V8 Torque user manual

V8 Torque is a language that allows developers contributing to the V8 project to express changes in the VM by focusing on the intent of their changes to the VM, rather than preoccupying themselves with unrelated implementation details. The language was designed to be simple enough to make it easy to directly translate the ECMAScript specification into an implementation in V8, but powerful enough to express the low-level V8 optimization tricks in a robust way, like creating fast-paths based on tests for specific object-shapes.

Torque will be familiar to V8 engineers and JavaScript developers, combining a TypeScript-like syntax that eases both writing and understanding V8 code with syntax and types that reflects concepts that are already common in the CodeStubAssembler. With a strong type system and structured control flow, Torque ensures correctness by construction. Torque’s expressiveness is sufficient to express almost all of the functionality that is currently found in V8’s builtins. It also is very interoperable with CodeStubAssembler builtins and macros written in C++, allowing Torque code to use hand-written CSA functionality and vice-versa.

Torque provides language constructs to represent high-level, semantically-rich tidbits of V8 implementation, and the Torque compiler converts these morsels into efficient assembly code using the CodeStubAssembler. Both Torque’s language structure and the Torque compiler’s error checking ensure correctness in ways that were previously laborious and error-prone with direct usage of the CodeStubAssembler. Traditionally, writing optimal code with the CodeStubAssembler required V8 engineers to carry a lot of specialized knowledge in their heads — much of which was never formally captured in any written documentation — to avoid subtle pitfalls in their implementation. Without that knowledge, the learning curve for writing efficient builtins was steep. Even armed with the necessary knowledge, non-obvious and non-policed gotchas often led to correctness or security bugs. With Torque, many of these pitfalls can be avoided and recognized automatically by the Torque compiler.

Getting started #

Most source written in Torque is checked into the V8 repository under the src/builtins directory, with the file extension .tq. Torque definitions of V8's heap-allocated classses are found alongside their C++ definitions, in .tq files with the same name as corresponding C++ files in src/objects. The actual Torque compiler can be found under src/torque. Tests for Torque functionality are checked in under test/torque, test/cctest/torque, and test/unittests/torque.

To give you a taste of the language, let’s write a V8 builtin that prints “Hello World!”. To do this, we’ll add a Torque macro in a test case and call it from the cctest test framework.

Begin by opening up the test/torque/test-torque.tq file and add the following code at the end (but before the last closing }):

@export
macro PrintHelloWorld(): void {
  Print('Hello world!');
}

Next, open up test/cctest/torque/test-torque.cc and add the following test case that uses the new Torque code to build a code stub:

TEST(HelloWorld) {
  Isolate* isolate(CcTest::InitIsolateOnce());
  CodeAssemblerTester asm_tester(isolate, 0);
  TestTorqueAssembler m(asm_tester.state());
  {
    m.PrintHelloWorld();
    m.Return(m.UndefinedConstant());
  }
  FunctionTester ft(asm_tester.GenerateCode(), 0);
  ft.Call();
}

Then build the cctest executable, and finally execute the cctest test to print ‘Hello world’:

$ out/x64.debug/cctest test-torque/HelloWorld
Hello world!

How Torque generates code #

The Torque compiler doesn’t create machine code directly, but rather generates C++ code that calls V8’s existing CodeStubAssembler interface. The CodeStubAssembler uses the TurboFan compiler’s backend to generate efficient code. Torque compilation therefore requires multiple steps:

  1. The gn build first runs the Torque compiler. It processes all *.tq files. Each Torque file path/to/file.tq causes the generation of the following files:

    • path/to/file-tq-csa.cc and path/to/file-tq-csa.h containing generated CSA macros.
    • path/to/file-tq.inc to be included in in a corresponding header path/to/file.h containing class definitions.
    • path/to/file-tq-inl.inc to be included in the corresponding inline header path/to/file-inl.h, containing C++ accessors of class definitions.
    • path/to/file-tq.cc containing generated heap verifiers, printers, etc.

    The Torque compiler also generates various other known .h files, meant to be consumed by the V8 build.

  2. The gn build then compiles the generated -csa.cc files from step 1 into the mksnapshot executable.

  3. When mksnapshot runs, all of V8’s builtins are generated and packaged in to the snapshot file, including those that are defined in Torque and any other builtins that use Torque-defined functionality.

  4. The rest of V8 is built. All of Torque-authored builtins are made accessible via the snapshot file which is linked into V8. They can be called like any other builtin. In addition, the d8 or chrome executable also includes the generated compilation units related to class definitions directly.

