1 // SPDX-License-Identifier: GPL-2.0-or-later
3 * lib/textsearch.c Generic text search interface
5 * Authors: Thomas Graf <tgraf@suug.ch>
6 * Pablo Neira Ayuso <pablo@netfilter.org>
8 * ==========================================================================
15 * The textsearch infrastructure provides text searching facilities for
16 * both linear and non-linear data. Individual search algorithms are
17 * implemented in modules and chosen by the user.
21 * .. code-block:: none
25 * | finish()|<--------------(6)-----------------+
26 * |get_next_block()|<--------------(5)---------------+ |
28 * | | +------------------------------+
29 * | | | init() find() destroy() |
30 * | | +------------------------------+
32 * | | +---------------+ (2) (4) (8)
33 * | (1)|----->| prepare() |---+ | |
34 * | (3)|----->| find()/next() |-----------+ |
35 * | (7)|----->| destroy() |----------------------+
36 * +----------------+ +---------------+
38 * (1) User configures a search by calling textsearch_prepare() specifying
39 * the search parameters such as the pattern and algorithm name.
40 * (2) Core requests the algorithm to allocate and initialize a search
41 * configuration according to the specified parameters.
42 * (3) User starts the search(es) by calling textsearch_find() or
43 * textsearch_next() to fetch subsequent occurrences. A state variable
44 * is provided to the algorithm to store persistent variables.
45 * (4) Core eventually resets the search offset and forwards the find()
46 * request to the algorithm.
47 * (5) Algorithm calls get_next_block() provided by the user continuously
48 * to fetch the data to be searched in block by block.
49 * (6) Algorithm invokes finish() after the last call to get_next_block
50 * to clean up any leftovers from get_next_block. (Optional)
51 * (7) User destroys the configuration by calling textsearch_destroy().
52 * (8) Core notifies the algorithm to destroy algorithm specific
53 * allocations. (Optional)
57 * Before a search can be performed, a configuration must be created
58 * by calling textsearch_prepare() specifying the searching algorithm,
59 * the pattern to look for and flags. As a flag, you can set TS_IGNORECASE
60 * to perform case insensitive matching. But it might slow down
61 * performance of algorithm, so you should use it at own your risk.
62 * The returned configuration may then be used for an arbitrary
63 * amount of times and even in parallel as long as a separate struct
64 * ts_state variable is provided to every instance.
66 * The actual search is performed by either calling
67 * textsearch_find_continuous() for linear data or by providing
68 * an own get_next_block() implementation and
69 * calling textsearch_find(). Both functions return
70 * the position of the first occurrence of the pattern or UINT_MAX if
71 * no match was found. Subsequent occurrences can be found by calling
72 * textsearch_next() regardless of the linearity of the data.
74 * Once you're done using a configuration it must be given back via
80 * struct ts_config *conf;
81 * struct ts_state state;
82 * const char *pattern = "chicken";
83 * const char *example = "We dance the funky chicken";
85 * conf = textsearch_prepare("kmp", pattern, strlen(pattern),
86 * GFP_KERNEL, TS_AUTOLOAD);
88 * err = PTR_ERR(conf);
92 * pos = textsearch_find_continuous(conf, \&state, example, strlen(example));
93 * if (pos != UINT_MAX)
94 * panic("Oh my god, dancing chickens at \%d\n", pos);
96 * textsearch_destroy(conf);
98 /* ========================================================================== */
100 #include <linux/module.h>
101 #include <linux/types.h>
102 #include <linux/string.h>
103 #include <linux/init.h>
104 #include <linux/rculist.h>
105 #include <linux/rcupdate.h>
106 #include <linux/err.h>
107 #include <linux/textsearch.h>
108 #include <linux/slab.h>
110 static LIST_HEAD(ts_ops
);
111 static DEFINE_SPINLOCK(ts_mod_lock
);
113 static inline struct ts_ops
*lookup_ts_algo(const char *name
)
118 list_for_each_entry_rcu(o
, &ts_ops
, list
) {
119 if (!strcmp(name
, o
->name
)) {
120 if (!try_module_get(o
->owner
))
132 * textsearch_register - register a textsearch module
133 * @ops: operations lookup table
135 * This function must be called by textsearch modules to announce
136 * their presence. The specified &@ops must have %name set to a
137 * unique identifier and the callbacks find(), init(), get_pattern(),
138 * and get_pattern_len() must be implemented.
