xref: /freebsd/contrib/unbound/util/module.h (revision 6ae1554a5d9b318f8ad53ccc39fa5a961403da73)
1 /*
2  * util/module.h - DNS handling module interface
3  *
4  * Copyright (c) 2007, NLnet Labs. All rights reserved.
5  *
6  * This software is open source.
7  *
8  * Redistribution and use in source and binary forms, with or without
9  * modification, are permitted provided that the following conditions
10  * are met:
11  *
12  * Redistributions of source code must retain the above copyright notice,
13  * this list of conditions and the following disclaimer.
14  *
15  * Redistributions in binary form must reproduce the above copyright notice,
16  * this list of conditions and the following disclaimer in the documentation
17  * and/or other materials provided with the distribution.
18  *
19  * Neither the name of the NLNET LABS nor the names of its contributors may
20  * be used to endorse or promote products derived from this software without
21  * specific prior written permission.
22  *
23  * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
24  * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
25  * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
26  * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
27  * HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
28  * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED
29  * TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
30  * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
31  * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
32  * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
33  * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
34  */
35 
36 /**
37  * \file
38  *
39  * This file contains the interface for DNS handling modules.
40  *
41  * The module interface uses the DNS modules as state machines.  The
42  * state machines are activated in sequence to operate on queries.  Once
43  * they are done, the reply is passed back.  In the usual setup the mesh
44  * is the caller of the state machines and once things are done sends replies
45  * and invokes result callbacks.
46  *
47  * The module provides a number of functions, listed in the module_func_block.
48  * The module is inited and destroyed and memory usage queries, for the
49  * module as a whole, for entire-module state (such as a cache).  And per-query
50  * functions are called, operate to move the state machine and cleanup of
51  * the per-query state.
52  *
53  * Most per-query state should simply be allocated in the query region.
54  * This is destroyed at the end of the query.
55  *
56  * The module environment contains services and information and caches
57  * shared by the modules and the rest of the system.  It also contains
58  * function pointers for module-specific tasks (like sending queries).
59  *
60  * *** Example module calls for a normal query
61  *
62  * In this example, the query does not need recursion, all the other data
63  * can be found in the cache.  This makes the example shorter.
64  *
65  * At the start of the program the iterator module is initialised.
66  * The iterator module sets up its global state, such as donotquery lists
67  * and private address trees.
68  *
69  * A query comes in, and a mesh entry is created for it.  The mesh
70  * starts the resolution process.  The validator module is the first
71  * in the list of modules, and it is started on this new query.  The
72  * operate() function is called.  The validator decides it needs not do
73  * anything yet until there is a result and returns wait_module, that
74  * causes the next module in the list to be started.
75  *
76  * The next module is the iterator.  It is started on the passed query and
77  * decides to perform a lookup.  For this simple example, the delegation
78  * point information is available, and all the iterator wants to do is
79  * send a UDP query.  The iterator uses env.send_query() to send the
80  * query.  Then the iterator suspends (returns from the operate call).
81  *
82  * When the UDP reply comes back (and on errors and timeouts), the
83  * operate function is called for the query, on the iterator module,
84  * with the event that there is a reply.  The iterator decides that this
85  * is enough, the work is done.  It returns the value finished from the
86  * operate call, which causes the previous module to be started.
87  *
88  * The previous module, the validator module, is started with the event
89  * that the iterator module is done.  The validator decides to validate
90  * the query.  Once it is done (which could take recursive lookups, but
91  * in this example no recursive lookups are needed), it returns from the
92  * operate function with finished.
93  *
94  * There is no previous module from the validator module, and the mesh
95  * takes this to mean that the query is finally done.  The mesh invokes
96  * callbacks and sends packets to queriers.
97  *
98  * If other modules had been waiting (recursively) on the answer to this
99  * query, then the mesh will tell them about it.  It calls the inform_super
100  * routine on all the waiting modules, and once that is done it calls all of
101  * them with the operate() call.  During inform_super the query that is done
102  * still exists and information can be copied from it (but the module should
103  * not really re-entry codepoints and services).  During the operate call
104  * the modules can use stored state to continue operation with the results.
