$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 20 2020 $ .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 n sys/param.h n 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: l -bullet -offset indent t The current process needs to await the results of a pending I/O operation. t The current process needs resources
q e.g. memory which are temporarily unavailable. t The current process wants access to data structures which are locked by other processes. .El
p 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.
p The .Fn tsleep function takes the following arguments: l -tag -width priority t 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 . t 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. t 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
q through the process structure field Li p_wmesg for user level utilities such as .Xr ps 1 . t 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
q if Dv PCATCH No was set was posted, .Fn tsleep will return .Er EWOULDBLOCK . .El
p The .Fn msleep function behaves just like .Fn tsleep , but takes an additional argument: l -tag -width priority t 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
p The .Fn rwsleep function behaves just like .Fn tsleep , but takes an additional argument: l -tag -width priority t 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
p 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.
p 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.
p 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
q 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
a sys/kern/kern_synch.c . .Sh SEE ALSO .Xr hz 9 , .Xr mi_switch 9 , .Xr timeout 9