1udns(3) Library Functions Manual udns(3)
3
7 udns - stub DNS resolver library
8
11 #include <udns.h>
12 struct dns_ctx;
13 struct dns_query;
14 extern struct dns_ctx dns_defctx;
15 struct dns_ctx *ctx;
16 typedef void dns_query_fn(ctx, void *result, void *data);
17 typedef int
18 dns_parse_fn(const unsigned char *qnd,
19 const unsigned char *pkt,
20 const unsigned char *cur,
21 const unsigned char *end,
22 void **resultp);
23 cc ... -ludns
25
28 The DNS library, udns, implements thread-safe stub DNS resolver func‐
29 tionality, which may be used both traditional, syncronous way and asyn‐
30 cronously, with application-supplied event loop.
31 While DNS works with both TCP and UDP, performing UDP query first and
34 if the result does not fit in UDP buffer (512 bytes max for original
35 DNS protocol), retrying the query over TCP, the library uses UDP only,
36 but uses EDNS0 (RFC2671) extensions which allows larger UDP buffers.
37 The library uses single UDP socket to perform all operations even when
40 asking multiple nameservers. This way, it is very simple to use the
41 library in asyncronous event-loop applications: an application should
42 add only single socket to the set of filedescriptors it monitors for
43 I/O.
44 The library uses two main objects, resolver context of type
47 struct dns_ctx, and query structure of type struct dns_query, both are
48 opaque for an application. Resolver context holds global information
49 about the resolver, such as list of nameservers to use, list of active
50 requests and the like. Query objects holds information about a single
51 DNS query in progress and are allocated/processed/freed by the library.
52 Pointer to query structure may be treated as an identifier of an in-
53 progress query and may be used to cancel the asyncronous query or to
54 wait for it to complete.
55 Asyncronous interface works as follows. An application initializes
58 resolver context, submits any number of queries for it using one of
59 supplied dns_submit_XXX() routines (each return the query identifier as
60 pointer to query structure), waits for input on the UDP socket used by
61 the library, and gives some control to the library by calling
62 dns_ioevent() and dns_timeouts() routines when appropriate. The
63 library performs all necessary processing and executes application sup‐
64 plied callback routine when a query completes (either successefully or
65 not), giving it the result if any, pointer to the resolver context
66 (from which completion status may be obtained), and the data pointer
67 supplied by an application when the query has been submitted. When
68 submitting a query, an application requests how to handle the reply --
69 to either return raw DNS reply packet for its own low-level processing,
70 or it may provide an address of parsing routine of type dns_parse_fn to
71 perform conversion of on-wire format into easy to use data structure
72 (the library provides parsing routines for several commonly used
73 resource record types, as well as type-safe higher-level inteface that
74 requests parsing automatically). The I/O monitoring and timeout han‐
75 dling may be either traditional select() or poll() based, or any call‐
76 back-driven technique may be used.
77 Additionally, the library provides traditional syncronous interface,
80 which may be intermixed with asyncronous calls (during syncronous query
81 processing, other asyncronous queries for the same resolver context
82 continued to be processed as usual). An application uses one of numer‐
83 ous dns_resolve_XXX() routines provided by the library to perform a
84 query. As with asyncronous interface, an application may either
85 request to return raw DNS packet or type-specific data structure by
86 providing the parsing routine to handle the reply. Every routine from
87 dns_resolve_XXX() series return pointer to result or NULL in case of
88 any error. Query completion status (or length of the raw DNS packet)
89 is available from the resolver context using dns_status() routine, the
90 same way as for the asyncronous interface.
91 Internally, library uses on-wire format of domain names, referred to as
94 DN format in this manual page. This is a series of domain labels whith
95 preceeding length byte, terminated by zero-length label wich is inte‐
96 gral part of the DN format. There are several routines provided to
97 convert from traditional asciiz string to DN and back. Higher-level
98 type-specific query interface hides the DN format from an application.
99
102 Every DNS Resource Record (RR) has a type and a class. The library
103 defines several integer constants, DNS_C_XXX and DNS_T_XXX, to use as
104 symbolic names for RR classes and types, such as DNS_C_IN for Internet
105 class, DNS_T_A for IPv4 address record type and so on. See udns.h
106 header file for complete list of all such constants.
107 The following constants are defined in udns.h header file:
110 DNS_MAXDN (255 bytes)
112 Maximum length of the domain name in internal (on-wire) DN for‐
113 mat.
114 DNS_MAXLABEL (63 bytes)
116 Maximum length of a single label in DN format.
117 DNS_MAXNAME (1024 bytes)
119 Maximum length of asciiz format of a domain name.
120 DNS_HSIZE (12 bytes)
122 Size of header in DNS packet.
123 DNS_PORT (53)
125 Default port to use when contacting a DNS server.
126 DNS_MAXSERV (6 servers)
128 Maximum number of DNS servers to use.
129 DNS_MAXPACKET (512 bytes)
131 Maximum length of DNS UDP packet as specified by original DNS
132 protocol
133 DNS_EDNS0PACKET (4096 bytes)
135 Default length of DNS UDP packet (with EDNS0 extensions) the
136 library uses. Note that recursive nameservers usually resides
137 near the client asking them to resolve names, e.g. on the same
138 LAN segment or even on the same host, so UDP packet fragmenta‐
139 tion isn't a problem in most cases. Note also that the size of
140 actual packets will be as many bytes as actual reply size
141 requires, which is smaller than this value in almost all cases.
142 Additionally, several constants are defined to simplify work with raw
145 DNS packets, such as DNS response codes (DNS_R_XXX), DNS header layout
146 (DNS_H_XXX) and others. Again, see udns.h for complete list. Library
147 error codes (DNS_E_XXX) are described later in this manual page.
148
151 Resolver context, of type struct dns_ctx, is an object which is opaque
152 to an application. Several routines provided by the library to ini‐
153 tialize, copy and free resolver contexts. Most other high-level rou‐
154 tines in this library expects a pointer to resolver context, ctx, as
155 the first argument. There is a default resolver context available,
156 named dns_defctx. When the context pointer ctx passed to a routine is
157 NULL, dns_defctx is used. Several resolver contexts may be active at
158 the same time, for example, when an application is multi-threaded and
159 each thread uses resolver.
160 In order to use the library, an application should initialize and open
162 one or more resolver context objects. These are two separate actions,
163 performed by dns_init() (or dns_reset()), and dns_open(). Between the
164 two calls, an application is free to pefrorm additional initialisation,
165 such as setting custom nameservers, options or domain search lists.
