|Main index||Section 3||日本語||Options|
The clnt_call(), rpc_call(), and rpc_broadcast() routines handle the client side of the procedure call. The remaining routines deal with error handling in the case of errors.
Some of the routines take a CLIENT handle as one of the arguments. A CLIENT handle can be created by an RPC creation routine such as clnt_create() (see rpc_clnt_create(3)).
These routines are safe for use in multithreaded applications. CLIENT handles can be shared between threads, however in this implementation requests by different threads are serialized (that is, the first request will receive its results before the second request is sent).
|A function macro that calls the remote procedure procnum associated with the client handle, clnt, which is obtained with an RPC client creation routine such as clnt_create() (see rpc_clnt_create(3)). The inproc argument is the XDR function used to encode the procedure's arguments, and outproc is the XDR function used to decode the procedure's results; in is the address of the procedure's argument(s), and out is the address of where to place the result(s). The tout argument is the time allowed for results to be returned, which is overridden by a time-out set explicitly through clnt_control(), see rpc_clnt_create(3). If the remote call succeeds, the status returned is RPC_SUCCESS, otherwise an appropriate status is returned.|
|A function macro that frees any data allocated by the RPC/XDR system when it decoded the results of an RPC call. The out argument is the address of the results, and outproc is the XDR routine describing the results. This routine returns 1 if the results were successfully freed, and 0 otherwise.|
|A function macro that copies the error structure out of the client handle to the structure at address errp.|
|Print a message to standard error corresponding to the condition indicated by stat. A newline is appended. Normally used after a procedure call fails for a routine for which a client handle is not needed, for instance rpc_call().|
|Print a message to the standard error indicating why an RPC call failed; clnt is the handle used to do the call. The message is prepended with string s and a colon. A newline is appended. Normally used after a remote procedure call fails for a routine which requires a client handle, for instance clnt_call().|
|Take the same arguments as clnt_perrno(), but instead of sending a message to the standard error indicating why an RPC call failed, return a pointer to a string which contains the message. The clnt_sperrno() function is normally used instead of clnt_perrno() when the program does not have a standard error (as a program running as a server quite likely does not), or if the programmer does not want the message to be output with printf() (see printf(3)), or if a message format different than that supported by clnt_perrno() is to be used. Note: unlike clnt_sperror() and clnt_spcreateerror() (see rpc_clnt_create(3)), clnt_sperrno() does not return pointer to static data so the result will not get overwritten on each call.|
|Like clnt_perror(), except that (like clnt_sperrno()) it returns a string instead of printing to standard error. However, clnt_sperror() does not append a newline at the end of the message. Warning: returns pointer to a buffer that is overwritten on each call.|
|Like rpc_call(), except the call message is broadcast to all the connectionless transports specified by nettype. If nettype is NULL, it defaults to "netpath". Each time it receives a response, this routine calls eachresult(), whose form is: bool_t eachresult(caddr_t out, const struct netbuf * addr, const struct netconfig * netconf) where out is the same as out passed to rpc_broadcast(), except that the remote procedure's output is decoded there; addr points to the address of the machine that sent the results, and netconf is the netconfig structure of the transport on which the remote server responded. If eachresult() returns 0, rpc_broadcast() waits for more replies; otherwise it returns with appropriate status. Warning: broadcast file descriptors are limited in size to the maximum transfer size of that transport. For Ethernet, this value is 1500 bytes. The rpc_broadcast() function uses AUTH_SYS credentials by default (see rpc_clnt_auth(3)).|
|Like rpc_broadcast(), except that the initial timeout, inittime and the maximum timeout, waittime are specified in milliseconds. The inittime argument is the initial time that rpc_broadcast_exp() waits before resending the request. After the first resend, the re-transmission interval increases exponentially until it exceeds waittime.|
|Call the remote procedure associated with prognum, versnum, and procnum on the machine, host. The inproc argument is used to encode the procedure's arguments, and outproc is used to decode the procedure's results; in is the address of the procedure's argument(s), and out is the address of where to place the result(s). The nettype argument can be any of the values listed on rpc(3). This routine returns RPC_SUCCESS if it succeeds, or an appropriate status is returned. Use the clnt_perrno() routine to translate failure status into error messages. Warning: rpc_call() uses the first available transport belonging to the class nettype, on which it can create a connection. You do not have control of timeouts or authentication using this routine.|
|RPC_CLNT_CALLS (3)||May 7, 1993|
|Main index||Section 3||日本語||Options|
|“||Today, the Unix equivalent of a power drill would have 20 dials and switches, come with a nonstandard plug, require the user to hand-wind the motor coil, and not accept 3/8" or 7/8" drill bits (though this would be documented in the BUGS section of its instruction manual).||”|
|— The Unix Haters' handbook|