From 4becc8ff62500c4cd3836c4581f2e2f7f1309776 Mon Sep 17 00:00:00 2001 From: Joyee Cheung Date: Sun, 12 Feb 2023 00:44:48 +0100 Subject: [PATCH] src: use an array for faster binding data lookup Locally the hashing of the binding names sometimes has significant presence in the profile of bindings, because there can be collisions, which makes the cost of adding a new binding data non-trivial, but it's wasteful to spend time on hashing them or dealing with collisions at all, since we can just use the EmbedderObjectType enum as the key, as the string names are not actually used beyond debugging purposes and can be easily matched with a macro. And since we can just use the enum as the key, we do not even need the map and can just use an array with the enum as indices for the lookup. --- node.gyp | 1 + src/README.md | 30 ++++++++++------- src/base_object.h | 1 + src/base_object_types.h | 69 ++++++++++++++++++++++++++++++++++++++++ src/node_blob.h | 4 +-- src/node_file.h | 4 +-- src/node_http2_state.h | 2 +- src/node_http_parser.cc | 2 +- src/node_process.h | 4 +-- src/node_realm-inl.h | 14 +++++--- src/node_realm.cc | 5 +-- src/node_realm.h | 6 ++-- src/node_snapshotable.cc | 8 ++--- src/node_snapshotable.h | 15 +-------- src/node_util.h | 4 +-- src/node_v8.h | 4 +-- 16 files changed, 117 insertions(+), 56 deletions(-) create mode 100644 src/base_object_types.h diff --git a/node.gyp b/node.gyp index f89091774721b7..2305dbbb202751 100644 --- a/node.gyp +++ b/node.gyp @@ -573,6 +573,7 @@ 'src/async_wrap-inl.h', 'src/base_object.h', 'src/base_object-inl.h', + 'src/base_object_types.h', 'src/base64.h', 'src/base64-inl.h', 'src/callback_queue.h', diff --git a/src/README.md b/src/README.md index 58718935820ada..bcd6b6e8ccc4d9 100644 --- a/src/README.md +++ b/src/README.md @@ -486,18 +486,32 @@ that state is through the use of `Realm::AddBindingData`, which gives binding functions access to an object for storing such state. That object is always a [`BaseObject`][]. -Its class needs to have a static `type_name` field based on a -constant string, in order to disambiguate it from other classes of this type, -and which could e.g. match the binding's name (in the example above, that would -be `cares_wrap`). +In the binding, call `SET_BINDING_ID()` with an identifier for the binding +type. For example, for `http_parser::BindingData`, the identifier can be +`http_parser_binding_data`. + +If the binding should be supported in a snapshot, the id and the +fully-specified class name should be added to the `SERIALIZABLE_BINDING_TYPES` +list in `base_object_types.h`, and the class should implement the serialization +and deserialization methods. See the comments of `SnapshotableObject` on how to +implement them. Otherwise, add the id and the class name to the +`UNSERIALIZABLE_BINDING_TYPES` list instead. ```cpp +// In base_object_types.h, add the binding to either +// UNSERIALIZABLE_BINDING_TYPES or SERIALIZABLE_BINDING_TYPES. +// The second parameter is a descriptive name of the class, which is +// usually the fully-specified class name. + +#define UNSERIALIZABLE_BINDING_TYPES(V) \ + V(http_parser_binding_data, http_parser::BindingData) + // In the HTTP parser source code file: class BindingData : public BaseObject { public: BindingData(Environment* env, Local obj) : BaseObject(env, obj) {} - static constexpr FastStringKey type_name { "http_parser" }; + SET_BINDING_ID(http_parser_binding_data) std::vector parser_buffer; bool parser_buffer_in_use = false; @@ -527,12 +541,6 @@ void InitializeHttpParser(Local target, } ``` -If the binding is loaded during bootstrap, add it to the -`SERIALIZABLE_OBJECT_TYPES` list in `src/node_snapshotable.h` and -inherit from the `SnapshotableObject` class instead. See the comments -of `SnapshotableObject` on