1 .. SPDX-License-Identifier: GPL-2.0
2
3 ======================
4 The seq_file Interface
5 ======================
6
7 Copyright 2003 Jonathan Corbet <corbet@lwn.net>
8
9 This file is originally from the LWN.net Driver Porting series at
10 https://lwn.net/Articles/driver-porting/
11
12
13 There are numerous ways for a device driver (or other kernel component) to
14 provide information to the user or system administrator. One useful
15 technique is the creation of virtual files, in debugfs, /proc or elsewhere.
16 Virtual files can provide human-readable output that is easy to get at
17 without any special utility programs; they can also make life easier for
18 script writers. It is not surprising that the use of virtual files has
19 grown over the years.
20
21 Creating those files correctly has always been a bit of a challenge,
22 however. It is not that hard to make a virtual file which returns a
23 string. But life gets trickier if the output is long - anything greater
24 than an application is likely to read in a single operation. Handling
25 multiple reads (and seeks) requires careful attention to the reader's
26 position within the virtual file - that position is, likely as not, in the
27 middle of a line of output. The kernel has traditionally had a number of
28 implementations that got this wrong.
29
30 The 2.6 kernel contains a set of functions (implemented by Alexander Viro)
31 which are designed to make it easy for virtual file creators to get it
32 right.
33
34 The seq_file interface is available via <linux/seq_file.h>. There are
35 three aspects to seq_file:
36
37 * An iterator interface which lets a virtual file implementation
38 step through the objects it is presenting.
39
40 * Some utility functions for formatting objects for output without
41 needing to worry about things like output buffers.
42
43 * A set of canned file_operations which implement most operations on
44 the virtual file.
45
46 We'll look at the seq_file interface via an extremely simple example: a
47 loadable module which creates a file called /proc/sequence. The file, when
48 read, simply produces a set of increasing integer values, one per line. The
49 sequence will continue until the user loses patience and finds something
50 better to do. The file is seekable, in that one can do something like the
51 following::
52
53 dd if=/proc/sequence of=out1 count=1
54 dd if=/proc/sequence skip=1 of=out2 count=1
55
56 Then concatenate the output files out1 and out2 and get the right
57 result. Yes, it is a thoroughly useless module, but the point is to show
58 how the mechanism works without getting lost in other details. (Those
59 wanting to see the full source for this module can find it at
60 https://lwn.net/Articles/22359/).
61
62 Deprecated create_proc_entry
63 ============================
64
65 Note that the above article uses create_proc_entry which was removed in
66 kernel 3.10. Current versions require the following update::
67
68 - entry = create_proc_entry("sequence", 0, NULL);
69 - if (entry)
70 - entry->proc_fops = &ct_file_ops;
71 + entry = proc_create("sequence", 0, NULL, &ct_file_ops);
72
73 The iterator interface
74 ======================
75
76 Modules implementing a virtual file with seq_file must implement an
77 iterator object that allows stepping through the data of interest
78 during a "session" (roughly one read() system call). If the iterator
79 is able to move to a specific position - like the file they implement,
80 though with freedom to map the position number to a sequence location
81 in whatever way is convenient - the iterator need only exist
82 transiently during a session. If the iterator cannot easily find a
83 numerical position but works well with a first/next interface, the
84 iterator can be stored in the private data area and continue from one
85 session to the next.
86
87 A seq_file implementation that is formatting firewall rules from a
88 table, for example, could provide a simple iterator that interprets
89 position N as the Nth rule in the chain. A seq_file implementation
90 that presents the content of a, potentially volatile, linked list
91 might record a pointer into that list, providing that can be done
92 without risk of the current location being removed.
93
94 Positioning can thus be done in whatever way makes the most sense for
95 the generator of the data, which need not be aware of how a position
96 translates to an offset in the virtual file. The one obvious exception
97 is that a position of zero should indicate the beginning of the file.
98
99 The /proc/sequence iterator just uses the count of the next number it
100 will output as its position.
101
102 Four functions must be implemented to make the iterator work. The
103 first, called start(), starts a session and takes a position as an
104 argument, returning an iterator which will start reading at that
105 position. The pos passed to start() will always be either zero, or
106 the most recent pos used in the previous session.
107
108 For our simple sequence example,
109 the start() function looks like::
110
111 static void *ct_seq_start(struct seq_file *s, loff_t *pos)
112 {
113 loff_t *spos = kmalloc(sizeof(loff_t), GFP_KERNEL);
114 if (! spos)
115 return NULL;
116 *spos = *pos;
117 return spos;
118 }
119
120 The entire data structure for this iterator is a single loff_t value
121 holding the current position. There is no upper bound for the sequence
122 iterator, but that will not be the case for most other seq_file
123 implementations; in most cases the start() function should check for a
124 "past end of file" condition and return NULL if need be.