166 Optionally, in case no additional custom initialisation is required,
167 dns_init() may open the context if do_open argument (see below) is non-
168 zero.
169 When initializing resolver context, the library uses information from
171 system file /etc/resolv.conf (see resolv.conf(5)), consults environment
172 variables $LOCALDOMAIN, $DNSCACHEIP, $NAMESERVERS and $RES_OPTIONS, and
173 local host name to obtain list of local nameservers, domain name search
174 list and various resolver options.
175 The following routines to initialize resolver context are available:
177 void dns_reset(ctx)
179 int dns_init(ctx, int do_open)
180 dns_reset() resets a given resolver context to default values,
181 preparing it to be opened by dns_open(). It is ok to call this
182 routine against opened and active context - all active queries
183 will be dropped, sockets will be closed and so on. This routine
184 does not initialize any parameters from system configuration
185 files, use dns_init() for this. There's no error return - oper‐
186 ation always succeeds. dns_init() does everything dns_reset()
187 does, plus initializes various parameters of the context accord‐
188 ing to system configuration and process environment variables.
189 If do_open is non-zero, dns_init() calls dns_open(), so that the
190 whole library initialisation is performed in a single step.
191 struct dns_ctx *dns_new(struct dns_ctx *copy)
193 void dns_free(ctx)
194 dns_new() allocates new resolver context and copies all parame‐
195 ters for a given resolver context copy, or default context if
196 copy is NULL, and returns pointer to the newly allocated con‐
197 text. The context being copied should be initialized.
198 dns_new() may fail if there's no memory available to make a copy
199 of copy, in which case the routine will return NULL pointer.
200 dns_free() is used to close assotiated socket and free resolver
201 context resources and cancelling (abandoming) all active queries
202 assotiated with it. It's an error to free dns_defctx, only
203 dynamically allocated contexts returned by dns_new() are allowed
204 to be freed by dns_free().
205 int dns_add_serv(ctx, const char *servaddr)
207 int dns_add_serv_s(ctx, const struct sockaddr *sa)
208 int dns_add_srch(ctx, const char *srch)
209 Add an element to list of nameservers (dns_add_serv(), as
210 asciiz-string servaddr with an IP address of the nameserver, and
211 dns_add_serv_s(), as initialized socket address sa), or search
212 list (dns_add_srch(), as a pointer to domain name) for the given
213 context ctx. If the last argument is a NULL pointer, the corre‐
214 sponding list (search or nameserver) is reset instead. Upon
215 successeful completion, each routine returns new number of ele‐
216 ments in the list in question. On error, negative value is
217 returned and global variable errno is set appropriately. It is
218 an error to call any of this functions if the context is opened
219 (after dns_open() or dns_init() with non-zero argument).
220 int dns_set_opts(ctx, const char *opts)
222 set resolver context options from opts string, in the same way
223 as processing options statement in resolv.conf and $RES_OPTIONS
224 environment variable.
225 void dns_set_opt(ctx, int opt, val)
227 TODO The flags argument is a bitmask with the following bits
228 defined:
229 DNS_NOSRCH
231 do not perform domain name search in search list.
232 DNS_NORD
234 do not request recursion when performing queries (i.e.
235 don't set RD flag in querues).
236 DNS_AAONLY
238 request authoritative answers only (i.e. set AA flag in
239 queries).
240 int dns_open(ctx)
243 int dns_sock(const ctx)
244 void dns_close(ctx)
245 dns_open() opens the UDP socket used for queries if not already
246 open, and return assotiated filedescriptor (or negative value in
247 case of error). Before any query can be submitted, the context
248 should be opened using this routine. And before opening, the
249 context should be initialized. dns_sock() return the UDP socket
250 if open, or -1 if not. dns_close() closes the UDP socket if it
251 was open, and drops all active queries if any.
252 int dns_active(const ctx)
255 return number of active queries queued for the given context
256 ctx, or zero if none.
257 int dns_status(const ctx)
260 return status code from last operation. When using syncronous
261 interface, this is the query completion status of the last
262 query. With asyncronous interface, from within the callback
263 routine, this is the query completion status of the query for
264 which the callback is being called. When query submission
265 fails, this is the error code indicating failure reason. All
266 error codes are negative and are represented by DNS_E_XXX con‐
267 stants described below.
268 void dns_ioevent(ctx, time_t now)
271 this routine may be called by an application to process I/O
272 events on the UDP socket used by the library, as returned by
273 dns_sock(). The routine tries to receive incoming UDP datagram
274 from the socket and process it. The socket is set up to be non-
275 blocking, so it is safe to call the routine even if there's no
276 data to read. The routine will process as many datagrams as are
277 queued for the socket, so it is safe to use it with either
278 level-triggered or edge-triggered I/O monitoring model. The now
279 argument is either a current time as returned by time(), or 0,
280 in which case the routine will obtain current time by it's own.
281 int dns_timeouts(ctx, int maxwait, time_t now)
284 process any pending timeouts and return number of secounds from
285 current time (now if it is not 0) to the time when the library
286 wants the application to pass it control to process more queued
287 requests. In case when there are no requests pending, this time
288 is -1. The routine will not request a time larger than maxwait
289 secounds if it is greather or equal to zero. If now is 0, the
290 routine will obtain current time by it's own; when it is not 0,
291 it should contain current time as returned by time().
292 typedef void dns_utm_fn(ctx, int timeout, void *data)
295 void dns_set_cbck(ctx, dns_utm_fn *utmfn, void *data)
296 An application may use custom callback-based I/O multiplexing
297 mechanism. Usually such a mechanism have concept of a timer,
298 and an ability to register a timer event in a form of a callback
299 routine which will be executed after certain amount of time. In
300 order to use such an event mechanism, udns provides an ability
301 to register and de-register timer events necessary for internal
302 processing using whatever event mechanism an application uses.
303 For this to work, it is possible to assotiate a pointer to a
304 routine that will perform necessary work for (de)registering
305 timer events with a given resolver context, and udns will call
306 that routine at appropriate times. Prototype of such a routine
307 is shown by dns_utm_fn typedef above. Libudns assotiates single
308 timer with resolver context. User-supplied utmfn routine will
309 be called by the library with the following arguments:
310 ctx == NULL
312 delete user timer, at context free time or when an appli‐
313 cation changes user timer request routine using
314 dns_set_cbck();
315 ctx != NULL, timeout < 0
317 don't fire timer anymore, when there are no active
318 requests;
319 ctx != NULL, timeout == 0
321 fire timer at the next possibility, but not immediately;
322 ctx != NULL, timeout > 0
324 fire timer after timeout seconds after now.
