Compare commits

...
Author SHA1 Message Date
Yorgos Thessalonikefs f6269baa60 - Additional fix for CVE-2025-11411 (possible domain hijacking attack),
to include YXDOMAIN and non-referral nodata answers in the mitigation as
  well, reported by TaoFei Guo from Peking University, Yang Luo and JianJun
  Chen from Tsinghua University.
2025-11-26 11:09:40 +01:00
Yorgos Thessalonikefs 19154c6e58 - Set version to 1.24.2. 2025-11-26 10:58:06 +01:00
Yorgos Thessalonikefs a33f0638e1 - Fix CVE-2025-11411 (possible domain hijacking attack), reported by Yuxiao Wu,
Yunyi Zhang, Baojun Liu and Haixin Duan from Tsinghua University.
2025-10-22 10:54:57 +02:00
Yorgos Thessalonikefs bbeee42e25 - Set version to 1.24.1. 2025-10-22 10:50:18 +02:00
Yorgos Thessalonikefs 2dd821c257 - Too many quotes for the EDE message debug printout. 2025-09-17 11:27:16 +02:00
W.C.A. Wijngaards 713b5db537 - Fix to print warning for when so-sndbuf setsockopt is not granted. 2025-09-15 16:11:27 +02:00
Yorgos Thessalonikefs c3a8d5251f - Small debug output improvement when attaching an EDE. 2025-09-15 12:06:49 +02:00
W.C.A. Wijngaards d71ead5598 - Update contrib/aaaa-filter-iterator.patch so it applies on 1.24.0. 2025-09-11 13:23:51 +02:00
W.C.A. Wijngaards cdbfadfb7b - version set to 1.24.0 for release.
- tag for 1.24.0rc1.
2025-09-11 09:03:40 +02:00
W.C.A. Wijngaards 4267de87b5 - Fix #1332: CNAME chains are sometimes not followed when RPZs add a
local CNAME rewrite.
2025-09-09 12:34:11 +02:00
Yorgos Thessalonikefs dd6200065f - Update man pages. 2025-09-08 14:50:27 +02:00
Yorgos Thessalonikefs a72177e73c - Update documentation for using "SET ... EX" in Redis.
- Document max buffer sizes for Redis commands.
2025-09-08 14:49:12 +02:00
W.C.A. Wijngaards 5588f66bdb - For #1328: make depend. 2025-09-03 14:04:24 +02:00
W.C.A. Wijngaards 85e916e7e0 - Fix indentation in tcp-mss option parsing. 2025-09-02 17:12:14 +02:00
W.C.A. Wijngaards af96824642 - Fix #1324: Memory leak in 'msgparse.c' in
'parse_edns_options_from_query(...)'.
2025-09-02 17:10:42 +02:00
W.C.A. Wijngaards 8faa95354d - Fix #1235: Outdated Python2 code in
unbound/pythonmod/examples/log.py.
2025-09-02 12:54:03 +02:00
W.C.A. Wijngaards c57c39833e - Fix for #1324: Fix to free edns options scratch in ratelimit case. 2025-09-01 09:28:29 +02:00
Yorgos Thessalonikefs 44da5eee66 - Limit the number of consecutive reads on an HTTP/2 session.
Thanks to Gal Bar Nahum for exposing the possibility of infinite
  reads on the session.
2025-08-29 15:35:32 +02:00
W.C.A. Wijngaards 74bc8c9e77 - Fix setup_listen_sslctx warning for nettle compile. 2025-08-28 14:19:52 +02:00
W.C.A. Wijngaards 0c558cb805 - Fix unbound-control dump_cache for double unlock of lruhash table.
Changelog entry.
2025-08-27 16:55:55 +02:00
W.C.A. Wijngaards 533291dce9 - Fix unbound-control dump_cache for double unlock of lruhash table. 2025-08-27 16:55:38 +02:00
W.C.A. Wijngaards e613e27f35 - Fix ports workflow to install expat for macos. 2025-08-26 14:41:13 +02:00
W.C.A. Wijngaards f2f36a2733 - Fix that the zone acquired timestamp is set after the
zonefile is read.
2025-08-22 14:06:51 +02:00
W.C.A. Wijngaards 78d9bcacb6 - Fix #1319: [FR] zone status for Unbound auth-zones. 2025-08-22 12:40:00 +02:00
W.C.A. Wijngaards c170ed1b30 - Fix sha1 enable environment variable in test code on windows. 2025-08-22 10:04:57 +02:00
W.C.A. Wijngaards ebfa09e04f - For #1318: Fix compile warnings for DoH compile on windows. 2025-08-22 10:04:00 +02:00
W.C.A. Wijngaards 65be1d0ec3 - Fix for #1317: Fix contrib/unbound.service comment path for
systemd network configuration.
2025-08-21 15:49:42 +02:00
W.C.A. Wijngaards ca36e21f71 - Fix #1317: Unbound starts too early. Add
Wants=network-online.target under [Unit] in unbound.service.
2025-08-21 15:14:42 +02:00
W.C.A. Wijngaards 458c793012 - Fix to check for extraneous command arguments for unbound-control,
when the command takes no arguments but there are arguments present.
Changelog note for the fix.
2025-08-21 10:00:41 +02:00
W.C.A. Wijngaards 1341905de8 - Fix to check for extraneous command arguments for unbound-control,
when the command takes no arguments but there are arguments present.
2025-08-21 09:47:24 +02:00
W.C.A. Wijngaards 1d877400ea - Fix cache_lookup subnet print to not print messages without rrsets
and perform in-depth check on node in the addrtree.
2025-08-15 16:04:34 +02:00
W.C.A. Wijngaards 8fd4b91afc - Fix cache_lookup subnet print to not print messages without rrsets
and perform in-depth check on node in the addrtree.
2025-08-15 16:04:23 +02:00
W.C.A. Wijngaards 523710f371 - Fix cache_lookup subnet printout to wipe zero part of the prefix.
Changelog entry.
2025-08-15 14:07:05 +02:00
W.C.A. Wijngaards 13bb78a740 - Fix cache_lookup subnet printout to wipe zero part of the prefix. 2025-08-15 14:06:54 +02:00
W.C.A. Wijngaards 1e37f86ef5 - unbound-control cache_lookup +t allows tld and root names. And
subnet cache contents are printed.
Changelog, documentation and unit test.
2025-08-15 13:03:31 +02:00
W.C.A. Wijngaards 81345fe1e3 - unbound-control cache_lookup +t allows tld and root names. And
subnet cache contents are printed.
2025-08-15 13:03:00 +02:00
W.C.A. Wijngaards 40877f46e5 - Fix to decouple file descriptor activity and cache lookups in
dump_cache.
Changelog note.
2025-08-14 12:20:22 +02:00
W.C.A. Wijngaards 991108af8d - Fix to decouple file descriptor activity and cache lookups in
dump_cache.
2025-08-14 12:20:10 +02:00
W.C.A. Wijngaards 4f790bd65e - Fix to increase responsiveness of dump_cache. 2025-08-14 11:25:40 +02:00
W.C.A. Wijngaards d122ae6490 - Fix to unlock cache_lookup message for malformed records. 2025-08-13 12:02:41 +02:00
W.C.A. Wijngaards 651a71fa76 - Fix to remove debug from cache_lookup. 2025-08-13 11:59:53 +02:00
W.C.A. Wijngaards 2f7890eb6e - unbound-control cache_lookup <domains> prints the cached rrsets
and messages for those.
Changelog and information.
2025-08-13 11:36:47 +02:00
W.C.A. Wijngaards fad747308f - unbound-control cache_lookup <domains> prints the cached rrsets
and messages for those.
2025-08-13 11:31:42 +02:00
W.C.A. Wijngaards d55f20fdcc - Fix that unbound-control dump_cache releases the cache locks
every so often, so that the server stays responsive.
Changelog entry for it.
2025-08-12 12:00:01 +02:00
W.C.A. Wijngaards 19b289feec - Fix that unbound-control dump_cache releases the cache locks
every so often, so that the server stays responsive.
2025-08-12 11:59:48 +02:00
W.C.A. Wijngaards 752a3f7f52 - Fix to whitespace in dname_str. 2025-08-07 16:19:10 +02:00
W.C.A. Wijngaards 3ec5d78ac9 - Fix that edns-subnet failure to create a subquery errors as
servfail, and not formerror.
2025-08-07 16:09:47 +02:00
W.C.A. Wijngaards 08d59c9a78 - Fix dname_str for printout of long names. Thanks to Jan Komissar
for the fix.
2025-08-07 09:45:02 +02:00
W.C.A. Wijngaards cdcc0337d1 - Fix edns subnet, so that the subquery without subnet is stored in
global cache if the querier used 0.0.0.0/0 and the name and address
  do not receive subnet treatment. If the name and address are
  configured for subnet, it is stored in the subnet cache.
2025-08-06 12:08:44 +02:00
W.C.A. Wijngaards 305adf12bf - Fix edns subnet, so that the subquery without subnet is stored in
global cache if the querier used 0.0.0.0/0 and the name and address
  do not receive subnet treatment. If the name and address are
  configured for subnet, it is stored in the subnet cache.
2025-08-06 12:01:42 +02:00
W.C.A. Wijngaards 3d7e847a5e - Fix to use assertions for consistency checks in #1309 reclaimed 2025-08-05 16:20:01 +02:00
W.C.A. Wijngaards da6b735ed9 - Fix #1309: incorrectly reclaimed tcp handler can cause data
corruption and segfault.
2025-08-05 15:46:54 +02:00
W.C.A. Wijngaards 5758427d86 - Fix testbound test program to accurately output packets from hex. 2025-08-01 10:34:12 +02:00
W.C.A. Wijngaards 4f12148af4 - Fix redis cachedb module gettimeofday init failure.
Changelog note for the fix.
2025-07-28 09:33:42 +02:00
W.C.A. Wijngaards f8f4779f1f - Fix redis cachedb module gettimeofday init failure. 2025-07-28 09:32:55 +02:00
W.C.A. Wijngaards da9ab59e10 - Redis checks for server down and throttles reconnects.
And unit test for redis reconnect interval.
2025-07-24 11:06:00 +02:00
W.C.A. Wijngaards 424f86466a - Redis checks for server down and throttles reconnects. 2025-07-24 11:05:25 +02:00
W.C.A. Wijngaards 910288c0d1 - iana portlist updated. 2025-07-17 14:50:29 +02:00
W.C.A. Wijngaards b6e52c0a52 - Fix #1303: [FR] Disable TLSv1.2. 2025-07-17 14:50:13 +02:00
W.C.A. Wijngaards b58af78b63 - Fix to not set rlimits in the unit tests. 2025-07-17 11:40:31 +02:00
W.C.A. Wijngaards a1150078f2 - Add unit tests for non-ecs aggregation. 2025-07-16 11:46:04 +02:00
W.C.A. Wijngaards f49e6ccecd - Fix for RebirthDay Attack CVE-2025-5994, reported by Xiang Li
from AOSP Lab Nankai University.
- Tag for 1.23.1 with the release of 1.23.0 and the CVE fix, the
  repository continues with the previous fixes, with 1.23.2.
2025-07-16 11:40:32 +02:00
W.C.A. Wijngaards ad0e0fc290 Merge branch 'branch-1.23.1' 2025-07-16 11:34:44 +02:00
W.C.A. Wijngaards 5bf82f2464 - Fix RebirthDay Attack CVE-2025-5994, reported by Xiang Li from AOSP
Lab Nankai University.
2025-07-16 10:02:01 +02:00
W.C.A. Wijngaards 20f862de51 - Set version to 1.23.1. 2025-07-16 10:00:39 +02:00
Yorgos Thessalonikefs 46823f7bc3 - Update man page templates from rst. 2025-07-12 17:37:54 +02:00
Yorgos Thessalonikefs dd3ac53ff1 - For #1289: add num.valops in the unbound-control man page. 2025-07-12 17:35:16 +02:00
Yorgos Thessalonikefs 1a6052fcac - For #1289: test num.valops in existing stat_values.tdir. 2025-07-12 17:33:43 +02:00
Yorgos Thessalonikefs 6689f0b4da Changelog entry for #1289:
- Merge #1289 from Roland van Rijswijk-Deij: Add extra statistic to
  track the number of signature validation operations.
  Adds 'num.valops' to extended statistics.
2025-07-12 16:35:02 +02:00
44ac818f87 Add extra statistic to track the number of signature validation operations (#1289)
* Add extra statistic to track the number of signature validation operations performed by the validator module

* Move validation operation statistic to mesh as suggested

* Fix NULL pointer dereference in case the mesh is not used (and is `NULL`)

Co-authored-by: Wouter Wijngaards <wcawijngaards@users.noreply.github.com>

* Fix NULL pointer dereference on qstate and qstate->env in unit test situation

---------

Co-authored-by: Wouter Wijngaards <wcawijngaards@users.noreply.github.com>
2025-07-12 16:29:38 +02:00
W.C.A. Wijngaards b4e12030e7 - For #1301: configure cant find SSL_is_quic in OpenSSL 3.5.1. 2025-07-11 15:58:50 +02:00
W.C.A. Wijngaards 9fe92d1119 - Fix detection of SSL_CTX_set_tmp_ecdh function. 2025-07-11 15:47:59 +02:00
W.C.A. Wijngaards 6ba2d6840b - Fix to improve dnstap discovery on Fedora. 2025-07-08 09:29:27 +02:00
W.C.A. Wijngaards 1de9d6ec66 - Fix layout of comm_point_udp_ancil_callback. 2025-07-03 15:57:49 +02:00
W.C.A. Wijngaards cb919d5126 - For #1300: implement sock-queue-timeout for FreeBSD as well. 2025-07-03 15:54:33 +02:00
W.C.A. Wijngaards 444c839474 - Fix #1300: Is 'sock-queue-timeout' a linux only feature. 2025-07-03 14:10:46 +02:00
W.C.A. Wijngaards 0276bcbceb Changelog note for #1299
- Generate ltmain.sh and configure again.
2025-07-02 10:57:23 +02:00
Jose Luis DuranandGitHub 41c55ffac1 Fix typos (#1299) 2025-07-02 10:50:49 +02:00
Yorgos Thessalonikefs 4200d23882 - For #1247, replay test (added tcp_transport to
outnet_serviced_query).
2025-06-25 14:02:47 +02:00
Yorgos Thessalonikefs e2814fe165 - For #1247, turn off fetch-policy for delegation when looking into
parent side name servers that may not update the addresses and hit
  NXNS limits.
2025-06-25 13:59:17 +02:00
Yorgos Thessalonikefs ca153f4657 - Fix #1247: forward-first: ssl handshake failed on root nameservers. 2025-06-25 13:56:50 +02:00
Yorgos Thessalonikefs ee607c0f72 Changelog entry for #1293:
- Fix #1293: EDE 6 is attached to insecure cached answers when client
  sends the CD bit.
2025-06-20 14:11:22 +02:00
Yorgos Thessalonikefs 2d90d5d729 - Fix #1293: EDE 6 is attached to insecure cached answers when client sends
the CD bit.
2025-06-20 14:09:30 +02:00
W.C.A. Wijngaards ce72770f61 - Fix rrset cache create allocation failure case. 2025-06-19 16:27:13 +02:00
W.C.A. Wijngaards de93a636da Changelog note for #1297
- Merge #1297: edns-subnet: fix NULL_AFTER_DEREF on subnetmod.
2025-06-19 16:24:54 +02:00
KostyaandGitHub d3f88d008f edns-subnet: fix NULL_AFTER_DEREF on subnetmod (#1297)
Found by static analyzer svace.
Static analyzer message: Redundant comparison with a NULL value at subnetmod.c:236 for pointer 'sn_env->subnet_msg_cache',
which was dereferenced at slabhash.c:228 by passing as 1st parameter to function 'slabhash_setmarkdel' at subnetmod.c:235.

Moved usage of sn_env->subnet_msg_cache in slabhash_setmarkdel after checking.

Signed-off-by: Konstantin Kamanin <bewflast@gmail.com>
2025-06-19 16:24:21 +02:00
W.C.A. Wijngaards a1d68cdc96 - Fix #1296: DNS over QUIC depends on a very outdated version of
ngtcp2. Fixed so it works with ngtcp2 1.13.0 and OpenSSL 3.5.0.
2025-06-19 14:39:45 +02:00
Yorgos Thessalonikefs 9201c75013 - Fix for consistent use of local zone CNAME alias for configured auth
zones. Now it also applies to downstream configured auth zones.
2025-06-17 15:03:29 +02:00
W.C.A. Wijngaards f066d6d453 - Fix #1295: Windows 32-bit binaries download seems to be missing dll
dependency.
2025-06-16 14:26:54 +02:00
W.C.A. Wijngaards a04bd5da29 - Fix to check control-interface addresses in unbound-checkconf. 2025-06-16 12:43:31 +02:00
W.C.A. Wijngaards e4cf7aeccf - Fix header return value description for skip_pkt_rrs and
parse_edns_from_query_pkt.
2025-06-12 12:17:01 +02:00
W.C.A. Wijngaards a8aa1dbbe1 - Fix conditional expressions with parentheses for bitwise and. 2025-06-11 16:42:43 +02:00
W.C.A. Wijngaards 9f29292839 - Fix bitwise operators in conditional expressions with parentheses. 2025-06-11 15:46:31 +02:00
W.C.A. Wijngaards 1cc1e0b89e - iana portlist updated. 2025-06-05 11:11:56 +02:00
W.C.A. Wijngaards 565bce670c - Fix comment for the dname_remove_label_limit_len function. 2025-06-05 11:11:32 +02:00
W.C.A. Wijngaards c0563f43b0 - Fix unbound-anchor certificate file read for line ends and end of
file.
2025-06-05 11:09:53 +02:00
Yorgos Thessalonikefs 81f3de4da2 - Small man page corrections for the 'disable-dnssec-lame-check' option. 2025-06-03 14:12:27 +02:00
W.C.A. Wijngaards ff7dfd52a2 - Fix #1288: [FR] Improve fuzzing of unbound by adapting the netbound
program.
2025-05-21 12:41:54 +02:00
Yorgos Thessalonikefs 342a0f48e3 - Add more checks about respip in unbound-checkconf.
Also fixes #310: unbound-checkconf not reporting RPZ configuration
  error.
2025-05-20 16:21:02 +02:00
Yorgos Thessalonikefs 71ac59e6f4 Changelog entry for #1285:
- Merge #1285:  RST man pages.
2025-05-20 12:23:32 +02:00
Yorgos ThessalonikefsandGitHub 6662f71732 RST man pages (#1285)
Introduce restructuredText man pages to sync the online and source code man page documentation.
The templated man pages (*.in) are still part of the repo but generated with docutils from their .rst counterpart.
Documentation on how to generate those (mainly for core developers) are in README.man.
2025-05-20 12:20:20 +02:00
W.C.A. Wijngaards 32644937b0 - Fix for cname chain length with qtype ANY and qname minimisation.
Thanks to Jim Greenwood from Nominet for the report.
2025-05-19 13:17:21 +02:00
W.C.A. Wijngaards 1634beb4a0 - Fix config of slab values when there is no config file. 2025-05-15 14:34:18 +02:00
W.C.A. Wijngaards 1ef7b4a246 - Adjusted so-sndbuf default to 4m. 2025-05-13 15:31:05 +02:00
W.C.A. Wijngaards 03772d10fb - Change default for so-sndbuf to 1m, to mitigate a cross-layer
issue where the UDP socket send buffers are exhausted waiting
  for ARP/NDP resolution. Thanks to Reflyable for the report.
2025-05-13 15:04:32 +02:00
Yorgos Thessalonikefs 9152c914af - Fix #1282: log-destaddr fail on long ipv6 addresses. 2025-05-13 11:02:58 +02:00
Yorgos Thessalonikefs a35ac5d82e - Fix #1284: NULL pointer deref in az_find_nsec_cover() (latent bug)
by adding a log_assert() to safeguard future development.
2025-05-13 11:00:23 +02:00
W.C.A. Wijngaards 21e3278400 - Fix #1283: Unsafe usage of atoi() while parsing the configuration
file.
2025-05-12 14:57:42 +02:00
Yorgos Thessalonikefs 8e1deede08 Changelog entry for #1280:
- Merge #1280: Fix auth nsec3 code. Fixes NSEC3 code to not break on
  broken auth zones that include unsigned out of zone (above apex)
  data. Could lead to hang while trying to prove a wildcard answer.
2025-05-12 14:29:11 +02:00
Yorgos ThessalonikefsandGitHub 4e23523d1a Fix auth nsec3 code (#1280)
- Fix NSEC3 code to not break on broken auth zones that include unsigned
  out of zone (above apex) data. Could lead to hang while trying to
  prove a wildcard answer.
  Reported by Dmitrii Kuvaiskii from Amazon Web Services.

- Tests for NSEC3 auth zones with out of zone data.
2025-05-12 14:26:47 +02:00
W.C.A. Wijngaards 8190526250 - Fix #1281: forward-zone "name: ." conflicts with auth-zone "name: ."
in 1.23.0, but worked in 1.22.0.
2025-05-09 16:01:41 +02:00
Yorgos Thessalonikefs 5dd14e2644 - Sync unbound and unbound-checkconf log output for unknown modules. 2025-05-05 14:47:12 +02:00
Yorgos Thessalonikefs b50faccb1d Changelog entry for #1276:
- Merge #1276: Auto-configure '-slabs' values.
2025-04-29 15:23:07 +02:00
fcc21885e4 Auto-configure '-slabs' values (#1276)
- Auto-configure '-slabs' values to a power of 2 value close to num-threads
  by default for multi-threaded environments.

Co-authored-by: Wouter Wijngaards <wcawijngaards@users.noreply.github.com>
2025-04-29 15:21:47 +02:00
W.C.A. Wijngaards a904a3a2c2 - Fix dnstap to use protoc. 2025-04-29 12:43:56 +02:00
W.C.A. Wijngaards c253c8367a - Fix for parallel build of dnstap protoc-c output. 2025-04-29 12:38:41 +02:00
Yorgos Thessalonikefs 0f95fae445 Changelog entry for #1275:
- Merge #1275: Use macros for the fr_check_changed* functions.
2025-04-28 15:50:03 +02:00
Yorgos ThessalonikefsandGitHub e5bbc36ae0 Use macros for the fr_check_changed* functions (#1275) 2025-04-28 15:48:45 +02:00
W.C.A. Wijngaards c88fa02c18 - Fix #1272: assertion failure testcode/unitverify.c:202. 2025-04-25 11:12:28 +02:00
W.C.A. Wijngaards e794234ac8 - Tag for 1.23.0rc2. This became the release of 1.23.0 on 24 April
2025. The code repository continues with 1.23.1 in development.
2025-04-24 10:17:45 +02:00
W.C.A. Wijngaards db53ebb798 Merge branch 'branch-1.23.0' 2025-04-24 10:14:02 +02:00
W.C.A. Wijngaards fe835f9d52 - Increase default to num-queries-per-thread: 2048, when unbound is
compiled with libevent. It makes saturation of the task queue more
  resource intensive and less practical. Thanks to Shiming Liu,
  Network and Information Security Lab, Tsinghua University for the
  report.
2025-04-16 12:03:08 +02:00
Yorgos Thessalonikefs 30c13d0351 Changelog entry for #1265:
- Merge #1265: Fix WSAPoll.
2025-04-11 15:10:46 +02:00
Yorgos Thessalonikefs d6e8ac1289 Fix WSAPoll (#1265)
* Fix calling WSAPoll.

* fast_reload: explicitly set tcp_wouldblock on Windows when there is no
  command to read from the fast_reload thread.

* For poll(), also check for ENOMEM (Linux).

* Remove ifdefs for ENOMEM.
* Some systems return EAGAIN for poll.
2025-04-11 15:09:07 +02:00
Yorgos Thessalonikefs 9c99b404a1 Changelog entry for #1265:
- Merge #1265: Fix WSAPoll.
2025-04-11 15:07:25 +02:00
Yorgos ThessalonikefsandGitHub 75e8fd7539 Fix WSAPoll (#1265)
* Fix calling WSAPoll.

* fast_reload: explicitly set tcp_wouldblock on Windows when there is no
  command to read from the fast_reload thread.

* For poll(), also check for ENOMEM (Linux).

* Remove ifdefs for ENOMEM.
* Some systems return EAGAIN for poll.
2025-04-11 15:05:52 +02:00
W.C.A. Wijngaards 16ee7cf944 - Fix for print of connection type in log-replies for dot and doh. 2025-04-10 09:33:51 +02:00
W.C.A. Wijngaards 38026a21ee Merge branch 'master' into branch-1.23.0 2025-04-09 14:20:22 +02:00
W.C.A. Wijngaards 4f06e658d1 - Fix #1264: unbound 1.22.0 leaks memory when doing DoH. 2025-04-09 14:13:58 +02:00
W.C.A. Wijngaards fca3ae0535 - Fix to detect if atomic_store links in configure. 2025-04-09 11:06:25 +02:00
W.C.A. Wijngaards 66b3abc803 Merge branch 'branch-1.23.0' of github.com:NLnetLabs/unbound into branch-1.23.0 2025-04-08 15:43:05 +02:00
W.C.A. Wijngaards 922c5c3ab3 - Fix fast_reload to print chroot with config file name. 2025-04-08 15:42:56 +02:00
Yorgos Thessalonikefs f9d740dfdc - Update to the manpage for the fast_reload part. 2025-04-08 15:34:51 +02:00
W.C.A. Wijngaards 5eb1382fc0 - Tag for 1.23.0rc1. 2025-04-08 08:37:30 +02:00
Yorgos Thessalonikefs e94e140f49 - More explicit text about memory usage during fast_reload. 2025-04-07 16:56:38 +02:00
Yorgos Thessalonikefs db6505eb8b Changelog entry for #902:
- Merge #902: DNS Error Reporting (RFC 9567). Introduces new
  configuration option 'dns-error-reporting' and new statistics for
  'num.dns_error_reports'.
2025-04-07 10:28:37 +02:00
a616437338 DNS Error Reporting (RFC 9567) (#902)
* v1 EDER poc

* remove superfluous edns_list_get_option function

* create an EDER configurable

* Hackathon 114

* Fixes for version -04

* Generated configparser and configlexer are not versioned in master anymore

* Remove NOERROR DNS Error Reporting; not part of final RFC.
* Use assigned IANA EDNS0 Option Code for Report-Channel.

* Fix buffer protection and agent domain validity

* Use DNS Error Reporting instead of the eder nickname

* Update documentation.

* Fix typo.

* Bail out early if ede is not present.

* Forget previous EDNS options from upstream; this is what was
  implicitly happening but not deterministacally.

* Don't report LDNS_EDE_OTHER and bail early if there is no reporting
  agent.

* Only do DNS error reporting when a client asked for something that
  went wrong.

* Add an error reporting agent in the parent that should be ignored.

* review feedback.

* fixup for fast reload

* Add 'num.dns_error_reports' to stats and test for it.

---------

Co-authored-by: TCY16 <tom@nlnetlabs.nl>
Co-authored-by: Yorgos Thessalonikefs <yorgos@nlnetlabs.nl>
2025-04-07 10:25:10 +02:00
W.C.A. Wijngaards eb390dd038 - Fix to update common.sh for speed of kill_pid. 2025-04-04 11:55:42 +02:00
W.C.A. Wijngaards 4074a5b2ba - Fix test to speed up common.sh script kill_pid. 2025-04-04 11:37:42 +02:00
W.C.A. Wijngaards a7e618b9c2 Fix typo in Changelog entry. 2025-04-04 11:00:16 +02:00
Yorgos Thessalonikefs 349721a39d Changelog entry for #1019:
- Merge #1019: Redis read-only replica support.
  Introduces new 'redis-replica-*' options for the Redis cache backend.
2025-04-04 10:58:59 +02:00
W.C.A. Wijngaards ed71f72baa - Fix redis_replicat test for unused option defaults and log printout. 2025-04-04 10:57:53 +02:00
W.C.A. Wijngaards 0e5d26807d - Fix nettle compile for warnings and ticket keys. 2025-04-04 10:30:22 +02:00
90243a694a Redis read-only replica support (#1019)
* Set version to 1.19.1 for point release.

* Initial work for Redis read-only replica support.

* Test for Redis replica.

* Documentation for the Redis replica timeouts.

* redis replica, rewrite set_timeout()

* clean merge.

* Add new options for fast reload.

* Apply suggestions from code review

Co-authored-by: Wouter Wijngaards <wcawijngaards@users.noreply.github.com>

* some more typos

---------

Co-authored-by: W.C.A. Wijngaards <wouter@nlnetlabs.nl>
Co-authored-by: Wouter Wijngaards <wcawijngaards@users.noreply.github.com>
2025-04-04 10:20:47 +02:00
W.C.A. Wijngaards 7fb6f9d613 - Fix comment name in the rpz nsdname test. 2025-04-04 08:58:54 +02:00
W.C.A. Wijngaards 7fb05c01c2 - Fix mesh_copy_client_info to omit null contents from copy. 2025-04-04 08:57:24 +02:00
W.C.A. Wijngaards cb5a019d9f - Fix parameter unused warning in net_help.c. 2025-04-03 15:21:16 +02:00
W.C.A. Wijngaards 53ecdb25ef - Fix test for stat_values for wait limit defaults for localhost. 2025-04-03 14:50:30 +02:00
W.C.A. Wijngaards 9939d5cdb0 - Fix ub_event and include dnstap and win_svc headers. 2025-04-03 10:59:41 +02:00
W.C.A. Wijngaards 5f91d5f74c - Fix wait-limit-netblock and wait-limit-cookie-netblock config parse
to allow two arguments.
2025-04-03 09:53:27 +02:00
W.C.A. Wijngaards c2ca679f5c - Fix #1263: Exempt loopback addresses from wait-limit. 2025-04-03 09:45:36 +02:00
W.C.A. Wijngaards ba18abcd35 - Fix that ub_event has the facility to deal with callbacks for
fast reload, doq, windows-stop and dnstap.
- Fix fast reload test to check if pid exists before acting on it.
2025-04-02 16:25:58 +02:00
W.C.A. Wijngaards a7704ad49f - Fix unbound-control test so it counts the new flush_negative output,
also answers the _ta probe from testns and prints command output
  and skip a thread specific test when no threads are available.
2025-04-02 15:52:20 +02:00
Yorgos Thessalonikefs 8b38fe5ab6 - For #1262, ifdef is no longer needed. 2025-04-02 11:53:50 +02:00
Yorgos Thessalonikefs fcdb0bbf91 Changelog entry for #1262:
- Merge #1262 from markyang92, fix build with
  'gcc-15 -Wbuiltin-declaration-mismatch' error in compat/malloc.c.
2025-04-02 11:52:04 +02:00
Yorgos ThessalonikefsandGitHub b738b354bd Merge pull request #1262 from markyang92/master
fix build with gcc-15 -Wbuiltin-declaration-mismatch error in compat/malloc.c
2025-04-02 11:49:49 +02:00
mark.yang 1c58ce0791 fix build with gcc-15 -Wbuiltin-declaration-mismatch error
See more details: http://errors.yoctoproject.org/Errors/Details/850313
../git/compat/malloc.c:9:7: warning: conflicting types for built-in function 'malloc'; expected 'void *(long unsigned int)' [-Wbuiltin-declaration-mismatch]
    9 | void *malloc ();
      |       ^~~~~~
../git/compat/malloc.c:5:1: note: 'malloc' is declared in header '<stdlib.h>'
    4 | #include "config.h"
  +++ |+#include <stdlib.h>
    5 | #undef malloc
../git/compat/malloc.c: In function 'rpl_malloc_unbound':
../git/compat/malloc.c:23:10: error: too many arguments to function 'malloc'; expected 0, have 1
   23 |   return malloc (n);
      |          ^~~~~~  ~
../git/compat/malloc.c:9:7: note: declared here
    9 | void *malloc ();
      |       ^~~~~~

* Seeing that there is '#undef malloc', it appears they don't want to
  use the malloc from stdlib.h.
  Therefore, we need to correctly define the parameters for malloc.

Signed-off-by: mark.yang <mark.yang@lge.com>
2025-04-02 15:25:42 +09:00
W.C.A. Wijngaards a7106fc58d - Enable the auth_tls.tdir and auth_tls_failcert.tdir tests. 2025-04-01 14:44:25 +02:00
W.C.A. Wijngaards 17f95ffac1 - Fix escape more characters when printing an RR type with an unquoted
string.
2025-04-01 13:53:29 +02:00
W.C.A. Wijngaards 0ca76b05e0 - Skip the unit tests for auth_tls.tdir and auth_tls_failcert.tdir. 2025-03-31 16:04:07 +02:00
218f5cfc92 Fast Reload Option (#1042)
* - fast-reload, add unbound-control fast_reload

* - fast-reload, make a thread to service the unbound-control command.

* - fast-reload, communication sockets for information transfer.

* - fast-reload, fix compile for unbound-dnstap-socket.

* - fast-reload, set nonblocking communication to keep the server thread
  responding to DNS requests.

* - fast-reload, poll routine to test for readiness, timeout fails connection.

* - fast-reload, detect loop in sock_poll_timeout routine.

* - fast-reload, send done and exited notification.

* - fast-reload, defines for constants in ipc.

* - fast-reload, ipc socket recv and send resists partial reads and writes and
  can continue byte by byte. Also it can continue after an interrupt.

* - fast-reload, send exit command to thread when done.

* - fast-reload, output strings for client on string list.

* - fast-reload, add newline to terminal output.

* - fast-reload, send client string to remote client.

* - fast-reload, better debug output.

* - fast-reload, print queue structure, for output to the remote client.

* - fast-reload, move print items to print queue from fast_reload_thread struct.

* - fast-reload, keep list of pending print queue items in daemon struct.

* - fast-reload, comment explains in_list for printq to print remainder.

* - fast-reload, unit test testdata/fast_reload_thread.tdir that tests the
  thread output.

* - fast-reload, fix test link for fast_reload_printq_list_delete function.

* - fast-reload, reread config file from disk.

* - fast-reload, unshare forwards, making the structure locked, with an rwlock.

* - fast-reload, for nonthreaded, the unbound-control commands forward,
  forward_add and forward_delete should be distributed to other processes,
  but when threaded, they should not be distributed to other threads because
  the structure is not thread specific any more.

* - fast-reload, unshared stub hints, making the structure locked, with an rwlock.

* - fast-reload, helpful comments for hints lookup function return value.

* - fast-reload, fix bug in fast reload printout, the strlist appendlist routine,
  and printout time statistics after the reload is done.

* - fast-reload, keep track of reloadtime and deletestime and print them.

* - fast-reload, keep track of constructtime and print it.

* - fast-reload, construct new items.

* - fast-reload, better comment.

* - fast-reload, reload the config and swap trees for forwards and stub hints.

* - fast-reload, in forwards_swap_tree set protection of trees with locks.

* - fast-reload, in hints_swap_tree also swap the node count of the trees.

* - fast-reload, reload ipc to stop and start threads.

* - fast-reload, unused forward declarations removed.

* - fast-reload, unit test that fast reload works with forwards and stubs.

* - fast-reload, fix clang analyzer warnings.

* - fast-reload, small documentation entry in unbound-control -h output.

* - fast-reload, printout memory use by fast reload, in bytes.

* - fast-reload, compile without threads.

* - fast-reload, document fast_reload in man page.

* - fast-reload, print ok when done successfully.

* - fast-reload, option for fast-reload commandline, +v verbosity option,
  with timing and memory use output.

* - fast-reload, option for fast-reload commandline, +p does not pause threads.

* - fast-reload, option for fast-reload commandline, +d drops mesh queries.

* - fast-reload, fix to poll every thread with nopause to make certain that
  resources are not held by the threads and can be deleted.

* - fast-reload, fix to use atomic store for config variables with nopause.

* - fast-reload, reload views.

* - fast-reload, when tag defines are different, it drops the queries.

* - fast-reload, fix tag define check.

* - fast-reload, document that tag change causes drop of queries.

* - fast-reload, fix space in documentation man page.

* - fast-reload, copy respip client information to query state, put views tree
  in module env for lookup.

* - fast-reload, nicer respip view comparison.

* - fast-reload, respip global set is in module env.

* - fast-reload, document that respip_client_info acl info is copied.

* - fast-reload, reload the respip_set.

* - fast-reload, document no pause and pick up of use_response_ip boolean.

* - fast-reload, fix test compile.

* - fast-reload, reload local zones.

* Update locking management for iter_fwd and iter_hints methods. (#1054)

fast reload, move most of the locking management to iter_fwd and
iter_hints methods. The caller still has the ability to handle its
own locking, if desired, for atomic operations on sets of different
structs.

Co-authored-by: Wouter Wijngaards <wcawijngaards@users.noreply.github.com>

* - fast-reload, reload access-control.

* - fast-reload, reload access control interface, such as interface-action.

* - fast-reload, reload tcp-connection-limit.

* - fast-reload, improve comments on acl_list and tcl_list swap tree.

* - fast-reload, fixup references to old tcp connection limits in open tcp
  connections.

* - fast-reload, fixup to clean tcp connection also for different linked order.

* - fast-reload, if no tcp connection limits existed, no need to remove
  references for that.

* - fast-reload, document more options that work and do not work.

* - fast-reload, reload auth_zone and rpz data.

* - fast-reload, fix auth_zones_get_mem.

* - fast-reload, fix compilation of testbound for the new comm_timer_get_mem
  reference in remote control.

* - fast-reload, change use_rpz with reload.

* - fast-reload, list changes in auth zones and stop zonemd callbacks for
  deleted auth zones.

* - fast-reload, note xtree is not swapped, and why it is not swapped.

* - fast-reload, for added auth zones, pick up zone transfer and zonemd tasks.

* - fast-reload, unlock xfr when done with transfer pick up.

* - fast-reload, unlock z when picking up the xfr for it during transfer task
  pick up.

* - fast-reload, pick up task changes for added, deleted and modified auth zones.

* - fast-reload, remove xfr of auth zone deletion without tasks.

* - fast-reload, pick up zone transfer config.

* - fast-reload, the main worker thread picks up the transfer tasks and also
  performs setup of the xfer struct.

* - fast-reload, keep writelock on newzone when auth zone changes.

* - fast-reload, change cachedb_enabled setting.

* - fast-reload, pick up edns-strings config.

* - fast-reload, note that settings are not updated.

* - fast-reload, pick up dnstap config.

* - fast-reload, dnstap options that need to be loaded without +p.

* - fast-reload, fix auth zone reload

* - fast-reload, remove debug for auth zone test.

* - fast-reload, fix auth zone reload with zone transfer.

* - fast-reload, fix auth zone reload lock order.

* - fast-reload, remove debug from fast reload test.

* - fast-reload, remove unused function.

* - fast-reload, fix the worker trust anchor probe timer lock acquisition in
  the probe answer callback routine for trust anchor probes.

* - fast-reload, reload trust anchors.

* - fast-reload, fix trust anchor reload lock on autr global data and test
  for trust anchor reload.

* - fast-reload, adjust cache sizes.

* - fast-reload, reload cache sizes when changed.

* - fast-reload, reload validator env changes.

* - fast-reload, reload mesh changes.

* - fast-reload, check for incompatible changes.

* - fast-reload, improve error text for incompatible change.

* - fast-reload, fix check config option compatibility.

* - fast-reload, improve error text for nopause change.

* - fast-reload, fix spelling of incompatible options.

* - fast-reload, reload target-fetch-policy, outbound-msg-retry, max-sent-count
  and max-query-restarts.

* - fast-reload, check nopause config change for target-fetch-policy.

* - fast-reload, reload do-not-query-address, private-address and capt-exempt.

* - fast-reload, check nopause config change for do-not-query-address,
  private-address and capt-exempt.

* - fast-reload, check fast reload not possible due to interface and
  outgoing-interface changes.

* - fast-reload, reload nat64 settings.

* - fast-reload, reload settings stored in the infra structure.

* - fast-reload, fix modstack lookup and remove outgoing-range check.

* - fast-reload, more explanation for config parse failure.

* - fast-reload, reload worker outside network changes.

* - fast-reload, detect incompatible changes in network settings.

* fast-reload, commit test files.

* - fast-reload, fix warnings for call types in windows compile.

* - fast-reload, fix warnings and comm_point_internal for tcp wouldblock calls.

* - fast-reload, extend lock checks for repeat thread ids.

* - fast-reload, additional test cases, cache change and tag changes.

* - fast-reload, fix documentation for auth_zone_verify_zonemd_with_key.

* - fast-reload, fix copy_cfg type casts and memory leak on config parse failure.

* - fast-reload, fix use of WSAPoll.

* Review comments for the fast reload feature (#1259)

* - fast-reload review, respip set can be null from a view.

* - fast-reload review, typos.

* - fast-reload review, keep clang static analyzer happy.

* - fast-reload review, don't forget to copy tag_actions.

* - fast-reload review, less indentation.

* - fast-reload review, don't leak respip_actions when reloading.

* - fast-reload review, protect NULL pointer dereference in get_mem
  functions.

* - fast-reload review, add fast_reload_most_options.tdir to test most
  options with high verbosity when fast reloading.

* - fast-reload review, don't skip new line on long error printouts.

* - fast-reload review, typo.

* - fast-reload review, use new_z for consistency.

* - fast-reload review, nit for unlock ordering to make eye comparison
  with the lock counterpart easier.

* - fast-reload review, in case of error the sockets are already closed.

* - fast-reload review, identation.

* - fast-reload review, add static keywords.

* - fast-reload review, update unbound-control usage text.

* - fast-reload review, updates to the man page.

* - fast-reload, the fast-reload command is experimental.

* - fast-reload, fix compile of doqclient for fast reload functions.

* Changelog comment for #1042
- Merge #1042: Fast Reload. The unbound-control fast_reload is added.
  It reads changed config in a thread, then only briefly pauses the
  service threads, that keep running. DNS service is only interrupted
  briefly, less than a second.

---------

Co-authored-by: Yorgos Thessalonikefs <yorgos@nlnetlabs.nl>
2025-03-31 15:25:24 +02:00
W.C.A. Wijngaards 914cef75f9 - iana portlist update. 2025-03-31 14:25:16 +02:00
W.C.A. Wijngaards f76365754e - Fix for ci test, expat is installed on the osx image. 2025-03-27 10:52:00 +01:00
W.C.A. Wijngaards aa77d02336 - Fix unit test dname log printout typecast. 2025-03-27 10:47:20 +01:00
Yorgos Thessalonikefs 23273d76a5 - For #1255, for ios try the latest expat version again. 2025-03-26 16:51:46 +01:00
Yorgos Thessalonikefs 94a84d3387 - For #1255, for ios disable building tests that require C++11. 2025-03-26 16:40:10 +01:00
Yorgos Thessalonikefs d91c857e22 - For #1255, for ios use an older expat version that does not require
C++11 language features.
2025-03-26 16:15:16 +01:00
Yorgos Thessalonikefs d03e4b1884 - Fix #1255: Multiple pinnings to vulnerable copies of libexpat. 2025-03-26 14:58:54 +01:00
W.C.A. Wijngaards 376f2ade2a - Fix #1254: send failed: Socket is not connected and
`remote address is 0.0.0.0 port 53`.
2025-03-24 08:54:54 +01:00
W.C.A. Wijngaards 0eabc8d0f1 - Fix for #1253: Fix for redis cachedb backend to expect an integer
reply for the EXPIRE command.
2025-03-21 14:07:22 +01:00
W.C.A. Wijngaards a42fb99508 - Fix #1253: Cache entries fail to be removed from Redis cachedb
backend with unbound-control flush* +c.
2025-03-21 12:56:21 +01:00
W.C.A. Wijngaards eccf52e39d - Fix print of RR type NSAP-PTR, it is an unquoted string. 2025-03-20 15:51:03 +01:00
W.C.A. Wijngaards 30b9cb5f81 - Fix for windows compile create ssl contexts. 2025-03-18 14:01:53 +01:00
W.C.A. Wijngaards b0d20e2d28 - Fix #1251: WSAPoll first argument cannot be NULL. 2025-03-18 13:59:53 +01:00
W.C.A. Wijngaards 13afde2cad - Fix representation of types GPOS and RESINFO, add rdf type for 2025-03-17 16:38:17 +01:00
Yorgos Thessalonikefs 71bb60e586 - Fix 'unbound-control flush_negative' when reporting removed data;
reported by David 'eqvinox' Lamparter.
2025-03-16 09:02:20 +01:00
W.C.A. Wijngaards 5c84bb573f Changelog nore for #1238 and add --help description.
- Merge #1238: Prefer SOURCE_DATE_EPOCH over actual time.
  Add --help output description for the SOURCE_DATE_EPOCH variable.
2025-02-28 09:32:49 +01:00
Sefa EyeogluandGitHub d9f1dae540 Prefer SOURCE_DATE_EPOCH over actual time (#1238)
* Add ax_build_date_epoch from Autoconf Archive

Signed-off-by: Sefa Eyeoglu <contact@scrumplex.net>

* Prefer SOURCE_DATE_EPOCH over actual time

Signed-off-by: Sefa Eyeoglu <contact@scrumplex.net>

---------

Signed-off-by: Sefa Eyeoglu <contact@scrumplex.net>
2025-02-28 09:22:19 +01:00
W.C.A. Wijngaards a3d750b7d8 Changelog note for #1243
- Merge #1243: Do not shadow tm on line 236.
2025-02-25 09:37:36 +01:00
Florian ObserandGitHub c371be3f9f Do not shadow tm on line 236. (#1243) 2025-02-25 09:36:44 +01:00
Yorgos Thessalonikefs c5c5486261 - Fix hash calculation for cachedb to ignore case. Previously, cached
records there were only relevant for same case queries (if not
  already in Unbound's internal cache).
2025-02-24 14:47:13 +01:00
Yorgos Thessalonikefs 1894c0a150 Changelog entry for #1241:
- Merge #1241: Fix infra-keep-probing for low infra-cache-max-rtt
  values.
2025-02-19 13:46:01 +01:00
Yorgos ThessalonikefsandGitHub edb20a2ad2 Merge pull request #1241 from NLnetLabs/bugfix/keep-probing-max-rtt
Fix infra-keep-probing for low infra-cache-max-rtt values
2025-02-19 13:44:40 +01:00
Yorgos Thessalonikefs bcf162abd7 - The maximum value of a probe rto was not aligned with the
(configurable) infra-cache-max-rtt value. That could result in
  infra-keep-probing not working if an infra-cache-max-rtt value was chosen
  that was below 12000 ms. This fix still uses a default value of 12000
  ms for the probe but caps it to the infra-cache-max-rtt if that is
  lower.
2025-02-19 12:14:59 +01:00
Yorgos Thessalonikefs 5e1f35b59b - Fix static analysis report about unhandled EOF on error conditions
when reading anchor key files.
2025-02-19 11:24:49 +01:00
Yorgos Thessalonikefs 72828ff81c - Consider reconfigurations when calculating the still_useful_timeout
for servers in the infrastructure cache.
2025-02-17 15:21:18 +01:00
W.C.A. Wijngaards 01cea4d5be - Fix #986: Resolving sas.com with dnssec-validation fails though
signed delegations seem to be (mostly) correct.
2025-01-30 16:26:31 +01:00
Yorgos Thessalonikefs 35dbbcb2f5 - Make the default value of module-config "validator iterator"
regardless of compilation options. --enable-subnet would implicitly
  change the value to enable the subnetcache module by default in the
  past.
2025-01-29 12:08:28 +01:00
Yorgos Thessalonikefs 911509fd59 Changelog entry for #1220:
- Merge #1220 from Petr Menšík, Add unbound members group access to
  control key.
2025-01-24 16:56:09 +01:00
Yorgos ThessalonikefsandGitHub b48958c983 Merge pull request #1220 from InfrastructureServices/unbound-control-group-key
Add unbound members group access to control key
2025-01-24 16:53:12 +01:00
Yorgos Thessalonikefs cc55beefc8 Changelog entry for #1224:
- Merge #1224 from Theo Buehler: Do not use DSA API unless USE_DSA is
  set.
2025-01-21 17:35:00 +01:00
Yorgos ThessalonikefsandGitHub a2bf32bb4e Merge pull request #1224 from botovq/improve-use-dsa
Do not use DSA API unless USE_DSA is set
2025-01-21 17:33:48 +01:00
W.C.A. Wijngaards d9b863ed76 Changelog note for #1229
- Merge #1229: check before use daemon->shm_info.
2025-01-21 15:48:46 +01:00
eaglegaiandGitHub 073c7301eb check before use daemon->shm_info (#1229)
fix core after the command `unbound-control stop unbound`

fix:https://github.com/NLnetLabs/unbound/issues/1228

Signed-off-by: eaglegai <eaglegai@163.com>
2025-01-21 15:47:51 +01:00
Yorgos Thessalonikefs f822042cd0 - Do not open unencrypted channels next to encrypted ones on the same
port.
2025-01-21 15:26:40 +01:00
W.C.A. Wijngaards 5f58ced71e - Fix to check length in ATMA string to wire. 2025-01-21 12:30:30 +01:00
W.C.A. Wijngaards 207ae97ff9 - Fix encoding of RR type ATMA. 2025-01-21 12:27:15 +01:00
W.C.A. Wijngaards 9a0de14aa1 - Fix compile of interface check code when dnscrypt or quic is
disabled.
2025-01-21 10:13:48 +01:00
Yorgos Thessalonikefs 048c193243 - Use the same interface listening port discovery code for all needed
protocols.
- Port to string only when needed before getaddrinfo().
2025-01-21 10:04:30 +01:00
Yorgos Thessalonikefs d62fff2c7c - Create the quic SSL listening context only when needed. 2025-01-20 15:49:37 +01:00
Yorgos Thessalonikefs 3f839cebc3 Changelog entry for #1222:
- Merge #1222: Unique DoT and DoH SSL contexts to allow for different
  ALPN.
2025-01-20 15:45:11 +01:00
Yorgos ThessalonikefsandGitHub e4483bbbd1 Unique DoT and DoH SSL contexts to allow for different ALPN (#1222) 2025-01-20 15:43:44 +01:00
Theo Buehler 8672b34fca Do not use DSA API unless USE_DSA is set
Even if USE_DSA is unset, unbound ends up linking against OpenSSL
DSA API because these guards are missing.
2025-01-18 10:40:43 +01:00
Yorgos Thessalonikefs 1d428f2d54 Changelog entry for #1221:
- Merge #1221: Consider auth zones when checking for forwarders.
2025-01-17 10:19:26 +01:00
Yorgos ThessalonikefsandGitHub 9882a395ab Merge pull request #1221 from NLnetLabs/bugfix/consider-auth-zones-when-forwarding
Consider auth zones when checking for forwarders
2025-01-17 10:18:32 +01:00
Yorgos Thessalonikefs 394588818f - Use correct RFC number for resolver.arpa. 2025-01-15 10:55:31 +01:00
Yorgos Thessalonikefs f52b2a6ea2 - Add resolver.arpa and service.arpa to the default locally served
zones.
2025-01-14 17:18:32 +01:00
Yorgos Thessalonikefs b2fec3be11 - Take configured auth zones into consideration when checking if a
request needs to be forwarded.
2025-01-14 16:38:53 +01:00
Petr Menšík f4881bd81a Add unbound members group access to control key
Recent openssl genrsa does not use umask for generated keys. There is no
strong reason why every member of unbound group should be able read
server key. But control key would be quite useful to be group readable
and to allow control access to whole group. Allowing access to control
by group membership, not via sudo.
2025-01-14 14:35:09 +01:00
Yorgos Thessalonikefs c3b5bff311 - Fix typo. 2025-01-13 12:32:16 +01:00
Yorgos Thessalonikefs 62a0e03801 - Fix #1213: Misleading error message on default access control causing
refuse.
2025-01-13 11:33:24 +01:00
Yorgos Thessalonikefs 716f3df385 Changelog entry for #1214:
- Merge #1214: Use TCP_NODELAY on TLS sockets to speed up the TLS
  handshake.
2025-01-10 13:54:49 +01:00
Yorgos ThessalonikefsandGitHub 7e4f7ec5be Merge pull request #1214 from NLnetLabs/bugfix/tls-handshake
Use TCP_NODELAY on TLS sockets to speed up the TLS handshake.
2025-01-10 13:53:46 +01:00
Yorgos Thessalonikefs 7559d26c93 - Use TCP_NODELAY on TLS sockets to speed up the TLS handshake. 2025-01-10 12:11:59 +01:00
Yorgos Thessalonikefs eb36c880de Changelog entry for #1174:
- Merge #1174: Serve expired cache update fixes. Fixes a regression bug
  with serve-expired that appeared in 1.22.0 and would not allow the
  iterator to update the cache with not-yet-validated entries resulting
  in increased outgoing traffic.
2024-12-31 16:30:35 +01:00
Yorgos ThessalonikefsandGitHub fff9f62a1e Serve expired cache update fixes (#1174)
- Fixes a regression bug with serve-expired that appeared in 1.22.0
  and would not allow the iterator to update the cache with
  not-yet-validated entries resulting in increased outgoing traffic.

