.\" $NetBSD: poll.2,v 1.38 2023/07/07 01:31:25 riastradh Exp $ .\" .\" Copyright (c) 1998, 2005, 2020 The NetBSD Foundation, Inc. .\" All rights reserved. .\" .\" This code is derived from software contributed to The NetBSD Foundation .\" by Charles M. Hannum. .\" .\" Redistribution and use in source and binary forms, with or without .\" modification, are permitted provided that the following conditions .\" are met: .\" 1. Redistributions of source code must retain the above copyright .\" notice, this list of conditions and the following disclaimer. .\" 2. Redistributions in binary form must reproduce the above copyright .\" notice, this list of conditions and the following disclaimer in the .\" documentation and/or other materials provided with the distribution. .\" .\" THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS .\" ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED .\" TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR .\" PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS .\" BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR .\" CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF .\" SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS .\" INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN .\" CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) .\" ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE .\" POSSIBILITY OF SUCH DAMAGE. .\" .Dd February 8, 2021 .Dt POLL 2 .Os .Sh NAME .Nm poll, pollts, ppoll .Nd synchronous I/O multiplexing .Sh LIBRARY .Lb libc .Sh SYNOPSIS .In poll.h .Ft int .Fn poll "struct pollfd *fds" "nfds_t nfds" "int timeout" .In poll.h .In signal.h .In time.h .Ft int .Fn pollts "struct pollfd * restrict fds" "nfds_t nfds" "const struct timespec * restrict ts" "const sigset_t * restrict sigmask" .In poll.h .In signal.h .In time.h .Ft int .Fn ppoll "struct pollfd * restrict fds" "nfds_t nfds" "const struct timespec * restrict ts" "const sigset_t * restrict sigmask" .Sh DESCRIPTION .Fn poll , .Fn pollts and .Fn ppoll examine a set of file descriptors to see if some of them are ready for I/O. For each object inspected, the caller provides a list of conditions (called ``events'') to check for, and the kernel returns a list of conditions that are true. The intent, as with .Xr select 2 , is to check for whether I/O is possible before performing any, so as to permit a top-level event loop to process input from many sources (and output to many destinations) without blocking on any of them and thus becoming stuck. .Ss Arguments The .Fa fds argument is a pointer to an array of pollfd structures, one per file to inspect, as defined in .In poll.h (shown below). The .Fa nfds argument gives the size of the .Fa fds array. .Pp If .Fa timeout is neither zero nor INFTIM (\-1), it specifies a maximum interval to wait for any file descriptor to become ready, in milliseconds. If .Fa timeout is INFTIM (\-1), then .Fn poll blocks indefinitely. If .Fa timeout is zero, then .Fn poll will return without blocking. .Pp Similarly, if .Fa ts is not a null pointer, it references a timespec structure which specifies a maximum interval to wait for any file descriptor to become ready. If .Fa ts is a null pointer, .Fn pollts and .Fn ppoll block indefinitely. If .Fa ts is not a null pointer, referencing a zero-valued timespec structure, then .Fn pollts and .Fn ppoll will return without blocking. .Pp If .Fa sigmask is not a null pointer, then the .Fn pollts and .Fn ppoll functions replace the signal mask of the caller by the set of signals pointed to by .Fa sigmask while the call is in progress, and restore the caller's original signal mask before returning. .Pp The .Vt pollfd structure: .Bd -literal struct pollfd { int fd; /* file descriptor */ short events; /* events to look for */ short revents; /* events returned */ }; .Ed .\" without this Pp there is no blank line after the struct which is ugly .Pp The fields of .Fa struct pollfd are as follows: .Pp .Bl -tag -width XXXrevents -compact .It Fa fd File descriptor to poll. (Input) .It Fa events Conditions to poll for. (Input) .It Fa revents Conditions that are true. (Output) .El .Ss Conditions There are four conditions that can be placed in .Fa events and reported in .Fa revents : .Pp .Bl -tag -width XXXPOLLWRNORM -compact .It POLLRDNORM Normal data may be read without blocking. .It POLLRDBAND Urgent/out-of-band data may be read without blocking. .It POLLWRNORM Normal data may be written without blocking. .It POLLWRBAND Urgent/out-of-band data may be written without blocking. .El .Pp There are three more conditions that are always checked for regardless of .Fa events and thus may always be reported in .Fa revents : .Pp .Bl -tag -width XXXPOLLWRNORM -compact .It POLLERR An exceptional condition has occurred on the object. .It POLLHUP The object has been disconnected. .It POLLNVAL The file descriptor is not open. .El .Pp The following additional flags are defined: .Pp .Bl -tag -width XXXPOLLWRNORM -compact .It POLLIN Synonym for POLLRDNORM. .It POLLOUT Synonym for POLLWRNORM. .It POLLPRI Synonym for POLLRDBAND. .El .Ss Notes If the value passed in .Fa fd is negative, the entry is ignored and .Fa revents is set to 0. (POLLNVAL is .Em not set.) .Pp No file descriptor will ever produce POLLHUP at the same time as POLLWRNORM. .\" (XXX but what about POLLWRBAND? POLLRDBAND? POLLRDNORM?) .Pp Sockets produce POLLIN rather than POLLHUP when the remote end is closed. .Sh RETURN VALUES .Fn poll returns the number of descriptors that are ready for I/O, or returns \-1 and sets .Dv errno if an error occurred. If the time limit expires, .Fn poll returns 0. If .Fn poll returns with an error, including one due to an interrupted call, the .Fa fds array will be unmodified. .Sh COMPATIBILITY This implementation differs from the historical one in that no individual file descriptor may cause .Fn poll to return with an error. In cases where this would have happened in the historical implementation (e.g. trying to poll a .Xr revoke 2 Ns d descriptor), this implementation instead copies the .Fa events bitmask to the .Fa revents bitmask. Attempting to perform I/O on this descriptor will then return an error. This behavior is believed to be more useful. .Pp The .Fn ppoll function is a wrapper for .Fn pollts to provide compatibility with the Linux implementation. .Sh ERRORS An error return from .Fn poll indicates: .Bl -tag -width Er .It Bq Er EFAULT .Fa fds points outside the process's allocated address space. .It Bq Er EINTR A signal was delivered before the time limit expired and before any of the selected events occurred. .It Bq Er EINVAL The specified time limit is negative or the number of pollfd structures specified is larger than the current file descriptor resource limit. .El .Sh SEE ALSO .Xr accept 2 , .Xr connect 2 , .Xr read 2 , .Xr recv 2 , .Xr select 2 , .Xr send 2 , .Xr write 2 .Sh HISTORY The .Fn poll function appeared in .At V.3 , and was added to .Nx in .Nx 1.3 . The .Fn pollts function first appeared in .Nx 3.0 . The .Fn ppoll function first appeared in .Nx 10.0 . .Sh BUGS As of this writing, in the underlying implementation, POLLIN and POLLPRI are distinct flags from POLLRDNORM and POLLRDBAND (respectively) and the behavior is not always exactly identical. .Pp The detailed behavior of specific flags is not very portable from one OS to another. .\" The old note, which is too vague to be helpful: .\" .\" The distinction between some of the fields in the .\" .Fa events .\" and .\" .Fa revents .\" bitmasks is really not useful without STREAMS. .\" The fields are defined for compatibility with existing software.