Wire Format Specification

This document is a specification of the Fuchsia Interface Definition Language (FIDL) data structure encoding.

See Overview for more information about FIDL's overall purpose, goals, and requirements, as well as links to related documents.

Design

Goals

  • Efficiently transfer messages between processes.
  • General purpose, for use with device drivers, high-level services, and applications.
  • Optimized for Zircon IPC only; portability is not a goal.
  • Optimized for direct memory access; inter-machine transport is not a goal.
  • Optimized for 64-bit only; no accommodation for 32-bit environments.
  • Uses uncompressed native data types with host-endianness and correct alignment to support in-place access of message contents.
  • Compatible with C structure in-memory layout (with suitable field ordering and packing annotations).
  • Structures are fixed size and inlined; variable-sized data is stored out-of-line.
  • Structures are not self-described; FIDL files describe their contents.
  • No versioning of structures, but interfaces can be extended with new methods for protocol evolution.
  • No offset calculations required, very little arithmetic which may overflow.
  • Support fast single-pass encoding and validation (as a combined operation).
  • Support fast single-pass decoding and validation (as a combined operation).

Messages

A message is a contiguous data structure represented using the FIDL Wire Format, consisting of a single in-line primary object followed by a sequence of out-of-line secondary objects stored in traversal order.

Objects

Messages are aggregates of objects.

The primary object of a message is simply the first object it contains. It is always a struct of fixed size whose type (and size) is known from the context (such as by examining the method ordinal in the interface call header).

To store variable-size or optional data, the primary object may refer to secondary objects, such as string content, vector content, structs, and unions. Secondary objects are stored out-of-line sequentially in traversal order following the object which reference them. In encoded messages, the presence of secondary objects is marked by a flag. In decoded messages, the flags are substituted with pointers to their location in memory (or null pointers when absent).

Primary and secondary objects are 8-byte aligned and stored sequentially in traversal order without gaps other than the minimum required padding to maintain object alignment.

Objects may also contain in-line objects which are aggregated within the body of the containing object, such as embedded structs and fixed-size arrays of structs. The alignment factor of in-line objects is determined by the alignment factor of their most restrictive member.

In the following example, each Rect structure contains two Point objects which are stored in-line whereas each Region structure contains a vector with a variable number of Rect objects which are stored sequentially out-of-line. In this case, the secondary object is the vector's content (as a unit).

struct Region {
    vector<Rect> rects;
};
struct Rect {
    Point top_left;
    Point bottom_right;
};
struct Point { uint32 x, y; };

drawing

Traversal Order

The traversal order of a message is a determined by a recursive depth-first walk of all of the objects it contains, as obtained by following the chain of references.

Given the following structure:

struct Cart {
    vector<Item> items;
};
struct Item {
    Product product;
    uint32 quantity;
};
struct Product {
    string sku;
    string name;
    string? description;
    uint32 price;
};

The depth-first traversal order for a Cart message is defined by the following pseudo-code:

visit Cart:
    for each Item in Cart.items vector data:
        visit Item.product:
            visit Product.sku
            visit Product.name
            visit Product.description
            visit Product.price
        visit Item.quantity

Dual Forms

The same message content can be expressed in two forms -- encoded and decoded -- which have the same size and overall layout but differ in terms of their representation of pointers (memory addresses) or handles (capabilities).

FIDL is designed such that encoding and decoding of messages can occur in place in memory assuming that objects have been stored in traversal order.

The representation of encoded messages is unambiguous. There is exactly one possible encoding for all messages of the same type with the same content.

drawing

Encoded Messages

An encoded message has been prepared for transfer to another process: it does not contain pointers (memory addresses) or handles (capabilities).

During encoding

  • all pointers to sub-objects within the message are replaced with flags which indicate whether their referent is present or not-present in traversal order
  • all handles within the message are extracted to an associated handle vector and replaced with flags which indicate whether their referent is present or not-present in traversal order

The resulting encoding message and handle vector can then be sent to another process using zx_channel_call() or a similar IPC mechanism.

Decoded Messages

A decoded message has been prepared for use within a process's address space: it may contain pointers (memory addresses) or handles (capabilities).

During decoding...

  • all pointers to sub-objects within the message are reconstructed using the encoded present / not-present flags in traversal order
  • all handles within the message are restored from the associated handle vector using the encoded present / not-present flags in traversal order

The resulting decoded message is ready to be consumed directly from memory.

Data Types

Primitives

  • Value stored in native machine format.
  • Packed with natural alignment.
    • Each m-byte primitive is stored on an m-byte boundary.
  • Not nullable.