- Treat serve_expired_norec_ttl as a backoff timer for failed updates of expired records.
- Try to use expired answers instead of SERVFAIL if serve-expired is
  enabled even without serve-expired-client-timeout.
- Add suggestion to refresh the cached norec_ttl and expired_ttl when a
  response cannot update the usable expired entry.
2024-12-31 16:28:12 +01:00
Yorgos Thessalonikefs e57e537c85 - For #1207: [FR] Support for RESINFO RRType 261 (RFC9606), add
LDNS_RR_TYPE_RESINFO similar to LDNS_RR_TYPE_TXT.
2024-12-20 15:04:34 +01:00
Yorgos Thessalonikefs 71d821fde9 Changelog entry for #1204:
- Merge #1204: ci: set persist-credentials: false for actions/checkout
  per zizmor suggestion.
2024-12-13 13:43:29 +01:00
Yorgos ThessalonikefsandGitHub df5ab5624d Merge pull request #1204 from NLnetLabs/zizmor-improvements 2024-12-13 13:42:31 +01:00
Maarten Aertsen eb08dc617a set persist-credentials: false per zizmor suggestion 2024-12-13 13:12:03 +01:00
Yorgos Thessalonikefs ded4c82ced - Fix typo in log_servfail.tdir test. 2024-12-03 16:03:05 +01:00
Yorgos Thessalonikefs e82a691efe Changelog entry for #1187:
- Merge #1187: Create the SSL_CTX for QUIC before chroot and privilege
  drop.
2024-12-03 14:21:34 +01:00
61d7250b96 Create the SSL_CTX for QUIC before chroot and privilege drop (#1187)
Fixes #1185 by creating the SSL_CTX for QUIC before chroot and
privilege drop, just like the other SSL_CTX creations.

---------

Co-authored-by: Wouter Wijngaards <wcawijngaards@users.noreply.github.com>
2024-12-03 14:20:33 +01:00
Yorgos Thessalonikefs b4a9c8bb05 - Safeguard alias loop while looking in the cache for expired answers. 2024-12-03 14:10:17 +01:00
Yorgos Thessalonikefs be92752368 - Merge #1198: Fix log-servfail with serve expired and no useful cache
contents.
2024-12-03 14:05:12 +01:00
Yorgos ThessalonikefsandGitHub 1512945c79 Merge pull request #1198 from NLnetLabs/bugfix/log-servfail-serve-expired
Fix log-servfail with serve expired and no useful cache contents
2024-12-03 14:02:03 +01:00
Yorgos Thessalonikefs 9de159b96b - For #1175, the default value of serve-expired-ttl is set to 86400
(1 day) as suggested by RFC8767.
2024-12-03 13:09:51 +01:00
Yorgos Thessalonikefs bd2e66de1e Changelog entry for #1189, #1197:
- Merge #1189: Fix the dname_str method to cause conversion errors
  when the domain name length is 255.
- Merge #1197: dname_str() fixes.
2024-12-03 11:58:06 +01:00
Yorgos ThessalonikefsandGitHub 9770e855d2 Merge pull request #1197 from NLnetLabs/dname_str-more-tests
dname_str() fixes
2024-12-03 11:55:41 +01:00
Yorgos Thessalonikefs c124f67f33 - For #1193, introduce log-servfail.tdir and cleanup the log-servfail
setting from other tests.
2024-12-02 12:30:11 +01:00
Yorgos Thessalonikefs c55490c1e6 - Fix #1193: log-servfail fails to log host SERVFAIL responses in
Unbound 1.19.2 on Ubuntu 24.04.1 LTS, by not considering cached
  failures when trying to reply with expired data.
2024-12-02 12:28:11 +01:00
Yorgos Thessalonikefs f46acec35f - For #1189, homogenize the input buffer size for dname_str(). 2024-12-02 11:53:56 +01:00
Yorgos Thessalonikefs 1cd2fb3b9d - For #1189, add unit tests for dname_str() and debug check the input
buffer size.
2024-12-02 10:03:35 +01:00
wenxuan70 06fb30d0a0 Fix the dname_str method to cause conversion errors when the domain name length is 255 2024-11-24 17:53:23 +08:00
Yorgos Thessalonikefs 9e3c50ec9e - For #1175, update serve-expired tests. 2024-11-22 16:14:02 +01:00
Yorgos Thessalonikefs eefdbb341f - Fix #1175: serve-expired does not adhere to secure-by-default
principle. The default value of serve-expired-client-timeout
  is set to 1800 as suggested by RFC8767.
2024-11-22 15:32:34 +01:00
Yorgos Thessalonikefs e75da7d954 - Fix comparison to help static analyzer. 2024-11-20 10:53:45 +01:00
Yorgos Thessalonikefs 9a3a1bc221 Changelog entry for #1169:
- Merge #1169 from Sergey Kacheev, fix: lock-free counters for
  auth_zone up/down queries.
2024-11-19 17:01:34 +01:00
Yorgos ThessalonikefsandGitHub c1e9d7be7f Merge pull request #1169 from sakateka/lock-free-az-counters
fix: lock-free counters for auth_zone up/down queries
2024-11-19 17:00:01 +01:00
Sergey Kacheev 2c72a4970b fix: lock-free counters for auth_zone up/down queries 2024-11-19 18:55:31 +03:00
W.C.A. Wijngaards 4cf7fae50c - Fix for #1183: release nsec3 hashes per test file. 2024-11-15 10:47:27 +01:00
W.C.A. Wijngaards a2ac980737 - Fix #1183: the data being used is released in method
nsec3_hash_test_entry.
2024-11-15 10:37:35 +01:00
Yorgos Thessalonikefs 733d5f7161 - Complete fix for max-global-quota to 200. 2024-11-08 17:34:28 +01:00
Yorgos Thessalonikefs fe288a9b06 - More descriptive text for 'harden-algo-downgrade'. 2024-11-08 13:56:04 +01:00
Yorgos Thessalonikefs fd1a1d5fa0 - Increase the default of max-global-quota to 200 from 128 after
operational feedback. Still keeping the possible amplification
  factor (CAMP related issues) in the hundreds.
2024-11-06 16:28:37 +01:00
Yorgos Thessalonikefs 3c4b87636a Changelog entry for:
- Fix SETEX check during Redis (re)initialization.
2024-11-05 12:20:25 +01:00
Yorgos Thessalonikefs 36d8c6e778 - Fix SETEX check during Redis (re)initialization. 2024-11-05 12:18:55 +01:00
W.C.A. Wijngaards 60fd77b8f9 - Fix to log redis timeout error string on failure. 2024-11-05 11:41:41 +01:00
W.C.A. Wijngaards d5e91d181b - Fix for the serve expired DNSSEC information fix, it would not allow
current delegation information be updated in cache. The fix allows
  current delegation and validation recursion information to be
  updated, but as a consequence no longer has certain expired
  information around for later dnssec valid expired responses.
2024-11-05 10:39:27 +01:00
W.C.A. Wijngaards 7985d17b57 Changelog note for #1167
- Merge #1167: Makefile.in: fix occasional parallel build failures
  around bison rule.
2024-11-04 13:26:27 +01:00
Sergei TrofimovichandGitHub 46cfbf313d Makefile.in: fix occasional parallel build failures around bison rule (#1167)
Without the change `make -j16 --shuffle` occasinally fails to build as:

    $ make -j16 --shuffle
    ...
    bison -y -d -o util/configparser.c ./util/configparser.y
    ...
    /libtool --tag=CC --mode=compile gcc -I.  -I...-openssl-3.3.2-dev/include -I...-libevent-2.1.12-dev/include -I...-expat-2.6.3-dev/include -DSRCDIR=. -g -O2 -flto -fPIE -pthread  -o configparser.lo -c util/configparser.c
    ...
    util/configparser.c:755:3: error: expected ',' or '}' at end of input
  755 |   YYSYMBOL_server_low_rtt = 626,           /* server_low_rtt  */
      |   ^

The build failure happens due to this `Makefile.in` rule:

    util/configparser.c util/configparser.h:  $(srcdir)/util/configparser.y
        @-if test ! -d util; then $(INSTALL) -d util; fi
        $(YACC) -d -o util/configparser.c $(srcdir)/util/configparser.y

For GNU make that means that each of the targets will attempt the rule
execution when the file is missing: one for .c file and another for .h
file:

    https://www.gnu.org/software/make/manual/html_node/Multiple-Targets.html

The workaround is to only run $(YACC) for .c target and use .c as a
pre-requisite for an .h file.

Before the change the build fails about every 10-th run.
After the change no build failures after 100 successful builds.
2024-11-04 13:26:05 +01:00
W.C.A. Wijngaards 533c3b0514 - Fix redis that during a reload it does not fail if the redis
server does not connect or does not respond. It still logs the
  errors and if the server is up checks expiration features.
2024-11-04 10:14:26 +01:00
W.C.A. Wijngaards 5f3f214da5 - Fix redis that during a reload it does not fail if the redis
server does not connect or does not respond. It still logs the
  errors and if the server is up checks expiration features.
2024-11-04 10:14:13 +01:00
Yorgos Thessalonikefs 11b8157a98 Changelog entry for #1157:
- Merge #1157 from Liang Zhu, Fix heap corruption when calling
  ub_ctx_delete in Windows.
2024-11-01 16:27:06 +01:00
Liang ZhuandGitHub 1c24cd79cc Fix heap corruption when calling ub_ctx_delete in Windows (#1157) 2024-11-01 16:26:05 +01:00
Yorgos Thessalonikefs d34fb3ed77 Changelog entry for #1170:
- Merge #1170 from Melroy van den Berg, Fix chroot manpage
  description.
2024-11-01 16:12:07 +01:00
Melroy van den BergandGitHub c37833c943 Fix chroot manpage description (#1170) 2024-11-01 16:10:57 +01:00
Yorgos Thessalonikefs 8a6a4bd7f3 - Add test case for #1159.
- Some clean up for stat_values.test.
2024-11-01 15:57:52 +01:00
Yorgos Thessalonikefs d23523e528 - Merge #1159: Stats for discard-timeout and wait-limit. 2024-11-01 15:54:24 +01:00
Wouter WijngaardsandGitHub 75967e7f5e Stats for discard-timeout and wait-limit (#1159)
* - Stats num.queries_discard_timeout and num.queries_wait_limit are used
  instead of the mesh dropped that of requests exceeded.
2024-11-01 15:52:58 +01:00
Yorgos Thessalonikefs f5580f0a63 - Fix #1163: Typos in unbound.conf documentation. 2024-10-25 21:25:16 +02:00
W.C.A. Wijngaards 0e2b2743d8 Add changelog entry for tag for 1.22.0rc1. 2024-10-17 10:57:07 +02:00
W.C.A. Wijngaards 018be1d089 - Tag for 1.22.0 release. This did not contain the 1154 fix
from 16 oct. The code repository continues with
  version 1.22.1 in development.
2024-10-17 10:48:58 +02:00
W.C.A. Wijngaards 9a63db344e - Fix #1154: Tag Incorrectly Applying for Other Interfaces
Using the Same IP. This fix is not for 1.22.0.
2024-10-16 15:56:33 +02:00
W.C.A. Wijngaards 0076736fc4 - Fix for dnstap with dnscrypt and dnstap without dnsoverquic. 2024-10-16 11:52:49 +02:00
Yorgos Thessalonikefs f8e45ed696 - Fix for dnsoverquic and dnstap to use the correct dnstap
environment.
2024-10-16 11:02:31 +02:00
W.C.A. Wijngaards 2a28c7389c - Fix dnsoverquic to extend the number of streams when one is closed. 2024-10-14 13:53:55 +02:00
W.C.A. Wijngaards 114edf2c38 - Fix to display warning if quic-port is set but dnsoverquic is not
enabled when compiled.
2024-10-14 11:34:26 +02:00
W.C.A. Wijngaards e0c93e300b - Fix contrib/aaaa-filter-iterator.patch for change in call
signature for cache_fill_missing.
2024-10-11 11:42:30 +02:00
W.C.A. Wijngaards bd1813b126 - Fix harden-unverified-glue for AAAA cache_fill_missing lookups. 2024-10-11 09:03:11 +02:00
W.C.A. Wijngaards 1b7e14dc39 - Fix to disable detection of quic configured ports when quic is
not compiled in.
2024-10-11 08:51:14 +02:00
W.C.A. Wijngaards 8b7782e8fc - Fix add reallocarray to alloc stats unit test, and disable
override of strdup in unbound-host, and the result of config
  get option is freed properly.
2024-10-10 10:43:23 +02:00
W.C.A. Wijngaards e0201435a4 - Fix cookie_file test sporadic fails for time change during
the test.
2024-10-10 09:45:48 +02:00
W.C.A. Wijngaards 66fb3ff670 - Fix for dnstap compile of doqclient with doq disabled. 2024-10-09 15:52:33 +02:00
W.C.A. Wijngaards 36461ea73d Changelog entry and unit test for fix of NSEC TTL and prefetch ttl.
- Fix to limit NSEC TTL for messages from cachedb. Fix to limit the
  prefetch ttl for messages after a CNAME with short TTL.
2024-10-09 15:29:23 +02:00
W.C.A. Wijngaards 5679c8b1df - Fix to limit NSEC TTL for messages from cachedb. Fix to limit the
prefetch ttl for messages after a CNAME with short TTL.
2024-10-09 15:28:55 +02:00
W.C.A. Wijngaards a4d8c0c43b Changelog note for #871
- Merge #871: DNS over QUIC. This adds `quic-port: 853` and
  `quic-size: 8m` that enable dnsoverquic, and the counters
  `num.query.quic` and `mem.quic` in the statistics output.
  The feature needs to be enabled by compiling with libngtcp2,
  with `--with-libngtcp2=path` and libngtcp2 needs openssl+quic,
  pass that with `--with-ssl=path` to compile unbound as well.
2024-10-09 10:35:45 +02:00
3d89c26d32 DNSoverQUIC (#871)
* - dnsoverquic, configure --with-libngtcp2 option.

* - dnsoverquic, create comm_point for doq and receive cmsg local address.

* - dnsoverquic, less obtrusive debug.

* - dnsoverquic, log and fix local port number. Neater subroutines and ifdefs.

* - dnsoverquic, add testcode/doqclient.

* - dnsoverquic, review fixes on doqclient.

* - dnsoverquic, fix unit test testbound link.

* - dnsoverquic, parse query in doqclient.

* - dnsoverquic, link with libngtcp2_crypto_openssl and code for doqclient.

* - dnsoverquic, random routine for doqclient and fix ngaddr allocation, and
  check ub_initstate return.

* - dnsoverquic, fix doqclient free of allocated ngaddr addresses.

* - dnsoverquic, enable debug output with -v for doqclient.

* - dnsoverquic, create and set TLS object and TLS context in doqclient.

* - dnsoverquic, work on quic tls context in doqclient.

* - dnsoverquic, set default dnsoverquic port to the standardized 853 port.

* - dnsoverquic, remove debug comment.

* - dnsoverquic, dns-over-quic quic-port: 853 config option.

* - dnsoverquic, log type of interface created at start of unbound.

* - dnsoverquic, log type of no tls https as https when interface is created.

* - dnsoverquic, setup client quic tls methods.

* - dnsoverquic, event work in doqclient.

* - dnsoverquic, explain in documentation that QUIC uses UDP.

* - dnsoverquic, make doqclient exit.

* - dnsoverquic, doqclient cleanup run routine.

* - dnsoverquic, doqclient code nicer.

* - dnsoverquic, doqclient read and timer.

* - dnsoverquic, doqclient write work.

* - dnsoverquic, review fixes.

* - dnsoverquic, detect openssl quic support at configure time.

* - dnsoverquic, do not allow QUIC on port 53 to stop confusion of DoQ and DNS.

* - dnsoverquic, in doqclient, when idle close is returned, drop the connection
  without calling ngtcp2_conn_write_connection_close.

* - dnsoverquic, in doqclient, log callbacks.

* - dnsoverquic, in doqclient add extend_max_local_streams_bidi callback.

* - dnsoverquic, in doqclient add client query lists.

* - dnsoverquic, in doqclient, code cleaner, log text nicer.

* - dnsoverquic, in doqclient, work on write_streams.

* - dnsoverquic, in doqclient, use signed int for stream_id, work on the
  ngtcp2_recv_stream_data callback.

* - dnsoverquic, in doqclient, print result and fixes for recv data.

* - dnsoverquic, in doqclient, add the event callbacks to fptr wlist.

* - dnsoverquic, in doqclient, when already expired, use zero timeout timer.

* - dnsoverquic, in doqclient, ignore unused return codes from
  ngtcp2_conn_writev_stream.

* - dnsoverquic, add doqclient event functions to the unbound-dnstap-socket
  test tool for linking.

* - dnsoverquic, in doqclient, fix multiple operands for the commandline.
  neater dns message output.

* - dnsoverquic, in doqclient, store packet when write blocks and try later.

* - dnsoverquic, in doqclient, limit number of packets and number of bytes sent.

* - dnsoverquic, in doqclient, better size estimate for outgoing packet.

* - dnsoverquic, in doqclient, fix that already written next packet is not
  counted for data length to send.

* - dnsoverquic, in doqclient, early data transmission and session resumption.

* - dnsoverquic, send version negotiation packet.

* - dnsoverquic, send retry and accept the connection.

* - dnsoverquic, storage structures.

* - dnsoverquic, doq connection setup.

* - dnsoverquic, neater code layout for new conn. Fix verbosity of log print.

* - dnsoverquic, doq conn callback functions.

* - dnsoverquic, doq_fill_rand routine in header file.

* - dnsoverquic, keep track of connection ids.

* - dnsoverquic, get_new_connection_id callback.

* - dnsoverquic, create doq_conid tree.

* - dnsoverquic, settings for server connection.

* - dnsoverquic, tls context.

* - dnsoverquic, sendmsg error handling.

* - dnsoverquic, neat code.

* - dnsoverquic, track doq connection last error.

* - dnsoverquic, neater packet address parameters.

* - dnsoverquic, fix uninitialized bytes in msg control in doq sendmsg, and
  fix tree cleanup of conid tree.

* - dnsoverquic, better usage text for doqclient.

* - dnsoverquic, neat code.

* - dnsoverquic, connection receive packet handling.

* - dnsoverquic, debug output.

* - dnsoverquic, debug switched meaning of scid and dcid gives
  ERR_TRANSPORT_PARAM.

* - dnsoverquic, remove debug output.

* - dnsoverquic, connection delete routine and error from connection read in
  more detail with less clutter.

* - dnsoverquic, write to stream, and receive stream data, log packet.

* - dnsoverquic, alpn set up.

* - dnsoverquic, connection close.

* - dnsoverquic, doq_table and locks.

* - dnsoverquic, fix tests.

* - dnsoverquic, better locking.

* - dnsoverquic, doq_stream.

* - dnsoverquic, remove compile warning.

* - dnsoverquic, doq_stream receive data.

* - dnsoverquic, fixes for locks and keep length bytes allocated.

* - dnsoverquic, lock connection on initial insertion.

* - dnsoverquic, reply information, and reply buffer.

* - dnsoverquic, reply info from cache, local-zone and recursion lookups.

* - dnsoverquic, spelling in comment about buffer storage.

* - dnsoverquic, stream write list and doqclient fixes to exit and printout.

* - dnsoverquic, doqclient -q option for short printout.

* - dnsoverquic, unit test with local data reply.

* - dnsoverquic, write connection and write event is set.

* - dnsoverquic, neater logging for write event connection stream writes.

* - dnsoverquic, log remote connection when the streams are written for it.

* - dnsoverquic, better threaded use, threads can write to doq connections at
  the same time.

* - dnsoverquic, unit test for the calculation of connection size with a query.

* - dnsoverquic, use less memory per connection.

* - dnsoverquic, remove unit test output.

* - dnsoverquic, add MSG_DONTWAIT so that there is no mistakenly blocking
  socket operations.

* - dnsoverquic, doqclient logs address on connection failures.

* - dnsoverquic, compat code for clock get time routine.

* - dnsoverquic, use skip_test for doq unit test.

* - dnsoverquic, fixes for proxyprotocol, use remote_addr and set proxyprotocol
  disabled on the doq connection.

* - dnsoverquic, doqclient sets log identity to its name, instead of "unbound".

* - dnsoverquic, handle blocked udp packet writes.

* - dnsoverquic, fix function documentation for verbose_print_addr from
  services/listen_dnsport.c.

* - dnsoverquic, fix doq_conn lock protection. The checklock allows to set
  the output file name, and doqclient uses that. Print place of lock_protect.

* - dnsoverquic, neater buffer clear when write of blocked packet fails, make
  sure that memory area does not overlap for blocked packet addresses when
  write of blocked packet fails, and size blocked packet buffer to the pkt buf.

* - dnsoverquic, move lock check after the test to test script in doq test.

* - dnsoverquic, the doq test uses valgrind when enabled.

* - dnsoverquic, git ignore the doqclient test.

* - dnsoverquic, limit the buffer for packets to max packet size with some more.

* - dnsoverquic, spelling fix.

* - dnsoverquic, timer work, structure and adds and deletes.

* - dnsoverquic, timer_tree uses table.lock.

* - dnsoverquic, fix timer tree remove and spelling in header file comment.

* - dnsoverquic, fix testbound for timer compare function linkage.

* - dnsoverquic, timer set add debug output.

* - dnsoverquic, doq_conn_check_timer function.

* - dnsoverquic, doq_done_setup_timer_and_write function.

* - dnsoverquic, fix that doq conn is not deleted whilst editing write and timer.

* - dnsoverquic, Fix #861 make ERROR netevent.h:1073:32: error: field 'blocked_pkt_pi' has incomplete type

* - dnsoverquic, timer element has timeout setup when socket callback complete.

* - dnsoverquic, fix unit test compile.

* - dnsoverquic, timer callback routine, handle timeout and close and delete the
  connection if necessary.

* - dnsoverquic, timer pickup stops at current time.

* - dnsoverquic, timer comparable with the event base time.

* - dnsoverquic, erase marked time when timer disabled.

* - dnsoverquic, fix timer to set correctly and lock popped write connection
  early, before it is modified.

* - dnsoverquic, fix to unlock connection lock when it is unlinked and deleted.

* - dnsoverquic, fix to unlock connection lock when it is deleted because it is
  a duplicate connection.

* - dnsoverquic, fix that doq timer is not disabled when not set.

* - dnsoverquic, quic-size: 8m maximum number of bytes for QUIC buffers.

* - dnsoverquic, flex and bison.

* - dnsoverquic, quic-size turn away new connections when full.

* - dnsoverquic, doqclient outputs stream reset information.

* - dnsoverquic, detect stream close and reset.

* - dnsoverquic, free stream buffers when data is acked and stream is closed.

* - dnsoverquic, delete stream when closed. Unlink it. Allow stream_id 4 as first.

* - dnsoverquic, stats output for mem.quic and num.query.quic.

* - dnsoverquic, review fix.

* - dnsoverquic, fix when compiled without ngtcp2.

* - dnsoverquic, fix to detect ngtcp2_crypto_quictls for openssl crypto, after
  change in libngtcp2.

* - dnsoverquic, fix for newer ngtcp2 versions. detect ngtcp2_ccerr_default,
  ngtcp2/ngtcp2_crypto_quictls.h, struct ngtcp2_pkt_hd.tokenlen,
  struct ngtcp2_settings.tokenlen and struct ngtcp2_version_cid.

* - dnsoverquic, fix for newer ngtcp2 version, detect number of arguments for
  ngtcp2_conn_shutdown_stream.

* - dnsoverquic, fix for newer ngtcp2.

* - dnsoverquic, use the functions from util/timeval_func.h.

* - dnsoverquic, fix in doqclient only write transport parameters once.

* - dnsoverquic, debug log output removed.

* - dnsoverquic, fix in doqclient to work with renamed NGTCP2_CC_ALGO_BBR_V2
  from ngtcp2.

* - dnsoverquic, fix to check in doq_server_socket_create that tls-service-key
  and tls-service-pem have a value.

* - dnsoverquic, fix to error when doq_server_socket_create fails.

* - dnsoverquic, improve linebreaks in configparser additions.

* - dnsoverquic, fix port from interface pickup after main branch change.

* Fix getting user data from SSL, fix calloc warning.

* Fix fwrite return value check in doqclient

* - timeval_substruct from timeval_func.h
- lock_protect also for HAVE_NGTCP2_CCERR_DEFAULT
- fix doq logging for inet_ntop failures

* - memset for consistency
- no value returned from msghdr_get_ecn when S_SPLINT_S is defined

* - dnsoverquic, rerun autoconf.

---------

Co-authored-by: Yorgos Thessalonikefs <yorgos@nlnetlabs.nl>
2024-10-09 10:32:03 +02:00
W.C.A. Wijngaards dcf7afd722 - Fix #1128: Cannot override tcp-upstream and tls-upstream with
forward-tcp-upstream and forward-tls-upstream.
2024-10-08 15:29:03 +02:00
W.C.A. Wijngaards e67171612b - Fix #1149: unbound-control-setup hangs sometimes depending on
the openssl version.
2024-10-08 11:54:07 +02:00
Yorgos Thessalonikefs a1b25f0296 - The fix for CVE-2024-8508 was part of 1.21.1, a security point release
on 1.21.0. The code repository continues with this fix and the version
  number 1.22.0.
2024-10-03 18:19:01 +02:00
Yorgos Thessalonikefs 490585bf29 Merge branch 'release-1.21.1' 2024-10-03 18:14:01 +02:00
Yorgos Thessalonikefs b7c61d7cc2 - Fix CVE-2024-8508, unbounded name compression could lead to denial of
service.
2024-10-03 17:41:20 +02:00
Yorgos Thessalonikefs 01883fb017 - Set version to 1.21.1 2024-10-03 14:11:57 +02:00
W.C.A. Wijngaards 5bb3b9cc83 - Fix unbound dnstap socket test program analyzer warnings about
unused variable assignments and variable initialization.
2024-09-30 16:36:01 +02:00
W.C.A. Wijngaards 3a1b79f6a1 - Fix negative cache NSEC3 parameter compares for zero length NSEC3
salt.
2024-09-30 09:25:51 +02:00
W.C.A. Wijngaards 84eeb9b97c - Fix #1144: [FR] log timestamps in ISO8601 format with timezone.
This adds the option `log-time-iso: yes` that logs in ISO8601
  format.
2024-09-25 11:16:46 +02:00
Yorgos Thessalonikefs d88eeb4c32 Changelog entry for #1143:
- Merge #1143: Fix cache update when serve expired is used. Expired
  records are favored over resolution and validation failures when
  serve-expired is used.
2024-09-24 16:49:34 +02:00
Yorgos ThessalonikefsandGitHub 2e398d51ba Fix cache update when serve expired is used (#1143)
- Fix cache update when serve expired is used in order to not evict
  still usable expired records. Modules are forbidden to update the
  cache if their answer is DNSSEC unchecked or bogus and a valid
  (expired) entry already exists. Bogus replies from the validator are
  also discarded in favor of existing (expired) valid replies.

- serve-expired-ttl-reset should try to keep expired records in the
  cache in case they are reset.
2024-09-24 16:47:04 +02:00
Yorgos Thessalonikefs 24ebca7df6 - More clear text for prefetch and minimal-responses in the
unbound.conf man page.
2024-09-24 15:10:21 +02:00
Yorgos Thessalonikefs 7f4a61e6fc - Attempt to further fix doh_downstream_buffer_size.tdir flakiness. 2024-09-24 12:21:03 +02:00
Yorgos Thessalonikefs db719d404f - Fix doxygen warnings by commenting out CLANG_ASSISTED_PARSING,
CLANG_ADD_INC_PATHS, CLANG_OPTIONS and CLANG_DATABASE_PATH; they were
  already disabled.
2024-09-23 15:31:32 +02:00
W.C.A. Wijngaards a35a0c49da - Fix dns64 with prefetch that the prefetch is stored in cache. 2024-09-23 12:19:43 +02:00
W.C.A. Wijngaards 5e9b6296b7 - Add redis-command-timeout: 20 and redis-connect-timeout: 200,
that can set the timeout separately for commands and the
  connection set up to the redis server. If they are not
  specified, the redis-timeout value is used.
2024-09-17 13:10:34 +02:00
W.C.A. Wijngaards 606e262fdd Changelog comment for #1140.
- Merge #1140: Fix spelling mistake in comments.
2024-09-16 12:15:04 +02:00
TochusandGitHub 819764663b Fix spelling mistake in comments (#1140)
I noticed a spelling mistake in the comments. The term “chain of trust” was incorrectly written as “chainoftrust”. This change corrects the spelling to “chain of trust” which is the correct term used in English.
2024-09-16 12:14:28 +02:00
Yorgos Thessalonikefs 6bf2b2ac56 - Fix and add comments in testdata/val_negcache_ttl.rpl. 2024-09-11 12:16:02 +02:00
W.C.A. Wijngaards 5767b0933f - Add unit test for ttl limit for aggressive nsec. 2024-09-10 10:17:31 +02:00
W.C.A. Wijngaards 24e0f0ab7e - Fix to limit NSEC and NSEC3 TTL when aggressive nsec is
enabled (RFC9077).
2024-09-10 10:13:48 +02:00
Yorgos Thessalonikefs d3fdbba877 - Fix comment to not trigger doxygen unknown command. 2024-09-06 16:03:20 +02:00
Yorgos Thessalonikefs c36ce2a390 - Fix alloc-size and calloc-transposed-args compiler warnings. 2024-09-06 16:01:30 +02:00
W.C.A. Wijngaards 7ecff4113c - Fix config file read for dnstap-sample-rate. 2024-09-05 09:35:54 +02:00
W.C.A. Wijngaards 99824bc0e6 Changelog note for #1135
- Merge #1135: Add new IANA trust anchor.
2024-09-02 09:25:44 +02:00
9f09c36401 Add new IANA trust anchor (#1135)
Signed-off-by: Keelan Cannoo <keelan.cannoo@cyberstorm.mu>
Co-authored-by: Keelan10 <keelan.cannoo@cyberstorm.mu>
2024-09-02 09:24:55 +02:00
W.C.A. Wijngaards a887284703 - Fix for #1132, comment about adjusted copy of reference check. 2024-08-30 08:56:00 +02:00
W.C.A. Wijngaards fb198b96f1 Changelog note for #1132 and fix for #1132.
- Merge #1132: b.root renumbering.
- Fix for #1132, adjusted unit test for change in the test file.
2024-08-30 08:51:56 +02:00
Loganaden VelvindronandGitHub 30bf996f39 b.root renumbering (#1132)
https://b.root-servers.org/news/2023/05/16/new-addresses.html

Worked together with Jaykishan Muktawoa <jay@cyberstorm.mu>
2024-08-30 08:48:31 +02:00
W.C.A. Wijngaards 52154e658a - Fix to print port number in logs for auth zone transfer activities. 2024-08-29 13:04:03 +02:00
W.C.A. Wijngaards c06d3646a9 - Unit test for auth zone transfer TLS, and TLS failure. 2024-08-29 10:40:31 +02:00
W.C.A. Wijngaards 42d421a305 - Fix that stub-zone and forward-zone clauses do not exhaust memory
for long content.
2024-08-28 13:16:29 +02:00
W.C.A. Wijngaards b5951ce1fa - Fix that when rpz is applied the message does not get picked up by
the validator. That stops validation failures for the message.
2024-08-28 10:51:22 +02:00
W.C.A. Wijngaards 6b37309705 - Fix #1130: Loads of logs: "validation failure: key for validation
<domain>. is marked as invalid because of a previous" for
  non-DNSSEC signed zone.
2024-08-27 17:00:27 +02:00
W.C.A. Wijngaards dc274fef9b - Fix documentation for cache_fill_missing function. 2024-08-23 13:19:15 +02:00
W.C.A. Wijngaards db1167c8b3 - Fix #1127: error: "memory exhausted" when defining more than 9994
local-zones.
2024-08-23 09:22:07 +02:00
W.C.A. Wijngaards 1e0cf1e86b - Merge patch to fix for glue that is outside of zone, with
`harden-unverified-glue`, from Karthik Umashankar (Microsoft).
  Enabling this option protects the Unbound resolver against bad
  glue, that is unverified out of zone glue, by resolving them.
  It uses the records as last resort if there is no other working
  glue.
2024-08-23 08:56:48 +02:00
W.C.A. Wijngaards 348df52e05 Enable ci back after debug. 2024-08-21 14:20:04 +02:00
W.C.A. Wijngaards 6b3266aaf8 - Fix for char signedness warnings on NetBSD. 2024-08-21 14:15:23 +02:00
W.C.A. Wijngaards 4f52461e81 - Add cross platform netbsd to github ci. 2024-08-21 14:03:11 +02:00
W.C.A. Wijngaards 06d5031d22 - Add cross platform openbsd to github ci. 2024-08-21 13:50:55 +02:00
W.C.A. Wijngaards 12119e2d3e ci for freebsd nicer, with libevent, faster without static compile, and
with grouped output, also the pkg install is conditional on the platform.
2024-08-21 13:37:42 +02:00
W.C.A. Wijngaards 19d53d5663 Fix for freebsd ci. 2024-08-21 13:24:54 +02:00
W.C.A. Wijngaards 04e6f9e03b - Add cross platform freebsd to github ci. 2024-08-21 13:20:00 +02:00
W.C.A. Wijngaards 3d350fa73d - Add iter-scrub-ns, iter-scrub-cname and max-global-quota
configuration options.
2024-08-20 14:08:52 +02:00
W.C.A. Wijngaards 015b2b0daf - Fix #1126: unbound-control-setup hangs while testing for openssl
presence starting from version 1.21.0.
2024-08-19 15:51:47 +02:00
W.C.A. Wijngaards 5fa84d50bf - Tag for release 1.21.0, the repository continues with 1.21.1
in development.
2024-08-15 11:01:41 +02:00
W.C.A. Wijngaards 79e4c57851 - Fix spelling for the cache-min-negative-ttl entry in the
example.conf.
2024-08-09 14:04:25 +02:00
W.C.A. Wijngaards 5abdd09095 - Fix that for windows the module startup is called and sets up
the module-config.
2024-08-08 16:14:09 +02:00
W.C.A. Wijngaards 158c1defe3 - Set version number to 1.21.0 for release. 2024-08-08 09:30:53 +02:00
W.C.A. Wijngaards b4519012dc - Fix CacheFlush issues with limit on NS RRs. Thanks to Yehuda Afek,
Anat Bremler-Barr, Shoham Danino and Yuval Shavitt (Tel-Aviv
  University and Reichman University).
2024-08-08 09:28:44 +02:00
W.C.A. Wijngaards ed883238fd - Fix CAMP issues with global quota. Thanks to Huayi Duan, Marco
Bearzi, Jodok Vieli, and Cagin Tanir from NetSec group, ETH Zurich.
2024-08-08 09:27:45 +02:00
W.C.A. Wijngaards 0f2f6025e7 - Fix that alloc stats for forwards and hints are printed, and when
alloc stats is enabled, the unit test for unbound control waits for
  reloads to complete.
2024-08-02 15:51:40 +02:00
W.C.A. Wijngaards 3cbf554e3b Changelog note for #1090
- Merge #1090: Cookie secret file. Adds
  `cookie-secret-file: "unbound_cookiesecrets.txt"` option to store
  cookie secrets for EDNS COOKIE secret rollover. The remote control
  add_cookie_secret, activate_cookie_secret and drop_cookie_secret
  commands can be used for rollover, the command print_cookie_secrets
  shows the values in use.
2024-08-02 13:36:06 +02:00
ad21dbd1c2 Cookie secret file (#1090)
* - cookie-secret-file, define struct.

* - cookie-secret-file, add config option, create, read and delete struct.

* - cookie-secret-file, check cookie secrets for cookie validation.

* - cookie-secret-file, unbound-control add_cookie_secret, drop_cookie_secret,
  activate_cookie_secret and print_cookie_secrets.

* - cookie-secret-file, test and fix locks, renew writes a fresh cookie,
  staging cookies get a fresh cookie and spelling in error message.

* - cookie-secret-file, remove unused variable from cookie file unit test.

* Remove unshare and faketime dependencies for cookie_file test; documentation nits.

---------

Co-authored-by: Yorgos Thessalonikefs <yorgos@nlnetlabs.nl>
2024-08-02 13:32:08 +02:00
W.C.A. Wijngaards 50cf55bdac Update changelog.
- Fix testbound for alloc stats strdup in util/alloc.c.
2024-08-02 08:59:47 +02:00
W.C.A. Wijngaards 6106528a50 - Fix testbound for alloc stats strdup in util/alloc.c. 2024-08-02 08:58:22 +02:00
W.C.A. Wijngaards befa7d8cd8 - Fix that alloc stats has strdup checks, it stops debuggers from
complaining about mismatch at free time.
2024-08-02 08:54:54 +02:00
W.C.A. Wijngaards 92be76fb89 - Fix that the worker mem report with alloc stats does not attempt
to print memory use of forwards and hints if they have been
  deleted already.
2024-08-01 17:15:07 +02:00
W.C.A. Wijngaards 9a6b6765cc - Fix dnstap test program, cleans up to have clean memory on exit,
for tap_data_free, does not delete NULL items. Also it does not try
  to free the tail, specifically in the free of the list since that
  picked up the next item in the list for its loop causing invalid
  free. Added internal unit test to unbound-dnstap-socket for that.
2024-08-01 16:12:04 +02:00
W.C.A. Wijngaards 03b511b1a2 - Fix for #1114: Fix that cache fill for forward-host names is
performed, so that with nonzero target-fetch-policy it fetches
  forwarder addresses and uses them from cache. Also updated that
  delegation point cache fill routines use CDflag for AAAA message
  lookups, so that its negative lookup stops a recursion since the
  cache uses the bit for disambiguation for dns64 but the recursion
  uses CDflag for the AAAA target lookups, so the check correctly
  stops a useless recursion by its cache lookup.
2024-07-31 11:42:44 +02:00
W.C.A. Wijngaards 6af28bed08 - Fix to document parameters of auth_zone_verify_zonemd_with_key. 2024-07-30 13:47:53 +02:00
W.C.A. Wijngaards f094f4ea3c - Add root key 38696 from 2024 for DNSSEC validation. It is added
to the default root keys in unbound-anchor. The content can be
  inspected with `unbound-anchor -l`.
2024-07-25 11:42:22 +02:00
Yorgos Thessalonikefs c717debace - For #935 and #1104, clarify RPZ order and semantics. 2024-07-24 01:54:02 +02:00
Yorgos Thessalonikefs 7d4d21764a - Cleanup ede.tdir test. 2024-07-23 20:22:25 +02:00
W.C.A. Wijngaards 83e6977f06 - Fix link of unbound-dnstap-socket without openssl. 2024-07-23 15:06:54 +02:00
W.C.A. Wijngaards 671e11552c - Fix link of dnstap without openssl. 2024-07-23 14:56:21 +02:00
W.C.A. Wijngaards c4541e634b - Fix uninitialized variable warning in create_tcp_accept_sock. 2024-07-23 10:42:36 +02:00
W.C.A. Wijngaards 30da725e67 - Fix to have empty definition when not supported for weak attribute. 2024-07-23 10:02:39 +02:00
W.C.A. Wijngaards 8de5ae3552 - Fix compile when the compiler does not support the noreturn
attribute.
2024-07-23 09:55:31 +02:00
W.C.A. Wijngaards 5bea29b01c - For #1110: Test for fallthrough attribute in configure and add
fallthrough attribute annotations.
2024-07-23 09:47:42 +02:00
Wouter WijngaardsandGitHub 89c9eafa44 Merge pull request #1110 from r-barnes/patch-1
Make fallthrough explicit for libworker.c
2024-07-23 09:44:28 +02:00
Yorgos Thessalonikefs 3512eaec48 - Fix #1106: ratelimit-below-domain logs the wrong FROM address. 2024-07-23 09:07:06 +02:00
Richard BarnesandGitHub f9bd35dcfa Make fallthrough explicit for libworker.c
The code currently doesn't compile with LLVM's `-Wimplicit-fallthrough` flag, but the attribute works for both GCC (>=7) and LLVM.
2024-07-22 16:37:38 -07:00
W.C.A. Wijngaards 3af4e44646 - Fix dnstap wakeup, a running wakeup timer is left to expire and not
increased, a timer is started when the dtio thread is sleeping,
  the timer set disabled when the dtio thread goes to sleep, and
  after sleep the thread checks to see if there are messages to log
  immediately.
2024-07-19 16:16:02 +02:00
W.C.A. Wijngaards c3dd6a2dbd - Add dnstap-sample-rate that logs only 1/N messages, for high volume
server environments. Thanks Dan Luther.
2024-07-19 10:04:40 +02:00
W.C.A. Wijngaards 8fca3e7c5b - For #1103: Fix to drop mesh state reference for the http2 stream
associated with the reply, not the currently active stream. And
  it does not remove it twice on a mesh_send_reply call. The reply
  h2_stream is NULL when not in use, for more initialisation.
2024-07-16 14:23:10 +02:00
W.C.A. Wijngaards 8947c2c764 - For #1103: fix to also drop mesh state reference when the discard
limit is reached, when there is an error making a new recursion
  state and when the connection is dropped with is_drop.
2024-07-15 14:51:20 +02:00
W.C.A. Wijngaards b1e3319a11 Merge branch 'master' of github.com:NLnetLabs/unbound 2024-07-12 16:41:58 +02:00
W.C.A. Wijngaards d52f501d90 - For #1103: fix to also drop mesh state reference when a h2 reply is
dropped.
2024-07-12 16:41:46 +02:00
Yorgos Thessalonikefs 7083d58c6b - For #1102: clearer text for using interface-* options for the
loopback interface.
2024-07-12 16:29:44 +02:00
W.C.A. Wijngaards 3adb9c8f92 - Fix #1103: unbound 1.20.0 segmentation fault with nghttp2. 2024-07-12 16:11:29 +02:00
Yorgos Thessalonikefs 51425b2388 - Add RPZ tag tests in acl_interface.tdir. 2024-07-12 15:38:12 +02:00
W.C.A. Wijngaards d43760a8cd - For #773: In contrib/unbound.service.in set unbound to start after
network-online.target. Also for contrib/unbound_portable.service.in.
2024-07-10 14:05:43 +02:00
Yorgos Thessalonikefs ea3e327006 - Update list of known EDE codes. 2024-07-09 15:58:30 +02:00
W.C.A. Wijngaards be09350eca - Fix shadowed error string variable in validator dnskey handling. 2024-07-08 16:50:16 +02:00
W.C.A. Wijngaards 169acfc546 - Fixup algo_needs_reason string buffer length. 2024-07-08 15:38:27 +02:00
W.C.A. Wijngaards bed7cc2a90 - Fix that validation reason failure that uses string print uses
separate buffer that is passed, from the scratch validation buffer.
2024-07-08 15:29:20 +02:00
Yorgos Thessalonikefs 02f4446833 - Don't check for message TTL changes if the RRsets remain the same. 2024-07-05 19:58:19 +02:00
W.C.A. Wijngaards c8a2289542 - Fix for #1099: Fix to check for deleted RRset when the contents
is updated and fetched after it is stored, and also check for a
  changed RRset.
2024-07-05 17:54:46 +02:00
W.C.A. Wijngaards b53d90053e - Fix #1099: Unbound core dump on SIGSEGV. 2024-07-05 17:18:01 +02:00
W.C.A. Wijngaards 978b0696d3 - Fix neater printout. 2024-07-05 14:11:26 +02:00
W.C.A. Wijngaards ec5f86b4eb - Fix for neater printout for error for missing DS response. 2024-07-05 08:49:52 +02:00
W.C.A. Wijngaards ec2f45c6fd - Fix to print details about the failure to lookup a DNSKEY record
when validation fails due to the missing DNSKEY. Also for key prime
  and DS lookups.
2024-07-04 14:51:18 +02:00
W.C.A. Wijngaards 6b319c97ee - Fix compile warnings in fptr_wlist.c. 2024-07-03 16:42:52 +02:00
W.C.A. Wijngaards 6eb3992c9e - Fix to remove unneeded linebreak in fptr_wlist.c. 2024-07-03 15:51:22 +02:00
W.C.A. Wijngaards 94a94fd8c8 - Fix to use modstack_init in zonemd unit test. 2024-07-03 15:49:13 +02:00
W.C.A. Wijngaards 36f9d1a2a9 - Add unit test skip files and bison and flex output to gitignore. 2024-07-03 14:59:39 +02:00
W.C.A. Wijngaards d3a2264272 Changelog entry for #144 and #1098
- Fix #144: Port ipset to BSD pf tables.
2024-07-03 14:53:42 +02:00
Wouter WijngaardsandGitHub 92b6c2a7b9 Merge pull request #1098 from NLnetLabs/ipset-pf-support
Ipset pf support
2024-07-03 14:49:16 +02:00
W.C.A. Wijngaards a335e601e4 ipset-pf-support, move startup and destartup to the front of the module
func block functions, modstack call deinit function names, and detect
module change when no startup functions are needed.
2024-07-03 13:53:44 +02:00
a19009df1d Apply suggestions from code review
Co-authored-by: Yorgos Thessalonikefs <yorgos@nlnetlabs.nl>
2024-07-03 13:08:51 +02:00
Yorgos Thessalonikefs 96f8a94c19 - Fix for repeated use of a DNAME record: first overallocate and then
move the exact size of the init value to avoid false positive heap
  overflow reads from address sanitizers.
2024-07-03 10:08:44 +02:00
W.C.A. Wijngaards 2fe4e2ec3e - Fix compile warning in worker pthread id printout. 2024-07-02 09:44:58 +02:00
W.C.A. Wijngaards e54928a628 - Fix unused variable warning in do_cache_remove. 2024-07-02 09:33:22 +02:00
W.C.A. Wijngaards 538434186e - Fix to remove unused include from the readzone test program. 2024-07-02 09:31:34 +02:00
W.C.A. Wijngaards 65e7253d19 ipset-pf-support, simplification of code. 2024-07-02 09:08:27 +02:00
W.C.A. Wijngaards 03ac902296 - ipset-pf-support, fix to skip unit test if no pf dev. 2024-07-01 17:11:20 +02:00
W.C.A. Wijngaards 97ad1df343 ipset-pf-support, fix to log error on failure to open pf. 2024-07-01 17:06:11 +02:00
W.C.A. Wijngaards 2279cde8f7 ipset-pf-support, fix to remove unused include, free at end, adjust
qname for comparison.
2024-07-01 17:02:49 +02:00
W.C.A. Wijngaards ff653a7ef8 Call module init init again, and new function startup and destartup.
NULL can be used if the function is not used. Open shared ports during
reload. Deinit is called during reload.
2024-07-01 16:10:07 +02:00
W.C.A. Wijngaards fd11cd9182 ipset-pf-support, fix compilation, close of pf socket. 2024-07-01 14:54:45 +02:00
W.C.A. Wijngaards 3953f827fb Merge branch 'master' of https://github.com/madroach/unbound into ipset-pf-support 2024-07-01 14:36:33 +02:00
W.C.A. Wijngaards 7fbc061846 - Fix ip-ratelimit-cookie setting, it was not applied. 2024-06-27 14:51:58 +02:00
Yorgos Thessalonikefs 70f73a33b3 - Explicitly set the RD bit for the mesh query flags when prefetching.
These queries have no waiting client but they need to be treated as
  recursive.
2024-06-26 15:51:58 +02:00
Yorgos Thessalonikefs b67fbb69e7 - Fix pkg-config availability check in dnstap/dnstap.m4 and
systemd.m4.
- autoconf.
2024-06-21 14:34:12 +02:00
Yorgos Thessalonikefs 902c79608c - Fix #1092: Ubuntu 22.04 Jammy fails to compile unbound 1.20.0; by
adding helpful text for the Python interpreter version and allowing
  the default pkg-config unavailability error message to be shown.
- autoconf.
2024-06-19 15:27:50 +02:00
W.C.A. Wijngaards 08050dc939 - Fix #1091: Build fails with OpenSSL >= 3.0 built with
OPENSSL_NO_DEPRECATED.
2024-06-17 12:28:45 +02:00
W.C.A. Wijngaards 9603924bb4 - Add unit test for validation of repeated use of a DNAME record. 2024-06-07 11:56:19 +02:00
W.C.A. Wijngaards 4c2da2b979 - Fix validation for repeated use of a DNAME record. 2024-06-06 15:28:21 +02:00
W.C.A. Wijngaards 1974732d19 - Fix typos for 'the the' in text. 2024-06-06 09:35:57 +02:00
W.C.A. Wijngaards 3cad5818a1 - Fix memory leak in setup of dsa sig. 2024-06-06 09:30:09 +02:00
Yorgos Thessalonikefs ad12109191 - Merge #1080: AddressSanitizer detection in tdir tests and memory leak
fixes.
2024-06-04 17:34:58 +02:00
Yorgos Thessalonikefs f611220eb8 - Skip unbound-dnstap-socket unit test when not compiled with
--enable-debug.
2024-06-04 16:59:58 +02:00
W.C.A. Wijngaards 86fe9cbce5 - Fix to squelch connection reset by peer errors from log. And fix
that the tcp read errors are labeled as initial for the first calls.
2024-06-03 12:14:51 +02:00
Yorgos Thessalonikefs ac609fcbfc - Fix memory leak on exit for unbound-dnstap-socket; creates false negatives
during testing.
2024-05-31 12:11:17 +02:00
Yorgos Thessalonikefs 486985fbdf - Fix memory leak when reload_keep_cache is used and num-threads changes. 2024-05-31 12:09:35 +02:00
Yorgos Thessalonikefs caab100207 - Enable AddressSanitizer error detection in tdir tests. 2024-05-31 12:08:43 +02:00
W.C.A. Wijngaards 4b30e88eec - Fix for #1079: fix RPZ taglist in iterator callback that no client
info is like no taglist intersection.
2024-05-30 12:44:26 +02:00
W.C.A. Wijngaards b6c7ea563f - Fix #1079: tags from tagged rpz zones are no longer honored after
upgrade from 1.19.3 to 1.20.0.
2024-05-30 12:11:30 +02:00
W.C.A. Wijngaards 910d7cf446 Changelog note for #1078.
- Merge #1078: Only check old pid if no username.
2024-05-29 14:45:01 +02:00
Wouter WijngaardsandGitHub cd485f2036 Merge pull request #1078 from vopatek/master
Only check old pid if no username
2024-05-29 14:44:31 +02:00
Martin Vopatek 8ff1baf585 Only check old pid if no username
Do as the comment says and only check old pid if there is no username
configured.
2024-05-29 14:16:18 +02:00
Yorgos Thessalonikefs 5fc4673901 - Update patch to remove 'command' shell builtin and update error
text.
2024-05-27 17:17:48 +02:00
Michael TokarevandYorgos Thessalonikefs faf3d358dc unbound-control-setup: check openssl
Before doing anything, check if openssl binary (which we will use)
is available, and print a useful error message if it is not found.
2024-05-27 16:35:34 +02:00
Yorgos Thessalonikefs f5a2160ba3 - Fix unused variable warning on compilation with no thread support. 2024-05-27 14:56:52 +02:00
W.C.A. Wijngaards 0c0c36f015 - Fix spelling of tcp-idle-timeout docs, from Michael Tokarev. 2024-05-27 14:36:35 +02:00
W.C.A. Wijngaards 47956de897 - Fix to enable that SERVFAIL is cached, for a short period, for more
cases. In the cases where limits are exceeded.
2024-05-27 13:53:16 +02:00
Yorgos Thessalonikefs b30c869a59 Changelog entry for #1059:
- Fix #1059: Intermittent DNS blocking failure with local-zone and
  always_nxdomain. Addition of local_zones dynamically via
  unbound-control was not finding the zone's parent correctly.
2024-05-24 15:24:52 +02:00
Yorgos ThessalonikefsandGitHub 6f030e9672 Proper parent identification for dynamically entered local zones (#1076)
- Fix #1059: Intermittent DNS blocking failure with local-zone and
  always_nxdomain. Addition of local_zones dynamically via
  unbound-control was not finding the zone's parent correctly.
2024-05-24 15:21:40 +02:00
W.C.A. Wijngaards 7107d3c9e7 - Fix #1064: Unbound 1.20 Cachedb broken?
Add unit test for validation status commit.
2024-05-24 09:06:48 +02:00
W.C.A. Wijngaards fbdc06ebc4 - Fix for #1064: Fix that cachedb expired messages are considered
insecure, and thus can be served to clients when dnssec is enabled.
2024-05-21 17:06:18 +02:00
W.C.A. Wijngaards d149e755fd - Fix for parse end of forward-zone, stub-zone and view. 2024-05-21 12:04:57 +02:00
W.C.A. Wijngaards 86ee8ccd12 - Fix to print a parse error when config is read with no name for
a forward-zone, stub-zone or view.
2024-05-21 11:54:18 +02:00
W.C.A. Wijngaards 8d6a1ba811 Changelog note for #1073.
- Merge #1073: fix null pointer dereference issue in function
  ub_ctx_set_fwd.
2024-05-21 11:52:47 +02:00
Wouter WijngaardsandGitHub 9ccb8d5f20 Merge pull request #1073 from xiaoxiaoafeifei/master
fix null pointer dereference issue in function ub_ctx_set_fwd
2024-05-21 11:51:52 +02:00
zhailiangliang 8e43e2574c fix null pointer dereference issue in function ub_ctx_set_fwd of file libunbound/libunbound.c 2024-05-21 08:40:35 +00:00
Yorgos Thessalonikefs 2e70506763 Changelog entry for #1069:
- Merge #1069: Fix unbound-control stdin commands for multi-process
  Unbounds.
2024-05-17 10:31:20 +02:00
7f184c8ca8 Fix unbound-control stdin commands for multi-process Unbounds (#1069)
- Fix unbound-control commands that read stdin in multi-process
  operation (local_zones_remove, local_zones, local_datas_remove,
  local_datas, view_local_datas_remove, view_local_datas). They will
  be properly distributed to all processes. dump_cache and load_cache
  are no longer supported in multi-process operation.

