1 .. SPDX-License-Identifier: GPL-2.0 1 .. SPDX-License-Identifier: GPL-2.0
2 !! 2 ==============
3 ==== <<
4 FUSE 3 FUSE
5 ==== !! 4 ==============
6 5
7 Definitions 6 Definitions
8 =========== 7 ===========
9 8
10 Userspace filesystem: 9 Userspace filesystem:
11 A filesystem in which data and metadata are 10 A filesystem in which data and metadata are provided by an ordinary
12 userspace process. The filesystem can be ac 11 userspace process. The filesystem can be accessed normally through
13 the kernel interface. 12 the kernel interface.
14 13
15 Filesystem daemon: 14 Filesystem daemon:
16 The process(es) providing the data and metad 15 The process(es) providing the data and metadata of the filesystem.
17 16
18 Non-privileged mount (or user mount): 17 Non-privileged mount (or user mount):
19 A userspace filesystem mounted by a non-priv 18 A userspace filesystem mounted by a non-privileged (non-root) user.
20 The filesystem daemon is running with the pr 19 The filesystem daemon is running with the privileges of the mounting
21 user. NOTE: this is not the same as mounts 20 user. NOTE: this is not the same as mounts allowed with the "user"
22 option in /etc/fstab, which is not discussed 21 option in /etc/fstab, which is not discussed here.
23 22
24 Filesystem connection: 23 Filesystem connection:
25 A connection between the filesystem daemon a 24 A connection between the filesystem daemon and the kernel. The
26 connection exists until either the daemon di 25 connection exists until either the daemon dies, or the filesystem is
27 umounted. Note that detaching (or lazy umou 26 umounted. Note that detaching (or lazy umounting) the filesystem
28 does *not* break the connection, in this cas 27 does *not* break the connection, in this case it will exist until
29 the last reference to the filesystem is rele 28 the last reference to the filesystem is released.
30 29
31 Mount owner: 30 Mount owner:
32 The user who does the mounting. 31 The user who does the mounting.
33 32
34 User: 33 User:
35 The user who is performing filesystem operat 34 The user who is performing filesystem operations.
36 35
37 What is FUSE? 36 What is FUSE?
38 ============= 37 =============
39 38
40 FUSE is a userspace filesystem framework. It 39 FUSE is a userspace filesystem framework. It consists of a kernel
41 module (fuse.ko), a userspace library (libfuse 40 module (fuse.ko), a userspace library (libfuse.*) and a mount utility
42 (fusermount). 41 (fusermount).
43 42
44 One of the most important features of FUSE is 43 One of the most important features of FUSE is allowing secure,
45 non-privileged mounts. This opens up new poss 44 non-privileged mounts. This opens up new possibilities for the use of
46 filesystems. A good example is sshfs: a secur 45 filesystems. A good example is sshfs: a secure network filesystem
47 using the sftp protocol. 46 using the sftp protocol.
48 47
49 The userspace library and utilities are availa 48 The userspace library and utilities are available from the
50 `FUSE homepage: <https://github.com/libfuse/>` !! 49 `FUSE homepage: <http://fuse.sourceforge.net/>`_
51 50
52 Filesystem type 51 Filesystem type
53 =============== 52 ===============
54 53
55 The filesystem type given to mount(2) can be o 54 The filesystem type given to mount(2) can be one of the following:
56 55
57 fuse 56 fuse
58 This is the usual way to mount a FUSE fi 57 This is the usual way to mount a FUSE filesystem. The first
59 argument of the mount system call may co 58 argument of the mount system call may contain an arbitrary string,
60 which is not interpreted by the kernel. 59 which is not interpreted by the kernel.
61 60
62 fuseblk 61 fuseblk
63 The filesystem is block device based. T 62 The filesystem is block device based. The first argument of the
64 mount system call is interpreted as the 63 mount system call is interpreted as the name of the device.
65 64
66 Mount options 65 Mount options
67 ============= 66 =============
68 67
69 fd=N 68 fd=N
70 The file descriptor to use for communication 69 The file descriptor to use for communication between the userspace
71 filesystem and the kernel. The file descrip 70 filesystem and the kernel. The file descriptor must have been
72 obtained by opening the FUSE device ('/dev/f 71 obtained by opening the FUSE device ('/dev/fuse').
