Chapter 11

Inter-Client Exchange (ICE) Library

Version 1.0

X Consortium Standard

X Version 11, Release 6

Ralph Mor

X Consortium

Copyright © 1993, 1994 X Consortium

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Soft ware, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE X CONSORTIUM BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHITER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

Except as contained in this notice, the name of the X Consortium shall not be used in advertising or otherwise to promote the sale, use or other dealings in this Software without prior written authorization from the X Consortium.

X Window System is a trademark of X Consortium, Inc.

Acknowledgments

Thanks to Bob Scheifler for his thoughtful input on the design of the ICE library. Thanks also to Jordan Brown, Larry Cable, Donna Converse, Clive Feather, Stephen Gildea, Vania Joloboff, Kaleb Keithley, Stuart Marks, Hiro Miyamoto, Ralph Swick, Jirn VanGilder, and Mike Wexler.

Overview of ICE

There are numerous possible "inter-client" protocols, with many similarities and common needs - authentication, version negotiation, byte order negotiation, etc. The ICE protocol is intended to provide a framework for building such protocols, allowing them to make use of common negotiation mechanisms and to be multiplexed over a single transport connection.

11.1 THE ICE LIBRARY -A "C"LANGUAGE INTERFACE TO ICE

A client that wishes to utilize ICE must first register the protocols it understands with the ICE library. Each protocol is dynamically assigned a major opcode ranging from 1-255 (two clients can use different major opcodes for the same protocol). The next step for the client is to either open a connection with another client, or to wait for connections made by other clients. Authentication may be required. A client can both initiate connections with other clients and be waiting for clients to connect to itself (a nested session manager is an example). Once an ICE connection is established between the two clients, one of the clients needs to initiate a Protocol Setup in order to "activate" a given protocol. Once the other client accepts the Protocol Setup (once again, authentication may be required), the two clients are ready to start passing messages specific to that protocol to each other. Multiple protocols may be active on a single ICE connection. Clients are responsible for notifying the ICE library when a protocol is no longer active on an ICE connection, although ICE does not define how each sub-protocol triggers a protocol shutdown.

The ICE library utilizes callbacks to process incoming messages. Using callbacks allows Protocol Setups and authentication to happen "behind the scenes." An additional benefit is that messages never need to be buffered up by the library when the client "blocks" waiting for a particular message.

11.2 INTENDED AUDIENCE

This document is intended primarily for implementors of protocol libraries layered on top of ICE. Typically, applications that wish to utilize ICE will make calls into individual protocol libraries rather than directly make calls into the ICE library. However, some applications will have to make some initial calls into the ICE library in order to accept ICE connections (for example, a session manager accepting connections from clients). But in general, protocol libraries should be designed to hide the inner details of ICE from applications.

11.3 HEADER FILES AND LIBRARY NAME

The header file <X11/ICE/ICE/ICElib.h>defines all of the ICE lib data structures and function prototypes. ICElib.hincludes the header file <X11/ICE/ICE.h>which defines all of the ICElib constants. Protocol libraries that need to read and write messages should include the header file <X11/ICE/ICEmsg.h>.

Applications should link against ICElib using -lICE.

11.4 NOTE ON PREFIXES

The following name prefixes are used in the library to distinguish between a client that initiates a Protocol Setup and a client which responds with a Protocol Reply:

IcePo - Ice Protocol Originator

IcePa - Ice Protocol Acceptor

11.5 PROTOCOL REGISTRATION

In order for two clients to exchange messages for a given protocol, each side must register the protocol with the ICE library. The purpose of registration is for each side to obtain a major opcode for the protocol, and to provide callbacks for processing messages and handling authentication. There are two separate registration functions - one to handle the side that does a Protocol Setup, and one to handle the side that responds with a Protocol Reply.

It is recommended that protocol registration occur before the two clients establish an ICE connection. If protocol registration occurs after an ICE connection is created, there can be a brief interval of time in which a Protocol Setup is received, but the protocol is not registered. If it is not possible to register a protocol before the creation of an ICE commotion, proper precautions should be taken to avoid the above race condition.

The IceRegisterForProtocolSetup function should be called for the client that initiate a Protocol Setup.

int IceRegisterForProtocolSetup(protocol_name,vendor, release, version count, version_recs,auth count, auth narnes, auth_procs, io_error_proc)

char*protocol name;
char*vendor;
char*release;
intversion_count;
IcePoVersionRec*version_recs;
intauth_count;
char**auth_narnes;
IcePoAuthProc*auth_procs;
IceIOErrorProcio_error_proc;

IceRegisterForProtocolSetup returns the major opcode reserved, or -1 if an error occurred. In order to actually activate the protocol, the IceProtocolSetup function needs to be called with this major opcode. Once the protocol is activated, all messages for the protocol should be sent using this major opcode.

A protocol library may support multiple versions of the same protocol. version recs. specifies a list of supported versions of the protocol, prioritized in decreasing order of preference. Each version record consists of a major and minor version of the protocol, as well as a callback to be used for processing incoming messages.

typedef struct {
	int major_version;
	int minor_version;
	IcePoProcessMsgProc process_msg_proc;
} IcePoVersionRec;

The IcePoProcessMsgProc callback is responsible for processing the set of messages that can be received

by the client that initiated the Protocol Setup. The details of how this callback works is described in the section titled Callbacks for Processing Messages.

Authentication may be required before the protocol can become active. The protocol library must register the authentication methods that it supports with the ICE library. auth_names and auth_procs are a list of authentication names and callbacks, prioritized in decreasing order of preference. The details of how the IcePoAuthProc callback works is described in the section titled Authentication Methods.

