ChorusOS man pages section 3RPC: RPC Services

rpcgen(3RPC)

NAME

rpcgen- an RPC protocol compiler

SYNOPSIS

rpcgen infile

rpcgen [ -a ] [ -A ] [ -b ] [ -C ] \
[ -D name [ = value ] ] [ -i size ] [ -I [ -K seconds ] ] \
[ -L ] [ -M ] [ -N ] [ -T ] [ -Y pathname ] infile
rpcgen [ -c | -h | -l | -m | -t | -Sc | -Ss | -Sm ] \
[ -o outfile ] [ infile ]
rpcgen [ -s nettype ] [ -o outfile ] [ infile ]
rpcgen [ -n netid ] [ -o outfile ] [ infile ]

DESCRIPTION

rpcgen is a tool that generates C code to implement an RPC protocol. The input to rpcgen is a language similar to C, known as RPC Language (Remote Procedure Call Language).

rpcgen is normally used as in the first synopsis where it takes an infile and generates three output files. Assuming the infile is named proto.x, rpcgen will generate:

a header in proto.h
XDR routines in proto_xdr.c
server-side stubs in proto_svc.c
client-side stubs in proto_clnt.c

proto_tbl.i

rpcgen can also generate sample client and server files that can be customized to suit a particular application. The -Sc, -Ss, and -Sm options generate a sample client, server and Makefile, respectively. The -a option generates all files, including sample files. If the infile is proto.x, then the client side sample file is written to the proto_client.c file, the server side sample file to the proto_server.c file and the sample Makefile to the makefile.proto file.

The server can be started either by itself or by the port monitors (for example, inetd or listen) . When it is started by a port monitor, it only creates servers for the transport for which the file descriptor 0 was passed. The name of the transport must be specified by setting up the environment variable PM_TRANSPORT. When the server generated by rpcgen is executed, it creates server handles for all the transports specified in the NETPATH environment variable, or if it is unset, it creates server handles for all the visible transports from the /etc/netconfig file.

Note -

The transports are chosen at run time and not at compile time. When the server is self-started, by default, it runs in the background. A special define symbol RPC_SVC_FG can be used to run the server process in foreground.

The second synopsis provides special features which allow for the creation of more sophisticated RPC servers. These features include support for user-provided #defines and RPC dispatch tables. The entries in the RPC dispatch table contain:

Pointers to the service routine corresponding to that procedure.
A pointer to the input and output arguments indicating the size of these routines.

A server can use the dispatch table to check authorization and then to execute the service routine; a client library may use it to deal with the details of storage management and XDR data conversion.

The synopses shown above are used when one does not want to generate all the output files, but only a particular one. See the EXAMPLES section below for examples of rpcgen usage. When rpcgen is executed with the -s option, it creates servers for that particular class of transports. When executed with the -n option, it creates a server for the transport specified by netid. If infile is not specified, rpcgen accepts the standard input.

All the options mentioned in the second synopsis can be used with the other three synopses, but the changes will be made only to the specified output file.

The C preprocessor cc -E is run on the input file before it is actually interpreted by rpcgen. For each type of output file, rpcgen defines a special preprocessor symbol for use by the rpcgen programmer:

RPC_HDR: defined when compiling into headers
RPC_XDR: defined when compiling into XDR routines
RPC_SVC: defined when compiling into server-side stubs
RPC_CLNT: defined when compiling into client-side stubs
RPC_TBL: defined when compiling into RPC dispatch tables

Any line beginning with ``%'' is passed directly into the output file, uninterpreted by rpcgen. To specify the path name of the C preprocessor, use the -Y flag.

For every data type referred to in the infile, rpcgen assumes that there is a routine with the string xdr_ associated to the name of the data type. If this routine does not exist in the RPC/XDR library, it must be provided. Providing an undefined data type allows customization of XDR routines.

OPTIONS

The following options are supported:

-a

Generate all files, including sample files.

-A

Enable the Automatic MT mode in the server main program. In this mode, the RPC library automatically creates threads to service client requests. This option generates multithread-safe stubs by implicitly turning on the -M option. Server multithreading modes and parameters can be set using the rpc_control(3N) call. rpcgen generated code does not change the default values for the Automatic MT mode. See RESTRICTIONS below.

-b

Backward compatibility mode. Generate transport-specific RPC code for older versions of the operating system (generates code for SunOS 4.X ChorusOS 4.0).