73 72
74 rootmode=M 73 rootmode=M
75 The file mode of the filesystem's root in oc 74 The file mode of the filesystem's root in octal representation.
76 75
77 user_id=N 76 user_id=N
78 The numeric user id of the mount owner. 77 The numeric user id of the mount owner.
79 78
80 group_id=N 79 group_id=N
81 The numeric group id of the mount owner. 80 The numeric group id of the mount owner.
82 81
83 default_permissions 82 default_permissions
84 By default FUSE doesn't check file access pe 83 By default FUSE doesn't check file access permissions, the
85 filesystem is free to implement its access p 84 filesystem is free to implement its access policy or leave it to
86 the underlying file access mechanism (e.g. i 85 the underlying file access mechanism (e.g. in case of network
87 filesystems). This option enables permissio 86 filesystems). This option enables permission checking, restricting
88 access based on file mode. It is usually us 87 access based on file mode. It is usually useful together with the
89 'allow_other' mount option. 88 'allow_other' mount option.
90 89
91 allow_other 90 allow_other
92 This option overrides the security measure r 91 This option overrides the security measure restricting file access
93 to the user mounting the filesystem. This o 92 to the user mounting the filesystem. This option is by default only
94 allowed to root, but this restriction can be 93 allowed to root, but this restriction can be removed with a
95 (userspace) configuration option. 94 (userspace) configuration option.
96 95
97 max_read=N 96 max_read=N
98 With this option the maximum size of read op 97 With this option the maximum size of read operations can be set.
99 The default is infinite. Note that the size 98 The default is infinite. Note that the size of read requests is
100 limited anyway to 32 pages (which is 128kbyt 99 limited anyway to 32 pages (which is 128kbyte on i386).
101 100
102 blksize=N 101 blksize=N
103 Set the block size for the filesystem. The 102 Set the block size for the filesystem. The default is 512. This
104 option is only valid for 'fuseblk' type moun 103 option is only valid for 'fuseblk' type mounts.
105 104
106 Control filesystem 105 Control filesystem
107 ================== 106 ==================
108 107
109 There's a control filesystem for FUSE, which c 108 There's a control filesystem for FUSE, which can be mounted by::
110 109
111 mount -t fusectl none /sys/fs/fuse/connectio 110 mount -t fusectl none /sys/fs/fuse/connections
112 111
113 Mounting it under the '/sys/fs/fuse/connection 112 Mounting it under the '/sys/fs/fuse/connections' directory makes it
114 backwards compatible with earlier versions. 113 backwards compatible with earlier versions.
115 114
116 Under the fuse control filesystem each connect 115 Under the fuse control filesystem each connection has a directory
117 named by a unique number. 116 named by a unique number.
118 117
119 For each connection the following files exist 118 For each connection the following files exist within this directory:
120 119
121 waiting 120 waiting
122 The number of requests which are wai 121 The number of requests which are waiting to be transferred to
123 userspace or being processed by the 122 userspace or being processed by the filesystem daemon. If there is
124 no filesystem activity and 'waiting' 123 no filesystem activity and 'waiting' is non-zero, then the
125 filesystem is hung or deadlocked. 124 filesystem is hung or deadlocked.
126 125
127 abort 126 abort
128 Writing anything into this file will 127 Writing anything into this file will abort the filesystem
129 connection. This means that all wai 128 connection. This means that all waiting requests will be aborted an
130 error returned for all aborted and n 129 error returned for all aborted and new requests.
131 130
132 Only the owner of the mount may read or write 131 Only the owner of the mount may read or write these files.
133 132
134 Interrupting filesystem operations 133 Interrupting filesystem operations
135 ################################## 134 ##################################
136 135
137 If a process issuing a FUSE filesystem request 136 If a process issuing a FUSE filesystem request is interrupted, the
138 following will happen: 137 following will happen:
139 138
140 - If the request is not yet sent to userspa 139 - If the request is not yet sent to userspace AND the signal is
141 fatal (SIGKILL or unhandled fatal signal) 140 fatal (SIGKILL or unhandled fatal signal), then the request is
142 dequeued and returns immediately. 141 dequeued and returns immediately.