325 The data argument passed to the routine will be the same as
327 passed to dns_set_cbck().
328 When a timer expires, an application should call dns_tmeouts()
330 routine (see below). Non-callback timer usage is provided too.
331 XXXX TODO: some more resolver context routines, like dns_set_dbgfn()
334 etc.
335
338 There are two ways to perform DNS queries: traditional syncronous way,
339 when udns performs all the necessary processing and return control to
340 the application only when the query completes, and asyncronous way,
341 when an application submits one or more queries to the library using
342 given resolver context, and waits for completion by monitoring
343 filedescriptor used by library and calling library routines to process
344 input on that filedescriptor. Asyncronous mode works with callback
345 routines: an application supplies an address of a routine to execute
346 when the query completes, and a data pointer, which is passed to the
347 callback routine.
348 Queries are submitted to the library in a form of struct dns_query. To
351 perform asyncronous query, an application calls one of the dns_sub‐
352 mit_XXX() rounines, and provides necessary information for a callback,
353 together with all the query parameters. When the query completes,
354 library will call application-supplied callback routine, giving it the
355 resolver context (wich holds query completion status), dynamically
356 allocated result (which will be either raw DNS packet or, if applicatin
357 requested parsing the result by specifying non-NULL parse routine,
358 ready-to-use type-specific structure), and a data pointer provided by
359 an application when it submitted the query. It is the application
360 who's responsible for freeing the result memory.
361 Generic query callback routine looks like this:
363 typedef void
364 dns_query_fn(ctx, void *result, void *data)
365 Type-specific query interface expects similar form of callback routine
366 with the only difference in type of result argument, which will be
367 pointer to specific data structure (decoded reply) instead of this void
368 pointer to raw DNS packet data.
369 Result parsing routine looks like this:
372 typedef int
373 dns_parse_fn(const unsigned char *qdn,
374 const unsigned char *pkt,
375 const unsigned char *cur,
376 const unsigned char *end,
377 void **resultp);
378 When called by the library, the arguments are as follows: pkt points to
379 the start of the packet received; end points past the end of the packet
380 received; cur points past the query DN in the query section of the
381 packet; qdn points to the original query DN. The routine should allo‐
382 cate a single buffer to hold the result, parse the reply filling in the
383 buffer, and return the buffer using resultp argument. It returns 0 in
384 case of error, or udns error code (DNS_E_XXX constants) in case of
385 error. Note that by the time when the parse routine is called by the
386 library, packet is already verified to be a reply to the original
387 query, by matching query DN, query class and query type.
388 Type-specific query inteface supplies necessary parsing routines auto‐
391 matically.
392 In case of error, query completion status as returned by dns_sta‐
395 tus(ctx), will contain one of the following values:
396 positive value
398 length of raw DNS packet if parsing is not requested.
399 0 the query was successeful and the reply points to type-specific
401 data structure.
402 DNS_E_TEMPFAIL
404 temporary error, the resolver nameserver was not able to process
405 our query or timed out.
406 DNS_E_PROTOCOL
408 protocol error, a nameserver returned malformed reply.
409 DNS_E_NXDOMAIN
411 the domain name does not exist.
412 DNS_E_NODATA
414 there is no data of requested type found.
415 DNS_E_NOMEM
417 out of memory while processing request.
418 DNS_E_BADQUERY
420 some aspect of the query (most common is the domain name in
421 question) is invalid, and the library can't even start a query.
422 Library provides two series of routines which uses similar interface --
425 one for asyncronous queries and another for syncronous queries. There
426 are two general low-level routines in each series to submit (asyn‐
427 cronous interface) and resolve (syncronous interface) queries, as well
428 as several type-specific routines with more easy-to-use interfaces. To
429 submit an asyncronous query, use one of dns_submit_XXX() routine, each
430 of which accepts query parameters, pointers to callback routine and to
431 callback data, and optional current time hint. Note type-specific
432 dns_submit_XXX() routines expects specific type of the callback routine
433 as well, which accepts reply as a pointer to corresponding structure,
434 not a void pointer). Every dns_submit_XXX() routine return pointer to
435 internal query structure of type struct dns_query, used as an identi‐
436 fier for the given query.
437 To resolve a query syncronously, use one of dns_resolve_XXX() routines,
440 which accepts the same query parameters (but not the callback pointers)
441 as corresponding dns_submit_XXX(), and return the query result, which
442 is the same as passed to the callback routine in case of asyncronous
443 interface.
444 In either case, the result memory (if the query completed successe‐
447 fully) is dynamically allocated and should be freed by an application.
448 If the query failed for any reason, the result will be NULL, and error
449 status will be available from dns_status(ctx) routine as shown above.
450 struct dns_query *
453 dns_submit_dn(ctx,
454 const unsigned char *dn, qcls, qtyp, flags,
455 parse, cbck, data)
456 struct dns_query *
457 dns_submit_p(ctx,
458 const char *name, qcls, qtyp, flags,
459 parse, cbck, data)
460 enum dns_class qcls;
461 enum dns_type qtyp;
462 int flags;
463 dns_parse_fn *parse;
464 dns_query_fn *cbck;
465 void *data;
466 submit a query for processing for the given resolver context
467 ctx. Two routines differs only in 3rd argument, which is domain
468 name in DN format (dn) or asciiz string (name). The query will
469 be performed for the given domain name, with type qtyp in class
470 qcls, using option bits in flags, using RR parsing routine
471 pointed by parse if not-NULL, and upon completion, cbck function
472 will be called with the data argument. In case of successeful
473 query submission, the routine return pointer to internal query
474 structure which may be treated as an identifier of the query as
475 used by the library, and may be used as an argument for dns_can‐
476 cel() routine. In case of error, NULL will be returned, and
477 context error status (available using dns_status() routine) will
478 be set to corresponding error code, which in this case may be
479 DNS_E_BADQUERY if the name of dn is invalid, DNS_E_NOMEM if
480 there's no memory available to allocate query structure, or
481 DNS_E_TEMPFAIL if an internal error occured.
482 void *dns_resolve_dn(ctx,
485 const unsigned char *dn, qcls, qtyp, flags, parse);
486 void *dns_resolve_p(ctx,
487 const char *name, qcls, qtyp, flags, parse)
488 enum dns_class qcls;
489 enum dns_type qtyp;
490 int flags;
491 dns_parse_fn *parse;
492 syncronous interface. The routines perform all the steps neces‐
493 sary to resolve the given query and return the result. If
494 there's no positive result for any reason, all the routines
495 return NULL, and set context error status (available using
496 dns_status() routine) to indicate the error code. If the query
497 was successeful, context status code will contain either the
498 length of the raw DNS reply packet if parse argument was NULL
499 (in which case the return value is pointer to the reply DNS
500 packet), or 0 (in which case the return value is the result of
501 parse routine). If the query successeful (return value is not
502 NULL), the memory returned was dynamically allocated by the
503 library and should be free()d by application after use.