Graphically, the build process looks like this:

Torque tooling #

Basic tooling and development environment support is available for Torque.

Troubleshooting builds involving Torque #

Why do you need to know this? Understanding how Torque files get converted into machine code is important because different problems (and bugs) can potentially arise in the different stages of translating Torque into the binary bits embedded in the snapshot:

constexpr: compile-time vs. run-time #

Understanding the Torque build process is also important to understanding a core feature in the Torque language: constexpr.

Torque allows evaluation of expressions in Torque code at runtime (i.e. when V8 builtins are executed as part of executing JavaScript). However, it also allows expressions to be executed at compile time (i.e. as part of the Torque build process and before the V8 library and d8 executable have even been created).

Torque uses the constexpr keyword to indicate that an expression must be evaluated at build-time. Its usage is somewhat analogous to C++’s constexpr: in addition to borrowing the constexpr keyword and some of its syntax from C++, Torque similarly uses constexpr to indicate the distinction between evaluation at compile-time and runtime.

However, there are some subtle differences in Torque’s constexpr semantics. In C++, constexpr expressions can be evaluated completely by the C++ compiler. In Torque constexpr expressions cannot fully be evaluated by the Torque compiler, but instead map to C++ types, variables and expressions that can be (and must be) fully evaluated when running mksnapshot. From the Torque-writer’s perspective, constexpr expressions do not generate code executed at runtime, so in that sense they are compile-time, even though they are technically evaluated by C++ code external to Torque that mksnapshot runs. So, in Torque, constexpr essentially means “mksnapshot-time”, not “compile time”.

In combination with generics, constexpr is a powerful Torque tool that can be used to automate the generation of multiple very efficient specialized builtins that differ from each other in a small number of specific details that can be anticipated by V8 developers in advance.

Files #

Torque code is packaged in individual source files. Each source file consists of a series of declarations, which themselves can optionally wrapped in a namespace declaration to separate the namespaces of declarations. The following description of the grammar is likely out-of-date. The source-of-truth is the grammar definition in the Torque compiler, which is written using contex-free grammar rules.

A Torque file is a sequence of declarations. The possible declarations are listed in torque-parser.cc.

Namespaces #

Torque namespaces allow declarations to be in independent namespaces. They are similar to C++ namespaces. They allow you to create declarations that are not automatically visible in other namespaces. They can be nested, and declarations inside a nested namespace can access the declarations in the namespace that contains them without qualification. Declarations that are not explicitly in a namespace declaration are put in a shared global default namespace that is visible to all namespaces. Namespaces can be reopened, allowing them to be defined over multiple files.

For example:

macro IsJSObject(o: Object): bool { … }  // In default namespace

namespace array {
  macro IsJSArray(o: Object): bool { … }  // In array namespace
};

namespace string {
  // …
  macro TestVisibility() {
    IsJsObject(o); // OK, global namespace visible here
    IsJSArray(o);  // ERROR, not visible in this namespace
    array::IsJSArray(o);  // OK, explicit namespace qualification
  }
  // …
};

namespace array {
  // OK, namespace has been re-opened.
  macro EnsureWriteableFastElements(array: JSArray){ … }
};

Declarations #

Types #

Torque is strongly typed. Its type system is the basis for many of the security and correctness guarantees it provides.

For many basic types, Torque doesn’t actually inherently know very much about them. Instead, many types are just loosely coupled with CodeStubAssembler and C++ types through explicit type mappings and rely on the C++ compiler to enforce the rigor of that mapping. Such types are realized as abstract types.

Abstract types #

Torque’s abstract types map directly to C++ compile-time and CodeStubAssembler runtime values. Their declarations specify a name and a relationship to C++ types:

AbstractTypeDeclaration :
  type IdentifierName ExtendsDeclaration opt GeneratesDeclaration opt ConstexprDeclaration opt

ExtendsDeclaration :
  extends IdentifierName ;

GeneratesDeclaration :
  generates StringLiteral ;

ConstexprDeclaration :
  constexpr StringLiteral ;

IdentifierName specifies the name of the abstract type, and ExtendsDeclaration optionally specifies the type from which the declared type derives. GeneratesDeclaration optionally specifies a string literal which corresponds to the C++ TNode type used in CodeStubAssembler code to contain a runtime value of its type. ConstexprDeclaration is a string literal specifying the C++ type corresponding to the constexpr version of the Torque type for build-time (mksnapshot-time) evaluation.