The IceIOErrorProc callback is invoked if the ICE connection unexpectedly breaks.Pass NULL for io_error_proc if not interested in being notified. See Section 11.12, Error Handling, for more details on this callback.

The IceRegisterForProtocolReply function should be called for the client that responds to a Protocol Setup with a Protocol Reply.

int IceRegisterForProtocolReply(protocol_name, vendor, release,version_count, version_recs, auth_count, auth_names,auth_process, host_based auth_proc, protocol_setup_proc,protocol_activate_proc, io_error_proc)

char *protocol_narne;
char *vendor;
char *release;
int version_count;
IcePoVersionRec *version recs;
int auth_count;
char **auth_names;
IcePoAuthProc *auth_procs;
IceHostBasedAuthProc host_based_auth_proc;
IceProtocolSetupProc protocol_setup_proc;
IceProlocolAclivateProcprotocol activate_proc;
IceIOErrorProc io_error proc;

IceRegisterForProtocolReply returns the major opcode reserved, or -1 if an error occurred. The major opcode should be used in all subsequent messages sent for this protocol.

A protocol library may support multiple versions of the same protocol. version recs specifies a list of supported versions of the protocol, prioritized in decreasing order of preference. Each version record consists

of a major and minor version of the protocol, as well as a callback to be used for processing incoming messages.

typedef struct {
	int major_version;
	int minor_version;
	IcePaProcessMsgProc process_msg_proc;
} IcePaVersionRec;

The IcePaProcessMsgProc callback is responsible for processing the set of messages that can be received by the client that accepted the Protocol Setup.The details of how this callback works is described in the section titled Callbacks for Processing Messages.

Authentication may be required before the protocol can become active. The protocol library must register the authentication methods that it supports with the ICE library. auth_names and auth_procs are a list of authentication names and callbacks, prioritized in decreasing order of preference. The details of how the IcePoAuthProc callback works is described in the section titled Authentication Methods.

If authentication fails and the client attempting to initiate the Protocol Setup has not required authentication, the IceHostBasedAuthProc callback is invoked with the host name of the originating client. If the callback returns True, the Protocol Setup will succeed, even though the original authentication failed. Note that authentication can effectively be disabled by registering an IceHostBasedAuthProc which always returns True. If no host based authentication is allowed, pass NULL for host_based_auth_proc.

typedef Bool (*IceHostBasedAuthProc) ();

Bool HostBasedAuthProc ( host_narne )

char *host_narne;

Since Protocol Setups and authentication happen "behind the scenes" via callbacks, the protocol library needs some way of being notified when the Protocol Setup has completed. This occurs in two phases. In the first phase, the IceProtocolSetupProc callback is invoked after authentication has successfully completed, before the ICE library sends a Protocol Reply. Any resources required for this protocol should be allocated at this time. If the IceProtocolSetupProc returns a successful status, the ICE library will send the Protocol Reply and then invoke the IceProtocolActivateProc callback. Otherwise, an error will be sent to the other client in response to the Protocol Setup.

The IceProtocolActivateProc is an optional callback, and should be registered only if the protocol library intends to generate a message immediately following the Protocol Reply. Pass NULL for protocol_activate proc if not interested in this callback.

typedef Status (*IceProtocolSetupProc) ();

Status ProtocolSetupProc(ice_conn, major_version, minor_version, vendor, release, c

lient_data_ret, failure_reason_ret )
IceConn ice_conn;
int major_version;
int minor version;
char *vendor;
char *release;
IcePointer *client data ret;
char **failure reason_ret;

The pointer stored in the client_data_ret argument will be passed to the IcePaProcessMsgProc callback whenever a message has arrived for this protocol on the ICE connection.

The vendor and release strings should be freed with free() when they are no longer needed.

If a failure occurs, the IceProtocolSetupProc should return a zero status, as well as allocate and return a failure reason string in failure reason_ret. The ICE library will be responsible for freeing this memory.

The IceProtocolActivateProc discussed above is defined as follows:

typedef void (*IceProtocolActivateProc) ();

void ProtocolActivateProc ( ice_conn, client_data)

ice connection ice_conn;
Ice Pointer client_data;

The IcelOErrorProc callback is invoked if the ICE connection unexpectedly breaks. Pass NULL for io error proc if not interested in being notified. See Section 11.12, Error Handling, for more details on this callback.

11.5.1 Callbacks for Processing Messages

When an application detects that there is a new data to read on an ICE connection (via select), it calls the IceProcessMessages function (discussed in the section titled Processing Messages). When IceProcessMessages reads an ICE message header with a major opcode other than zero (reserved for the ICE protocol), it needs to call a function which will read the rest of the message, unpack it, and process it accordingly.

If the message arrives at the client which initiated the Protocol Setup, the IcePoProcessMsgProc callback is invoked.

typed void (*IcePoProcessMsgProc) ();

void PoProcessMsgProc(ice_conn, client data, opcode, length, swap, reply_wait, reply ready ret)

Ice Conn ice_conn;
IcePointer client_data;
int opcode;
unsigned long length;
Bool swap;
IceReplyWaitInfo *reply_wait;
Bool *reply_ready_ret;

If the message arrives at the client which accepted the Protocol Setup, the IcePaProcessMsgProc callback is invoked.

typedef void (*IcePaProcessMsgProc)();

void PaProcessMsgProc(ice_conn, client_data, opcode, length, swap)

IceConn ice conn;
IcePointer client_data;
int opcode;
unsigned long length;
Bool swap;

In order to read the message, both of the above callbacks should use the macros defined in the section of this document titled Reading ICE Messages. Note that byte swapping may be necessary. As a convenience, the length field in the ICE header will be swapped by ICElib if necessary.