The following primitive types are supported:

CategoryTypes
Booleanbool
Signed integerint8 int16 int32 int64
Unsigned integeruint8 uint16 uint32 uint64
IEEE 754 Floating-pointfloat32 float64

Number types are suffixed with their size in bits, bool is 1 byte.

Enums

  • Primitive value representing a proper enumerated type; bit fields are not valid enums.
  • Stored directly using their underlying primitive type.
  • Not nullable.

From the perspective of the wire format, enums are just fancy names for primitive types.

For example, an enum whose underlying type is int32 is stored in exactly the same way as any ordinary C int32_t would be.

Arrays

  • Fixed length sequence of homogeneous elements.
  • Packed with natural alignment of their elements.
    • Alignment of array is the same as the alignment of its elements.
    • Each subsequent element is aligned on element's alignment boundary.
  • The stride of the array is exactly equal to the size of the element (which includes the padding required to satisfy element alignment constraints).
  • Not nullable.
  • There is no special case for arrays of bools. Each bool element takes one byte as usual.

Arrays are denoted:

  • array<T>:n: where T can be any FIDL type (including an array) and n is the number of elements in the array.

drawing

Strings

  • Variable-length sequence of UTF-8 encoded characters representing text.

  • Nullable; null strings and empty strings are distinct.

  • Can specify a maximum size, e.g. string:40 for a maximum 40 byte string.

  • String content does not have a null-terminator.[^1]

  • Stored as a 16 byte record consisting of:

    • size : 64-bit unsigned number of code units (bytes)
    • data : 64-bit presence indication or pointer to out-of-line string data
  • When encoded for transfer, data indicates presence of content:

    • 0 : string is null
    • UINTPTR_MAX : string is non-null, data is the next out-of-line object
  • When decoded for consumption, data is a pointer to content.

    • 0 : string is null
    • <valid pointer> : string is non-null, data is at indicated memory address

Strings are denoted as follows:

  • string : non-nullable string (validation error occurs if null data is encountered)
  • string? : nullable string
  • string:N, string:N? : string with maximum length of N bytes

drawing

Vectors

  • Variable-length sequence of homogeneous elements.
  • Nullable; null vectors and empty vectors are distinct.
  • Can specify a maximum size, e.g. vector<T>:40 for a maximum 40 element vector.
  • Stored as a 16 byte record consisting of:
    • size : 64-bit unsigned number of elements
    • data : 64-bit presence indication or pointer to out-of-line element data
  • When encoded for transfer, data indicates presence of content:
    • 0 : vector is null
    • UINTPTR_MAX : vector is non-null, data is the next out-of-line object
  • When decoded for consumption, data is a pointer to content.
    • 0 : vector is null
    • <valid pointer> : vector is non-null, data is at indicated memory address
  • There is no special case for vectors of bools. Each bool element takes one byte as usual.

Vectors are denoted as follows:

  • vector<T> : non-nullable vector of element type T (validation error occurs if null data is encountered)
  • vector<T>? : nullable vector of element type T
  • vector<T>:N, vector<T>:N? : vector with maximum length of N elements

T can be any FIDL type.

drawing

Handles

  • Transfers a Zircon capability by handle value.

  • Stored as a 32-bit unsigned integer.

  • Nullable by encoding as a zero-valued[^2] handle (equivalent to ZX_HANDLE_INVALID).

  • When encoded for transfer, the stored integer indicates presence of handle:

    • 0 : handle is null
    • UINT32_MAX : handle is non-null, handle value is the next entry in handle table
  • When decoded for consumption, the stored integer is handle value itself.

    • 0 : handle is null
    • <valid handle> : handle is non-null, handle value is provided in-line

Handles are denoted:

  • handle : non-nullable Zircon handle of unspecified type
  • handle? : nullable Zircon handle of unspecified type
  • handle<H> : non-nullable Zircon handle of type H
  • handle<H>? : nullable Zircon handle of type H
  • Interface : non-nullable FIDL interface (client endpoint of channel)
  • Interface? : nullable FIDL interface (client endpoint of channel)
  • request<Interface> : non-nullable FIDL interface request (server endpoint of channel)
  • request<Interface>? : nullable FIDL interface request (server endpoint of channel)

H can be one of[^3]: channel, event, eventpair, fifo, job, process, port, resource, socket, thread, vmo

Structs

  • Record type consisting of a sequence of typed fields.

  • Alignment factor of structure is defined by maximal alignment factor of any of its fields.

  • Structure is padded with zeroes so that its size is a multiple of its alignment factor.