 - Remove testdata/remote-threaded.tdir. testdata/09-unbound-control.tdir
  now checks both single and multi process/thread operation.

---------

Co-authored-by: Wouter Wijngaards <wcawijngaards@users.noreply.github.com>
2024-05-17 10:25:24 +02:00
W.C.A. Wijngaards da2b307aa3 - Fix #1071: [FR] Clear both in-memory and cachedb module cache with
`unbound-control flush*` commands.
2024-05-16 16:56:58 +02:00
Yorgos Thessalonikefs 739a88ceed Changelog entry for #1070:
- Merge #1070: Fix rtt assignement for low values of
  infra-cache-max-rtt.
2024-05-16 13:43:24 +02:00
Yorgos ThessalonikefsandGitHub 3ff5c7a74d Fix rtt assignement for low values of infra-cache-max-rtt (#1070)
* Fix rtt assignement for still useful servers when a low value for
  infra-cache-max-rtt is configured.
2024-05-16 13:42:32 +02:00
Yorgos Thessalonikefs 1048c4a28c - Add missing common functions to tdir tests. 2024-05-15 11:20:36 +02:00
W.C.A. Wijngaards 7de009f99a - Fix when the mesh jostle is exceeded that nameserver targets are
marked as resolved, so that the lookup is not stuck on the
  requestlist.
2024-05-10 09:50:35 +02:00
W.C.A. Wijngaards 95669855fb - Fix to squelch udp connect errors in the log at low verbosity about
invalid argument for IPv6 link local addresses.
2024-05-08 16:40:41 +02:00
W.C.A. Wijngaards 56e7cade28 The code repository continues with version 1.20.1. 2024-05-08 11:10:53 +02:00
W.C.A. Wijngaards c085a53268 - Fix for #1062: declaration before statement, avoid print of null,
and redundant check for array size.
And changelog note for merge of #1062.
2024-05-07 14:05:21 +02:00
Wouter WijngaardsandGitHub 49569b81aa Merge pull request #1062 from xiaoxiaoafeifei/master
Fix potential overflow bug while parsing port in function cfg_mark_ports
2024-05-07 14:02:21 +02:00
zhailiangliang 4497e8a154 Fix potential overflow bug while parsing port in function cfg_mark_ports 2024-05-07 11:35:52 +00:00
W.C.A. Wijngaards b9525c5fd4 - Set version number to 1.20.0 for release. 2024-05-01 10:15:12 +02:00
W.C.A. Wijngaards c3206f4568 - Fix for the DNSBomb vulnerability CVE-2024-33655. Thanks to Xiang Li
from the Network and Information Security Lab of Tsinghua University
  for reporting it.
2024-05-01 10:10:58 +02:00
W.C.A. Wijngaards 9abed3fc83 - Fix doxygen comment for errinf_to_str_bogus. 2024-04-29 13:42:26 +02:00
Yorgos Thessalonikefs 63a6b7b255 - Cleanup unnecessary strdup calls for EDE strings. 2024-04-29 10:15:19 +02:00
W.C.A. Wijngaards 15dc8e8a3f - Man page entry for unbound-checkconf -q. 2024-04-26 14:54:25 +02:00
Yorgos Thessalonikefs cd4a017e96 - Fix #876: [FR] can unbound-checkconf be silenced when configuration
is valid?
2024-04-26 14:50:39 +02:00
W.C.A. Wijngaards 82c0207fa6 - Add unit tests for cachedb and subnet cache expired data. 2024-04-26 13:33:26 +02:00
W.C.A. Wijngaards 7c5e765b3b - Fix cachedb with serve-expired-client-timeout disabled. The edns
subnet module deletes global cache and cachedb cache when it
  stores a result, and serve-expired is enabled, so that the global
  reply, that is older than the ecs reply, does not return after
  the ecs reply expires.
2024-04-26 13:32:15 +02:00
W.C.A. Wijngaards f456d97a34 - Fix doc unit test for out of directory build. 2024-04-25 17:06:06 +02:00
W.C.A. Wijngaards 8b490b1540 - Fix to disable fragmentation on systems with IP_DONTFRAG,
with a nonzero value for the socket option argument.
2024-04-25 12:53:05 +02:00
W.C.A. Wijngaards b3951e5885 Changelog note for #1041 and #1038.
- Merge #1041: Stub and Forward unshare. This has one structure
  for them and fixes #1038: fatal error: Could not initialize
  thread / error: reading root hints.
2024-04-25 11:12:27 +02:00
Wouter WijngaardsandGitHub ced9762b14 Merge pull request #1041 from NLnetLabs/stubfwd-unshare
Stub and Forward unshare
2024-04-25 11:11:00 +02:00
9b9bba9f02 Update locking management for iter_fwd and iter_hints methods. (#1054)
fast reload, move most of the locking management to iter_fwd and
iter_hints methods. The caller still has the ability to handle its
own locking, if desired, for atomic operations on sets of different
structs.

Co-authored-by: Wouter Wijngaards <wcawijngaards@users.noreply.github.com>
2024-04-25 11:05:58 +02:00
W.C.A. Wijngaards 07859a9ef3 - Fix configure flto check error, by finding grep for it. 2024-04-25 10:53:35 +02:00
W.C.A. Wijngaards cb74467acb - Fix ci workflow for macos for moved install locations. 2024-04-24 16:31:44 +02:00
Yorgos Thessalonikefs 62dad42152 - Merge #1053: Remove child delegations from cache when grandchild
delegations are returned from parent.
2024-04-23 14:24:07 +02:00
Yorgos Thessalonikefs 3ec74d1e3a - When a granchild delegation is returned, remove any cached child delegations
up to parent to not cause delegation invalidation because of an
  expired child delegation that would never be updated. Most likely to
  happen without qname-minimisation. Reported by Roland van Rijswijk-Deij.
2024-04-22 15:46:06 +02:00
W.C.A. Wijngaards 52aff65e35 - Fix edns subnet to sort rrset references when storing messages
in the cache. This fixes a race condition in the rrset locks.
2024-04-22 13:44:42 +02:00
W.C.A. Wijngaards 5994fb3db5 - Add checklock feature verbose_locking to trace locks and unlocks. 2024-04-22 13:42:35 +02:00
Yorgos Thessalonikefs 0dbcb45d28 Changelog entry for #1049:
- Merge #1049 from Petr Menšík: Py_NoSiteFlag is not needed since
  Python 3.8
2024-04-15 14:49:14 +02:00
Yorgos ThessalonikefsandGitHub d71f257215 Merge pull request #1049 from InfrastructureServices/python3.12-Py_NoSiteFlag-warning
Py_NoSiteFlag is not needed since Python 3.8
2024-04-15 14:47:45 +02:00
Petr Mensik b11d234f77 Py_NoSiteFlag is not needed since Python 3.8
Python since 3.12 prints warning about Py_NoSiteFlag is deprecated. It
seems that variable is not needed since Python 3.8, since it sets in
such cases directly config.site_import variable few moments later.
Move using deprecated variable to versions before that flag in config
could be used only.

This should fix warning like:

pythonmod/pythonmod.c: In function 'pythonmod_init':
pythonmod/pythonmod.c:359:7: warning: 'Py_NoSiteFlag' is deprecated [-Wdeprecated-declarations]
  359 |       Py_NoSiteFlag = 1;
      |       ^~~~~~~~~~~~~
In file included from /usr/include/python3.12/Python.h:48,
                 from pythonmod/pythonmod.c:54:
/usr/include/python3.12/cpython/pydebug.h:14:37: note: declared here
   14 | Py_DEPRECATED(3.12) PyAPI_DATA(int) Py_NoSiteFlag;
      |                                     ^~~~~~~~~~~~~

https://docs.python.org/3/c-api/init.html#c.Py_NoSiteFlag
2024-04-15 13:52:41 +02:00
W.C.A. Wijngaards 0d4c5aa421 - Fix configure, autoconf for #1048. 2024-04-15 12:17:56 +02:00
W.C.A. Wijngaards 9e60f93b84 Changelog note for #1048.
- Fix #1048: Update ax_pkg_swig.m4 and ax_pthread.m4.
2024-04-15 12:15:54 +02:00
Wouter WijngaardsandGitHub 9d63e8f093 Merge pull request #1048 from InfrastructureServices/configure-m4-refresh
Update ax_pkg_swig.m4 and ax_pthread.m4
2024-04-15 12:15:26 +02:00
Petr Mensik 4c40253df7 Update ax_pkg_swig.m4 and ax_pthread.m4
Use vanilla m4 files with known source. Prepared for possible removal at
build time if the system already has autoconf-archive source present.
Switch to AX_PKG_SWIG macro for versioned or unversioned swig detection.
2024-04-15 11:54:46 +02:00
W.C.A. Wijngaards 491b56d051 - Fixup cachedb to not refetch when serve-expired-client-timeout is
used.
2024-04-12 14:22:18 +02:00
W.C.A. Wijngaards 4d530920e0 - Fixup unit test for cachedb server expired client timeout with
a check if response if from upstream or from cachedb.
2024-04-12 11:51:00 +02:00
W.C.A. Wijngaards 08fb9a9209 - Fix cachedb for serve-expired with serve-expired-client-timeout. 2024-04-12 11:26:53 +02:00
W.C.A. Wijngaards 04ff2672b5 - Fix to not reply serve expired unless enabled for cachedb. 2024-04-10 17:06:01 +02:00
W.C.A. Wijngaards d47849a26e - Fix cachedb for serve-expired with serve-expired-reply-ttl. 2024-04-10 17:01:57 +02:00
W.C.A. Wijngaards 63ee97d0fd - Fix makefile dependencies for fake_event.c. 2024-04-10 14:04:39 +02:00
W.C.A. Wijngaards bd74a32b79 - Extended test for cachedb serve expired. 2024-04-10 13:08:23 +02:00
W.C.A. Wijngaards b990be88ef - Add test for cachedb serve expired. 2024-04-10 12:36:21 +02:00
W.C.A. Wijngaards cccf5e73c0 - Fixup compile without cachedb. 2024-04-10 11:33:52 +02:00
W.C.A. Wijngaards f3f85e5a11 - Fixup compile without cachedb. 2024-04-10 11:29:10 +02:00
W.C.A. Wijngaards d55511f1dd - Fixup compile without cachedb. 2024-04-10 11:27:08 +02:00
W.C.A. Wijngaards d98c7b9ae3 - Implement cachedb-check-when-serve-expired: yes option, default
is enabled. When serve expired is enabled with cachedb, it first
  checks cachedb before serving the expired response.
2024-04-10 11:21:28 +02:00
Yorgos Thessalonikefs a30221c5bb - Merge #1043 from xiaoxiaoafeifei: Add loongarch support; updates
config.guess(2024-01-01) and config.sub(2024-01-01), verified
  with upstream.
2024-04-09 17:00:59 +02:00
zhailiangliang 411e843fc4 Add loongarch support:
1. update config.guess(2024-01-01) and config.sub(2024-01-01)
  2. add loongarch to little endian support
2024-04-09 11:15:14 +08:00
Yorgos Thessalonikefs 8575d5b35c - Fix #595: unbound-anchor cannot deal with full disk; it will now
first write out to a temp file before replacing the original one,
  like Unbound already does for auto-trust-anchor-file.
2024-04-08 14:15:03 +02:00
W.C.A. Wijngaards ba16e41160 - Fix comment syntax for view function views_find_view. 2024-04-05 16:11:29 +02:00
Yorgos Thessalonikefs 708d5229ae - Merge #1027: Introduce 'cache-min-negative-ttl' option. 2024-04-05 11:44:37 +02:00
Yorgos Thessalonikefs fb4a7d65d7 - Fix #369: dnstap showing extra responses; for client responses
right from the cache when replying with expired data or
  prefetching.
2024-04-03 15:18:13 +02:00
W.C.A. Wijngaards d7353e6e99 - fast-reload, helpful comments for hints lookup function return value. 2024-04-03 13:55:57 +02:00
W.C.A. Wijngaards 48113cfaba - fast-reload, unshared stub hints, making the structure locked, with an rwlock. 2024-04-03 13:55:57 +02:00
W.C.A. Wijngaards c0b5754ef7 - fast-reload, for nonthreaded, the unbound-control commands forward,
forward_add and forward_delete should be distributed to other processes,
  but when threaded, they should not be distributed to other threads because
  the structure is not thread specific any more.
2024-04-03 13:55:57 +02:00
W.C.A. Wijngaards f2fb498c69 - fast-reload, unshare forwards, making the structure locked, with an rwlock. 2024-04-03 13:55:54 +02:00
Yorgos Thessalonikefs 91e8e0e511 - Fix #1035: Potential Bug while parsing port from the "stub-host"
string; also affected forward-zones and remote-control host
  directives.
2024-04-03 13:37:57 +02:00
W.C.A. Wijngaards dfff8d23cf - For #1040: adjust error text and disallow negative ports in other
parts of cfg_mark_ports.
2024-04-03 10:16:18 +02:00
W.C.A. Wijngaards 103d9a68fa Changelog note for #1040
- Fix #1040: fix heap-buffer-overflow issue in function cfg_mark_ports
  of file util/config_file.c.
2024-04-03 10:03:04 +02:00
Wouter WijngaardsandGitHub a814fe8c5f Merge pull request #1040 from xiaoxiaoafeifei/master
fix heap-buffer-overflow issue in function cfg_mark_ports of file util/config_file.c
2024-04-03 10:02:45 +02:00
zhailiangliang 193401e754 fix heap-buffer-overflow issue in function cfg_mark_ports of file util/config_file.c 2024-04-03 15:40:58 +08:00
W.C.A. Wijngaards e1aeabde44 - Fix for crypto related failures to have a better error string. 2024-03-28 09:58:57 +01:00
W.C.A. Wijngaards 6d1e61173b - Fix #1034: DoT forward-zone via unbound-control. 2024-03-28 09:58:03 +01:00
W.C.A. Wijngaards 6f82b5be4a - Fix that the server does not chown the pidfile. 2024-03-27 14:52:25 +01:00
W.C.A. Wijngaards 192f1b0e2b - Fix that when the server truncates the pidfile, it does not follow
symbolic links.
2024-03-27 14:07:54 +01:00
W.C.A. Wijngaards 238a796e38 - Fix to add unit test for lruhash space that exercises the routines. 2024-03-27 13:33:46 +01:00
W.C.A. Wijngaards fe393ac355 - Fix comment in lruhash space function. 2024-03-27 12:30:00 +01:00
W.C.A. Wijngaards 3ea078baf6 - Fix for #1032, add safeguard to make table space positive. 2024-03-27 11:49:20 +01:00
W.C.A. Wijngaards eb3e1ae24f - Fix #1032: The size of subnet_msg_cache calculation mistake cause
memory usage increased beyond expectations.
2024-03-27 11:45:34 +01:00
W.C.A. Wijngaards c2b20c585e - Fix name of unit test for subnet cache response. 2024-03-27 11:43:55 +01:00
Yorgos Thessalonikefs 07561964fc - For #831: Format text, use exclamation icon and explicit label
names.
2024-03-25 22:02:08 +01:00
Yorgos Thessalonikefs ce8c1ce5b0 Changelog entry for #831
- Merge #831 from Pierre4012: Improve Windows NSIS installer
  script (setup.nsi).
2024-03-25 16:46:25 +01:00
Pierre4012andGitHub ef60dcac31 Improve Windows NSIS installer script (setup.nsi) (#831)
* Improve Windows NSIS installer script (setup.nsi)

Two improvements of installer script :
- avoid error message when Unbound is running,
- add "DisplayVersion" in registry thus Windows package manager (Winget) can handle Unbound.

* Update setup.nsi ask user to stop unbound service + DisplayVersion in Windows registry
2024-03-25 16:43:49 +01:00
W.C.A. Wijngaards 73bd5a19aa - Fix localdata and rpz localdata to match CNAME only if no direct
type match is available.
2024-03-19 10:21:10 +01:00
W.C.A. Wijngaards fef974ca5c - Fix rpz so that rpz CNAME can apply after rpz CNAME. And fix that
clientip and nsip can give a CNAME.
2024-03-19 09:32:53 +01:00
W.C.A. Wijngaards 8dbf46913b - Fix rpz for qtype CNAME after nameserver trigger. 2024-03-18 14:36:29 +01:00
W.C.A. Wijngaards e46b188fe8 - Add rpz unit test for nsip action override. 2024-03-18 14:11:43 +01:00
W.C.A. Wijngaards e6b1f9a4c3 - Fix rpz that copies the cname override completely to the temp
region, so there are no references to the rpz region.
2024-03-18 13:52:59 +01:00
W.C.A. Wijngaards 39cfc8c1c0 - Fix rpz, it follows iterator CNAMEs for nsip and nsdname and sets
the reply query_info values, that is better for debug logging.
2024-03-18 12:45:00 +01:00
W.C.A. Wijngaards 79e25e192c - Fix that rpz CNAME content is limited to the max number of cnames. 2024-03-18 11:25:29 +01:00
Yorgos Thessalonikefs 792089f523 Merge branch 'features/makedist-persist-windir' 2024-03-15 17:22:00 +01:00
Yorgos Thessalonikefs 34636caa2d - For windows build, persist the openssl and expat directories for
repeated builds while debugging.
2024-03-15 16:59:46 +01:00
W.C.A. Wijngaards 2993437eaa - Fix that addrinfo is not kept around but copied and freed, so that
log-destaddr uses a copy of the information, much like NSD does.
2024-03-15 13:39:49 +01:00
W.C.A. Wijngaards 0bcc8c0211 - The code repository continues with version 1.19.4. 2024-03-14 10:33:13 +01:00
W.C.A. Wijngaards 4b54d8e15e - Fix rpz for cname override action after nsdname and nsip triggers. 2024-03-13 17:14:14 +01:00
W.C.A. Wijngaards afe52595a9 - Fix to unify codepath for local alias for rpz cname action override. 2024-03-13 16:12:48 +01:00
W.C.A. Wijngaards 4f417262e3 - Fix rpz that the rpz override is taken in case of clientip triggers.
Fix that the clientip passthru action is logged. Fix that the
  clientip localdata action is logged. Fix rpz override action cname
  for the clientip trigger.
2024-03-13 16:04:58 +01:00
W.C.A. Wijngaards 1db3b38104 - Fix #1029: rpz trigger clientip and action rpz-passthru not working
as expected.
2024-03-13 13:45:04 +01:00
Yorgos Thessalonikefs bc47f50926 Changelog entry for #1028:
- Merge #1028: Clearer documentation for tcp-idle-timeout and
  edns-tcp-keepalive-timeout.
2024-03-12 14:52:57 +01:00
Yorgos ThessalonikefsandGitHub e36b5a099c Clearer documentation for tcp-idle-timeout and edns-tcp-keepalive-timeout (#1028)
* - Clearer documentation for tcp-idle-timeout and
  edns-tcp-keepalive-timeout.

* - Address review comment.
2024-03-12 14:52:00 +01:00
Yorgos Thessalonikefs 025881d0e9 - Introduce 'cache-min-negative-ttl' option to bound the minimum TTL for
negative answers overriding 'cache-min-ttl'.
2024-03-12 11:24:59 +01:00
W.C.A. Wijngaards 320d0a5f1b - Fix #1021 Inconsistent Behavior with Changing rpz-cname-override
and doing a unbound-control reload.
2024-03-11 16:31:58 +01:00
W.C.A. Wijngaards d382210fce Update doc/Changelog to note the fixes included in 1.19.3rc2. 2024-03-11 12:30:24 +01:00
W.C.A. Wijngaards 7b62767e16 - Fix unbound-control-setup.cmd to have CA v3 basicConstraints,
like unbound-control-setup.sh has.
2024-03-08 17:18:05 +01:00
W.C.A. Wijngaards 6568841bb0 - Fix doc test so it ignores but outputs unsupported doxygen options. 2024-03-08 16:43:24 +01:00
W.C.A. Wijngaards e361f6b284 - Fix qname minimisation for reply with a DNAME for qtype CNAME that
answers it.
2024-03-08 16:33:17 +01:00
Yorgos Thessalonikefs 53766917ef - Update doc/unbound.doxygen with 'doxygen -u'. Fixes option
deprecation warnings and updates with newer defaults.
2024-03-08 16:13:36 +01:00
W.C.A. Wijngaards 2a255076f5 - Fix validator classification of qtype DNAME for positive and
redirection answers, and fix validator signature routine for dealing
  with the synthesized CNAME for a DNAME without previously
  encountering it and also for when the qtype is DNAME.
2024-03-08 14:10:06 +01:00
W.C.A. Wijngaards fb080e7853 - Remove unused portion from iter_dname_ttl unit test. 2024-03-08 09:51:37 +01:00
W.C.A. Wijngaards 0818841038 - Fix TTL of synthesized CNAME when a DNAME is used from cache. 2024-03-08 09:47:59 +01:00
W.C.A. Wijngaards 939baebfe7 - Fix unbound-control-setup.cmd to use 3072 bits so that certificates
are long enough for newer OpenSSL versions.
2024-03-08 09:07:36 +01:00
W.C.A. Wijngaards 326ba26522 - Version set to 1.19.3 for release. After 1.19.2 point release with
security fix for CVE-2024-1931, Denial of service when trimming
  EDE text on positive replies. The code repo includes the fix and
  is for version 1.19.3.
2024-03-07 11:06:42 +01:00
W.C.A. Wijngaards ec0b510f1c - Fix for #1022: Fix ede prohibited in access control refused answers. 2024-03-05 13:39:29 +01:00
W.C.A. Wijngaards be626f7c53 - Fix edns subnet replies for scope zero answers to not get stored
in the global cache, and in cachedb, when the upstream replies
  without an EDNS record.
2024-03-04 13:20:13 +01:00
W.C.A. Wijngaards 3096e4930e - Move github workflows to use checkoutv4. 2024-02-28 11:44:52 +01:00
Yorgos Thessalonikefs 33bdf44a04 - Document the suspend argument for process_ds_response(). 2024-02-23 14:34:33 +01:00
W.C.A. Wijngaards ccbe31c21f - Fix trim of EDE text from large udp responses from spinning cpu. 2024-02-22 16:22:31 +01:00
Yorgos Thessalonikefs c6746499c1 Changelog entry for #1010:
- Merge #1010: Mention REFUSED has the TC bit set with unmatched
  allow_cookie acl in the manpage. It also fixes the code to match the
  documentation about clients with a valid cookie that bypass the
  ratelimit regardless of the allow_cookie acl.
2024-02-20 15:33:18 +01:00
Willem TooropandGitHub e1229e375f Mention REFUSED has the TC bit set with unmatched allow_cookie acl in the manpage (#1010)
* Mention REFUSED with TC with unmatched allow_cookie acl in manpage

Also moved the part about bypassing ip-ratelimit to the ip-ratelimit
description as it will be bypassed with a valid DNS-Cookie regardless of the
allow_cookie acl.

* Apply suggestions from code review

* Update doc/unbound.conf.5.in

* DNS-Cookies should bypass ip-ratelimit setting
2024-02-20 15:29:34 +01:00
W.C.A. Wijngaards be27499d39 - These fixes are part of the 1.19.1 release, that is a security
point release on 1.19.0, the code repository continues with these
  fixes, with version number 1.19.2.
2024-02-13 14:03:30 +01:00
W.C.A. Wijngaards 56a2b564ef Merge commit '92f2a1ca690a44880f4c4fa70a4b5a4b029aaf1c' 2024-02-13 13:58:09 +01:00
W.C.A. Wijngaards 9a00877af9 Merge commit '882903f2fa800c4cb6f5e225b728e2887bb7b9ae' 2024-02-13 13:57:56 +01:00
W.C.A. Wijngaards 92f2a1ca69 - Fix CVE-2023-50868, NSEC3 closest encloser proof can exhaust CPU. 2024-02-13 13:02:43 +01:00
W.C.A. Wijngaards 882903f2fa - Fix CVE-2023-50387, DNSSEC verification complexity can be exploited to
exhaust CPU resources and stall DNS resolvers.
2024-02-13 13:02:08 +01:00
W.C.A. Wijngaards 54d86dd73b - Fix documentation for access-control in the unbound.conf man page. 2024-02-08 14:36:18 +01:00
Yorgos Thessalonikefs b496714caa - autoconf. 2024-02-07 10:51:16 +01:00
Yorgos Thessalonikefs 3f5175584b - For #1006: fix logic error introduced by previous fix. 2024-02-07 10:49:28 +01:00
Yorgos Thessalonikefs 11fff226f3 - autoheader, autoconf. 2024-02-07 10:42:39 +01:00
Yorgos Thessalonikefs 93490a0fc1 - Fix #1006: Can't find protobuf-c package since #999. 2024-02-07 10:38:52 +01:00
W.C.A. Wijngaards 0585c3e5fd Autoconf and changelog note for #999
- Merge #999: Search for protobuf-c with pkg-config.
2024-01-30 16:24:41 +01:00
Wouter WijngaardsandGitHub 0b74f2a007 Merge pull request #999 from NickCao/master
Search for protobuf-c with pkg-config
2024-01-30 16:23:43 +01:00
Nick Cao 59d98b9ef6 Search for protobuf-c with pkg-config 2024-01-26 17:52:24 -05:00
Yorgos Thessalonikefs 3522451600 - Update message TTL when using cached RRSETs. It could result in
non-expired messages with expired RRSETs (non-usable messages by
  Unbound).
2024-01-23 10:10:37 +01:00
Yorgos Thessalonikefs fe03bacd6c - Update error printout for duplicate trust anchors to include the
trust anchor name (relates to #920).
2024-01-22 15:54:36 +01:00
W.C.A. Wijngaards 1f46d5945b - Fix for #997: Print details for SSL certificate failure. 2024-01-22 09:40:36 +01:00
W.C.A. Wijngaards 585d73bf7c For analysis workflow, clean up the script to use OpenSSL Configure
without change.
2024-01-17 16:23:18 +01:00
W.C.A. Wijngaards 9e84cebfdb - workflow for analysis, cleanup of windows compile with msys2 perl. 2024-01-17 16:20:22 +01:00
W.C.A. Wijngaards d1a2bd67da - Fix warning for windres on resource files due to redefinition. 2024-01-17 16:19:56 +01:00
W.C.A. Wijngaards 7708429d35 For workflow, set perl interpreter for build. 2024-01-17 15:08:56 +01:00
W.C.A. Wijngaards 6045911d95 - Fix for workflow 2024-01-17 14:58:44 +01:00
W.C.A. Wijngaards 1b212aa073 For workflow, look for pacman. 2024-01-17 14:56:10 +01:00
W.C.A. Wijngaards 4d48166835 For workflow, use msys2 perl. 2024-01-17 14:53:23 +01:00
W.C.A. Wijngaards 437bac370a In workflow, use strawberry perl to run configure script. 2024-01-17 14:40:57 +01:00
W.C.A. Wijngaards 546062d3a3 For workflow, change path separator. 2024-01-17 14:36:35 +01:00
W.C.A. Wijngaards 576b93c99f For windows runner, look at perl contents. 2024-01-17 14:30:48 +01:00
W.C.A. Wijngaards 180275c4e0 Fix to install with cpanmin a missing perl module for the windows workflow. 2024-01-17 14:07:57 +01:00
W.C.A. Wijngaards 379e4b68f5 Fix for workflow to install perl module. 2024-01-17 14:03:30 +01:00
W.C.A. Wijngaards 74b4d81992 - Update workflow for ports to use newer openssl on windows compile. 2024-01-17 13:45:59 +01:00
W.C.A. Wijngaards fea8f0d5fd Changelog note for #993
- Merge #993: Update b.root-servers.net also in example config file.
2024-01-16 16:44:15 +01:00
Wouter WijngaardsandGitHub 0e5dab5eaf Merge pull request #993 from InfrastructureServices/b.root-servers.net-conf
Update b.root-servers.net also in example config file
2024-01-16 16:44:02 +01:00
W.C.A. Wijngaards c550bc154f - Fix to link with libssp for libcrypto and getaddrinfo check for
only header. Also update crosscompile to remove ssp for 32bit.
2024-01-16 16:40:14 +01:00
Petr Mensik 40fcb91206 Update b.root-servers.net also in example config file
Addition to commit a8739bad76, which
updated only address specified in code. But addresses provided in
example configuration were not updated, I think they should be updated
too.
2024-01-16 16:14:13 +01:00
W.C.A. Wijngaards c8554ff48c - Fix to link with -lcrypt32 for OpenSSL 3.2.0 on Windows. 2024-01-15 16:44:27 +01:00
W.C.A. Wijngaards 3d95cef08c Changelog note for #988.
- Merge #988: Fix NLnetLabs#981: dump_cache truncates large records.
2024-01-09 08:41:52 +01:00
Wouter WijngaardsandGitHub 9cd724cf5e Merge pull request #988 from dyunwei/master
Fix NLnetLabs#981: dump_cache truncates large records.
2024-01-09 08:41:30 +01:00
dyunwei eb7eb5ce68 Fix NLnetLabs#981: dump_cache truncates large records. 2024-01-09 14:17:31 +08:00
W.C.A. Wijngaards 418eeb642c - Fix unit test for #987 change in udp1xxx retry packet send. 2024-01-05 14:11:55 +01:00
W.C.A. Wijngaards 8ac56d004d Changelog note for #987
- Merge #987: skip edns frag retry if advertised udp payload size is
  not smaller.
2024-01-05 13:47:30 +01:00
Wouter WijngaardsandGitHub 52a76583c5 Merge pull request #987 from borisVanhoof/skip_edns_frag_retry
skip edns frag retry if advertised udp payload size is not smaller
2024-01-05 13:47:15 +01:00
sahnalys12 b1d02cc94f skip edns frag retry if advertised udp payload size is not smaller
If serviced query is in UDP_EDNS_FRAG mode, and EDNS_ADVERTISED_SIZE
is 1232 (the default) or more, then the retry will have the same edns
udp payload size with the same result.
2024-01-05 12:16:23 +01:00
W.C.A. Wijngaards b9b488b6d3 - Remove unneeded newlines and improve indentation in remote control
code.
2024-01-04 17:06:15 +01:00
W.C.A. Wijngaards 9a2d0238a8 - Fix #983: Sha1 runtime insecure change was incomplete. 2024-01-03 13:33:43 +01:00
W.C.A. Wijngaards 5cc21690eb Changelog note for #985.
- Merge #985: Add DoH and DoT to dnstap message.
2024-01-03 10:37:44 +01:00
Wouter WijngaardsandGitHub f80f65d58c Merge pull request #985 from k-akashi/dnstap_dot_doh
Add DoH and DoT to dnstap message
2024-01-03 10:36:38 +01:00
W.C.A. Wijngaards df284fbe65 Changelog note for #979 and #980.
- Merge #980: DoH: reject non-h2 early. To fix #979: Improve errors
  for non-HTTP/2 DoH clients.
2024-01-03 10:04:06 +01:00
Wouter WijngaardsandGitHub b632ebf129 Merge pull request #980 from jsha/reject-non-h2
DoH: reject non-h2 early
2024-01-03 09:59:15 +01:00
k-akashi 4b9cd8e81d Add DoH and DoT to dnstap message 2023-12-27 07:26:21 +09:00
Yorgos Thessalonikefs 6c82f4ae9b - Update example.conf with cookie options. 2023-12-22 11:42:20 +01:00
Jacob Hoffman-Andrews 87644fa46b DoH: reject non-h2 early
Previously, non-h2 connections would be accepted, and then error out
with a verbose message "http2: session_recv from ____ failed,
error: Received bad client magic byte string". Instead, we can detect
absence of h2 support at connection time and reject with a clearer
verbose message.
2023-12-13 16:58:44 -08:00
W.C.A. Wijngaards 22dc376392 Fixup doc/Changelog. 2023-12-08 17:17:23 +01:00
W.C.A. Wijngaards 4ef1fb5a24 - Fix root_zonemd unit test, it checks that the root ZONEMD verifies,
now that the root has a valid ZONEMD.
2023-12-08 17:15:35 +01:00
Yorgos Thessalonikefs be6fd80a1c - Merge PR #973: Use the origin (DNAME) TTL for synthesized CNAMEs as per RFC 6672. 2023-12-08 09:23:26 +01:00
W.C.A. Wijngaards c48a467b8e Changelog note for #975
- Merge #975: Fixed some syntax errors in rpl files.
2023-12-07 12:11:03 +01:00
Wouter WijngaardsandGitHub 557edbb84a Merge pull request #975 from NLnetLabs/testdata-syntax
Fixed some syntax errors in rpl files.
2023-12-07 12:10:40 +01:00
Philip Homburg 0cfc6e6d95 Fixed some syntax errors in rpl files. 2023-12-07 11:38:01 +01:00
W.C.A. Wijngaards 2978106991 - Fix #974: doc: default number of outgoing ports without libevent. 2023-12-07 09:41:03 +01:00
Yorgos Thessalonikefs 8517f49745 - Use the origin (DNAME) TTL for syntesized CNAMEs as per RFC 6672. 2023-12-06 23:40:01 +01:00
W.C.A. Wijngaards 6c76b4e2f8 - Fix tests to use new common.sh functions, wait_logfile and
kill_from_pidfile.
2023-12-06 16:32:06 +01:00
W.C.A. Wijngaards c91bd60051 - Update test script file common.sh. 2023-12-06 16:14:24 +01:00
W.C.A. Wijngaards a8739bad76 - Updated IPv4 and IPv6 address for b.root-servers.net in root hints. 2023-12-06 13:25:58 +01:00
W.C.A. Wijngaards 2b97442f2e - iana portlist update. 2023-12-06 13:22:35 +01:00
W.C.A. Wijngaards 5997355e22 - Fix to sync the tests script file common.sh. 2023-12-06 11:58:14 +01:00
W.C.A. Wijngaards d8bd9845cc - Fix dnstap that assertion failed on logging other than UDP and TCP
traffic. It lists it as TCP traffic.
2023-12-05 13:14:08 +01:00
W.C.A. Wijngaards 3d1bc143af - Fix #969: [FR] distinguish Do53, DoT and DoH in the logs. 2023-12-05 10:05:51 +01:00
W.C.A. Wijngaards 72a25f825c Changelog note for #971
- Merge #971: fix 'WARNING: Message has 41 extra bytes at end'.
2023-12-05 09:09:42 +01:00
Wouter WijngaardsandGitHub c4e0354876 Merge pull request #971 from dukeartem/master
fix 'WARNING: Message has 41 extra bytes at end'
2023-12-05 09:09:23 +01:00
ArtemandGitHub 389e820878 fix 'WARNING: Message has 41 extra bytes at end' 2023-12-05 09:43:15 +03:00
Yorgos Thessalonikefs a2f0186427 - Merge #968: Replace the obsolescent fgrep with grep -F in tests. 2023-11-27 16:47:13 +01:00
W.C.A. Wijngaards 58d670a258 - Fix #964: config.h.in~ backup file in release tar balls. 2023-11-27 16:04:33 +01:00
Yorgos Thessalonikefs ca88669435 - Replace the obsolescent fgrep with grep -F in tests. 2023-11-24 15:51:17 +01:00
Yorgos Thessalonikefs 15a9b0f2be - Use 127.0.0.1 explicitly in tests to avoid delays and errors on newer
systems.
2023-11-24 15:34:25 +01:00
W.C.A. Wijngaards 197bf15402 - Fix unit test parse of origin syntax. 2023-11-09 15:26:46 +01:00
W.C.A. Wijngaards 0bae242cbf - The repository continues with 1.19.1. 2023-11-08 10:59:25 +01:00
W.C.A. Wijngaards 3352b1090e - Set version number to 1.19.0.
- Tag for 1.19.0rc1 release.
2023-11-02 08:40:20 +01:00
W.C.A. Wijngaards c4d17dd231 - Fix compilation without openssl, remove unused function warning. 2023-11-01 17:09:37 +01:00
W.C.A. Wijngaards 5f78f67e39 - Fix SSL compile failure for other missing definitions in
log_crypto_err_io_code_arg.
2023-11-01 14:20:52 +01:00
W.C.A. Wijngaards b1d99bb6b6 - Fix SSL compile failure for definition in log_crypto_err_io_code_arg. 2023-11-01 14:14:02 +01:00
George Thessalonikefs 8914e9fd03 - Mention flex and bison in README.md when building from repository
source.
2023-11-01 13:57:06 +01:00
George Thessalonikefs 8d1d728d88 - Fix #941: dnscrypt doesn't work after upgrade to 1.18 with
suggestion by dukeartem to also fix the udp_ancil with dnscrypt.
2023-10-31 22:41:06 +01:00
George Thessalonikefs 59c14c747a Changelog entry for #930
- Merge #930 from Stuart Henderson: add void to
  log_ident_revert_to_default declaration.
2023-10-30 12:18:01 +01:00
Yorgos ThessalonikefsandGitHub ccdf29a5f8 Merge pull request #930 from sthen/patch-1
add void to log_ident_revert_to_default declaration
2023-10-30 11:53:39 +01:00
W.C.A. Wijngaards a7e079ea16 - autoconf. 2023-10-30 10:44:23 +01:00
George Thessalonikefs a97bed9d22 - Clearer configure text for missing protobuf-c development libraries. 2023-10-24 16:34:12 +02:00
W.C.A. Wijngaards 0ce68e97a7 Changelog entry for #951.
- Merge #951: Cachedb no store. The cachedb-no-store: yes option is
  used to stop cachedb from writing messages to the backend storage.
  It reads messages when data is available from the backend. The
  default is no.
2023-10-20 17:01:13 +02:00
Wouter WijngaardsandGitHub 3f66230874 Merge pull request #951 from NLnetLabs/cachedb-no-store
Cachedb no store
2023-10-20 17:00:13 +02:00
W.C.A. Wijngaards 35d0a8a843 - Fix to print detailed errors when an SSL IO routine fails via
SSL_get_error.
2023-10-19 11:17:32 +02:00
George Thessalonikefs 44c3d4d2dc - Changelog entry for:
Merge #955 from buevsan: fix ipset wrong behavior.
- Update testdata/ipset.tdir test for ipset fix.
2023-10-18 15:11:38 +02:00
Yorgos ThessalonikefsandGitHub 167772fbca Merge pull request #955 from buevsan/fix-ipset
fix ipset wrong behavior
2023-10-18 15:08:08 +02:00
George Thessalonikefs 2f0b11673a - Update the dns64_lookup.rpl test for the DNS64 fallback patch. 2023-10-18 12:59:54 +02:00
George Thessalonikefs c5aa6a2286 - Changelog entry for DNS64 patches from Daniel Gröber. 2023-10-18 12:16:35 +02:00
George Thessalonikefs d5522c3480 Fixes for dns64 fallback to plain AAAA when no A records:
- Cleanup if condition.
- Rename variable for readability.
2023-10-18 12:03:40 +02:00
Daniel Gröber via Unbound-usersandGeorge Thessalonikefs c1e5e6781e dns64: Fall back to plain AAAA query with synthall but no A records
Networks which only have tunneled IPv6 access but still want to go
IPv6-only internally can use unbound's DNS64 module together with the
dns64-synthall or dns64-ignore-aaaa options to direct most traffic (any
dualstack domain) to their NAT64.

There is only one problem with this setup, currently domains with only AAAA
records will fail to resolve.

To allow for this use-case arrange for the A sub-query to make the AAAA
super query advance along the module stack when no records are returned.

Signed-off-by: Daniel Gröber <dxld@darkboxed.org>
2023-10-18 12:03:40 +02:00
George Thessalonikefs dd086e5bfd Fixes for dns64 readability refactoring:
- Move declarations to the top for C90 compliance.
- Save cycles by not calling (yet) unneeded functions.
- Possible use of uninitialised value.
- Consistent formatting.
2023-10-18 12:03:40 +02:00
Daniel Gröber via Unbound-usersandGeorge Thessalonikefs 213bb7c6ed dns64: Fix misleading indentation
Signed-off-by: Daniel Gröber <dxld@darkboxed.org>
2023-10-18 12:03:40 +02:00
Daniel Gröber via Unbound-usersandGeorge Thessalonikefs 0c88f98a3b dns64: Refactor handle_event checks for readability
No functional change intended.

Signed-off-by: Daniel Gröber <dxld@darkboxed.org>
2023-10-18 12:03:40 +02:00
Alexander V. Buev 23ae0a9838 fix ipset wrong behavior
Issue description:

    If local zone directive is assigned as following:

      "local-zone domain.com ipset"

    then any answers for query with names such as
    "anytext1domain.com" or "example2domain.com"
    will be added to ipset.