In both of the above callbacks, client_data is a pointer to client data that was registered at Protocol Setup time. In the case of IcePoProcessMsgProc, the client data was set in the call to IceProtocolSetup. In the case of IcePaProcessMsgProc,the client data was set in the IceProtocolSetupProc callback.

The IcePoProcessMsgProc callback needs to check the reply_wait argument. If reply_wait is NULL,the ICE library expects the function to pass the message to the client via a callback. For example, if this is a Session Management Save Yourself message, this function should notify the client of the Save Yourself via a callback. The details of how such a callback would be defined is implementation dependent.

However, if reply_wait is not NULL, then the client is waiting for a reply or an error for a message it previously sent. reply_wait is of type IceReplyWaitInfo.

typedef struct {
	unsigned long sequence_of_request;
	int major_opcode_of_request;
	int minor_opcode_of_request;
	IcePointer reply;
} IceReplyWaitInfo;

IceReplyWaitlnfo contains the major/minor opcodes and sequence number of the message for which a reply is being awaited. it also contains a pointer to the reply message to be filled in (the protocol library should cast this IcePOINTER to the appropriate reply type). In most cases, the reply will have some fixed-size part, and the client waiting for the reply will have provided a pointer to a structure to hold this fixed-size data. If there is variable-length data, it would be expected that the IcePoProcessMsgProc callback will have to allocate additional memory and store pointer(s) to that memory in the fixed-size structure. If the entire data is variable length (e.g., a single variable-length string), then the client waiting for the reply would probably just pass a pointer to fixed-size space to hold a pointer, and the IcePoProcessMsgProc callback would allocate the storage and store the pointer. It is the responsibility of the client receiving the reply to free any memory allocated on its behalf.

If reply_wait is not NULL and IcePoProcessMsgProc has a reply or error to return in response to this reply_wait (i.e. no callback was generated), then the reply_ready_ret argument should be set to True. Note that an error should only be returned if it corresponds to the reply being waited for. Otherwise, the IcePoProcessMsgProc should either handle the error internally, or invoke an error handler for its library.

If reply_wait is NULL, then care must be taken not to store any value in reply_ready_ret since this pointer may also be NULL.

The IcePaProcessMsgProc callback, on the other hand, should always pass the message to the client via a callback. For example, if this is a Session Management Interact Request message, this function should notify the client of the Interact Request via a callback.

The reason the IcePaProcessMsgProc callback does not have a reply_wait like IcePoProcessMsgProc does, is because a process that is acting as a "server" should never block for a reply (infinite blocking can occur if the connecting client does not act properly, denying access to other clients).

11.5.2 Authentication Methods

As discussed earlier, a protocol library must register the authentication methods that it soupiness with the ICE library. For each authentication method, there are two callbacks that may be registered - one to handle the side that initiates a Protocol Setup, and one to handle the side that accepts or rejects this request.

IcePoAuthProc is the callback invoked for the client that initiated the Protocol Setup. This callback must be able to respond to the initial Authentication Required message or subsequent Authentication Next Phase messages sent by the other client.

typedef IcePoAuthStatus (*IcePoAuthProc)();

IcePoAuthStatus PoAuthProc(ice_conn, aut_state_ptr, clean_up, swap, auth_datalen, auth_data, reply_datalen_ret, reply_data_ret, error_string_ret)

IceCoM ice_conn;
IcePoinler *auth state_ptr;
Bool clean_up;
Bool swap;
int auth_datakn;
IcePointer auth_data;
int *reply_datalen_ret;
IcePointer *reply_data_ret;
char **error_string_ret;

Authentication may require several phases, depending on the authentication method. As a result, the IcePoAuthProc may be called more than once when authenticating a client, and some state will have to be maintained between each invocation. At the start of each Protocol Setup, *auth_state_ptr is NULL, and the function should initialize its state and set this pointer. In subsequent invocations of the callback, the pointer should be used to get at any state previously stored by the callback.

If needed, the network ID of the client accepting the Protocol Setup can be obtained by calling the IceConnectionString function.

ICElib will be responsible for freeing the reply data_ret and error_string_ret pointers with free().

The auth_data pointer may point to a volatile block of memory. If the data must be kept beyond this invocation of the callback, be sure to make a copy of it.

The IcePoAuthProc should return one of four values:

IcePaAuthProc is the callback invoked for the client that received the Protocol Setup.

typedef IcePaAuthStatus (*IcePaAuthProc) ();

IcePaAuthStatus PaAuthProc(ice_conn, auth_stale ptr, swap, auth_datalen, auth_data, reply_datalen_ret, reply_data_ret, error_string_ret)

IceConn ice_conn;
IcePointer *auth_state_ptr;
Bool swap;
int auth_datalen;
IcePointer auth_data;
int *reply_datalen_ret;
IcePointer *reply_data_ret;
char **error_string_Kt;

Authentication may require several phases, depending on the authentication method. As a result, the IcePaAuthProc may be called more than once when authenticating a client, and some state will have to be maintained between each invocation. At the start of each Proxocol Setup, auth_datalen is zero, *auth_state_ptr is NULL, and the function should initialize its state and set this pointer. In subsequent invocations of the callback, the pointer should be used to get at any state previously stored by the callback.

If needed, the network ID of the client accepting the Protocol Setup can be obtained by calling the IceConnectionString function.

The auth data pointer may point to a volatile block of memory. If the data must be kept beyond this invocation of the callback, be sure to make a copy of it.

ICElib will be responsible for freeing the reply_data_ret and error_string_ret pointers with free().