504 void *dns_resolve(ctx, struct dns_query *q)
507 wait for the given query q, as returned by one of dns_sub‐
508 mit_XXX() routines, for completion, and return the result. The
509 callback routine will not be called for this query. After com‐
510 pletion, the query identifier q is not valid. Both
511 dns_resolve_dn() and dns_resolve_p() are just wrappers around
512 corresponding submit routines and this dns_resolve() routine.
513 void dns_cancel(ctx, struct dns_query *q)
516 cancel an active query q, without calling a callback routine.
517 After completion, the query identifier q is not valid.
518
521 In addition to the generic low-level query interface, the library pro‐
522 vides a set of routines to perform specific queries in a type-safe man‐
523 ner, as well as parsers for several well-known resource record types.
524 The library implements high-level interface for A, AAAA, PTR, MX and
525 TXT records and DNSBL and RHSBL functionality. These routines returns
526 specific types as result of a query, instead of raw DNS packets. The
527 following types and routines are available.
528 struct dns_rr_null {
531 char *dnsn_qname; /* original query name */
532 char *dnsn_cname; /* canonical name */
533 unsigned dnsn_ttl; /* Time-To-Live (TTL) value */
534 int dnsn_nrr; /* number of records in the set */
535 };
536 NULL RR set, used as a base for all other RR type structures. Every RR
538 structure as used by the library have four standard fields as in
539 struct dns_rr_null.
540 IN A Queries
543 struct dns_rr_a4 { /* IN A RRset */
544 char *dnsa4_qname; /* original query name */
545 char *dnsa4_cname; /* canonical name */
546 unsigned dnsa4_ttl; /* Time-To-Live (TTL) value */
547 int dnsa4_nrr; /* number of addresses in the set */
548 struct in_addr dnsa4_addr[]; /* array of addresses */
549 };
550 typedef void
551 dns_query_a4_fn(ctx, struct dns_rr_a4 *result, data)
552 dns_parse_fn dns_parse_a4;
553 struct dns_query *
554 dns_submit_a4(ctx, const char *name, int flags,
555 dns_query_a4_fn *cbck, data);
556 struct dns_rr_a4 *
557 dns_resolve_a4(ctx, const char *name, int flags);
558 The dns_rr_a4 structure holds a result of an IN A query, which is an
560 array of IPv4 addresses. Callback routine for IN A queries expected to
561 be of type dns_query_a4_fn, which expects pointer to dns_rr_a4 struc‐
562 ture as query result instead of raw DNS packet. The dns_parse_a4() is
563 used to convert raw DNS reply packet into dns_rr_a4 structure (it is
564 used internally and may be used directly too with generic query inter‐
565 face). Routines dns_submit_a4() and dns_resolve_a4() are used to per‐
566 form A IN queries in a type-safe manner. The name parameter is the
567 domain name in question, and flags is query flags bitmask, with one
568 bit, DNS_NOSRCH, of practical interest (if the name is absolute, that
569 is, it ends up with a dot, DNS_NOSRCH flag will be set automatically).
570 IN AAAA Queries
573 struct dns_rr_a6 { /* IN AAAA RRset */
574 char *dnsa6_qname; /* original query name */
575 char *dnsa6_cname; /* canonical name */
576 unsigned dnsa6_ttl; /* Time-To-Live (TTL) value */
577 int dnsa6_nrr; /* number of addresses in the set */
578 struct in6_addr dnsa6_addr[]; /* array of addresses */
579 };
580 typedef void
581 dns_query_a6_fn(ctx, struct dns_rr_a6 *result, data)
582 dns_parse_fn dns_parse_a6;
583 struct dns_query *
584 dns_submit_a6(ctx, const char *name, int flags,
585 dns_query_a6_fn *cbck, data);
586 struct dns_rr_a6 *
587 dns_resolve_a6(ctx, const char *name, int flags);
588 The dns_rr_a6 structure holds a result of an IN AAAA query, which is an
590 array of IPv6 addresses. Callback routine for IN AAAA queries expected
591 to be of type dns_query_a6_fn, which expects pointer to dns_rr_a6
592 structure as query result instead of raw DNS packet. The
593 dns_parse_a6() is used to convert raw DNS reply packet into dns_rr_a6
594 structure (it is used internally and may be used directly too with
595 generic query interface). Routines dns_submit_a6() and
596 dns_resolve_a6() are used to perform AAAA IN queries in a type-safe
597 manner. The name parameter is the domain name in question, and flags
598 is query flags bitmask, with one bit, DNS_NOSRCH, of practical interest
599 (if the name is absolute, that is, it ends up with a dot, DNS_NOSRCH
600 flag will be set automatically).
601 IN PTR Queries
604 struct dns_rr_ptr { /* IN PTR RRset */
605 char *dnsptr_qname; /* original query name */
606 char *dnsptr_cname; /* canonical name */
607 unsigned dnsptr_ttl; /* Time-To-Live (TTL) value */
608 int dnsptr_nrr; /* number of domain name pointers */
609 char *dnsptr_ptr[]; /* array of domain name pointers */
610 };
611 typedef void
612 dns_query_ptr_fn(ctx, struct dns_rr_ptr *result, data)
613 dns_parse_fn dns_parse_ptr;
614 struct dns_query *
615 dns_submit_a4ptr(ctx, const struct in_addr *addr,
616 dns_query_ptr_fn *cbck, data);
617 struct dns_rr_ptr *
618 dns_resolve_a4ptr(ctx, const struct in_addr *addr);
619 struct dns_query *
620 dns_submit_a6ptr(ctx, const struct in6_addr *addr,
621 dns_query_ptr_fn *cbck, data);
622 struct dns_rr_ptr *
623 dns_resolve_a6ptr(ctx, const struct in6_addr *addr);
624 The dns_rr_ptr structure holds a result of an IN PTR query, which is an
626 array of domain name pointers for a given IPv4 or IPv6 address. Call‐
627 back routine for IN PTR queries expected to be of type
628 dns_query_ptr_fn, which expects pointer to dns_rr_ptr structure as
629 query result instead of raw DNS packet. The dns_parse_ptr() is used to
630 convert raw DNS reply packet into dns_rr_ptr structure (it is used
631 internally and may be used directly too with generic query interface).