140 * Returns 0 or -EEXISTS if another module has already registered
143 int textsearch_register(struct ts_ops
*ops
)
148 if (ops
->name
== NULL
|| ops
->find
== NULL
|| ops
->init
== NULL
||
149 ops
->get_pattern
== NULL
|| ops
->get_pattern_len
== NULL
)
152 spin_lock(&ts_mod_lock
);
153 list_for_each_entry(o
, &ts_ops
, list
) {
154 if (!strcmp(ops
->name
, o
->name
))
158 list_add_tail_rcu(&ops
->list
, &ts_ops
);
161 spin_unlock(&ts_mod_lock
);
164 EXPORT_SYMBOL(textsearch_register
);
167 * textsearch_unregister - unregister a textsearch module
168 * @ops: operations lookup table
170 * This function must be called by textsearch modules to announce
171 * their disappearance for examples when the module gets unloaded.
172 * The &ops parameter must be the same as the one during the
175 * Returns 0 on success or -ENOENT if no matching textsearch
176 * registration was found.
178 int textsearch_unregister(struct ts_ops
*ops
)
183 spin_lock(&ts_mod_lock
);
184 list_for_each_entry(o
, &ts_ops
, list
) {
186 list_del_rcu(&o
->list
);
193 spin_unlock(&ts_mod_lock
);
196 EXPORT_SYMBOL(textsearch_unregister
);
198 struct ts_linear_state
204 static unsigned int get_linear_data(unsigned int consumed
, const u8
**dst
,
205 struct ts_config
*conf
,
206 struct ts_state
*state
)
208 struct ts_linear_state
*st
= (struct ts_linear_state
*) state
->cb
;
210 if (likely(consumed
< st
->len
)) {
211 *dst
= st
->data
+ consumed
;
212 return st
->len
- consumed
;
219 * textsearch_find_continuous - search a pattern in continuous/linear data
220 * @conf: search configuration
221 * @state: search state
222 * @data: data to search in
223 * @len: length of data
225 * A simplified version of textsearch_find() for continuous/linear data.
226 * Call textsearch_next() to retrieve subsequent matches.
228 * Returns the position of first occurrence of the pattern or
229 * %UINT_MAX if no occurrence was found.
231 unsigned int textsearch_find_continuous(struct ts_config
*conf
,
232 struct ts_state
*state
,
233 const void *data
, unsigned int len
)
235 struct ts_linear_state
*st
= (struct ts_linear_state
*) state
->cb
;
237 conf
->get_next_block
= get_linear_data
;
241 return textsearch_find(conf
, state
);
243 EXPORT_SYMBOL(textsearch_find_continuous
);
246 * textsearch_prepare - Prepare a search
247 * @algo: name of search algorithm
248 * @pattern: pattern data
249 * @len: length of pattern
250 * @gfp_mask: allocation mask
251 * @flags: search flags
253 * Looks up the search algorithm module and creates a new textsearch
254 * configuration for the specified pattern.
256 * Note: The format of the pattern may not be compatible between
257 * the various search algorithms.
259 * Returns a new textsearch configuration according to the specified
260 * parameters or a ERR_PTR(). If a zero length pattern is passed, this
261 * function returns EINVAL.
263 struct ts_config
*textsearch_prepare(const char *algo
, const void *pattern
,
264 unsigned int len
, gfp_t gfp_mask
, int flags
)
267 struct ts_config
*conf
;
271 return ERR_PTR(-EINVAL
);
273 ops
= lookup_ts_algo(algo
);
274 #ifdef CONFIG_MODULES
276 * Why not always autoload you may ask. Some users are
277 * in a situation where requesting a module may deadlock,
278 * especially when the module is located on a NFS mount.
280 if (ops
== NULL
&& flags
& TS_AUTOLOAD
) {
281 request_module("ts_%s", algo
);
282 ops
= lookup_ts_algo(algo
);
289 conf
= ops
->init(pattern
, len
, gfp_mask
, flags
);
300 module_put(ops
->owner
);
304 EXPORT_SYMBOL(textsearch_prepare
);
307 * textsearch_destroy - destroy a search configuration
308 * @conf: search configuration
310 * Releases all references of the configuration and frees
313 void textsearch_destroy(struct ts_config
*conf
)
316 if (conf
->ops
->destroy
)
317 conf
->ops
->destroy(conf
);
318 module_put(conf
->ops
->owner
);
323 EXPORT_SYMBOL(textsearch_destroy
);