xref: /freebsd/share/man/man9/kasan.9 (revision 2e3f49888ec8851bafb22011533217487764fdb0)
1.\"-
2.\" Copyright (c) 2021 The FreeBSD Foundation
3.\"
4.\" This documentation was written by Mark Johnston under sponsorship from
5.\" the FreeBSD Foundation.
6.\"
7.\" Redistribution and use in source and binary forms, with or without
8.\" modification, are permitted provided that the following conditions
9.\" are met:
10.\" 1. Redistributions of source code must retain the above copyright
11.\"    notice, this list of conditions and the following disclaimer.
12.\" 2. Redistributions in binary form must reproduce the above copyright
13.\"    notice, this list of conditions and the following disclaimer in the
14.\"    documentation and/or other materials provided with the distribution.
15.\"
16.\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR 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 AUTHOR 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.Dd October 13, 2023
29.Dt KASAN 9
30.Os
31.Sh NAME
32.Nm KASAN
33.Nd Kernel Address SANitizer
34.Sh SYNOPSIS
35The
36.Pa GENERIC-KASAN
37kernel configuration can be used to compile a KASAN-enabled kernel using
38.Pa GENERIC
39as a base configuration.
40Alternately, to compile KASAN into the kernel, place the following line in your
41kernel configuration file:
42.Bd -ragged -offset indent
43.Cd "options KASAN"
44.Ed
45.Pp
46.In sys/asan.h
47.Ft void
48.Fn kasan_mark "const void *addr" "size_t size" "size_t redzsize" "uint8_t code"
49.Sh DESCRIPTION
50.Nm
51is a subsystem which leverages compiler instrumentation to detect invalid
52memory accesses in the kernel.
53Currently it is implemented on the amd64 and arm64 platforms.
54.Pp
55When
56.Nm
57is compiled into the kernel, the compiler is configured to emit function
58calls upon every memory access.
59The functions are implemented by
60.Nm
61and permit run-time detection of several types of bugs including
62use-after-frees, double frees and frees of invalid pointers, and out-of-bounds
63accesses.
64These protections apply to memory allocated by
65.Xr uma 9 ,
66.Xr malloc 9
67and related functions, and
68.Fn kmem_malloc
69and related functions,
70as well as global variables and kernel stacks.
71.Nm
72is conservative and will not detect all instances of these types of bugs.
73Memory accesses through the kernel map are sanitized, but accesses via the
74direct map are not.
75When
76.Nm
77is configured, the kernel aims to minimize its use of the direct map.
78.Sh IMPLEMENTATION NOTES
79.Nm
80is implemented using compiler instrumentation and a kernel runtime.
81When a
82kernel is built with the KASAN option enabled, the compiler inserts function calls
83before most memory accesses in the generated code.
84The runtime implements the corresponding functions, which decide whether a
85given access is valid.
86If not, the runtime prints a warning or panics the kernel, depending on the
87value of the
88.Sy debug.kasan.panic_on_violation
89sysctl/tunable.
90.Pp
91The
92.Nm
93runtime in a KASAN-configured kernel can be disabled by
94setting the loader tunable
95.Sy debug.kasan.disable=1 .
96.Pp
97The
98.Nm
99runtime works by maintaining a shadow map for the kernel map.
100There exists a linear mapping between addresses in the kernel map and addresses
101in the shadow map.
102The shadow map is used to store information about the current state of
103allocations from the kernel map.
104For example, when a buffer is returned by
105.Xr malloc 9 ,
106the corresponding region of the shadow map is marked to indicate that the
107buffer is valid.
108When it is freed, the shadow map is updated to mark the buffer as invalid.
109Accesses to the buffer are intercepted by the
110.Nm
111runtime and validated using the contents of the shadow map.
112.Pp
113Upon booting, all kernel memory is marked as valid.
114Kernel allocators must mark cached but free buffers as invalid, and must mark
115them valid before freeing the kernel virtual address range.
116This slightly reduces the effectiveness of
117.Nm
118but simplifies its maintenance and integration into the kernel.
119.Pp
120Updates to the shadow map are performed by calling
121.Fn kasan_mark .
122Parameter
123.Fa addr
124is the address of the buffer whose shadow is to be updated,
125.Fa size
126is the usable size of the buffer, and
127.Fa redzsize
128is the full size of the buffer allocated from lower layers of the system.
129.Fa redzsize
130must be greater than or equal to
131.Fa size .
132In some cases kernel allocators will return a buffer larger than that requested
133by the consumer; the unused space at the end is referred to as a red zone and is
134always marked as invalid.
135.Fa code
136allows the caller to specify an identifier used when marking a buffer as invalid.
137The identifier is included in any reports generated by
138.Nm
139and helps identify the source of the invalid access.
140For instance, when an item is freed to a
141.Xr uma 9
142zone, the item is marked with
143.Dv KASAN_UMA_FREED .
144See
145.In sys/asan.h
146for the available identifiers.
147If the entire buffer is to be marked valid, i.e.,
148.Fa size
149and
150.Fa redzsize
151are equal,
152.Fa code
153should be 0.
154.Sh SEE ALSO
155.Xr build 7 ,
156.Xr KMSAN 9 ,
157.Xr malloc 9 ,
158.Xr memguard 9 ,
159.Xr redzone 9 ,
160.Xr uma 9
161.Sh HISTORY
162.Nm
163was ported from
164.Nx
165and first appeared in
166.Fx 13.1 .
167.Sh BUGS
168Accesses to kernel memory outside of the kernel map are ignored by the
169.Nm
170runtime.
171When
172.Nm
173is configured, the kernel memory allocators are configured to use the kernel
174map, but some uses of the direct map remain.
175For example, on amd64 and arm64, accesses to page table pages are not tracked.
176.Pp
177Some kernel memory allocators explicitly permit accesses after an object has
178been freed.
179These cannot be sanitized by
180.Nm .
181For example, memory from all
182.Xr uma 9
183zones initialized with the
184.Dv UMA_ZONE_NOFREE
185flag are not sanitized.
186