xref: /freebsd/share/man/man9/pmap_enter.9 (revision 5ca8e32633c4ffbbcd6762e5888b6a4ba0708c6c)
1.\"
2.\" Copyright (c) 2003 Bruce M Simpson <bms@spc.org>
3.\" Copyright (c) 2014 The FreeBSD Foundation
4.\" All rights reserved.
5.\"
6.\" Redistribution and use in source and binary forms, with or without
7.\" modification, are permitted provided that the following conditions
8.\" are met:
9.\" 1. Redistributions of source code must retain the above copyright
10.\"    notice, this list of conditions and the following disclaimer.
11.\" 2. Redistributions in binary form must reproduce the above copyright
12.\"    notice, this list of conditions and the following disclaimer in the
13.\"    documentation and/or other materials provided with the distribution.
14.\"
15.\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
16.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
17.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
18.\" ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
19.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
20.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
21.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
22.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
23.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
24.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
25.\" SUCH DAMAGE.
26.\"
27.Dd December 16, 2018
28.Dt PMAP_ENTER 9
29.Os
30.Sh NAME
31.Nm pmap_enter
32.Nd insert a virtual page into a physical map
33.Sh SYNOPSIS
34.In sys/param.h
35.In vm/vm.h
36.In vm/pmap.h
37.Ft int
38.Fo pmap_enter
39.Fa "pmap_t pmap" "vm_offset_t va" "vm_page_t m" "vm_prot_t prot"
40.Fa "u_int flags" "int8_t psind"
41.Fc
42.Sh DESCRIPTION
43The
44.Fn pmap_enter
45function creates a mapping in the physical map
46.Fa pmap
47from the virtual address
48.Fa va
49to the physical page
50.Fa m
51with the protection
52.Fa prot .
53Any previous mapping at the virtual address
54.Fa va
55is destroyed.
56.Pp
57The
58.Fa flags
59argument may have the following values:
60.Bl -tag -width ".Dv PMAP_ENTER_NOSLEEP"
61.It Dv VM_PROT_READ
62A read access to the given virtual address triggered the call.
63.It Dv VM_PROT_WRITE
64A write access to the given virtual address triggered the call.
65.It Dv VM_PROT_EXECUTE
66An execute access to the given virtual address triggered the call.
67.It Dv PMAP_ENTER_WIRED
68The mapping should be marked as wired.
69.It Dv PMAP_ENTER_NOSLEEP
70This function may not sleep during creation of the mapping.
71If the mapping cannot be created without sleeping, an appropriate
72Mach VM error is returned.
73.El
74If the
75.Dv PMAP_ENTER_NOSLEEP
76flag is not specified, this function must create the requested mapping
77before returning.
78It may not fail.
79In order to create the requested mapping, this function may destroy
80any non-wired mapping in any pmap.
81.Pp
82The
83.Fa psind
84parameter specifies the page size that should be used by the mapping.
85The supported page sizes are described by the global array
86.Dv pagesizes[] .
87The desired page size is specified by passing the index of the array
88element that equals the desired page size.
89.Pp
90When the
91.Fn pmap_enter
92function destroys or updates a managed mapping, including an existing
93mapping at virtual address
94.Fa va ,
95it updates the
96.Ft vm_page
97structure corresponding to the previously mapped physical page.
98If the physical page was accessed through the managed mapping,
99then the
100.Ft vm_page
101structure's
102.Dv PGA_REFERENCED
103aflag is set.
104If the physical page was modified through the managed mapping, then the
105.Fn vm_page_dirty
106function is called on the
107.Ft vm_page
108structure.
109.Pp
110The
111.Dv PGA_WRITEABLE
112aflag must be set for the page
113.Fa m
114if the new mapping is managed and writeable.
115It is advised to clear
116.Dv PGA_WRITEABLE
117for destroyed mappings if the implementation can ensure
118that no other writeable managed mappings for the previously
119mapped pages exist.
120.Pp
121If the request modifies an existing mapping to use a different physical
122page, an implementation of
123.Nm
124must invalidate the previous mapping before installing the new one.
125This ensures that all threads sharing the pmap keep a consistent
126view of the mapping, which is necessary for the correct handling
127of CoW (copy on write) faults.
128.Pp
129If the page
130.Fa m
131is managed, the page must be busied by the caller
132or the owning object must be locked.
133In the later case, the
134.Dv PMAP_ENTER_NOSLEEP
135must be specified by the caller.
136.Pp
137The
138.Fn pmap_enter
139function must handle the multiprocessor TLB consistency for the
140given address.
141.Sh NOTES
142On arm and i386 architectures the existing implementation
143of the
144.Nm
145function is incomplete, only value 0 for
146.Fa psind
147is supported.
148Other supported architectures, except amd64, have
149.Dv pagesizes[]
150array of size 1.
151.Sh RETURN VALUES
152If successful, the
153.Fn pmap_enter
154function returns
155.Er KERN_SUCCESS .
156If the
157.Dv PMAP_ENTER_NOSLEEP
158flag was specified and the resources required for the mapping cannot
159be acquired without sleeping,
160.Dv KERN_RESOURCE_SHORTAGE
161is returned.
162.Sh SEE ALSO
163.Xr pmap 9
164.Sh AUTHORS
165This manual page was first written by
166.An Bruce M Simpson Aq Mt bms@spc.org
167and then rewritten by
168.An Alan Cox Aq Mt alc@FreeBSD.org
169and
170.An Konstantin Belousov Aq Mt kib@FreeBSD.org .
171