    • e.g. 1. a struct with an int32 and an int8 field has an alignment of 4 bytes (due to int32) and a size of 8 bytes (3 bytes of padding)
    • e.g. 2. a struct with a bool and a string field has an alignment of 8 bytes (due to string) and a size of 24 bytes (7 bytes of padding)
    • e.g. 3. a struct with a bool and a uint8[2] field has an alignment of 1 byte and a size of 3 bytes (no padding!)
  • In general, changing the definition of a struct will break binary compatibility; instead prefer to extend interfaces by adding new methods which use new structs.

  • A struct with no fields (an “empty” struct) has a size of 1 and an alignment of 1.

    • An empty struct is exactly equivalent to a struct with a single uint8 field that contains a value of zero.

Storage of a structure depends on whether it is nullable at point of reference.

  • Non-nullable structures:
    • Contents are stored in-line within their containing type, enabling very efficient aggregate structures to be constructed.
    • The structure layout does not change when inlined; its fields are not repacked to fill gaps in its container.
  • Nullable structures:
    • Contents are stored out-of-line and accessed through an indirect reference.
    • When encoded for transfer, stored reference indicates presence of structure:
      • 0 : reference is null
      • UINTPTR_MAX : reference is non-null, structure content is the next out-of-line object
    • When decoded for consumption, stored reference is a pointer.
      • 0 : reference is null
      • <valid pointer> : reference is non-null, structure content is at indicated memory address

Structs are denoted by their declared name (e.g. Circle) and nullability:

  • Circle : non-nullable Circle
  • Circle? : nullable Circle

The following example shows how structs are laid out according to their fields.

struct Circle {
    bool filled;
    Point center;    // Point will be stored in-line
    float32 radius;
    Color? color;    // Color will be stored out-of-line
    bool dashed;
};
struct Point { float32 x, y; };
struct Color { float32 r, g, b; };

The Color content is padded to the 8 byte secondary object alignment boundary.

drawing

Unions

  • Tagged option type consisting of tag field and variadic contents.
  • Tag field is represented with a uint32 enum.
  • Size of union is the size of the tag field plus the size of the largest option including padding necessary to satisfy its alignment requirements.
  • Alignment factor of union is defined by the maximal alignment factor of the tag field and any of its options.
  • Union is padded so that its size is a multiple of its alignment factor.
    • e.g. 1. a union with an int32 and an int8 option has an alignment of 4 bytes (due to int32) and a size of 8 bytes including the 4 byte tag (0 to 3 bytes of padding).
    • e.g. 2. a union with a bool and a string option has an alignment of 8 bytes (due to string) and a size of 24 bytes (4 to 19 bytes of padding).
  • In general, changing the definition of a union will break binary compatibility; instead prefer to extend interfaces by adding new methods which use new unions.

Storage of a union depends on whether it is nullable at point of reference.

  • Non-nullable unions:
    • Contents are stored in-line within their containing type, enabling very efficient aggregate structures to be constructed.
    • The union layout does not change when inlined; its options are not repacked to fill gaps in its container.
  • Nullable unions:
    • Contents are stored out-of-line and accessed through an indirect reference.
    • When encoded for transfer, stored reference indicates presence of union:
      • 0 : reference is null
      • UINTPTR_MAX : reference is non-null, union content is the next out-of-line object
    • When decoded for consumption, stored reference is a pointer.
      • 0 : reference is null
      • <valid pointer> : reference is non-null, union content is at indicated memory address

Union are denoted by their declared name (e.g. Pattern) and nullability:

  • Pattern : non-nullable Shape
  • Pattern? : nullable Shape

The following example shows how unions are laid out according to their options.

struct Paint {
    Pattern fg;
    Pattern? bg;
};
union Pattern {
    Color color;
    Texture texture;
};
struct Color { float32 r, g, b; };
struct Texture { string name; };

When laying out Pattern, space is first allotted to the tag (4 bytes) then to the selected option.