Here’s an example from base.tq for Torque’s 31- and 32-bit signed integer types:

type int32 generates 'TNode<Int32T>' constexpr 'int32_t';
type int31 extends int32 generates 'TNode<Int32T>' constexpr 'int31_t';

Union types #

Union types express that a value belongs to one of several possible types. We only allow union types for tagged values, because they can be distinguished at runtime using the map pointer. For example, JavaScript numbers are either Smi values or allocated HeapNumber objects.

type Number = Smi | HeapNumber;

Union types satisfy the following equalities:

It is only allowed to form union types from tagged types because untagged types cannot be distinguished at runtime.

When mapping union types to CSA, the most specific common supertype of all the types of the union type is selected, with the exception of Number and Numeric, which are mapped to the corresponding CSA union types.

Class types #

Class types make it possible to define, allocate and manipulate structured objects on the V8 GC heap from Torque code. Each Torque class type must correspond to a subclass of HeapObject in C++ code. In order to minimize the expense of maintaining boilerplate object-accessing code between V8’s C++ and Torque implementation, the Torque class definitions are used to generate the required C++ object-accessing code whenever possible (and appropriate) to reduce the hassle of keeping C++ and Torque synchronized by hand.

ClassDeclaration :
  ClassAnnotation* extern opt transient opt class IdentifierName ExtendsDeclaration opt GeneratesDeclaration opt {
    ClassMethodDeclaration*
    ClassFieldDeclaration*
  }

ClassAnnotation :
  @doNotGenerateCppClass
  @generateBodyDescriptor
  @generatePrint
  @abstract
  @export
  @noVerifier
  @hasSameInstanceTypeAsParent
  @highestInstanceTypeWithinParentClassRange
  @lowestInstanceTypeWithinParentClassRange
  @reserveBitsInInstanceType ( NumericLiteral )
  @apiExposedInstanceTypeValue ( NumericLiteral )

ClassMethodDeclaration :
  transitioning opt IdentifierName ImplicitParameters opt ExplicitParameters ReturnType opt LabelsDeclaration opt StatementBlock

ClassFieldDeclaration :
  ClassFieldAnnotation* weak opt const opt FieldDeclaration;

ClassFieldAnnotation :
  @noVerifier
  @if ( Identifier )
  @ifnot ( Identifier )

FieldDeclaration :
  Identifier ArraySpecifier opt : Type ;

ArraySpecifier :
  [ Expression ]

An example class:

extern class JSProxy extends JSReceiver {
  target: JSReceiver|Null;
  handler: JSReceiver|Null;
}

extern signifies that this class is defined in C++, rather than defined only in Torque.

The field declarations in classes implicitly generate field getters and setters that can be used from CodeStubAssembler, e.g.:

// In TorqueGeneratedExportedMacrosAssembler:
TNode<HeapObject> LoadJSProxyTarget(TNode<JSProxy> p_o);
void StoreJSProxyTarget(TNode<JSProxy> p_o, TNode<HeapObject> p_v);

As described above, the fields defined in Torque classes generate C++ code that removes the need for duplicate boilerplate accessor and heap visitor code. The hand-written definition of JSProxy must inherit from a generated class template, like this:

// In js-proxy.h:
class JSProxy : public TorqueGeneratedJSProxy<JSProxy, JSReceiver> {

  // Whatever the class needs beyond Torque-generated stuff goes here...

  // At the end, because it messes with public/private:
  TQ_OBJECT_CONSTRUCTORS(JSProxy)
}

// In js-proxy-inl.h:
TQ_OBJECT_CONSTRUCTORS_IMPL(JSProxy)

The generated class provides cast functions, field accessor functions, and field offset constants (e.g. kTargetOffset and kHandlerOffset in this case) representing the byte offset of each field from the beginning of the class.

Class type annotations #

Some classes can't use the inheritance pattern shown in the example above. In those cases, the class can specify @doNotGenerateCppClass, inherit directly from its superclass type, and include a Torque-generated macro for its field offset constants. Such classes must implement their own accessors and cast functions. Using that macro looks like this:

class JSProxy : public JSReceiver {
 public:
  DEFINE_FIELD_OFFSET_CONSTANTS(
      JSReceiver::kHeaderSize, TORQUE_GENERATED_JS_PROXY_FIELDS)
  // Rest of class omitted...
}

@generateBodyDescriptor causes Torque to emit a class BodyDescriptor within the generated class, which represents how the garbage collector should visit the object. Otherwise the C++ code must either define its own object visitation, or use one of the existing patterns (for example, inheriting from Struct and including the class in STRUCT_LIST means that the class is expected to contain only tagged values).