The IcePaAuthProc should return one of four values:

11.6 ICE CONNECTIONS

In order for two clients to establish an ICE connection, one client has to be "waiting" for connections, and the other client has to initiate the connection. Most clients will initiate connections, so we discuss that first.

11.6.1 Opening an ICE Connection

In order to open an ICE connection with another client (that is waiting for connections), call the Ice Open Connection function.

IceConn lceOpenConnection(network_ids_list, context,must_authenticate, major_opcode_check, error_length, error_string_ret)

char *network_ids_list;
IcePointer context;
Bool must_authenticate;
int major_opcode_check;
int error_length;
char *error_string_ret;

IceOpenConnection retums an opaque ICE connection object if it succeeds, NULL otherwise.

network_ids_list contains a list of network IDs separated by commas. An attempt will be made to use the first network ID. If that fails, an attempt will be made using the second network ID, and so on. Each network ID has the form...

Most protocol libraries will have some sort of "open" function which should internally make a call into Ice Open Connection. When Ice Open Connection is called, it may be possible to use a previously opened ICE connection (if the target client is the same). However, there are cases in which shared ICE connection are not desired.

The context argument is used to determine if an ICE connection can be shared. If context is NULL, then the caller is always willing to share the connection. If context is not NULL, then the caller is not willing to use a previously opened ICE connection that has a different non-NULL context associated with it.

In addition, if major_opcode_check contains a non-zero major opcode value, a previously created ICE connection will be used only if the major opcode is not active on the connection. This can be used to force multiple ICE connection between two clients for the same protocol.

Any authentication requirements are handled internally by the ICE library. The method by which the authentication data is obtained is implementation dependent.*

After IceOpenConnection is called, the client is ready to send a Protocol Setup (provided that IceRegisterForProtocolSetup was called), or receive a Protocol Setup (provided that IceRegisterForProtocolReply was called).

The X Consortium's ICElib implementation uses an .ICEauthority file (see Appendix C).

11.6.2 Listening for ICE Connections

Clients wishing to accept ICE connections must first call IceListenForConnections so they can listen for connections. A list of opaque "listen" objects are returned, one for each type of transport method that is available (for example, Unix Domain, TCP, DECnet, etc.).

Status Ice Listen For Connections (count_ret, listen_objs_ret, error_length, error_string_ret)

int *count_ret;
IceListenObj **listen_objs_ret;
int error length;
char *error_string_ret;

The return value of Ice Listen For Connections is zero for failure, and a positive value for success.

Call Ice FreeListenObjs to close and free the listen objects.

void IceFreeListenObjs ( count, listen_obis )

int count;
IceListenObj listen_objs;

In order to detect a new connection on a listen object, select() must be called on the description associated with the listen object. To obtain the descriptor, call the IceGetListenConnectionNumber ffunction.

int IceGetListenConnectionNumber(listen_obj)

IceListenObj listen_obj;

To obtain the network ID string associated with a listen object, call the IceGetListenConnectionString function.

char *IceGetListenConnectionString (listen obj)

IceListenObj listen_obj;

A network ID has the form...

To compose a string containing a list of network IDs separated by commas (the format recognized by IceOpenConnection), call the IceComposeNetworkldList function.

char *IceComposeNetworkldList (count, listen_objs )

int count;
IceListenObj *listen_objs;

Home

---

11.6.3 Host Based Authentication for ICE Connections

If authentication fails when a client attempts to open an ICE connection, and the initiating client has not required authentication, a host based authentication procedure may be invoked to provide a last chance for the client to connect. Each listen object has such a callback associated with it, and this callback is set using the IceSetHostBasedAuthProc function.

void IceSetHostBasedAuthProc (listen_obj, host_based_auth_proc)

IceListenObj listen obj;
IceHostBasedAuthProc host_based_auth_proc;

By default, each listen object has no host based authentication procedure associated with it. Passing NULL for host_based_auth_proc turns off host based authentication if it was previously set.

typedef Bool (*IceHostBasedAuthProc) ();

Bool HostBasedAuthProc(host_name)

char *host_name;

If IceHostBasedAuthProc returns True, access will be granted, even though the original authentication failed. Note that authentication can effectively be disabled by registering an IceHostBasedAuthProc which always returns True.

Host based authentication is also allowed al Protocol Setup lime. The callback is specified in the IceRegisterForProtocolReply function discussed earlier.

11.6.4 Accepting ICE Connections

After a connection attempt is detected on a listen object returned by IceListenForConnections, Ice Accept Connection should be called. This returns a new opaque ICE connection object.

IceConn Ice Accept Connection(listen_obj, status_ret)

IceListenObj listen_obj;
IceAcceptStatus *status_ret;

The status_ret argument is set to one of the following values:

In general, in order to detect new connections, the application will call select() on the file descriptors associated with the listen objects. When a new connection is detected, the IceAcceptConnection function should be called. IceAcceptConnection may return a new ICE connection that is in a "pending" state. This is because before the connection can become valid, authentication may be necessary. Since the ICE library cannot block and wait for the connection to become valid (infinite blocking can occur if the connecting client does not act properly), the application must wait for the connection status to become "valid".

