xref: /openbsd-src/share/man/man9/tsleep.9 (revision 41ce3b17e73f6b7d2d9e1a3d961e4bab2d895cb5)

.\"	$OpenBSD: tsleep.9,v 1.16 2022/03/31 17:27:23 naddy Exp $
.\"	$NetBSD: sleep.9,v 1.11 1999/03/24 06:15:12 mycroft Exp $
.\"
.\" Copyright (c) 1996 The NetBSD Foundation, Inc.
.\" All rights reserved.
.\"
.\" This code is derived from software contributed to The NetBSD Foundation
.\" by Paul Kranenburg.
.\"
.\" 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 $Mdocdate: March 31 2022 $
.Dt TSLEEP 9
.Os
.Sh NAME
.Nm tsleep ,
.Nm tsleep_nsec ,
.Nm msleep ,
.Nm msleep_nsec ,
.Nm rwsleep ,
.Nm rwsleep_nsec ,
.Nm wakeup ,
.Nm wakeup_n ,
.Nm wakeup_one
.Nd process context sleep and wakeup
.Sh SYNOPSIS
.In sys/param.h
.In sys/systm.h
.Fd #define INFSLP	UINT64_MAX
.Fd #define MAXTSLP	(UINT64_MAX - 1)
.Ft int
.Fo tsleep
.Fa "void *ident"
.Fa "int priority"
.Fa "const char *wmesg"
.Fa "int timo"
.Fc
.Ft int
.Fo tsleep_nsec
.Fa "void *ident"
.Fa "int priority"
.Fa "const char *wmesg"
.Fa "uint64_t nsecs"
.Fc
.Ft int
.Fo msleep
.Fa "void *ident"
.Fa "struct mutex *mtx"
.Fa "int priority"
.Fa "const char *wmesg"
.Fa "int timo"
.Fc
.Ft int
.Fo msleep_nsec
.Fa "void *ident"
.Fa "struct mutex *mtx"
.Fa "int priority"
.Fa "const char *wmesg"
.Fa "uint64_t nsecs"
.Fc
.Ft int
.Fo rwsleep
.Fa "void *ident"
.Fa "struct rwlock *rwl"
.Fa "int priority"
.Fa "const char *wmesg"
.Fa "int timo"
.Fc
.Ft int
.Fo rwsleep_nsec
.Fa "void *ident"
.Fa "struct rwlock *rwl"
.Fa "int priority"
.Fa "const char *wmesg"
.Fa "uint64_t nsecs"
.Fc
.Ft void
.Fn wakeup "void *ident"
.Ft void
.Fn wakeup_n "void *ident" "int count"
.Ft void
.Fn wakeup_one "void *ident"
.Sh DESCRIPTION
These functions implement voluntary context switching.
.Fn tsleep ,
.Fn msleep
and
.Fn rwsleep
are used throughout the kernel whenever processing in the current context
cannot continue for any of the following reasons:
.Bl -bullet -offset indent
.It
The current process needs to await the results of a pending I/O operation.
.It
The current process needs resources
.Pq e.g. memory
which are temporarily unavailable.
.It
The current process wants access to data structures which are locked by
other processes.
.El
.Pp
The
.Fn wakeup ,
.Fn wakeup_n ,
and
.Fn wakeup_one
functions are used to notify sleeping processes of possible changes to the
condition that caused them to go to sleep.
Typically, an awakened process will -- after it has acquired a context
again -- retry the action that blocked its operation to see if the
.Dq blocking
condition has cleared.
.Pp
The
.Fn tsleep
function takes the following arguments:
.Bl -tag -width priority
.It Fa ident
An identifier of the
.Dq wait channel
representing the resource for which the current process needs to wait.
This typically is the virtual address of some kernel data structure related
to the resource for which the process is contending.
The same identifier must be used in a call to
.Fn wakeup
to get the process going again.
.Fa ident
should not be
.Dv NULL .
.It Fa priority
The process priority to be used when the process is awakened and put on
the queue of runnable processes.
This mechanism is used to optimize
.Dq throughput
of processes executing in kernel mode.
If the flag
.Dv PCATCH
is OR'ed into
.Fa priority ,
the process checks for posted signals before and after sleeping.
.It Fa wmesg
A pointer to a character string indicating the reason a process is sleeping.
The kernel does not use the string, but makes it available
.Pq through the process structure field Li p_wmesg
for user level utilities such as
.Xr ps 1 .
.It Fa timo
If non-zero, the process will sleep for at most
.Li timo/hz
seconds.
If this amount of time elapses and no
.Fn wakeup "ident"
has occurred, and no signal
.Pq if Dv PCATCH No was set
was posted,
.Fn tsleep
will return
.Er EWOULDBLOCK .
.El
.Pp
The
.Fn msleep
function behaves just like
.Fn tsleep ,
but takes an additional argument:
.Bl -tag -width priority
.It Fa mtx
A mutex that will be unlocked when the process is safely
on the sleep queue.
The mutex will be relocked at the end of msleep unless the
.Dv PNORELOCK
flag is set in the
.Fa priority
argument.
.El
.Pp
The
.Fn rwsleep
function behaves just like
.Fn tsleep ,
but takes an additional argument:
.Bl -tag -width priority
.It Fa rwl
A read- or write-lock that will be unlocked when the process is safely
on the sleep queue.
The lock will be relocked at the end of rwsleep unless the
.Dv PNORELOCK
flag is set in the
.Fa priority
argument.
.El
.Pp
The
.Fn tsleep_nsec ,
.Fn msleep_nsec ,
and
.Fn rwsleep_nsec
functions behave like their unsuffixed counterparts except that they
accept a timeout in terms of nanoseconds.
These functions will always sleep for at least one tick,
even if
.Fa nsecs
is zero.
If
.Fa nsecs
is equal to
.Dv INFSLP
these functions do not time out,
otherwise they sleep for at least
.Fa nsecs
nanoseconds.
.Pp
The
.Fn wakeup
function will mark all processes which are currently sleeping on the identifier
.Fa ident
as runnable.
Eventually, each of the processes will resume execution in the kernel
context, causing a return from
.Fn tsleep .
Note that processes returning from sleep should always re-evaluate the
conditions that blocked them, since a call to
.Fn wakeup
merely signals a
.Em possible
change to the blocking conditions.
For example, when two or more processes are waiting for an exclusive lock,
only one of them will succeed in acquiring the lock when it is released.
All others will have to go back to sleep and wait for the next opportunity.
.Pp
The
.Fn wakeup_n
and
.Fn wakeup_one
functions behave similarly to
.Fn wakeup
except that only
.Fa count
or one process, respectively, is marked runnable.
.Sh RETURN VALUES
.Fn tsleep ,
.Fn tsleep_nsec ,
.Fn msleep ,
.Fn msleep_nsec ,
.Fn rwsleep ,
and
.Fn rwsleep_nsec
return 0 if they return as a result of a
.Fn wakeup .
If they return as a result of a signal, the return value is
.Er ERESTART
if the signal has the
.Dv SA_RESTART
property
.Pq see Xr sigaction 2 ,
and
.Er EINTR
otherwise.
If they return as a result of a timeout, the return value is
.Er EWOULDBLOCK .
.Sh CODE REFERENCES
These functions are implemented in the file
.Pa sys/kern/kern_synch.c .
.Sh SEE ALSO
.Xr hz 9 ,
.Xr mi_switch 9 ,
.Xr timeout 9