If the @generatePrint annotation is added, then the generator will implement a C++ function that prints the field values as defined by the Torque layout. Using the JSProxy example, the signature would be void TorqueGeneratedJSProxy<JSProxy, JSReceiver>::JSProxyPrint(std::ostream& os), which can be inherited by JSProxy.

The Torque compiler also generates verification code for all extern classes, unless the class opts out with the @noVerifier annotation. For example, the JSProxy class definition above will generate a C++ method void TorqueGeneratedClassVerifiers::JSProxyVerify(JSProxy o, Isolate* isolate) which verifies that its fields are valid according to the Torque type definition. It will also generate a corresponding function on the generated class, TorqueGeneratedJSProxy<JSProxy, JSReceiver>::JSProxyVerify, which calls the static function from TorqueGeneratedClassVerifiers. If you want to add extra verification for a class (such as a range of acceptable values on a number, or a requirement that field foo is true if field bar is non-null, etc.), then add a DECL_VERIFIER(JSProxy) to the C++ class (which hides the inherited JSProxyVerify) and implement it in src/objects-debug.cc. The first step of any such custom verifier should be to call the generated verifier, such as TorqueGeneratedClassVerifiers::JSProxyVerify(*this, isolate);. (To run those verifiers before and after every GC, build with v8_enable_verify_heap = true and run with --verify-heap.)

@abstract indicates that the class itself is not instantiated, and does not have its own instance type: the instance types that logically belong to the class are the instance types of the derived classes.

The @export annotation causes the Torque compiler to generate a concrete C++ class (such as JSProxy in the example above). This is obviously only useful if you don't want to add any C++ functionality beyond that provided by the Torque-generated code. Cannot be used in conjunction with extern. For a class that is defined and used only within Torque, it is most appropriate to use neither extern nor @export.

@hasSameInstanceTypeAsParent indicates classes that have the same instance types as their parent class, but rename some fields, or possibly have a different map. In such cases, the parent class is not abstract.

The annotations @highestInstanceTypeWithinParentClassRange, @lowestInstanceTypeWithinParentClassRange, @reserveBitsInInstanceType, and @apiExposedInstanceTypeValue all affect generation of instance types. Generally you can ignore these and be okay. Torque is responsible for assigning a unique value in the enum v8::internal::InstanceType for every class so that V8 can determine at runtime the type any object in the JS heap. Torque's assignment of instance types should be adequate in the vast majority of cases, but there are a few cases where we want an instance type for a particular class to be stable across builds, or to be at the beginning or end of the range of instance types assigned to its superclass, or to be a range of reserved values that can be defined outside of Torque.

Class fields #

As well as plain values, as in the example above, class fields may contain indexed data. Here's an example:

extern class CoverageInfo extends HeapObject {
  const slot_count: int32;
  slots[slot_count]: CoverageInfoSlot;
}

This means that instances of CoverageInfo are of varying sizes based on the data in slot_count.

Unlike C++, Torque will not implicitly add padding between fields; instead, it will fail and emit an error if fields are not properly aligned. Torque also requires that strong fields, weak fields, and scalar fields be together with other fields of the same category in the field order.

const means that a field cannot be altered at runtime (or at least not easily; Torque will fail compilation if you attempt to set it). This is a good idea for length fields, which should only be reset with great care because they would require freeing any released space and might cause data races with a marking thread.
In fact, Torque requires length fields used for indexed data to be const.

weak at the beginning of a field declaration means that the field is a custom weak reference, as opposed to the MaybeObject tagging mechanism for weak fields.
In addition weak affects generation of constants such as kEndOfStrongFieldsOffset and kStartOfWeakFieldsOffset, which is a legacy feature used in some custom BodyDescriptors and currently also still requires grouping fields marked as weak together. We hope to remove this keyword once Torque is fully capable of generating all BodyDescriptors.

If the object stored in a field may be a MaybeObject-style weak reference (with the second bit set), then Weak<T> should be used in the type and the weak keyword should not be used. There are still some exceptions to this rule, like this field from Map, which can contain some strong and some weak types, and is also marked as weak for inclusion in the weak section:

  weak transitions_or_prototype_info: Map|Weak<Map>|TransitionArray|
      PrototypeInfo|Smi;

@if and @ifnot mark fields that should be included in some build configurations but not others. They accept values from the list in BuildFlags, in src/torque/torque-parser.cc.