The following pseudo-code demonstrates how connections are accepted:

new_ice_conn = IceAcceptConnection (listen_obj);
status = IceConnectionStatus (new_ice_conn);
time_start = time_now;

while (status == IceConnectPending)
{
	selectO on {new_ice_conn, all open connections}

	for (each ice_conn in the list of open connections)
		if (data ready on ice_com)
		{
			status = IceProcessMessages (ice_conn, NULL, NULL);
			if (status == IceProcessMessagesIOError)
				IceCloseConnection (ice_conn);
		}

	if (data ready on new_ice_conn)
	{
		/*
		 *IceProcessMessages is called until the connection
		 *is non-pending. Doing so handles the connection
		 *setup request and any authentication requirements.
		 */

		IceProcessMessages (new_ice_conn, NULL, NULL);
		status = IceConnectionStatus (new_ice_com);
	}
	else
	{
		if ( time now - time_start > MAX_WAIT_TIME)
			status = Ice Connect Rejecled;
	}
}

if (status == Ice Connect Accepted)
{
	Add new_ice_conn to the list of open connections
}
else
{
	IceCloseConnection (new_ice_conn);
}

After IceAcceptConnection is called and the connection has been validated, the client is ready to receive a Protocol Setup (provided that IceRegisterForProtocolReply was called), or send a Protocol Setup (provided that IceRegisterForProtocolSetup was called).

11.6.5 Closing ICE Connections

To close an ICE connection created with IceOpenConnection or IceAcceptConnection, call the IceCloseConnection function.

IceCloseStatus IceCloseConnection (ice conn )

IceConn ice_conn;

In order to actually close an ICE connection, the following conditions must be met:

The open reference count must have reached zero on this ICE connection. When IceOpenConnection is called, it tries to use a previously opened ICE connection. If it is able to use an existing connection, it increments the open reference count on the connection by one. So in order to close an ICE connection, each call to IceOpenConnection must be matched with a call to IceCloseConnection. The connection can be closed only on the last call to IceCloseConnection.

-

The active protocol count must have reached zero. Each time a Protocol Setup succeeds on the connection the active protocol count is incremented by one. When the client no longer expects to use the protocol on the connection, the IceProtocolShutdown function should be called, which decrements the active protocol count by one (see the Protocol Setup and Shutdown section).

-

If shutdown negotiation is enabled on the connection, the client on the other side of the ICE connection must agree to have the connection closed.

IceCloseConnection returns one of the following values:

When it is known that the client on the other side of the ICE connection has terminated the Connection without initiating shutdown negotiation, the IceSetShutdownNegotiation function should be called to turn off shutdown negotiation. This will prevent IceCloseConnection from writing to a broken connection.

void IceSetShutdownNegotiation (ice_conn, negotiate )

IceConnice ice_conn;
Bool negotiate;

In order to check the shutdown negotiation status of an ICE connection, call the IceCheckShutdown Negotiation function.

Bool IceCheckShutdownNegotiation (ice conn )

IceConnice ice_conn;

IceCheckShutdownNegotiation returns True if shutdown negotiation will take place on the connection, False otherwise. Negotiation is on by default for a connection. It can only be changed with the IceSetShutdownNegotiation function.

11.6.6 Connection Watch Procedures

In order to add a watch procedure which will be called each time ICElib opens a new connection via IceOpenConnection or IceAcceptConnection, or closes a connection via IceCloseConnection, call the IceAddConnectionWatch function.

Status IceAddConnectionWatch(watch_proc,client_data)

IceWatchProc watch_proc;
IcePointer client_data;

The return value of IceAddConnectionWatch is zero for failure, and a positive value for success.

Note that several calls to IceOpenConnection might share the same ICE connection. In such a case, the watch procedure is only invoked when the connection is first created (after authentication succeeds). Similarly, since connections might be shared, the watch procedure is called only if IceCloseConnection actually closes the connection (right before the IceConn is freed).

The watch procedures are very useful for applications which need to add a file description to a select mask when a new connection is created, and remove the file descriptor when the connection is destroyed. Since connections are shared, knowing when to add and remove the file descriptor from the select mask would be difficult without the watch procedures.

Multiple watch procedures may be registered with the ICE library. No assumptions should be made about their order of invocation.

If one or more ICE connections were already created by the ICE library al the lime the watch procedure is registered, the watch procedure will instantly be invoked for each of these ICE connections (with the opening flag set to True).

The watch procedure is of type IceWatchProc.

typedef void (*IceWatchProc)();

void WatchProc(ice_conn, client_data, opening, watch_data)

IceConn ice_conn;
IcePointer client_data;
Bool openig;
IcePointer *watch_data;

If opening is True, the client should set the *watch_data pointer to any data it may need to save until the connection is closed and the watch procedure is invoked again with opening set to False.

To remove a watch procedure, call the IceRemoveConnectionWatch function.

void IceRemoveConnectionWatch(watch_proc,client_data)

IceWatchProc watch_porc;
IcePointer client_data;

In order to activate a protocol on a given ICE connection, call the IceProtocolSetup function.

IceProlocolSetupStatus IceProtocolSetup(ice_conn, my_opcode, client_data, must_authenticate, major_version ret, minor version_ret, vendor ret, release ret, error_length, error_string ret)
IceConn ice_conn;
int my_opcode;
IcePointer client_data;
Bool must_authenticate;
int *major_version_ret;
int *minor_version_ret;
char **vendor_ret;
char **release_ret;
int error_length;
char *error_string_ret;

The vendor_ret and release_ret strings should be freed with free() when no longer needed.

IceProtocolSetup returns one of the following values:

In order to notify the ICE library when a given protocol will no longer be used on an ICE connection, call the IceProtocolShutdown function.

Status IceProLocolShutdown (ice_conn, major_opcode )

IceConn ice_conn;
int major_opcode;

The return value of IceProtocolShutdown is zero for failure, and a positive value for success.

Failure will occur if the major opcode was never registered OR the protocol of the major opcode was never "activated" on the connection. By "activated" we mean that a Protocol Setup succeeded on the connection. Note that ICE does not define how each sub-protocol triggers a protocol shutdown.

