3 Copyright 2003 Jonathan Corbet <corbet@lwn.net>
4 This file is originally from the LWN.net Driver Porting series at
5 http://lwn.net/Articles/driver-porting/
8 There are numerous ways for a device driver (or other kernel component) to
9 provide information to the user or system administrator. One useful
10 technique is the creation of virtual files, in debugfs, /proc or elsewhere.
11 Virtual files can provide human-readable output that is easy to get at
12 without any special utility programs; they can also make life easier for
13 script writers. It is not surprising that the use of virtual files has
16 Creating those files correctly has always been a bit of a challenge,
17 however. It is not that hard to make a virtual file which returns a
18 string. But life gets trickier if the output is long - anything greater
19 than an application is likely to read in a single operation. Handling
20 multiple reads (and seeks) requires careful attention to the reader's
21 position within the virtual file - that position is, likely as not, in the
22 middle of a line of output. The kernel has traditionally had a number of
23 implementations that got this wrong.
25 The 2.6 kernel contains a set of functions (implemented by Alexander Viro)
26 which are designed to make it easy for virtual file creators to get it
29 The seq_file interface is available via <linux/seq_file.h>. There are
30 three aspects to seq_file:
32 * An iterator interface which lets a virtual file implementation
33 step through the objects it is presenting.
35 * Some utility functions for formatting objects for output without
36 needing to worry about things like output buffers.
38 * A set of canned file_operations which implement most operations on
41 We'll look at the seq_file interface via an extremely simple example: a
42 loadable module which creates a file called /proc/sequence. The file, when
43 read, simply produces a set of increasing integer values, one per line. The
44 sequence will continue until the user loses patience and finds something
45 better to do. The file is seekable, in that one can do something like the
48 dd if=/proc/sequence of=out1 count=1
49 dd if=/proc/sequence skip=1 out=out2 count=1
51 Then concatenate the output files out1 and out2 and get the right
52 result. Yes, it is a thoroughly useless module, but the point is to show
53 how the mechanism works without getting lost in other details. (Those
54 wanting to see the full source for this module can find it at
55 http://lwn.net/Articles/22359/).
58 The iterator interface
60 Modules implementing a virtual file with seq_file must implement a simple
61 iterator object that allows stepping through the data of interest.
62 Iterators must be able to move to a specific position - like the file they
63 implement - but the interpretation of that position is up to the iterator
64 itself. A seq_file implementation that is formatting firewall rules, for
65 example, could interpret position N as the Nth rule in the chain.
66 Positioning can thus be done in whatever way makes the most sense for the
67 generator of the data, which need not be aware of how a position translates
68 to an offset in the virtual file. The one obvious exception is that a
69 position of zero should indicate the beginning of the file.
71 The /proc/sequence iterator just uses the count of the next number it
72 will output as its position.
74 Four functions must be implemented to make the iterator work. The first,
75 called start() takes a position as an argument and returns an iterator
76 which will start reading at that position. For our simple sequence example,
77 the start() function looks like:
79 static void *ct_seq_start(struct seq_file *s, loff_t *pos)
81 loff_t *spos = kmalloc(sizeof(loff_t), GFP_KERNEL);
88 The entire data structure for this iterator is a single loff_t value
89 holding the current position. There is no upper bound for the sequence
90 iterator, but that will not be the case for most other seq_file
91 implementations; in most cases the start() function should check for a
92 "past end of file" condition and return NULL if need be.
94 For more complicated applications, the private field of the seq_file
95 structure can be used. There is also a special value which can be returned
96 by the start() function called SEQ_START_TOKEN; it can be used if you wish
97 to instruct your show() function (described below) to print a header at the
98 top of the output. SEQ_START_TOKEN should only be used if the offset is
101 The next function to implement is called, amazingly, next(); its job is to
102 move the iterator forward to the next position in the sequence. The
103 example module can simply increment the position by one; more useful
104 modules will do what is needed to step through some data structure. The
105 next() function returns a new iterator, or NULL if the sequence is
106 complete. Here's the example version:
108 static void *ct_seq_next(struct seq_file *s, void *v, loff_t *pos)
115 The stop() function is called when iteration is complete; its job, of
116 course, is to clean up. If dynamic memory is allocated for the iterator,
117 stop() is the place to free it.
119 static void ct_seq_stop(struct seq_file *s, void *v)
124 Finally, the show() function should format the object currently pointed to
125 by the iterator for output. The example module's show() function is:
127 static int ct_seq_show(struct seq_file *s, void *v)
130 seq_printf(s, "%lld\n", (long long)*spos);
134 If all is well, the show() function should return zero. A negative error
135 code in the usual manner indicates that something went wrong; it will be
136 passed back to user space. This function can also return SEQ_SKIP, which
137 causes the current item to be skipped; if the show() function has already
138 generated output before returning SEQ_SKIP, that output will be dropped.