Classes defined entirely outside Torque #

Some classes are not defined in Torque, but Torque must know about every class because it is responsible for assigning instance types. For this case, classes can be declared with no body, and Torque will generate nothing for them except the instance type. Example:

extern class OrderedHashMap extends HashTable;

Shapes #

Defining a shape looks just like defining a class except that it uses the keyword shape instead of class. A shape is a subtype of JSObject representing a point-in-time arrangement of in-object properties (in spec-ese, these are "data properties" rather than "internal slots"). A shape does not have its own instance type. An object with a particular shape may change and lose that shape at any time because the object might go into dictionary mode and move all of its properties out to a separate backing store.

Structs #

structs are collections of data that can easily be passed around together. (Completely unrelated to the class named Struct.) Like classes, they can include macros that operate on the data. Unlike classes, they also support generics. The syntax looks similar to a class:

@export
struct PromiseResolvingFunctions {
  resolve: JSFunction;
  reject: JSFunction;
}

struct ConstantIterator<T: type> {
  macro Empty(): bool {
    return false;
  }
  macro Next(): T labels _NoMore {
    return this.value;
  }

  value: T;
}
Struct annotations #

Any struct marked as @export will be included with a predictable name in the generated file gen/torque-generated/csa-types.h. The name is prepended with TorqueStruct, so PromiseResolvingFunctions becomes TorqueStructPromiseResolvingFunctions.

Struct fields can be marked as const, which means they shouldn't be written to. The entire struct can still be overwritten.

Structs as class fields #

A struct may be used as the type of a class field. In that case, it represents packed, ordered data within the class (otherwise, structs have no alignment requirements). This is particularly useful for indexed fields in classes. As an example, DescriptorArray contains an array of three-value structs:

struct DescriptorEntry {
  key: Name|Undefined;
  details: Smi|Undefined;
  value: JSAny|Weak<Map>|AccessorInfo|AccessorPair|ClassPositions;
}

extern class DescriptorArray extends HeapObject {
  const number_of_all_descriptors: uint16;
  number_of_descriptors: uint16;
  raw_number_of_marked_descriptors: uint16;
  filler16_bits: uint16;
  enum_cache: EnumCache;
  descriptors[number_of_all_descriptors]: DescriptorEntry;
}
References and Slices #

Reference<T> and Slice<T> are special structs representing pointers to data held within heap objects. They both contain an object and an offset; Slice<T> also contains a length. Rather than constructing these structs directly, you can use special syntax: &o.x will create a Reference to the field x within the object o, or a Slice to the data if x is an indexed field. For both references and slices, there are const and mutable versions. For references, these types are written as &T and const &T for mutable and constant references, respectively. The mutability refers to the data they point to and might not hold globally, that is, you can create const references to mutable data. For slices, there is no special syntax for the types and the two versions are written ConstSlice<T> and MutableSlice<T>. References can be dereferenced with * or ->, consistent with C++.

References and slices to untagged data can also point to off-heap data.

Bitfield structs #

A bitfield struct represents a collection of numeric data that is packed into a single numeric value. Its syntax looks similar to a normal struct, with the addition of the number of bits for each field.

bitfield struct DebuggerHints extends uint31 {
  side_effect_state: int32: 2 bit;
  debug_is_blackboxed: bool: 1 bit;
  computed_debug_is_blackboxed: bool: 1 bit;
  debugging_id: int32: 20 bit;
}

If a bitfield struct (or any other numeric data) is stored within a Smi, it can be represented using the type SmiTagged<T>.

Function pointer types #

Function pointers can only point to builtins defined in Torque, since this guarantees the default ABI. They are especially useful to reduce binary code size.

While function pointer types are anonymous (like in C), they can be bound to a type alias (like a typedef in C).

type CompareBuiltinFn = builtin(implicit context: Context)(Object, Object, Object) => Number;

Special types #

There are two special types indicated by the keywords void and never. void is used as the return type for callables that do not return a value, and never is used as the return type for callables that never actually return (i.e. only exit through exceptional paths).

Transient types #

In V8, heap objects can change layout at runtime. To express object layouts that are subject to change or other temporary assumptions in the type system, Torque supports the concept of a “transient type”. When declaring an abstract type, adding the keyword transient marks it as a transient type.