drawing

Transactional Messages

  • Transactions consist of sequences of correlated messages sent between the client and implementation of an interface over a Zircon channel.
  • Each message is prefixed with a simple 16 byte header, the body immediately follows header.
    • zx_txid_t txid, transaction ID (32 bits)
      • txids with the high bit set are reserved for use by zx_channel_call
      • txids with the high bit unset are reserved for use by userspace
      • See the channel call manpage for more details on txid allocation
    • uint32 reserved0, reserved for future use.
    • uint32 flags, all unused bits must be set to zero
    • uint32 ordinal
      • The zero ordinal is invalid.
      • Ordinals with the most significant bit set are reserved.
        • Ordinals 0x80001xxx are “control” messages
        • Ordinals 0x80002xxx are “fileio” messages
  • A non-zero transaction ID is used to correlate sequences of messages which involve a request and a response, e.g. in a two-way method call. The initiator of the request is responsible for assigning a unique transaction ID to the request. The handler of the request is responsible for echoing the transaction ID it received back into the response which it sends. The initiator can reuse transaction IDs once it receives their corresponding responses.
  • A zero transaction ID is reserved for messages which do not require a response from the other side, e.g. one-way calls or system messages.
  • There are three kinds of messages: method calls, method results, and control messages.
  • Ordinals indicate the purpose of the message.
    • Ordinals with the most significant bit set are reserved for control messages and future expansion.
    • Ordinals without the most significant bit set indicate method calls and responses.
  • Flags control the interpretation of the message. All unused bits must be set to zero.
    • Currently there are no flags, so all bits must be zero.

Messages which are sent directly through Zircon channels have a maximum total size (header + body) which is defined by the kernel (currently 64 KB, eventual intent may be 16 KB).

It is possible to extend interfaces by declaring additional methods with unique ordinals. The language also supports creating derived interfaces provided the method ordinals remain unique down the hierarchy. Interface derivation is purely syntactic; it does not affect the wire format).

We'll use the following interface for the next few examples.

    interface Calculator {
        1: Add(int32 a, int32 b) -> (int32 sum);
        2: Divide(int32 dividend, int32 divisor)
        -> (int32 quotient, int32 remainder);
        3: Clear();
    };

FIDL does not provide a mechanism to determine the “version” of an interface; interface capabilities must be determined out-of-band such as by querying a ServiceProvider for an interface “version” by name or by other means.

Method Call Messages

The client of an interface sends method call messages to the implementor of the interface to invoke methods of that interface.

If a server receives an unknown, unsupported, or unexpected method call message, it must close the channel.

The message indicates the method to invoke by means of its ordinal index. The body of the message contains the method arguments as if they were packed in a struct.

drawing

Method Result Messages

The implementor of an interface sends method result messages to the client of the interface to indicate completion of a method invocation and to provide a (possibly empty) result.

If a client receives an unknown, unsupported, or unexpected method call message, it must close the channel.

Only two-way method calls which are defined to provide a (possibly empty) result in the FIDL interface declaration will elicit a method result message. One-way method calls must not produce a method result message.

A method result message provides the result associated with a prior method call. The body of the message contains the method results as if they were packed in a struct.

The message result header consists of uint32 txid, uint32 reserved, uint32 flags, uint32 ordinal. The txid must be equal to the txid of the method call to which this message is a response. The flags must be zero. The ordinal must be equal to the ordinal of the method call to which this message is a response.

drawing

Event Messages

These support sending unsolicited messages from the server back to the client.

interface Calculator {
    1: Add(int32 a, int32 b) -> (int32 sum);
    2: Divide(int32 dividend, int32 divisor) -> (int32 quotient, int32 remainder);
    3: Clear();
    4: -> Error(uint32 status_code);
};

The implementor of an interface sends unsolicited event messages to the client of the interface to indicate that an asynchronous event occurred as specified by the interface declaration.

Events may be used to let the client observe significant state changes without having to create an additional channel to receive the response.

In the Calculator example, we can imagine that an attempt to divide by zero would cause the Error() event to be sent with a “divide by zero” status code prior to the connection being closed. This allows the client to distinguish between the connection being closed due to an error as opposed to for other reasons (such as the calculator process terminating abnormally).

The body of the message contains the event arguments as if they were packed in a struct, just as with method result messages.

drawing

Control Messages

Control messages support in-band signaling of events other than method calls and responses.

If a client or server receives an unknown, unsupported, or unexpected control message, it must discard it. This allows for future expansion of control messages in the protocol.

The maximum size of a valid control message is 512 bytes, including the header.

Epitaph (Control Message Ordinal 0xFFFFFFFF)

An epitaph is a message with ordinal 0xFFFFFFFF. A server may send an epitaph as the last message prior to closing the connection, to provide an indication of why the connection is being closed. No further messages may be sent through the channel after the epitaph. Epitaphs are not sent from clients to servers.

The epitaph contains an error status. The error status of the epitaph is stored in the reserved uint32 of the message header. The reserved word is treated as being of type zx_status_t: negative numbers are reserved for system error codes, positive numbers are reserved for application error codes, and ZX_OK is used to indicate normal connection closure. The message is otherwise empty.

Details

Size and Alignment

sizeof(T) denotes the size in bytes for an object of type T.

alignof(T) denotes the alignment factor in bytes to store an object of type T.