140 We will look at seq_printf() in a moment. But first, the definition of the
141 seq_file iterator is finished by creating a seq_operations structure with
142 the four functions we have just defined:
144 static const struct seq_operations ct_seq_ops = {
145 .start = ct_seq_start,
151 This structure will be needed to tie our iterator to the /proc file in
154 It's worth noting that the iterator value returned by start() and
155 manipulated by the other functions is considered to be completely opaque by
156 the seq_file code. It can thus be anything that is useful in stepping
157 through the data to be output. Counters can be useful, but it could also be
158 a direct pointer into an array or linked list. Anything goes, as long as
159 the programmer is aware that things can happen between calls to the
160 iterator function. However, the seq_file code (by design) will not sleep
161 between the calls to start() and stop(), so holding a lock during that time
162 is a reasonable thing to do. The seq_file code will also avoid taking any
163 other locks while the iterator is active.
168 The seq_file code manages positioning within the output created by the
169 iterator and getting it into the user's buffer. But, for that to work, that
170 output must be passed to the seq_file code. Some utility functions have
171 been defined which make this task easy.
173 Most code will simply use seq_printf(), which works pretty much like
174 printk(), but which requires the seq_file pointer as an argument. It is
175 common to ignore the return value from seq_printf(), but a function
176 producing complicated output may want to check that value and quit if
177 something non-zero is returned; an error return means that the seq_file
178 buffer has been filled and further output will be discarded.
180 For straight character output, the following functions may be used:
182 int seq_putc(struct seq_file *m, char c);
183 int seq_puts(struct seq_file *m, const char *s);
184 int seq_escape(struct seq_file *m, const char *s, const char *esc);
186 The first two output a single character and a string, just like one would
187 expect. seq_escape() is like seq_puts(), except that any character in s
188 which is in the string esc will be represented in octal form in the output.
190 There is also a pair of functions for printing filenames:
192 int seq_path(struct seq_file *m, struct path *path, char *esc);
193 int seq_path_root(struct seq_file *m, struct path *path,
194 struct path *root, char *esc)
196 Here, path indicates the file of interest, and esc is a set of characters
197 which should be escaped in the output. A call to seq_path() will output
198 the path relative to the current process's filesystem root. If a different
199 root is desired, it can be used with seq_path_root(). Note that, if it
200 turns out that path cannot be reached from root, the value of root will be
201 changed in seq_file_root() to a root which *does* work.
206 So far, we have a nice set of functions which can produce output within the
207 seq_file system, but we have not yet turned them into a file that a user
208 can see. Creating a file within the kernel requires, of course, the
209 creation of a set of file_operations which implement the operations on that
210 file. The seq_file interface provides a set of canned operations which do
211 most of the work. The virtual file author still must implement the open()
212 method, however, to hook everything up. The open function is often a single
213 line, as in the example module:
215 static int ct_open(struct inode *inode, struct file *file)
217 return seq_open(file, &ct_seq_ops);
220 Here, the call to seq_open() takes the seq_operations structure we created
221 before, and gets set up to iterate through the virtual file.
223 On a successful open, seq_open() stores the struct seq_file pointer in
224 file->private_data. If you have an application where the same iterator can
225 be used for more than one file, you can store an arbitrary pointer in the
226 private field of the seq_file structure; that value can then be retrieved
227 by the iterator functions.
229 The other operations of interest - read(), llseek(), and release() - are
230 all implemented by the seq_file code itself. So a virtual file's
231 file_operations structure will look like:
233 static const struct file_operations ct_file_ops = {
234 .owner = THIS_MODULE,
238 .release = seq_release
241 There is also a seq_release_private() which passes the contents of the
242 seq_file private field to kfree() before releasing the structure.
244 The final step is the creation of the /proc file itself. In the example
245 code, that is done in the initialization code in the usual way:
247 static int ct_init(void)
249 struct proc_dir_entry *entry;
251 entry = create_proc_entry("sequence", 0, NULL);
253 entry->proc_fops = &ct_file_ops;
257 module_init(ct_init);
259 And that is pretty much it.
264 If your file will be iterating through a linked list, you may find these
267 struct list_head *seq_list_start(struct list_head *head,
269 struct list_head *seq_list_start_head(struct list_head *head,
271 struct list_head *seq_list_next(void *v, struct list_head *head,
274 These helpers will interpret pos as a position within the list and iterate
275 accordingly. Your start() and next() functions need only invoke the
276 seq_list_* helpers with a pointer to the appropriate list_head structure.
279 The extra-simple version
281 For extremely simple virtual files, there is an even easier interface. A
282 module can define only the show() function, which should create all the
283 output that the virtual file will contain. The file's open() method then
286 int single_open(struct file *file,
287 int (*show)(struct seq_file *m, void *p),
290 When output time comes, the show() function will be called once. The data
291 value given to single_open() can be found in the private field of the
292 seq_file structure. When using single_open(), the programmer should use
293 single_release() instead of seq_release() in the file_operations structure
294 to avoid a memory leak.