632 Routines dns_submit_a4ptr() and dns_resolve_a4ptr() are used to perform
633 IN PTR queries for IPv4 addresses in a type-safe manner. Routines
634 dns_submit_a6ptr() and dns_resolve_a6ptr() are used to perform IN PTR
635 queries for IPv6 addresses.
636 IN MX Queries
639 struct dns_mx { /* single MX record */
640 int priority; /* priority value of this MX */
641 char *name; /* domain name of this MX */
642 };
643 struct dns_rr_mx { /* IN MX RRset */
644 char *dnsmx_qname; /* original query name */
645 char *dnsmx_cname; /* canonical name */
646 unsigned dnsmx_ttl; /* Time-To-Live (TTL) value */
647 int dnsmx_nrr; /* number of mail exchangers in the set */
648 struct dns_mx dnsmx_mx[]; /* array of mail exchangers */
649 };
650 typedef void
651 dns_query_mx_fn(ctx, struct dns_rr_mx *result, data)
652 dns_parse_fn dns_parse_mx;
653 struct dns_query *
654 dns_submit_mx(ctx, const char *name, int flags,
655 dns_query_mx_fn *cbck, data);
656 struct dns_rr_mx *
657 dns_resolve_mx(ctx, const char *name, int flags);
658 The dns_rr_mx structure holds a result of an IN MX query, which is an
660 array of mail exchangers for a given domain. Callback routine for IN
661 MX queries expected to be of type dns_query_mx_fn, which expects
662 pointer to dns_rr_mx structure as query result instead of raw DNS
663 packet. The dns_parse_mx() is used to convert raw DNS reply packet
664 into dns_rr_mx structure (it is used internally and may be used
665 directly too with generic query interface). Routines dns_submit_mx()
666 and dns_resolve_mx() are used to perform IN MX queries in a type-safe
667 manner. The name parameter is the domain name in question, and flags
668 is query flags bitmask, with one bit, DNS_NOSRCH, of practical interest
669 (if the name is absolute, that is, it ends up with a dot, DNS_NOSRCH
670 flag will be set automatically).
671 TXT Queries
674 struct dns_txt { /* single TXT record */
675 int len; /* length of the text */
676 unsigned char *txt; /* pointer to the text */
677 };
678 struct dns_rr_txt { /* TXT RRset */
679 char *dnstxt_qname; /* original query name */
680 char *dnstxt_cname; /* canonical name */
681 unsigned dnstxt_ttl; /* Time-To-Live (TTL) value */
682 int dnstxt_nrr; /* number of text records in the set */
683 struct dns_txt dnstxt_txt[]; /* array of TXT records */
684 };
685 typedef void
686 dns_query_txt_fn(ctx, struct dns_rr_txt *result, data)
687 dns_parse_fn dns_parse_txt;
688 struct dns_query *
689 dns_submit_txt(ctx, const char *name, enum dns_class qcls,
690 int flags, dns_query_txt_fn *cbck, data);
691 struct dns_rr_txt *
692 dns_resolve_txt(ctx, const char *name,
693 enum dns_class qcls, int flags);
694 The dns_rr_txt structure holds a result of a TXT query, which is an
696 array of text records for a given domain name. Callback routine for
697 TXT queries expected to be of type dns_query_txt_fn, which expects
698 pointer to dns_rr_txt structure as query result instead of raw DNS
699 packet. The dns_parse_txt() is used to convert raw DNS reply packet
700 into dns_rr_txt structure (it is used internally and may be used
701 directly too with generic query interface). Routines dns_submit_txt()
702 and dns_resolve_txt() are used to perform IN MX queries in a type-safe
703 manner. The name parameter is the domain name in question, and flags
704 is query flags bitmask, with one bit, DNS_NOSRCH, of practical interest
705 (if the name is absolute, that is, it ends up with a dot, DNS_NOSRCH
706 flag will be set automatically). Note that each TXT string is repre‐
707 sented by struct dns_txt, while zero-terminated (and the len field of
708 the structure does not include the terminator), may contain embedded
709 null characters -- content of TXT records is not interpreted by the
710 library in any way.
711 SRV Queries
714 struct dns_srv { /* single SRV record */
715 int priority; /* priority of the record */
716 int weight; /* weight of the record */
717 int port; /* the port number to connect to */
718 char *name; /* target host name */
719 };
720 struct dns_rr_srv { /* SRV RRset */
721 char *dnssrv_qname; /* original query name */
722 char *dnssrv_cname; /* canonical name */
723 unsigned dnssrv_ttl; /* Time-To-Live (TTL) value */
724 int dnssrv_nrr; /* number of text records in the set */
725 struct dns_srv dnssrv_srv[]; /* array of SRV records */
726 };
727 typedef void
728 dns_query_srv_fn(ctx, struct dns_rr_srv *result, data)
729 dns_parse_fn dns_parse_srv;
730 struct dns_query *
731 dns_submit_srv(ctx, const char *name, const char *service, const char *protocol,
732 int flags, dns_query_txt_fn *cbck, data);
733 struct dns_rr_srv *
734 dns_resolve_srv(ctx, const char *name, const char *service, const char *protocol,
735 int flags);
736 The dns_rr_srv structure holds a result of an IN SRV (rfc2782) query,
738 which is an array of servers (together with port numbers) which are
739 performing operations for a given service using given protocol on a
740 target domain name. Callback routine for IN SRV queries expected to be
741 of type dns_query_srv_fn, which expects pointer to dns_rr_srv structure
742 as query result instead of raw DNS packet. The dns_parse_srv() is used
743 to convert raw DNS reply packet into dns_rr_srv structure (it is used
744 internally and may be used directly too with generic query interface).
745 Routines dns_submit_srv() and dns_resolve_srv() are used to perform IN
746 SRV queries in a type-safe manner. The name parameter is the domain
747 name in question, service and protocl specifies the service and the
748 protocol in question (the library will construct query DN according to
749 rfc2782 rules) and may be NULL (in this case the library assumes name
750 parameter holds the complete SRV query), and flags is query flags bit‐
751 mask, with one bit, DNS_NOSRCH, of practical interest (if the name is
752 absolute, that is, it ends up with a dot, DNS_NOSRCH flag will be set
753 automatically).