// A HeapObject with a JSArray map, and either fast packed elements, or fast
// holey elements when the global NoElementsProtector is not invalidated.
transient type FastJSArray extends JSArray
    generates 'TNode<JSArray>';

For example, in the case of FastJSArray, the transient type is invalidated if the array changes to dictionary elements or if the global NoElementsProtector is invalidated. To express this in Torque, annotate all callables that could potentially do that as transitioning. For example, calling a JavaScript function can execute arbitrary JavaScript, so it is transitioning.

extern transitioning macro Call(implicit context: Context)
                               (Callable, Object): Object;

The way this is policed in the type system is that it is illegal to access a value of a transient type across a transitioning operation.

const fastArray : FastJSArray = Cast<FastJSArray>(array) otherwise Bailout;
Call(f, Undefined);
return fastArray; // Type error: fastArray is invalid here.

Enums #

Enumerations provide a means to define a set of constants and group them under a name similar to
the enum classes in C++. A declaration is introduced by the enum keyword and adheres to the following
syntactical structure:

EnumDeclaration :
  extern enum IdentifierName ExtendsDeclaration opt ConstexprDeclaration opt { IdentifierName list+ (, ...) opt }

A basic example looks like this:

extern enum LanguageMode extends Smi {
  kStrict,
  kSloppy
}

This declaration defines a new type LanguageMode, where the extends clause specifies the underlying
type, that is the runtime type used to represent a value of the enum. In this example, this is TNode<Smi>,
since this is what the type Smi generates. A constexpr LanguageMode converts to LanguageMode
in the generated CSA files since no constexpr clause is specified on the enum to replace the default name.
If the extends clause is omitted, Torque will generate only the constexpr version of the type. The extern keyword tells Torque that there is a C++ definition of this enum. Currently, only extern enums are supported.

Torque generates a distinct type and constant for each of the enum's entries. Those are defined
inside a namespace that matches the enum's name. Necessary specializations of FromConstexpr<> are
generated to convert from the entry's constexpr types to the enum type. The value generated for an entry in the C++ files is <enum-constexpr>::<entry-name> where <enum-constexpr> is the constexpr name generated for the enum. In the above example, those are LanguageMode::kStrict and LanguageMode::kSloppy.

Torque's enumerations work very well together with the typeswitch construct, because the
values are defined using distinct types:

typeswitch(language_mode) {
  case (LanguageMode::kStrict): {
    // ...
  }
  case (LanguageMode::kSloppy): {
    // ...
  }
}

If the C++ definition of the enum contains more values than those used in .tq files, Torque needs to know that. This is done by declaring the enum 'open' by appending a ... after the last entry. Consider the ExtractFixedArrayFlag for example, where only some of the options are available/used from within
Torque:

enum ExtractFixedArrayFlag constexpr 'CodeStubAssembler::ExtractFixedArrayFlag' {
  kFixedDoubleArrays,
  kAllFixedArrays,
  kFixedArrays,
  ...
}

Callables #

Callables are conceptually like functions in JavaScript or C++, but they have some additional semantics that allow them to interact in useful ways with CSA code and with the V8 runtime. Torque provides several different types of callables: macros, builtins, runtimes and intrinsics.

CallableDeclaration :
  MacroDeclaration
  BuiltinDeclaration
  RuntimeDeclaration
  IntrinsicDeclaration

macro callables #

Macros are a callable that correspond to a chunk of generated CSA-producing C++. macros can either be fully defined in Torque, in which case the CSA code is generated by Torque, or marked extern, in which case the implementation must be provided as hand-written CSA code in a CodeStubAssembler class. Conceptually, it’s useful to think of macros of chunks of inlinable CSA code that are inlined at callsites.

macro declarations in Torque take the following form:

MacroDeclaration :
   transitioning opt macro IdentifierName ImplicitParameters opt ExplicitParameters ReturnType opt LabelsDeclaration opt StatementBlock
  extern transitioning opt macro IdentifierName ImplicitParameters opt ExplicitTypes ReturnType opt LabelsDeclaration opt ;

Every non-extern Torque macro uses the StatementBlock body of the macro to create a CSA-generating function in its namespace’s generated Assembler class. This code looks just like other code that you might find in code-stub-assembler.cc, albeit a bit less readable because it’s machine-generated. macros that are marked extern have no body written in Torque and simply provide the interface to hand-written C++ CSA code so that it’s usable from Torque.

macro definitions specify implicit and explict parameters, an optional return type and optional labels. Parameters and return types will be discussed in more detail below, but for now it suffices to know that they work somewhat like TypeScript parameters, which as discussed in the Function Types section of the TypeScript documentation here.

Labels are a mechanism for exceptional exit from a macro. They map 1:1 to CSA labels and are added as CodeStubAssemblerLabels*-typed parameters to the C++ method generated for the macro. Their exact semantics are discussed below, but for the purpose of a macro declartion, the comma-separated list of a macro’s labels is optionally provided with the labels keywords and positioned after the macro’s parameter lists and return type.