  This commit fixes this issue.
2023-10-17 18:35:42 +03:00
W.C.A. Wijngaards 0f78bea4a3 - Fix #954: Inconsistent RPZ handling for A record returned along with
CNAME.
2023-10-17 16:47:04 +02:00
George Thessalonikefs 4b627bd29e - Update pymod tests for the new Python script variable. 2023-10-16 16:32:09 +02:00
George Thessalonikefs e4510c76e5 - For multi Python module setups, clean previously parsed module
functions in __main__'s dictionary, if any, so that only current
  module functions are registered.
2023-10-16 16:03:11 +02:00
George Thessalonikefs 122dd6c11e - Expose the configured listening and outgoing interfaces, if any, as
a list of strings in the Python 'config_file' class instead of the
  current Swig object proxy; fixes #79.
2023-10-16 15:53:47 +02:00
George Thessalonikefs 63a5280f8f - Expose the script filename in the Python module environment 'mod_env'
instead of the config_file structure which includes the linked list
  of scripts in a multi Python module setup; fixes #79.
2023-10-16 15:47:18 +02:00
George Thessalonikefs 07149f576a - Better fix for infinite loop when reading multiple lines of input on
a broken remote control socket, by treating a zero byte line the
  same as transmission end. Addesses #947 and #948.
2023-10-13 14:58:16 +02:00
dbd2a43ab1 Apply suggestions from code review
Co-authored-by: Yorgos Thessalonikefs <george@nlnetlabs.nl>
2023-10-13 13:46:52 +02:00
W.C.A. Wijngaards 4a211a9117 - cachedb-no-store, example conf and man page documentation. 2023-10-13 11:37:18 +02:00
W.C.A. Wijngaards 18ebe165ba Merge branch 'master' into cachedb-no-store 2023-10-12 14:51:12 +02:00
W.C.A. Wijngaards 908e1cb11a Changelog note for #944.
- Merge #944: Disable EDNS DO.
  Disable the EDNS DO flag in upstream requests. This can be helpful
  for devices that cannot handle DNSSEC information. But it should not
  be enabled otherwise, because that would stop DNSSEC validation. The
  DNSSEC validation would not work for Unbound itself, and also not
  for downstream users. Default is no. The option
  is disable-edns-do: no
2023-10-12 14:05:31 +02:00
Wouter WijngaardsandGitHub 5c6c57ed89 Merge pull request #944 from NLnetLabs/disable-edns-do
Disable EDNS DO
2023-10-12 14:04:29 +02:00
W.C.A. Wijngaards 47094fd83f Merge branch 'master' into cachedb-no-store 2023-10-11 13:51:34 +02:00
W.C.A. Wijngaards 67153f897e - Fix for #949: Fix pythonmod/ubmodule-tst.py for Python 3.x. 2023-10-11 13:47:28 +02:00
W.C.A. Wijngaards f2528dc3ac - Fix that cachedb does not warn when serve-expired is disabled about
use of serve-expired-reply-ttl and serve-expired-client-timeout.
2023-10-11 13:29:56 +02:00
W.C.A. Wijngaards 935bc162e1 - cachedb-no-store, unit test cachedb_no_store.tdir. 2023-10-11 12:01:55 +02:00
W.C.A. Wijngaards d5954aff08 - Fix #949: "could not create control compt". 2023-10-11 11:59:26 +02:00
George Thessalonikefs e98b89651e - Fix #850: [FR] Ability to use specific database in Redis, with new
redis-logical-db configuration option.
2023-10-11 11:44:55 +02:00
George Thessalonikefs 516f90abdb - Fix infinite loop when reading multiple lines of input on a broken
remote control socket. Addesses #947 and #948.
2023-10-10 15:17:48 +02:00
W.C.A. Wijngaards c09320c651 - Fix that printout of EDNS options shows the EDNS cookie option by
name.
2023-10-09 12:36:54 +02:00
W.C.A. Wijngaards 6d0812b567 - Fix edns subnet so that queries with a source prefix of zero cause
the recursor send no edns subnet option to the upstream.
2023-10-09 12:21:22 +02:00
b05154218c Update doc/unbound.conf.5.in
Co-authored-by: Yorgos Thessalonikefs <george@nlnetlabs.nl>
2023-10-06 16:40:34 +02:00
c8ae3de610 Update validator/validator.c
Co-authored-by: Yorgos Thessalonikefs <george@nlnetlabs.nl>
2023-10-06 16:39:33 +02:00
W.C.A. Wijngaards ae96aa0a6d - cachedb-no-store, implement cachedb-no-store: yes configuration option. 2023-10-06 13:22:10 +02:00
W.C.A. Wijngaards 1e85749e6e Merge branch 'master' into disable-edns-do 2023-10-05 15:57:41 +02:00
W.C.A. Wijngaards b624ed5050 - disable-edns-do, validator init prints warning when disable-edns-do is
turned on, but there are trust anchors, and then turns off disable-edns-do.
2023-10-05 14:33:22 +02:00
W.C.A. Wijngaards b865aca03a - Fix #946: Forwarder returns servfail on upstream response noerror no
data.
2023-10-04 18:16:22 +02:00
W.C.A. Wijngaards 2e7714e80c - disable-edns-do, unbound-checkconf warns about disable-edns-do and DNSSEC
validation that is enabled, and suggests to turn one off.
2023-10-04 15:28:52 +02:00
W.C.A. Wijngaards 39df4f0923 - disable-edns-do, queriers receive no EDNS in response if the
disable-edns-do option is enabled and they set the DO flag. And unit test
  for that.
2023-10-04 13:54:05 +02:00
W.C.A. Wijngaards eff3e01ec3 Merge branch 'master' into disable-edns-do 2023-10-04 13:34:47 +02:00
George Thessalonikefs 13d4504dfc - Merge #881: Generalise the proxy protocol code. 2023-10-03 14:51:50 +02:00
George Thessalonikefs 9342bf685e - Fix misplaced comment. 2023-10-02 16:13:23 +02:00
George Thessalonikefs f804c087e4 proxy-protocol, review comments:
- more generic switch statement for address families;
- comment the protocol values as such in their definitions;
- less hardcoded values for address family and protocol combinations.
2023-09-29 17:31:52 +02:00
W.C.A. Wijngaards 0102360e92 root-zonemd-test, add test for ZONEMD usage from the root zone,
currently with the unsupported algorithm.
2023-09-29 16:03:08 +02:00
W.C.A. Wijngaards ca481322d4 Merge branch 'master' into disable-edns-do 2023-09-28 16:49:01 +02:00
W.C.A. Wijngaards 654a7eab62 - Fix #942: 1.18.0 libunbound DNS regression when built without
OpenSSL.
2023-09-22 13:15:35 +02:00
W.C.A. Wijngaards 4e5b0b7eec - disable-edns-do, unit test checks lookup without EDNS DO flag. 2023-09-22 11:39:39 +02:00
W.C.A. Wijngaards 9aaafddf04 - disable-edns-do, the option turns of the EDNS DO flag when a message is
sent from the iterator.
2023-09-20 13:28:06 +02:00
W.C.A. Wijngaards 9cd282e001 Merge branch 'master' into disable-edns-do 2023-09-20 13:18:26 +02:00
W.C.A. Wijngaards bd5dc855af - Fix rpz tcp-only action with rpz triggers nsdname and nsip. 2023-09-18 09:55:39 +02:00
W.C.A. Wijngaards 31218166fc - Fix to remove two c99 notations. 2023-09-15 13:30:30 +02:00
W.C.A. Wijngaards 6bdecdbc5a Changelog note and autoconf for #936
- Merge #936: Check for c99 with autoconf versions prior to 2.70.
2023-09-15 13:24:20 +02:00
Wouter WijngaardsandGitHub 45e6e71652 Merge pull request #936 from fobser/c99
Check for c99 with autoconf versions prior to 2.70
2023-09-15 13:23:28 +02:00
Florian Obser 3abb32c3ba Check for c99 with autoconf versions prior to 2.70 2023-09-15 13:00:17 +02:00
W.C.A. Wijngaards 6e65343895 - Fix authority zone answers for obscured DNAMEs and delegations. 2023-09-14 11:37:49 +02:00
W.C.A. Wijngaards d1977c679b - disable-edns-do, doc and add option disable-edns-do: no. 2023-09-13 13:11:53 +02:00
W.C.A. Wijngaards 0ee44ef384 - Fix send of udp retries when ENOBUFS is returned. It stops looping
and also waits for the condition to go away. Reported by Florian
  Obser.
2023-09-08 13:35:42 +02:00
Philip Homburg 1c8f0e0fc5 Avoid calling comm_point_udp_ancil_callback from comm_point_create_udp 2023-09-07 16:35:22 +02:00
Philip Homburg 17a557dfd5 Fix #928 (1.18 doesn't start on macOS/SunOS) 2023-09-07 16:35:22 +02:00
W.C.A. Wijngaards 8c751d48a2 - Fix to print EDE text in readable form in output logs. 2023-09-07 15:28:01 +02:00
W.C.A. Wijngaards 5b8a7340bb - Fix to set ede match in unit test for rr length removal.
Changelog note.
2023-09-07 14:59:13 +02:00
W.C.A. Wijngaards 33784e612a - Fix to set ede match in unit test for rr length removal. 2023-09-07 14:58:51 +02:00
W.C.A. Wijngaards fdd5f8ff83 - Fix to add EDE text when RRs have been removed due to length. 2023-09-07 14:44:48 +02:00
W.C.A. Wijngaards 63616a5fce - Fix to move msgparse_rrset_remove_rr code to util/msgparse.c. 2023-09-07 11:29:53 +02:00
W.C.A. Wijngaards dfc00271d1 - Fix to scrub resource records of type A and AAAA that have an
inappropriate size. They are removed from responses.
2023-09-07 11:08:04 +02:00
W.C.A. Wijngaards 1143050ea6 Changelog note for #931
- Merge #931: Prevent warnings from -Wmissing-prototypes.
2023-09-06 09:49:35 +02:00
Wouter WijngaardsandGitHub 5ab4a9330a Merge pull request #931 from fobser/missing-prototypes
Prevent warnings from -Wmissing-prototypes.
2023-09-06 09:49:23 +02:00
Florian Obser 2cd7c719ef Prevent warnings from -Wmissing-prototypes. 2023-09-05 17:35:30 +02:00
Stuart HendersonandGitHub ed00129866 add void to log_ident_revert_to_default declaration
Avoid warning from LLVM 16:

util_log.c:190:33: warning: a function declaration without a prototype is deprecated in all versions of C             
[-Wstrict-prototypes]                                                                                                 
void log_ident_revert_to_default()                                                                                    
                                ^                                                                                     
                                 void
2023-09-05 14:28:06 +00:00
W.C.A. Wijngaards 10843805ac - Fix #927: unbound 1.18.0 make test error. Fix make test without SHA1. 2023-08-31 13:54:03 +02:00
W.C.A. Wijngaards f0e15272b2 - Fix autoconf 2.69 warnings in configure. 2023-08-31 09:19:24 +02:00
W.C.A. Wijngaards d4c4537b33 - Fix for WKS call to getservbyname that creates allocation on exit
in unit test by testing numbers first and testing from the services
  list later.
2023-08-30 14:31:24 +02:00
W.C.A. Wijngaards ba1183bb6e Tag 1.18.0rc1 became the 1.18.0 release on
30 aug 2023, with the fix from 25 aug, fix compile on NetBSD
included. The repository continues with version 1.18.1.
2023-08-30 11:03:42 +02:00
W.C.A. Wijngaards 85ee5284f5 - Fix for version generation race condition that ignored changes. 2023-08-28 13:25:04 +02:00
W.C.A. Wijngaards 3795e37410 - Fix compile error on NetBSD in util/netevent.h. 2023-08-25 08:43:27 +02:00
W.C.A. Wijngaards d396d8dae3 - Tag for 1.18.0rc1 release. 2023-08-23 10:49:45 +02:00
W.C.A. Wijngaards 8da117b086 - Set version number to 1.18.0. 2023-08-22 14:32:51 +02:00
W.C.A. Wijngaards 6e43145382 - Fix unit test for unbound-control to work when threads are disabled,
and fix cache dump check.
2023-08-21 16:58:10 +02:00
W.C.A. Wijngaards be53e37b15 - Fix #923: processQueryResponse() THROWAWAY should be mindful of
fail_reply.
2023-08-21 14:32:13 +02:00
W.C.A. Wijngaards 3160d6ac08 - Fix for #925: unbound.service: Main process exited, code=killed,
status=11/SEGV. Fixes cachedb configuration handling.
2023-08-21 11:28:49 +02:00
W.C.A. Wijngaards d1f388ec93 - Fix windows ci workflow to install bison and flex. 2023-08-21 10:43:02 +02:00
W.C.A. Wijngaards 5bd07f8e88 Further debug for windows ci workflow. 2023-08-21 10:20:23 +02:00
W.C.A. Wijngaards 3ea6f730ac - Debug Windows ci workflow. 2023-08-21 10:16:00 +02:00
W.C.A. Wijngaards 083770cb39 - Fix stat_values test to work with dig that enables DNS cookies. 2023-08-18 13:39:27 +02:00
W.C.A. Wijngaards 8756ad63dd - Fix uninitialized memory passed in padding bytes of cmsg to sendmsg. 2023-08-18 13:18:46 +02:00
W.C.A. Wijngaards c4566aa5d3 Changelog for commit.
- Fix for iter_dec_attempts that could cause a hang, part of
  capsforid and qname minimisation, depending on the settings.
2023-08-18 13:10:51 +02:00
W.C.A. Wijngaards 2791ccbe02 - Fix for iter_dec_attempts that could cause a hang, part of
capsforid and qname minimisation, depending on the settings.
2023-08-18 09:11:06 +02:00
W.C.A. Wijngaards 5f423906de - Fix ip_ratelimit test to work with dig that enables DNS cookies. 2023-08-17 15:43:14 +02:00
W.C.A. Wijngaards 4844fa3481 - Fix regional_alloc_init for potential unaligned source of the copy. 2023-08-17 15:22:54 +02:00
W.C.A. Wijngaards 0f5fecd516 Changelog note for #762.
- Merge PR #762: Downstream DNS Server Cookies a la RFC7873 and
  RFC9018. Create server cookies for clients that send client cookies.
  This needs to be explicitly turned on in the config file with:
  `answer-cookie: yes`. A `cookie-secret:` can be configured for
  anycast setups. Without one, a random cookie secret is generated.
  The acl option `allow_cookie` allows queries with either a valid
  cookie or over a stateful transport. The statistics output has
  `queries_cookie_valid` and `queries_cookie_client` and
  `queries_cookie_invalid` information. The `ip\-ratelimit\-cookie:`
  value determines a rate limit for queries with cookies, if desired.
2023-08-17 15:18:46 +02:00
Wouter WijngaardsandGitHub a1c82ac5fd Merge pull request #762 from NLnetLabs/features/downstream-cookies
Downstream DNS Server Cookies a la RFC7873 and RFC9018
2023-08-17 15:14:11 +02:00
W.C.A. Wijngaards 1c85901cc4 - Fix out of bounds read in parse_edns_options_from_query, it would read
8 bytes after a client option of length 8, and then ignore them to
  recreate a 24 byte response. The fixup does not read out of bounds,
  and puts zeroes in the buffer at that point, that then are ignored.
2023-08-16 16:58:49 +02:00
W.C.A. Wijngaards b1c707e551 - Fix possibly unaligned memory access. 2023-08-16 14:57:38 +02:00
W.C.A. Wijngaards 2b1028bdad - Fix possibly unaligned memory access. 2023-08-16 10:06:06 +02:00
George Thessalonikefs bab5ad623c - For #762: Introduce stat counters for downstream DNS Cookies per
thread and total: num.queries_cookie_valid, num.queries_cookie_client,
  num.queries.cookie_invalid.
2023-08-08 15:19:56 +02:00
George Thessalonikefs 49e4258102 - For #762: Interaction between DNS Cookies and source IP ratelimiting
by allowing Cookies to bypass the ratelimit, but still allowing
  ratelimit to valid DNS Cookie clients via the new
  ip-ratelimit-cookie option.
2023-08-08 10:14:03 +02:00
George Thessalonikefs 81e219827e - For #762: Silence maybe-uninitialized compiler warning. 2023-08-07 11:20:48 +02:00
George Thessalonikefs 02ac374640 - For #762: Remove re-introduced files from merge (configlexer.c, configparser.c, configparser.h). 2023-08-07 11:12:49 +02:00
George Thessalonikefs 025d810b45 - For #762: annotate case statement fallthrough for gcc. 2023-08-07 11:04:23 +02:00
George Thessalonikefs 4ccb613396 Merge branch 'master' into features/downstream-cookies 2023-08-05 20:37:48 +02:00
George Thessalonikefs 9025be814a - For #762: remove uneeded include. 2023-08-05 20:18:56 +02:00
George Thessalonikefs fbc0256825 - For #762: Cleaner manpage text and uniform use of the term DNS
Cookies.
2023-08-05 20:00:37 +02:00
George Thessalonikefs 8580a74b37 - For #762: Introduce rpl testing for DNS Cookies. 2023-08-05 19:50:57 +02:00
George Thessalonikefs b6e2f4dbf8 - For #762: Formatting. 2023-08-04 19:03:23 +02:00
George Thessalonikefs 702f485587 - For #762: relocate EDNS cookie code to util/edns and introduce unit
tests.
2023-08-04 14:26:08 +02:00
George Thessalonikefs 6e47c1e05b - For #762: remove relocated code. 2023-08-02 15:51:05 +02:00
George Thessalonikefs d4145772b5 - Move a cache reply callback in worker.c closer to the cache reply
generation.
2023-08-02 12:33:52 +02:00
George Thessalonikefs 00a08beee9 - For #762: Ignore util/siphash.c for the 01-doc test. 2023-08-01 17:40:49 +02:00
George Thessalonikefs 5b55a46550 - For #762: relocate RFC 1982 serial number arithmetic functions to their own
file in util/rfc_1982.[ch].
2023-08-01 17:26:14 +02:00
George Thessalonikefs 2cc9563cf8 - Fix to use the now cached EDE, if any, for CD_bit queries. 2023-08-01 15:23:25 +02:00
George Thessalonikefs 8aec671860 - More braces and formatting for Fix for EDNS EDE size calculation to
avoid future bugs.
2023-08-01 15:15:33 +02:00
W.C.A. Wijngaards 990b12bc8e - Fix for EDNS EDE size calculation. 2023-08-01 15:08:50 +02:00
George Thessalonikefs 11f12bc981 - Merge #911 from natalie-reece: Exclude EDE before other EDNS options when
there isn't enough space.
- For #911: Try to trim EXTRA-TEXT (and LDNS_EDE_OTHER options
  altogether) before giving up on attaching EDE options.
2023-08-01 10:01:18 +02:00
George Thessalonikefs 08e11284fb - For #911: Try to trim EXTRA-TEXT (and LDNS_EDE_OTHER options
altogether) before giving up on attaching EDE options.
2023-08-01 09:55:28 +02:00
W.C.A. Wijngaards f531011e85 - iana portlist update. 2023-07-31 10:24:43 +02:00
George Thessalonikefs d18813be30 - Merge #790 from Tom Carpay: Add support for EDE caching in cachedb
and subnetcache.
2023-07-31 10:18:10 +02:00
5f76e201f0 - For #790: Update formatting in cachedb/cachedb.c
Co-authored-by: Wouter Wijngaards <wcawijngaards@users.noreply.github.com>
2023-07-31 10:13:01 +02:00
George Thessalonikefs dcd0191d7e Merge branch 'master' of https://github.com/natalie-reece/unbound into natalie-reece-master 2023-07-31 09:57:21 +02:00
George Thessalonikefs 8936f3496e - For #790: clean testcase. 2023-07-31 09:41:57 +02:00
George Thessalonikefs f97927a47e Merge branch 'master' into features/ede-caching-cachedb 2023-07-30 14:17:52 +02:00
George Thessalonikefs 6819c1e444 - Merge #759 from Tom Carpay: Add EDE (RFC8914) caching. 2023-07-30 11:48:04 +02:00
George Thessalonikefs 6487d6febe - For #759: fix doc string. 2023-07-30 11:43:06 +02:00
George Thessalonikefs 373904f865 - Fix unused variable compile warning for kernel timestamps in
netevent.c
2023-07-28 20:17:07 +02:00
George Thessalonikefs 0912015fb9 - Review for #790: Fix memory leak. 2023-07-28 17:21:15 +02:00
George Thessalonikefs c15cfb4bd9 - Review for #790: Address Wouter's comments. 2023-07-28 16:55:51 +02:00
George Thessalonikefs 66c95e8081 - Review for #790: replace tdir tests with simpler rpl tests. 2023-07-28 16:54:05 +02:00
George Thessalonikefs 843fc69927 Address review comments for #759:
- Clear error text when an expected signature is missing.
2023-07-28 14:05:25 +02:00
George Thessalonikefs 50ea4a1072 Address review comments for #759:
- Decrease allocations for "" EDE strings when loading the cachedump.
- Check for existence of EDE code before attaching.
2023-07-28 12:56:13 +02:00
George Thessalonikefs 1e47eea6e3 - Merge #889 from borisVanhoof: Free memory in error case + remove
unused function.
2023-07-21 21:13:43 +02:00
George Thessalonikefs 6289238cd6 - For #889: Account for num_detached_states before possible
mesh_state_delete when erroring out.
2023-07-21 21:05:38 +02:00
George Thessalonikefs 97fdd0e2eb - For #889: use netcat-openbsd instead of netcat-traditional. 2023-07-21 21:04:40 +02:00
George Thessalonikefs 201da1f50a Merge branch 'free_memory_in_error_case' of https://github.com/borisVanhoof/unbound into borisVanhoof-free_memory_in_error_case 2023-07-21 17:04:33 +02:00
George Thessalonikefs 51c189394d - Cleaner failure code for callback functions in interface.i. 2023-07-21 16:53:36 +02:00
George Thessalonikefs 2e257fff2a - Merge #390 from Frank Riley: Add missing callbacks to the python
module.
2023-07-21 16:50:57 +02:00
George Thessalonikefs 4776f749be Merge branch 'python_add_missing_callbacks' of https://github.com/fhriley/unbound into fhriley-python_add_missing_callbacks 2023-07-21 15:19:12 +02:00
George Thessalonikefs 04053d39a0 - Merge #118 from mibere: Changed verbosity level for Redis init &
deinit.
2023-07-21 15:01:48 +02:00
mibereandGeorge Thessalonikefs ef9f7f113f Log established connection to Redis 2023-07-21 14:41:26 +02:00
mibereandGeorge Thessalonikefs 2d33bba3c0 Changed verbosity level for Redis init & deinit
Redis init & deinit are basic (operational) information
2023-07-21 14:39:34 +02:00
George Thessalonikefs 36b38cea74 - Merge #857 from eaglegai: fix potential memory leaks when errors
happen.
2023-07-21 14:04:38 +02:00
George Thessalonikefs 40f446a499 - For #857: fix mixed declarations and code. 2023-07-21 14:02:01 +02:00
George Thessalonikefs e839771ddf Merge branch 'master' of https://github.com/eaglegai/unbound into eaglegai-master 2023-07-21 12:43:47 +02:00
George Thessalonikefs ae2c2be1a5 Merge branch 'master' of github.com:NLnetLabs/unbound 2023-07-20 12:56:31 +02:00
George Thessalonikefs bf37487dca - Merge #901 from Sergei Trofimovich: config: improve handling of
unknown modules.
2023-07-20 12:56:20 +02:00
W.C.A. Wijngaards 8d45c1592b - For #909: Fix RR class comparison. 2023-07-20 12:16:24 +02:00
George Thessalonikefs 862fa0d514 Merge branch 'module-error' of https://github.com/trofi/unbound into trofi-module-error 2023-07-20 12:08:27 +02:00
George Thessalonikefs 27c028cf27 - Merge #909 from headshog: Numeric truncation when parsing TYPEXX and
CLASSXX representation.
- For #909: Fix return values.
2023-07-20 11:57:17 +02:00
George Thessalonikefs 5b7faca7db For #909: Numeric truncation when parsing TYPEXX and CLASSXX representation
- Fix return values.
- Formatting nits.
2023-07-20 11:42:05 +02:00
George Thessalonikefs d29bc712b1 Merge branch 'fix-numtrunc' of https://github.com/headshog/unbound into headshog-fix-numtrunc 2023-07-20 11:07:19 +02:00
headshog 0b131d5a31 parse sldns_get_rr_class_by_name and sldns_get_rr_type_by_name return value 0 2023-07-19 18:09:03 +03:00
George Thessalonikefs 90b434c260 - For #759: add support for cached EDEs to cachedump 2023-07-19 17:06:10 +02:00
George Thessalonikefs 846b158304 - Remove redundant checks when attaching EDE to a SERVFAIL answer. 2023-07-19 15:26:08 +02:00
George Thessalonikefs 95604a90e8 Review for #759:
- Keep EDE information for keys close to key creation.
- Fix inconsistencies between reply and cached EDEs.
- Incorporate EDE caching checks in EDE tests.
- Fix some EDE cases where missing DNSKEY was wrongly reported.
2023-07-19 15:20:44 +02:00
George Thessalonikefs f5a2a58ce3 Review for #759:
- Fix SEGFAULT in load_cache control command.
- Change reason_bogus_str to an explicit NULL-terminated string.
- Fix potential memory leak when discarding a message for referrals and
  0 TTL answers.
- Fix reason_bogus initialization in localzone answers.
- reply_info creation in validator is always regional.
2023-07-17 17:26:31 +02:00
headshog 78c284e05d fix numtrunc in str2wire.c 2023-07-17 12:58:52 +03:00
George Thessalonikefs 299f55b0d1 - More clear description of the different auth-zone behaviors on the
man page.
2023-07-14 15:28:42 +02:00
George Thessalonikefs 52f0387cac - Merge #880 from chipitsine: services/authzone.c: remove redundant
check.
2023-07-13 11:52:14 +02:00
George Thessalonikefs 606e5a0a5f Merge branch 'cleanup_cppcheck' of https://github.com/chipitsine/unbound into chipitsine-cleanup_cppcheck 2023-07-13 11:49:41 +02:00
George Thessalonikefs 15b8d8b96a Merge branch 'master' into features/ede-caching 2023-07-13 11:25:59 +02:00
Natalie Reece 67e52ea9c5 Exclude EDE before other EDNS options when there isn't enough space 2023-07-11 17:01:26 -06:00
George Thessalonikefs 7240ecbeb0 - Merge #664 from tilan7763: Add prefetch support for subnet cache
entries.
- For #664: Easier code flow for subnetcache prefetching.
- For #664: Add testcase.
- For #664: Rename subnet_prefetch tests to subnet_global_prefetch to
  differentiate from the new subnet prefetch support.
2023-07-11 14:31:49 +02:00
George Thessalonikefs a952ac17be Merge branch 'tilan7663-subnet_cache_prefetch' into subnet_cache_prefetch 2023-07-07 16:50:58 +02:00
George Thessalonikefs 40e47bf767 - For #664: easier code flow for subnetcache prefetching.
- For #664: add testcase.
2023-07-06 22:22:21 +02:00
George Thessalonikefs 0afe58a06e - Skip the 00-lint test. splint is not maintained; it either does not
work or produces false positives. Static analysis is handled in the
  clang test.
2023-07-03 15:38:16 +02:00
George Thessalonikefs 2069271384 - Merge #802: add validation EDEs to queries where the CD bit is set.
- For #802: Cleanup comments and add RCODE check for CD bit test case.
2023-07-03 14:48:39 +02:00
George Thessalonikefs 014db3fb03 - For #802: Cleanup comments and add RCODE check for CD bit test case. 2023-07-03 14:40:01 +02:00
W.C.A. Wijngaards 5aa47fb1fa - Fix dereference of NULL variable warning in mesh_do_callback. 2023-07-03 13:50:39 +02:00
George Thessalonikefs 1962991937 - Merge #739: Add SVCB dohpath support.
- Code cleanup for sldns_str2wire_svcparam_key_lookup.
2023-07-03 11:02:05 +02:00
George Thessalonikefs 5be7f1ef8a - Code cleanup for sldns_str2wire_svcparam_key_lookup. 2023-07-03 10:51:34 +02:00
W.C.A. Wijngaards 48a6ff14a4 - Fix #906: warning: ‘Py_SetProgramName’ is deprecated. 2023-07-03 10:23:37 +02:00
George Thessalonikefs ade710a9fd - For #739: minor cleanup for testcases. 2023-07-03 10:10:16 +02:00
George Thessalonikefs 31d38d2cfe Merge branch 'master' into features/dohpath-for-unbound 2023-06-30 11:50:23 +02:00
George Thessalonikefs 380e3de140 - Merge #827 from rcmcdonald91: Eliminate unnecessary Python reloading
which causes memory leaks.
2023-06-29 13:02:27 +02:00
George Thessalonikefs fc8bf269e9 - More fixes for reference counting for python module and clean up
failure code.
2023-06-29 12:26:49 +02:00
W.C.A. Wijngaards 7696074fa9 - Fix python modules with multiple scripts, by incrementing reference
counts.
2023-06-29 10:16:37 +02:00
George Thessalonikefs bea61fc37c - Remove warning about unknown cast-function-type warning pragma. 2023-06-27 16:44:29 +02:00
George Thessalonikefs 41dac805f5 - Merge #892: Add cachedb hit stat. Introduces 'num.query.cachedb' as
a new statistical counter.
2023-06-27 12:46:26 +02:00
W.C.A. Wijngaards 2207a55107 Add changelog and contrib/README mention for #903
- Merge #903: contrib: add yocto compatible init script.
2023-06-22 15:41:17 +02:00
Wouter WijngaardsandGitHub 196f8a68f6 Merge pull request #903 from beni-sandu/master
contrib: add yocto compatible init script
2023-06-22 15:37:07 +02:00
Beniamin Sandu 0d13b4ec4c contrib: add yocto compatible init script
Signed-off-by: Beniamin Sandu <beniaminsandu@gmail.com>
2023-06-22 14:18:45 +03:00
George Thessalonikefs f1537e2041 - For #762: please doxygen. 2023-06-22 12:21:27 +02:00
George Thessalonikefs 47cf44cc70 - For #762: relocate edns_opt_list_append_keepalive. 2023-06-22 12:11:28 +02:00
George Thessalonikefs 1cd75cccfc - For #762: More generic integration for siphash.c 2023-06-22 11:45:08 +02:00
Sergei Trofimovich d10a889a68 config: improve handling of unknown modules
The change fixes module print when specified module is unknown. On
example config:

    server:
      module-config: "respip valdator iterator"

Before the change printed error looked like:

    error: Unknown value in module-config, module: ''. This module is
    not present (not compiled in), See the list of linked modules with
    unbound -V

After the change module is printed as expected:

    error: Unknown value in module-config, module: 'valdator'. This
    module is not present (not compiled in), See the list of linked
    modules with unbound -V

Module truncation happens because parse error does not guarantee that
leading whitespace is removed by `module_factory()` call.

The change always removes leading whitespace (if present).
2023-06-19 18:20:22 +01:00
Philip Homburg 64476280ed Changelog for #887 and #516 2023-06-15 11:09:08 +02:00
Philip Homburg b1bcde3137 Merge branch 'freebsd-aslr-issue' 2023-06-15 11:06:29 +02:00
George Thessalonikefs b02f9befcd - For #762: fix compiler C90 warning. 2023-06-14 16:41:01 +02:00
George Thessalonikefs 0f1ea7e490 - Properly handle all return values of worker_check_request during
early EDE code.
- Do not check the incoming request more than once.
2023-06-14 11:40:59 +02:00
W.C.A. Wijngaards f9317d65b3 - Fix for uncertain unit test for doh buffer size events. 2023-06-12 12:39:00 +02:00
W.C.A. Wijngaards 2cf0359ffe Changelog note for #895
- Fix #895: python + sysconfig gives ANOTHER path comparing to
  distutils.
2023-06-12 10:39:44 +02:00
W.C.A. Wijngaards 35885e5a70 - Merge #896: Fix: #895: pythonmodule: add all site-packages
directories to sys.path.
2023-06-12 10:30:50 +02:00
Wouter WijngaardsandGitHub 0585b1d75a Merge pull request #896 from socketpair/bugfix
Fix: #895: pythonmodule: add all site-packages directories to sys.path
2023-06-12 10:30:05 +02:00
Коренберг Марк a25fc52818 Fix: #895: pythonmodule: add all site-packages directories to sys.path 2023-06-11 10:39:13 +03:00
Philip Homburg 52581f8644 Fix for issue #887 (Timeouts to forward servers on BSD based system with ASLR)
and proabbly #516 (Stream reuse does not work on Windows)
2023-06-09 13:59:31 +02:00
9412b9c2ca Review comment for daemon/stats.c
Co-authored-by: Wouter Wijngaards <wcawijngaards@users.noreply.github.com>
2023-06-02 12:39:23 +02:00
65230bd117 Review comment for testdata/stat_values.tdir/stat_values.test
Co-authored-by: Wouter Wijngaards <wcawijngaards@users.noreply.github.com>
2023-05-31 17:09:16 +02:00
George Thessalonikefs db5cf5851d - More efficient mesh accounting per client. 2023-05-30 23:34:31 +02:00
George Thessalonikefs 3c3fd7a795 - More predictable testing for cachedb. 2023-05-30 23:33:48 +02:00
George Thessalonikefs 4f52be4db9 - Introduce num.query.cachedb to track cache hits for the external cache. 2023-05-30 17:49:50 +02:00
W.C.A. Wijngaards 512236d705 - Fix unbound-dnstap-socket time fraction conversion for printout. 2023-05-25 16:27:19 +02:00
W.C.A. Wijngaards d57986724e - Fix unbound-dnstap-socket printout when no query is present. 2023-05-25 14:30:25 +02:00
Boris VANHOOF 4e39d3dfe1 could not find package netcat for docker container 2023-05-23 10:11:07 +02:00
Boris VANHOOF a21bc23139 free memory in error case 2023-05-23 09:23:03 +02:00
Boris VANHOOF 62d54d8091 remove unused function 2023-05-23 09:22:35 +02:00
Boris VANHOOF 17559c737b typo in comments 2023-05-23 09:21:58 +02:00
W.C.A. Wijngaards 59fd48c226 - Fix to remove unused variables from RPZ clientip data structure. 2023-05-19 16:36:31 +02:00
W.C.A. Wijngaards da78c42f88 - Fix RPZ removal of client-ip, nsip, nsdname triggers from IXFR. 2023-05-19 14:38:41 +02:00
W.C.A. Wijngaards a07ccbae9a - Fix to print debug log for ancillary data with correct IP address. 2023-05-16 09:21:21 +02:00
W.C.A. Wijngaards 2a2598dbf2 - Fix #888: [FR] Use kernel timestamps for dnstap. 2023-05-16 08:50:38 +02:00
W.C.A. Wijngaards 1fb78afc29 - Fix warning in windows compile, in set_recvtimestamp. 2023-05-11 09:32:59 +02:00
W.C.A. Wijngaards b2cba7b707 - Fix doxygen in addr_to_nat64 header definition. 2023-05-04 15:53:05 +02:00
W.C.A. Wijngaards 8dd09e31d2 - Fix to remove unused whitespace from acx_nlnetlabs.m4 and config.h. 2023-05-04 11:17:06 +02:00
W.C.A. Wijngaards 806c3d7330 - Fix #885: Error: util/configlexer.c: No such file or directory,
adds error messages explaining to install flex and bison.
2023-05-04 11:12:11 +02:00
W.C.A. Wijngaards cac1d13fda - Fix proxy-protocol to read header in multiple reads and check buffer size. 2023-05-02 14:54:51 +02:00
W.C.A. Wijngaards 80153decd1 - Fix proxy-protocol buffer checks when writing and read from buffer. 2023-05-02 14:36:29 +02:00
George Thessalonikefs 2695eb9d1a Changelog entry for #722:
- Merge #722 from David 'eqvinox' Lamparter: NAT64 support.
- For #722: minor fixes, formatting, refactoring.
2023-05-01 18:32:40 +02:00
George Thessalonikefs 20184483df Merge branch 'eqvinox-nat64' 2023-05-01 18:24:05 +02:00
George Thessalonikefs adb4aeb609 - For #722: Minor fixes, formatting and refactoring. 2023-05-01 18:23:13 +02:00
W.C.A. Wijngaards 70c2b587fc - Fix RPZ IP responses with trigger rpz-drop on cache entries, that
they are dropped.
2023-05-01 09:26:17 +02:00
Philip Homburg a50ddd7ab9 Changelog for #860 2023-04-26 17:15:59 +02:00
Philip Homburg 4a04ba813a Merge branch '0ttl' 2023-04-26 17:14:15 +02:00
Philip Homburg 1aa2c318e7 Remove msg_del_for_0ttl, call msg_cache_remove directly 2023-04-26 17:11:29 +02:00
George Thessalonikefs e1ec3cf893 Merge branch 'nat64' of https://github.com/eqvinox/unbound into eqvinox-nat64 2023-04-26 15:14:39 +02:00
W.C.A. Wijngaards 8058dc9127 - Fix for #882: document variable to stop doxygen warning. 2023-04-26 14:07:33 +02:00
W.C.A. Wijngaards 144f29638c - Fix for #882: small changes, date updated in Copyright for
util/timeval_func.c and util/timeval_func.h. Man page entries and
  example entry.
2023-04-26 13:49:33 +02:00
Wouter WijngaardsandGitHub 7081b0340f Merge pull request #882 from vvfedorenko/features/dropqueuedpackets
Features/dropqueuedpackets
2023-04-26 13:25:32 +02:00
Vadim Fedorenko 263096d1f6 stats: add query max wait time metric
Add new statistic value to know the size of the queue in microseconds.

Signed-off-by: Vadim Fedorenko <vadfed@meta.com>
2023-04-26 03:27:25 -07:00
Vadim Fedorenko e577ab105e stats: add counter for timed out queries
Add counter `num_queries_timed_out` meaning queries that were sitting in the
socket queue and waiting to being processed too long. There is no reason
to process such queries, so let's drop it in the very beginning of the
pipeline.

Signed-off-by: Vadim Fedorenko <vadfed@meta.com>
2023-04-26 03:27:25 -07:00
Vadim Fedorenko 04540f82e5 config: add sock_queue_timeout configuration
Add sock_queue_timeout config option to have queue timeout configurable.

Signed-off-by: Vadim Fedorenko <vadfed@meta.com>
2023-04-26 03:27:19 -07:00
Vadim Fedorenko 2e6ddd6032 netevent: parse and store rcv timestamp from sock
Add special field in comm_point to store the software receive timestamp
for every particular UDP packet. Aux data parser is updated to read
values and the whole callback is switched to use recvmsg form.

Signed-off-by: Vadim Fedorenko <vadfed@meta.com>
2023-04-26 03:26:51 -07:00
Vadim Fedorenko a197aac2f6 timeval_func: move all timeval manipulation to separate file
There are several definitions of the same functions manipulating timeval
structures. Let's move them to separate file and arrange the code
preperly.

Signed-off-by: Vadim Fedorenko <vadfed@meta.com>
2023-04-26 03:23:41 -07:00
W.C.A. Wijngaards 15a2add0f8 streamtcp, implement NOTIFY[=N] that sends a notify packet. 2023-04-26 11:57:10 +02:00
Vadim Fedorenko 648ad4db6f Linting change.
Remove config parser/lexer code as it's rebuilded every time but can
break adding new config options.
Also clean up the code base to avoid mixing actual code changes and lint
issues.

Signed-off-by: Vadim Fedorenko <vadfed@meta.com>
2023-04-25 17:05:00 -07:00
W.C.A. Wijngaards 4bcc0a0a7a streamtcp, implement IXFR=N queries, add documentation for proxy option. 2023-04-25 16:44:58 +02:00
W.C.A. Wijngaards 542f717bf9 - adjust generic proxy-protocol header for IPv6 support with ifdef. 2023-04-25 08:16:19 +02:00
George Thessalonikefs b5cc8b6c59 - Generalise the proxy protocol code 2023-04-24 16:15:56 +02:00
Ilya ShipitsinandGitHub cba1350886 services/authzone.c: remove redundant check
found by cppcheck

services\authzone.c:7513:12: style: Condition 'rrlist[i]' is always true [knownConditionTrueFalse]
2023-04-24 11:45:41 +02:00
W.C.A. Wijngaards fe46bc47d7 - Fix for #878: Invalid IP address in unbound.conf causes Segmentation
Fault on OpenBSD.
2023-04-19 09:56:31 +02:00
W.C.A. Wijngaards e11d206a82 Changelog entry for #875 and #874.
- Merge #875: change obsolete txt URL in unbound-anchor.c to point
  to RFC 7958, and Fix #874.
2023-04-14 11:19:25 +02:00
Wouter WijngaardsandGitHub 8de8c8f6f7 Merge pull request #875 from dyunwei/master
Changelog entry for #874
2023-04-14 11:17:43 +02:00
yunweiandGitHub 4954df5859 Changelog entry for #874
unbound-anchor is compliant with RFC 7958, and the XML format remains unchanged between the old draft and RFC 7958.  Update the comments to improve clarity.
2023-04-14 16:45:11 +08:00
W.C.A. Wijngaards a3ef9dd53b - Show build status for branch=master. 2023-04-13 11:29:53 +02:00
W.C.A. Wijngaards d6c33e1757 - Fix build badge, from failing travis link to github ci action link. 2023-04-13 11:22:11 +02:00
W.C.A. Wijngaards 7033234a48 - Fix for #870: Add test case for the qname minimisation and CNAME. 2023-04-06 10:04:04 +02:00
W.C.A. Wijngaards c7618a9b80 - Fix #870: NXDOMAIN instead of NOERROR rcode when asked for existing
CNAME record.
2023-04-04 10:06:16 +02:00
Philip Homburg 312035f58a Changelog for issue #676 2023-03-24 14:54:14 +01:00
Philip Homburg 7e6a7f310d Fix issue #676: Unencrypted query is sent when forward-tls-upstream: yes is
used without tls-cert-bundle

Model the behavior of unbound in unbound-host: always create a SSL context
2023-03-24 14:51:37 +01:00
Philip Homburg 1ac9b7548b Small fixes from Wouter's review 2023-03-23 15:15:54 +01:00
Philip Homburg eb7eff4fc7 Extra consistency check to make sure that when TLS is requested, either we
set up a TLS connection or we return an error.
2023-03-23 15:00:10 +01:00
Philip Homburg 072be3300f Tests for serve-expired in combination with new 0 TTL data. 2023-03-22 15:23:47 +01:00
Philip Homburg 9d7b1d3127 Fix issue #860: Bad interaction with 0 TTL records and serve-expired 2023-03-22 15:23:47 +01:00
Philip Homburg e850ca67f1 Changelog for issue #851 2023-03-21 13:54:27 +01:00
Philip Homburg 2a100ee9ee Fix issue #851: reserved identifier violation 2023-03-21 13:51:51 +01:00
W.C.A. Wijngaards 8f83c0a2cb - iana portlist update. 2023-03-20 14:55:55 +01:00
George Thessalonikefs d7e7761141 - Fix #812, fix #846, by using the SSL_OP_IGNORE_UNEXPECTED_EOF option
to ignore the unexpected eof while reading in openssl >= 3.
2023-03-17 14:39:37 +01:00
W.C.A. Wijngaards a97d7175a6 - Fix ssl.h include brackets, instead of quotes. 2023-03-16 15:40:43 +01:00
W.C.A. Wijngaards 4f25d75d4b - Fix unbound-dnstap-socket test program to reply the finish frame
over a TLS connection correctly.
2023-03-14 16:57:37 +01:00
eaglegai 184248eb0e fix memory leak in unbound-streamtcp when open_svr failed
==1927474== Memcheck, a memory error detector
==1927474== Copyright (C) 2002-2017, and GNU GPL'd, by Julian Seward et al.
==1927474== Using Valgrind-3.16.0 and LibVEX; rerun with -h for copyright info
==1927474== Command: unbound-streamtcp -f localhost
==1927474==
fatal: bad server specs 'localhost'
==1927474==
==1927474== HEAP SUMMARY:
==1927474==     in use at exit: 131,186 bytes in 4 blocks
==1927474==   total heap usage: 5 allocs, 1 frees, 132,210 bytes allocated
==1927474==
==1927474== 40 bytes in 1 blocks are still reachable in loss record 1 of 4
==1927474==    at 0x483F751: malloc (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so)
==1927474==    by 0x1E0573: sldns_buffer_new (sbuffer.c:21)
==1927474==    by 0x11ECED: send_em (streamtcp.c:374)
==1927474==    by 0x11E6C1: main (streamtcp.c:585)
==1927474==
==1927474== 40 bytes in 1 blocks are still reachable in loss record 2 of 4
==1927474==    at 0x483F751: malloc (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so)
==1927474==    by 0x1E0573: sldns_buffer_new (sbuffer.c:21)
==1927474==    by 0x11ECFA: send_em (streamtcp.c:375)
==1927474==    by 0x11E6C1: main (streamtcp.c:585)
==1927474==
==1927474== 65,553 bytes in 1 blocks are still reachable in loss record 3 of 4
==1927474==    at 0x483F751: malloc (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so)
==1927474==    by 0x1E0583: sldns_buffer_new (sbuffer.c:27)
==1927474==    by 0x11ECED: send_em (streamtcp.c:374)
==1927474==    by 0x11E6C1: main (streamtcp.c:585)
==1927474==
==1927474== 65,553 bytes in 1 blocks are still reachable in loss record 4 of 4
==1927474==    at 0x483F751: malloc (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so)
==1927474==    by 0x1E0583: sldns_buffer_new (sbuffer.c:27)
==1927474==    by 0x11ECFA: send_em (streamtcp.c:375)
==1927474==    by 0x11E6C1: main (streamtcp.c:585)
==1927474==
==1927474== LEAK SUMMARY:
==1927474==    definitely lost: 0 bytes in 0 blocks
==1927474==    indirectly lost: 0 bytes in 0 blocks
==1927474==      possibly lost: 0 bytes in 0 blocks
==1927474==    still reachable: 131,186 bytes in 4 blocks
==1927474==         suppressed: 0 bytes in 0 blocks
==1927474==
==1927474== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 0 from 0)

Signed-off-by: eaglegai <eaglegai@163.com>
2023-03-07 21:49:54 +08:00
eaglegai 319119943f fix potential memory leak in unbound-host when errors happen
==3709953== HEAP SUMMARY:
==3709953==     in use at exit: 276,541 bytes in 23 blocks
==3709953==   total heap usage: 29 allocs, 6 frees, 280,682 bytes allocated
==3709953==
==3709953== 1 bytes in 1 blocks are still reachable in loss record 1 of 23
==3709953==    at 0x4866EC0: malloc (in /usr/lib64/valgrind/vgpreload_memcheck-arm64-linux.so)
==3709953==    by 0x48E2BC3: ub_initstate (random.c:85)
==3709953==    by 0x489B067: ub_ctx_create_nopipe (libunbound.c:114)
==3709953==    by 0x489B31F: ub_ctx_create (libunbound.c:180)
==3709953==    by 0x10E203: main (unbound-host.c:433)
==3709953==
......
==3709953== 8,192 bytes in 1 blocks are still reachable in loss record 22 of 23
==3709953==    at 0x4866EC0: malloc (in /usr/lib64/valgrind/vgpreload_memcheck-arm64-linux.so)
==3709953==    by 0x48E427B: regional_create_custom (regional.c:94)
==3709953==    by 0x48DEA03: edns_strings_create (edns.c:57)
==3709953==    by 0x489B0F3: ub_ctx_create_nopipe (libunbound.c:157)
==3709953==    by 0x489B31F: ub_ctx_create (libunbound.c:180)
==3709953==    by 0x10E203: main (unbound-host.c:433)
==3709953==
==3709953== 262,144 bytes in 1 blocks are still reachable in loss record 23 of 23
==3709953==    at 0x486933C: calloc (in /usr/lib64/valgrind/vgpreload_memcheck-arm64-linux.so)
==3709953==    by 0x48C826F: config_create (config_file.c:179)
==3709953==    by 0x48C85AF: config_create_forlib (config_file.c:383)
==3709953==    by 0x489B0BB: ub_ctx_create_nopipe (libunbound.c:130)
==3709953==    by 0x489B31F: ub_ctx_create (libunbound.c:180)
==3709953==    by 0x10E203: main (unbound-host.c:433)
==3709953==
==3709953== LEAK SUMMARY:
==3709953==    definitely lost: 0 bytes in 0 blocks
==3709953==    indirectly lost: 0 bytes in 0 blocks
==3709953==      possibly lost: 0 bytes in 0 blocks
==3709953==    still reachable: 276,541 bytes in 23 blocks
==3709953==         suppressed: 0 bytes in 0 blocks
==3709953==
==3709953== For lists of detected and suppressed errors, rerun with: -s
==3709953== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 0 from 0)

