xref: /dragonfly/lib/libc/sys/madvise.2 (revision ef2b2b9d)
1.\" Copyright (c) 1991, 1993
2.\"	The Regents of the University of California.  All rights reserved.
3.\"
4.\" Redistribution and use in source and binary forms, with or without
5.\" modification, are permitted provided that the following conditions
6.\" are met:
7.\" 1. Redistributions of source code must retain the above copyright
8.\"    notice, this list of conditions and the following disclaimer.
9.\" 2. Redistributions in binary form must reproduce the above copyright
10.\"    notice, this list of conditions and the following disclaimer in the
11.\"    documentation and/or other materials provided with the distribution.
12.\" 3. Neither the name of the University nor the names of its contributors
13.\"    may be used to endorse or promote products derived from this software
14.\"    without specific prior written permission.
15.\"
16.\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
17.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
18.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
19.\" ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
20.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
21.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
22.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
23.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
24.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
25.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
26.\" SUCH DAMAGE.
27.\"
28.\"	@(#)madvise.2	8.1 (Berkeley) 6/9/93
29.\" $FreeBSD: src/lib/libc/sys/madvise.2,v 1.17.2.8 2003/01/06 23:33:59 trhodes Exp $
30.\"
31.Dd May 4, 2019
32.Dt MADVISE 2
33.Os
34.Sh NAME
35.Nm madvise ,
36.Nm posix_madvise ,
37.Nm mcontrol
38.Nd give advice about use of memory
39.Sh LIBRARY
40.Lb libc
41.Sh SYNOPSIS
42.In sys/types.h
43.In sys/mman.h
44.Ft int
45.Fn madvise "void *addr" "size_t len" "int behav"
46.Ft int
47.Fn posix_madvise "void *addr" "size_t len" "int behav"
48.Ft int
49.Fn mcontrol "void *addr" "size_t len" "int behav" "off_t value"
50.Sh DESCRIPTION
51The
52.Fn madvise
53system call
54allows a process that has knowledge of its memory behavior
55to describe it to the system.
56The
57.Fn posix_madvise
58interface is identical and is provided for standards conformance.
59The
60.Fn mcontrol
61system call is an extension of
62.Fn madvise
63that takes an additional
64.Fa value
65argument (see the description of the
66.Dv MADV_SETMAP
67behavior below).
68.Pp
69The known behaviors are:
70.Bl -tag -width ".Dv MADV_SEQUENTIAL"
71.It Dv MADV_NORMAL
72Tells the system to revert to the default paging
73behavior.
74.It Dv MADV_RANDOM
75Is a hint that pages will be accessed randomly, and prefetching
76is likely not advantageous.
77.It Dv MADV_SEQUENTIAL
78Causes the VM system to depress the priority of
79pages immediately preceding a given page when it is faulted in.
80.It Dv MADV_WILLNEED
81Causes pages that are in a given virtual address range
82to temporarily have higher priority, and if they are in
83memory, decrease the likelihood of them being freed.  Additionally,
84the pages that are already in memory will be immediately mapped into
85the process, thereby eliminating unnecessary overhead of going through
86the entire process of faulting the pages in.  This WILL NOT fault
87pages in from backing store, but quickly map the pages already in memory
88into the calling process.
89.It Dv MADV_DONTNEED
90Allows the VM system to decrease the in-memory priority
91of pages in the specified range.  Additionally future references to
92this address range will incur a page fault.
93.It Dv MADV_FREE
94Gives the VM system the freedom to free pages,
95and tells the system that information in the specified page range
96is no longer important.  This is an efficient way of allowing
97.Xr malloc 3
98to free pages anywhere in the address space, while keeping the address space
99valid.  The next time that the page is referenced, the page might be demand
100zeroed, or might contain the data that was there before the
101.Dv MADV_FREE
102call.
103References made to that address space range will not make the VM system
104page the information back in from backing store until the page is
105modified again.
106.It Dv MADV_NOSYNC
107Request that the system not flush the data associated with this map to
108physical backing store unless it needs to.  Typically this prevents the
109filesystem update daemon from gratuitously writing pages dirtied
110by the VM system to physical disk.  Note that VM/filesystem coherency is
111always maintained, this feature simply ensures that the mapped data is
112only flush when it needs to be, usually by the system pager.
113.Pp
114This feature is typically used when you want to use a file-backed shared
115memory area to communicate between processes (IPC) and do not particularly
116need the data being stored in that area to be physically written to disk.
117With this feature you get the equivalent performance with mmap that you
118would expect to get with SysV shared memory calls, but in a more controllable
119and less restrictive manner.  However, note that this feature is not portable
120across
121.Ux
122platforms (though some may do the right thing by default).
123For more information see the
124.Dv MAP_NOSYNC
125section of
126.Xr mmap 2
127.It Dv MADV_AUTOSYNC
128Undoes the effects of
129.Dv MADV_NOSYNC
130for any future pages dirtied within the
131address range.  The effect on pages already dirtied is indeterminate - they
132may or may not be reverted.  You can guarantee reversion by using the
133.Xr msync 2
134or
135.Xr fsync 2
136system calls.
137.It Dv MADV_NOCORE
138Region is not included in a core file.
139.It Dv MADV_CORE
140Include region in a core file.
141.It Dv MADV_INVAL
142Invalidate the hardware page table for a region of memory, forcing
143accesses to re-fault the pages.
144This command is primarily meant to be used in areas of memory
145governed by a virtual page table after modifications have been made
146to it.
147.It Dv MADV_SETMAP
148Set the offset of the page directory page to
149.Fa value
150for the virtual page table governing
151the specified area of memory.  The entire memory area under virtual page table
152management should be specified.  You may encounter unexpected effects
153if you only set the page directory page for part of the mapping.
154.El
155.Pp
156Portable programs that call the
157.Fn posix_madvise
158interface should use the aliases
159.Dv POSIX_MADV_NORMAL , POSIX_MADV_SEQUENTIAL ,
160.Dv POSIX_MADV_RANDOM , POSIX_MADV_WILLNEED ,
161and
162.Dv POSIX_MADV_DONTNEED
163rather than the flags described above.
164.Sh RETURN VALUES
165.Rv -std madvise posix_madvise mcontrol
166.Sh ERRORS
167The
168.Fn madvise ,
169.Fn posix_madvise ,
170and
171.Fn mcontrol
172functions will fail if:
173.Bl -tag -width Er
174.It Bq Er EINVAL
175The
176.Fa behav
177argument is not valid or the virtual address range specified by the
178.Fa addr
179and
180.Fa len
181arguments is not valid.
182.El
183.Sh SEE ALSO
184.Xr mincore 2 ,
185.Xr mprotect 2 ,
186.Xr msync 2 ,
187.Xr munmap 2 ,
188.Xr posix_fadvise 3
189.Sh STANDARDS
190The
191.Fn posix_madvise
192interface conforms to
193.St -p1003.1-2001 .
194.Sh HISTORY
195The
196.Fn madvise
197function first appeared in
198.Bx 4.4 .
199The
200.Fn mcontrol
201function was added in
202.Dx 1.7 .
203