3c74dc9f4d
This patch adds two macros to declare per-type allocators: - JS_DECLARE_ALLOCATOR(TypeName) - JS_DEFINE_ALLOCATOR(TypeName) When used, they add a type-specific CellAllocator that the Heap will delegate allocation requests to. The result of this is that GC objects of the same type always end up within the same HeapBlock, drastically reducing the ability to perform type confusion attacks. It also improves HeapBlock utilization, since each block now has cells sized exactly to the type used within that block. (Previously we only had a handful of block sizes available, and most GC allocations ended up with a large amount of slack in their tails.) There is a small performance hit from this, but I'm sure we can make up for it elsewhere. Note that the old size-based allocators still exist, and we fall back to them for any type that doesn't have its own CellAllocator.
67 lines
2.3 KiB
C++
67 lines
2.3 KiB
C++
/*
|
|
* Copyright (c) 2022, Linus Groh <linusg@serenityos.org>
|
|
* Copyright (c) 2023, Luke Wilde <lukew@serenityos.org>
|
|
*
|
|
* SPDX-License-Identifier: BSD-2-Clause
|
|
*/
|
|
|
|
#pragma once
|
|
|
|
#include <AK/Variant.h>
|
|
#include <LibJS/Bytecode/Interpreter.h>
|
|
#include <LibJS/Runtime/ExecutionContext.h>
|
|
#include <LibJS/Runtime/Object.h>
|
|
|
|
namespace JS {
|
|
|
|
// 27.6.2 Properties of AsyncGenerator Instances, https://tc39.es/ecma262/#sec-properties-of-asyncgenerator-intances
|
|
class AsyncGenerator final : public Object {
|
|
JS_OBJECT(AsyncGenerator, Object);
|
|
JS_DECLARE_ALLOCATOR(AsyncGenerator);
|
|
|
|
public:
|
|
enum class State {
|
|
SuspendedStart,
|
|
SuspendedYield,
|
|
Executing,
|
|
AwaitingReturn,
|
|
Completed,
|
|
};
|
|
|
|
static ThrowCompletionOr<NonnullGCPtr<AsyncGenerator>> create(Realm&, Value, ECMAScriptFunctionObject*, ExecutionContext, Bytecode::CallFrame);
|
|
|
|
virtual ~AsyncGenerator() override = default;
|
|
|
|
void async_generator_enqueue(Completion, NonnullGCPtr<PromiseCapability>);
|
|
ThrowCompletionOr<void> resume(VM&, Completion completion);
|
|
void await_return();
|
|
void complete_step(Completion, bool done, Realm* realm = nullptr);
|
|
void drain_queue();
|
|
|
|
State async_generator_state() const { return m_async_generator_state; }
|
|
void set_async_generator_state(Badge<AsyncGeneratorPrototype>, State value);
|
|
|
|
Optional<String> const& generator_brand() const { return m_generator_brand; }
|
|
|
|
private:
|
|
AsyncGenerator(Realm&, Object& prototype, ExecutionContext);
|
|
|
|
virtual void visit_edges(Cell::Visitor&) override;
|
|
|
|
void execute(VM&, Completion completion);
|
|
ThrowCompletionOr<void> await(Value);
|
|
|
|
// At the time of constructing an AsyncGenerator, we still need to point to an
|
|
// execution context on the stack, but later need to 'adopt' it.
|
|
State m_async_generator_state { State::SuspendedStart }; // [[AsyncGeneratorState]]
|
|
ExecutionContext m_async_generator_context; // [[AsyncGeneratorContext]]
|
|
Vector<AsyncGeneratorRequest> m_async_generator_queue; // [[AsyncGeneratorQueue]]
|
|
Optional<String> m_generator_brand; // [[GeneratorBrand]]
|
|
|
|
GCPtr<ECMAScriptFunctionObject> m_generating_function;
|
|
Value m_previous_value;
|
|
Optional<Bytecode::CallFrame> m_frame;
|
|
GCPtr<Promise> m_current_promise;
|
|
};
|
|
|
|
}
|