Signed-off-by: eaglegai <eaglegai@163.com>
2023-03-06 22:04:06 +08:00
Christian McDonald 60304f972e #827 review response 2023-02-24 09:31:19 -05:00
W.C.A. Wijngaards d97c174f50 - Fix for #852: Completion of error handling. 2023-02-23 13:38:29 +01:00
Philip Homburg ed07c5424d Changelog entry for issue #825 2023-02-21 09:29:39 +01:00
Philip Homburg fb187d1a60 Merge branch 'master' into ecs-serve-expired-bug 2023-02-21 09:28:13 +01:00
Philip Homburg 71e0ddc94a Improved comment 2023-02-21 09:27:03 +01:00
Philip Homburg b89d0d1cce Test cache update from serve-expired and client-subnet-always-forward 2023-02-21 09:21:43 +01:00
Philip Homburg d1f5ded1d9 ifdef CLIENT_SUBNET 2023-02-21 09:21:24 +01:00
Philip Homburg fb06364014 Fix issue #825: interaction between ECS and serve-expired. 2023-02-21 09:20:28 +01:00
George Thessalonikefs 02a77f0567 Changelog entry for
- Clean up iterator/iterator.c::error_response_cache() and allow for
  better interaction with serve-expired, prefetch and cached error
  responses.
2023-02-10 16:54:44 +01:00
George Thessalonikefs eb81761b13 - Clean up iterator/iterator.c::error_response_cache() and allow for
better interaction with serve-expired, prefetch and cached error
  responses.
2023-02-10 16:51:07 +01:00
George Thessalonikefs 96c70d91ca - Add testcase for refreshing expired error responses. 2023-02-09 12:44:01 +01:00
W.C.A. Wijngaards 488811157e - Fix to git ignore the library symbol file that configure can create. 2023-02-09 12:08:27 +01:00
W.C.A. Wijngaards e225e4bcab - Fix consistency of unit test without roundrobin answers for the
cnametooptout unit test.
2023-02-09 11:46:33 +01:00
George Thessalonikefs 1c1c5d72d3 Changelog entry for
- Allow TTL refresh of expired error responses.
2023-02-09 10:52:56 +01:00
George Thessalonikefs 87a8c80fcb - Allow TTL refresh of expired error responses. 2023-02-09 10:47:46 +01:00
W.C.A. Wijngaards a8977df4d9 - Fix unit tests for spurious empty messages. 2023-02-09 10:38:55 +01:00
W.C.A. Wijngaards 4953daa016 - Fix to ignore entirely empty responses, and try at another authority.
This turns completely empty responses, a type of noerror/nodata into
  a servfail, but they do not conform to RFC2308, and the retry can
  fetch improved content.
2023-02-09 09:56:40 +01:00
W.C.A. Wijngaards 5ab5b3b43a - Fix #841: Unbound won't build with aaaa-filter-iterator.patch. 2023-02-08 11:36:52 +01:00
Christian McDonald 0bf55e6063 set modinfo to null last 2023-02-01 15:17:59 -05:00
George Thessalonikefs 24e6d1e18e - Add duration variable for speed_local.test. 2023-01-30 11:33:58 +01:00
R. Christian McDonaldandGitHub 45142868ec Fix copyright date 2023-01-28 14:02:51 -05:00
W.C.A. Wijngaards 8b9382998d Regenerate configure for the fix acx_nlnetlabs.m4 for -Wstrict-prototypes. 2023-01-26 10:59:16 +01:00
W.C.A. Wijngaards c482999898 - Fix acx_nlnetlabs.m4 for -Wstrict-prototypes. 2023-01-26 10:54:38 +01:00
George Thessalonikefs 6bf677e7de Fix #833: [FR] Ability to set the Redis password. 2023-01-23 11:45:07 +01:00
W.C.A. Wijngaards d666e9bd13 - Fix #835: [FR] Ability to use Redis unix sockets. 2023-01-23 10:10:23 +01:00
W.C.A. Wijngaards 77f15428c9 - Add #835: [FR] Ability to use Redis unix sockets. 2023-01-23 10:09:28 +01:00
W.C.A. Wijngaards 111e66ae64 Changelog note for #819, generate configparser.c and comment syntax change.
- Merge #819: Added new static zone type block_a to suppress all A
  queries for specific zones.
2023-01-20 16:19:20 +01:00
Wouter WijngaardsandGitHub 6a4a9435d1 Merge pull request #819 from pavel-odintsov/pavel/suppress_a
Added new static zone type block_a to suppress all A queries for specific zones
2023-01-20 16:18:05 +01:00
R. Christian McDonaldandGitHub 55a28d6946 cleanup callbacks that might have been registered by Python script 2023-01-19 16:30:47 -05:00
R. Christian McDonaldandGitHub ddecffd918 Merge branch 'NLnetLabs:master' into rcm-python-reloading 2023-01-19 16:28:12 -05:00
W.C.A. Wijngaards 6afdc336ba - Fix test for new default. 2023-01-19 16:06:30 +01:00
W.C.A. Wijngaards c9233f8429 - Set default for harden-unknown-additional to no. So that it does
not hamper future protocol developments.
2023-01-19 15:45:10 +01:00
W.C.A. Wijngaards 8df1e58209 - Add harden-unknown-additional option. Default on and it removes
unknown records from the authority section and additional section.
  Thanks to Xiang Li, from NISL Lab, Tsinghua University.
2023-01-19 14:59:18 +01:00
W.C.A. Wijngaards d69f875261 - Set max-udp-size default to 1232. This is the same default value as
the default value for edns-buffer-size. It restricts client edns
  buffer size choices, and makes unbound behave similar to other DNS
  resolvers. The new choice, down from 4096 means it is harder to get
  large responses from Unbound. Thanks to Xiang Li, from NISL Lab,
  Tsinghua University.
2023-01-19 14:16:17 +01:00
W.C.A. Wijngaards b12ab31ae3 - Fix not following cleared RD flags potentially enables amplification
DDoS attacks, reported by Xiang Li and Wei Xu from NISL Lab,
  Tsinghua University. The fix stops query loops, by refusing to send
  RD=0 queries to a forwarder, they still get answered from cache.
2023-01-18 13:18:47 +01:00
Christian McDonald cfd3bcb21e eliminate unnecessary Python reloading which causes memory leaks 2023-01-13 16:39:28 -05:00
W.C.A. Wijngaards 90d42148de - Improve documentation for #826, describe the large collisions amount. 2023-01-13 11:22:47 +01:00
W.C.A. Wijngaards 469133e8df Changelog note and documentation for #826
- Merge #826: Аdd a metric about the maximum number of collisions in
  lrushah.
2023-01-13 11:01:46 +01:00
Wouter WijngaardsandGitHub 785c938106 Merge pull request #826 from sakateka/lruhash_max_collisions
Аdd a metric about the maximum number of collisions in lrushah
2023-01-13 10:57:44 +01:00
Sergey Kacheev 52a4ccee18 add a metric about the maximum number of collisions in lrushah 2023-01-13 13:33:38 +07:00
W.C.A. Wijngaards aa621f1c04 Code repository continues with version 1.17.2. 2023-01-12 10:21:28 +01:00
W.C.A. Wijngaards 0fed35a4b7 - Fix python version detection in configure. 2023-01-09 15:10:00 +01:00
W.C.A. Wijngaards 4517dcd439 - Fix python module install path detection. 2023-01-09 15:03:38 +01:00
W.C.A. Wijngaards c5c4f6d40b Changelog note for 1.17.1rc2 fix.
- Fix wildcard in hyperlocal zone service degradation, reported
  by Sergey Kacheev. This fix is included in 1.17.1rc2.
2023-01-06 13:23:02 +01:00
W.C.A. Wijngaards 0419b06ca4 Merge branch 'branch-1.17.1' 2023-01-06 13:22:06 +01:00
W.C.A. Wijngaards 90831af981 - Fix wildcard in hyperlocal zone service degradation, reported
by Sergey Kacheev.
2023-01-06 13:21:39 +01:00
W.C.A. Wijngaards ba6325f24f - Fix #823: Response change to NODATA for some ANY queries since
1.12, tested on 1.16.1.
2023-01-06 09:16:59 +01:00
W.C.A. Wijngaards 2025946247 Changelog note for tag for 1.17.1rc1.
- Tag for 1.17.1 release.
2023-01-05 11:06:07 +01:00
Alex BandandGitHub e92bd614c0 Add Mastodon link 2023-01-04 16:21:43 +01:00
Alex BandandGitHub f93fdb5e09 Add Mastodon 2023-01-04 15:57:45 +01:00
Pavel Odintsov d5b9a790fe Added new static zone type block_a to suppress all A queries for specific zones 2023-01-03 19:17:51 +00:00
W.C.A. Wijngaards 70260273a4 - Update github workflows to use checkout v3. 2023-01-02 13:30:03 +01:00
W.C.A. Wijngaards 1224cd9d25 - Fix windows compile for libunbound subprocess reap comm point closes. 2023-01-02 13:06:39 +01:00
George Thessalonikefs 0682d4371e - Merge #569 from JINMEI Tatuya: add keep-cache option to
'unbound-control reload' to keep caches.
2022-12-14 16:40:00 +01:00
George Thessalonikefs 7716d26d46 - Use an explicit 'reload_keep_cache' command and introduce test cases
for #569.
2022-12-14 16:33:28 +01:00
George Thessalonikefs 857d6ce3a1 Merge branch 'reuse-caches-on-reload' of https://github.com/jinmeiib/unbound-1 into jinmeiib-reuse-caches-on-reload 2022-12-13 16:42:38 +01:00
George Thessalonikefs af2ef61c49 - Merge #461 from Christian Allred: Add max-query-restarts option. 2022-12-13 15:54:51 +01:00
George Thessalonikefs df411b3f28 - Updates for #461 (Add max-query-restarts option). 2022-12-13 15:29:22 +01:00
George Thessalonikefs 71db243b0d Merge branch 'restart_conf' of https://github.com/cgallred/unbound into cgallred-restart_conf 2022-12-13 14:35:01 +01:00
George Thessalonikefs 67cf625608 Merge branch 'master' of github.com:NLnetLabs/unbound 2022-12-13 13:59:42 +01:00
George Thessalonikefs c61b2121b5 - Expose 'max-sent-count' as a configuration option; the
default value retains Unbound's behavior.
2022-12-13 13:57:07 +01:00
George Thessalonikefs 859d0f2dfe - Expose 'statistics-inhibit-zero' as a configuration option; the
default value retains Unbound's behavior.
2022-12-13 10:47:37 +01:00
W.C.A. Wijngaards 1a2e6aabac - Fix to wrap Makefile scripts directory in quotes for uninstall. 2022-12-13 09:03:52 +01:00
W.C.A. Wijngaards 726aa5b0f5 Changelog note for #808
- Merge #808: Wrap Makefile script's directory variables in quotes.
2022-12-13 08:53:44 +01:00
Wouter WijngaardsandGitHub 5204c14e6d Merge pull request #808 from sneurlax/fix/windows
Wrap Makefile script's directory variables in quotes
2022-12-13 08:53:07 +01:00
sneurlax 2daaebf3aa wrap directory variables in quotes
see https://github.com/NLnetLabs/unbound/issues/807
2022-12-12 19:27:19 -06:00
TCY16 dd3984eae9 add validation EDEs to CD bit queries 2022-12-05 11:41:17 +01:00
W.C.A. Wijngaards 6b8642b662 Fix date. 2022-12-01 13:05:02 +01:00
W.C.A. Wijngaards 5c041c0ba9 - Fix #773: When used with systemd-networkd, unbound does not start
until systemd-networkd-wait-online.service times out.
2022-12-01 13:04:05 +01:00
George Thessalonikefs d7a9def160 - Clear documentation for interactivity between the subnet module and
the serve-expired and prefetch configuration options.
2022-11-30 14:45:36 +01:00
George Thessalonikefs ef8111ece7 Merge branch 'master' of github.com:NLnetLabs/unbound 2022-11-30 14:34:00 +01:00
George Thessalonikefs 90f6cb1158 - Add SVCB and HTTPS to the types removed by 'unbound-control flush'. 2022-11-30 14:33:16 +01:00
W.C.A. Wijngaards effbf99281 - Fix #782: Segmentation fault in stats.c:404. 2022-11-30 10:18:27 +01:00
Philip Homburg 81861aee05 Changelog entry for #720 2022-11-29 16:20:52 +01:00
W.C.A. Wijngaards 6f7da59b77 - Fix for the ignore of tcp events for closed comm points, preserve
the use after free protection features.
2022-11-28 10:04:52 +01:00
TCY16 a96c64d966 add subnetcache test and rename/improve cachedb test 2022-11-24 16:14:37 +01:00
TCY16 79d39e7f4d Add subnetcache support 2022-11-24 16:14:12 +01:00
Philip-NLnetLabsandGitHub b86a97019f Merge pull request #720 from jonathangray/winsock_uaf
fix use after free when WSACreateEvent() fails
2022-11-23 14:08:01 +01:00
George Thessalonikefs 896f7a8306 - Ignore expired error responses. 2022-11-22 17:44:55 +01:00
tcarpayandGitHub a7ac109e4e Remove erroneous filepath in test 2022-11-22 12:11:55 +01:00
TCY16 79108f4c3d add ede cachedb test 2022-11-21 15:26:49 +01:00
TCY16 6dcba49ff1 add cachedb support 2022-11-21 13:23:00 +01:00
TCY16 8b4a8493d0 Merge branch 'master' of github.com:NLnetLabs/unbound into features/ede-caching 2022-11-21 11:34:36 +01:00
TCY16 b65ff768bc remove superfluous variables 2022-11-18 11:30:11 +01:00
W.C.A. Wijngaards f72116883b - Fix #779: [doc] Missing documention in ub_resolve_event() for
callback parameter was_ratelimited.
2022-11-11 11:28:15 +01:00
George Thessalonikefs 4e305e644b - Complementary fix for distutils.sysconfig deprecation in Python 3.10
to commit 62c5039ab9.
2022-11-09 11:41:28 +01:00
W.C.A. Wijngaards 89d9b25090 - iana portlist update. 2022-11-08 15:24:24 +01:00
W.C.A. Wijngaards dda1d9544c - Fix #775: libunbound: subprocess reap causes parent process reap
to hang.
2022-11-08 15:04:05 +01:00
W.C.A. Wijngaards 52a9e6268e - Fix to make sure to not read again after a tcp comm point is closed. 2022-11-08 13:23:44 +01:00
W.C.A. Wijngaards 8367b24bc5 - Fix to ignore tcp events for closed comm points. 2022-11-08 12:02:48 +01:00
Willem Toorop 8df26b132b Merge branch 'master' into devel/merge-master-into-downstream-cookies 2022-11-07 17:09:20 +00:00
David Lamparter 64fb06f892 NAT64 support
This implements #721.  Includes documentation and some very basic tests.
Please refer to doc for further detail.
2022-11-07 11:37:50 +00:00
TCY16 5bf4c505db remove value check 2022-10-28 17:40:18 +02:00
TCY16 5158876bfd Merge branch 'master' of github.com:NLnetLabs/unbound into features/dohpath-for-unbounid 2022-10-28 11:14:59 +02:00
George Thessalonikefs f531faf163 Changelog entry for #767
- Merge #767 from jonathangray: consistently use IPv4/IPv6 in
    unbound.conf.5.
2022-10-21 15:49:56 +02:00
Yorgos ThessalonikefsandGitHub c0c9acccfd Merge pull request #767 from jonathangray/man
consistently use IPv4/IPv6 in unbound.conf.5
2022-10-21 15:46:45 +02:00
W.C.A. Wijngaards 17e5dd6131 - Fix that cachedb does not store failures in the external cache. 2022-10-21 10:11:47 +02:00
George Thessalonikefs e9107907e5 - Clarify the use of MAX_SENT_COUNT in the iterator code. 2022-10-18 12:29:07 +02:00
W.C.A. Wijngaards ba8642aeb7 - testcode/dohclient sets log identity to its name. 2022-10-17 16:00:43 +02:00
W.C.A. Wijngaards 5ffa4d7232 - In unit test, print python script name list correctly. 2022-10-14 16:49:57 +02:00
W.C.A. Wijngaards 2571d00535 Changelog note for #768
- Merge #768 from fobser: Arithmetic on a pointer to void is a GNU
  extension.
2022-10-14 16:22:17 +02:00
Wouter WijngaardsandGitHub f9a563d495 Merge pull request #768 from fobser/void_math
Arithmetic on a pointer to void is a GNU extension.
2022-10-14 16:21:14 +02:00
Florian Obser 08dcae0dab Arithmetic on a pointer to void is a GNU extension. 2022-10-14 13:56:32 +02:00
W.C.A. Wijngaards 5ac1bc13cb - Tag for 1.17.0 release. The code repository continues with 1.17.1. 2022-10-13 09:34:44 +02:00
W.C.A. Wijngaards f5e1ef650d Merge branch 'branch-1.17.0' 2022-10-13 09:32:22 +02:00
George Thessalonikefs d25e0cd9b0 - Fix PROXYv2 header read for TCP connections when no proxied addresses
are provided.
2022-10-11 17:39:30 +02:00
Jonathan Gray 4f27799456 consistently use IPv4/IPv6 2022-10-10 19:14:58 +11:00
W.C.A. Wijngaards 97d1cff315 Changelog note for tag for 1.17.0rc1 release. 2022-10-07 13:29:33 +02:00
George Thessalonikefs a4631a3ecf - Fix unit test to properly test the reuse_write_wait_pop function. 2022-10-07 11:29:46 +02:00
George Thessalonikefs 2569b12b9c - Fix to stop possible loops in the tcp reuse code (write_wait list
and tcp_wait list). Based on analysis and patch from Prad Seniappan
  and Karthik Umashankar.
2022-10-07 11:25:36 +02:00
W.C.A. Wijngaards bf1cce6f9b - Fix proxy length debug output printout typecasts. 2022-10-06 15:53:21 +02:00
W.C.A. Wijngaards b043bc5eb4 - Fix to stop responses with TC flag from resulting in partial
responses. It retries to fetch the data elsewhere, or fails the
  query and in depth fix removes the TC flag from the cached item.
2022-10-06 10:01:09 +02:00
George Thessalonikefs d122617dd4 - Fix checkconf test for dnscrypt and proxy port. 2022-10-05 22:03:01 +02:00
W.C.A. Wijngaards 6b8181acb7 - Fix dnscrypt compile for proxy protocol code changes. 2022-10-05 14:09:12 +02:00
George Thessalonikefs f609a45354 - Make ede.tdir test more predictable by using static data. 2022-10-05 02:44:50 +02:00
George Thessalonikefs 60db1111c0 - Use DEBUG_TDIR from environment in mini_tdir.sh for debugging.
- Fix string comparison in mini_tdir.sh.
2022-10-05 01:13:29 +02:00
George Thessalonikefs 40b2b3a6f3 Changelog entry for #764
- Merge #764: Leniency for target discovery when under load (for
  NRDelegation changes).
2022-10-04 22:24:18 +02:00
Yorgos ThessalonikefsandGitHub f1d263a318 Leniency for target discovery when under load (for NRDelegation changes) (#764)
* - Introduce leniency for target discovery when under load.

* - Allow for easier testing (to be reverted).

* - Happy compiler.

* - Precheck access to target_fetch_policy.

* - Do not mark a nameserver as resolved when one of A/AAAA is negative.

* - Update fetch_glue.rpl test for (possible) outstanding queries.

* - Update fetch_glue_cname.rpl test for possible outstanding queries.

* - Better fix for fetch_glue_cname.rpl.

* - Fix iter_emptydp_for_glue.rpl to match the referral.

* - Disabled the nxns tests for now (to be reverted).

* - Update iter_recurse.rpl for possible outstanding queries.

* Revert "- Disabled the nxns tests for now (to be reverted)."

This reverts commit 34a9c13a90.

* Revert "- Allow for easier testing (to be reverted)."

This reverts commit b6dfe35e1d.
2022-10-04 22:21:08 +02:00
W.C.A. Wijngaards f0614a57f8 - Fix to clean up after the acl_interface unit test. 2022-10-04 16:59:10 +02:00
W.C.A. Wijngaards bf7a2884fb - Fix static analysis report to remove dead code from the
rpz_callback_from_iterator_module function.
2022-10-04 09:08:11 +02:00
W.C.A. Wijngaards c0eaadfc42 - Fix to close errno block in comm_point_tcp_handle_read outside of
ifdef.
2022-10-03 16:21:39 +02:00
George Thessalonikefs 22e43aa631 Changelog entry for #760
- Merge #760: PROXYv2 downstream support. (New proxy-protocol-port
  configuration option).
2022-10-03 15:34:22 +02:00
Yorgos ThessalonikefsandGitHub c4e51a4cfe PROXYv2 downstream support (#760) 2022-10-03 15:29:47 +02:00
W.C.A. Wijngaards 7d96a7e3fe - Fix windows compile, the identifier interface is defined in headers. 2022-10-03 15:03:50 +02:00
W.C.A. Wijngaards 9842fbf760 - Fix test tdir skip report printout. 2022-10-03 10:26:30 +02:00
W.C.A. Wijngaards a102fb1df8 - Fix to remove erroneous TC flag from TCP upstream. 2022-10-03 09:53:41 +02:00
Willem Toorop bd2c202674 The generated lexer and parser sources for configuring cookies 2022-09-28 10:34:06 +02:00
Willem Toorop 75f3fbdd65 Downstream DNS Cookies a la RFC7873 and RFC9018
Create server cookies for clients that send client cookies.
Needs to be turned on in the config file with:

	answer-cookie: yes

A cookie-secret can be configured for anycast setups.
Also adds an access control list that will allow queries with
either a valid cookie or over a stateful transport.
2022-09-28 10:28:19 +02:00
Willem Toorop 71f23ef354 extended_error_encode() for extended errors 2022-09-28 09:57:56 +02:00
George Thessalonikefs 5b98816751 - Better output for skipped tdir tests. 2022-09-26 15:51:28 +02:00
TCY16 0b176750bd add @wcawijngaards' review comments 2022-09-26 12:14:17 +02:00
TCY16 f0989fc754 differentiate between malloc and regional_alloc 2022-09-26 11:49:49 +02:00
TCY16 c9f90def0a swap malloc for regional_alloc and add free 2022-09-26 11:18:58 +02:00
TCY16 57c4f28f20 add happy-flow test 2022-09-21 16:21:48 +02:00
TCY16 d731fa2e21 use correct edns_list attach function 2022-09-21 16:19:38 +02:00
W.C.A. Wijngaards e93c75a5d4 - Fix doxygen warning in respip.h. 2022-09-21 15:23:04 +02:00
W.C.A. Wijngaards 8e18f11965 - This patch was released in 1.16.3, the code repository continues
with the previous features and fixes for 1.17.0.
2022-09-21 12:16:13 +02:00
W.C.A. Wijngaards e3871ca907 Merge branch 'branch-1.16.3' 2022-09-21 12:11:26 +02:00
TCY16 dcfcde2ec8 add cached EDE strings 2022-09-21 11:21:33 +02:00
W.C.A. Wijngaards 137719522a - Patch for CVE-2022-3204 Non-Responsive Delegation Attack. 2022-09-21 11:10:38 +02:00
W.C.A. Wijngaards bd3c5702a7 branch for 1.16.3 release. 2022-09-21 11:09:03 +02:00
Philip-NLnetLabsandGitHub f68e252345 Merge pull request #757 from NLnetLabs/philip-issue-749
Philip issue 749
2022-09-20 15:07:14 +02:00
Philip Homburg 1e5158045f Merge branch 'ryndia-issue-749' into philip-issue-749 2022-09-20 14:48:57 +02:00
George Thessalonikefs 99e12ae4b5 - Remove unused testcode/mini_tpkg.sh file. 2022-09-20 14:47:24 +02:00
Philip Homburg 5f3b460586 Align with version 1.58 on cvsweb.openbsd.org 2022-09-20 14:47:19 +02:00
George Thessalonikefs 9b1647ebae - Convert tdir tests to use the new skip_test functionality. 2022-09-20 14:45:20 +02:00
George Thessalonikefs 307805b64f Changelog entry for #753:
- Merge #753: ACL per interface. (New interface-* configuration
  options).
2022-09-20 11:36:01 +02:00
Yorgos ThessalonikefsandGitHub f02d6bddd1 Merge pull request #753 from NLnetLabs/acl_interface
ACL per interface
2022-09-16 16:45:59 +02:00
eb02170338 Apply suggestions from code review
Co-authored-by: Wouter Wijngaards <wcawijngaards@users.noreply.github.com>
2022-09-16 14:43:23 +02:00
George Thessalonikefs d301bfe4a2 - ACL per interface: refactor, complete testing and a bugfix for
interface names.
2022-09-11 20:57:41 +02:00
George Thessalonikefs fc123303ac - Add functionality to skip tdir tests from the .pre file;
- Initial tests for interface-* options.
2022-09-11 20:21:32 +02:00
George Thessalonikefs aec33b3d63 Documentation for interface-* options. 2022-09-11 20:21:32 +02:00
George Thessalonikefs 7e9fd2114b Cleared error messages for interface-* options. 2022-09-11 20:21:32 +02:00
George Thessalonikefs c30bdff939 Initial commit for interface based ACL. 2022-09-11 20:21:32 +02:00
ryndiaandGitHub cfc656294e Update arc4random.c 2022-09-07 20:16:20 +04:00
W.C.A. Wijngaards 007db2c327 - Fix to check pthread_t size after pthread has been detected. 2022-09-02 10:21:00 +02:00
W.C.A. Wijngaards 5bbaf78c3f - Remove include that was there for debug purposes. 2022-09-02 10:11:23 +02:00
W.C.A. Wijngaards 57230d7f22 - Fix to log a verbose message at operational notice level if a
thread is not responding, to stats requests. It is logged with
  thread identifiers.
2022-09-01 15:14:20 +02:00
TCY16 5f309d0018 Add caching EDEs 2022-09-01 14:10:14 +02:00
W.C.A. Wijngaards d66e1cccf8 - Fix to set out of file descriptor warning to operational verbosity. 2022-09-01 14:01:56 +02:00
W.C.A. Wijngaards 2450b4653a - Slow down log frequency of write wait failures. 2022-09-01 14:00:29 +02:00
W.C.A. Wijngaards eb3378396f - Fix to update config tests to fix checking if nonblocking sockets
work on OpenBSD.
2022-09-01 09:16:05 +02:00
W.C.A. Wijngaards 1f5cc25974 - Fix for wait for udp send to stop when packet is successfully sent. 2022-08-31 16:45:15 +02:00
W.C.A. Wijngaards ec5812a748 - Fix to wait for blocked write on UDP sockets, with a timeout if it
takes too long the packet is dropped.
2022-08-31 11:54:11 +02:00
W.C.A. Wijngaards 10a5a5880a - Patch from Vadim Fedorenko that adds MSG_DONTWAIT to receive
operations, so that instruction reordering does not cause mistakenly
  blocking socket operations.
2022-08-31 10:11:25 +02:00
W.C.A. Wijngaards 2fa1c17cd9 - Fix to avoid process wide fcntl calls mixed with nonblocking
operations after a blocked write.
2022-08-31 10:09:39 +02:00
TCY16 b642c5fe1f add better URI template checking 2022-08-25 14:06:13 +02:00
TCY16 f3fa363443 implement @gthess' review comments; fix check on compulsory text and add tests 2022-08-24 12:38:08 +02:00
W.C.A. Wijngaards e6f878ee71 - Fix #741: systemd socket activation fails on IPv6. 2022-08-22 09:12:08 +02:00
TCY16 8d939691a3 implement @wcawijngaards' review comment 2022-08-19 14:48:47 +02:00
TCY16 b465e0cfc0 add testcase and fix comment 2022-08-15 16:12:22 +02:00
TCY16 73b3b32062 fix styling nits 2022-08-15 14:41:06 +02:00
TCY16 6e31d1f5be add dohpath parsing 2022-08-15 14:36:35 +02:00
TCY16 14fe4669e7 fix testcase comment 2022-08-12 14:09:00 +02:00
W.C.A. Wijngaards dc6c04b243 - Fix to log accept error ENFILE and EMFILE errno, but slowly, once
per 10 seconds. Also log accept failures when no slow down is used.
2022-08-12 09:54:29 +02:00
W.C.A. Wijngaards ef57f8bd51 - Fix #734 [FR] enable unbound-checkconf to detect more (basic)
errors.
2022-08-05 14:41:05 +02:00
W.C.A. Wijngaards fbe8e3b0b2 - Fix ratelimit inconsistency, for ip-ratelimits the value is the
amount allowed, like for ratelimits.
2022-08-04 11:33:37 +02:00
W.C.A. Wijngaards 1c164ab442 Changelog note for #730
- Merge #730 from luisdallos: Fix startup failure on Windows 8.1 due
  to unsupported IPV6_USER_MTU socket option being set.
2022-08-02 16:12:31 +02:00
Wouter WijngaardsandGitHub e738ec31ca Merge pull request #730 from luisdallos/win81-unsup-sockopt-fix
Fix startup failure on Windows 8.1 due to unsupported IPV6_USER_MTU socket option being set
2022-08-02 16:12:16 +02:00
W.C.A. Wijngaards 07b073ddb3 - Fix unittest for edns subnet change. 2022-08-02 14:43:57 +02:00
W.C.A. Wijngaards 0f08cc6d55 - Fix edns subnet so that scope 0 answers only match sourcemask 0
queries for answers from cache if from a query with sourcemask 0.
2022-08-02 14:13:55 +02:00
Luis Dallos 7d3c6f1c43 Fix startup failure on Windows 8.1 due to unsupported IPV6_USER_MTU socket option being set
Newer mingw-w64 (starting from 8.0.1) introduces support for `IPV6_USER_MTU` socket
option [1], which is not supported on Windows 8.1 and older [2]. As there is no way
to avoid this socket option from being picked at compile time when targeting older
versions of Windows, check for `setsockopt(..., IPV6_USER_MTU, ...)` failures at
runtime in order to avoid startup failure on those versions of Windows where the
`IPV6_USER_MTU` socket option is unsupported.

[1]: mirror/mingw-w64@e30bff4
[2]: `WSAGetLastError()` returns `WSAENOPROTOOPT` (`Bad protocol option`) error code
2022-08-01 23:03:24 -04:00
W.C.A. Wijngaards cd22fdc28d - Fix #728: alloc_reg_obtain() core dump. Stop double
alloc_reg_release when serviced_create fails.
2022-08-01 16:45:41 +02:00
W.C.A. Wijngaards 5ae48f85c0 - The code repo continues with 1.16.3. 2022-08-01 15:17:10 +02:00
W.C.A. Wijngaards cbed768b8f - Tests for ghost domain fixes. 2022-08-01 13:26:22 +02:00
W.C.A. Wijngaards f6753a0f10 - Fix the novel ghost domain issues CVE-2022-30698 and CVE-2022-30699. 2022-08-01 13:24:40 +02:00
George Thessalonikefs efdd70c7b5 - Cleanup some comments and TODO text. 2022-07-23 19:55:15 +02:00
JINMEI Tatuya fccb2eb2e8 prevent memory leak in case cache isn't reused 2022-07-22 14:33:21 -07:00
Jonathan Gray 1464b166a4 fix use after free when WSACreateEvent() fails 2022-07-22 18:23:59 +10:00
W.C.A. Wijngaards f993ffbb07 Changelog note for #718
- Merge #718: Introduce infra-cache-max-rtt option to config max
  retransmit timeout.
2022-07-19 15:19:12 +02:00
Wouter WijngaardsandGitHub ef9bb0213b Merge pull request #718 from hunts/make_rtt_max_timeout_configurable
Introduce infra-cache-max-rtt option to config max retransmit timeout
2022-07-19 15:15:34 +02:00
George Thessalonikefs 309e23515e - Update documentation for 'outbound-msg-retry:'. 2022-07-19 12:47:01 +02:00
Minghang Chen 249efd4285 Introduce infra-cache-max-rtt option to config max retransmit timeout
Added the option and let it default to 120 seconds so that it won't change
current behavior.

Related-to #717
2022-07-16 01:46:18 -07:00
W.C.A. Wijngaards 12cd495d55 - iana portlist update. 2022-07-15 09:20:25 +02:00
W.C.A. Wijngaards 33bd49af81 - Merge PR 714: Avoid treat normal hosts as unresponsive servers.
And fixup the lock code.
2022-07-15 08:51:31 +02:00
Wouter WijngaardsandGitHub 6e50c46102 Merge pull request #714 from hunts/rtt-of-expired-host
Avoid treat normal hosts as unresponsive servers
2022-07-15 08:49:51 +02:00
Hunts Chen 88bf803297 Avoid treat normal hosts as unresponsive servers
This is a fix for issue #713

When infra-keep-probing is on, all hosts with expired entries were treated as
unresponsive servers and thus causing problems (see #713).

This commit change that, so that normal hosts with expired entries are treated
as unknown servers.
2022-07-14 10:16:13 -07:00
George Thessalonikefs 9e4a17baaf - For windows crosscompile, fix setting the IPV6_MTU socket option
equivalent (IPV6_USER_MTU); allows cross compiling with latest
  cross-compiler versions.
2022-07-12 17:17:59 +02:00
W.C.A. Wijngaards 2abd6f7d58 - Fix dname count in sldns parse type descriptor for SVCB and HTTPS. 2022-07-12 13:29:51 +02:00
W.C.A. Wijngaards 7696398231 - Fix verbose EDE error printout. 2022-07-11 13:13:51 +02:00
Wouter WijngaardsandGitHub 886ffcc155 Merge pull request #711 from NLnetLabs/add-security.md
add SECURITY.md, based on krill
2022-07-11 11:37:28 +02:00
W.C.A. Wijngaards a45ced7739 - This became 1.16.1 on 11 July 2022.
The code repo continues with version 1.16.2 under development.
2022-07-11 11:34:54 +02:00
maertsenandGitHub 125983558b add SECURITY.md, based on krill 2022-07-05 16:39:04 +02:00
W.C.A. Wijngaards 903538c76e - Tag for 1.16.1rc1 release. 2022-07-04 13:48:56 +02:00
George Thessalonikefs d447d60fc6 Merge branch 'crrodriguez-IP_BIND_ADDRESS_NO_PORT' 2022-07-04 12:49:00 +02:00
George Thessalonikefs e5f66b4902 - For #668: relocate and make code more portable. 2022-07-04 12:46:17 +02:00
George Thessalonikefs 0f4c4c1163 Merge branch 'IP_BIND_ADDRESS_NO_PORT' of https://github.com/crrodriguez/unbound into crrodriguez-IP_BIND_ADDRESS_NO_PORT 2022-07-04 11:15:58 +02:00
George Thessalonikefs eda0c0c194 - Fix bug introduced in 'improve val_sigcrypt.c::algo_needs_missing for
one loop pass'.
2022-07-04 09:34:45 +02:00
George Thessalonikefs 309b1d368b - Reintroduce documentation and more EDE support for
val_sigcrypt.c::dnskeyset_verify_rrset_sig.
2022-07-04 00:06:26 +02:00
George Thessalonikefs c513119bba - Improve val_sigcrypt.c::algo_needs_missing for one loop pass. 2022-07-03 23:32:18 +02:00
George Thessalonikefs a30286502c - Fix for correct openssl error when adding windows CA certificates to
the openssl trust store.
2022-07-03 22:41:39 +02:00
George Thessalonikefs 317bab9f1d For #660: formatting, less verbose logging, add EDE information. 2022-07-03 22:32:56 +02:00
George Thessalonikefs 2fba248ebe Changelog entry for #660
- Merge PR #660 from Petr Menšík: Sha1 runtime insecure.
2022-07-03 22:28:17 +02:00
Yorgos ThessalonikefsandGitHub e102aea751 Merge pull request #660 from InfrastructureServices/sha1-runtime-insecure
Sha1 runtime insecure
2022-07-03 22:24:58 +02:00
George Thessalonikefs 24cb764d61 - Add missing changelog entries from PR #706. 2022-07-03 22:21:56 +02:00
George Thessalonikefs cb28bdcdb0 Changelog entry for #671:
- Merge PR #671 from Petr Menšík: Disable ED25519 and ED448 in FIPS
  mode on openssl3.
2022-07-03 22:08:47 +02:00
George Thessalonikefs bcf7a8aed2 Merge branch 'InfrastructureServices-fips-mode-algo-ed25519' 2022-07-03 21:59:20 +02:00
George Thessalonikefs 391dd86c3b Merge branch 'master' into InfrastructureServices-fips-mode-algo-ed25519 2022-07-01 17:34:09 +02:00
Yorgos ThessalonikefsandGitHub b0ce31b4e4 Merge pull request #706 from NLnetLabs/nxns-fallback
NXNS fallback
2022-07-01 16:24:33 +02:00
George Thessalonikefs 2dbaba7d73 - Improved logging for NXNS fallback. 2022-07-01 16:18:33 +02:00
George Thessalonikefs 923eb7d474 - Allow fallback to the parent side when MAX_TARGET_NX is reached.
This will also allow MAX_TARGET_NX more NXDOMAINs.
2022-06-29 17:32:29 +02:00
George Thessalonikefs 58b21e4fca - Fix to not count cached NXDOMAIN for MAX_TARGET_NX. 2022-06-29 17:26:09 +02:00
W.C.A. Wijngaards ef718f02e4 Remove debug output from windows compile workflow. 2022-06-29 12:01:39 +02:00
W.C.A. Wijngaards c8add463c4 - Fix compile warning for windows compile. 2022-06-29 12:00:44 +02:00
W.C.A. Wijngaards 203f606c38 - Fix detection of libz on windows compile with static option. 2022-06-29 11:41:20 +02:00
W.C.A. Wijngaards e7badb9428 Debug output for windows compile workflow. 2022-06-29 11:17:12 +02:00
George Thessalonikefs b816318106 - Fix #704: [FR] Statistics counter for number of outgoing UDP queries
sent; introduces 'num.query.udpout' to the 'unbound-control stats'
  command.
2022-06-29 10:51:54 +02:00
George Thessalonikefs 1ceb031b58 - Add debug option to the mini_tdir.sh test code. 2022-06-29 10:47:18 +02:00
George Thessalonikefs 352e1b28a1 Merge branch 'master' of github.com:NLnetLabs/unbound 2022-06-29 10:44:43 +02:00
W.C.A. Wijngaards 80dbc7dd2c - iana portlist update. 2022-06-29 09:38:31 +02:00
George Thessalonikefs c8e7539313 - Formatting. 2022-06-28 18:31:27 +02:00
George Thessalonikefs 12796d0af8 - Fix for cached 0 TTL records to not trigger prefetching when
serve-expired-client-timeout is set.
2022-06-28 17:05:28 +02:00
W.C.A. Wijngaards b057d2127a - Fix test program dohclient close to use portability routine. 2022-06-28 09:23:43 +02:00
George Thessalonikefs 9177867d07 - Show the output of the exact .rpl run that failed with 'make test'. 2022-06-28 00:22:30 +02:00
tcarpayandGitHub 9e6b838d0b Update changelog with #705 change 2022-06-23 17:21:17 +02:00
tcarpayandGitHub 12552fddf0 clarify addition of verbosity level mentioned in #705 2022-06-23 15:00:48 +02:00
Philip Homburg 3bade62c8a Fix use after free issue with edns options (https://github.com/NLnetLabs/unbound/issues/663) 2022-06-22 15:00:28 +02:00
Philip Homburg c806b5a2df Merge branch 'zone-parsers-blank-line-issue' 2022-06-21 15:32:35 +02:00
Philip Homburg 7f67f7118c Change log entry for lines with blanks issue 2022-06-21 15:30:04 +02:00
George Thessalonikefs 7c9177095f - Remove unused LDNS function check for GOST Engine unloading. 2022-06-20 16:27:15 +02:00
George Thessalonikefs 233cb5c218 - Note in the unbound.conf text that NOTIFY is allowed from the url:
addresses for auth and rpz zones.
2022-06-14 17:59:56 +02:00
George Thessalonikefs e6a0afae73 Changelog entry for #688
- Merge PR #688: Rpz url notify issue.
2022-06-14 17:56:23 +02:00
Yorgos ThessalonikefsandGitHub 90767fce98 Merge pull request #688 from NLnetLabs/rpz-url-notify-issue
Rpz url notify issue
2022-06-14 17:53:01 +02:00
George Thessalonikefs 187bc72633 - Add testcase for allowing NOTIFY on URL addresses. 2022-06-14 17:44:37 +02:00
Philip Homburg 78b4b12713 Test loading a zone with blank lines over https 2022-06-14 17:05:03 +02:00
Philip Homburg b8e111e379 Avoid network traffic during test, a bit of cleanup 2022-06-14 17:04:23 +02:00
Philip Homburg 0e1c266f86 Fix issue with lines that only consist of blanks with optional comment 2022-06-13 11:41:14 +02:00
Philip Homburg c6ef39a22d Test loading a cached zone that has lines consisting of blanks 2022-06-08 14:45:30 +02:00
George Thessalonikefs 459b73018f - Fix for edns client subnet to respect not looking in its cache when
instructed to do so (e.g., prefetch).
2022-06-03 16:11:35 +02:00
W.C.A. Wijngaards 97b7224885 - makedist.sh picks up 32bit libssp-0.dll when 32bit compile. 2022-06-03 15:18:27 +02:00
W.C.A. Wijngaards a3ab0921b0 Continue with 1.16.1 under development. 2022-06-02 13:56:52 +02:00
Philip Homburg 16dd802c2e Add url 'master' to allow notify list 2022-05-31 15:10:38 +02:00
Philip Homburg 6dad2d2fc6 allow-notify doesn't work for url on rpz zones (https://github.com/NLnetLabs/unbound/issues/679) 2022-05-31 15:10:38 +02:00
W.C.A. Wijngaards edc1d07718 - Version is set to 1.16.0 for release. Release tag 1.16.0rc1. 2022-05-27 16:09:28 +02:00
W.C.A. Wijngaards 2d61706171 - Fix #684: [FTBS] configure script error with libmnl on openSUSE 15.3 (and possibly other distributions) 2022-05-27 09:10:16 +02:00
W.C.A. Wijngaards 11d077c826 - Fix some lint type warnings. 2022-05-20 15:32:27 +02:00
W.C.A. Wijngaards d19d7b81ec - Fix ede test to not use default pidfile, and use local interface. 2022-05-20 15:14:11 +02:00
W.C.A. Wijngaards 714200ad0e - Fix to silence test for ede error output to the console from the
test setup script.
2022-05-20 09:08:29 +02:00
George Thessalonikefs c76420664a Merge branch 'fips-mode-algo-ed25519' of https://github.com/InfrastructureServices/unbound into InfrastructureServices-fips-mode-algo-ed25519 2022-05-19 12:53:48 +03:00
George Thessalonikefs 7e506bb477 - Fix typos in config_set_option for the 'num-threads' and
'ede-serve-expired' options.
2022-05-18 19:56:26 +03:00
George Thessalonikefs 91c298c901 Merge branch 'subnet_cache_prefetch' of https://github.com/tilan7663/unbound into tilan7663-subnet_cache_prefetch 2022-05-17 16:16:09 +03:00
George Thessalonikefs f73e548934 - Fix #678: [FR] modify behaviour of unbound-control rpz_enable zone,
by updating unbound-control's documentation.
2022-05-15 22:53:17 +03:00
W.C.A. Wijngaards e62b309959 - For #677: Added tls-system-cert to config parser and documentation.
- Changelog note for #677.
2022-05-12 16:30:19 +02:00
Wouter WijngaardsandGitHub 2132e67b36 Merge pull request #677 from InfrastructureServices/use-system-cas
Allow using system certificates not only on Windows
2022-05-12 16:16:49 +02:00
Petr Mensik 0abfddd279 Allow using system certificates not only on Windows
OpenSSL has a way to load default file. That file might contain usable
certificates to verify common connections. Allow similar trust as on
windows and leave it on openssl package to provide sane defaults.

Also provide use-system-cert alias, because it is not windows specific
anymore.
2022-05-12 16:07:41 +02:00
Yorgos ThessalonikefsandGitHub ba4564b4ef Merge pull request #674 from NLnetLabs/prefetch_when_ecs_enabled
Fix #417: prefetch and ECS causing cache corruption when used
2022-05-12 01:17:42 +02:00
George Thessalonikefs daf316ea1b - Fix #417: prefetch and ECS causing cache corruption when used
together.
2022-05-12 00:56:01 +02:00
W.C.A. Wijngaards f0d91950ad - Fix #673: DNS over TLS: error: SSL_handshake syscall: No route to
host.
2022-05-11 17:10:42 +02:00
Petr Mensik 917c30a46a Disable ED25519 and ED448 in FIPS mode on openssl3
Both crypto functions are not allowed by FIPS 140-3. Use openssl 3.0
function to check FIPS mode presence and use it to make those algorithms
unsupported.
2022-05-11 16:19:25 +02:00
George Thessalonikefs 77149fc2aa - Fix Python build in non-source directory; based on patch by
Michael Tokarev.
2022-05-10 15:57:17 +02:00
tcarpayandGitHub 47ce372f13 Changelog entry for #604: Add the basic EDE (RFC8914) cases 2022-05-06 12:53:49 +02:00
tcarpayandGitHub 0ce36e8289 Add the basic EDE (RFC8914) cases (#604) 2022-05-06 12:48:53 +02:00
Tian Lan 8afbc0944f Add prefetch support for subnet cache entries
- Entries in the subnet cache should now be prefetched.

- Rename testdata subnet_*.crpl to subnet_*.rpl so they are visible to
  make test

