHECK_EQ(out->data_ownership, SnapshotData::DataOwnership::kOwned);

diff --git a/src/util.cc b/src/util.cc
@@ -267,6 +267,21 @@ int ReadFileSync(std::string* result, const char* path) {
   return 0;
 }
 
+std::vector<char> ReadFileSync(FILE* fp) {
+  CHECK_EQ(ftell(fp), 0);
+  int err = fseek(fp, 0, SEEK_END);
+  CHECK_EQ(err, 0);
+  size_t size = ftell(fp);
+  CHECK_NE(size, static_cast<size_t>(-1L));
+  err = fseek(fp, 0, SEEK_SET);
+  CHECK_EQ(err, 0);
+
+  std::vector<char> contents(size);
+  size_t num_read = fread(contents.data(), size, 1, fp);
+  CHECK_EQ(num_read, 1);
+  return contents;
+}
+
 void DiagnosticFilename::LocalTime(TIME_TYPE* tm_struct) {
 #ifdef _WIN32
   GetLocalTime(tm_struct);

diff --git a/src/util.h b/src/util.h
@@ -867,6 +867,8 @@ std::unique_ptr<T> static_unique_pointer_cast(std::unique_ptr<U>&& ptr) {
 // Returns a non-zero code if it fails to open or read the file,
 // aborts if it fails to close the file.
 int ReadFileSync(std::string* result, const char* path);
+// Reads all contents of a FILE*, aborts if it fails.
+std::vector<char> ReadFileSync(FILE* fp);
 
 v8::Local<v8::FunctionTemplate> NewFunctionTemplate(
     v8::Isolate* isolate,

diff --git a/test/embedding/embedtest.cc b/test/embedding/embedtest.cc
@@ -65,11 +65,28 @@ int RunNodeInstance(MultiIsolatePlatform* platform,
       std::find(args.begin(), args.end(), "--embedder-snapshot-create");
   auto snapshot_arg_it =
       std::find(args.begin(), args.end(), "--embedder-snapshot-blob");
+  auto snapshot_as_file_it =
+      std::find(args.begin(), args.end(), "--embedder-snapshot-as-file");
   if (snapshot_arg_it < args.end() - 1 &&
       snapshot_build_mode_it == args.end()) {
-    FILE* fp = fopen((snapshot_arg_it + 1)->c_str(), "r");
+    const char* filename = (snapshot_arg_it + 1)->c_str();
+    FILE* fp = fopen(filename, "r");
     assert(fp != nullptr);
-    snapshot = node::EmbedderSnapshotData::FromFile(fp);
+    if (snapshot_as_file_it != args.end()) {
+      snapshot = node::EmbedderSnapshotData::FromFile(fp);
+    } else {
+      uv_fs_t req;
+      int statret = uv_fs_stat(nullptr, &req, filename, nullptr);
+      assert(statret == 0);
+      size_t filesize = req.statbuf.st_size;
+      uv_fs_req_cleanup(&req);
+
+      std::vector<char> vec(filesize);
+      size_t read = fread(vec.data(), filesize, 1, fp);
+      assert(read == 1);
+      snapshot = node::EmbedderSnapshotData::FromBlob(vec);
+    }
+    assert(snapshot);
     fclose(fp);
   }
 
@@ -125,7 +142,13 @@ int RunNodeInstance(MultiIsolatePlatform* platform,
 
     FILE* fp = fopen((snapshot_arg_it + 1)->c_str(), "w");
     assert(fp != nullptr);
-    snapshot->ToFile(fp);
+    if (snapshot_as_file_it != args.end()) {
+      snapshot->ToFile(fp);
+    } else {
+      const std::vector<char> vec = snapshot->ToBlob();
+      size_t written = fwrite(vec.data(), vec.size(), 1, fp);
+      assert(written == 1);
+    }
     fclose(fp);
   }
 

diff --git a/test/embedding/test-embedding.js b/test/embedding/test-embedding.js
@@ -55,15 +55,18 @@ function getReadFileCodeForPath(path) {
 }
 
 // Basic snapshot support
-{
+for (const extraSnapshotArgs of [[], ['--embedder-snapshot-as-file']]) {
   // readSync + eval since snapshots don't support userland require() (yet)
   const snapshotFixture = fixtures.path('snapshot', 'echo-args.js');
   const blobPath = path.join(tmpdir.path, 'embedder-snapshot.blob');
   const buildSnapshotArgs = [
     `eval(${getReadFileCodeForPath(snapshotFixture)})`, 'arg1', 'arg2',
     '--embedder-snapshot-blob', blobPath, '--embedder-snapshot-create',
+    ...extraSnapshotArgs,
+  ];
+  const runEmbeddedArgs = [
+    '--embedder-snapshot-blob', blobPath, ...extraSnapshotArgs, 'arg3', 'arg4',
   ];
-  const runEmbeddedArgs = ['--embedder-snapshot-blob', blobPath, 'arg3', 'arg4'];
 
   fs.rmSync(blobPath, { force: true });
   assert.strictEqual(child_process.spawnSync(binary, [