125
126 For more complicated applications, the private field of the seq_file
127 structure can be used to hold state from session to session. There is
128 also a special value which can be returned by the start() function
129 called SEQ_START_TOKEN; it can be used if you wish to instruct your
130 show() function (described below) to print a header at the top of the
131 output. SEQ_START_TOKEN should only be used if the offset is zero,
132 however. SEQ_START_TOKEN has no special meaning to the core seq_file
133 code. It is provided as a convenience for a start() function to
134 communicate with the next() and show() functions.
135
136 The next function to implement is called, amazingly, next(); its job is to
137 move the iterator forward to the next position in the sequence. The
138 example module can simply increment the position by one; more useful
139 modules will do what is needed to step through some data structure. The
140 next() function returns a new iterator, or NULL if the sequence is
141 complete. Here's the example version::
142
143 static void *ct_seq_next(struct seq_file *s, void *v, loff_t *pos)
144 {
145 loff_t *spos = v;
146 *pos = ++*spos;
147 return spos;
148 }
149
150 The next() function should set ``*pos`` to a value that start() can use
151 to find the new location in the sequence. When the iterator is being
152 stored in the private data area, rather than being reinitialized on each
153 start(), it might seem sufficient to simply set ``*pos`` to any non-zero
154 value (zero always tells start() to restart the sequence). This is not
155 sufficient due to historical problems.
156
157 Historically, many next() functions have *not* updated ``*pos`` at
158 end-of-file. If the value is then used by start() to initialise the
159 iterator, this can result in corner cases where the last entry in the
160 sequence is reported twice in the file. In order to discourage this bug
161 from being resurrected, the core seq_file code now produces a warning if
162 a next() function does not change the value of ``*pos``. Consequently a
163 next() function *must* change the value of ``*pos``, and of course must
164 set it to a non-zero value.
165
166 The stop() function closes a session; its job, of course, is to clean
167 up. If dynamic memory is allocated for the iterator, stop() is the
168 place to free it; if a lock was taken by start(), stop() must release
169 that lock. The value that ``*pos`` was set to by the last next() call
170 before stop() is remembered, and used for the first start() call of
171 the next session unless lseek() has been called on the file; in that
172 case next start() will be asked to start at position zero::
173
174 static void ct_seq_stop(struct seq_file *s, void *v)
175 {
176 kfree(v);
177 }
178
179 Finally, the show() function should format the object currently pointed to
180 by the iterator for output. The example module's show() function is::
181
182 static int ct_seq_show(struct seq_file *s, void *v)
183 {
184 loff_t *spos = v;
185 seq_printf(s, "%lld\n", (long long)*spos);
186 return 0;
187 }
188
189 If all is well, the show() function should return zero. A negative error
190 code in the usual manner indicates that something went wrong; it will be
191 passed back to user space. This function can also return SEQ_SKIP, which
192 causes the current item to be skipped; if the show() function has already
193 generated output before returning SEQ_SKIP, that output will be dropped.
194
195 We will look at seq_printf() in a moment. But first, the definition of the
196 seq_file iterator is finished by creating a seq_operations structure with
197 the four functions we have just defined::
198
199 static const struct seq_operations ct_seq_ops = {
200 .start = ct_seq_start,
201 .next = ct_seq_next,
202 .stop = ct_seq_stop,
203 .show = ct_seq_show
204 };
205
206 This structure will be needed to tie our iterator to the /proc file in
207 a little bit.
208
209 It's worth noting that the iterator value returned by start() and
210 manipulated by the other functions is considered to be completely opaque by
211 the seq_file code. It can thus be anything that is useful in stepping
212 through the data to be output. Counters can be useful, but it could also be
213 a direct pointer into an array or linked list. Anything goes, as long as
214 the programmer is aware that things can happen between calls to the
215 iterator function. However, the seq_file code (by design) will not sleep
216 between the calls to start() and stop(), so holding a lock during that time
217 is a reasonable thing to do. The seq_file code will also avoid taking any
218 other locks while the iterator is active.
219
220 The iterator value returned by start() or next() is guaranteed to be
221 passed to a subsequent next() or stop() call. This allows resources
222 such as locks that were taken to be reliably released. There is *no*
223 guarantee that the iterator will be passed to show(), though in practice
224 it often will be.
225
226
227 Formatted output
228 ================
229
230 The seq_file code manages positioning within the output created by the
231 iterator and getting it into the user's buffer. But, for that to work, that
232 output must be passed to the seq_file code. Some utility functions have
233 been defined which make this task easy.
234
235 Most code will simply use seq_printf(), which works pretty much like
236 printk(), but which requires the seq_file pointer as an argument.
237
238 For straight character output, the following functions may be used::
239
240 seq_putc(struct seq_file *m, char c);
241 seq_puts(struct seq_file *m, const char *s);
242 seq_escape(struct seq_file *m, const char *s, const char *esc);
243
244 The first two output a single character and a string, just like one would
245 expect. seq_escape() is like seq_puts(), except that any character in s
246 which is in the string esc will be represented in octal form in the output.