105  * (network buffers are used to contain the answer packet during the
106  * inform_super phase, but after that the network buffers will be cleared
107  * of their contents so that other tasks can be performed).
108  *
109  * *** Example module calls for recursion
110  *
111  * A module is called in operate, and it decides that it wants to perform
112  * recursion.  That is, it wants the full state-machine-list to operate on
113  * a different query.  It calls env.attach_sub() to create a new query state.
114  * The routine returns the newly created state, and potentially the module
115  * can edit the module-states for the newly created query (i.e. pass along
116  * some information, like delegation points).  The module then suspends,
117  * returns from the operate routine.
118  *
119  * The mesh meanwhile will have the newly created query (or queries) on
120  * a waiting list, and will call operate() on this query (or queries).
121  * It starts again at the start of the module list for them.  The query
122  * (or queries) continue to operate their state machines, until they are
123  * done.  When they are done the mesh calls inform_super on the module that
124  * wanted the recursion.  After that the mesh calls operate() on the module
125  * that wanted to do the recursion, and during this phase the module could,
126  * for example, decide to create more recursions.
127  *
128  * If the module decides it no longer wants the recursive information
129  * it can call detach_subs.  Those queries will still run to completion,
130  * potentially filling the cache with information.  Inform_super is not
131  * called any more.
132  *
133  * The iterator module will fetch items from the cache, so a recursion
134  * attempt may complete very quickly if the item is in cache.  The calling
135  * module has to wait for completion or eventual timeout.  A recursive query
136  * that times out returns a servfail rcode (servfail is also returned for
137  * other errors during the lookup).
138  *
139  * Results are passed in the qstate, the rcode member is used to pass
140  * errors without requiring memory allocation, so that the code can continue
141  * in out-of-memory conditions.  If the rcode member is 0 (NOERROR) then
142  * the dns_msg entry contains a filled out message.  This message may
143  * also contain an rcode that is nonzero, but in this case additional
144  * information (query, additional) can be passed along.
145  *
146  * The rcode and dns_msg are used to pass the result from the the rightmost
147  * module towards the leftmost modules and then towards the user.
148  *
149  * If you want to avoid recursion-cycles where queries need other queries
150  * that need the first one, use detect_cycle() to see if that will happen.
151  *
152  */
153 
154 #ifndef UTIL_MODULE_H
155 #define UTIL_MODULE_H
156 #include "util/storage/lruhash.h"
157 #include "util/data/msgreply.h"
158 #include "util/data/msgparse.h"
159 struct sldns_buffer;
160 struct alloc_cache;
161 struct rrset_cache;
162 struct key_cache;
163 struct config_file;
164 struct slabhash;
165 struct query_info;
166 struct edns_data;
167 struct regional;
168 struct worker;
169 struct module_qstate;
170 struct ub_randstate;
171 struct mesh_area;
172 struct mesh_state;
173 struct val_anchors;
174 struct val_neg_cache;
175 struct iter_forwards;
176 struct iter_hints;
177 
178 /** Maximum number of modules in operation */
179 #define MAX_MODULE 5
180 
181 /**
182  * Module environment.
183  * Services and data provided to the module.
184  */
185 struct module_env {
186 	/* --- data --- */
187 	/** config file with config options */
188 	struct config_file* cfg;
189 	/** shared message cache */
190 	struct slabhash* msg_cache;
191 	/** shared rrset cache */
192 	struct rrset_cache* rrset_cache;
193 	/** shared infrastructure cache (edns, lameness) */
194 	struct infra_cache* infra_cache;
195 	/** shared key cache */
196 	struct key_cache* key_cache;
197 
198 	/* --- services --- */
199 	/**
200 	 * Send serviced DNS query to server. UDP/TCP and EDNS is handled.
201 	 * operate() should return with wait_reply. Later on a callback
202 	 * will cause operate() to be called with event timeout or reply.
203 	 * The time until a timeout is calculated from roundtrip timing,
204 	 * several UDP retries are attempted.
205 	 * @param qname: query name. (host order)
206 	 * @param qnamelen: length in bytes of qname, including trailing 0.
207 	 * @param qtype: query type. (host order)
208 	 * @param qclass: query class. (host order)
209 	 * @param flags: host order flags word, with opcode and CD bit.