11.8 PROCESSING MESSAGES

In order to process incoming messages on an ICE connection, the IceProcessMessages function should be called.

IceProcessMessagesStatus IceProcessMessages(ice_conn, reply_wait, reply_ready_ret)

IceConn ice_conn;
IceReplyWaitlnfo *reply_wait;
Bool *reply_ready_ret;

This function is used in two ways. In the first a client may generate a message and "block" by calling IceProcessMessages repeatedly until it gets its reply. In the second case, a client calls IceProcessMessages with reply_wait set to NULL in response to select() showing that there is data to read on the ICE connection. The ICE library may process zero or more complete messages. Note that messages which are not "blocked" for are always processed by invoking callbacks.

IceReplyWaitlnfo contains the major/minor opcodes and sequence number of the message for which a reply is being awaited. It also contains a pointer to the reply message to be filled in (the protocol library should cast this IcePointer to the appropriate reply type). In most cases, the reply will have some fixed-size part, and the client waiting for the reply will have provided a pointer to a structure to hold this fixed-size data. If there is variable-length data, it would be expected that the IcePoProcessMsgProc callback will have to allocate additional memory and store pointer(s) to that memory in the fixed-size structure. If the entire data is variable length (e.g., a single variable-length string), then the client waiting for the reply would probably just pass a pointer to fixed-size space to hold a pointer, and the IcePoProcessMsgProc callback would allocate the storage and store the pointer. lt is the responsibility of the client receiving the reply to free up any memory allocated on its behalf.

typedef struct {
	unsigned long sequence_of_request;
	int major_opcode_of_request;
	int minor_opcode_of_request;
	IcePointer reply;
} IceReplyWaitInfo;

If reply_wait is not NULL and IceProcessMessages has a reply or error to return in response to this reply_wait (i.e. no callback was generated), then the reply_ready_ret argument will be set to True.

If reply wait is NULL, then the caller may also pass NULL for reply_ready_ret and be guaranteed that no value will be stored in this pointer.

IceProcessMessages returns one of the following values:

11.9 PING

To send a Ping message to the client on the other side of the ICE connection, call the IcePing function.

Status incepting(ice_conn, ping_reply_proc, clienl_data)

IceConn ice_conn;
IcePingReplyProc ping_reply_proc;
IcePointer client_data;

The return value of IcePing is zero for failure, and a positive value for success.

When IceProcessMessages processes the Ping reply, it will invoke the IcePingReplyProc callback.

typedef void (*IcePingReplyProc)();

void PingReplyProc(ice_conn, client_data)

IceConn ice_conn;
IcePointer client_data;

11.10 Informational Functions

IceConnectStatus IceConnectionStatus ( ice conn )
IceConn ice_conn;

Returns the status of an ICE connection. The possible return values are:

char *lceVendor(ice conn)

IceConn ice_conn;

Returns the ICE library vendor identification for the other side of the connection. The string should be freed with a call to free() when no longer needed.

char *IceRelease(ice_conn)

IceConn ice_conn;

Returns the release identification of the ICE library on the other side of the connection. The string should be freed with a call to free() when no longer needed.

int IceProtocolVersion (ice_conn )

IceConn ice_conn;

Returns the major version of the ICE protocol on this connection.

int IceProtocolRevision (ice_conn )

IceConn ice_conn;

Returns the minor version of the ICE protocol on this connection.

int IceConnectionNumber (ice_conn )

IceConn ice_conn;

Returns the file descriptor of this ICE connection.

char *IceConnectionString(ice_conn)

IceConn ice_conn;

Returns the network ID of the client which accepted this connection. The string should be freed with a call to free() when no longer needed.

unsigned long IceLastSentSequenceNumber(ice_conn)

IceConnice_conn;

Returns the sequence number of the last message sent on this ICE connection.

unsigned long IceLastReceivedSequenceNumber (ice_conn )

IceConn ice_conn;

Returns the sequence number of the last message received on this ICE connection.

Bool IceSwapping(ice_conn)

IceConn ice_conn;

Returns True if byte swapping is necessary when reading messages on the ICE connection.

IcePointer IceGetContext (ice_conn )

IceConn ice_conn;

Returns the context associated with a connection created by IceOpenConnection.

11.11 ICE MESSAGES

All ICE messages have a standard 8 byte header. The ICElib macros which read and write messages rely on the following naming convention for message headers:

The 3rd and 4th bytes of the message header can be used as needed. The length field is specified in units of 8 bytes.

11.11.1. Sending ICE Messages

The ICE library maintains an output buffer used for generating messages. Protocol libraries layered on top of ICE may choose to batch messages together and flush the output buffer at appropriate times.

If an IO error has occurred on an ICE connection, all write operations will be ignored. Refer to the section titled Error Handling for more discussion on handling IO errors.

To get the size of the ICE output buffer, call the IceGetOutBufSize function.

int IceGetOutBufSize(ice_conn)

IceConn ice_conn;

To flush the ICE output buffer, call the IceFlush function.

IceFlush (ice_conn )

IceConn ice_conn;

Note that the output buffer may be implicitly flushed if there is insufficient space to generate a message.

The following macros can be used to generate ICE messages:

IceGetHeader(ice_conn, major_opcode, minor_opcode, header_size, <C_data_t

ype>, pmsg)
IceConn ice_conn;
int major_opcode;
int minor_opcode;
int header_size;
<C_data_type>*pmsg;

IceGetHeader is used to set up a message header on an ICE connection. It sets the major and minor opcodes of the message, and initializes the message's length to the length of the header. If additional variable length data follows, the message's length field should be updated.