Here’s an example from base.tq of external and Torque-defined macros:

extern macro BranchIfFastJSArrayForCopy(Object, Context): never
    labels Taken, NotTaken;
macro BranchIfNotFastJSArrayForCopy(implicit context: Context)(o: Object):
    never
    labels Taken, NotTaken {
  BranchIfFastJSArrayForCopy(o, context) otherwise NotTaken, Taken;
}

builtin callables #

builtins are similar to macros in that they can either be fully defined in Torque or marked extern. In the Torque-based builtin case, the body for the builtin is used to generate a V8 builtin that can be called just like any other V8 builtin, including automatically adding the relevant information in builtin-definitions.h. Like macros, Torque builtins that are marked extern have no Torque-based body and simply provide an interface to existing V8 builtins so that they can be used from Torque code.

builtin declarations in Torque have the following form:

MacroDeclaration :
  transitioning opt javascript opt builtin IdentifierName ImplicitParameters opt ExplicitParametersOrVarArgs ReturnType opt StatementBlock
  extern transitioning opt javascript opt builtin IdentifierName ImplicitParameters opt ExplicitTypesOrVarArgs ReturnType opt ;

There is only one copy of the code for a Torque builtin, and that is in the generated builtin code object. Unlike macros, when builtins are called from Torque code, the CSA code is not inlined at the callsite, but instead a call is generated to the builtin.

builtins cannot have labels.

If you are coding the implementation of a builtin, you can craft a tailcall to a builtin or a runtime function iff (if and only if) it's the final call in the builtin. The compiler may be able to avoid creating a new stack frame in this case. Simply add tail before the call, as in tail MyBuiltin(foo, bar);.

runtime callables #

runtimes are similar to builtins in that they can expose an interface to external functionality to Torque. However, instead of being implemented in CSA, the functionality provided by a runtime must always be implemented in the V8 as a standard runtime callback.

runtime declarations in Torque have the following form:

MacroDeclaration :
  extern transitioning opt runtime IdentifierName ImplicitParameters opt ExplicitTypesOrVarArgs ReturnType opt ;

The extern runtime specified with name IdentifierName corresponds to the runtime function specified by Runtime::kIdentifierName.

Like builtins, runtimes cannot have labels.

You can also call a runtime function as a tailcall when appropriate. Simply include the tail keyword before the call.

Runtime function declarations are often placed in a namespace called runtime. This disambiguates them from builtins of the same name and makes it easier to see at the callsite that we are calling a runtime funtion. We should consider making this mandatory.

intrinsic callables #

intrinsics are builtin Torque callables that provide access to internal funtionality that can’t be otherwise implemented in Torque. They are declared in Torque, but not defined, since the implementation is provided by the Torque compiler. intrinsic declarations use the following grammar:

IntrinsicDeclaration :
  intrinsic % IdentifierName ImplicitParameters opt ExplicitParameters ReturnType opt ;

For the most part, “user” Torque code should rarely have to use intrinsics directly.
The following are some of the supported intrinsics:

// %RawObjectCast downcasts from Object to a subtype of Object without
// rigorous testing if the object is actually the destination type.
// RawObjectCasts should *never* (well, almost never) be used anywhere in
// Torque code except for in Torque-based UnsafeCast operators preceeded by an
// appropriate type assert()
intrinsic %RawObjectCast<A: type>(o: Object): A;

// %RawPointerCast downcasts from RawPtr to a subtype of RawPtr without
// rigorous testing if the object is actually the destination type.
intrinsic %RawPointerCast<A: type>(p: RawPtr): A;

// %RawConstexprCast converts one compile-time constant value to another.
// Both the source and destination types should be 'constexpr'.
// %RawConstexprCast translate to static_casts in the generated C++ code.
intrinsic %RawConstexprCast<To: type, From: type>(f: From): To;

// %FromConstexpr converts a constexpr value into into a non-constexpr
// value. Currently, only conversion to the following non-constexpr types
// are supported: Smi, Number, String, uintptr, intptr, and int32
intrinsic %FromConstexpr<To: type, From: type>(b: From): To;

// %Allocate allocates an unitialized object of size 'size' from V8's
// GC heap and "reinterpret casts" the resulting object pointer to the
// specified Torque class, allowing constructors to subsequently use
// standard field access operators to initialize the object.
// This intrinsic should never be called from Torque code. It's used
// internally when desugaring the 'new' operator.
intrinsic %Allocate<Class: type>(size: intptr): Class;

Like builtins and runtimes, intrinsics cannot have labels.

