Copyright (c) 1996 Joerg Wunsch

All rights reserved.

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 DEVELOPERS ``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 DEVELOPERS 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.

$Id: sleep.9,v 1.10 1998/01/16 18:12:57 bde Exp $
"
.Dd December 17, 1998 .Os .Dt SLEEP 9 .Sh NAME .Nm sleep , .Nm tsleep , .Nm asleep , .Nm await , .Nm wakeup .Nd wait for events .Sh SYNOPSIS .Fd #include <sys/param.h> .Fd #include <sys/systm.h> .Fd #include <sys/proc.h> .Ft int .Fn tsleep "void *ident" "int priority" "const char *wmesg" "int timo" .Ft int .Fn asleep "void *ident" "int priority" "const char *wmesg" "int timo" .Ft int .Fn await "int priority" "int timo" .Ft void .Fn wakeup "void *ident" .Ft void .Fn wakeup_one "void *ident" .Sh DESCRIPTION The functions .Fn tsleep and .Fn wakeup handle event-based process blocking. If a process must wait for an external event, it is put on sleep by .Nm tsleep . The parameter .Ar ident is an arbitrary address that uniquely identifies the event on which the process is being asleep. All processes sleeping on a single .Ar ident are woken up later by .Nm wakeup , often called from inside an interrupt routine, to indicate that the resource the process was blocking on is available now.

p The parameter .Ar wmesg is a string describing the sleep condition for tools like .Xr ps 1 . Due to the limited space of those programs to display arbitrary strings, this message should not be longer than 6 characters.

p The .Fn wakeup_one function is used to make the first process in the queue that is sleeping on the parameter .Fa ident runnable. This can prevent the system from becoming saturated when a large number of processes are sleeping on the same address, but only one of them can actually do any useful work when made runnable.

p .Nm Tsleep is the general sleep call. Suspends the current process until a wakeup is performed on the specified identifier. The process will then be made runnable with the specified .Ar priority . Sleeps at most .Ar timo / hz seconds (0 means no timeout). If .Ar pri includes the .Dv PCATCH flag, signals are checked before and after sleeping, else signals are not checked. Returns 0 if awakened, .Dv EWOULDBLOCK if the timeout expires. If .Dv PCATCH is set and a signal needs to be delivered, .Dv ERESTART is returned if the current system call should be restarted if possible, and .Dv EINTR is returned if the system call should be interrupted by the signal

q return Dv EINTR .

p .Nm Sleep is the traditional form. It doesn't let you specify a timeout nor a .Ar wmesg , hence its use is deprecated.

p .Nm Asleep implements the new asynchronous sleep function. It takes the same arguments as .Fn tsleep and places the process on the appropriate wait queue, but .Fn asleep leaves the process runnable and returns immediately. The caller is then expected to, at some point in the future, call .Fn await to actually wait for the previously queued wait condition. If .Fn asleep is called several times, only the most recent call is effective. .Fn asleep may be called with an .Ar ident value of NULL to remove any previously queued condition.

p .Nm Await implements the new asynchronous wait function. If you .Fn asleep on an identifier, .Fn await will actually block the process until someone calls .Fn wakeup on that identifier. If someone calls .Fn wakeup after you .Fn asleep but before you .Fn await then the .Fn await call is effectively a NOP. If .Fn await is called multiple times without an intervening .Fn asleep the .Fn await is effective a NOP, but will call .Fn mswitch for safety. The .Fn await function allows you to override the priority and timeout values to be used. If the value -1 is specified for an argument, the value is taken from the previous .Fn asleep call. If you pass -1 for the priority you must be prepared to catch signal conditions if the prior call to .Fn asleep specified it in its priority. If you pass -1 for the timeout you must be prepared to catch a timeout condition if the prior call to .Fn asleep specified a timeout. When you use -1, you should generally not make assumptions as to the arguments used by the prior .Fn asleep call.

p The .Fn asleep and .Fn await functions are used by the kernel code for various purposes but the main one is to allow complex interlocking code to 'backout' of a temporary resource failure (such as lack of memory or trying to access a block that is not in the buffer cache) in order to release major locks prior to blocking, and to then retry the call that failed on wakeup. This involves subroutines deep in the kernel calling .Fn asleep and returning a temporary failure, then popping back up through a number of call levels before calling .Fn await , then retrying. The kernel might also use these functions to avoid using spinlocks in a check-condition interlock. That is, in case the case where the kernel wishes to check the condition of something and then block on it. To avoid the race between the check and the blocking, the kernel can first check the condition, then call .Fn asleep , then check the condition a second time before calling .Fn await . The overlap makes the race condition impossible. .Sh RETURN VALUES See above. .Sh SEE ALSO .Xr ps 1 .Sh HISTORY The sleep/wakeup process synchronization mechanism is very old. It appeared in a very early version of Unix.

p .Nm Tsleep appeared in x 4.4 . .Sh AUTHORS This man page has been written by