754 NAPTR Queries
757 struct dns_naptr { /* single NAPTR record */
758 int order; /* record order */
759 int preference; /* preference of this record */
760 char *flags; /* application-specific flags */
761 char *services; /* service parameters */
762 char *regexp; /* substitutional regular expression */
763 char *replacement; /* replacement string */
764 };
765 struct dns_rr_naptr { /* NAPTR RRset */
766 char *dnsnaptr_qname; /* original query name */
767 char *dnsnaptr_cname; /* canonical name */
768 unsigned dnsnaptr_ttl; /* Time-To-Live (TTL) value */
769 int dnsnaptr_nrr; /* number of text records in the set */
770 struct dns_naptr dnsnaptr_naptr[]; /* array of NAPTR records */
771 };
772 typedef void
773 dns_query_naptr_fn(ctx, struct dns_rr_naptr *result, data)
774 dns_parse_fn dns_parse_naptr;
775 struct dns_query *
776 dns_submit_naptr(ctx, const char *name, int flags,
777 dns_query_txt_fn *cbck, data);
778 struct dns_rr_naptr *
779 dns_resolve_naptr(ctx, const char *name, int flags);
780 The dns_rr_naptr structure holds a result of an IN NAPTR (rfc3403)
782 query. Callback routine for IN NAPTR queries expected to be of type
783 dns_query_naptr_fn, expects pointer to dns_rr_naptr structure as query
784 result instead of raw DNS packet. The dns_parse_naptr() is used to
785 convert raw DNS reply packet into dns_rr_naptr structure (it is used
786 internally and may be used directly too with generic query interface).
787 Routines dns_submit_naptr() and dns_resolve_naptr() are used to perform
788 IN NAPTR queries in a type-safe manner. The name parameter is the
789 domain name in question, and flags is query flags bitmask, with one
790 bit, DNS_NOSRCH, of practical interest (if the name is absolute, that
791 is, it ends up with a dot, DNS_NOSRCH flag will be set automatically).
792 DNSBL Interface
795 A DNS-based blocklists, or a DNSBLs, are in wide use nowadays, espe‐
796 cially to protect mailservers from spammers. The library provides
797 DNSBL interface, a set of routines to perform queries against DNSBLs.
798 Routines accepts an IP address (IPv4 and IPv6 are both supported) and a
799 base DNSBL zone as query parameters, and returns either dns_rr_a4 or
800 dns_rr_txt structure. Note that IPv6 interface return IPv4 RRset.
801 struct dns_query *
803 dns_submit_a4dnsbl(ctx,
804 const struct in_addr *addr, const char *dnsbl,
805 dns_query_a4_fn *cbck, void *data);
806 struct dns_query *
807 dns_submit_a4dnsbl_txt(ctx,
808 const struct in_addr *addr, const char *dnsbl,
809 dns_query_txt_fn *cbck, void *data);
810 struct dns_query *
811 dns_submit_a6dnsbl(ctx,
812 const struct in6_addr *addr, const char *dnsbl,
813 dns_query_a4_fn *cbck, void *data);
814 struct dns_query *
815 dns_submit_a6dnsbl_txt(ctx,
816 const struct in6_addr *addr, const char *dnsbl,
817 dns_query_txt_fn *cbck, void *data);
818 struct dns_rr_a4 *dns_resolve_a4dnsbl(ctx,
819 const struct in_addr *addr, const char *dnsbl)
820 struct dns_rr_txt *dns_resolve_a4dnsbl_txt(ctx,
821 const struct in_addr *addr, const char *dnsbl)
822 struct dns_rr_a4 *dns_resolve_a6dnsbl(ctx,
823 const struct in6_addr *addr, const char *dnsbl)
824 struct dns_rr_txt *dns_resolve_a6dnsbl_txt(ctx,
825 const struct in6_addr *addr, const char *dnsbl)
826 Perform (submit or resolve) a DNSBL query for the given dnsbl domain
827 and an IP addr in question, requesting either A or TXT records.
828 RHSBL Interface
831 RHSBL is similar to DNSBL, but instead of an IP address, the parameter
832 is a domain name.
833 struct dns_query *
835 dns_submit_rhsbl(ctx, const char *name, const char *rhsbl,
836 dns_query_a4_fn *cbck, void *data);
837 struct dns_query *
838 dns_submit_rhsbl_txt(ctx, const char *name, const char *rhsbl,
839 dns_query_txt_fn *cbck, void *data);
840 struct dns_rr_a4 *
841 dns_resolve_rhsbl(ctx, const char *name, const char *rhsbl);
842 struct dns_rr_txt *
843 dns_resolve_rhsbl_txt(ctx, const char *name, const char *rhsbl);
844 Perform (submit or resolve) a RHSBL query for the given rhsbl domain
845 and name in question, requesting either A or TXT records.
846
850 Domain Names (DNs)
851 A DN is a series of domain name labels each starts with length byte,
852 followed by empty label (label with zero length). The following rou‐
853 tines to work with DNs are provided.
854 unsigned dns_dnlen(const unsigned char *dn)
857 return length of the domain name dn, including the terminating
858 label.
859 unsigned dns_dnlabels(const unsigned char *dn)
862 return number of non-zero labels in domain name dn.
863 unsigned dns_dnequal(dn1, dn2)
866 const unsigned char *dn1, *dn2;
867 test whenever the two domain names, dn1 and dn2, are equal
868 (case-insensitive). Return domain name length if equal or 0 if
869 not.
870 unsigned dns_dntodn(sdn, ddn, dnsiz)
873 const unsigned char *sdn;
874 unsigned char *ddn;
875 unsigned dnsiz;
876 copies the source domain name sdn to destination buffer ddn of
877 size dnsiz. Return domain name length or 0 if ddn is too small.
878 int dns_ptodn(name, namelen, dn, dnsiz, isabs)
881 int dns_sptodn(name, dn, dnsiz)
882 const char *name; unsigned namelen;
883 unsigned char *dn; unsigned dnsiz;
884 int *isabs;
885 convert asciiz name name of length namelen to DN format, placing
886 result into buffer dn of size dnsiz. Return length of the DN if
887 successeful, 0 if the dn buffer supplied is too small, or nega‐
888 tive value if name is invalid. If isabs is non-NULL and conver‐
889 sion was successeful, *isabs will be set to either 1 or 0
890 depending whenever name was absolute (i.e. ending with a dot) or
891 not. Name length, namelength, may be zero, in which case
892 strlen(name) will be used. Second form, dns_sptodn(), is a sim‐
893 plified form of dns_ptodn(), equivalent to
894 dns_ptodn(name, 0, dn, dnlen, 0).