143 142
144 - If the request is not yet sent to userspa 143 - If the request is not yet sent to userspace AND the signal is not
145 fatal, then an interrupted flag is set fo 144 fatal, then an interrupted flag is set for the request. When
146 the request has been successfully transfe 145 the request has been successfully transferred to userspace and
147 this flag is set, an INTERRUPT request is 146 this flag is set, an INTERRUPT request is queued.
148 147
149 - If the request is already sent to userspa 148 - If the request is already sent to userspace, then an INTERRUPT
150 request is queued. 149 request is queued.
151 150
152 INTERRUPT requests take precedence over other 151 INTERRUPT requests take precedence over other requests, so the
153 userspace filesystem will receive queued INTER 152 userspace filesystem will receive queued INTERRUPTs before any others.
154 153
155 The userspace filesystem may ignore the INTERR 154 The userspace filesystem may ignore the INTERRUPT requests entirely,
156 or may honor them by sending a reply to the *o 155 or may honor them by sending a reply to the *original* request, with
157 the error set to EINTR. 156 the error set to EINTR.
158 157
159 It is also possible that there's a race betwee 158 It is also possible that there's a race between processing the
160 original request and its INTERRUPT request. T 159 original request and its INTERRUPT request. There are two possibilities:
161 160
162 1. The INTERRUPT request is processed before 161 1. The INTERRUPT request is processed before the original request is
163 processed 162 processed
164 163
165 2. The INTERRUPT request is processed after 164 2. The INTERRUPT request is processed after the original request has
166 been answered 165 been answered
167 166
168 If the filesystem cannot find the original req 167 If the filesystem cannot find the original request, it should wait for
169 some timeout and/or a number of new requests t 168 some timeout and/or a number of new requests to arrive, after which it
170 should reply to the INTERRUPT request with an 169 should reply to the INTERRUPT request with an EAGAIN error. In case
171 1) the INTERRUPT request will be requeued. In 170 1) the INTERRUPT request will be requeued. In case 2) the INTERRUPT
172 reply will be ignored. 171 reply will be ignored.
173 172
174 Aborting a filesystem connection 173 Aborting a filesystem connection
175 ================================ 174 ================================
176 175
177 It is possible to get into certain situations 176 It is possible to get into certain situations where the filesystem is
178 not responding. Reasons for this may be: 177 not responding. Reasons for this may be:
179 178
180 a) Broken userspace filesystem implementatio 179 a) Broken userspace filesystem implementation
181 180
182 b) Network connection down 181 b) Network connection down
183 182
184 c) Accidental deadlock 183 c) Accidental deadlock
185 184
186 d) Malicious deadlock 185 d) Malicious deadlock
187 186
188 (For more on c) and d) see later sections) 187 (For more on c) and d) see later sections)
189 188
190 In either of these cases it may be useful to a 189 In either of these cases it may be useful to abort the connection to
191 the filesystem. There are several ways to do 190 the filesystem. There are several ways to do this:
192 191
193 - Kill the filesystem daemon. Works in case 192 - Kill the filesystem daemon. Works in case of a) and b)
194 193
195 - Kill the filesystem daemon and all users o 194 - Kill the filesystem daemon and all users of the filesystem. Works
196 in all cases except some malicious deadloc 195 in all cases except some malicious deadlocks
197 196
198 - Use forced umount (umount -f). Works in a 197 - Use forced umount (umount -f). Works in all cases but only if
199 filesystem is still attached (it hasn't be 198 filesystem is still attached (it hasn't been lazy unmounted)
200 199
201 - Abort filesystem through the FUSE control 200 - Abort filesystem through the FUSE control filesystem. Most
202 powerful method, always works. 201 powerful method, always works.
203 202
204 How do non-privileged mounts work? 203 How do non-privileged mounts work?
205 ================================== 204 ==================================
206 205
207 Since the mount() system call is a privileged 206 Since the mount() system call is a privileged operation, a helper
208 program (fusermount) is needed, which is insta 207 program (fusermount) is needed, which is installed setuid root.
