.\" .\" epoll by Davide Libenzi ( efficient event notification retrieval ) .\" Copyright (C) 2003 Davide Libenzi .\" .\" This program is free software; you can redistribute it and/or modify .\" it under the terms of the GNU General Public License as published by .\" the Free Software Foundation; either version 2 of the License, or .\" (at your option) any later version. .\" .\" This program is distributed in the hope that it will be useful, .\" but WITHOUT ANY WARRANTY; without even the implied warranty of .\" MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the .\" GNU General Public License for more details. .\" .\" You should have received a copy of the GNU General Public License .\" along with this program; if not, write to the Free Software .\" Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA .\" .\" Davide Libenzi .\" .TH EPOLL_CTL 2 "2002-10-23" Linux "Linux Programmer's Manual" .SH NAME epoll_ctl \- control interface for an epoll descriptor .SH SYNOPSIS .B #include .sp .BI "int epoll_ctl(int " epfd ", int " op ", int " fd ", struct epoll_event *" event ) .SH DESCRIPTION Control an .B epoll descriptor, .IR epfd , by requesting that the operation .IR op be performed on the target file descriptor, .IR fd . The .IR event describes the object linked to the file descriptor .IR fd . The .I struct epoll_event is defined as : .sp .in +0.5i .nf typedef union epoll_data { void *ptr; int fd; __uint32_t u32; __uint64_t u64; } epoll_data_t; struct epoll_event { __uint32_t events; /* Epoll events */ epoll_data_t data; /* User data variable */ }; .fi .in -0.5i The .I events member is a bit set composed using the following available event types : .TP .B EPOLLIN The associated file is available for .BR read (2) operations. .TP .B EPOLLOUT The associated file is available for .BR write (2) operations. .TP .BR EPOLLRDHUP " (since kernel 2.6.17)" Stream socket peer closed connection, or shut down writing half of connection. (This flag is especially useful for writing simple code to detect peer shutdown when using Edge Triggered monitoring.) .TP .B EPOLLPRI There is urgent data available for .BR read (2) operations. .TP .B EPOLLERR Error condition happened on the associated file descriptor. .BR epoll_wait (2) will always wait for this event; it is not necessary to set it in .IR events . .TP .B EPOLLHUP Hang up happened on the associated file descriptor. .BR epoll_wait (2) will always wait for this event; it is not necessary to set it in .IR events . .TP .B EPOLLET Sets the Edge Triggered behaviour for the associated file descriptor. The default behaviour for .B epoll is Level Triggered. See .BR epoll (7) for more detailed information about Edge and Level Triggered event distribution architectures. .TP .BR EPOLLONESHOT " (since kernel 2.6.2)" Sets the one-shot behaviour for the associated file descriptor. This means that after an event is pulled out with .BR epoll_wait (2) the associated file descriptor is internally disabled and no other events will be reported by the .B epoll interface. The user must call .BR epoll_ctl (2) with .B EPOLL_CTL_MOD to re-enable the file descriptor with a new event mask. .PP The .B epoll interface supports all file descriptors that support .BR poll (2). Valid values for the .IR op parameter are : .RS .TP .B EPOLL_CTL_ADD Add the target file descriptor .I fd to the .B epoll descriptor .I epfd and associate the event .I event with the internal file linked to .IR fd . .TP .B EPOLL_CTL_MOD Change the event .I event associated with the target file descriptor .IR fd . .TP .B EPOLL_CTL_DEL Remove the target file descriptor .I fd from the .B epoll file descriptor, .IR epfd . The .IR event is ignored and can be NULL (but see BUGS below). .RE .SH "RETURN VALUE" When successful, .BR epoll_ctl (2) returns zero. When an error occurs, .BR epoll_ctl (2) returns \-1 and .I errno is set appropriately. .SH ERRORS .TP .B EBADF .I epfd or .I fd is not a valid file descriptor. .TP .B EEXIST .I op was EPOLL_CTL_ADD, and the supplied file descriptor .IR fd is already in .IR epfd . .TP .B EINVAL .IR epfd is not an .B epoll file descriptor, or .IR fd is the same as .IR epfd , or the requested operation .I op is not supported by this interface. .TP .B ENOENT .I op was EPOLL_CTL_MOD or EPOLL_CTL_DEL, and .IR fd is not in .IR epfd . .TP .B ENOMEM There was insufficient memory to handle the requested .I op control operation. .TP .B EPERM The target file .I fd does not support .BR epoll . .SH CONFORMING TO .BR epoll_ctl (2) is Linux specific, and was introduced in kernel 2.5.44. .\" The interface should be finalized by Linux kernel 2.5.66. .SH BUGS In kernel versions before 2.6.9, the .B EPOLL_CTL_DEL operation required a non-NULL pointer in .IR event , even though this argument is ignored. Since kernel 2.6.9, .I event can be specified as NULL when using .BR EPOLL_CTL_DEL . .SH "SEE ALSO" .BR epoll_create (2), .BR epoll_wait (2), .BR poll (2), .BR epoll (7)