FIDL primitive types are stored at offsets in the message which are a multiple of their size in bytes. Thus for primitives T_,_ alignof(T) == sizeof(T). This is called natural alignment. It has the nice property of satisfying typical alignment requirements of modern CPU architectures.

FIDL complex types, such as structs and arrays, are stored at offsets in the message which are a multiple of the maximum alignment factor of any of their fields. Thus for complex types T, alignof(T) == max(alignof(F:T)) over all fields F in T. It has the nice property of satisfying typical C structure packing requirements (which can be enforced using packing attributes in the generated code). The size of a complex type is the total number of bytes needed to store its members properly aligned plus padding up to the type's alignment factor.

FIDL primary and secondary objects are aligned at 8-byte offsets within the message, regardless of their contents. The primary object of a FIDL message starts at offset 0. Secondary objects, which are the only possible referent of pointers within the message, always start at offsets which are a multiple of 8. (So all pointers within the message point at offsets which are a multiple of 8.)

FIDL in-line objects (complex types embedded within primary or secondary objects) are aligned according to their type. They are not forced to 8 byte alignment.

Padding

The creator of a message must fill all alignment padding gaps with zeros.

The consumer of a message may verify that padding contains zeroes (and generate an error if not) but it is not required to check as long as it does not actually read the padding bytes.

Maximum Recursion Depth

FIDL arrays, vectors, structures, and unions enable the construction of recursive messages. Left unchecked, processing excessively deep messages could lead to resource exhaustion of the consumer.

For safety, the maximum recursion depth for all FIDL messages is limited to 32 levels of nested complex objects. The FIDL validator must enforce this by keeping track of the current nesting level during message validation.

Complex objects are arrays, vectors, structures, or unions which contain pointers or handles which require fix-up. These are precisely the kinds of objects for which encoding tables must be generated. See C Language Bindings for information about encoding tables. Therefore limiting the nesting depth of complex objects has the effect of limiting the recursion depth for traversal of encoding tables.

Formal definition:

  • The message body is defined to be at nesting level 0.
  • Each time the validator encounters a complex object, it increments the nesting level, recursively validates the object, then decrements the nesting level.
  • If at any time the nesting level exceeds 31, a validation error is raised and validation terminates.

Validation

The purpose of message validation is to discover wire format errors early before they have a chance to induce security or stability problems.

Message validation is required when decoding messages received from a peer to prevent bad data from propagating beyond the service entry point.

Message validation is optional but recommended when encoding messages to send to a peer to help localize violated integrity constraints.

To minimize runtime overhead, validation should generally be performed as part of a single pass message encoding or decoding process such that only a single traversal is needed. Since messages are encoded in depth-first traversal order, traversal exhibits good memory locality and should therefore be quite efficient.

For simple messages, validation may be very trivial, amounting to no more than a few size checks. While programmers are encouraged to rely on their FIDL bindings library to validate messages on their behalf, validation can also be done manually if needed.

Conformant FIDL bindings must check all of the following integrity constraints:

  • The total size of the message including all of its out-of-line sub-objects exactly equals the total size of the buffer that contains it. All sub-objects are accounted for.
  • The total number of handles referenced by the message exactly equals the total size of the handle table. All handles are accounted for.
  • The maximum recursion depth for complex objects is not exceeded.
  • All enum values fall within their defined range.
  • All union tag values fall within their defined range.
  • Encoding only:
    • All pointers to sub-objects encountered during traversal refer precisely to the next buffer position where a sub-object is expected to appear. As a corollary, pointers never refer to locations outside of the buffer.
  • Decoding only:
    • All present / not-present flags for referenced sub-objects hold the value 0 or UINTPTR_MAX.
    • All present / not-present flags for referenced handles hold the value 0 or UINT32_MAX.

Stricter FIDL bindings may perform some or all of the following additional safety checks:

  • All padding is filled with zeroes.
  • All floating point values represent valid IEEE 754 bit patterns.
  • All bools have the value 0 or 1.

Notes

[^1]: Justification for unterminated strings. Since strings can contain embedded null characters, it is safer to encode the size explicitly and to make no guarantees about null-termination, thereby defeating incorrect assumptions that clients might make. Modern APIs generally use sized strings as a security precaution. It's important that data always have one unambiguous interpretation. [^2]: Defining the zero handle to mean “there is no handle” makes it is safe to default-initialize wire format structures to all zeroes. Zero is also the value of the ZX_HANDLE_INVALID constant. [^3]: New handle types can easily be added to the language without affecting the wire format since all handles are transferred the same way.