209 208
210 The implication of providing non-privileged mo 209 The implication of providing non-privileged mounts is that the mount
211 owner must not be able to use this capability 210 owner must not be able to use this capability to compromise the
212 system. Obvious requirements arising from thi 211 system. Obvious requirements arising from this are:
213 212
214 A) mount owner should not be able to get elev 213 A) mount owner should not be able to get elevated privileges with the
215 help of the mounted filesystem 214 help of the mounted filesystem
216 215
217 B) mount owner should not get illegitimate ac 216 B) mount owner should not get illegitimate access to information from
218 other users' and the super user's processe 217 other users' and the super user's processes
219 218
220 C) mount owner should not be able to induce u 219 C) mount owner should not be able to induce undesired behavior in
221 other users' or the super user's processes 220 other users' or the super user's processes
222 221
223 How are requirements fulfilled? 222 How are requirements fulfilled?
224 =============================== 223 ===============================
225 224
226 A) The mount owner could gain elevated privil 225 A) The mount owner could gain elevated privileges by either:
227 226
228 1. creating a filesystem containing a devi 227 1. creating a filesystem containing a device file, then opening this device
229 228
230 2. creating a filesystem containing a suid 229 2. creating a filesystem containing a suid or sgid application, then executing this application
231 230
232 The solution is not to allow opening devic 231 The solution is not to allow opening device files and ignore
233 setuid and setgid bits when executing prog 232 setuid and setgid bits when executing programs. To ensure this
234 fusermount always adds "nosuid" and "nodev 233 fusermount always adds "nosuid" and "nodev" to the mount options
235 for non-privileged mounts. 234 for non-privileged mounts.
236 235
237 B) If another user is accessing files or dire 236 B) If another user is accessing files or directories in the
238 filesystem, the filesystem daemon serving 237 filesystem, the filesystem daemon serving requests can record the
239 exact sequence and timing of operations pe 238 exact sequence and timing of operations performed. This
240 information is otherwise inaccessible to t 239 information is otherwise inaccessible to the mount owner, so this
241 counts as an information leak. 240 counts as an information leak.
242 241
243 The solution to this problem will be prese 242 The solution to this problem will be presented in point 2) of C).
244 243
245 C) There are several ways in which the mount 244 C) There are several ways in which the mount owner can induce
246 undesired behavior in other users' process 245 undesired behavior in other users' processes, such as:
247 246
248 1) mounting a filesystem over a file or d 247 1) mounting a filesystem over a file or directory which the mount
249 owner could otherwise not be able to m 248 owner could otherwise not be able to modify (or could only
250 make limited modifications). 249 make limited modifications).
251 250
252 This is solved in fusermount, by check 251 This is solved in fusermount, by checking the access
253 permissions on the mountpoint and only 252 permissions on the mountpoint and only allowing the mount if
254 the mount owner can do unlimited modif 253 the mount owner can do unlimited modification (has write
255 access to the mountpoint, and mountpoi 254 access to the mountpoint, and mountpoint is not a "sticky"
256 directory) 255 directory)
257 256
258 2) Even if 1) is solved the mount owner c 257 2) Even if 1) is solved the mount owner can change the behavior
259 of other users' processes. 258 of other users' processes.
260 259
261 i) It can slow down or indefinitely d 260 i) It can slow down or indefinitely delay the execution of a
262 filesystem operation creating a Do 261 filesystem operation creating a DoS against the user or the
263 whole system. For example a suid 262 whole system. For example a suid application locking a
264 system file, and then accessing a 263 system file, and then accessing a file on the mount owner's
265 filesystem could be stopped, and t 264 filesystem could be stopped, and thus causing the system
266 file to be locked forever. 265 file to be locked forever.
267 266
268 ii) It can present files or directori 267 ii) It can present files or directories of unlimited length, or
269 directory structures of unlimited 268 directory structures of unlimited depth, possibly causing a
270 system process to eat up diskspac 269 system process to eat up diskspace, memory or other
271 resources, again causing *DoS*. 270 resources, again causing *DoS*.
272 271
273 The solution to this as well as B) is 272 The solution to this as well as B) is not to allow processes
274 to access the filesystem, which could 273 to access the filesystem, which could otherwise not be
275 monitored or manipulated by the mount 274 monitored or manipulated by the mount owner. Since if the
276 mount owner can ptrace a process, it c 275 mount owner can ptrace a process, it can do all of the above
277 without using a FUSE mount, the same c 276 without using a FUSE mount, the same criteria as used in
278 ptrace can be used to check if a proce 277 ptrace can be used to check if a process is allowed to access
279 the filesystem or not. 278 the filesystem or not.
