xref: /freebsd/share/man/man9/unr.9 (revision fa9896e082a1046ff4fbc75fcba4d18d1f2efc19)

.\" Copyright (c) 2005 Gleb Smirnoff <glebius@FreeBSD.org>
.\" 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 AUTHOR 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 AUTHOR 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 April 21, 2022
.Dt UNR 9
.Os
.Sh NAME
.Nm new_unrhdr ,
.Nm clean_unrhdr ,
.Nm clear_unrhdr ,
.Nm delete_unrhdr ,
.Nm alloc_unr ,
.Nm alloc_unr_specific ,
.Nm free_unr ,
.Nm create_iter_unr ,
.Nm next_iter_unr ,
.Nm free_iter_unr
.Nd "kernel unit number allocator"
.Sh SYNOPSIS
.In sys/systm.h
.Ft "struct unrhdr *"
.Fn new_unrhdr "int low" "int high" "struct mtx *mutex"
.Ft void
.Fn clean_unrhdr "struct unrhdr *uh"
.Ft void
.Fn clean_unrhdrl "struct unrhdr *uh"
.Ft void
.Fn clear_unrhdr "struct unrhdr *uh"
.Ft void
.Fn delete_unrhdr "struct unrhdr *uh"
.Ft int
.Fn alloc_unr "struct unrhdr *uh"
.Ft int
.Fn alloc_unrl "struct unrhdr *uh"
.Ft int
.Fn alloc_unr_specific "struct unrhdr *uh" "u_int item"
.Ft void
.Fn free_unr "struct unrhdr *uh" "u_int item"
.Ft void *
.Fn create_iter_unr "struct unrhdr *uh"
.Ft int
.Fn next_iter_unr "void *handle"
.Ft void
.Fn free_iter_unr "void *handle"
.Sh DESCRIPTION
The kernel unit number allocator is a generic facility, which allows to allocate
unit numbers within a specified range.
.Bl -tag -width indent
.It Fn new_unrhdr low high mutex
Initialize a new unit number allocator entity.
The
.Fa low
and
.Fa high
arguments
specify minimum and maximum number of unit numbers.
There is no cost associated with the range of unit numbers, so unless the resource
really is finite,
.Dv INT_MAX
can be used.
If
.Fa mutex
is not
.Dv NULL ,
it is used for locking when allocating and freeing units.
If the passed value is the token
.Va UNR_NO_MTX ,
then no locking is applied internally.
Otherwise, internal mutex is used.
.It Fn clear_unrhdr uh
Clear all units from the specified unit number allocator entity.
This function resets the entity as if it were just initialized with
.Fn new_unrhdr .
.It Fn delete_unrhdr uh
Delete specified unit number allocator entity.
This function frees the memory associated with the entity, it does not free
any units.
To free all units use
.Fn clear_unrhdr .
.It Fn clean_unrhdr uh
Freeing unit numbers might result in some internal memory becoming unused.
There are
.Nm unit
allocator consumers that cannot tolerate taking
.Xr malloc 9
locks to free the memory, while having their unit mutex locked.
For this reason, free of the unused memory after delete is postponed
until the consumer can afford calling into the
.Xr malloc 9
subsystem.
Call
.Fn clean_unrhdr uh
to do the cleanup.
In particular, this needs to be done before freeing a unr, if
a deletion of units could have been performed.
.It Fn clean_unrhdrl
Same as
.Fn clean_unrhdr ,
but assumes that the unr mutex is already owned, if any.
.It Fn alloc_unr uh
Return a new unit number.
The lowest free number is always allocated.
This function does not allocate memory and never sleeps, however it may
block on a mutex.
If no free unit numbers are left,
.Li \-1
is returned.
.It Fn alloc_unrl uh
Same as
.Fn alloc_unr
except that mutex is assumed to be already locked and thus is not used.
.It Fn alloc_unr_specific uh item
Allocate a specific unit number.
This function allocates memory and thus may sleep.
The allocated unit number is returned on success.
If the specified number is already allocated or out of the range,
.Li \-1
is returned.
.It Fn free_unr uh item
Free a previously allocated unit number.
This function may require allocating memory, and thus it can sleep.
There is no pre-locked variant.
.El
.Sh ITERATOR INTERFACE
The
.Nm unr
facility provides an interface to iterate over all allocated units
for the given
.Dv unrhdr .
Iterators are identified by an opaque handle.
More than one iterators can operate simultaneously; the iterator position
data is recorded only in the iterator handle.
.Pp
Consumers must ensure that the unit allocator is not modified between
calls to the iterator functions.
In particular, the internal allocator mutex cannot provide consistency,
because it is acquired and dropped inside the
.Fn next_iter_unr
function.
If the allocator was modified, it is safe to free the iterator with
.Fn free_iter_unr
method nevertheless.
.Bl -tag -width indent
.It Fn create_iter_unr uh
Create an iterator.
Return the handle that should be passed to other iterator functions.
.It Fn next_iter_unr handle
Return the value of the next unit.
Units are returned in ascending order.
A return value of
.Li \-1
indicates the end of iteration, in which
case
.Li \-1
is returned for all future calls.
.It Fn free_iter_unr handle
Free the iterator, handle is no longer valid.
.El
.Sh CODE REFERENCES
The above functions are implemented in
.Pa sys/kern/subr_unit.c .
.Sh HISTORY
Kernel unit number allocator first appeared in
.Fx 6.0 .
.Sh AUTHORS
.An -nosplit
Kernel unit number allocator was written by
.An Poul-Henning Kamp .
This manpage was written by
.An Gleb Smirnoff .