-c

Compile into XDR routines.

-C

Generate header and stub files which can be used with ANSI C compilers. Headers generated with this flag can also be used with C++ programs.

-D name[=value]

Define a symbol name. Equivalent to the #define directive in the source. If no value is given, the value is defined as 1 . This option may be specified more than once.

-h

Compile into C data-definitions (a header). The -T option can be used in conjunction to produce a header which supports RPC dispatch tables.

-i size

Size at which to start generating inline code. This option is useful for optimization. The default size is 5.

-I

Compile support for inetd(1M) in the server side stubs. Such servers can be self-started or can be started by inetd. When the server is selfstarted, it backgrounds itself by default. A special define symbol, RPC_SVC_FG, can be used to run the server process in foreground, or the user may simply compile without the -I option.

If there are no pending client requests, the inetd servers exit after 120 seconds (default). The default can be changed with the -K option. All of the error messages for inetd servers are always logged with syslog(3).

Note -

This option is supported for backward compatibility only. It should always be used in conjunction with the -b option which generates backward compatibility code. By default (that is, when -b is not specified), rpcgen generates servers that can be invoked through port monitors. See RESTRICTIONS below.

-K

By default, services created using rpcgen and invoked through port monitors wait 120 seconds after servicing a request before exiting. That interval can be changed using the -K flag. To create a server that exits immediately upon servicing a request, use -K 0 . To create a server that never exits, the appropriate argument is -K -1 .

When monitoring for a server, some portmonitors, like listen(1M), always spawn a new process in response to a service request. If it is known that a server will be used with this type of monitor, the server should exit immediately on completion. For these servers, rpcgen should be used with -K 0 . See RESTRICTIONS.

-l

Compile into client-side stubs.

-L

When the servers are started in foreground, use syslog(3) to log the server errors instead of printing them on the standard error. See RESTRICTIONS.

-m

Compile into server-side stubs, but do not generate a "main" routine. This option is useful for doing callback-routines and for users who need to write their own "main" routine to do initialization.

-M

Generate multithread-safe stubs for passing arguments and results between rpcgen-generated code and user written code. This option is useful for users who want to use threads in their code. See RESTRICTIONS.

-N

This option allows procedures to have multiple arguments. It also uses a style of parameter passing that closely resembles C. So, when passing an argument to a remote procedure, you do not have to pass a pointer to the argument, but can pass the argument itself. This behavior is different from the old style of rpcgen-generated code. To maintain backward compatibility, this option is not the default.

-n netid

Compile into server-side stubs for the transport specified by netid. There should be an entry for netid in the netconfig database. This option may be specified more than once, to compile a server that serves multiple transports.

-o outfile

Specify the name of the output file. If none is specified, standard output is used (-c, -h, -l, -m, -n, -s, -Sc, -Sm ,-Ss, and -t modes only).

-s nettype

Compile into server-side stubs for all the transports belonging to the class nettype. The supported classes are netpath, visible, circuit_n, circuit_v, datagram_n, datagram_v, tcp, and udp (see rpc(3N) for the meanings associated with these classes). This option may be specified more than once. Note: the transports are chosen at run time and not at compile time.

-Sc

Generate sample client code that uses remote procedure calls.

-Sm

Generate a sample Makefile which can be used for compiling the application.

-Ss

Generate sample server code that uses remote procedure calls.

-t

Compile into RPC dispatch table.

-T

Generate the code to support RPC dispatch tables. The options -c, -h, -l, -m, -s, -Sc, -Sm ,-Ss, and -t are used exclusively to generate a particular type of file, while the options -D and -T are global and can be used with the other options.

-Y pathname

Give the name of the directory where rpcgen will start looking for the C preprocessor.

OPERANDS

infile input file

EXAMPLES

The following example,

example% rpcgen -T prot.x

prot.h

prot_clnt.c

prot_svc.c

prot_xdr.c

prot_tbl.i

The following example sends the C data-definitions (header) to the standard output:

example% rpcgen -h prot.x

To send the test version of the -DTEST, server side stubs for all the transport belonging to the class datagram_n to standard output, use:

example% rpcgen -s datagram_n -DTEST prot.x

To create the server side stubs for the transport indicated by netid tcp, use:

example% rpcgen -n tcp -o prot_svc.c prot.x