280 279
281 Note that the *ptrace* check is not st 280 Note that the *ptrace* check is not strictly necessary to
282 prevent C/2/i, it is enough to check i !! 281 prevent B/2/i, it is enough to check if mount owner has enough
283 privilege to send signal to the proces 282 privilege to send signal to the process accessing the
284 filesystem, since *SIGSTOP* can be use 283 filesystem, since *SIGSTOP* can be used to get a similar effect.
285 284
286 I think these limitations are unacceptable? 285 I think these limitations are unacceptable?
287 =========================================== 286 ===========================================
288 287
289 If a sysadmin trusts the users enough, or can 288 If a sysadmin trusts the users enough, or can ensure through other
290 measures, that system processes will never ent 289 measures, that system processes will never enter non-privileged
291 mounts, it can relax the last limitation in se !! 290 mounts, it can relax the last limitation with a 'user_allow_other'
292 !! 291 config option. If this config option is set, the mounting user can
293 - With the 'user_allow_other' config option. !! 292 add the 'allow_other' mount option which disables the check for other
294 set, the mounting user can add the 'allow_ !! 293 users' processes.
295 disables the check for other users' proces <<
296 <<
297 User namespaces have an unintuitive intera <<
298 an unprivileged user - normally restricted <<
299 'allow_other' - could do so in a user name <<
300 privileged. If any process could access su <<
301 this would give the mounting user the abil <<
302 processes in user namespaces where they're <<
303 reason 'allow_other' restricts access to u <<
304 or a descendant. <<
305 <<
306 - With the 'allow_sys_admin_access' module o <<
307 set, super user's processes have unrestric <<
308 irrespective of allow_other setting or use <<
309 mounting user. <<
310 <<
311 Note that both of these relaxations expose the <<
312 information leak or *DoS* as described in poin <<
313 preceding section. <<
314 294
315 Kernel - userspace interface 295 Kernel - userspace interface
316 ============================ 296 ============================
317 297
318 The following diagram shows how a filesystem o 298 The following diagram shows how a filesystem operation (in this
319 example unlink) is performed in FUSE. :: 299 example unlink) is performed in FUSE. ::
320 300
321 301
322 | "rm /mnt/fuse/file" | FUSE 302 | "rm /mnt/fuse/file" | FUSE filesystem daemon
323 | | 303 | |
324 | | >sys_ 304 | | >sys_read()
325 | | >fu 305 | | >fuse_dev_read()
326 | | > 306 | | >request_wait()
327 | | 307 | | [sleep on fc->waitq]
328 | | 308 | |
329 | >sys_unlink() | 309 | >sys_unlink() |
330 | >fuse_unlink() | 310 | >fuse_unlink() |
331 | [get request from | 311 | [get request from |
332 | fc->unused_list] | 312 | fc->unused_list] |
333 | >request_send() | 313 | >request_send() |
334 | [queue req on fc->pending] | 314 | [queue req on fc->pending] |
335 | [wake up fc->waitq] | 315 | [wake up fc->waitq] | [woken up]
336 | >request_wait_answer() | 316 | >request_wait_answer() |
337 | [sleep on req->waitq] | 317 | [sleep on req->waitq] |
338 | | < 318 | | <request_wait()
339 | | [ 319 | | [remove req from fc->pending]
340 | | [ 320 | | [copy req to read buffer]
341 | | [ 321 | | [add req to fc->processing]
342 | | <fu 322 | | <fuse_dev_read()
343 | | <sys_ 323 | | <sys_read()
344 | | 324 | |
345 | | [perf 325 | | [perform unlink]
346 | | 326 | |
347 | | >sys_ 327 | | >sys_write()
348 | | >fu 328 | | >fuse_dev_write()
349 | | [ 329 | | [look up req in fc->processing]
350 | | [ 330 | | [remove from fc->processing]
351 | | [ 331 | | [copy write buffer to req]
352 | [woken up] | [ 332 | [woken up] | [wake up req->waitq]
353 | | <fu 333 | | <fuse_dev_write()
354 | | <sys_ 334 | | <sys_write()
355 | <request_wait_answer() | 335 | <request_wait_answer() |
356 | <request_send() | 336 | <request_send() |
357 | [add request to | 337 | [add request to |
358 | fc->unused_list] | 338 | fc->unused_list] |
359 | <fuse_unlink() | 339 | <fuse_unlink() |
360 | <sys_unlink() | 340 | <sys_unlink() |
361 341
362 .. note:: Everything in the description above 342 .. note:: Everything in the description above is greatly simplified
363 343
364 There are a couple of ways in which to deadloc 344 There are a couple of ways in which to deadlock a FUSE filesystem.