210 	 * @param dnssec: if set, EDNS record will have bits set.
211 	 *	If EDNS_DO bit is set, DO bit is set in EDNS records.
212 	 *	If BIT_CD is set, CD bit is set in queries with EDNS records.
213 	 * @param want_dnssec: if set, the validator wants DNSSEC.  Without
214 	 * 	EDNS, the answer is likely to be useless for this domain.
215 	 * @param nocaps: do not use caps_for_id, use the qname as given.
216 	 *	(ignored if caps_for_id is disabled).
217 	 * @param addr: where to.
218 	 * @param addrlen: length of addr.
219 	 * @param zone: delegation point name.
220 	 * @param zonelen: length of zone name.
221 	 * @param q: wich query state to reactivate upon return.
222 	 * @return: false on failure (memory or socket related). no query was
223 	 *	sent. Or returns an outbound entry with qsent and qstate set.
224 	 *	This outbound_entry will be used on later module invocations
225 	 *	that involve this query (timeout, error or reply).
226 	 */
227 	struct outbound_entry* (*send_query)(uint8_t* qname, size_t qnamelen,
228 		uint16_t qtype, uint16_t qclass, uint16_t flags, int dnssec,
229 		int want_dnssec, int nocaps, struct sockaddr_storage* addr,
230 		socklen_t addrlen, uint8_t* zone, size_t zonelen,
231 		struct module_qstate* q);
232 
233 	/**
234 	 * Detach-subqueries.
235 	 * Remove all sub-query references from this query state.
236 	 * Keeps super-references of those sub-queries correct.
237 	 * Updates stat items in mesh_area structure.
238 	 * @param qstate: used to find mesh state.
239 	 */
240 	void (*detach_subs)(struct module_qstate* qstate);
241 
242 	/**
243 	 * Attach subquery.
244 	 * Creates it if it does not exist already.
245 	 * Keeps sub and super references correct.
246 	 * Updates stat items in mesh_area structure.
247 	 * Pass if it is priming query or not.
248 	 * return:
249 	 * o if error (malloc) happened.
250 	 * o need to initialise the new state (module init; it is a new state).
251 	 *   so that the next run of the query with this module is successful.
252 	 * o no init needed, attachment successful.
253 	 *
254 	 * @param qstate: the state to find mesh state, and that wants to
255 	 * 	receive the results from the new subquery.
256 	 * @param qinfo: what to query for (copied).
257 	 * @param qflags: what flags to use (RD, CD flag or not).
258 	 * @param prime: if it is a (stub) priming query.
259 	 * @param valrec: validation lookup recursion, does not need validation
260 	 * @param newq: If the new subquery needs initialisation, it is
261 	 * 	returned, otherwise NULL is returned.
262 	 * @return: false on error, true if success (and init may be needed).
263 	 */
264 	int (*attach_sub)(struct module_qstate* qstate,
265 		struct query_info* qinfo, uint16_t qflags, int prime,
266 		int valrec, struct module_qstate** newq);
267 
268 	/**
269 	 * Kill newly attached sub. If attach_sub returns newq for
270 	 * initialisation, but that fails, then this routine will cleanup and
271 	 * delete the fresly created sub.
272 	 * @param newq: the new subquery that is no longer needed.
273 	 * 	It is removed.
274 	 */
275 	void (*kill_sub)(struct module_qstate* newq);
276 
277 	/**
278 	 * Detect if adding a dependency for qstate on name,type,class will
279 	 * create a dependency cycle.
280 	 * @param qstate: given mesh querystate.
281 	 * @param qinfo: query info for dependency.
282 	 * @param flags: query flags of dependency, RD/CD flags.
283 	 * @param prime: if dependency is a priming query or not.
284 	 * @param valrec: validation lookup recursion, does not need validation
285 	 * @return true if the name,type,class exists and the given
286 	 * 	qstate mesh exists as a dependency of that name. Thus
287 	 * 	if qstate becomes dependent on name,type,class then a
288 	 * 	cycle is created.