Signed-off-by: Tian Lan <tian.lan@twosigma.com>
2022-04-29 11:46:05 -04:00
W.C.A. Wijngaards b61b0af5d6 - Fix #670: SERVFAIL problems with unbound 1.15.0 running on
OpenBSD 7.1.
2022-04-28 14:51:47 +02:00
Cristian Rodríguez 6a4ea692d4 Set IP_BIND_ADDRESS_NO_PORT on outbound tcp sockets
When bound to a local address the kernel does not know if the socket
will listen() or connect() and must reserve a port inmediately after
bind() effectively limiting the random port range to ~32k.
when IP_BIND_ADDRESS_NO_PORT is set, the kernel delays source port
allocation until the time the complete 4-tuple is known allowing
a much larger number of combinations
2022-04-23 23:08:45 +00:00
Christian Allred d19e12ab5d Merge branch 'master' of https://github.com/NLnetLabs/unbound into restart_conf 2022-04-18 12:16:40 -07:00
Petr Mensik 74c6cf5ac6 Log detailed openssl error also for digests failures
Make output still only shown in verbose detail. But provide openssl
error details to make a reason more obvious.
2022-04-12 16:13:49 +02:00
Petr Mensik 33c8baaaba Forward indeterminate status higher
Create a path where it can result in insecure.
2022-04-08 16:26:50 +02:00
Petr Mensik 6cfcf21451 Make SHA-1 signed domains insecure if openssl refuses the digest
RHEL9/CentOS 9 would fail in default crypto policy. If call to openssl
returns invalid digest then report the name insecure. If all tested
signatures return the same issue, then make the reply insecure.
2022-04-08 16:26:50 +02:00
W.C.A. Wijngaards 1289c53c1a - Fix zonemd unsupported algo check to set reason to NULL before the
check routine, but after malformed checks, to get the correct NULL
  output when the digest matches.
2022-04-08 11:19:40 +02:00
W.C.A. Wijngaards d02e83ae2b - Fix zonemd unsupported algo check to print unsupported reason before
zeroing it.
2022-04-08 11:10:11 +02:00
W.C.A. Wijngaards 8f2847ba69 - Fix zonemd unsupported algo check reason to not copy to next record,
and check for success for debug printout.
2022-04-08 10:54:57 +02:00
W.C.A. Wijngaards 730a03e9bd - Fix zonemd unsupported algo check. 2022-04-08 09:36:01 +02:00
W.C.A. Wijngaards e4ca71e85b - Fix zonemd check to allow unsupported algorithms to load.
If there are only unsupported algorithms, or unsupported schemes,
  and no failed or successful other ZONEMD records, or malformed
  or bad ZONEMD records, the unsupported records allow the zone load.
2022-04-08 09:29:37 +02:00
W.C.A. Wijngaards 8f8a8a341a - Fix spelling error in comment in sldns_str2wire_svcparam_key_lookup. 2022-03-25 16:07:23 +01:00
W.C.A. Wijngaards 722391baf1 - Fix #651: [FR] Better logging for refused queries. 2022-03-23 13:56:52 +01:00
George Thessalonikefs ee63ac8f9b - Minor formatting. 2022-03-18 16:04:16 +02:00
George Thessalonikefs 1e23c4a376 Merge PR #648 from eaglegai: fix -q doesn't work when use with
'unbound-control stats_shm'
2022-03-18 16:03:35 +02:00
eaglegai c60bed8eef fix -q doesn't work when use with 'unbound-control stats_shm'
Signed-off-by: eaglegai <eaglegai@163.com>
2022-03-18 19:08:50 +08:00
W.C.A. Wijngaards 9484ddee2e - Fix to describe auth-zone and other configuration at the local-zone
configuration option, to allow for more broadly view of the options.
2022-03-17 14:24:37 +01:00
W.C.A. Wijngaards 402135b41a - Fix to ensure uniform handling of spaces and tabs when parsing RRs. 2022-03-16 09:54:53 +01:00
W.C.A. Wijngaards b0dbfa37f9 Changelog note for #644, move commands together for library binary. 2022-03-09 09:43:07 +01:00
Wouter WijngaardsandGitHub 79f8fc1472 Merge pull request #644 from FGasper/issue_643_pkgconfig_install_lib
Make `install-lib` make target install the pkg-config file.
2022-03-09 09:41:54 +01:00
Felipe Gasper 516f4f9cf4 Make install-lib make target install the pkg-config file.
Issue #643
2022-03-08 14:01:09 -05:00
W.C.A. Wijngaards 62c5039ab9 - Fix configure for python to use sysutils, because distutils is
deprecated. It uses sysutils when available, distutils otherwise.
2022-03-07 14:02:31 +01:00
W.C.A. Wijngaards b202b0874c - Fix for #637: fix integer overflow checks in sldns_str2period. 2022-03-03 16:24:46 +01:00
W.C.A. Wijngaards debe5c665f - Fix #637: Integer Overflow in sldns_str2period function. 2022-03-03 14:19:59 +01:00
W.C.A. Wijngaards f81420d77f - Fix compile warnings for printf ll format on mingw compile. 2022-03-02 14:34:36 +01:00
George Thessalonikefs b8e7dfa01e - Various fixes for #632: variable initialisation, convert the qinfo
to str once, accept trailing dot in the local-zone ipset option.
2022-03-02 14:29:56 +01:00
George Thessalonikefs f7ca447368 Changelog entry for #632
- Merge PR #632 from scottrw93: Match cnames in ipset.
2022-03-02 13:22:17 +01:00
gthessandGitHub a6880644aa Merge pull request #632 from scottrw93/cnames-ipset
Match cnames in ipset
2022-03-02 13:17:35 +01:00
George Thessalonikefs 0b8e6dafe5 - Added tests for ipset. 2022-03-02 13:08:16 +01:00
W.C.A. Wijngaards 99a3f5ee85 - Fix pythonmod for change in iter_dp_is_useless function prototype. 2022-03-01 10:27:24 +01:00
W.C.A. Wijngaards 7749d98a14 - Fix for edns client subnet option add fix in removal code, from review. 2022-02-28 14:51:35 +01:00
W.C.A. Wijngaards 8e8ccfe3c3 - Fix edns client subnet to add the option based on the option list,
so that it is not state dependent, after the state fix of #605 for
  double EDNS options.
2022-02-28 14:39:48 +01:00
George Thessalonikefs ad158ed5cc Changelog entry for #623:
- Merge #623 from rex4539: Fix typos.
2022-02-28 12:37:59 +01:00
gthessandGitHub 6e79237dc8 Merge pull request #623 from rex4539/typos
Fix typos
2022-02-28 12:36:11 +01:00
George Thessalonikefs 82adcfb971 - Fix #630: Unify the RPZ log messages. 2022-02-28 12:07:25 +01:00
W.C.A. Wijngaards 6e64356175 - Fix for #633: updated fix with new text. 2022-02-28 09:56:30 +01:00
W.C.A. Wijngaards e11cf2d964 - Fix #633: Document unix domain socket support for unbound-control. 2022-02-28 09:39:33 +01:00
W.C.A. Wijngaards c084c27b39 - Fix check interface existence for support detection in remote lookup. 2022-02-25 15:24:40 +01:00
W.C.A. Wijngaards 2c3225e6c7 - update Makefile dependencies. 2022-02-25 10:29:33 +01:00
W.C.A. Wijngaards 4b772ed571 - Fix to detect that no IPv6 support means that IPv6 addresses are
useless for delegation point lookups.
2022-02-25 10:27:56 +01:00
scottrw93 4f3df7ee86 Match cnames in ipset 2022-02-20 22:37:49 +00:00
Dimitris Apostolou c7be51a11b Fix typos 2022-02-18 15:51:03 +02:00
W.C.A. Wijngaards c4d8104db3 Changelog note for #631.
- Merge #631 from mollyim: Replace OpenSSL's ERR_PACK with
  ERR_GET_REASON.
2022-02-18 09:37:44 +01:00
Wouter WijngaardsandGitHub fbbb42c9d4 Merge pull request #631 from mollyim/boringssl-compat
Replace OpenSSL's ERR_PACK with ERR_GET_REASON
2022-02-18 09:37:34 +01:00
W.C.A. Wijngaards a746d9693a - Fix that address not available is squelched from the logs for
udp connect failures. It is visible on verbosity 4 and more.
2022-02-18 09:03:56 +01:00
Oscar Mira 78aee89201 Replace OpenSSL's ERR_PACK with ERR_GET_REASON 2022-02-17 20:20:18 +01:00
W.C.A. Wijngaards 6de5310728 - Fix for #628: fix rpz-passthru for qname trigger by localzone type. 2022-02-16 09:51:25 +01:00
W.C.A. Wijngaards 2b90181d3a - Fix #628: A rpz-passthru action is not ending RPZ zone processing. 2022-02-15 16:20:12 +01:00
W.C.A. Wijngaards 91a5cc9a08 - Remove debug info from #618 fix. 2022-02-11 16:34:24 +01:00
W.C.A. Wijngaards a0feea393a - Fix #618: enabling interface-automatic disables DNS-over-TLS.
Adds the option to list interface-automatic-ports.
2022-02-11 10:58:53 +01:00
W.C.A. Wijngaards 4c6b59fa47 - Fix #624: Unable to stop Unbound in Windows console (does not
respond to CTRL+C command).
2022-02-11 08:53:24 +01:00
W.C.A. Wijngaards ed4ce7b398 Release 1.15.0 on 10 feb 2022.
The repository continues with version 1.15.1.
And Changelog note.
2022-02-10 09:54:51 +01:00
W.C.A. Wijngaards 58ac43c754 Note 1.15.0rc1 tag creation in Changelog.
- Tag for 1.15.0rc1 created.
2022-02-07 11:31:41 +01:00
W.C.A. Wijngaards 5f724da8c5 - Fix that TCP interface does not use TLS when TLS is also configured. 2022-02-07 09:31:10 +01:00
JINMEI Tatuya 5b2eda28e3 add keep-cache option to unbound-control reload to keep caches 2021-11-11 10:47:08 -08:00
Christian Allred 766244bd81 Document max-query-restarts option 2021-04-05 18:59:09 -07:00
Christian Allred 07c0d04a14 Use max-query-restarts in iterative resolver 2021-04-05 16:25:43 -07:00
Christian Allred 0e3068559c Add max-query-restarts to grammar and lexer 2021-04-05 16:24:49 -07:00
Christian Allred 41fa45c99e Add max-query-restarts config parameter 2021-04-05 15:41:53 -07:00
Christopher Zimmermann 1d23e0c920 Merge remote-tracking branch 'upstream/master' 2021-02-03 13:19:19 +01:00
Frank Riley 28b45e1d87 Add missing callbacks to the python module 2021-01-01 10:19:32 -07:00
Christopher Zimmermann ad51795314 Don't try to run daemon_privileged on reload. 2020-05-10 22:30:25 +02:00
Christopher Zimmermann 7d76e84953 Port ipset to BSD pf tables 2020-05-10 22:30:25 +02:00
Christopher Zimmermann c96e4ca121 allow privileged initialisation of modules 2020-05-10 22:30:25 +02:00
919 changed files with 104648 additions and 33603 deletions
+81 -18
View File
@@ -32,12 +32,12 @@ jobs:
- name: OS X
os: macos-latest
install_expat: "yes"
config: "--enable-debug --disable-flto --with-ssl=/usr/local/opt/openssl --with-libexpat=/usr/local/opt/expat"
config: "--enable-debug --disable-flto --with-ssl=/opt/homebrew/opt/openssl --with-libexpat=/opt/homebrew/opt/expat"
make_test: "yes"
- name: Clang on OS X
os: macos-latest
install_expat: "yes"
config: "CC=clang --enable-debug --disable-flto --with-ssl=/usr/local/opt/openssl --with-libexpat=/usr/local/opt/expat --disable-static"
config: "CC=clang --enable-debug --disable-flto --with-ssl=/opt/homebrew/opt/openssl --with-libexpat=/opt/homebrew/opt/expat --disable-static"
make_test: "yes"
clang_analysis: "yes"
- name: ubsan (gcc undefined behaviour sanitizer)
@@ -161,13 +161,44 @@ jobs:
test_windows: "yes"
config: "no"
make: "no"
- name: FreeBSD
os: ubuntu-latest
config: "no"
make: "no"
with_cross_platform_action: "yes"
cross_platform_os: "freebsd"
cross_platform_arch: "x86-64"
cross_platform_version: "14.1"
cross_platform_config: "--enable-debug --disable-flto --with-libevent --disable-static"
- name: OpenBSD
os: ubuntu-latest
config: "no"
make: "no"
with_cross_platform_action: "yes"
cross_platform_os: "openbsd"
cross_platform_arch: "x86-64"
cross_platform_version: "7.5"
cross_platform_config: "--enable-debug --disable-flto --with-libevent --disable-static"
- name: NetBSD
os: ubuntu-latest
config: "no"
make: "no"
with_cross_platform_action: "yes"
cross_platform_os: "netbsd"
cross_platform_arch: "x86-64"
cross_platform_version: "10.0"
cross_platform_config: "--enable-debug --disable-flto --with-libevent --disable-static"
steps:
- uses: actions/checkout@v2
- uses: actions/checkout@v4
with:
submodules: false
persist-credentials: false
- name: test_windows
if: ${{ matrix.test_windows == 'yes' }}
env:
LIBEXPAT_FNAME: expat-2.7.0
LIBEXPAT_VERSION_DIR: R_2_7_0
shell: bash
run: |
export unboundpath=`pwd`
@@ -175,21 +206,26 @@ jobs:
cd ..
export prepath=`pwd`
echo prepath=${prepath}
#echo "curl cpanm"
#curl -L -k -s -S -o cpanm https://cpanmin.us/
#echo "perl cpanm Pod::Usage"
#perl cpanm Pod::Usage
echo "choco install winflexbison3"
choco install winflexbison3
echo 'LEX="win_flex"'
export LEX="win_flex"
echo 'YACC="win_bison -y"'
export YACC="win_bison -y"
mkdir openssl
echo "curl openssl"
curl -L -k -s -S -o openssl-1.1.1j.tar.gz https://www.openssl.org/source/openssl-1.1.1j.tar.gz
tar xzf openssl-1.1.1j.tar.gz
cd openssl-1.1.1j
curl -L -k -s -S -o openssl-3.2.0.tar.gz https://www.openssl.org/source/openssl-3.2.0.tar.gz
tar xzf openssl-3.2.0.tar.gz
cd openssl-3.2.0
# remove pod::Usage because we do not need -help or -man output
# from the Configure script
echo "Fixup ./Configure by removing use Pod::Usage require"
sed -e 's/use Pod::Usage//' < Configure > Configure.fix
echo "./Configure.fix no-shared no-asm -DOPENSSL_NO_CAPIENG mingw64 --prefix=\""$prepath/openssl\"""
./Configure.fix no-shared no-asm -DOPENSSL_NO_CAPIENG mingw64 --prefix="$prepath/openssl"
# from the Configure script, this fixes it for openssl 1.1.1
#echo "Fixup ./Configure by removing use Pod::Usage require"
#sed -e 's/use Pod::Usage//' < Configure > Configure.fix
# ./Configure.fix no-shared no-asm -DOPENSSL_NO_CAPIENG mingw64 --prefix="$prepath/openssl"
# pacman is used to install for msys2, with
# C:/msys64/usr/bin/pacman -S perl
echo "C:/msys64/usr/bin/perl ./Configure no-shared no-asm -DOPENSSL_NO_CAPIENG mingw64 --prefix=\""$prepath/openssl"\" PERL=\"C:/msys64/usr/bin/perl\""
C:/msys64/usr/bin/perl ./Configure no-shared no-asm -DOPENSSL_NO_CAPIENG mingw64 --prefix="$prepath/openssl" PERL="C:/msys64/usr/bin/perl"
# make the libs only, build faster
echo "make build_libs"
#make
@@ -205,9 +241,9 @@ jobs:
cd ..
mkdir expat
echo "curl expat"
curl -L -k -s -S -o expat-2.2.10.tar.gz https://github.com/libexpat/libexpat/releases/download/R_2_2_10/expat-2.2.10.tar.gz
tar xzf expat-2.2.10.tar.gz
cd expat-2.2.10
curl -L -k -s -S -o $LIBEXPAT_FNAME.tar.gz https://github.com/libexpat/libexpat/releases/download/$LIBEXPAT_VERSION_DIR/$LIBEXPAT_FNAME.tar.gz
tar xzf $LIBEXPAT_FNAME.tar.gz
cd $LIBEXPAT_FNAME
echo "./configure SHELL=/usr/bin/bash CONFIG_SHELL=/usr/bin/bash --prefix=\"$prepath/expat\" --exec-prefix=\"$prepath/expat\" --bindir=\"$prepath/expat/bin\" --includedir=\"$prepath/expat/include\" --mandir=\"$prepath/expat/man\" --libdir=\"$prepath/expat/lib\""
./configure SHELL=/usr/bin/bash CONFIG_SHELL=/usr/bin/bash --prefix="$prepath/expat" --exec-prefix="$prepath/expat" --bindir="$prepath/expat/bin" --includedir="$prepath/expat/include" --mandir="$prepath/expat/man" --libdir="$prepath/expat/lib"
# fixup SHELL is treated specially, but SHELZZ is not by make.
@@ -326,6 +362,33 @@ jobs:
echo "::group::make install"
make install
echo "::endgroup::"
- name: cross-platform-action on ${{ matrix.cross_platform_os }} ${{ matrix.cross_platform_version }}
if: ${{ matrix.with_cross_platform_action == 'yes' }}
uses: cross-platform-actions/action@v0.25.0
env:
CROSS_PLATFORM_OS: ${{ matrix.cross_platform_os }}
with:
environment_variables: CROSS_PLATFORM_OS
operating_system: ${{ matrix.cross_platform_os }}
architecture: ${{ matrix.cross_platform_arch }}
version: ${{ matrix.cross_platform_version }}
shell: bash
memory: 4G
cpu_count: 2
run: |
set -e -x
if test "$CROSS_PLATFORM_OS" = "freebsd"; then sudo pkg install -y openssl libevent expat; fi
if test "$CROSS_PLATFORM_OS" = "openbsd"; then sudo pkg_add libevent; fi
if test "$CROSS_PLATFORM_OS" = "netbsd"; then sudo pkgin -y install libevent; fi
echo "::group::configure"
./configure ${{ matrix.cross_platform_config }}
echo "::endgroup::"
echo "::group::make"
make
echo "::endgroup::"
echo "::group::make test"
make test
echo "::endgroup::"
- name: install libevent
if: ${{ matrix.install_libevent == 'yes' }}
run: sudo apt-get install libevent-dev
+3 -1
View File
@@ -12,7 +12,9 @@ jobs:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v2
- uses: actions/checkout@v4
with:
persist-credentials: false
- name: configure
run: ./configure --enable-debug
- name: make
+6
View File
@@ -10,6 +10,10 @@
/config.status
/dnstap/dnstap_config.h
/dnscrypt/dnscrypt_config.h
/util/configlexer.c
/util/configparser.c
/util/configparser.h
/clubsyms.def
/doc/example.conf
/doc/libunbound.3
/doc/unbound-anchor.8
@@ -32,6 +36,7 @@
/asynclook
/delayer
/dohclient
/doqclient
/lock-verify
/memstats
/perf
@@ -54,6 +59,7 @@
/pythonmod/unboundmodule.py
/testdata/result.*
/testdata/.done-*
/testdata/.skip-*
/testdata/.perfstats.txt
/doc/html
/doc/xml
+861 -572
View File
File diff suppressed because it is too large Load Diff
+6 -3
View File
@@ -1,9 +1,10 @@
# Unbound
[![Travis Build Status](https://travis-ci.org/NLnetLabs/unbound.svg?branch=master)](https://travis-ci.org/NLnetLabs/unbound)
[![Github Build Status](https://github.com/NLnetLabs/unbound/actions/workflows/ci.yml/badge.svg?branch=master)](https://github.com/NLnetLabs/unbound/actions)
[![Packaging status](https://repology.org/badge/tiny-repos/unbound.svg)](https://repology.org/project/unbound/versions)
[![Fuzzing Status](https://oss-fuzz-build-logs.storage.googleapis.com/badges/unbound.svg)](https://bugs.chromium.org/p/oss-fuzz/issues/list?sort=-opened&can=1&q=proj:unbound)
[![Documentation Status](https://readthedocs.org/projects/unbound/badge/?version=latest)](https://unbound.readthedocs.io/en/latest/?badge=latest)
[![Mastodon Follow](https://img.shields.io/mastodon/follow/109262826617293067?domain=https%3A%2F%2Ffosstodon.org&style=social)](https://fosstodon.org/@nlnetlabs)
Unbound is a validating, recursive, caching DNS resolver. It is designed to be
fast and lean and incorporates modern features based on open standards. If you
@@ -16,7 +17,9 @@ You can learn more about Unbound by reading our
## Compiling
Make sure you have the C toolchain, OpenSSL and its include files, and libexpat
installed. Unbound can be compiled and installed using:
installed.
If building from the repository source you also need flex and bison installed.
Unbound can be compiled and installed using:
```
./configure && make && make install
@@ -26,7 +29,7 @@ You can use libevent if you want. libevent is useful when using many (10000)
outgoing ports. By default max 256 ports are opened at the same time and the
builtin alternative is equally capable and a little faster.
Use the `--with-libevent=dir` configure option to compile Unbound with libevent
Use the `--with-libevent` configure option to compile Unbound with libevent
support.
## Unbound configuration
+31
View File
@@ -0,0 +1,31 @@
# Security Policy
## Supported Versions
NLnet Labs adheres to the straightforward, semantic versioning scheme that is
commonly used in the software industry.
Support is provided in respect of the latest release, i.e. releases with the
highest minor and patch version level. We do not backport security fixes to
older (minor) versions. In the event a new major version is released (e.g. from
3.2.18 to 4.0.0), support will also be provided on the latest minor version of
the previous major version (3.2.18) for a period of one year from the release of
the new major version (4.0.0).
In the event that, during this period, a new patch or minor version of the
previous major version is released, then support on these versions will only be
provided for the remainder of the one-year-period.
You can find detailed information on our software support policy here:
https://www.nlnetlabs.nl/support/software-support-policy/
## Reporting a Vulnerability
We take security very seriously. If you have discovered a security vulnerability
in one of our projects and you would like to report it to us, you can send an
encrypted message to our Security Entry Point.
Details are described here:
https://www.nlnetlabs.nl/security-report/
+110 -49
View File
@@ -2,7 +2,13 @@
# Copyright 2009, Wouter Wijngaards, NLnet Labs.
# BSD licensed.
#
# Version 43
# Version 48
# 2024-01-16 fix to add -l:libssp.a to -lcrypto link check.
# and check for getaddrinfo with only header.
# 2024-01-15 fix to add crypt32 to -lcrypto link check when checking for gdi32.
# 2023-05-04 fix to remove unused whitespace.
# 2023-01-26 fix -Wstrict-prototypes.
# 2022-09-01 fix checking if nonblocking sockets work on OpenBSD.
# 2021-08-17 fix sed script in ssldir split handling.
# 2021-08-17 fix for openssl to detect split version, with ssldir_include
# and ssldir_lib output directories.
@@ -186,7 +192,7 @@ dnl cache=`echo $1 | sed 'y%.=/+- %___p__%'`
AC_CACHE_VAL(cv_prog_cc_flag_needed_$cache,
[
echo '$2' > conftest.c
echo 'void f(){}' >>conftest.c
echo 'void f(void){}' >>conftest.c
if test -z "`$CC $CPPFLAGS $CFLAGS $ERRFLAG -c conftest.c 2>&1`"; then
eval "cv_prog_cc_flag_needed_$cache=no"
else
@@ -232,7 +238,7 @@ dnl DEPFLAG: set to flag that generates dependencies.
AC_DEFUN([ACX_DEPFLAG],
[
AC_MSG_CHECKING([$CC dependency flag])
echo 'void f(){}' >conftest.c
echo 'void f(void){}' >conftest.c
if test "`$CC -MM conftest.c 2>&1`" = "conftest.o: conftest.c"; then
DEPFLAG="-MM"
else
@@ -271,7 +277,7 @@ ACX_CHECK_COMPILER_FLAG_NEEDED($C99FLAG -D__EXTENSIONS__ -D_BSD_SOURCE -D_DEFAUL
#include <getopt.h>
#endif
int test() {
int test(void) {
int a;
char **opts = NULL;
struct timeval tv;
@@ -308,7 +314,7 @@ ACX_CHECK_COMPILER_FLAG_NEEDED($C99FLAG -D__EXTENSIONS__ -D_BSD_SOURCE -D_DEFAUL
#include <getopt.h>
#endif
int test() {
int test(void) {
int a;
char **opts = NULL;
struct timeval tv;
@@ -334,7 +340,7 @@ ACX_CHECK_COMPILER_FLAG_NEEDED($C99FLAG,
[
#include <stdbool.h>
#include <ctype.h>
int test() {
int test(void) {
int a = 0;
return a;
}
@@ -344,7 +350,7 @@ ACX_CHECK_COMPILER_FLAG_NEEDED(-D_BSD_SOURCE -D_DEFAULT_SOURCE,
[
#include <ctype.h>
int test() {
int test(void) {
int a;
a = isascii(32);
return a;
@@ -355,7 +361,7 @@ ACX_CHECK_COMPILER_FLAG_NEEDED(-D_GNU_SOURCE,
[
#include <netinet/in.h>
int test() {
int test(void) {
struct in6_pktinfo inf;
int a = (int)sizeof(inf);
return a;
@@ -369,7 +375,7 @@ ACX_CHECK_COMPILER_FLAG_NEEDED(-D_GNU_SOURCE -D_FRSRESGID,
[
#include <unistd.h>
int test() {
int test(void) {
int a = setresgid(0,0,0);
a = setresuid(0,0,0);
return a;
@@ -384,7 +390,7 @@ ACX_CHECK_COMPILER_FLAG_NEEDED(-D_POSIX_C_SOURCE=200112,
#endif
#include <netdb.h>
int test() {
int test(void) {
int a = 0;
char *t;
time_t time = 0;
@@ -412,7 +418,7 @@ ACX_CHECK_COMPILER_FLAG_NEEDED(-D__EXTENSIONS__,
#include <getopt.h>
#endif
int test() {
int test(void) {
int a;
char **opts = NULL;
struct timeval tv;
@@ -474,7 +480,7 @@ fi
dnl Setup ATTR_FORMAT config.h parts.
dnl make sure you call ACX_CHECK_FORMAT_ATTRIBUTE also.
AC_DEFUN([AHX_CONFIG_FORMAT_ATTRIBUTE],
[
[
#ifdef HAVE_ATTR_FORMAT
# define ATTR_FORMAT(archetype, string_index, first_to_check) \
__attribute__ ((format (archetype, string_index, first_to_check)))
@@ -704,7 +710,7 @@ AC_DEFUN([ACX_SSL_CHECKS], [
LIBSSL_LDFLAGS="$LIBSSL_LDFLAGS -L$ssldir_lib"
ACX_RUNTIME_PATH_ADD([$ssldir_lib])
fi
AC_MSG_CHECKING([for EVP_sha256 in -lcrypto])
LIBS="$LIBS -lcrypto"
LIBSSL_LIBS="$LIBSSL_LIBS -lcrypto"
@@ -729,40 +735,73 @@ AC_DEFUN([ACX_SSL_CHECKS], [
]])],[
AC_DEFINE([HAVE_EVP_SHA256], 1,
[If you have EVP_sha256])
AC_MSG_RESULT(yes)
AC_MSG_RESULT(yes)
],[
AC_MSG_RESULT(no)
LIBS="$BAKLIBS"
LIBSSL_LIBS="$BAKSSLLIBS"
LIBS="$LIBS -ldl"
LIBSSL_LIBS="$LIBSSL_LIBS -ldl"
AC_MSG_CHECKING([if -lcrypto needs -ldl])
AC_LINK_IFELSE([AC_LANG_PROGRAM([[]], [[
int EVP_sha256(void);
(void)EVP_sha256();
]])],[
AC_DEFINE([HAVE_EVP_SHA256], 1,
[If you have EVP_sha256])
AC_MSG_RESULT(yes)
],[
AC_MSG_RESULT(no)
LIBS="$BAKLIBS"
LIBSSL_LIBS="$BAKSSLLIBS"
LIBS="$LIBS -ldl -pthread"
LIBSSL_LIBS="$LIBSSL_LIBS -ldl -pthread"
AC_MSG_CHECKING([if -lcrypto needs -ldl -pthread])
AC_LINK_IFELSE([AC_LANG_PROGRAM([[]], [[
int EVP_sha256(void);
(void)EVP_sha256();
]])],[
AC_DEFINE([HAVE_EVP_SHA256], 1,
[If you have EVP_sha256])
AC_MSG_RESULT(yes)
],[
AC_MSG_RESULT(no)
AC_MSG_ERROR([OpenSSL found in $ssldir, but version 0.9.7 or higher is required])
LIBS="$LIBS -lgdi32 -lws2_32 -lcrypt32"
LIBSSL_LIBS="$LIBSSL_LIBS -lgdi32 -lws2_32 -lcrypt32"
AC_MSG_CHECKING([if -lcrypto needs -lgdi32 -lws2_32 -lcrypt32])
AC_LINK_IFELSE([AC_LANG_PROGRAM([[]], [[
int EVP_sha256(void);
(void)EVP_sha256();
]])],[
AC_DEFINE([HAVE_EVP_SHA256], 1,
[If you have EVP_sha256])
AC_MSG_RESULT(yes)
],[
AC_MSG_RESULT(no)
LIBS="$BAKLIBS"
LIBSSL_LIBS="$BAKSSLLIBS"
LIBS="$LIBS -lgdi32 -lws2_32 -lcrypt32 -l:libssp.a"
LIBSSL_LIBS="$LIBSSL_LIBS -lgdi32 -lws2_32 -lcrypt32 -l:libssp.a"
AC_MSG_CHECKING([if -lcrypto needs -lgdi32 -lws2_32 -lcrypt32 -l:libssp.a])
AC_LINK_IFELSE([AC_LANG_PROGRAM([[]], [[
int EVP_sha256(void);
(void)EVP_sha256();
]])],[
AC_DEFINE([HAVE_EVP_SHA256], 1,
[If you have EVP_sha256])
AC_MSG_RESULT(yes)
],[
AC_MSG_RESULT(no)
LIBS="$BAKLIBS"
LIBSSL_LIBS="$BAKSSLLIBS"
LIBS="$LIBS -ldl"
LIBSSL_LIBS="$LIBSSL_LIBS -ldl"
AC_MSG_CHECKING([if -lcrypto needs -ldl])
AC_LINK_IFELSE([AC_LANG_PROGRAM([[]], [[
int EVP_sha256(void);
(void)EVP_sha256();
]])],[
AC_DEFINE([HAVE_EVP_SHA256], 1,
[If you have EVP_sha256])
AC_MSG_RESULT(yes)
],[
AC_MSG_RESULT(no)
LIBS="$BAKLIBS"
LIBSSL_LIBS="$BAKSSLLIBS"
LIBS="$LIBS -ldl -pthread"
LIBSSL_LIBS="$LIBSSL_LIBS -ldl -pthread"
AC_MSG_CHECKING([if -lcrypto needs -ldl -pthread])
AC_LINK_IFELSE([AC_LANG_PROGRAM([[]], [[
int EVP_sha256(void);
(void)EVP_sha256();
]])],[
AC_DEFINE([HAVE_EVP_SHA256], 1,
[If you have EVP_sha256])
AC_MSG_RESULT(yes)
],[
AC_MSG_RESULT(no)
AC_MSG_ERROR([OpenSSL found in $ssldir, but version 0.9.7 or higher is required])
])
])
])
])
])
])
])
fi
@@ -776,7 +815,7 @@ AC_CHECK_HEADERS([openssl/rand.h],,, [AC_INCLUDES_DEFAULT])
dnl Check for SSL, where SSL is mandatory
dnl Adds --with-ssl option, searches for openssl and defines HAVE_SSL if found
dnl Setup of CPPFLAGS, CFLAGS. Adds -lcrypto to LIBS.
dnl Setup of CPPFLAGS, CFLAGS. Adds -lcrypto to LIBS.
dnl Checks main header files of SSL.
dnl
AC_DEFUN([ACX_WITH_SSL],
@@ -833,7 +872,7 @@ dnl try to see if an additional _LARGEFILE_SOURCE 1 is needed to get fseeko
ACX_CHECK_COMPILER_FLAG_NEEDED(-D_LARGEFILE_SOURCE=1,
[
#include <stdio.h>
int test() {
int test(void) {
int a = fseeko(stdin, 0, 0);
return a;
}
@@ -858,7 +897,7 @@ char* (*f) () = getaddrinfo;
#ifdef __cplusplus
}
#endif
int main() {
int main(void) {
;
return 0;
}
@@ -869,7 +908,7 @@ dnl see if on windows
if test "$ac_cv_header_windows_h" = "yes"; then
AC_DEFINE(USE_WINSOCK, 1, [Whether the windows socket API is used])
USE_WINSOCK="1"
if echo $LIBS | grep 'lws2_32' >/dev/null; then
if echo "$LIBS" | grep 'lws2_32' >/dev/null; then
:
else
LIBS="$LIBS -lws2_32"
@@ -877,6 +916,24 @@ if test "$ac_cv_header_windows_h" = "yes"; then
fi
],
dnl no quick getaddrinfo, try mingw32 and winsock2 library.
dnl perhaps getaddrinfo needs only the include
AC_LINK_IFELSE(
[AC_LANG_PROGRAM(
[
#ifdef HAVE_WS2TCPIP_H
#include <ws2tcpip.h>
#endif
],
[
(void)getaddrinfo(NULL, NULL, NULL, NULL);
]
)],
[
ac_cv_func_getaddrinfo="yes"
AC_DEFINE(USE_WINSOCK, 1, [Whether the windows socket API is used])
USE_WINSOCK="1"
],
ORIGLIBS="$LIBS"
LIBS="$LIBS -lws2_32"
AC_LINK_IFELSE(
@@ -901,6 +958,7 @@ ac_cv_func_getaddrinfo="no"
LIBS="$ORIGLIBS"
])
)
)
AC_MSG_RESULT($ac_cv_func_getaddrinfo)
if test $ac_cv_func_getaddrinfo = yes; then
@@ -922,7 +980,7 @@ cache=`echo $1 | sed 'y%.=/+-%___p_%'`
AC_CACHE_VAL(cv_cc_deprecated_$cache,
[
echo '$3' >conftest.c
echo 'void f(){ $2 }' >>conftest.c
echo 'void f(void){ $2 }' >>conftest.c
if test -z "`$CC $CPPFLAGS $CFLAGS -c conftest.c 2>&1 | grep -e deprecated -e unavailable`"; then
eval "cv_cc_deprecated_$cache=no"
else
@@ -963,6 +1021,9 @@ AC_LANG_SOURCE([[
#ifdef HAVE_SYS_TYPES_H
#include <sys/types.h>
#endif
#ifdef HAVE_SYS_SELECT_H
#include <sys/select.h>
#endif
#ifdef HAVE_SYS_SOCKET_H
#include <sys/socket.h>
#endif
@@ -1313,7 +1374,7 @@ AC_DEFUN([AHX_CONFIG_W32_FD_SET_T],
#ifdef HAVE_WINSOCK2_H
#define FD_SET_T (u_int)
#else
#define FD_SET_T
#define FD_SET_T
#endif
])
@@ -1351,7 +1412,7 @@ dnl $3: define value, 1
AC_DEFUN([AHX_CONFIG_FLAG_OMITTED],
[#if defined($1) && !defined($2)
#define $2 $3
[#]endif ])
[#]endif])
dnl Wrapper for AHX_CONFIG_FLAG_OMITTED for -D style flags
dnl $1: the -DNAME or -DNAME=value string.
+70 -18
View File
@@ -9,7 +9,7 @@ AC_DEFUN([AC_PYTHON_DEVEL],[
AC_PATH_PROG([PYTHON],[python[$PYTHON_VERSION]])
if test -z "$PYTHON"; then
AC_MSG_ERROR([Cannot find python$PYTHON_VERSION in your system path])
AC_MSG_ERROR([Cannot find 'python$PYTHON_VERSION' in your system path. You can use the environment variable 'PYTHON_VERSION=version_number' for an explicit version.])
PYTHON_VERSION=""
fi
@@ -17,28 +17,75 @@ AC_DEFUN([AC_PYTHON_DEVEL],[
PYTHON_VERSION=`$PYTHON -c "import sys; \
print(sys.version.split()[[0]])"`
fi
# calculate the version number components.
[
v="$PYTHON_VERSION"
PYTHON_VERSION_MAJOR=`echo $v | sed 's/[^0-9].*//'`
if test -z "$PYTHON_VERSION_MAJOR"; then PYTHON_VERSION_MAJOR="0"; fi
v=`echo $v | sed -e 's/^[0-9]*$//' -e 's/[0-9]*[^0-9]//'`
PYTHON_VERSION_MINOR=`echo $v | sed 's/[^0-9].*//'`
if test -z "$PYTHON_VERSION_MINOR"; then PYTHON_VERSION_MINOR="0"; fi
v=`echo $v | sed -e 's/^[0-9]*$//' -e 's/[0-9]*[^0-9]//'`
PYTHON_VERSION_PATCH=`echo $v | sed 's/[^0-9].*//'`
if test -z "$PYTHON_VERSION_PATCH"; then PYTHON_VERSION_PATCH="0"; fi
]
#
# Check if you have distutils, else fail
#
AC_MSG_CHECKING([for the distutils Python package])
if ac_distutils_result=`$PYTHON -c "import distutils" 2>&1`; then
# For some systems, sysconfig exists, but has the wrong paths,
# on Debian 10, for python 2.7 and 3.7. So, we check the version,
# and for older versions try distutils.sysconfig first. For newer
# versions>=3.10, where distutils.sysconfig is deprecated, use
# sysconfig first and then attempt the other one.
py_distutils_first="no"
if test $PYTHON_VERSION_MAJOR -lt 3; then
py_distutils_first="yes"
fi
if test $PYTHON_VERSION_MAJOR -eq 3 -a $PYTHON_VERSION_MINOR -lt 10; then
py_distutils_first="yes"
fi
# Check if you have the first module
if test "$py_distutils_first" = "yes"; then m="distutils"; else m="sysconfig"; fi
sysconfig_module=""
AC_MSG_CHECKING([for the $m Python module])
if ac_modulecheck_result1=`$PYTHON -c "import $m" 2>&1`; then
AC_MSG_RESULT([yes])
else
sysconfig_module="$m"
else
AC_MSG_RESULT([no])
AC_MSG_ERROR([cannot import Python module "distutils".
Please check your Python installation. The error was:
$ac_distutils_result])
PYTHON_VERSION=""
fi
fi
# if not found, try the other one.
if test -z "$sysconfig_module"; then
if test "$py_distutils_first" = "yes"; then m2="sysconfig"; else m2="distutils"; fi
AC_MSG_CHECKING([for the $m2 Python module])
if ac_modulecheck_result2=`$PYTHON -c "import $m2" 2>&1`; then
AC_MSG_RESULT([yes])
sysconfig_module="$m2"
else
AC_MSG_RESULT([no])
AC_MSG_ERROR([cannot import Python module "$m", or "$m2".
Please check your Python installation. The errors are:
$m
$ac_modulecheck_result1
$m2
$ac_modulecheck_result2])
PYTHON_VERSION=""
fi
fi
if test "$sysconfig_module" = "distutils"; then sysconfig_module="distutils.sysconfig"; fi
#
# Check for Python include path
#
AC_MSG_CHECKING([for Python include path])
if test -z "$PYTHON_CPPFLAGS"; then
python_path=`$PYTHON -c "import distutils.sysconfig; \
print(distutils.sysconfig.get_python_inc());"`
if test "$sysconfig_module" = "sysconfig"; then
python_path=`$PYTHON -c 'import sysconfig; \
print(sysconfig.get_path("include"));'`
else
python_path=`$PYTHON -c "import distutils.sysconfig; \
print(distutils.sysconfig.get_python_inc());"`
fi
if test -n "${python_path}"; then
python_path="-I$python_path"
fi
@@ -52,14 +99,14 @@ $ac_distutils_result])
#
AC_MSG_CHECKING([for Python library path])
if test -z "$PYTHON_LDFLAGS"; then
PYTHON_LDFLAGS=`$PYTHON -c "from distutils.sysconfig import *; \
PYTHON_LDFLAGS=`$PYTHON -c "from $sysconfig_module import *; \
print('-L'+get_config_var('LIBDIR')+' -L'+get_config_var('LIBDEST')+' '+get_config_var('BLDLIBRARY'));"`
fi
AC_MSG_RESULT([$PYTHON_LDFLAGS])
AC_SUBST([PYTHON_LDFLAGS])
if test -z "$PYTHON_LIBDIR"; then
PYTHON_LIBDIR=`$PYTHON -c "from distutils.sysconfig import *; \
PYTHON_LIBDIR=`$PYTHON -c "from $sysconfig_module import *; \
print(get_config_var('LIBDIR'));"`
fi
@@ -68,8 +115,13 @@ $ac_distutils_result])
#
AC_MSG_CHECKING([for Python site-packages path])
if test -z "$PYTHON_SITE_PKG"; then
PYTHON_SITE_PKG=`$PYTHON -c "import distutils.sysconfig; \
print(distutils.sysconfig.get_python_lib(1,0));"`
if test "$sysconfig_module" = "sysconfig"; then
PYTHON_SITE_PKG=`$PYTHON -c 'import sysconfig; \
print(sysconfig.get_path("platlib"));'`
else
PYTHON_SITE_PKG=`$PYTHON -c "import distutils.sysconfig; \
print(distutils.sysconfig.get_python_lib(1,0));"`
fi
fi
AC_MSG_RESULT([$PYTHON_SITE_PKG])
AC_SUBST([PYTHON_SITE_PKG])
+70
View File
@@ -0,0 +1,70 @@
# ===========================================================================
# https://www.gnu.org/software/autoconf-archive/ax_build_date_epoch.html
# ===========================================================================
#
# SYNOPSIS
#
# AX_BUILD_DATE_EPOCH(VARIABLE[, FORMAT[, ACTION-IF-FAIL]])
#
# DESCRIPTION
#
# Sets VARIABLE to a string representing the current time. It is
# formatted according to FORMAT if specified, otherwise it is formatted as
# the number of seconds (excluding leap seconds) since the UNIX epoch (01
# Jan 1970 00:00:00 UTC).
#
# If the SOURCE_DATE_EPOCH environment variable is set, it uses the value
# of that variable instead of the current time. See
# https://reproducible-builds.org/specs/source-date-epoch). If
# SOURCE_DATE_EPOCH is set but cannot be properly interpreted as a UNIX
# timestamp, then execute ACTION-IF-FAIL if specified, otherwise error.
#
# VARIABLE is AC_SUBST-ed.
#
# LICENSE
#
# Copyright (c) 2016 Eric Bavier <bavier@member.fsf.org>
#
# This program is free software: you can redistribute it and/or modify it
# under the terms of the GNU General Public License as published by the
# Free Software Foundation, either version 3 of the License, or (at your
# option) any later version.
#
# This program is distributed in the hope that it will be useful, but
# WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General
# Public License for more details.
#
# You should have received a copy of the GNU General Public License along
# with this program. If not, see <https://www.gnu.org/licenses/>.
#
# As a special exception, the respective Autoconf Macro's copyright owner
# gives unlimited permission to copy, distribute and modify the configure
# scripts that are the output of Autoconf when processing the Macro. You
# need not follow the terms of the GNU General Public License when using
# or distributing such scripts, even though portions of the text of the
# Macro appear in them. The GNU General Public License (GPL) does govern
# all other use of the material that constitutes the Autoconf Macro.
#
# This special exception to the GPL applies to versions of the Autoconf
# Macro released by the Autoconf Archive. When you make and distribute a
# modified version of the Autoconf Macro, you may extend this special
# exception to the GPL to apply to your modified version as well.
#serial 2
AC_DEFUN([AX_BUILD_DATE_EPOCH],
[dnl
AC_MSG_CHECKING([for build time])
ax_date_fmt="m4_default($2,%s)"
AS_IF([test x"$SOURCE_DATE_EPOCH" = x],
[$1=`date "+$ax_date_fmt"`],
[ax_build_date=`date -u -d "@$SOURCE_DATE_EPOCH" "+$ax_date_fmt" 2>/dev/null \
|| date -u -r "$SOURCE_DATE_EPOCH" "+$ax_date_fmt" 2>/dev/null`
AS_IF([test x"$ax_build_date" = x],
[m4_ifval([$3],
[$3],
[AC_MSG_ERROR([malformed SOURCE_DATE_EPOCH])])],
[$1=$ax_build_date])])
AC_MSG_RESULT([$$1])
])dnl AX_BUILD_DATE_EPOCH
+58 -52
View File
@@ -1,39 +1,43 @@
# ===========================================================================
# http://autoconf-archive.cryp.to/ac_pkg_swig.html
# https://www.gnu.org/software/autoconf-archive/ax_pkg_swig.html
# ===========================================================================
#
# SYNOPSIS
#
# AC_PROG_SWIG([major.minor.micro])
# AX_PKG_SWIG([major.minor.micro], [action-if-found], [action-if-not-found])
#
# DESCRIPTION
#
# This macro searches for a SWIG installation on your system. If found you
# should call SWIG via $(SWIG). You can use the optional first argument to
# check if the version of the available SWIG is greater than or equal to
# the value of the argument. It should have the format: N[.N[.N]] (N is a
# number between 0 and 999. Only the first N is mandatory.)
# This macro searches for a SWIG installation on your system. If found,
# then SWIG is AC_SUBST'd; if not found, then $SWIG is empty. If SWIG is
# found, then SWIG_LIB is set to the SWIG library path, and AC_SUBST'd.
#
# If the version argument is given (e.g. 1.3.17), AC_PROG_SWIG checks that
# the swig package is this version number or higher.
# You can use the optional first argument to check if the version of the
# available SWIG is greater than or equal to the value of the argument. It
# should have the format: N[.N[.N]] (N is a number between 0 and 999. Only
# the first N is mandatory.) If the version argument is given (e.g.
# 1.3.17), AX_PKG_SWIG checks that the swig package is this version number
# or higher.
#
# As usual, action-if-found is executed if SWIG is found, otherwise
# action-if-not-found is executed.
#
# In configure.in, use as:
#
# AC_PROG_SWIG(1.3.17)
# SWIG_ENABLE_CXX
# SWIG_MULTI_MODULE_SUPPORT
# SWIG_PYTHON
# AX_PKG_SWIG(1.3.17, [], [ AC_MSG_ERROR([SWIG is required to build..]) ])
# AX_SWIG_ENABLE_CXX
# AX_SWIG_MULTI_MODULE_SUPPORT
# AX_SWIG_PYTHON
#
# LAST MODIFICATION
#
# 2008-04-12
#
# COPYLEFT
# LICENSE
#
# Copyright (c) 2008 Sebastian Huber <sebastian-huber@web.de>
# Copyright (c) 2008 Alan W. Irwin <irwin@beluga.phys.uvic.ca>
# Copyright (c) 2008 Alan W. Irwin
# Copyright (c) 2008 Rafael Laboissiere <rafael@laboissiere.net>
# Copyright (c) 2008 Andrew Collier <colliera@ukzn.ac.za>
# Copyright (c) 2008 Andrew Collier
# Copyright (c) 2011 Murray Cumming <murrayc@openismus.com>
# Copyright (c) 2018 Reini Urban <rurban@cpan.org>
# Copyright (c) 2021 Vincent Danjean <Vincent.Danjean@ens-lyon.org>
#
# This program is free software; you can redistribute it and/or modify it
# under the terms of the GNU General Public License as published by the
@@ -46,7 +50,7 @@
# Public License for more details.
#
# You should have received a copy of the GNU General Public License along
# with this program. If not, see <http://www.gnu.org/licenses/>.
# with this program. If not, see <https://www.gnu.org/licenses/>.
#
# As a special exception, the respective Autoconf Macro's copyright owner
# gives unlimited permission to copy, distribute and modify the configure
@@ -57,17 +61,21 @@
# all other use of the material that constitutes the Autoconf Macro.
#
# This special exception to the GPL applies to versions of the Autoconf
# Macro released by the Autoconf Macro Archive. When you make and
# distribute a modified version of the Autoconf Macro, you may extend this
# special exception to the GPL to apply to your modified version as well.
# Macro released by the Autoconf Archive. When you make and distribute a
# modified version of the Autoconf Macro, you may extend this special
# exception to the GPL to apply to your modified version as well.
AC_DEFUN([AC_PROG_SWIG],[
AC_PATH_PROG([SWIG],[swig])
#serial 15
AC_DEFUN([AX_PKG_SWIG],[
# Find path to the "swig" executable.
AC_PATH_PROGS([SWIG],[swig swig3.0 swig2.0])
if test -z "$SWIG" ; then
AC_MSG_WARN([cannot find 'swig' program. You should look at http://www.swig.org])
SWIG='echo "Error: SWIG is not installed. You should look at http://www.swig.org" ; false'
elif test -n "$1" ; then
AC_MSG_CHECKING([for SWIG version])
m4_ifval([$3],[$3],[:])
elif test -z "$1" ; then
m4_ifval([$2],[$2],[:])
else
AC_MSG_CHECKING([SWIG version])
[swig_version=`$SWIG -version 2>&1 | grep 'SWIG Version' | sed 's/.*\([0-9][0-9]*\.[0-9][0-9]*\.[0-9][0-9]*\).*/\1/g'`]
AC_MSG_RESULT([$swig_version])
if test -n "$swig_version" ; then
@@ -77,12 +85,12 @@ AC_DEFUN([AC_PROG_SWIG],[
if test -z "$required_major" ; then
[required_major=0]
fi
[required=`echo $required | sed 's/[0-9]*[^0-9]//'`]
[required=`echo $required. | sed 's/[0-9]*[^0-9]//'`]
[required_minor=`echo $required | sed 's/[^0-9].*//'`]
if test -z "$required_minor" ; then
[required_minor=0]
fi
[required=`echo $required | sed 's/[0-9]*[^0-9]//'`]
[required=`echo $required. | sed 's/[0-9]*[^0-9]//'`]
[required_patch=`echo $required | sed 's/[^0-9].*//'`]
if test -z "$required_patch" ; then
[required_patch=0]
@@ -103,30 +111,28 @@ AC_DEFUN([AC_PROG_SWIG],[
if test -z "$available_patch" ; then
[available_patch=0]
fi
[badversion=0]
if test $available_major -lt $required_major ; then
[badversion=1]
fi
if test $available_major -eq $required_major \
-a $available_minor -lt $required_minor ; then
[badversion=1]
fi
if test $available_major -eq $required_major \
-a $available_minor -eq $required_minor \
-a $available_patch -lt $required_patch ; then
[badversion=1]
fi
if test $badversion -eq 1 ; then
AC_MSG_WARN([SWIG version >= $1 is required. You have $swig_version. You should look at http://www.swig.org])
SWIG='echo "Error: SWIG version >= $1 is required. You have '"$swig_version"'. You should look at http://www.swig.org" ; false'
# Convert the version tuple into a single number for easier comparison.
# Using base 100 should be safe since SWIG internally uses BCD values
# to encode its version number.
required_swig_vernum=`expr $required_major \* 10000 \
\+ $required_minor \* 100 \+ $required_patch`
available_swig_vernum=`expr $available_major \* 10000 \
\+ $available_minor \* 100 \+ $available_patch`
if test $available_swig_vernum -lt $required_swig_vernum; then
AC_MSG_WARN([SWIG version >= $1 is required. You have $swig_version.])
SWIG=''
m4_ifval([$3],[$3],[])
else
AC_MSG_NOTICE([SWIG executable is '$SWIG'])
SWIG_LIB=`$SWIG -swiglib`
AC_MSG_NOTICE([SWIG library directory is '$SWIG_LIB'])
AC_MSG_CHECKING([for SWIG library])
SWIG_LIB=`$SWIG -swiglib | tr '\r\n' ' '`
AC_MSG_RESULT([$SWIG_LIB])
m4_ifval([$2],[$2],[])
fi
else
AC_MSG_WARN([cannot determine SWIG version])
SWIG='echo "Error: Cannot determine SWIG version. You should look at http://www.swig.org" ; false'
SWIG=''
m4_ifval([$3],[$3],[])
fi
fi
AC_SUBST([SWIG_LIB])
+317 -127
View File
@@ -1,5 +1,5 @@
# ===========================================================================
# http://www.gnu.org/software/autoconf-archive/ax_pthread.html
# https://www.gnu.org/software/autoconf-archive/ax_pthread.html
# ===========================================================================
#
# SYNOPSIS
@@ -14,24 +14,28 @@
# flags that are needed. (The user can also force certain compiler
# flags/libs to be tested by setting these environment variables.)
#
# Also sets PTHREAD_CC to any special C compiler that is needed for
# multi-threaded programs (defaults to the value of CC otherwise). (This
# is necessary on AIX to use the special cc_r compiler alias.)
# Also sets PTHREAD_CC and PTHREAD_CXX to any special C compiler that is
# needed for multi-threaded programs (defaults to the value of CC
# respectively CXX otherwise). (This is necessary on e.g. AIX to use the
# special cc_r/CC_r compiler alias.)
#
# NOTE: You are assumed to not only compile your program with these flags,
# but also link it with them as well. e.g. you should link with
# but also to link with them as well. For example, you might link with
# $PTHREAD_CC $CFLAGS $PTHREAD_CFLAGS $LDFLAGS ... $PTHREAD_LIBS $LIBS
# $PTHREAD_CXX $CXXFLAGS $PTHREAD_CFLAGS $LDFLAGS ... $PTHREAD_LIBS $LIBS
#
# If you are only building threads programs, you may wish to use these
# If you are only building threaded programs, you may wish to use these
# variables in your default LIBS, CFLAGS, and CC:
#
# LIBS="$PTHREAD_LIBS $LIBS"
# CFLAGS="$CFLAGS $PTHREAD_CFLAGS"
# CXXFLAGS="$CXXFLAGS $PTHREAD_CFLAGS"
# CC="$PTHREAD_CC"
# CXX="$PTHREAD_CXX"
#
# In addition, if the PTHREAD_CREATE_JOINABLE thread-attribute constant
# has a nonstandard name, defines PTHREAD_CREATE_JOINABLE to that name
# (e.g. PTHREAD_CREATE_UNDETACHED on AIX).
# has a nonstandard name, this macro defines PTHREAD_CREATE_JOINABLE to
# that name (e.g. PTHREAD_CREATE_UNDETACHED on AIX).
#
# Also HAVE_PTHREAD_PRIO_INHERIT is defined if pthread is found and the
# PTHREAD_PRIO_INHERIT symbol is defined when compiling with
@@ -55,6 +59,7 @@