247
248 There are also a pair of functions for printing filenames::
249
250 int seq_path(struct seq_file *m, const struct path *path,
251 const char *esc);
252 int seq_path_root(struct seq_file *m, const struct path *path,
253 const struct path *root, const char *esc)
254
255 Here, path indicates the file of interest, and esc is a set of characters
256 which should be escaped in the output. A call to seq_path() will output
257 the path relative to the current process's filesystem root. If a different
258 root is desired, it can be used with seq_path_root(). If it turns out that
259 path cannot be reached from root, seq_path_root() returns SEQ_SKIP.
260
261 A function producing complicated output may want to check::
262
263 bool seq_has_overflowed(struct seq_file *m);
264
265 and avoid further seq_<output> calls if true is returned.
266
267 A true return from seq_has_overflowed means that the seq_file buffer will
268 be discarded and the seq_show function will attempt to allocate a larger
269 buffer and retry printing.
270
271
272 Making it all work
273 ==================
274
275 So far, we have a nice set of functions which can produce output within the
276 seq_file system, but we have not yet turned them into a file that a user
277 can see. Creating a file within the kernel requires, of course, the
278 creation of a set of file_operations which implement the operations on that
279 file. The seq_file interface provides a set of canned operations which do
280 most of the work. The virtual file author still must implement the open()
281 method, however, to hook everything up. The open function is often a single
282 line, as in the example module::
283
284 static int ct_open(struct inode *inode, struct file *file)
285 {
286 return seq_open(file, &ct_seq_ops);
287 }
288
289 Here, the call to seq_open() takes the seq_operations structure we created
290 before, and gets set up to iterate through the virtual file.
291
292 On a successful open, seq_open() stores the struct seq_file pointer in
293 file->private_data. If you have an application where the same iterator can
294 be used for more than one file, you can store an arbitrary pointer in the
295 private field of the seq_file structure; that value can then be retrieved
296 by the iterator functions.
297
298 There is also a wrapper function to seq_open() called seq_open_private(). It
299 kmallocs a zero filled block of memory and stores a pointer to it in the
300 private field of the seq_file structure, returning 0 on success. The
301 block size is specified in a third parameter to the function, e.g.::
302
303 static int ct_open(struct inode *inode, struct file *file)
304 {
305 return seq_open_private(file, &ct_seq_ops,
306 sizeof(struct mystruct));
307 }
308
309 There is also a variant function, __seq_open_private(), which is functionally
310 identical except that, if successful, it returns the pointer to the allocated
311 memory block, allowing further initialisation e.g.::
312
313 static int ct_open(struct inode *inode, struct file *file)
314 {
315 struct mystruct *p =
316 __seq_open_private(file, &ct_seq_ops, sizeof(*p));
317
318 if (!p)
319 return -ENOMEM;
320
321 p->foo = bar; /* initialize my stuff */
322 ...
323 p->baz = true;
324
325 return 0;
326 }
327
328 A corresponding close function, seq_release_private() is available which
329 frees the memory allocated in the corresponding open.
330
331 The other operations of interest - read(), llseek(), and release() - are
332 all implemented by the seq_file code itself. So a virtual file's
333 file_operations structure will look like::
334
335 static const struct file_operations ct_file_ops = {
336 .owner = THIS_MODULE,
337 .open = ct_open,
338 .read = seq_read,
339 .llseek = seq_lseek,
340 .release = seq_release
341 };
342
343 There is also a seq_release_private() which passes the contents of the
344 seq_file private field to kfree() before releasing the structure.
345
346 The final step is the creation of the /proc file itself. In the example
347 code, that is done in the initialization code in the usual way::
348
349 static int ct_init(void)
350 {
351 struct proc_dir_entry *entry;
352
353 proc_create("sequence", 0, NULL, &ct_file_ops);
354 return 0;
355 }
356
357 module_init(ct_init);
358
359 And that is pretty much it.
360
361
362 seq_list
363 ========
364
365 If your file will be iterating through a linked list, you may find these
366 routines useful::
367
368 struct list_head *seq_list_start(struct list_head *head,
369 loff_t pos);
370 struct list_head *seq_list_start_head(struct list_head *head,
371 loff_t pos);
372 struct list_head *seq_list_next(void *v, struct list_head *head,
373 loff_t *ppos);
374
375 These helpers will interpret pos as a position within the list and iterate
376 accordingly. Your start() and next() functions need only invoke the
377 ``seq_list_*`` helpers with a pointer to the appropriate list_head structure.
378
379
380 The extra-simple version
381 ========================
382
383 For extremely simple virtual files, there is an even easier interface. A
384 module can define only the show() function, which should create all the
385 output that the virtual file will contain. The file's open() method then
386 calls::
387
388 int single_open(struct file *file,
389 int (*show)(struct seq_file *m, void *p),
390 void *data);
391
392 When output time comes, the show() function will be called once. The data
393 value given to single_open() can be found in the private field of the
394 seq_file structure. When using single_open(), the programmer should use
395 single_release() instead of seq_release() in the file_operations structure
396 to avoid a memory leak.
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.