289 	 */
290 	int (*detect_cycle)(struct module_qstate* qstate,
291 		struct query_info* qinfo, uint16_t flags, int prime,
292 		int valrec);
293 
294 	/** region for temporary usage. May be cleared after operate() call. */
295 	struct regional* scratch;
296 	/** buffer for temporary usage. May be cleared after operate() call. */
297 	struct sldns_buffer* scratch_buffer;
298 	/** internal data for daemon - worker thread. */
299 	struct worker* worker;
300 	/** mesh area with query state dependencies */
301 	struct mesh_area* mesh;
302 	/** allocation service */
303 	struct alloc_cache* alloc;
304 	/** random table to generate random numbers */
305 	struct ub_randstate* rnd;
306 	/** time in seconds, converted to integer */
307 	time_t* now;
308 	/** time in microseconds. Relatively recent. */
309 	struct timeval* now_tv;
310 	/** is validation required for messages, controls client-facing
311 	 * validation status (AD bits) and servfails */
312 	int need_to_validate;
313 	/** trusted key storage; these are the configured keys, if not NULL,
314 	 * otherwise configured by validator. These are the trust anchors,
315 	 * and are not primed and ready for validation, but on the bright
316 	 * side, they are read only memory, thus no locks and fast. */
317 	struct val_anchors* anchors;
318 	/** negative cache, configured by the validator. if not NULL,
319 	 * contains NSEC record lookup trees. */
320 	struct val_neg_cache* neg_cache;
321 	/** the 5011-probe timer (if any) */
322 	struct comm_timer* probe_timer;
323 	/** Mapping of forwarding zones to targets.
324 	 * iterator forwarder information. per-thread, created by worker */
325 	struct iter_forwards* fwds;
326 	/**
327 	 * iterator forwarder information. per-thread, created by worker.
328 	 * The hints -- these aren't stored in the cache because they don't
329 	 * expire. The hints are always used to "prime" the cache. Note
330 	 * that both root hints and stub zone "hints" are stored in this
331 	 * data structure.
332 	 */
333 	struct iter_hints* hints;
334 	/** module specific data. indexed by module id. */
335 	void* modinfo[MAX_MODULE];
336 };
337 
338 /**
339  * External visible states of the module state machine
340  * Modules may also have an internal state.
341  * Modules are supposed to run to completion or until blocked.
342  */
343 enum module_ext_state {
344 	/** initial state - new query */
345 	module_state_initial = 0,
346 	/** waiting for reply to outgoing network query */
347 	module_wait_reply,
348 	/** module is waiting for another module */
349 	module_wait_module,
350 	/** module is waiting for another module; that other is restarted */
351 	module_restart_next,
352 	/** module is waiting for sub-query */
353 	module_wait_subquery,
354 	/** module could not finish the query */
355 	module_error,
356 	/** module is finished with query */
357 	module_finished
358 };
359 
360 /**
361  * Events that happen to modules, that start or wakeup modules.
362  */
363 enum module_ev {
364 	/** new query */
365 	module_event_new = 0,
366 	/** query passed by other module */
367 	module_event_pass,
368 	/** reply inbound from server */
369 	module_event_reply,
370 	/** no reply, timeout or other error */
371 	module_event_noreply,
372 	/** reply is there, but capitalisation check failed */
373 	module_event_capsfail,
374 	/** next module is done, and its reply is awaiting you */
375 	module_event_moddone,
376 	/** error */
377 	module_event_error
378 };
379 
380 /**
381  * Linked list of sockaddrs
382  * May be allocated such that only 'len' bytes of addr exist for the structure.
383  */
384 struct sock_list {
385 	/** next in list */
386 	struct sock_list* next;
387 	/** length of addr */
388 	socklen_t len;
389 	/** sockaddr */
390 	struct sockaddr_storage addr;
391 };
392 
393 /**
394  * Module state, per query.