#
# Copyright (c) 2008 Steven G. Johnson <stevenj@alum.mit.edu>
# Copyright (c) 2011 Daniel Richard G. <skunk@iSKUNK.ORG>
# Copyright (c) 2019 Marc Stevens <marc.stevens@cwi.nl>
#
# This program is free software: you can redistribute it and/or modify it
# under the terms of the GNU General Public License as published by the
@@ -67,7 +72,7 @@
# Public License for more details.
#
# You should have received a copy of the GNU General Public License along
# with this program. If not, see <http://www.gnu.org/licenses/>.
# with this program. If not, see <https://www.gnu.org/licenses/>.
#
# As a special exception, the respective Autoconf Macro's copyright owner
# gives unlimited permission to copy, distribute and modify the configure
@@ -82,35 +87,41 @@
# modified version of the Autoconf Macro, you may extend this special
# exception to the GPL to apply to your modified version as well.
#serial 21
#serial 31
AU_ALIAS([ACX_PTHREAD], [AX_PTHREAD])
AC_DEFUN([AX_PTHREAD], [
AC_REQUIRE([AC_CANONICAL_HOST])
AC_REQUIRE([AC_PROG_CC])
AC_REQUIRE([AC_PROG_SED])
AC_LANG_PUSH([C])
ax_pthread_ok=no
# We used to check for pthread.h first, but this fails if pthread.h
# requires special compiler flags (e.g. on True64 or Sequent).
# requires special compiler flags (e.g. on Tru64 or Sequent).
# It gets checked for in the link test anyway.
# First of all, check if the user has set any of the PTHREAD_LIBS,
# etcetera environment variables, and if threads linking works using
# them:
if test x"$PTHREAD_LIBS$PTHREAD_CFLAGS" != x; then
save_CFLAGS="$CFLAGS"
if test "x$PTHREAD_CFLAGS$PTHREAD_LIBS" != "x"; then
ax_pthread_save_CC="$CC"
ax_pthread_save_CFLAGS="$CFLAGS"
ax_pthread_save_LIBS="$LIBS"
AS_IF([test "x$PTHREAD_CC" != "x"], [CC="$PTHREAD_CC"])
AS_IF([test "x$PTHREAD_CXX" != "x"], [CXX="$PTHREAD_CXX"])
CFLAGS="$CFLAGS $PTHREAD_CFLAGS"
save_LIBS="$LIBS"
LIBS="$PTHREAD_LIBS $LIBS"
AC_MSG_CHECKING([for pthread_join in LIBS=$PTHREAD_LIBS with CFLAGS=$PTHREAD_CFLAGS])
AC_TRY_LINK_FUNC([pthread_join], [ax_pthread_ok=yes])
AC_MSG_CHECKING([for pthread_join using $CC $PTHREAD_CFLAGS $PTHREAD_LIBS])
AC_LINK_IFELSE([AC_LANG_CALL([], [pthread_join])], [ax_pthread_ok=yes])
AC_MSG_RESULT([$ax_pthread_ok])
if test x"$ax_pthread_ok" = xno; then
if test "x$ax_pthread_ok" = "xno"; then
PTHREAD_LIBS=""
PTHREAD_CFLAGS=""
fi
LIBS="$save_LIBS"
CFLAGS="$save_CFLAGS"
CC="$ax_pthread_save_CC"
CFLAGS="$ax_pthread_save_CFLAGS"
LIBS="$ax_pthread_save_LIBS"
fi
# We must check for the threads library under a number of different
@@ -118,12 +129,14 @@ fi
# (e.g. DEC) have both -lpthread and -lpthreads, where one of the
# libraries is broken (non-POSIX).
# Create a list of thread flags to try. Items starting with a "-" are
# C compiler flags, and other items are library names, except for "none"
# which indicates that we try without any flags at all, and "pthread-config"
# which is a program returning the flags for the Pth emulation library.
# Create a list of thread flags to try. Items with a "," contain both
# C compiler flags (before ",") and linker flags (after ","). Other items
# starting with a "-" are C compiler flags, and remaining items are
# library names, except for "none" which indicates that we try without
# any flags at all, and "pthread-config" which is a program returning
# the flags for the Pth emulation library.
ax_pthread_flags="pthreads none -Kthread -kthread lthread -pthread -pthreads -mthreads pthread --thread-safe -mt pthread-config"
ax_pthread_flags="pthreads none -Kthread -pthread -pthreads -mthreads pthread --thread-safe -mt pthread-config"
# The ordering *is* (sometimes) important. Some notes on the
# individual items follow:
@@ -132,82 +145,163 @@ ax_pthread_flags="pthreads none -Kthread -kthread lthread -pthread -pthreads -mt
# none: in case threads are in libc; should be tried before -Kthread and
# other compiler flags to prevent continual compiler warnings
# -Kthread: Sequent (threads in libc, but -Kthread needed for pthread.h)
# -kthread: FreeBSD kernel threads (preferred to -pthread since SMP-able)
# lthread: LinuxThreads port on FreeBSD (also preferred to -pthread)
# -pthread: Linux/gcc (kernel threads), BSD/gcc (userland threads)
# -pthreads: Solaris/gcc
# -mthreads: Mingw32/gcc, Lynx/gcc
# -pthread: Linux/gcc (kernel threads), BSD/gcc (userland threads), Tru64
# (Note: HP C rejects this with "bad form for `-t' option")
# -pthreads: Solaris/gcc (Note: HP C also rejects)
# -mt: Sun Workshop C (may only link SunOS threads [-lthread], but it
# doesn't hurt to check since this sometimes defines pthreads too;
# also defines -D_REENTRANT)
# ... -mt is also the pthreads flag for HP/aCC
# doesn't hurt to check since this sometimes defines pthreads and
# -D_REENTRANT too), HP C (must be checked before -lpthread, which
# is present but should not be used directly; and before -mthreads,
# because the compiler interprets this as "-mt" + "-hreads")
# -mthreads: Mingw32/gcc, Lynx/gcc
# pthread: Linux, etcetera
# --thread-safe: KAI C++
# pthread-config: use pthread-config program (for GNU Pth library)
case ${host_os} in
case $host_os in
freebsd*)
# -kthread: FreeBSD kernel threads (preferred to -pthread since SMP-able)
# lthread: LinuxThreads port on FreeBSD (also preferred to -pthread)
ax_pthread_flags="-kthread lthread $ax_pthread_flags"
;;
hpux*)
# From the cc(1) man page: "[-mt] Sets various -D flags to enable
# multi-threading and also sets -lpthread."
ax_pthread_flags="-mt -pthread pthread $ax_pthread_flags"
;;
openedition*)
# IBM z/OS requires a feature-test macro to be defined in order to
# enable POSIX threads at all, so give the user a hint if this is
# not set. (We don't define these ourselves, as they can affect
# other portions of the system API in unpredictable ways.)
AC_EGREP_CPP([AX_PTHREAD_ZOS_MISSING],
[
# if !defined(_OPEN_THREADS) && !defined(_UNIX03_THREADS)
AX_PTHREAD_ZOS_MISSING
# endif
],
[AC_MSG_WARN([IBM z/OS requires -D_OPEN_THREADS or -D_UNIX03_THREADS to enable pthreads support.])])
;;
solaris*)
# On Solaris (at least, for some versions), libc contains stubbed
# (non-functional) versions of the pthreads routines, so link-based
# tests will erroneously succeed. (We need to link with -pthreads/-mt/
# -lpthread.) (The stubs are missing pthread_cleanup_push, or rather
# a function called by this macro, so we could check for that, but
# who knows whether they'll stub that too in a future libc.) So,
# we'll just look for -pthreads and -lpthread first:
# tests will erroneously succeed. (N.B.: The stubs are missing
# pthread_cleanup_push, or rather a function called by this macro,
# so we could check for that, but who knows whether they'll stub
# that too in a future libc.) So we'll check first for the
# standard Solaris way of linking pthreads (-mt -lpthread).
ax_pthread_flags="-pthreads pthread -mt -pthread $ax_pthread_flags"
;;
darwin*)
ax_pthread_flags="-pthread $ax_pthread_flags"
ax_pthread_flags="-mt,-lpthread pthread $ax_pthread_flags"
;;
esac
# Clang doesn't consider unrecognized options an error unless we specify
# -Werror. We throw in some extra Clang-specific options to ensure that
# this doesn't happen for GCC, which also accepts -Werror.
# Are we compiling with Clang?
AC_MSG_CHECKING([if compiler needs -Werror to reject unknown flags])
save_CFLAGS="$CFLAGS"
ax_pthread_extra_flags="-Werror"
CFLAGS="$CFLAGS $ax_pthread_extra_flags -Wunknown-warning-option -Wsizeof-array-argument"
AC_COMPILE_IFELSE([AC_LANG_PROGRAM([int foo(void);],[foo()])],
[AC_MSG_RESULT([yes])],
[ax_pthread_extra_flags=
AC_MSG_RESULT([no])])
CFLAGS="$save_CFLAGS"
AC_CACHE_CHECK([whether $CC is Clang],
[ax_cv_PTHREAD_CLANG],
[ax_cv_PTHREAD_CLANG=no
# Note that Autoconf sets GCC=yes for Clang as well as GCC
if test "x$GCC" = "xyes"; then
AC_EGREP_CPP([AX_PTHREAD_CC_IS_CLANG],
[/* Note: Clang 2.7 lacks __clang_[a-z]+__ */
# if defined(__clang__) && defined(__llvm__)
AX_PTHREAD_CC_IS_CLANG
# endif
],
[ax_cv_PTHREAD_CLANG=yes])
fi
])
ax_pthread_clang="$ax_cv_PTHREAD_CLANG"
if test x"$ax_pthread_ok" = xno; then
for flag in $ax_pthread_flags; do
case $flag in
# GCC generally uses -pthread, or -pthreads on some platforms (e.g. SPARC)
# Note that for GCC and Clang -pthread generally implies -lpthread,
# except when -nostdlib is passed.
# This is problematic using libtool to build C++ shared libraries with pthread:
# [1] https://gcc.gnu.org/bugzilla/show_bug.cgi?id=25460
# [2] https://bugzilla.redhat.com/show_bug.cgi?id=661333
# [3] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=468555
# To solve this, first try -pthread together with -lpthread for GCC
AS_IF([test "x$GCC" = "xyes"],
[ax_pthread_flags="-pthread,-lpthread -pthread -pthreads $ax_pthread_flags"])
# Clang takes -pthread (never supported any other flag), but we'll try with -lpthread first
AS_IF([test "x$ax_pthread_clang" = "xyes"],
[ax_pthread_flags="-pthread,-lpthread -pthread"])
# The presence of a feature test macro requesting re-entrant function
# definitions is, on some systems, a strong hint that pthreads support is
# correctly enabled
case $host_os in
darwin* | hpux* | linux* | osf* | solaris*)
ax_pthread_check_macro="_REENTRANT"
;;
aix*)
ax_pthread_check_macro="_THREAD_SAFE"
;;
*)
ax_pthread_check_macro="--"
;;
esac
AS_IF([test "x$ax_pthread_check_macro" = "x--"],
[ax_pthread_check_cond=0],
[ax_pthread_check_cond="!defined($ax_pthread_check_macro)"])
if test "x$ax_pthread_ok" = "xno"; then
for ax_pthread_try_flag in $ax_pthread_flags; do
case $ax_pthread_try_flag in
none)
AC_MSG_CHECKING([whether pthreads work without any flags])
;;
*,*)
PTHREAD_CFLAGS=`echo $ax_pthread_try_flag | sed "s/^\(.*\),\(.*\)$/\1/"`
PTHREAD_LIBS=`echo $ax_pthread_try_flag | sed "s/^\(.*\),\(.*\)$/\2/"`
AC_MSG_CHECKING([whether pthreads work with "$PTHREAD_CFLAGS" and "$PTHREAD_LIBS"])
;;
-*)
AC_MSG_CHECKING([whether pthreads work with $flag])
PTHREAD_CFLAGS="$flag"
AC_MSG_CHECKING([whether pthreads work with $ax_pthread_try_flag])
PTHREAD_CFLAGS="$ax_pthread_try_flag"
;;
pthread-config)
AC_CHECK_PROG([ax_pthread_config], [pthread-config], [yes], [no])
if test x"$ax_pthread_config" = xno; then continue; fi
AS_IF([test "x$ax_pthread_config" = "xno"], [continue])
PTHREAD_CFLAGS="`pthread-config --cflags`"
PTHREAD_LIBS="`pthread-config --ldflags` `pthread-config --libs`"
;;
*)
AC_MSG_CHECKING([for the pthreads library -l$flag])
PTHREAD_LIBS="-l$flag"
AC_MSG_CHECKING([for the pthreads library -l$ax_pthread_try_flag])
PTHREAD_LIBS="-l$ax_pthread_try_flag"
;;
esac
save_LIBS="$LIBS"
save_CFLAGS="$CFLAGS"
ax_pthread_save_CFLAGS="$CFLAGS"
ax_pthread_save_LIBS="$LIBS"
CFLAGS="$CFLAGS $PTHREAD_CFLAGS"
LIBS="$PTHREAD_LIBS $LIBS"
CFLAGS="$CFLAGS $PTHREAD_CFLAGS $ax_pthread_extra_flags"
# Check for various functions. We must include pthread.h,
# since some functions may be macros. (On the Sequent, we
@@ -218,8 +312,18 @@ for flag in $ax_pthread_flags; do
# pthread_cleanup_push because it is one of the few pthread
# functions on Solaris that doesn't have a non-functional libc stub.
# We try pthread_create on general principles.
AC_LINK_IFELSE([AC_LANG_PROGRAM([#include <pthread.h>
static void routine(void *a) { *((int*)a) = 0; }
# if $ax_pthread_check_cond
# error "$ax_pthread_check_macro must be defined"
# endif
static void *some_global = NULL;
static void routine(void *a)
{
/* To avoid any unused-parameter or
unused-but-set-parameter warning. */
some_global = a;
}
static void *start_routine(void *a) { return a; }],
[pthread_t th; pthread_attr_t attr;
pthread_create(&th, 0, start_routine, 0);
@@ -227,101 +331,187 @@ for flag in $ax_pthread_flags; do
pthread_attr_init(&attr);
pthread_cleanup_push(routine, 0);
pthread_cleanup_pop(0) /* ; */])],
[ax_pthread_ok=yes],
[])
[ax_pthread_ok=yes],
[])
LIBS="$save_LIBS"
CFLAGS="$save_CFLAGS"
CFLAGS="$ax_pthread_save_CFLAGS"
LIBS="$ax_pthread_save_LIBS"
AC_MSG_RESULT([$ax_pthread_ok])
if test "x$ax_pthread_ok" = xyes; then
break;
fi
AS_IF([test "x$ax_pthread_ok" = "xyes"], [break])
PTHREAD_LIBS=""
PTHREAD_CFLAGS=""
done
fi
# Clang needs special handling, because older versions handle the -pthread
# option in a rather... idiosyncratic way
if test "x$ax_pthread_clang" = "xyes"; then
# Clang takes -pthread; it has never supported any other flag
# (Note 1: This will need to be revisited if a system that Clang
# supports has POSIX threads in a separate library. This tends not
# to be the way of modern systems, but it's conceivable.)
# (Note 2: On some systems, notably Darwin, -pthread is not needed
# to get POSIX threads support; the API is always present and
# active. We could reasonably leave PTHREAD_CFLAGS empty. But
# -pthread does define _REENTRANT, and while the Darwin headers
# ignore this macro, third-party headers might not.)
# However, older versions of Clang make a point of warning the user
# that, in an invocation where only linking and no compilation is
# taking place, the -pthread option has no effect ("argument unused
# during compilation"). They expect -pthread to be passed in only
# when source code is being compiled.
#
# Problem is, this is at odds with the way Automake and most other
# C build frameworks function, which is that the same flags used in
# compilation (CFLAGS) are also used in linking. Many systems
# supported by AX_PTHREAD require exactly this for POSIX threads
# support, and in fact it is often not straightforward to specify a
# flag that is used only in the compilation phase and not in
# linking. Such a scenario is extremely rare in practice.
#
# Even though use of the -pthread flag in linking would only print
# a warning, this can be a nuisance for well-run software projects
# that build with -Werror. So if the active version of Clang has
# this misfeature, we search for an option to squash it.
AC_CACHE_CHECK([whether Clang needs flag to prevent "argument unused" warning when linking with -pthread],
[ax_cv_PTHREAD_CLANG_NO_WARN_FLAG],
[ax_cv_PTHREAD_CLANG_NO_WARN_FLAG=unknown
# Create an alternate version of $ac_link that compiles and
# links in two steps (.c -> .o, .o -> exe) instead of one
# (.c -> exe), because the warning occurs only in the second
# step
ax_pthread_save_ac_link="$ac_link"
ax_pthread_sed='s/conftest\.\$ac_ext/conftest.$ac_objext/g'
ax_pthread_link_step=`AS_ECHO(["$ac_link"]) | sed "$ax_pthread_sed"`
ax_pthread_2step_ac_link="($ac_compile) && (echo ==== >&5) && ($ax_pthread_link_step)"
ax_pthread_save_CFLAGS="$CFLAGS"
for ax_pthread_try in '' -Qunused-arguments -Wno-unused-command-line-argument unknown; do
AS_IF([test "x$ax_pthread_try" = "xunknown"], [break])
CFLAGS="-Werror -Wunknown-warning-option $ax_pthread_try -pthread $ax_pthread_save_CFLAGS"
ac_link="$ax_pthread_save_ac_link"
AC_LINK_IFELSE([AC_LANG_SOURCE([[int main(void){return 0;}]])],
[ac_link="$ax_pthread_2step_ac_link"
AC_LINK_IFELSE([AC_LANG_SOURCE([[int main(void){return 0;}]])],
[break])
])
done
ac_link="$ax_pthread_save_ac_link"
CFLAGS="$ax_pthread_save_CFLAGS"
AS_IF([test "x$ax_pthread_try" = "x"], [ax_pthread_try=no])
ax_cv_PTHREAD_CLANG_NO_WARN_FLAG="$ax_pthread_try"
])
case "$ax_cv_PTHREAD_CLANG_NO_WARN_FLAG" in
no | unknown) ;;
*) PTHREAD_CFLAGS="$ax_cv_PTHREAD_CLANG_NO_WARN_FLAG $PTHREAD_CFLAGS" ;;
esac
fi # $ax_pthread_clang = yes
# Various other checks:
if test "x$ax_pthread_ok" = xyes; then
save_LIBS="$LIBS"
LIBS="$PTHREAD_LIBS $LIBS"
save_CFLAGS="$CFLAGS"
if test "x$ax_pthread_ok" = "xyes"; then
ax_pthread_save_CFLAGS="$CFLAGS"
ax_pthread_save_LIBS="$LIBS"
CFLAGS="$CFLAGS $PTHREAD_CFLAGS"
LIBS="$PTHREAD_LIBS $LIBS"
# Detect AIX lossage: JOINABLE attribute is called UNDETACHED.
AC_MSG_CHECKING([for joinable pthread attribute])
attr_name=unknown
for attr in PTHREAD_CREATE_JOINABLE PTHREAD_CREATE_UNDETACHED; do
AC_LINK_IFELSE([AC_LANG_PROGRAM([#include <pthread.h>],
[int attr = $attr; return attr /* ; */])],
[attr_name=$attr; break],
[])
done
AC_MSG_RESULT([$attr_name])
if test "$attr_name" != PTHREAD_CREATE_JOINABLE; then
AC_DEFINE_UNQUOTED([PTHREAD_CREATE_JOINABLE], [$attr_name],
[Define to necessary symbol if this constant
uses a non-standard name on your system.])
fi
AC_CACHE_CHECK([for joinable pthread attribute],
[ax_cv_PTHREAD_JOINABLE_ATTR],
[ax_cv_PTHREAD_JOINABLE_ATTR=unknown
for ax_pthread_attr in PTHREAD_CREATE_JOINABLE PTHREAD_CREATE_UNDETACHED; do
AC_LINK_IFELSE([AC_LANG_PROGRAM([#include <pthread.h>],
[int attr = $ax_pthread_attr; return attr /* ; */])],
[ax_cv_PTHREAD_JOINABLE_ATTR=$ax_pthread_attr; break],
[])
done
])
AS_IF([test "x$ax_cv_PTHREAD_JOINABLE_ATTR" != "xunknown" && \
test "x$ax_cv_PTHREAD_JOINABLE_ATTR" != "xPTHREAD_CREATE_JOINABLE" && \
test "x$ax_pthread_joinable_attr_defined" != "xyes"],
[AC_DEFINE_UNQUOTED([PTHREAD_CREATE_JOINABLE],
[$ax_cv_PTHREAD_JOINABLE_ATTR],
[Define to necessary symbol if this constant
uses a non-standard name on your system.])
ax_pthread_joinable_attr_defined=yes
])
AC_MSG_CHECKING([if more special flags are required for pthreads])
flag=no
case ${host_os} in
aix* | freebsd* | darwin*) flag="-D_THREAD_SAFE";;
osf* | hpux*) flag="-D_REENTRANT";;
solaris*)
if test "$GCC" = "yes"; then
flag="-D_REENTRANT"
else
# TODO: What about Clang on Solaris?
flag="-mt -D_REENTRANT"
fi
;;
esac
AC_MSG_RESULT([$flag])
if test "x$flag" != xno; then
PTHREAD_CFLAGS="$flag $PTHREAD_CFLAGS"
fi
AC_CACHE_CHECK([whether more special flags are required for pthreads],
[ax_cv_PTHREAD_SPECIAL_FLAGS],
[ax_cv_PTHREAD_SPECIAL_FLAGS=no
case $host_os in
solaris*)
ax_cv_PTHREAD_SPECIAL_FLAGS="-D_POSIX_PTHREAD_SEMANTICS"
;;
esac
])
AS_IF([test "x$ax_cv_PTHREAD_SPECIAL_FLAGS" != "xno" && \
test "x$ax_pthread_special_flags_added" != "xyes"],
[PTHREAD_CFLAGS="$ax_cv_PTHREAD_SPECIAL_FLAGS $PTHREAD_CFLAGS"
ax_pthread_special_flags_added=yes])
AC_CACHE_CHECK([for PTHREAD_PRIO_INHERIT],
[ax_cv_PTHREAD_PRIO_INHERIT], [
AC_LINK_IFELSE([AC_LANG_PROGRAM([[#include <pthread.h>]],
[[int i = PTHREAD_PRIO_INHERIT;]])],
[ax_cv_PTHREAD_PRIO_INHERIT=yes],
[ax_cv_PTHREAD_PRIO_INHERIT=no])
[ax_cv_PTHREAD_PRIO_INHERIT],
[AC_LINK_IFELSE([AC_LANG_PROGRAM([[#include <pthread.h>]],
[[int i = PTHREAD_PRIO_INHERIT;
return i;]])],
[ax_cv_PTHREAD_PRIO_INHERIT=yes],
[ax_cv_PTHREAD_PRIO_INHERIT=no])
])
AS_IF([test "x$ax_cv_PTHREAD_PRIO_INHERIT" = "xyes"],
[AC_DEFINE([HAVE_PTHREAD_PRIO_INHERIT], [1], [Have PTHREAD_PRIO_INHERIT.])])
AS_IF([test "x$ax_cv_PTHREAD_PRIO_INHERIT" = "xyes" && \
test "x$ax_pthread_prio_inherit_defined" != "xyes"],
[AC_DEFINE([HAVE_PTHREAD_PRIO_INHERIT], [1], [Have PTHREAD_PRIO_INHERIT.])
ax_pthread_prio_inherit_defined=yes
])
LIBS="$save_LIBS"
CFLAGS="$save_CFLAGS"
CFLAGS="$ax_pthread_save_CFLAGS"
LIBS="$ax_pthread_save_LIBS"
# More AIX lossage: compile with *_r variant
if test "x$GCC" != xyes; then
if test "x$GCC" != "xyes"; then
case $host_os in
aix*)
AS_CASE(["x/$CC"],
[x*/c89|x*/c89_128|x*/c99|x*/c99_128|x*/cc|x*/cc128|x*/xlc|x*/xlc_v6|x*/xlc128|x*/xlc128_v6],
[#handle absolute path differently from PATH based program lookup
AS_CASE(["x$CC"],
[x/*],
[AS_IF([AS_EXECUTABLE_P([${CC}_r])],[PTHREAD_CC="${CC}_r"])],
[AC_CHECK_PROGS([PTHREAD_CC],[${CC}_r],[$CC])])])
[x*/c89|x*/c89_128|x*/c99|x*/c99_128|x*/cc|x*/cc128|x*/xlc|x*/xlc_v6|x*/xlc128|x*/xlc128_v6],
[#handle absolute path differently from PATH based program lookup
AS_CASE(["x$CC"],
[x/*],
[
AS_IF([AS_EXECUTABLE_P([${CC}_r])],[PTHREAD_CC="${CC}_r"])
AS_IF([test "x${CXX}" != "x"], [AS_IF([AS_EXECUTABLE_P([${CXX}_r])],[PTHREAD_CXX="${CXX}_r"])])
],
[
AC_CHECK_PROGS([PTHREAD_CC],[${CC}_r],[$CC])
AS_IF([test "x${CXX}" != "x"], [AC_CHECK_PROGS([PTHREAD_CXX],[${CXX}_r],[$CXX])])
]
)
])
;;
esac
fi
fi
test -n "$PTHREAD_CC" || PTHREAD_CC="$CC"
test -n "$PTHREAD_CXX" || PTHREAD_CXX="$CXX"
AC_SUBST([PTHREAD_LIBS])
AC_SUBST([PTHREAD_CFLAGS])
AC_SUBST([PTHREAD_CC])
AC_SUBST([PTHREAD_CXX])
# Finally, execute ACTION-IF-FOUND/ACTION-IF-NOT-FOUND:
if test x"$ax_pthread_ok" = xyes; then
if test "x$ax_pthread_ok" = "xyes"; then
ifelse([$1],,[AC_DEFINE([HAVE_PTHREAD],[1],[Define if you have POSIX threads libraries and header files.])],[$1])
:
else
+216 -45
View File
@@ -47,9 +47,12 @@
#include "util/regional.h"
#include "util/net_help.h"
#include "util/config_file.h"
#include "util/data/dname.h"
#include "util/data/msgreply.h"
#include "util/data/msgencode.h"
#include "services/cache/dns.h"
#include "services/mesh.h"
#include "services/modstack.h"
#include "validator/val_neg.h"
#include "validator/val_secalgo.h"
#include "iterator/iter_utils.h"
@@ -102,7 +105,6 @@ static int
testframe_init(struct module_env* env, struct cachedb_env* cachedb_env)
{
struct testframe_moddata* d;
(void)env;
verbose(VERB_ALGO, "testframe_init");
d = (struct testframe_moddata*)calloc(1,
sizeof(struct testframe_moddata));
@@ -111,6 +113,15 @@ testframe_init(struct module_env* env, struct cachedb_env* cachedb_env)
log_err("out of memory");
return 0;
}
/* Register an EDNS option (65534) to bypass the worker cache lookup
* for testing */
if(!edns_register_option(LDNS_EDNS_UNBOUND_CACHEDB_TESTFRAME_TEST,
1 /* bypass cache */,
0 /* no aggregation */, env)) {
log_err("testframe_init, could not register test opcode");
free(d);
return 0;
}
lock_basic_init(&d->lock);
lock_protect(&d->lock, d, sizeof(*d));
return 1;
@@ -218,6 +229,8 @@ static int
cachedb_apply_cfg(struct cachedb_env* cachedb_env, struct config_file* cfg)
{
const char* backend_str = cfg->cachedb_backend;
if(!backend_str || *backend_str==0)
return 1;
cachedb_env->backend = cachedb_find_backend(backend_str);
if(!cachedb_env->backend) {
log_err("cachedb: cannot find backend name '%s'", backend_str);
@@ -228,7 +241,7 @@ cachedb_apply_cfg(struct cachedb_env* cachedb_env, struct config_file* cfg)
return 1;
}
int
int
cachedb_init(struct module_env* env, int id)
{
struct cachedb_env* cachedb_env = (struct cachedb_env*)calloc(1,
@@ -255,31 +268,19 @@ cachedb_init(struct module_env* env, int id)
return 0;
}
cachedb_env->enabled = 1;
if(env->cfg->serve_expired_reply_ttl)
log_warn(
"cachedb: serve-expired-reply-ttl is set but not working for data "
"originating from the external cache; 0 TLL is used for those.");
if(env->cfg->serve_expired_client_timeout)
log_warn(
"cachedb: serve-expired-client-timeout is set but not working for "
"data originating from the external cache; expired data are used "
"in the reply without first trying to refresh the data.");
return 1;
}
void
void
cachedb_deinit(struct module_env* env, int id)
{
struct cachedb_env* cachedb_env;
if(!env || !env->modinfo[id])
return;
cachedb_env = (struct cachedb_env*)env->modinfo[id];
/* free contents */
/* TODO */
if(cachedb_env->enabled) {
(*cachedb_env->backend->deinit)(env, cachedb_env);
}
free(cachedb_env);
env->modinfo[id] = NULL;
}
@@ -322,30 +323,31 @@ error_response(struct module_qstate* qstate, int id, int rcode)
/**
* Hash the query name, type, class and dbacess-secret into lookup buffer.
* @param qstate: query state with query info
* and env->cfg with secret.
* @param qinfo: query info
* @param env: with env->cfg with secret.
* @param buf: returned buffer with hash to lookup
* @param len: length of the buffer.
*/
static void
calc_hash(struct module_qstate* qstate, char* buf, size_t len)
calc_hash(struct query_info* qinfo, struct module_env* env, char* buf,
size_t len)
{
uint8_t clear[1024];
size_t clen = 0;
uint8_t hash[CACHEDB_HASHSIZE/8];
const char* hex = "0123456789ABCDEF";
const char* secret = qstate->env->cfg->cachedb_secret;
const char* secret = env->cfg->cachedb_secret;
size_t i;
/* copy the hash info into the clear buffer */
if(clen + qstate->qinfo.qname_len < sizeof(clear)) {
memmove(clear+clen, qstate->qinfo.qname,
qstate->qinfo.qname_len);
clen += qstate->qinfo.qname_len;
if(clen + qinfo->qname_len < sizeof(clear)) {
memmove(clear+clen, qinfo->qname, qinfo->qname_len);
query_dname_tolower(clear+clen);
clen += qinfo->qname_len;
}
if(clen + 4 < sizeof(clear)) {
uint16_t t = htons(qstate->qinfo.qtype);
uint16_t c = htons(qstate->qinfo.qclass);
uint16_t t = htons(qinfo->qtype);
uint16_t c = htons(qinfo->qclass);
memmove(clear+clen, &t, 2);
memmove(clear+clen+2, &c, 2);
clen += 4;
@@ -390,6 +392,15 @@ prep_data(struct module_qstate* qstate, struct sldns_buffer* buf)
if(!qstate->return_msg || !qstate->return_msg->rep)
return 0;
/* do not store failures like SERVFAIL in the cachedb, this avoids
* overwriting expired, valid, content with broken content. */
if(FLAGS_GET_RCODE(qstate->return_msg->rep->flags) !=
LDNS_RCODE_NOERROR &&
FLAGS_GET_RCODE(qstate->return_msg->rep->flags) !=
LDNS_RCODE_NXDOMAIN &&
FLAGS_GET_RCODE(qstate->return_msg->rep->flags) !=
LDNS_RCODE_YXDOMAIN)
return 0;
/* We don't store the reply if its TTL is 0 unless serve-expired is
* enabled. Such a reply won't be reusable and simply be a waste for
* the backend. It's also compatible with the default behavior of
@@ -397,6 +408,14 @@ prep_data(struct module_qstate* qstate, struct sldns_buffer* buf)
if(qstate->return_msg->rep->ttl == 0 &&
!qstate->env->cfg->serve_expired)
return 0;
/* The EDE is added to the out-list so it is encoded in the cached message */
if (qstate->env->cfg->ede && qstate->return_msg->rep->reason_bogus != LDNS_EDE_NONE) {
edns_opt_list_append_ede(&edns.opt_list_out, qstate->env->scratch,
qstate->return_msg->rep->reason_bogus,
qstate->return_msg->rep->reason_bogus_str);
}
if(verbosity >= VERB_ALGO)
log_dns_msg("cachedb encoding", &qstate->return_msg->qinfo,
qstate->return_msg->rep);
@@ -487,12 +506,42 @@ adjust_msg_ttl(struct dns_msg* msg, time_t adjust)
}
}
/* Set the TTL of the given RRset to fixed value. */
static void
packed_rrset_ttl_set(struct packed_rrset_data* data, time_t ttl)
{
size_t i;
size_t total = data->count + data->rrsig_count;
data->ttl = ttl;
for(i=0; i<total; i++) {
data->rr_ttl[i] = ttl;
}
data->ttl_add = 0;
}
/* Set the TTL of a DNS message and its RRs by to a fixed value. */
static void
set_msg_ttl(struct dns_msg* msg, time_t ttl)
{
size_t i;
msg->rep->ttl = ttl;
msg->rep->prefetch_ttl = PREFETCH_TTL_CALC(msg->rep->ttl);
msg->rep->serve_expired_ttl = msg->rep->ttl + SERVE_EXPIRED_TTL;
for(i=0; i<msg->rep->rrset_count; i++) {
packed_rrset_ttl_set((struct packed_rrset_data*)msg->
rep->rrsets[i]->entry.data, ttl);
}
}
/** convert dns message in buffer to return_msg */
static int
parse_data(struct module_qstate* qstate, struct sldns_buffer* buf)
parse_data(struct module_qstate* qstate, struct sldns_buffer* buf,
int* msg_expired)
{
struct msg_parse* prs;
struct edns_data edns;
struct edns_option* ede;
uint64_t timestamp, expiry;
time_t adjust;
size_t lim = sldns_buffer_limit(buf);
@@ -530,6 +579,24 @@ parse_data(struct module_qstate* qstate, struct sldns_buffer* buf)
if(!qstate->return_msg)
return 0;
/* We find the EDE in the in-list after parsing */
if(qstate->env->cfg->ede &&
(ede = edns_opt_list_find(edns.opt_list_in, LDNS_EDNS_EDE))) {
if(ede->opt_len >= 2) {
qstate->return_msg->rep->reason_bogus =
sldns_read_uint16(ede->opt_data);
}
/* allocate space and store the error string and it's size */
if(ede->opt_len > 2) {
size_t ede_len = ede->opt_len - 2;
qstate->return_msg->rep->reason_bogus_str = regional_alloc(
qstate->region, sizeof(char) * (ede_len+1));
memcpy(qstate->return_msg->rep->reason_bogus_str,
ede->opt_data+2, ede_len);
qstate->return_msg->rep->reason_bogus_str[ede_len] = 0;
}
}
qstate->return_rcode = LDNS_RCODE_NOERROR;
/* see how much of the TTL expired, and remove it */
@@ -540,15 +607,25 @@ parse_data(struct module_qstate* qstate, struct sldns_buffer* buf)
adjust = *qstate->env->now - (time_t)timestamp;
if(qstate->return_msg->rep->ttl < adjust) {
verbose(VERB_ALGO, "cachedb msg expired");
*msg_expired = 1;
/* If serve-expired is enabled, we still use an expired message
* setting the TTL to 0. */
if(qstate->env->cfg->serve_expired)
adjust = -1;
else
if(!qstate->env->cfg->serve_expired ||
(FLAGS_GET_RCODE(qstate->return_msg->rep->flags)
!= LDNS_RCODE_NOERROR &&
FLAGS_GET_RCODE(qstate->return_msg->rep->flags)
!= LDNS_RCODE_NXDOMAIN &&
FLAGS_GET_RCODE(qstate->return_msg->rep->flags)
!= LDNS_RCODE_YXDOMAIN))
return 0; /* message expired */
else
adjust = -1;
}
verbose(VERB_ALGO, "cachedb msg adjusted down by %d", (int)adjust);
adjust_msg_ttl(qstate->return_msg, adjust);
if(qstate->env->cfg->aggressive_nsec) {
limit_nsec_ttl(qstate->return_msg);
}
/* Similar to the unbound worker, if serve-expired is enabled and
* the msg would be considered to be expired, mark the state so a
@@ -556,6 +633,7 @@ parse_data(struct module_qstate* qstate, struct sldns_buffer* buf)
* 'now' should be redundant given how these values were calculated,
* but we check it just in case as does good_expiry_and_qinfo(). */
if(qstate->env->cfg->serve_expired &&
!qstate->env->cfg->serve_expired_client_timeout &&
(adjust == -1 || (time_t)expiry < *qstate->env->now)) {
qstate->need_refetch = 1;
}
@@ -568,10 +646,11 @@ parse_data(struct module_qstate* qstate, struct sldns_buffer* buf)
* return true if lookup was successful.
*/
static int
cachedb_extcache_lookup(struct module_qstate* qstate, struct cachedb_env* ie)
cachedb_extcache_lookup(struct module_qstate* qstate, struct cachedb_env* ie,
int* msg_expired)
{
char key[(CACHEDB_HASHSIZE/8)*2+1];
calc_hash(qstate, key, sizeof(key));
calc_hash(&qstate->qinfo, qstate->env, key, sizeof(key));
/* call backend to fetch data for key into scratch buffer */
if( !(*ie->backend->lookup)(qstate->env, ie, key,
@@ -585,7 +664,7 @@ cachedb_extcache_lookup(struct module_qstate* qstate, struct cachedb_env* ie)
}
/* parse dns message into return_msg */
if( !parse_data(qstate, qstate->env->scratch_buffer) ) {
if( !parse_data(qstate, qstate->env->scratch_buffer, msg_expired) ) {
return 0;
}
return 1;
@@ -598,7 +677,7 @@ static void
cachedb_extcache_store(struct module_qstate* qstate, struct cachedb_env* ie)
{
char key[(CACHEDB_HASHSIZE/8)*2+1];
calc_hash(qstate, key, sizeof(key));
calc_hash(&qstate->qinfo, qstate->env, key, sizeof(key));
/* prepare data in scratch buffer */
if(!prep_data(qstate, qstate->env->scratch_buffer))
@@ -615,13 +694,18 @@ cachedb_extcache_store(struct module_qstate* qstate, struct cachedb_env* ie)
* See if unbound's internal cache can answer the query
*/
static int
cachedb_intcache_lookup(struct module_qstate* qstate)
cachedb_intcache_lookup(struct module_qstate* qstate, struct cachedb_env* cde)
{
uint8_t dpname_storage[LDNS_MAX_DOMAINLEN+1];
uint8_t* dpname=NULL;
size_t dpnamelen=0;
struct dns_msg* msg;
/* for testframe bypass this lookup */
if(cde->backend == &testframe_backend) {
return 0;
}
if(iter_stub_fwd_no_cache(qstate, &qstate->qinfo,
&dpname, &dpnamelen))
&dpname, &dpnamelen, dpname_storage, sizeof(dpname_storage)))
return 0; /* no cache for these queries */
msg = dns_cache_lookup(qstate->env, qstate->qinfo.qname,
qstate->qinfo.qname_len, qstate->qinfo.qtype,
@@ -652,17 +736,44 @@ cachedb_intcache_lookup(struct module_qstate* qstate)
* Store query into the internal cache of unbound.
*/
static void
cachedb_intcache_store(struct module_qstate* qstate)
cachedb_intcache_store(struct module_qstate* qstate, int msg_expired)
{
uint32_t store_flags = qstate->query_flags;
int serve_expired = qstate->env->cfg->serve_expired;
if(qstate->env->cfg->serve_expired)
store_flags |= DNSCACHE_STORE_ZEROTTL;
if(!qstate->return_msg)
return;
if(serve_expired && msg_expired) {
/* Set TTLs to a value such that value + *env->now is
* going to be now-3 seconds. Making it expired
* in the cache. */
set_msg_ttl(qstate->return_msg, (time_t)-3);
/* The expired entry does not get checked by the validator
* and we need a validation value for it. */
if(qstate->env->cfg->cachedb_check_when_serve_expired)
qstate->return_msg->rep->security = sec_status_insecure;
}
(void)dns_cache_store(qstate->env, &qstate->qinfo,
qstate->return_msg->rep, 0, qstate->prefetch_leeway, 0,
qstate->region, store_flags);
qstate->region, store_flags, qstate->qstarttime,
qstate->is_valrec);
if(serve_expired && msg_expired) {
if(qstate->env->cfg->serve_expired_client_timeout) {
/* No expired response from the query state, the
* query resolution needs to continue and it can
* pick up the expired result after the timer out
* of cache. */
return;
}
/* set TTLs to zero again */
adjust_msg_ttl(qstate->return_msg, -1);
/* Send serve expired responses based on the cachedb
* returned message, that was just stored in the cache.
* It can then continue to work on this query. */
mesh_respond_serve_expired(qstate->mesh_info);
}
}
/**
@@ -678,6 +789,8 @@ cachedb_handle_query(struct module_qstate* qstate,
struct cachedb_qstate* ATTR_UNUSED(iq),
struct cachedb_env* ie, int id)
{
int msg_expired = 0;
qstate->is_cachedb_answer = 0;
/* check if we are enabled, and skip if so */
if(!ie->enabled) {
/* pass request to next module */
@@ -694,7 +807,7 @@ cachedb_handle_query(struct module_qstate* qstate,
/* lookup inside unbound's internal cache.
* This does not look for expired entries. */
if(cachedb_intcache_lookup(qstate)) {
if(cachedb_intcache_lookup(qstate, ie)) {
if(verbosity >= VERB_ALGO) {
if(qstate->return_msg->rep)
log_dns_msg("cachedb internal cache lookup",
@@ -711,31 +824,46 @@ cachedb_handle_query(struct module_qstate* qstate,
}
/* ask backend cache to see if we have data */
if(cachedb_extcache_lookup(qstate, ie)) {
if(cachedb_extcache_lookup(qstate, ie, &msg_expired)) {
if(verbosity >= VERB_ALGO)
log_dns_msg(ie->backend->name,
&qstate->return_msg->qinfo,
qstate->return_msg->rep);
/* store this result in internal cache */
cachedb_intcache_store(qstate);
cachedb_intcache_store(qstate, msg_expired);
/* In case we have expired data but there is a client timer for expired
* answers, pass execution to next module in order to try updating the
* data first.
* TODO: this needs revisit. The expired data stored from cachedb has
* 0 TTL which is picked up by iterator later when looking in the cache.
* Document that ext cachedb does not work properly with
* serve_stale_reply_ttl yet. */
*/
if(qstate->env->cfg->serve_expired && msg_expired) {
qstate->return_msg = NULL;
qstate->ext_state[id] = module_wait_module;
/* The expired reply is sent with
* mesh_respond_serve_expired, and so
* the need_refetch is not used. */
qstate->need_refetch = 0;
return;
}
if(qstate->need_refetch && qstate->serve_expired_data &&
qstate->serve_expired_data->timer) {
qstate->return_msg = NULL;
qstate->ext_state[id] = module_wait_module;
return;
}
qstate->is_cachedb_answer = 1;
/* we are done with the query */
qstate->ext_state[id] = module_finished;
return;
}
if(qstate->serve_expired_data &&
qstate->env->cfg->cachedb_check_when_serve_expired &&
!qstate->env->cfg->serve_expired_client_timeout) {
/* Reply with expired data if any to client, because cachedb
* also has no useful, current data */
mesh_respond_serve_expired(qstate->mesh_info);
}
/* no cache fetches */
/* pass request to next module */
qstate->ext_state[id] = module_wait_module;
@@ -753,12 +881,18 @@ static void
cachedb_handle_response(struct module_qstate* qstate,
struct cachedb_qstate* ATTR_UNUSED(iq), struct cachedb_env* ie, int id)
{
qstate->is_cachedb_answer = 0;
/* check if we are not enabled or instructed to not cache, and skip */
if(!ie->enabled || qstate->no_cache_store) {
/* we are done with the query */
qstate->ext_state[id] = module_finished;
return;
}
if(qstate->env->cfg->cachedb_no_store) {
/* do not store the item in the external cache */
qstate->ext_state[id] = module_finished;
return;
}
/* store the item into the backend cache */
cachedb_extcache_store(qstate, ie);
@@ -853,7 +987,7 @@ cachedb_get_mem(struct module_env* env, int id)
*/
static struct module_func_block cachedb_block = {
"cachedb",
&cachedb_init, &cachedb_deinit, &cachedb_operate,
NULL, NULL, &cachedb_init, &cachedb_deinit, &cachedb_operate,
&cachedb_inform_super, &cachedb_clear, &cachedb_get_mem
};
@@ -862,4 +996,41 @@ cachedb_get_funcblock(void)
{
return &cachedb_block;
}
int
cachedb_is_enabled(struct module_stack* mods, struct module_env* env)
{
struct cachedb_env* ie;
int id = modstack_find(mods, "cachedb");
if(id == -1)
return 0;
ie = (struct cachedb_env*)env->modinfo[id];
if(ie && ie->enabled)
return 1;
return 0;
}
void cachedb_msg_remove(struct module_qstate* qstate)
{
cachedb_msg_remove_qinfo(qstate->env, &qstate->qinfo);
}
void cachedb_msg_remove_qinfo(struct module_env* env, struct query_info* qinfo)
{
char key[(CACHEDB_HASHSIZE/8)*2+1];
int id = modstack_find(env->modstack, "cachedb");
struct cachedb_env* ie = (struct cachedb_env*)env->modinfo[id];
log_query_info(VERB_ALGO, "cachedb msg remove", qinfo);
calc_hash(qinfo, env, key, sizeof(key));
sldns_buffer_clear(env->scratch_buffer);
sldns_buffer_write_u32(env->scratch_buffer, 0);
sldns_buffer_flip(env->scratch_buffer);
/* call backend */
(*ie->backend->store)(env, ie, key,
sldns_buffer_begin(env->scratch_buffer),
sldns_buffer_limit(env->scratch_buffer),
0);
}
#endif /* USE_CACHEDB */
+24
View File
@@ -41,6 +41,7 @@
*/
#include "util/module.h"
struct cachedb_backend;
struct module_stack;
/**
* The global variable environment contents for the cachedb
@@ -110,3 +111,26 @@ size_t cachedb_get_mem(struct module_env* env, int id);
*/
struct module_func_block* cachedb_get_funcblock(void);
/**
* See if the cachedb is enabled.
* @param mods: module stack. It finds the cachedb module environment.
* @param env: module environment.
* @return true if exists and enabled.
*/
int cachedb_is_enabled(struct module_stack* mods, struct module_env* env);
/**
* Remove a message from the global cache. Because edns subnet has a more
* specific entry, and if not removed when everything expires, the global
* entry is used, instead of a fresh lookup of the edns subnet entry.
* @param qstate: query state.
*/
void cachedb_msg_remove(struct module_qstate* qstate);
/**
* Remove message from the cachedb cache, by query info.
* @param env: module environment to look up cachedb state.
* @param qinfo: the message to remove.
*/
void cachedb_msg_remove_qinfo(struct module_env* env,
struct query_info* qinfo);
+370 -69
View File
@@ -46,89 +46,335 @@
#include "cachedb/cachedb.h"
#include "util/alloc.h"
#include "util/config_file.h"
#include "util/locks.h"
#include "util/timeval_func.h"
#include "sldns/sbuffer.h"
#ifdef USE_REDIS
#include "hiredis/hiredis.h"
struct redis_moddata {
redisContext** ctxs; /* thread-specific redis contexts */
int numctxs; /* number of ctx entries */
const char* server_host; /* server's IP address or host name */
int server_port; /* server's TCP port */
struct timeval timeout; /* timeout for connection setup and commands */
/* thread-specific redis contexts */
redisContext** ctxs;
redisContext** replica_ctxs;
/* number of ctx entries */
int numctxs;
/* server's IP address or host name */
const char* server_host;
const char* replica_server_host;
/* server's TCP port */
int server_port;
int replica_server_port;
/* server's unix path, or "", NULL if unused */
const char* server_path;
const char* replica_server_path;
/* server's AUTH password, or "", NULL if unused */
const char* server_password;
const char* replica_server_password;
/* timeout for commands */
struct timeval command_timeout;
struct timeval replica_command_timeout;
/* timeout for connection setup */
struct timeval connect_timeout;
struct timeval replica_connect_timeout;
/* the reconnect interval time. */
struct timeval reconnect_interval;
struct timeval replica_reconnect_interval;
/* reconnect attempts, 0 if connected, counts up failed reconnects. */
int reconnect_attempts;
int replica_reconnect_attempts;
/* Lock on reconnect_wait time. */
lock_basic_type wait_lock;
lock_basic_type replica_wait_lock;
/* reconnect wait time, wait until it has passed before reconnect. */
struct timeval reconnect_wait;
struct timeval replica_reconnect_wait;
/* the redis logical database to use */
int logical_db;
int replica_logical_db;
/* if the SET with EX command is supported */
int set_with_ex_available;
};
/** The limit on the number of redis connect attempts. After failure if
* the number is exceeded, the reconnects are throttled by the wait time. */
#define REDIS_RECONNECT_ATTEMPT_LIMIT 3
static redisReply* redis_command(struct module_env*, struct cachedb_env*,
const char*, const uint8_t*, size_t);
const char*, const uint8_t*, size_t, int);
static void
moddata_clean(struct redis_moddata** moddata) {
if(!moddata || !*moddata)
return;
if((*moddata)->ctxs) {
int i;
for(i = 0; i < (*moddata)->numctxs; i++) {
if((*moddata)->ctxs[i])
redisFree((*moddata)->ctxs[i]);
}
free((*moddata)->ctxs);
}
if((*moddata)->replica_ctxs) {
int i;
for(i = 0; i < (*moddata)->numctxs; i++) {
if((*moddata)->replica_ctxs[i])
redisFree((*moddata)->replica_ctxs[i]);
}
free((*moddata)->replica_ctxs);
}
lock_basic_destroy(&(*moddata)->wait_lock);
lock_basic_destroy(&(*moddata)->replica_wait_lock);
free(*moddata);
*moddata = NULL;
}
static redisContext*
redis_connect(const struct redis_moddata* moddata)
redis_connect(const char* host, int port, const char* path,
const char* password, int logical_db,
const struct timeval connect_timeout,
const struct timeval command_timeout,
const struct timeval* reconnect_interval,
int* reconnect_attempts,
struct timeval* reconnect_wait,
lock_basic_type* wait_lock,
struct timeval* now_tv,
const char* infostr)
{
struct timeval now_val;
redisContext* ctx;
ctx = redisConnectWithTimeout(moddata->server_host,
moddata->server_port, moddata->timeout);
/* See if the redis server is down, and reconnect has to wait. */
if(*reconnect_attempts > REDIS_RECONNECT_ATTEMPT_LIMIT) {
/* Acquire lock to look at timeval, the integer has atomic
* integrity. */
struct timeval wait_tv;
if(now_tv) {
now_val = *now_tv;
} else {
if(gettimeofday(&now_val, NULL) < 0)
log_err("redis: gettimeofday: %s",
strerror(errno));
}
lock_basic_lock(wait_lock);
wait_tv = *reconnect_wait;
lock_basic_unlock(wait_lock);
if(timeval_smaller(&now_val, &wait_tv)) {
verbose(VERB_ALGO, "redis %sdown, reconnect wait",
infostr);
return NULL;
}
}
if(path && path[0]!=0) {
ctx = redisConnectUnixWithTimeout(path, connect_timeout);
} else {
ctx = redisConnectWithTimeout(host, port, connect_timeout);
}
if(!ctx || ctx->err) {
const char *errstr = "out of memory";
if(ctx)
errstr = ctx->errstr;
log_err("failed to connect to redis server: %s", errstr);
log_err("failed to connect to redis %sserver: %s", infostr, errstr);
goto fail;
}
if(redisSetTimeout(ctx, moddata->timeout) != REDIS_OK) {
log_err("failed to set redis timeout");
if(redisSetTimeout(ctx, command_timeout) != REDIS_OK) {
log_err("failed to set redis %stimeout, %s", infostr, ctx->errstr);
goto fail;
}
if(password && password[0]!=0) {
redisReply* rep;
rep = redisCommand(ctx, "AUTH %s", password);
if(!rep || rep->type == REDIS_REPLY_ERROR) {
log_err("failed to authenticate %swith password", infostr);
freeReplyObject(rep);
goto fail;
}
freeReplyObject(rep);
}
if(logical_db > 0) {
redisReply* rep;