IceGet Header Extra(ice_conn, major_opcode, minor_opcode, header_size, extra, <C_data_type>, pmsg, pdata)

IceConn ice_conn;
int major_opcode;
int minor_opcode;
int header_size;
int extra;
<C_data_type> *pmsg;
char pdata;

IceGetHeaderExtra is used to generate a message with a fixed (and relatively small) amount of variable length data. The complete message must fit in the ICE output buffer.

IceSimpleMessage(ice_conn,major_opcode,minor_opcode)

IceConn ice_conn;
int major_opcode;
int minor_pcode;

IceSimpleMessage is used to generate a message which is identical in size to the ICE header message, and has no additional data.

IceErrorHeader(ice_conn, offending_major_opcode, offending_minor_opcode, offending_sequence_num, severity, error_class, data_length)

IceConnice_conn;
int offiending_major_opcode;
int offending_minor_opcode;
int offending_sequence_num;
int severity;
int error_class;
int data_length;

IceErrorHeader sets up an error message header.

Note that the two clients connected by ICE may be using different major opcodes for a given protocol. The offending_major_opcode passed to this macro is the major opcode of the protocol for the client sending the error message.

Generic errors which are common to all protocols have classes in the range 0x8000..0xFFFF. See the Inter-Client Exchange Protocol document for more details.

Per-protocol errors have classes in the range 0x0000-0x7fff.

To write data to an ICE connection, use the IceWriteData macro. If the data fits into the ICE output buffer, it is copied there. Otherwise, the ICE output buffer is flushed and the data is directly sent.

This macro is used in conjunction with IceGetHeader and IceErrorHeader.

IceWriteData(ice_conn, bytes, data)

IceConn ice_conn;
int bytes;
char *data;

To write data as 16 bit quantities, use the IceWriteDatal6 macro.

IceWriteDatal6(ice_conn,byes, data)

IceConn ice_coon;
int bytes;
short *data;

To write data as 32 bit quantities, use the IceWriteData32 macro.

IceWrileData32(ice_conn, bytes, data)

IceConn ice_conn;
int bytes;
long *data;

To bypass copying data to the ICE output buffer, use the IceSendData to directly send date over the network connection. If necessary, the ICE output buffer is first flushed.

IceSendData(ice_conn, bytes, (char *) data )

Ice Conn ice_conn;
int bytes;
char *data;

To force 32 or 64 bit alignment, use the IceWritePad macro. A maximum of 7 pad bytes can be specified.

IceWritePad(ice conn, bytes)

IceConn ice_conn;
int bytes;

Home

---

11.11.2 Reading ICE Messages

The ICE library maintains an input buffer used for reading messages. If the ICE library chooses to perform non-blocking reads (this is implementation dependent), then for every read operation that it makes, zero or more complete messages may be read into the input buffer. As a result, for all of the macros described in this section which "read" messages, an actual read operation will occur on the connection only if the data is not already present in the input buffer.

To get the size of the ICE input buffer, call the IceGetlnBufSize function.

int IceGetlnBufSize (ice_conn )

IceConn ice_conn;

When reading messages, care must be taken to check for IO errors. If any IO error occurs in reading any part of a message, the message should be thrown out. After using any of the macros described below for reading messages, the IceValidIO macro can be used to check if an IO error occurred on the connection. After an IO error has occurred on an ICE connection, all read operations will be ignored. Refer to the section tilled Error Handling for more discussion on handling IO errors.

Bool IceValidIO(ice_conn)

IceConn ice_conn;

The following macros can be used to read ICE messages:

IceReadSimpleMessage(ice conn, <C data_type>, pmsg )

IceConn ice_conn;
<C_data_type> *pmsg;

Ice Read Simple Message is used for messages which are identical in size to the 8 byte ICE header, but use the spare 2 bytes in the header to encode additional data. Note that the ICE library always reads in these first 8 bytes so it can obtain the major opcode of the message. IceReadSimpleMessage simply returns a pointer to these 8 bytes, it does not actually read any data into the input buffer.

For a message with variable length data, there are two ways of reading the message. One method involves reading the complete message in one pass using IceReadCompleteMessage. The second method involves reading the message header (note that this may be larger than the 8 byte ICE header), then reading the variable length data in chunks (see IceReadMessageHeader and IceReadData).

IceReadCompleteMessage(ice_conn, header_size, <C_data_type>, pmsg, pdata)

IceConn ice_conn;
int header_size;
<C_data_type> *pmsg;
char *pdata;

If the ICE input buffer has sufficient space, IceReadCompleteMessage will read the complete message into the ICE input buffer. Otherwise, a buffer will be allocated to hold the variable length data. After the call, the pdata argument should be checked against NULL to make sure that there was sufficient memory to allocate the buffer.

After calling IceReadCompleteMessage and processing the message, IceDisposeCompleteMessage should be called.

IceDisposeCompleteMessage(ice_conn,pdata)

IceConnice_conn;
char *pdata;

If a buffer had to be allocated to hold the variable length data (because it didn't fit in the ICE input buffer), it is freed here by ICElib.

IceReadMessageHeader(ice conn, header_size, <C_data_type>, pmsg)

IceConn ice_conn;
int header_size;
<C_data_type> *pmsg;

IceReadMessageHeader reads just the message header. The rest of the data should be read with the IceReadData family of macros. This method of reading a message should be used when the variable length data must be read in chunks.

To read data directly into a user supplied buffer, use the IceReadData macro.

