t_optmgmt(3N)
NAME
t_optmgmt − manage options for a transport endpoint
SYNOPSIS
#include <tiuser.h>
int t_optmgmt (fd, req, ret)
int fd;
struct t_optmgmt ∗req;
struct t_optmgmt ∗ret;
DESCRIPTION
The t_optmgmt function enables a transport user to retrieve, verify, or negotiate protocol options with the transport provider. fd identifies a bound transport endpoint.
The req and ret arguments point to a t_optmgmt structure containing the following members:
struct netbuf opt;
long flags;
The opt field identifies protocol options and the flags field is used to specify the action to take with those options.
The options are represented by a netbuf [see intro(3N)] structure in a manner similar to the address in t_bind. req is used to request a specific action of the provider and to send options to the provider. len specifies the number of bytes in the options, buf points to the options buffer, and maxlen has no meaning for the req argument. The transport provider may return options and flag values to the user through ret. For ret, maxlen specifies the maximum size of the options buffer and buf points to the buffer where the options are to be placed. On return, len specifies the number of bytes of options returned. maxlen has no meaning for the req argument, but must be set in the ret argument to specify the maximum number of bytes the options buffer can hold.
The actual structure and content of the options is imposed by the transport provider. Under CX/UX, the Internet Protocols (ip, udp, and tcp) share a common options structure. The opt.buf field in req points to a list of options defined as follows:
struct opthdr opt1;
<value for opt1>
struct opthdr opt2;
<value for opt2>
...etc...
The total length of this list is set in opt.len in req. The opthdr structure is defined in <sys/socket.h> and contains the following members:
int level;
int name;
int len;
level is set to SOL_SOCKET for options defined at the socket level, and to the protocol number [see getprotobyname(3N)] for other levels. name defines the option name; see ip(4P), tcp(4P), and udp(4P). Len is the length of the value, normally 4 for ints.
The flags field of req can specify one of the following actions:
T_NEGOTIATE
This action enables the user to negotiate the values of the options specified in req with the transport provider. The provider will evaluate the requested options and negotiate the values, returning the negotiated values through ret.
T_CHECK This action enables the user to verify whether the options specified in req are supported by the transport provider. On return, the flags field of ret will have either T_SUCCESS or T_FAILURE set to indicate to the user whether the options are supported. These flags are only meaningful for the T_CHECK request.
T_DEFAULT This action enables a user to retrieve the default options supported by the transport provider into the opt field of ret. In req, the len field of opt must be zero and the buf field may be NULL.
If issued as part of the connectionless-mode service, t_optmgmt may block due to flow control constraints. The function will not complete until the transport provider has processed all previously sent data units.
On failure, t_errno may be set to one of the following:
[TBADF] The specified file descriptor does not refer to a transport endpoint.
[TOUTSTATE] The function was issued in the wrong sequence.
[TACCES] The user does not have permission to negotiate the specified options.
[TBADOPT] The specified protocol options were in an incorrect format or contained illegal information.
[TBADFLAG] An invalid flag was specified.
[TBUFOVFLW] The number of bytes allowed for an incoming argument is not sufficient to store the value of that argument. The information to be returned in ret will be discarded.
[TSYSERR] A system error has occurred during execution of this function.
SEE ALSO
intro(3N), t_getinfo(3N), t_open(3N), ip(4P), tcp(4P), udp(4P). CX/UX Networking Reference Manual
DIAGNOSTICS
t_optmgmt returns 0 on success and −1 on failure and t_errno is set to indicate the error.
CX/UX Networking