365 Since we are talking about unprivileged usersp 345 Since we are talking about unprivileged userspace programs,
366 something must be done about these. 346 something must be done about these.
367 347
368 **Scenario 1 - Simple deadlock**:: 348 **Scenario 1 - Simple deadlock**::
369 349
370 | "rm /mnt/fuse/file" | FUSE 350 | "rm /mnt/fuse/file" | FUSE filesystem daemon
371 | | 351 | |
372 | >sys_unlink("/mnt/fuse/file") | 352 | >sys_unlink("/mnt/fuse/file") |
373 | [acquire inode semaphore | 353 | [acquire inode semaphore |
374 | for "file"] | 354 | for "file"] |
375 | >fuse_unlink() | 355 | >fuse_unlink() |
376 | [sleep on req->waitq] | 356 | [sleep on req->waitq] |
377 | | <sys_ 357 | | <sys_read()
378 | | >sys_ 358 | | >sys_unlink("/mnt/fuse/file")
379 | | [ac 359 | | [acquire inode semaphore
380 | | fo 360 | | for "file"]
381 | | *DE 361 | | *DEADLOCK*
382 362
383 The solution for this is to allow the filesyst 363 The solution for this is to allow the filesystem to be aborted.
384 364
385 **Scenario 2 - Tricky deadlock** 365 **Scenario 2 - Tricky deadlock**
386 366
387 367
388 This one needs a carefully crafted filesystem. 368 This one needs a carefully crafted filesystem. It's a variation on
389 the above, only the call back to the filesyste 369 the above, only the call back to the filesystem is not explicit,
390 but is caused by a pagefault. :: 370 but is caused by a pagefault. ::
391 371
392 | Kamikaze filesystem thread 1 | Kamik 372 | Kamikaze filesystem thread 1 | Kamikaze filesystem thread 2
393 | | 373 | |
394 | [fd = open("/mnt/fuse/file")] | [requ 374 | [fd = open("/mnt/fuse/file")] | [request served normally]
395 | [mmap fd to 'addr'] | 375 | [mmap fd to 'addr'] |
396 | [close fd] | [FLUS 376 | [close fd] | [FLUSH triggers 'magic' flag]
397 | [read a byte from addr] | 377 | [read a byte from addr] |
398 | >do_page_fault() | 378 | >do_page_fault() |
399 | [find or create page] | 379 | [find or create page] |
400 | [lock page] | 380 | [lock page] |
401 | >fuse_readpage() | 381 | >fuse_readpage() |
402 | [queue READ request] | 382 | [queue READ request] |
403 | [sleep on req->waitq] | 383 | [sleep on req->waitq] |
404 | | [read 384 | | [read request to buffer]
405 | | [crea 385 | | [create reply header before addr]
406 | | >sys_ 386 | | >sys_write(addr - headerlength)
407 | | >fu 387 | | >fuse_dev_write()
408 | | [ 388 | | [look up req in fc->processing]
409 | | [ 389 | | [remove from fc->processing]
410 | | [ 390 | | [copy write buffer to req]
411 | | 391 | | >do_page_fault()
412 | | 392 | | [find or create page]
413 | | 393 | | [lock page]
414 | | 394 | | * DEADLOCK *
415 395
416 The solution is basically the same as above. 396 The solution is basically the same as above.
417 397
418 An additional problem is that while the write 398 An additional problem is that while the write buffer is being copied
419 to the request, the request must not be interr 399 to the request, the request must not be interrupted/aborted. This is
420 because the destination address of the copy ma 400 because the destination address of the copy may not be valid after the
421 request has returned. 401 request has returned.
422 402
423 This is solved with doing the copy atomically, 403 This is solved with doing the copy atomically, and allowing abort
424 while the page(s) belonging to the write buffe 404 while the page(s) belonging to the write buffer are faulted with
425 get_user_pages(). The 'req->locked' flag indi 405 get_user_pages(). The 'req->locked' flag indicates when the copy is
426 taking place, and abort is delayed until this 406 taking place, and abort is delayed until this flag is unset.
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.