IceReadData(ice conn, bytes, pdata)

IceConnice_conn;
int bytes;
char *pdata;

To read data as 16 bit quantities, use the IceReadData16 macro

IceReadDatal6(ice_conn, swap, bytes, pdata)

IceConn ice_conn;
Bool swap;
int bytes;
short *pdata;

To read data as 32 bit quantities, use the IceReadData32 macro

IceReadData32(ice_conn, swap, bytes, pdata)

IceConn ice_conn;
Bool swap;
int bytes;
long *pdata;

To force 32 or 64 bit alignment, use the IceReadPad macro A maximum of 7 pad bytes can be specified

IceReadPad(ice_conn, bytes)

IceConn ice_conn;
int bytes;

There are two default error handlers in ICElib: one to handle typically fatal conditions (for example, a connection dying because a machine crashed) and one to handle ICE-specific protocol errors These error handlers can be changed to user-supplied routines if you prefer your own error handling and can be changed as often as you like.

To set the ICE error handler, use IceSetErrorHandler.

IceErrorHandler IceSetErrorHandler(handler)

IceErrorHandler handler;

IceSetErrorHandler returns the previous error handler.

The ICE error handler is invoked when an unexpected ICE protocol error (major opcode 0) is encountered The action of the default handler is to print an explanatory message to stderr and if the severity is fatal, call exit() with a non-zero value If exiting is undesirable, the application should register its own error handler

Note that errors in other protocol domains should be handled by their respective libraries (these libraries should have their own error handlers)

An ICE error handler has the type of IceErrorHandler

typedef void (*IceErrorHandler)();

void ErrorHandler(ice_conn, swap, offending_minor_opcode, offending_sequence_num, error_class, severiry, values)

IceConn ice_conn;
Bool swap;
int offending minor_opcode;
unsigned long offending sequence_nwn;
int error_class;
int severity;
IcePointer values;

The following error classes are defined at the ICE level Refer to the Inter-Client Exchange Protocol document for more details.

IceBadMinor, IceBadState, IceBadLength, IceBadValue, IceBadMajor, IceNoAuth, IceNoVersion, IceSetupFailed, IceAuthRejected, IceAuthFailed, IceProtocolDuplicate, IceMajorOpcodeDuplicate, or IceUnknownProtocol.

To handle fatal I/O errors, use IceSetlOErrorHandler

IcelOErrorHandler IceSetlOErrorHandler(handler)

IceIOErrorHandler handler;

IceSetIOErrorHandler returns the previous IO error handler.

An ICE I/O error handler has the type of IceIOErrorHandler:

typedef void (*IceIOErrorHandler)();

voidIOErrorHandler(ice conn)

IceConn ice_conn;

There are two ways of handling IO errors in ICElib

In the first model, the IO error handler does whatever is necessary to respond to the IO error and then returns, but it does not call IceCloseConnection The ICE connection is given a "bad IO" status, and all future reads and writes to the connection are ignored The next time IceProcessMessages is called it A ill return a status of IceProcessMessagesIOError At that time, the application should call IceCloseConnection.

In the second model, the IO error handler does call IceCloseConnection, and then uses the longjmp() call to get back to the application's main event loop setjmp() and longjmp() may not work; properly on all platforms and special care must be taken to avoid memory leaks, so this second model is less desirable.

Before the application I/O error handler is invoked, protocol libraries that were interested in being notified of I/O errors will have their IceIOErrorProc handlers invoked. This handler is set up in the protocol registration functions (see IceRegisterForProtocolSetup and IceRegisterForProtocolReply), and could be used to clean up state specific to the protocol.

typedef void (*IceIOErrorProc)();

void IOErrorProc(ice_conn)

IceConn ice_conn;

Note that every IceIOErrorProc callback must return This is required because each active protocol must be notified of the broken connection, and the application IO error handler must be invoked afterwards.

11.13 MULTI-THREADING SUPPORT

To declare that multiple threads in an application will be using the ICE library, call IcelnitThreads.

Status IceInitThreads()

The IceInitThreads function must be the first ICElib function a multi-threaded program calls It must complete before any other ICElib call is made IcelnilThreads returns a non-zero status if and only if it was able to successfully initialize the threads package It is safe to call this function more than once, although the threads package will only be initialized once.

Protocol libraries layered on top of ICElib will have to lock critical sections of code that access an ICE connection (for example, when generating messages) Two calls, which are generally implemented as macros, are provided:

IceLockConn(ice_conn)

IceConn ice_conn;

IceUnlockConn(ice_conn)

IceConn ice_conn;

To keep an ICE connection locked across several ICElib calls, applications use IceAppLockConn and IceAppUnlockConn .

void IceAppLockConn(ice_conn)

IceCcnn ice_conn;

The IceAppLockConn function completely locks out other threads from ICElib until IceAppLockConn is called. Other threads attempting to use ICElib will block. If the program has not previously called IceInitThreads, IceAppLockConn has no effect.

void IceAppUnlockConn ( ice_conn )

IceConn ice_conn;

The IceAppUnlockConn function allows other threads to complete ICElib calls which were blocked by a previous call to IceAppLockConn from this thread. If the program has not previously called IceInitThreads, IceAppUnlockConn has no effect.

11.14 MISCELLANEOUS FUNCTIONS

To allocate scratch space (for example, when generating messages with variable data), use the IceAllocScratch function. Each ICE connection has one scratch space associated with it. The scratch space starts off as empty and grows as needed. The contents of the scratch space is not guaranteed to be preserved after any ICElib function is called.

char *IceAllocScratch(ice_conn,size)

IceConn ice_conn;

unsigned long size;

The memory returned by IceAllocScratch should not be freed by the caller! The ICE library will free the memory when the ICE connection is closed.