395  */
396 struct module_qstate {
397 	/** which query is being answered: name, type, class */
398 	struct query_info qinfo;
399 	/** flags uint16 from query */
400 	uint16_t query_flags;
401 	/** if this is a (stub or root) priming query (with hints) */
402 	int is_priming;
403 	/** if this is a validation recursion query that does not get
404 	 * validation itself */
405 	int is_valrec;
406 
407 	/** comm_reply contains server replies */
408 	struct comm_reply* reply;
409 	/** the reply message, with message for client and calling module */
410 	struct dns_msg* return_msg;
411 	/** the rcode, in case of error, instead of a reply message */
412 	int return_rcode;
413 	/** origin of the reply (can be NULL from cache, list for cnames) */
414 	struct sock_list* reply_origin;
415 	/** IP blacklist for queries */
416 	struct sock_list* blacklist;
417 	/** region for this query. Cleared when query process finishes. */
418 	struct regional* region;
419 	/** failure reason information if val-log-level is high */
420 	struct config_strlist* errinf;
421 
422 	/** which module is executing */
423 	int curmod;
424 	/** module states */
425 	enum module_ext_state ext_state[MAX_MODULE];
426 	/** module specific data for query. indexed by module id. */
427 	void* minfo[MAX_MODULE];
428 	/** environment for this query */
429 	struct module_env* env;
430 	/** mesh related information for this query */
431 	struct mesh_state* mesh_info;
432 	/** how many seconds before expiry is this prefetched (0 if not) */
433 	time_t prefetch_leeway;
434 };
435 
436 /**
437  * Module functionality block
438  */
439 struct module_func_block {
440 	/** text string name of module */
441 	const char* name;
442 
443 	/**
444 	 * init the module. Called once for the global state.
445 	 * This is the place to apply settings from the config file.
446 	 * @param env: module environment.
447 	 * @param id: module id number.
448 	 * return: 0 on error
449 	 */
450 	int (*init)(struct module_env* env, int id);
451 
452 	/**
453 	 * de-init, delete, the module. Called once for the global state.
454 	 * @param env: module environment.
455 	 * @param id: module id number.
456 	 */
457 	void (*deinit)(struct module_env* env, int id);
458 
459 	/**
460 	 * accept a new query, or work further on existing query.
461 	 * Changes the qstate->ext_state to be correct on exit.
462 	 * @param ev: event that causes the module state machine to
463 	 *	(re-)activate.
464 	 * @param qstate: the query state.
465 	 *	Note that this method is not allowed to change the
466 	 *	query state 'identity', that is query info, qflags,
467 	 *	and priming status.
468 	 *	Attach a subquery to get results to a different query.
469 	 * @param id: module id number that operate() is called on.
470 	 * @param outbound: if not NULL this event is due to the reply/timeout
471 	 *	or error on this outbound query.
472 	 * @return: if at exit the ext_state is:
473 	 *	o wait_module: next module is started. (with pass event).
474 	 *	o error or finished: previous module is resumed.
475 	 *	o otherwise it waits until that event happens (assumes
476 	 *	  the service routine to make subrequest or send message
477 	 *	  have been called.
478 	 */
479 	void (*operate)(struct module_qstate* qstate, enum module_ev event,
480 		int id, struct outbound_entry* outbound);
481 
482 	/**
483 	 * inform super querystate about the results from this subquerystate.
484 	 * Is called when the querystate is finished.  The method invoked is
485 	 * the one from the current module active in the super querystate.
486 	 * @param qstate: the query state that is finished.
487 	 *	Examine return_rcode and return_reply in the qstate.
488 	 * @param id: module id for this module.
489 	 *	This coincides with the current module for the super qstate.
490 	 * @param super: the super querystate that needs to be informed.
491 	 */
492 	void (*inform_super)(struct module_qstate* qstate, int id,
493 		struct module_qstate* super);
494 
495 	/**
496 	 * clear module specific data
497 	 */
498 	void (*clear)(struct module_qstate* qstate, int id);
499 
500 	/**
501 	 * How much memory is the module specific data using.
502 	 * @param env: module environment.
503 	 * @param id: the module id.
504 	 * @return the number of bytes that are alloced.
505 	 */
506 	size_t (*get_mem)(struct module_env* env, int id);
507 };
508 
509 /**
510  * Debug utility: module external qstate to string
511  * @param s: the state value.
512  * @return descriptive string.
513  */
514 const char* strextstate(enum module_ext_state s);
515 
516 /**
517  * Debug utility: module event to string
518  * @param e: the module event value.
519  * @return descriptive string.
520  */
521 const char* strmodulevent(enum module_ev e);
522 
523 #endif /* UTIL_MODULE_H */
524