895 extern const unsigned char dns_inaddr_arpa_dn[]
898 int dns_a4todn(const struct in_addr *addr, const unsigned char *tdn,
899 unsigned char *dn, unsigned dnsiz)
900 int dns_a4ptodn(const struct in_addr *addr, const char *tname,
901 unsigned char *dn, unsigned dnsiz)
902 extern const unsigned char dns_ip6_arpa_dn[]
903 int dns_a6todn(const struct in6_addr *addr, const unsigned char *tdn,
904 unsigned char *dn, unsigned dnsiz)
905 int dns_a6ptodn(const struct in6_addr *addr, const char *tname,
906 unsigned char *dn, unsigned dnsiz)
907 several variants of routines to convert IPv4 and IPv6 address
908 addr into reverseDNS-like domain name in DN format, storing
909 result in dn of size dnsiz. tdn (or tname) is the base zone
910 name, like in-addr.arpa for IPv4 or in6.arpa for IPv6. If tdn
911 (or tname) is NULL, dns_inaddr_arpa_dn (or dns_ip6_arpa_dn) will
912 be used. The routines may be used to construct a DN for a DNSBL
913 lookup for example. All routines return length of the resulting
914 DN on success, -1 if resulting DN is invalid, or 0 if the dn
915 buffer (dnsiz) is too small. To hold standard rDNS DN, a buffer
916 of size DNS_A4RSIZE (30 bytes) for IPv4 address, or DNS_A6RSIZE
917 (74 bytes) for IPv6 address, is sufficient.
918 int dns_dntop(dn, name, namesiz)
921 const unsigned char *dn;
922 const char *name; unsigned namesiz;
923 convert domain name dn in DN format to asciiz string, placing
924 result into name buffer of size namesiz. Maximum length of
925 asciiz representation of domain name is DNS_MAXNAME (1024)
926 bytes. Root domain is represented as empty string. Return
927 length of the resulting name (including terminating character,
928 i.e. strlen(name)+1) on success, 0 if the name buffer is too
929 small, or negative value if dn is invalid (last case should
930 never happen since all routines in this library which produce
931 domain names ensure the DNs generated are valid).
932 const char *dns_dntosp(const unsigned char *dn)
935 convert domain name dn in DN format to asciiz string using
936 static buffer. Return the resulting asciiz string on success or
937 NULL on failure. Note since this routine uses static buffer, it
938 is not thread-safe.
939 unsigned dns_dntop_size(const unsigned char *dn)
942 return the buffer size needed to convert the dn domain name in
943 DN format to asciiz string, for dns_dntop(). The routine return
944 either the size of buffer required, including the trailing zero
945 byte, or 0 if dn is invalid.
946 Working with DNS Packets
949 The following routines are provided to encode and decode DNS on-wire
950 packets. This is low-level interface.
951 DNS response codes (returned by dns_rcode() routine) are defined as
954 constants prefixed with DNS_R_. See udns.h header file for the com‐
955 plete list. In particular, constants DNS_R_NOERROR [22m(0), DNS_R_SERV‐
956 FAIL, DNS_R_NXDOMAIN may be of interest to an application.
957 unsigned dns_get16(const unsigned char *p)
960 unsigned dns_get32(const unsigned char *p)
961 helper routines, convert 16-bit or 32-bit integer in on-wire
962 format pointed to by p to unsigned.
963 unsigned char *dns_put16(unsigned char *d, unsigned n)
966 unsigned char *dns_put32(unsigned char *d, unsigned n)
967 helper routine, convert unsigned 16-bit or 32-bit integer n to
968 on-wire format to buffer pointed to by d, return d+2 or d+4.
969 DNS_HSIZE (12)
972 defines size of DNS header. Data section in the DNS packet
973 immediately follows the header. In the header, there are query
974 identifier (id), various flags and codes, and number of resource
975 records in various data sections. See udns.h header file for
976 complete list of DNS header definitions.
977 unsigned dns_qid(const unsigned char *pkt)
980 int dns_rd(const unsigned char *pkt)
981 int dns_tc(const unsigned char *pkt)
982 int dns_aa(const unsigned char *pkt)
983 int dns_qr(const unsigned char *pkt)
984 int dns_ra(const unsigned char *pkt)
985 unsigned dns_opcode(const unsigned char *pkt)
986 unsigned dns_rcode(const unsigned char *pkt)
987 unsigned dns_numqd(const unsigned char *pkt)
988 unsigned dns_numan(const unsigned char *pkt)
989 unsigned dns_numns(const unsigned char *pkt)
990 unsigned dns_numar(const unsigned char *pkt)
991 const unsigned char *dns_payload(const unsigned char *pkt)
992 return various parts from the DNS packet header pkt: query iden‐
993 tifier (qid), recursion desired (rd) flag, truncation occured
994 (tc) flag, authoritative answer (aa) flag, query response (qr)
995 flag, recursion available (ra) flag, operation code (opcode),
996 result code (rcode), number of entries in question section
997 (numqd), number of answers (numan), number of authority records
998 (numns), number of additional records (numar), and the pointer
999 to the packet data (payload).
1000 int dns_getdn(pkt, curp, pkte, dn, dnsiz)
1003 const unsigned char *dns_skipdn(cur, pkte)
1004 const unsigned char *pkt, *pkte, **curp, *cur;
1005 unsigned char *dn; unsigned dnsiz;
1006 dns_getdn() extract DN from DNS packet pkt which ends before
1007 pkte starting at position *curp into buffer pointed to by dn of
1008 size dnsiz. Upon successeful completion, *curp will point to
1009 the next byte in the packet after the extracted domain name. It
1010 return positive number (length of the DN if dn) upon successeful
1011 completion, negative value on error (when the packet contains
1012 invalid data), or zero if the dnsiz is too small (maximum length
1013 of a domain name is DNS_MAXDN). dns_skipdn() return pointer to
1014 the next byte in DNS packet which ends up before pkte after a
1015 domain name which starts at the cur byte, or NULL if the packet
1016 is invalid. dns_skipdn() is more or less equivalent to what
1017 dns_getdn() does, except it does not actually extract the domain
1018 name in question, and uses simpler interface.
