2 .\" The Regents of the University of California. All rights reserved.
4 .\" Redistribution and use in source and binary forms, with or without
5 .\" modification, are permitted provided that the following conditions
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. All advertising materials mentioning features or use of this software
13 .\" must display the following acknowledgement:
14 .\" This product includes software developed by the University of
15 .\" California, Berkeley and its contributors.
16 .\" 4. Neither the name of the University nor the names of its contributors
17 .\" may be used to endorse or promote products derived from this software
18 .\" without specific prior written permission.
20 .\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
21 .\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
22 .\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
23 .\" ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
24 .\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
25 .\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
26 .\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
27 .\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
28 .\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
29 .\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
32 .\" $FreeBSD: src/share/man/man9/buf.9,v 1.5.2.5 2001/12/17 11:30:18 ru Exp $
33 .\" $DragonFly: src/share/man/man9/buf.9,v 1.8 2007/09/13 10:55:55 swildner Exp $
40 .Nd "kernel buffer I/O scheme used in DragonFly VM system"
42 The kernel implements a KVM abstraction of the buffer cache which allows it
43 to map potentially disparate vm_page's into contiguous KVM for use by
44 (mainly filesystem) devices and device I/O.
45 This abstraction supports block sizes from
47 (usually 512) to upwards of several pages or more.
48 It also supports a relatively primitive byte-granular valid range and dirty
49 range currently hardcoded for use by NFS.
50 The code implementing the VM Buffer abstraction is mostly concentrated in
51 .Pa /usr/src/sys/kern/vfs_bio.c
53 .Pa /usr/src/sys/sys/buf.h .
55 One of the most important things to remember when dealing with buffer pointers
57 is that the underlying pages are mapped directly from the buffer cache.
58 No data copying occurs in the scheme proper, though some filesystems
59 such as UFS do have to copy a little when dealing with file fragments.
60 The second most important thing to remember is that due to the underlying page
63 base pointer in a buf is always
68 When you have a VM buffer representing some
72 the actual start of the buffer is
73 .Fa ( b_data + ( Fa b_offset & Dv PAGE_MASK ) )
76 Finally, the VM system's core buffer cache supports valid and dirty bits
77 .Fa ( m->valid , m->dirty )
80 Thus a platform with a hardware page size of 4096 bytes has 8 valid and 8
82 These bits are generally set and cleared in groups based on the device
83 block size of the device backing the page.
84 Complete page's worth are often referred to using the
86 bitmask (i.e. 0xFF if the hardware page size is 4096).
88 VM buffers also keep track of a byte-granular dirty range and valid range.
89 This feature is normally only used by the NFS subsystem.
90 I'm not sure why it is used at all, actually, since we have
92 valid/dirty granularity within the VM buffer.
93 If a buffer dirty operation creates a
95 the dirty range will extend to cover the hole.
96 If a buffer validation operation creates a
98 the byte-granular valid range is left alone and will not take into account
100 Thus the whole byte-granular abstraction is considered a bad hack and it
101 would be nice if we could get rid of it completely.
103 A VM buffer is capable of mapping the underlying VM cache pages into KVM in
104 order to allow the kernel to directly manipulate the data associated with
106 .Ft ( vnode , Fa b_offset , Fa b_size ) .
107 The kernel typically unmaps VM buffers the moment they are no longer needed
110 structure instantiated and even
112 array instantiated despite having unmapped them from KVM.
113 If a page making up a VM buffer is about to undergo I/O, the system typically
114 unmaps it from KVM and replaces the page in the
116 array with a placemarker called
118 The placemarker forces any kernel subsystems referencing the associated
120 to re-lookup the associated page.
121 I believe the placemarker hack is used to allow sophisticated devices
122 such as filesystem devices to remap underlying pages in order to deal with,
123 for example, remapping a file fragment into a file block.
125 VM buffers are used to track I/O operations within the kernel.
126 Unfortunately, the I/O implementation is also somewhat of a hack because
127 the kernel wants to clear the dirty bit on the underlying pages the moment
128 it queues the I/O to the VFS device, not when the physical I/O is actually
130 This can create confusion within filesystem devices that use delayed-writes
131 because you wind up with pages marked clean that are actually still dirty.
132 If not treated carefully, these pages could be thrown away!
133 Indeed, a number of serious bugs related to this hack were not fixed until
137 The kernel uses an instantiated VM buffer (i.e.
139 to placemark pages in this special state.
140 The buffer is typically flagged
142 When a device no longer needs a buffer it typically flags it as
144 Due to the underlying pages being marked clean, the
145 .Dv B_DELWRI | B_RELBUF
146 combination must be interpreted to mean that the buffer is still actually
147 dirty and must be written to its backing store before it can actually be
151 is not set, the underlying dirty pages are still properly marked as dirty
152 and the buffer can be completely freed without losing that clean/dirty state
154 .\"( XXX do we have to check other flags in regards to this situation ??? ).
156 The kernel reserves a portion of its KVM space to hold VM Buffer's data
158 Even though this is virtual space (since the buffers are mapped from the
159 buffer cache), we cannot make it arbitrarily large because instantiated
161 .Ft ( struct buf Ap s )
162 prevent their underlying pages in the buffer cache from being freed.
163 This can complicate the life of the paging system.
169 manual page was originally written by
171 and first appeared in