how to implement its serialization and -deserialization. - ### Exception handling diff --git a/src/base_object.h b/src/base_object.h index 779573362268a6..719f1d38ddf739 100644 --- a/src/base_object.h +++ b/src/base_object.h @@ -25,6 +25,7 @@ #if defined(NODE_WANT_INTERNALS) && NODE_WANT_INTERNALS #include // std::remove_reference +#include "base_object_types.h" #include "memory_tracker.h" #include "v8.h" diff --git a/src/base_object_types.h b/src/base_object_types.h new file mode 100644 index 00000000000000..f4c70a89177975 --- /dev/null +++ b/src/base_object_types.h @@ -0,0 +1,69 @@ +#ifndef SRC_BASE_OBJECT_TYPES_H_ +#define SRC_BASE_OBJECT_TYPES_H_ + +#include + +#if defined(NODE_WANT_INTERNALS) && NODE_WANT_INTERNALS + +namespace node { +// List of internalBinding() data wrappers. The first argument should match +// what the class passes to SET_BINDING_ID(), the second argument should match +// the C++ class name. +#define SERIALIZABLE_BINDING_TYPES(V) \ + V(fs_binding_data, fs::BindingData) \ + V(v8_binding_data, v8_utils::BindingData) \ + V(blob_binding_data, BlobBindingData) \ + V(process_binding_data, process::BindingData) + +#define UNSERIALIZABLE_BINDING_TYPES(V) \ + V(http2_binding_data, http2::BindingData) \ + V(http_parser_binding_data, http_parser::BindingData) + +// List of (non-binding) BaseObjects that are serializable in the snapshot. +// The first argument should match what the type passes to +// SET_OBJECT_ID(), the second argument should match the C++ class +// name. +#define SERIALIZABLE_NON_BINDING_TYPES(V) \ + V(util_weak_reference, util::WeakReference) + +// Helper list of all binding data wrapper types. +#define BINDING_TYPES(V) \ + SERIALIZABLE_BINDING_TYPES(V) \ + UNSERIALIZABLE_BINDING_TYPES(V) + +// Helper list of all BaseObjects that implement snapshot support. +#define SERIALIZABLE_OBJECT_TYPES(V) \ + SERIALIZABLE_BINDING_TYPES(V) \ + SERIALIZABLE_NON_BINDING_TYPES(V) + +#define V(TypeId, NativeType) k_##TypeId, +enum class BindingDataType : uint8_t { BINDING_TYPES(V) kBindingDataTypeCount }; +// Make sure that we put the bindings first so that we can also use the enums +// for the bindings as index to the binding data store. +enum class EmbedderObjectType : uint8_t { + BINDING_TYPES(V) SERIALIZABLE_NON_BINDING_TYPES(V) + // We do not need to know about all the unserializable non-binding types for + // now so we do not list them. + kEmbedderObjectTypeCount +}; +#undef V + +// For now, BaseObjects only need to call this when they implement snapshot +// support. +#define SET_OBJECT_ID(TypeId) \ + static constexpr EmbedderObjectType type_int = EmbedderObjectType::k_##TypeId; + +// Binding data should call this so that they can be looked up from the binding +// data store. +#define SET_BINDING_ID(TypeId) \ + static constexpr BindingDataType binding_type_int = \ + BindingDataType::k_##TypeId; \ + SET_OBJECT_ID(TypeId) \ + static_assert(static_cast(type_int) == \ + static_cast(binding_type_int)); + +} // namespace node + +#endif // defined(NODE_WANT_INTERNALS) && NODE_WANT_INTERNALS + +#endif // SRC_BASE_OBJECT_TYPES_H_ diff --git a/src/node_blob.h b/src/node_blob.h index f6d5ad89f69792..205bb14568af8f 100644 --- a/src/node_blob.h +++ b/src/node_blob.h @@ -117,9 +117,7 @@ class BlobBindingData : public SnapshotableObject { SERIALIZABLE_OBJECT_METHODS() - static constexpr FastStringKey type_name{"node::BlobBindingData"}; - static constexpr EmbedderObjectType type_int = - EmbedderObjectType::k_blob_binding_data; + SET_BINDING_ID(blob_binding_data) void MemoryInfo(MemoryTracker* tracker) const override; SET_SELF_SIZE(BlobBindingData) diff --git a/src/node_file.h b/src/node_file.h index e44b58c94ff74c..2325b6b26f756e 100644 --- a/src/node_file.h +++ b/src/node_file.h @@ -70,9 +70,7 @@ class