1019 struct dns_rr {
1022 unsigned char dnsrr_dn[DNS_MAXDN]; /* the RR DN name */
1023 enum dns_class dnsrr_cls; /* class of the RR */
1024 enum dns_type dnsrr_typ; /* type of the RR */
1025 unsigned dnsrr_ttl; /* TTL value */
1026 unsigned dnsrr_dsz; /* size of data in bytes */
1027 const unsigned char *dnsrr_dptr; /* pointer to the first data byte */
1028 const unsigned char *dnsrr_dend; /* next byte after RR */
1029 };
1030 The dns_rr structure is used to hold information about single
1031 DNS Resource Record (RR) in an easy to use form.
1032 struct dns_parse {
1035 const unsigned char *dnsp_pkt; /* pointer to the packet being parsed */
1036 const unsigned char *dnsp_end; /* end of the packet pointer */
1037 const unsigned char *dnsp_cur; /* current packet positionn */
1038 const unsigned char *dnsp_ans; /* pointer to the answer section */
1039 int dnsp_rrl; /* number of RRs left */
1040 int dnsp_nrr; /* number of relevant RRs seen so far */
1041 unsigned dnsp_ttl; /* TTL value so far */
1042 const unsigned char *dnsp_qdn; /* the domain of interest or NULL */
1043 enum dns_class dnsp_qcls; /* class of interest or 0 for any */
1044 enum dns_type dnsp_qtyp; /* type of interest or 0 for any */
1045 unsigned char dnsp_dnbuf[DNS_MAXDN]; /* domain name buffer */
1046 };
1047 The dns_parse structure is used to parse DNS reply packet. It
1048 holds information about the packet being parsed (dnsp_pkt,
1049 dnsp_end and dnsp_cur fields), number of RRs in the current sec‐
1050 tion left to do, and the information about specific RR which
1051 we're looking for (dnsp_qdn, dnsp_qcls and dnsp_qtyp fields).
1052 int dns_initparse(struct dns_parse *p,
1055 const unsigned char *qdn,
1056 const unsigned char *pkt,
1057 const unsigned char *cur,
1058 const unsigned char *end)
1059 initializes the RR parsing structure p. Arguments pkt, cur and
1060 end should describe the received packet: pkt is the start of the
1061 packet, end points to the next byte after the end of the packet,
1062 and cur points past the query DN in query section (to query
1063 class+type information). And qdn points to the query DN. This
1064 is the arguments passed to dns_parse_fn() routine. dns_init‐
1065 parse() initializes dnsp_pkt, dnsp_end and dnsp_qdn fields to
1066 the corresponding arguments, extracts and initializes dnsp_qcls
1067 and dnsp_qtyp fields to the values found at cur pointer, ini‐
1068 tializes dnsp_cur and dnsp_ans fields to be cur+4 (to the start
1069 of answer section), and initializes dnsp_rrl field to be number
1070 of entries in answer section. dnsp_ttl will be set to max TTL
1071 value, 0xffffffff, and dnsp_nrr to 0.
1072 int dns_nextrr(struct dns_parse *p, struct dns_rr *rr);
1075 searches for next RR in the packet based on the criteria pro‐
1076 vided in the p structure, filling in the rr structure and
1077 advancing p->dnsp_cur to the next RR in the packet. RR selec‐
1078 tion is based on dnsp_qdn, dnsp_qcls and dnsp_qtyp fields in the
1079 dns_parse structure. Any (or all) of the 3 fields may be 0,
1080 which means any actual value from the packet is acceptable. In
1081 case the field isn't 0 (or NULL for dnsp_qdn), only RRs with
1082 corresponding characteristics are acceptable. Additionally,
1083 when dnsp_qdn is non-NULL, dns_nextrr() performs automatic CNAME
1084 expansion. Routine will return positive value on success, 0 in
1085 case it reached the end of current section in the packet
1086 (p->dnsp_rrl is zero), or negative value if next RR can not be
1087 decoded (packet format is invalid). The routine updates
1088 p->dnsp_qdn automatically when this field is non-NULL and it
1089 encounters appropriate CNAME RRs (saving CNAME target in
1090 p->dnsp_dnbuf), so after end of the process, p->dnsp_qdn will
1091 point to canonical name of the domain in question. The routine
1092 updates p->dnsp_ttl value to be the minimum TTL of all RRs
1093 found.
1094 void dns_rewind(struct dns_parse *p, const unsigned char *qdn)
1097 this routine "rewinds" the packet parse state structure to be at
1098 the same state as after a call to dns_initparse(), i.e. reposi‐
1099 tion the parse structure p to the start of answer section and
1100 initialize p->dnsp_rrl to the number of entries in answer sec‐
1101 tion.
1102 int dns_stdrr_size(const struct dns_parse *p);
1105 return size to hold standard RRset structure information, as
1106 shown in dns_rr_null structure (for the query and canonical
1107 names). Used to calculate amount of memory to allocate for com‐
1108 mon part of type-specific RR structures in parsing routines.
1109 void *dns_stdrr_finish(struct dns_rr_null *ret, char *cp,
1112 const struct dns_parse *p);
1113 initializes standard RRset fields in ret structure using buffer
1114 pointed to by cp, which should have at least as many bytes as
1115 dns_stdrr_size(p) returned. Used to finalize common part of
1116 type-specific RR structures in parsing routines.
1117 See library source for usage examples of all the above low-level rou‐
1120 tines, especially source of the parsing routines.
1121 Auxilary Routines
1124 int dns_pton(int af, const char *src, void *dst);
1125 privides functionality similar to standard inet_pton() routine,
1126 to convert printable representation of an IP address of family
1127 af (either AF_INET or AF_INET6) pointed to by src into binary
1128 form suitable for socket addresses and transmission over net‐
1129 work, in buffer pointed to by dst. The destination buffer
1130 should be of size 4 for AF_INET family or 16 for AF_INET6. The
1131 return value is positive on success, 0 if src is not a valid
1132 text representation of an address of family af, or negative if
1133 the given address family is not supported.
1134 const char *dns_ntop(int af, const void *src,
1137 char *dst, int dstsize)
1138 privides functionality similar to standard inet_ntop() routine,
1139 to convert binary representation of an IP address of family af
1140 (either AF_INET or AF_INET6) pointed to by src (either 4 or 16
1141 bytes) into printable form in buffer in buffer pointed to by dst
1142 of size dstsize. The destination buffer should be at least of
1143 size 16 bytes for AF_INET family or 46 bytes for AF_INET6. The
1144 return value is either dst, or NULL pointer if dstsize is too
1145 small to hold this address or if the given address family is not
1146 supported.
1147
1150 The udns library has been written by Michael Tokarev, mjt@corpit.ru.
1151
1154 This manual page corresponds to udns version 0.0.9, released Jan-2007.
1155Library Functions Jan 2007 udns(3)