Explicit parameters #

Declarations of Torque-defined Callables, e.g. Torque macros and builtins, have explicit parameter lists. They are a list of identifier and type pairs using a syntax reminiscent of typed TypeScript function parameter lists, with the exception that Torque doesn’t support optional parameters or default parameters. Moreover, Torque-implement builtins can optionally support rest parameters if the builtin uses V8’s internal JavaScript calling convention (e.g. is marked with the javascript keyword).

ExplicitParameters :
  ( ( IdentifierName : TypeIdentifierName ) list* )
  ( ( IdentifierName : TypeIdentifierName ) list+ (, ... IdentifierName ) opt )

As an example:

javascript builtin ArraySlice(
    (implicit context: Context)(receiver: Object, ...arguments): Object {
  // …
}

Implicit parameters #

Torque callables can specify implicit parameters using something similar to Scala’s implicit parameters:

ImplicitParameters :
  ( implicit ( IdentifierName : TypeIdentifierName ) list* )

Concretely: A macro can declare implicit parameters in addition to explicit ones:

macro Foo(implicit context: Context)(x: Smi, y: Smi)

When mapping to CSA, implicit parameters and explicit parameters are treated the same and form a joint parameter list.

Implicit parameters are not mentioned at the callsite, but instead are passed implicitly: Foo(4, 5). For this to work, Foo(4, 5) must be called in a context that provides a value named context. Example:

macro Bar(implicit context: Context)() {
  Foo(4, 5);
}

In contrast to Scala, we forbid this if the names of the implicit parameters are not identical.

Since overload resolution can cause confusing behavior, we ensure that implicit parameters do not influence overload resolution at all. That is: when comparing candidates of an overload set, we do not consider the available implicit bindings at the call-site. Only after we found a single best overload, we check if implicit bindings for the implicit parameters are available.

Having the implicit parameters left of the explicit parameters is different from Scala, but maps better to the existing convention in CSA to have the context parameter first.

js-implicit #

For builtins with JavaScript linkage defined in Torque, you should use the keyword js-implicit instead of implicit. The arguments are limited to these four components of the calling convention:

They don’t all have to be declared, only the ones you want to use. For an example, here is our code for Array.prototype.shift:

  // https://tc39.es/ecma262/#sec-array.prototype.shift
  transitioning javascript builtin ArrayPrototypeShift(
      js-implicit context: NativeContext, receiver: JSAny)(...arguments): JSAny {
  ...

Note that the context argument is a NativeContext. This is because builtins in V8 always embed the native context in their closures. Encoding this in the js-implicit convention allows the programmer to eliminate an operation to load the native context from the function context.

Overload resolution #

Torque macros and operators (which are just aliases for macros) allow for argument-type overloading. The overloading rules are inspired by the ones of C++: an overload is selected if it is strictly better than all alternatives. This means that it has to be strictly better in at least one parameter, and better or equally good in all others.

When comparing a pair of corresponding parameters of two overloads…

If no overload is strictly better than all alternatives, this results in a compile error.

Deferred blocks #

A statement block can optionally be marked as deferred, which is a signal to the compiler that it's entered less often. The compiler may choose to locate these blocks at the end of the function, thus improving cache locality for the non-deferred regions of code. For example, in this code from the Array.prototype.forEach implementation, we expect to remain on the "fast" path, and only rarely take the bailout case:

  let k: Number = 0;
  try {
    return FastArrayForEach(o, len, callbackfn, thisArg)
        otherwise Bailout;
  }
  label Bailout(kValue: Smi) deferred {
    k = kValue;
  }

Here is another example, where the dictionary elements case is marked as deferred to improve code generation for the more likely cases (from the Array.prototype.join implementation):

  if (IsElementsKindLessThanOrEqual(kind, HOLEY_ELEMENTS)) {
    loadFn = LoadJoinElement<FastSmiOrObjectElements>;
  } else if (IsElementsKindLessThanOrEqual(kind, HOLEY_DOUBLE_ELEMENTS)) {
    loadFn = LoadJoinElement<FastDoubleElements>;
  } else if (kind == DICTIONARY_ELEMENTS)
    deferred {
      const dict: NumberDictionary =
          UnsafeCast<NumberDictionary>(array.elements);
      const nofElements: Smi = GetNumberDictionaryNumberOfElements(dict);
      // <etc>...

Porting CSA code to Torque #

The patch that ported Array.of serves as a minimal example of porting CSA code to Torque.