BindingData : public SnapshotableObject { using InternalFieldInfo = InternalFieldInfoBase; SERIALIZABLE_OBJECT_METHODS() - static constexpr FastStringKey type_name{"node::fs::BindingData"}; - static constexpr EmbedderObjectType type_int = - EmbedderObjectType::k_fs_binding_data; + SET_BINDING_ID(fs_binding_data) void MemoryInfo(MemoryTracker* tracker) const override; SET_SELF_SIZE(BindingData) diff --git a/src/node_http2_state.h b/src/node_http2_state.h index 75a98bf476b2f7..f9ac6b40c3410a 100644 --- a/src/node_http2_state.h +++ b/src/node_http2_state.h @@ -125,7 +125,7 @@ class Http2State : public BaseObject { SET_SELF_SIZE(Http2State) SET_MEMORY_INFO_NAME(Http2State) - static constexpr FastStringKey type_name { "http2" }; + SET_BINDING_ID(http2_binding_data) private: struct http2_state_internal { diff --git a/src/node_http_parser.cc b/src/node_http_parser.cc index a502fa0ccd9018..7c7f6dcd974da3 100644 --- a/src/node_http_parser.cc +++ b/src/node_http_parser.cc @@ -95,7 +95,7 @@ class BindingData : public BaseObject { public: BindingData(Realm* realm, Local obj) : BaseObject(realm, obj) {} - static constexpr FastStringKey type_name { "http_parser" }; + SET_BINDING_ID(http_parser_binding_data) std::vector parser_buffer; bool parser_buffer_in_use = false; diff --git a/src/node_process.h b/src/node_process.h index 30f655dfc71f23..5af062e63ea49b 100644 --- a/src/node_process.h +++ b/src/node_process.h @@ -54,9 +54,7 @@ class BindingData : public SnapshotableObject { using InternalFieldInfo = InternalFieldInfoBase; SERIALIZABLE_OBJECT_METHODS() - static constexpr FastStringKey type_name{"node::process::BindingData"}; - static constexpr EmbedderObjectType type_int = - EmbedderObjectType::k_process_binding_data; + SET_BINDING_ID(process_binding_data) BindingData(Realm* realm, v8::Local object); diff --git a/src/node_realm-inl.h b/src/node_realm-inl.h index 01aca9382eebe1..2fc3eef97729d1 100644 --- a/src/node_realm-inl.h +++ b/src/node_realm-inl.h @@ -66,9 +66,11 @@ inline T* Realm::GetBindingData(v8::Local context) { static_cast(context->GetAlignedPointerFromEmbedderData( ContextEmbedderIndex::kBindingDataStoreIndex)); DCHECK_NOT_NULL(map); - auto it = map->find(T::type_name); - if (UNLIKELY(it == map->end())) return nullptr; - T* result = static_cast(it->second.get()); + constexpr size_t binding_index = static_cast(T::binding_type_int); + static_assert(binding_index < std::tuple_size_v); + auto ptr = (*map)[binding_index]; + if (UNLIKELY(!ptr)) return nullptr; + T* result = static_cast(ptr.get()); DCHECK_NOT_NULL(result); DCHECK_EQ(result->realm(), GetCurrent(context)); return result; @@ -84,8 +86,10 @@ inline T* Realm::AddBindingData(v8::Local context, static_cast(context->GetAlignedPointerFromEmbedderData( ContextEmbedderIndex::kBindingDataStoreIndex)); DCHECK_NOT_NULL(map); - auto result = map->emplace(T::type_name, item); - CHECK(result.second); + constexpr size_t binding_index = static_cast(T::binding_type_int); + static_assert(binding_index < std::tuple_size_v); + CHECK(!(*map)[binding_index]); // Should not insert the binding twice. + (*map)[binding_index] = item; DCHECK_EQ(GetBindingData(context), item.get()); return item.get(); } diff --git a/src/node_realm.cc b/src/node_realm.cc index 9452e336ea789f..187323397bc733 100644 --- a/src/node_realm.cc +++ b/src/node_realm.cc @@ -302,8 +302,9 @@ void Realm::DoneBootstrapping() { void Realm::RunCleanup() { TRACE_EVENT0(TRACING_CATEGORY_NODE1(realm), "RunCleanup"); - binding_data_store_.clear(); - + for (size_t i = 0; i < binding_data_store_.size(); ++i) { + binding_data_store_[i].reset(); + } cleanup_queue_.Drain(); } diff --git a/src/node_realm.h b/src/node_realm.h index cf37fe31bdd6e4..04129eec47d551 100644 --- a/src/node_realm.h +++ b/src/node_realm.h @@ -21,9 +21,9 @@ struct RealmSerializeInfo { friend std::ostream& operator<<(std::ostream& o, const RealmSerializeInfo& i); }; -using BindingDataStore = std::unordered_map, - FastStringKey::Hash>; +using BindingDataStore = std::array, + static_cast( + BindingDataType::kBindingDataTypeCount)>; /** * node::Realm is a container for a set of JavaScript objects and functions diff --git a/src/node_snapshotable.cc b/src/node_snapshotable.cc index e1ceb78adce0ca..d9d4f162e9cfd2 100644 --- a/src/node_snapshotable.cc +++ b/src/node_snapshotable.cc @@ -1289,11 +1289,11 @@ SnapshotableObject::SnapshotableObject(Realm* realm, EmbedderObjectType type) : BaseObject(realm, wrap), type_(type) {} -std::string_view SnapshotableObject::GetTypeName() const { +std::string SnapshotableObject::GetTypeName() const { switch (type_) { #define V(PropertyName, NativeTypeName) \ case EmbedderObjectType::k_##PropertyName: { \ - return NativeTypeName::type_name.as_string_view(); \ + return #NativeTypeName; \ } SERIALIZABLE_OBJECT_TYPES(V) #undef V @@ -1334,7 +1334,7 @@ void DeserializeNodeInternalFields(Local holder, per_process::Debug(DebugCategory::MKSNAPSHOT, \ "Object %p is %s\n", \ (*holder), \ - NativeTypeName::type_name.as_string_view()); \ + #NativeTypeName); \ env_ptr->EnqueueDeserializeRequest( \ NativeTypeName::Deserialize, \ holder, \ @@ -1421,7 +1421,7 @@ void SerializeSnapshotableObjects(Realm* realm, } SnapshotableObject* ptr = static_cast(obj); - std::string type_name{ptr->GetTypeName()}; + std::string type_name = ptr->GetTypeName(); per_process::Debug(DebugCategory::MKSNAPSHOT, "Serialize snapshotable object %i (%p), " "object=%p, type=%s\n", diff --git a/src/node_snapshotable.h b/src/node_snapshotable.h index 3f4d0780131e20..28d9dd8c0aee14 100644 --- a/src/node_snapshotable.h +++ b/src/node_snapshotable.h @@ -22,19 +22,6 @@ struct PropInfo { SnapshotIndex index; // In the snapshot }; -#define SERIALIZABLE_OBJECT_TYPES(V) \ - V(fs_binding_data, fs::BindingData) \ - V(v8_binding_data, v8_utils::BindingData) \ - V(blob_binding_data, BlobBindingData) \ - V(process_binding_data, process::BindingData) \ - V(util_weak_reference, util::WeakReference) - -enum class EmbedderObjectType : uint8_t { -#define V(PropertyName, NativeType) k_##PropertyName, - SERIALIZABLE_OBJECT_TYPES(V) -#undef V -}; - typedef size_t SnapshotIndex; // When serializing an embedder object, we'll serialize the native states @@ -101,7 +88,7 @@ class SnapshotableObject : public BaseObject { SnapshotableObject(Realm* realm, v8::Local wrap, EmbedderObjectType type); - std::string_view GetTypeName() const; + std::string GetTypeName() const; // If returns false, the object will not be serialized. virtual bool PrepareForSerialization(v8::Local context, diff --git a/src/node_util.h b/src/node_util.h index 7192e080f2b08e..fa0faa618a61bc 100644 --- a/src/node_util.h +++ b/src/node_util.h @@ -14,9 +14,7 @@ class WeakReference : public SnapshotableObject { public: SERIALIZABLE_OBJECT_METHODS() - static constexpr FastStringKey type_name{"node::util::WeakReference"}; - static constexpr EmbedderObjectType type_int = - EmbedderObjectType::k_util_weak_reference; + SET_OBJECT_ID(util_weak_reference) WeakReference(Realm* realm, v8::Local object, diff --git a/src/node_v8.h b/src/node_v8.h index 2d95002bcfc473..002f506d20833e 100644 --- a/src/node_v8.h +++ b/src/node_v8.h @@ -23,9 +23,7 @@ class BindingData : public SnapshotableObject { using InternalFieldInfo = InternalFieldInfoBase; SERIALIZABLE_OBJECT_METHODS() - static constexpr FastStringKey type_name{"node::v8::BindingData"}; - static constexpr EmbedderObjectType type_int = - EmbedderObjectType::k_v8_binding_data; + SET_BINDING_ID(v8_binding_data) AliasedFloat64Array heap_statistics_buffer; AliasedFloat64Array heap_space_statistics_buffer;