rep = redisCommand(ctx, "SELECT %d", logical_db);
if(!rep || rep->type == REDIS_REPLY_ERROR) {
log_err("failed %sto set logical database (%d)",
infostr, logical_db);
freeReplyObject(rep);
goto fail;
}
freeReplyObject(rep);
}
*reconnect_attempts = 0;
if(verbosity >= VERB_OPS) {
char port_str[6+1];
port_str[0] = ' ';
(void)snprintf(port_str+1, sizeof(port_str)-1, "%d", port);
verbose(VERB_OPS, "Connection to Redis %sestablished (%s%s)",
infostr,
path&&path[0]!=0?path:host,
path&&path[0]!=0?"":port_str);
}
return ctx;
fail:
fail:
if(ctx)
redisFree(ctx);
(*reconnect_attempts)++;
if(*reconnect_attempts > REDIS_RECONNECT_ATTEMPT_LIMIT) {
/* Wait for the reconnect interval before trying again. */
struct timeval tv;
if(now_tv) {
now_val = *now_tv;
} else {
if(gettimeofday(&now_val, NULL) < 0)
log_err("redis: gettimeofday: %s",
strerror(errno));
}
tv = now_val;
timeval_add(&tv, reconnect_interval);
lock_basic_lock(wait_lock);
*reconnect_wait = tv;
lock_basic_unlock(wait_lock);
verbose(VERB_ALGO, "redis %sreconnect wait until %d.%6.6d",
infostr, (int)tv.tv_sec, (int)tv.tv_usec);
}
return NULL;
}
static void
set_timeout(struct timeval* timeout, int value, int explicit_value)
{
int v = explicit_value != 0 ? explicit_value : value;
timeout->tv_sec = v / 1000;
timeout->tv_usec = (v % 1000) * 1000;
}
static int
redis_init(struct module_env* env, struct cachedb_env* cachedb_env)
{
int i;
struct redis_moddata* moddata = NULL;
verbose(VERB_ALGO, "redis_init");
verbose(VERB_OPS, "Redis initialization");
moddata = calloc(1, sizeof(struct redis_moddata));
if(!moddata) {
log_err("out of memory");
return 0;
goto fail;
}
lock_basic_init(&moddata->wait_lock);
lock_protect(&moddata->wait_lock, &moddata->reconnect_wait,
sizeof(moddata->reconnect_wait));
lock_basic_init(&moddata->replica_wait_lock);
lock_protect(&moddata->replica_wait_lock,
&moddata->replica_reconnect_wait,
sizeof(moddata->replica_reconnect_wait));
moddata->numctxs = env->cfg->num_threads;
/* note: server_host and similar string configuration options are
* shallow references to configured strings; we don't have to free them
* in this module. */
moddata->server_host = env->cfg->redis_server_host;
moddata->replica_server_host = env->cfg->redis_replica_server_host;
moddata->server_port = env->cfg->redis_server_port;
moddata->replica_server_port = env->cfg->redis_replica_server_port;
moddata->server_path = env->cfg->redis_server_path;
moddata->replica_server_path = env->cfg->redis_replica_server_path;
moddata->server_password = env->cfg->redis_server_password;
moddata->replica_server_password = env->cfg->redis_replica_server_password;
set_timeout(&moddata->command_timeout,
env->cfg->redis_timeout,
env->cfg->redis_command_timeout);
set_timeout(&moddata->replica_command_timeout,
env->cfg->redis_replica_timeout,
env->cfg->redis_replica_command_timeout);
set_timeout(&moddata->connect_timeout,
env->cfg->redis_timeout,
env->cfg->redis_connect_timeout);
set_timeout(&moddata->replica_connect_timeout,
env->cfg->redis_replica_timeout,
env->cfg->redis_replica_connect_timeout);
set_timeout(&moddata->reconnect_interval, 1000, 0);
set_timeout(&moddata->replica_reconnect_interval, 1000, 0);
moddata->logical_db = env->cfg->redis_logical_db;
moddata->replica_logical_db = env->cfg->redis_replica_logical_db;
moddata->ctxs = calloc(env->cfg->num_threads, sizeof(redisContext*));
if(!moddata->ctxs) {
log_err("out of memory");
free(moddata);
return 0;
goto fail;
}
if((moddata->replica_server_host && moddata->replica_server_host[0]!=0)
|| (moddata->replica_server_path && moddata->replica_server_path[0]!=0)) {
/* There is a replica configured, allocate ctxs */
moddata->replica_ctxs = calloc(env->cfg->num_threads, sizeof(redisContext*));
if(!moddata->replica_ctxs) {
log_err("out of memory");
goto fail;
}
}
for(i = 0; i < moddata->numctxs; i++) {
redisContext* ctx = redis_connect(
moddata->server_host,
moddata->server_port,
moddata->server_path,
moddata->server_password,
moddata->logical_db,
moddata->connect_timeout,
moddata->command_timeout,
&moddata->reconnect_interval,
&moddata->reconnect_attempts,
&moddata->reconnect_wait,
&moddata->wait_lock,
env->now_tv,
"");
if(!ctx) {
log_err("redis_init: failed to init redis "
"(for thread %d)", i);
/* And continue, the context can be established
* later, just like after a disconnect. */
}
moddata->ctxs[i] = ctx;
}
if(moddata->replica_ctxs) {
for(i = 0; i < moddata->numctxs; i++) {
redisContext* ctx = redis_connect(
moddata->replica_server_host,
moddata->replica_server_port,
moddata->replica_server_path,
moddata->replica_server_password,
moddata->replica_logical_db,
moddata->replica_connect_timeout,
moddata->replica_command_timeout,
&moddata->replica_reconnect_interval,
&moddata->replica_reconnect_attempts,
&moddata->replica_reconnect_wait,
&moddata->replica_wait_lock,
env->now_tv,
"replica ");
if(!ctx) {
log_err("redis_init: failed to init redis "
"replica (for thread %d)", i);
/* And continue, the context can be established
* later, just like after a disconnect. */
}
moddata->replica_ctxs[i] = ctx;
}
}
/* note: server_host is a shallow reference to configured string.
* we don't have to free it in this module. */
moddata->server_host = env->cfg->redis_server_host;
moddata->server_port = env->cfg->redis_server_port;
moddata->timeout.tv_sec = env->cfg->redis_timeout / 1000;
moddata->timeout.tv_usec = (env->cfg->redis_timeout % 1000) * 1000;
for(i = 0; i < moddata->numctxs; i++)
moddata->ctxs[i] = redis_connect(moddata);
cachedb_env->backend_data = moddata;
if(env->cfg->redis_expire_records) {
if(env->cfg->redis_expire_records &&
moddata->ctxs[env->alloc->thread_num] != NULL) {
redisReply* rep = NULL;
int redis_reply_type = 0;
/** check if setex command is supported */
/** check if set with ex command is supported */
rep = redis_command(env, cachedb_env,
"SETEX __UNBOUND_REDIS_CHECK__ 1 none", NULL, 0);
"SET __UNBOUND_REDIS_CHECK__ none EX 1", NULL, 0, 1);
if(!rep) {
/** init failed, no response from redis server*/
log_err("redis_init: failed to init redis, the "
"redis-expire-records option requires the SETEX command "
"(redis >= 2.0.0)");
return 0;
goto set_with_ex_fail;
}
redis_reply_type = rep->type;
freeReplyObject(rep);
@@ -136,15 +382,21 @@ redis_init(struct module_env* env, struct cachedb_env* cachedb_env)
case REDIS_REPLY_STATUS:
break;
default:
/** init failed, setex command not supported */
log_err("redis_init: failed to init redis, the "
"redis-expire-records option requires the SETEX command "
"(redis >= 2.0.0)");
return 0;
/** init failed, set_with_ex command not supported */
goto set_with_ex_fail;
}
moddata->set_with_ex_available = 1;
}
return 1;
set_with_ex_fail:
log_err("redis_init: failure during redis_init, the "
"redis-expire-records option requires the SET with EX command "
"(redis >= 2.6.12)");
return 1;
fail:
moddata_clean(&moddata);
return 0;
}
static void
@@ -154,19 +406,8 @@ redis_deinit(struct module_env* env, struct cachedb_env* cachedb_env)
cachedb_env->backend_data;
(void)env;
verbose(VERB_ALGO, "redis_deinit");
if(!moddata)
return;
if(moddata->ctxs) {
int i;
for(i = 0; i < moddata->numctxs; i++) {
if(moddata->ctxs[i])
redisFree(moddata->ctxs[i]);
}
free(moddata->ctxs);
}
free(moddata);
verbose(VERB_OPS, "Redis deinitialization");
moddata_clean(&moddata);
}
/*
@@ -183,9 +424,9 @@ redis_deinit(struct module_env* env, struct cachedb_env* cachedb_env)
*/
static redisReply*
redis_command(struct module_env* env, struct cachedb_env* cachedb_env,
const char* command, const uint8_t* data, size_t data_len)
const char* command, const uint8_t* data, size_t data_len, int write)
{
redisContext* ctx;
redisContext* ctx, **ctx_selector;
redisReply* rep;
struct redis_moddata* d = (struct redis_moddata*)
cachedb_env->backend_data;
@@ -196,17 +437,50 @@ redis_command(struct module_env* env, struct cachedb_env* cachedb_env,
* assumption throughout the unbound architecture, so we simply assert
* it. */
log_assert(env->alloc->thread_num < d->numctxs);
ctx = d->ctxs[env->alloc->thread_num];
ctx_selector = !write && d->replica_ctxs
?d->replica_ctxs
:d->ctxs;
ctx = ctx_selector[env->alloc->thread_num];
/* If we've not established a connection to the server or we've closed
* it on a failure, try to re-establish a new one. Failures will be
* logged in redis_connect(). */
if(!ctx) {
ctx = redis_connect(d);
d->ctxs[env->alloc->thread_num] = ctx;
if(!write && d->replica_ctxs) {
ctx = redis_connect(
d->replica_server_host,
d->replica_server_port,
d->replica_server_path,
d->replica_server_password,
d->replica_logical_db,
d->replica_connect_timeout,
d->replica_command_timeout,
&d->replica_reconnect_interval,
&d->replica_reconnect_attempts,
&d->replica_reconnect_wait,
&d->replica_wait_lock,
env->now_tv,
"replica ");
} else {
ctx = redis_connect(
d->server_host,
d->server_port,
d->server_path,
d->server_password,
d->logical_db,
d->connect_timeout,
d->command_timeout,
&d->reconnect_interval,
&d->reconnect_attempts,
&d->reconnect_wait,
&d->wait_lock,
env->now_tv,
"");
}
ctx_selector[env->alloc->thread_num] = ctx;
}
if(!ctx)
return NULL;
if(!ctx) return NULL;
/* Send the command and get a reply, synchronously. */
rep = (redisReply*)redisCommand(ctx, command, data, data_len);
@@ -216,7 +490,7 @@ redis_command(struct module_env* env, struct cachedb_env* cachedb_env,
log_err("redis_command: failed to receive a reply, "
"closing connection: %s", ctx->errstr);
redisFree(ctx);
d->ctxs[env->alloc->thread_num] = NULL;
ctx_selector[env->alloc->thread_num] = NULL;
return NULL;
}
@@ -234,7 +508,14 @@ redis_lookup(struct module_env* env, struct cachedb_env* cachedb_env,
char* key, struct sldns_buffer* result_buffer)
{
redisReply* rep;
char cmdbuf[4+(CACHEDB_HASHSIZE/8)*2+1]; /* "GET " + key */
/* Supported commands:
* - "GET " + key
*/
#define REDIS_LOOKUP_MAX_BUF_LEN \
4 /* "GET " */ \
+(CACHEDB_HASHSIZE/8)*2 /* key hash */ \
+ 1 /* \0 */
char cmdbuf[REDIS_LOOKUP_MAX_BUF_LEN];
int n;
int ret = 0;
@@ -246,7 +527,7 @@ redis_lookup(struct module_env* env, struct cachedb_env* cachedb_env,
return 0;
}
rep = redis_command(env, cachedb_env, cmdbuf, NULL, 0);
rep = redis_command(env, cachedb_env, cmdbuf, NULL, 0, 0);
if(!rep)
return 0;
switch(rep->type) {
@@ -283,26 +564,45 @@ redis_store(struct module_env* env, struct cachedb_env* cachedb_env,
{
redisReply* rep;
int n;
int set_ttl = (env->cfg->redis_expire_records &&
struct redis_moddata* moddata = (struct redis_moddata*)
cachedb_env->backend_data;
int set_ttl = (moddata->set_with_ex_available &&
env->cfg->redis_expire_records &&
(!env->cfg->serve_expired || env->cfg->serve_expired_ttl > 0));
/* Supported commands:
* - "SET " + key + " %b"
* - "SETEX " + key + " " + ttl + " %b"
* - "SET " + key + " %b EX " + ttl
* older redis 2.0.0 was "SETEX " + key + " " + ttl + " %b"
* - "EXPIRE " + key + " 0"
*/
char cmdbuf[6+(CACHEDB_HASHSIZE/8)*2+11+3+1];
#define REDIS_STORE_MAX_BUF_LEN \
7 /* "EXPIRE " */ \
+(CACHEDB_HASHSIZE/8)*2 /* key hash */ \
+ 7 /* " %b EX " */ \
+ 20 /* ttl (uint64_t) */ \
+ 1 /* \0 */
char cmdbuf[REDIS_STORE_MAX_BUF_LEN];
if (!set_ttl) {
verbose(VERB_ALGO, "redis_store %s (%d bytes)", key, (int)data_len);
/* build command to set to a binary safe string */
n = snprintf(cmdbuf, sizeof(cmdbuf), "SET %s %%b", key);
} else if(ttl == 0) {
/* use the EXPIRE command, SET with EX 0 is an invalid time. */
/* Replies with REDIS_REPLY_INTEGER of 1. */
verbose(VERB_ALGO, "redis_store expire %s (%d bytes)",
key, (int)data_len);
n = snprintf(cmdbuf, sizeof(cmdbuf), "EXPIRE %s 0", key);
data = NULL;
data_len = 0;
} else {
/* add expired ttl time to redis ttl to avoid premature eviction of key */
ttl += env->cfg->serve_expired_ttl;
verbose(VERB_ALGO, "redis_store %s (%d bytes) with ttl %u",
key, (int)data_len, (uint32_t)ttl);
key, (int)data_len, (unsigned)(uint32_t)ttl);
/* build command to set to a binary safe string */
n = snprintf(cmdbuf, sizeof(cmdbuf), "SETEX %s %u %%b", key,
(uint32_t)ttl);
n = snprintf(cmdbuf, sizeof(cmdbuf), "SET %s %%b EX %u", key,
(unsigned)(uint32_t)ttl);
}
@@ -311,11 +611,12 @@ redis_store(struct module_env* env, struct cachedb_env* cachedb_env,
return;
}
rep = redis_command(env, cachedb_env, cmdbuf, data, data_len);
rep = redis_command(env, cachedb_env, cmdbuf, data, data_len, 1);
if(rep) {
verbose(VERB_ALGO, "redis_store set completed");
if(rep->type != REDIS_REPLY_STATUS &&
rep->type != REDIS_REPLY_ERROR) {
rep->type != REDIS_REPLY_ERROR &&
rep->type != REDIS_REPLY_INTEGER) {
log_err("redis_store: unexpected type of reply (%d)",
rep->type);
}
+7 -1
View File
@@ -57,6 +57,8 @@
#define BLOCKSZ 64
#define RSBUFSZ (16*BLOCKSZ)
#define REKEY_BASE (1024*1024) /* NB. should be a power of 2 */
/* Marked MAP_INHERIT_ZERO, so zero'd out in fork children. */
static struct {
size_t rs_have; /* valid bytes at end of rs_buf */
@@ -179,6 +181,7 @@ static void
_rs_stir(void)
{
u_char rnd[KEYSZ + IVSZ];
uint32_t rekey_fuzz = 0;
if (getentropy(rnd, sizeof rnd) == -1) {
if(errno != ENOSYS ||
@@ -201,7 +204,10 @@ _rs_stir(void)
rs->rs_have = 0;
memset(rsx->rs_buf, 0, sizeof(rsx->rs_buf));
rs->rs_count = 1600000;
/* rekey interval should not be predictable */
chacha_encrypt_bytes(&rsx->rs_chacha, (uint8_t *)&rekey_fuzz,
(uint8_t *)&rekey_fuzz, sizeof(rekey_fuzz));
rs->rs_count = REKEY_BASE + (rekey_fuzz % REKEY_BASE);
}
static inline void
+3 -3
View File
@@ -57,7 +57,7 @@ int getnameinfo(const struct sockaddr *sa, size_t ATTR_UNUSED(salen), char *host
}
if (host != NULL) {
if (flags & NI_NUMERICHOST) {
if ((flags & NI_NUMERICHOST)) {
if (strlcpy(host, inet_ntoa(sin->sin_addr),
hostlen) >= hostlen)
return (EAI_MEMORY);
@@ -168,7 +168,7 @@ getaddrinfo(const char *hostname, const char *servname,
port = 0;
}
if (hints && hints->ai_flags & AI_PASSIVE) {
if (hints && (hints->ai_flags & AI_PASSIVE)) {
addr = htonl(0x00000000);
if (hostname && inet_aton(hostname, &in) != 0)
addr = in.s_addr;
@@ -193,7 +193,7 @@ getaddrinfo(const char *hostname, const char *servname,
}
/* Don't try DNS if AI_NUMERICHOST is set */
if (hints && hints->ai_flags & AI_NUMERICHOST)
if (hints && (hints->ai_flags & AI_NUMERICHOST))
return (EAI_NONAME);
hp = gethostbyname(hostname);
+1 -1
View File
@@ -47,7 +47,7 @@
#define SHA512_Update SHA512Update
#define SHA512_Final SHA512Final
#else
#include "openssl/sha.h"
#include <openssl/sha.h>
#endif
#include <sys/vfs.h>
-4
View File
@@ -5,12 +5,8 @@
#undef malloc
#include <sys/types.h>
#ifndef USE_WINSOCK
void *malloc ();
#else
/* provide a prototype */
void *malloc (size_t n);
#endif
/* Allocate an N-byte block of memory from the heap.
If N is zero, allocate a 1-byte block. */
+704 -559
View File
File diff suppressed because it is too large Load Diff
+244 -34
View File
@@ -1,5 +1,8 @@
/* config.h.in. Generated from configure.ac by autoheader. */
/* apply the fallthrough attribute. */
#undef ATTR_FALLTHROUGH
/* apply the noreturn attribute to a function that exits the program */
#undef ATTR_NORETURN
@@ -57,6 +60,9 @@
/* Define to 1 if you have the <arpa/inet.h> header file. */
#undef HAVE_ARPA_INET_H
/* Whether the C compiler accepts the "fallthrough" attribute */
#undef HAVE_ATTR_FALLTHROUGH
/* Whether the C compiler accepts the "format" attribute */
#undef HAVE_ATTR_FORMAT
@@ -123,6 +129,14 @@
and to 0 if you don't. */
#undef HAVE_DECL_NGHTTP2_SESSION_SERVER_NEW
/* Define to 1 if you have the declaration of `ngtcp2_conn_server_new', and to
0 if you don't. */
#undef HAVE_DECL_NGTCP2_CONN_SERVER_NEW
/* Define to 1 if you have the declaration of `ngtcp2_crypto_encrypt_cb', and
to 0 if you don't. */
#undef HAVE_DECL_NGTCP2_CRYPTO_ENCRYPT_CB
/* Define to 1 if you have the declaration of `NID_ED25519', and to 0 if you
don't. */
#undef HAVE_DECL_NID_ED25519
@@ -159,6 +173,10 @@
0 if you don't. */
#undef HAVE_DECL_SSL_CTX_SET_ECDH_AUTO
/* Define to 1 if you have the declaration of `SSL_CTX_set_tmp_ecdh', and to 0
if you don't. */
#undef HAVE_DECL_SSL_CTX_SET_TMP_ECDH
/* Define to 1 if you have the declaration of `strlcat', and to 0 if you
don't. */
#undef HAVE_DECL_STRLCAT
@@ -222,6 +240,10 @@
/* Define to 1 if you have the `EVP_cleanup' function. */
#undef HAVE_EVP_CLEANUP
/* Define to 1 if you have the `EVP_default_properties_is_fips_enabled'
function. */
#undef HAVE_EVP_DEFAULT_PROPERTIES_IS_FIPS_ENABLED
/* Define to 1 if you have the `EVP_DigestVerify' function. */
#undef HAVE_EVP_DIGESTVERIFY
@@ -294,6 +316,9 @@
/* Define to 1 if you have the `getrlimit' function. */
#undef HAVE_GETRLIMIT
/* Define to 1 if you have the `gettid' function. */
#undef HAVE_GETTID
/* Define to 1 if you have the `glob' function. */
#undef HAVE_GLOB
@@ -357,6 +382,12 @@
/* Define if we have LibreSSL */
#undef HAVE_LIBRESSL
/* If we have atomic_store */
#undef HAVE_LINK_ATOMIC_STORE
/* Define to 1 if you have the <linux/net_tstamp.h> header file. */
#undef HAVE_LINUX_NET_TSTAMP_H
/* Define to 1 if you have the `localtime_r' function. */
#undef HAVE_LOCALTIME_R
@@ -369,8 +400,8 @@
/* Define to 1 if you have the `memmove' function. */
#undef HAVE_MEMMOVE
/* Define to 1 if you have the <memory.h> header file. */
#undef HAVE_MEMORY_H
/* Define to 1 if you have the <minix/config.h> header file. */
#undef HAVE_MINIX_CONFIG_H
/* Define to 1 if you have the <netdb.h> header file. */
#undef HAVE_NETDB_H
@@ -396,12 +427,80 @@
/* Define to 1 if you have the <net/if.h> header file. */
#undef HAVE_NET_IF_H
/* Define to 1 if you have the <net/pfvar.h> header file. */
#undef HAVE_NET_PFVAR_H
/* Define this to use nghttp2 client. */
#undef HAVE_NGHTTP2
/* Define to 1 if you have the <nghttp2/nghttp2.h> header file. */
#undef HAVE_NGHTTP2_NGHTTP2_H
/* Define this to use ngtcp2. */
#undef HAVE_NGTCP2
/* Define to 1 if you have the `ngtcp2_ccerr_default' function. */
#undef HAVE_NGTCP2_CCERR_DEFAULT
/* Define to 1 if you have the `ngtcp2_conn_encode_0rtt_transport_params'
function. */
#undef HAVE_NGTCP2_CONN_ENCODE_0RTT_TRANSPORT_PARAMS
/* Define to 1 if you have the `ngtcp2_conn_get_max_local_streams_uni'
function. */
#undef HAVE_NGTCP2_CONN_GET_MAX_LOCAL_STREAMS_UNI
/* Define to 1 if you have the `ngtcp2_conn_get_num_scid' function. */
#undef HAVE_NGTCP2_CONN_GET_NUM_SCID
/* Define to 1 if you have the `ngtcp2_conn_in_closing_period' function. */
#undef HAVE_NGTCP2_CONN_IN_CLOSING_PERIOD
/* Define to 1 if you have the `ngtcp2_conn_in_draining_period' function. */
#undef HAVE_NGTCP2_CONN_IN_DRAINING_PERIOD
/* Define if ngtcp2_conn_shutdown_stream has 4 arguments. */
#undef HAVE_NGTCP2_CONN_SHUTDOWN_STREAM4
/* Define to 1 if you have the `ngtcp2_conn_tls_early_data_rejected' function.
*/
#undef HAVE_NGTCP2_CONN_TLS_EARLY_DATA_REJECTED
/* Define to 1 if you have the `ngtcp2_crypto_encrypt_cb' function. */
#undef HAVE_NGTCP2_CRYPTO_ENCRYPT_CB
/* Define to 1 if you have the
`ngtcp2_crypto_quictls_configure_client_context' function. */
#undef HAVE_NGTCP2_CRYPTO_QUICTLS_CONFIGURE_CLIENT_CONTEXT
/* Define to 1 if you have the
`ngtcp2_crypto_quictls_configure_server_context' function. */
#undef HAVE_NGTCP2_CRYPTO_QUICTLS_CONFIGURE_SERVER_CONTEXT
/* Define to 1 if you have the
`ngtcp2_crypto_quictls_from_ossl_encryption_level' function. */
#undef HAVE_NGTCP2_CRYPTO_QUICTLS_FROM_OSSL_ENCRYPTION_LEVEL
/* Define to 1 if you have the `ngtcp2_crypto_quictls_init' function. */
#undef HAVE_NGTCP2_CRYPTO_QUICTLS_INIT
/* Define to 1 if the system has the type `ngtcp2_encryption_level'. */
#undef HAVE_NGTCP2_ENCRYPTION_LEVEL
/* Define to 1 if you have the <ngtcp2/ngtcp2_crypto_openssl.h> header file.
*/
#undef HAVE_NGTCP2_NGTCP2_CRYPTO_OPENSSL_H
/* Define to 1 if you have the <ngtcp2/ngtcp2_crypto_ossl.h> header file. */
#undef HAVE_NGTCP2_NGTCP2_CRYPTO_OSSL_H
/* Define to 1 if you have the <ngtcp2/ngtcp2_crypto_quictls.h> header file.
*/
#undef HAVE_NGTCP2_NGTCP2_CRYPTO_QUICTLS_H
/* Define to 1 if you have the <ngtcp2/ngtcp2.h> header file. */
#undef HAVE_NGTCP2_NGTCP2_H
/* Use libnss for crypto */
#undef HAVE_NSS
@@ -453,6 +552,12 @@
/* Define to 1 if you have the `OSSL_PARAM_BLD_new' function. */
#undef HAVE_OSSL_PARAM_BLD_NEW
/* Define to 1 if you have the `poll' function. */
#undef HAVE_POLL
/* Define to 1 if you have the <poll.h> header file. */
#undef HAVE_POLL_H
/* Define if you have POSIX threads libraries and header files. */
#undef HAVE_PTHREAD
@@ -559,18 +664,27 @@
/* Define to 1 if you have the `SSL_get1_peer_certificate' function. */
#undef HAVE_SSL_GET1_PEER_CERTIFICATE
/* Define to 1 if you have the `SSL_is_quic' function. */
#undef HAVE_SSL_IS_QUIC
/* Define to 1 if you have the `SSL_set1_host' function. */
#undef HAVE_SSL_SET1_HOST
/* Define to 1 if you have the <stdarg.h> header file. */
#undef HAVE_STDARG_H
/* Define to 1 if you have the <stdatomic.h> header file. */
#undef HAVE_STDATOMIC_H
/* Define to 1 if you have the <stdbool.h> header file. */
#undef HAVE_STDBOOL_H
/* Define to 1 if you have the <stdint.h> header file. */
#undef HAVE_STDINT_H
/* Define to 1 if you have the <stdio.h> header file. */
#undef HAVE_STDIO_H
/* Define to 1 if you have the <stdlib.h> header file. */
#undef HAVE_STDLIB_H
@@ -598,6 +712,23 @@
/* Define to 1 if `ipi_spec_dst' is a member of `struct in_pktinfo'. */
#undef HAVE_STRUCT_IN_PKTINFO_IPI_SPEC_DST
/* Define to 1 if `tokenlen' is a member of `struct ngtcp2_pkt_hd'. */
#undef HAVE_STRUCT_NGTCP2_PKT_HD_TOKENLEN
/* Define to 1 if `max_tx_udp_payload_size' is a member of `struct
ngtcp2_settings'. */
#undef HAVE_STRUCT_NGTCP2_SETTINGS_MAX_TX_UDP_PAYLOAD_SIZE
/* Define to 1 if `tokenlen' is a member of `struct ngtcp2_settings'. */
#undef HAVE_STRUCT_NGTCP2_SETTINGS_TOKENLEN
/* Define to 1 if `original_dcid_present' is a member of `struct
ngtcp2_transport_params'. */
#undef HAVE_STRUCT_NGTCP2_TRANSPORT_PARAMS_ORIGINAL_DCID_PRESENT
/* Define to 1 if the system has the type `struct ngtcp2_version_cid'. */
#undef HAVE_STRUCT_NGTCP2_VERSION_CID
/* Define to 1 if `sun_len' is a member of `struct sockaddr_un'. */
#undef HAVE_STRUCT_SOCKADDR_UN_SUN_LEN
@@ -673,6 +804,9 @@
/* Define to 1 if you have the <vfork.h> header file. */
#undef HAVE_VFORK_H
/* Define to 1 if you have the <wchar.h> header file. */
#undef HAVE_WCHAR_H
/* Define to 1 if you have the <windows.h> header file. */
#undef HAVE_WINDOWS_H
@@ -796,19 +930,27 @@
/* Shared data */
#undef SHARE_DIR
/* The size of `pthread_t', as computed by sizeof. */
#undef SIZEOF_PTHREAD_T
/* The size of `size_t', as computed by sizeof. */
#undef SIZEOF_SIZE_T
/* The size of `time_t', as computed by sizeof. */
#undef SIZEOF_TIME_T
/* The size of `unsigned long', as computed by sizeof. */
#undef SIZEOF_UNSIGNED_LONG
/* define if (v)snprintf does not return length needed, (but length used) */
#undef SNPRINTF_RET_BROKEN
/* Define to 1 if libsodium supports sodium_set_misuse_handler */
#undef SODIUM_MISUSE_HANDLER
/* Define to 1 if you have the ANSI C header files. */
/* Define to 1 if all of the C90 standard headers exist (not just the ones
required in a freestanding environment). This macro is provided for
backward compatibility; new code need not use it. */
#undef STDC_HEADERS
/* use default strptime. */
@@ -888,6 +1030,9 @@
/* Define this to enable client TCP Fast Open. */
#undef USE_MSG_FASTOPEN
/* Define this to use ngtcp2_crypto_ossl. */
#undef USE_NGTCP2_CRYPTO_OSSL
/* Define this to enable client TCP Fast Open. */
#undef USE_OSX_MSG_FASTOPEN
@@ -904,21 +1049,87 @@
#ifndef _ALL_SOURCE
# undef _ALL_SOURCE
#endif
/* Enable general extensions on macOS. */
#ifndef _DARWIN_C_SOURCE
# undef _DARWIN_C_SOURCE
#endif
/* Enable general extensions on Solaris. */
#ifndef __EXTENSIONS__
# undef __EXTENSIONS__
#endif
/* Enable GNU extensions on systems that have them. */
#ifndef _GNU_SOURCE
# undef _GNU_SOURCE
#endif
/* Enable threading extensions on Solaris. */
/* Enable X/Open compliant socket functions that do not require linking
with -lxnet on HP-UX 11.11. */
#ifndef _HPUX_ALT_XOPEN_SOCKET_API
# undef _HPUX_ALT_XOPEN_SOCKET_API
#endif
/* Identify the host operating system as Minix.
This macro does not affect the system headers' behavior.
A future release of Autoconf may stop defining this macro. */
#ifndef _MINIX
# undef _MINIX
#endif
/* Enable general extensions on NetBSD.
Enable NetBSD compatibility extensions on Minix. */
#ifndef _NETBSD_SOURCE
# undef _NETBSD_SOURCE
#endif
/* Enable OpenBSD compatibility extensions on NetBSD.
Oddly enough, this does nothing on OpenBSD. */
#ifndef _OPENBSD_SOURCE
# undef _OPENBSD_SOURCE
#endif
/* Define to 1 if needed for POSIX-compatible behavior. */
#ifndef _POSIX_SOURCE
# undef _POSIX_SOURCE
#endif
/* Define to 2 if needed for POSIX-compatible behavior. */
#ifndef _POSIX_1_SOURCE
# undef _POSIX_1_SOURCE
#endif
/* Enable POSIX-compatible threading on Solaris. */
#ifndef _POSIX_PTHREAD_SEMANTICS
# undef _POSIX_PTHREAD_SEMANTICS
#endif
/* Enable extensions specified by ISO/IEC TS 18661-5:2014. */
#ifndef __STDC_WANT_IEC_60559_ATTRIBS_EXT__
# undef __STDC_WANT_IEC_60559_ATTRIBS_EXT__
#endif
/* Enable extensions specified by ISO/IEC TS 18661-1:2014. */
#ifndef __STDC_WANT_IEC_60559_BFP_EXT__
# undef __STDC_WANT_IEC_60559_BFP_EXT__
#endif
/* Enable extensions specified by ISO/IEC TS 18661-2:2015. */
#ifndef __STDC_WANT_IEC_60559_DFP_EXT__
# undef __STDC_WANT_IEC_60559_DFP_EXT__
#endif
/* Enable extensions specified by ISO/IEC TS 18661-4:2015. */
#ifndef __STDC_WANT_IEC_60559_FUNCS_EXT__
# undef __STDC_WANT_IEC_60559_FUNCS_EXT__
#endif
/* Enable extensions specified by ISO/IEC TS 18661-3:2015. */
#ifndef __STDC_WANT_IEC_60559_TYPES_EXT__
# undef __STDC_WANT_IEC_60559_TYPES_EXT__
#endif
/* Enable extensions specified by ISO/IEC TR 24731-2:2010. */
#ifndef __STDC_WANT_LIB_EXT2__
# undef __STDC_WANT_LIB_EXT2__
#endif
/* Enable extensions specified by ISO/IEC 24747:2009. */
#ifndef __STDC_WANT_MATH_SPEC_FUNCS__
# undef __STDC_WANT_MATH_SPEC_FUNCS__
#endif
/* Enable extensions on HP NonStop. */
#ifndef _TANDEM_SOURCE
# undef _TANDEM_SOURCE
#endif
/* Enable general extensions on Solaris. */
#ifndef __EXTENSIONS__
# undef __EXTENSIONS__
/* Enable X/Open extensions. Define to 500 only if necessary
to make mbstate_t available. */
#ifndef _XOPEN_SOURCE
# undef _XOPEN_SOURCE
#endif
@@ -944,11 +1155,6 @@
`char[]'. */
#undef YYTEXT_POINTER
/* Enable large inode numbers on Mac OS X 10.5. */
#ifndef _DARWIN_USE_64_BIT_INODE
# define _DARWIN_USE_64_BIT_INODE 1
#endif
/* Number of bits in a file offset, on hosts where this is settable. */
#undef _FILE_OFFSET_BITS
@@ -958,18 +1164,12 @@
/* Define for large files, on AIX-style hosts. */
#undef _LARGE_FILES
/* Define to 1 if on MINIX. */
#undef _MINIX
/* Enable for compile on Minix */
#undef _NETBSD_SOURCE
/* Define to 2 if the system does not provide POSIX.1 features except with
this defined. */
#undef _POSIX_1_SOURCE
/* Define to 1 if you need to in order for `stat' and other things to work. */
#undef _POSIX_SOURCE
/* defined to use gcc ansi snprintf and sscanf that understands %lld when
compiled for windows. */
#undef __USE_MINGW_ANSI_STDIO
/* Define to empty if `const' does not conform to ANSI C. */
#undef const
@@ -1007,7 +1207,7 @@
/* Define to `long int' if <sys/types.h> does not define. */
#undef off_t
/* Define to `int' if <sys/types.h> does not define. */
/* Define as a signed integer type capable of holding a process identifier. */
#undef pid_t
/* Define to 'int' if not defined */
@@ -1045,39 +1245,39 @@
#if defined(OMITTED__D_GNU_SOURCE) && !defined(_GNU_SOURCE)
#define _GNU_SOURCE 1
#endif
#endif
#if defined(OMITTED__D_BSD_SOURCE) && !defined(_BSD_SOURCE)
#define _BSD_SOURCE 1
#endif
#endif
#if defined(OMITTED__D_DEFAULT_SOURCE) && !defined(_DEFAULT_SOURCE)
#define _DEFAULT_SOURCE 1
#endif
#endif
#if defined(OMITTED__D__EXTENSIONS__) && !defined(__EXTENSIONS__)
#define __EXTENSIONS__ 1
#endif
#endif
#if defined(OMITTED__D_POSIX_C_SOURCE_200112) && !defined(_POSIX_C_SOURCE)
#define _POSIX_C_SOURCE 200112
#endif
#endif
#if defined(OMITTED__D_XOPEN_SOURCE_600) && !defined(_XOPEN_SOURCE)
#define _XOPEN_SOURCE 600
#endif
#endif
#if defined(OMITTED__D_XOPEN_SOURCE_EXTENDED_1) && !defined(_XOPEN_SOURCE_EXTENDED)
#define _XOPEN_SOURCE_EXTENDED 1
#endif
#endif
#if defined(OMITTED__D_ALL_SOURCE) && !defined(_ALL_SOURCE)
#define _ALL_SOURCE 1
#endif
#endif
#if defined(OMITTED__D_LARGEFILE_SOURCE_1) && !defined(_LARGEFILE_SOURCE)
#define _LARGEFILE_SOURCE 1
#endif
#endif
@@ -1150,7 +1350,7 @@
#include <ws2tcpip.h>
#endif
#ifndef USE_WINSOCK
#if !defined(USE_WINSOCK) || !defined(HAVE_SNPRINTF) || defined(SNPRINTF_RET_BROKEN) || defined(__USE_MINGW_ANSI_STDIO)
#define ARG_LL "%ll"
#else
#define ARG_LL "%I64"
@@ -1161,7 +1361,7 @@
#endif
#ifdef HAVE_ATTR_FORMAT
# define ATTR_FORMAT(archetype, string_index, first_to_check) \
__attribute__ ((format (archetype, string_index, first_to_check)))
@@ -1271,7 +1471,7 @@ void* reallocarray(void *ptr, size_t nmemb, size_t size);
#ifdef HAVE_WINSOCK2_H
#define FD_SET_T (u_int)
#else
#define FD_SET_T
#define FD_SET_T
#endif
@@ -1399,6 +1599,10 @@ struct sockaddr_storage;
# define calloc(n,s) unbound_stat_calloc_log(n, s, __FILE__, __LINE__, __func__)
# define free(p) unbound_stat_free_log(p, __FILE__, __LINE__, __func__)
# define realloc(p,s) unbound_stat_realloc_log(p, s, __FILE__, __LINE__, __func__)
# define strdup(s) unbound_stat_strdup_log(s, __FILE__, __LINE__, __func__)
#ifdef HAVE_REALLOCARRAY
# define reallocarray(p,n,s) unbound_stat_reallocarray_log(p, n, s, __FILE__, __LINE__, __func__)
#endif
void *unbound_stat_malloc(size_t size);
void *unbound_stat_calloc(size_t nmemb, size_t size);
void unbound_stat_free(void *ptr);
@@ -1411,6 +1615,10 @@ void unbound_stat_free_log(void *ptr, const char* file, int line,
const char* func);
void *unbound_stat_realloc_log(void *ptr, size_t size, const char* file,
int line, const char* func);
void *unbound_stat_reallocarray_log(void *ptr, size_t nmemb, size_t size,
const char* file, int line, const char* func);
char *unbound_stat_strdup_log(const char *s, const char* file, int line,
const char* func);
#elif defined(UNBOUND_ALLOC_LITE)
# include "util/alloc.h"
#endif /* UNBOUND_ALLOC_LITE and UNBOUND_ALLOC_STATS */
@@ -1421,6 +1629,8 @@ void *unbound_stat_realloc_log(void *ptr, size_t size, const char* file,
#define UNBOUND_DNS_OVER_TLS_PORT 853
/** default port for DNS over HTTPS traffic. */
#define UNBOUND_DNS_OVER_HTTPS_PORT 443
/** default port for DNS over QUIC traffic. */
#define UNBOUND_DNS_OVER_QUIC_PORT 853
/** default port for unbound control traffic, registered port with IANA,
ub-dns-control 8953/tcp unbound dns nameserver control */
#define UNBOUND_CONTROL_PORT 8953
Vendored
+503 -325
View File
File diff suppressed because it is too large Load Diff
Vendored
+8488 -5066
View File
File diff suppressed because it is too large Load Diff
+307 -76
View File
@@ -2,23 +2,24 @@
# Process this file with autoconf to produce a configure script.
AC_PREREQ([2.56])
sinclude(acx_nlnetlabs.m4)
sinclude(ax_build_date_epoch.m4)
sinclude(ax_pthread.m4)
sinclude(acx_python.m4)
sinclude(ac_pkg_swig.m4)
sinclude(ax_pkg_swig.m4)
sinclude(dnstap/dnstap.m4)
sinclude(dnscrypt/dnscrypt.m4)
# must be numbers. ac_defun because of later processing
m4_define([VERSION_MAJOR],[1])
m4_define([VERSION_MINOR],[15])
m4_define([VERSION_MICRO],[0])
m4_define([VERSION_MINOR],[24])
m4_define([VERSION_MICRO],[2])
AC_INIT([unbound],m4_defn([VERSION_MAJOR]).m4_defn([VERSION_MINOR]).m4_defn([VERSION_MICRO]),[unbound-bugs@nlnetlabs.nl or https://github.com/NLnetLabs/unbound/issues],[unbound])
AC_SUBST(UNBOUND_VERSION_MAJOR, [VERSION_MAJOR])
AC_SUBST(UNBOUND_VERSION_MINOR, [VERSION_MINOR])
AC_SUBST(UNBOUND_VERSION_MICRO, [VERSION_MICRO])
LIBUNBOUND_CURRENT=9
LIBUNBOUND_REVISION=15
LIBUNBOUND_REVISION=34
LIBUNBOUND_AGE=1
# 1.0.0 had 0:12:0
# 1.0.1 had 0:13:0
@@ -101,6 +102,26 @@ LIBUNBOUND_AGE=1
# 1.13.2 had 9:13:1
# 1.14.0 had 9:14:1
# 1.15.0 had 9:15:1
# 1.16.0 had 9:16:1
# 1.16.1 had 9:17:1
# 1.16.2 had 9:18:1
# 1.16.3 had 9:19:1
# 1.17.0 had 9:20:1
# 1.17.1 had 9:21:1
# 1.18.0 had 9:22:1
# 1.19.0 had 9:23:1
# 1.19.1 had 9:24:1
# 1.19.2 had 9:25:1
# 1.19.3 had 9:26:1
# 1.20.0 had 9:27:1
# 1.21.0 had 9:28:1
# 1.21.1 had 9:29:1
# 1.22.0 had 9:30:1
# 1.23.0 had 9:31:1
# 1.23.1 had 9:32:1
# 1.24.0 had 9:33:1
# 1.24.1 had 9:34:1
# 1.24.2 had 9:35:1
# Current -- the number of the binary API that we're implementing
# Revision -- which iteration of the implementation of the binary
@@ -150,7 +171,7 @@ esac
# are we on MinGW?
if uname -s 2>&1 | grep MINGW >/dev/null; then on_mingw="yes"
else
else
if echo $host | grep mingw >/dev/null; then on_mingw="yes"
else on_mingw="no"; fi
fi
@@ -179,9 +200,9 @@ ub_conf_dir=`AS_DIRNAME(["$ub_conf_file"])`
AC_SUBST(ub_conf_dir)
# Determine run, chroot directory and pidfile locations
AC_ARG_WITH(run-dir,
AS_HELP_STRING([--with-run-dir=path],[set default directory to chdir to (by default dir part of cfg file)]),
UNBOUND_RUN_DIR="$withval",
AC_ARG_WITH(run-dir,
AS_HELP_STRING([--with-run-dir=path],[set default directory to chdir to (by default dir part of cfg file)]),
UNBOUND_RUN_DIR="$withval",
if test $on_mingw = no; then
UNBOUND_RUN_DIR=`dirname "$ub_conf_file"`
else
@@ -192,9 +213,9 @@ AC_SUBST(UNBOUND_RUN_DIR)
ACX_ESCAPE_BACKSLASH($UNBOUND_RUN_DIR, hdr_run)
AC_DEFINE_UNQUOTED(RUN_DIR, ["$hdr_run"], [Directory to chdir to])
AC_ARG_WITH(chroot-dir,
AS_HELP_STRING([--with-chroot-dir=path],[set default directory to chroot to (by default same as run-dir)]),
UNBOUND_CHROOT_DIR="$withval",
AC_ARG_WITH(chroot-dir,
AS_HELP_STRING([--with-chroot-dir=path],[set default directory to chroot to (by default same as run-dir)]),
UNBOUND_CHROOT_DIR="$withval",
if test $on_mingw = no; then
UNBOUND_CHROOT_DIR="$UNBOUND_RUN_DIR"
else
@@ -212,9 +233,9 @@ AC_ARG_WITH(share-dir,
AC_SUBST(UNBOUND_SHARE_DIR)
AC_DEFINE_UNQUOTED(SHARE_DIR, ["$UNBOUND_SHARE_DIR"], [Shared data])
AC_ARG_WITH(pidfile,
AS_HELP_STRING([--with-pidfile=filename],[set default pathname to unbound pidfile (default run-dir/unbound.pid)]),
UNBOUND_PIDFILE="$withval",
AC_ARG_WITH(pidfile,
AS_HELP_STRING([--with-pidfile=filename],[set default pathname to unbound pidfile (default run-dir/unbound.pid)]),
UNBOUND_PIDFILE="$withval",
if test $on_mingw = no; then
UNBOUND_PIDFILE="$UNBOUND_RUN_DIR/unbound.pid"
else
@@ -225,9 +246,9 @@ AC_SUBST(UNBOUND_PIDFILE)
ACX_ESCAPE_BACKSLASH($UNBOUND_PIDFILE, hdr_pid)
AC_DEFINE_UNQUOTED(PIDFILE, ["$hdr_pid"], [default pidfile location])
AC_ARG_WITH(rootkey-file,
AS_HELP_STRING([--with-rootkey-file=filename],[set default pathname to root key file (default run-dir/root.key). This file is read and written.]),
UNBOUND_ROOTKEY_FILE="$withval",
AC_ARG_WITH(rootkey-file,
AS_HELP_STRING([--with-rootkey-file=filename],[set default pathname to root key file (default run-dir/root.key). This file is read and written.]),
UNBOUND_ROOTKEY_FILE="$withval",
if test $on_mingw = no; then
UNBOUND_ROOTKEY_FILE="$UNBOUND_RUN_DIR/root.key"
else
@@ -238,9 +259,9 @@ AC_SUBST(UNBOUND_ROOTKEY_FILE)
ACX_ESCAPE_BACKSLASH($UNBOUND_ROOTKEY_FILE, hdr_rkey)
AC_DEFINE_UNQUOTED(ROOT_ANCHOR_FILE, ["$hdr_rkey"], [default rootkey location])
AC_ARG_WITH(rootcert-file,
AS_HELP_STRING([--with-rootcert-file=filename],[set default pathname to root update certificate file (default run-dir/icannbundle.pem). This file need not exist if you are content with the builtin.]),
UNBOUND_ROOTCERT_FILE="$withval",
AC_ARG_WITH(rootcert-file,
AS_HELP_STRING([--with-rootcert-file=filename],[set default pathname to root update certificate file (default run-dir/icannbundle.pem). This file need not exist if you are content with the builtin.]),
UNBOUND_ROOTCERT_FILE="$withval",
if test $on_mingw = no; then
UNBOUND_ROOTCERT_FILE="$UNBOUND_RUN_DIR/icannbundle.pem"
else
@@ -251,9 +272,9 @@ AC_SUBST(UNBOUND_ROOTCERT_FILE)
ACX_ESCAPE_BACKSLASH($UNBOUND_ROOTCERT_FILE, hdr_rpem)
AC_DEFINE_UNQUOTED(ROOT_CERT_FILE, ["$hdr_rpem"], [default rootcert location])
AC_ARG_WITH(username,
AS_HELP_STRING([--with-username=user],[set default user that unbound changes to (default user is unbound)]),
UNBOUND_USERNAME="$withval",
AC_ARG_WITH(username,
AS_HELP_STRING([--with-username=user],[set default user that unbound changes to (default user is unbound)]),
UNBOUND_USERNAME="$withval",
UNBOUND_USERNAME="unbound")
AC_SUBST(UNBOUND_USERNAME)
AC_DEFINE_UNQUOTED(UB_USERNAME, ["$UNBOUND_USERNAME"], [default username])
@@ -262,6 +283,9 @@ AC_DEFINE(WINVER, 0x0502, [the version of the windows API enabled])
ACX_RSRC_VERSION(wnvs)
AC_DEFINE_UNQUOTED(RSRC_PACKAGE_VERSION, [$wnvs], [version number for resource files])
# Check for 'grep -e' program, here, since ACX_CHECK_FLTO needs that.
AC_PROG_GREP
# Checks for typedefs, structures, and compiler characteristics.
AC_C_CONST
AC_LANG([C])
@@ -272,14 +296,14 @@ ACX_CHECK_COMPILER_FLAG(g, [CFLAGS="$CFLAGS -g"])
ACX_CHECK_COMPILER_FLAG(O2, [CFLAGS="$CFLAGS -O2"])
default_cflags=yes
fi
AC_PROG_CC
m4_version_prereq([2.70], [AC_PROG_CC], [AC_PROG_CC_STDC])
ACX_DEPFLAG
ACX_DETERMINE_EXT_FLAGS_UNBOUND
# debug mode flags warnings
AC_ARG_ENABLE(checking, AS_HELP_STRING([--enable-checking],[Enable warnings, asserts, makefile-dependencies]))
AC_ARG_ENABLE(debug, AS_HELP_STRING([--enable-debug],[same as enable-checking]))
if test "$enable_debug" = "yes"; then debug_enabled="$enable_debug";
if test "$enable_debug" = "yes"; then debug_enabled="$enable_debug";
else debug_enabled="$enable_checking"; fi
AC_SUBST(debug_enabled)
case "$debug_enabled" in
@@ -323,6 +347,8 @@ AC_MSG_RESULT($ac_cv_c_weak_attribute)
if test $ac_cv_c_weak_attribute = yes; then
AC_DEFINE(HAVE_ATTR_WEAK, 1, [Whether the C compiler accepts the "weak" attribute])
AC_DEFINE(ATTR_WEAK, [__attribute__((weak))], [apply the weak attribute to a symbol])
else
AC_DEFINE(ATTR_WEAK,[], [apply the weak attribute to a symbol])
fi
])dnl End of CHECK_WEAK_ATTRIBUTE
@@ -344,11 +370,54 @@ AC_MSG_RESULT($ac_cv_c_noreturn_attribute)
if test $ac_cv_c_noreturn_attribute = yes; then
AC_DEFINE(HAVE_ATTR_NORETURN, 1, [Whether the C compiler accepts the "noreturn" attribute])
AC_DEFINE(ATTR_NORETURN, [__attribute__((__noreturn__))], [apply the noreturn attribute to a function that exits the program])
else
AC_DEFINE(ATTR_NORETURN,[], [apply the noreturn attribute to a function that exits the program])
fi
])dnl End of CHECK_NORETURN_ATTRIBUTE
CHECK_NORETURN_ATTRIBUTE
AC_DEFUN([CHECK_FALLTHROUGH_ATTRIBUTE],
[AC_REQUIRE([AC_PROG_CC])
AC_MSG_CHECKING(whether the C compiler (${CC-cc}) accepts the "fallthrough" attribute)
BAKCFLAGS="$CFLAGS"
CFLAGS="$CFLAGS -Werror"
AC_CACHE_VAL(ac_cv_c_fallthrough_attribute,
[ac_cv_c_fallthrough_attribute=no
AC_COMPILE_IFELSE([AC_LANG_PROGRAM([[ #include <stdio.h>
void f(int x) {
int y = 0;
switch(x) {
case 1:
y = 1;
__attribute__((fallthrough));
/* fallthrough */
case 2:
y++;
break;
case 3:
y = 3;
break;
}
printf("%d", y);
}
]], [[
f(1);
]])],[ac_cv_c_fallthrough_attribute="yes"],[ac_cv_c_fallthrough_attribute="no"])
])
CFLAGS="$BAKCFLAGS"
AC_MSG_RESULT($ac_cv_c_fallthrough_attribute)
if test $ac_cv_c_fallthrough_attribute = yes; then
AC_DEFINE(HAVE_ATTR_FALLTHROUGH, 1, [Whether the C compiler accepts the "fallthrough" attribute])
AC_DEFINE(ATTR_FALLTHROUGH, [__attribute__((fallthrough));], [apply the fallthrough attribute.])
else
AC_DEFINE(ATTR_FALLTHROUGH,[], [apply the fallthrough attribute.])
fi
])dnl End of CHECK_FALLTHROUGH_ATTRIBUTE
CHECK_FALLTHROUGH_ATTRIBUTE
if test "$srcdir" != "."; then
CPPFLAGS="$CPPFLAGS -I$srcdir"
fi
@@ -382,18 +451,28 @@ fi
if test "$LEX" != "" -a "$LEX" != ":"; then
ACX_YYLEX_OPTION
fi
if test "$LEX" = "" -o "$LEX" = ":"; then
if test ! -f util/configlexer.c; then
AC_MSG_ERROR([no lex and no util/configlexer.c: need flex and bison to compile from source repository.])
fi
fi
AC_PROG_YACC
if test "$YACC" = "" -o "$YACC" = ":"; then
if test ! -f util/configparser.c; then
AC_MSG_ERROR([no yacc and no util/configparser.c: need flex and bison to compile from source repository.])
fi
fi
AC_CHECK_PROG(doxygen, doxygen, doxygen)
AC_CHECK_TOOL(STRIP, strip)
ACX_LIBTOOL_C_ONLY
# pkg-config is only needed for these options, do not require it otherwise
if test "$enable_systemd" = "yes" -o "$with_pyunbound" = "yes" -o "$with_pythonmod" = "yes"; then
if test "$enable_systemd" = "yes" -o "$enable_dnstap" = "yes" -o "$with_pyunbound" = "yes" -o "$with_pythonmod" = "yes"; then
PKG_PROG_PKG_CONFIG
fi
# Checks for header files.
AC_CHECK_HEADERS([stdarg.h stdbool.h netinet/in.h netinet/tcp.h sys/param.h sys/select.h sys/socket.h sys/un.h sys/uio.h sys/resource.h arpa/inet.h syslog.h netdb.h sys/wait.h pwd.h glob.h grp.h login_cap.h winsock2.h ws2tcpip.h endian.h sys/endian.h libkern/OSByteOrder.h sys/ipc.h sys/shm.h ifaddrs.h],,, [AC_INCLUDES_DEFAULT])
AC_CHECK_HEADERS([stdarg.h stdbool.h netinet/in.h netinet/tcp.h sys/param.h sys/select.h sys/socket.h sys/un.h sys/uio.h sys/resource.h arpa/inet.h syslog.h netdb.h sys/wait.h pwd.h glob.h grp.h login_cap.h winsock2.h ws2tcpip.h endian.h sys/endian.h libkern/OSByteOrder.h sys/ipc.h sys/shm.h ifaddrs.h poll.h],,, [AC_INCLUDES_DEFAULT])
# net/if.h portability for Darwin see:
# https://www.gnu.org/software/autoconf/manual/autoconf-2.69/html_node/Header-Portability.html
AC_CHECK_HEADERS([net/if.h],,, [
@@ -447,6 +526,10 @@ AC_CHECK_HEADERS([netioapi.h],,, [AC_INCLUDES_DEFAULT
#endif
])
# Check for Linux timestamping headers
AC_CHECK_HEADERS([linux/net_tstamp.h],,, [AC_INCLUDES_DEFAULT])
AC_CHECK_HEADERS([stdatomic.h],,, [AC_INCLUDES_DEFAULT])
# check for types.
# Using own tests for int64* because autoconf builtin only give 32bit.
AC_CHECK_TYPE(int8_t, signed char)
@@ -505,7 +588,8 @@ res = strptime("2010-07-15T00:00:00+00:00", "%t%Y%t-%t%m%t-%t%d%tT%t%H%t:%t%M%t:
if (!res) return 2;
res = strptime("20070207111842", "%Y%m%d%H%M%S", &tm);
if (!res) return 1; return 0; }
]])] , [eval "ac_cv_c_strptime_works=yes"], [eval "ac_cv_c_strptime_works=no"])
]])] , [eval "ac_cv_c_strptime_works=yes"], [eval "ac_cv_c_strptime_works=no"],
[eval "ac_cv_c_strptime_works=maybe"])
else
eval "ac_cv_c_strptime_works=maybe"
fi
@@ -542,11 +626,11 @@ sinclude(systemd.m4)
# Include systemd.m4 - end
# set memory allocation checking if requested
AC_ARG_ENABLE(alloc-checks, AS_HELP_STRING([--enable-alloc-checks],[ enable to memory allocation statistics, for debug purposes ]),
AC_ARG_ENABLE(alloc-checks, AS_HELP_STRING([--enable-alloc-checks],[ enable to memory allocation statistics, for debug purposes ]),
, )
AC_ARG_ENABLE(alloc-lite, AS_HELP_STRING([--enable-alloc-lite],[ enable for lightweight alloc assertions, for debug purposes ]),
AC_ARG_ENABLE(alloc-lite, AS_HELP_STRING([--enable-alloc-lite],[ enable for lightweight alloc assertions, for debug purposes ]),
, )
AC_ARG_ENABLE(alloc-nonregional, AS_HELP_STRING([--enable-alloc-nonregional],[ enable nonregional allocs, slow but exposes regional allocations to other memory purifiers, for debug purposes ]),
AC_ARG_ENABLE(alloc-nonregional, AS_HELP_STRING([--enable-alloc-nonregional],[ enable nonregional allocs, slow but exposes regional allocations to other memory purifiers, for debug purposes ]),
, )
if test x_$enable_alloc_nonregional = x_yes; then
AC_DEFINE(UNBOUND_ALLOC_NONREGIONAL, 1, [use malloc not regions, for debug use])
@@ -579,7 +663,7 @@ if test "$on_mingw" = "yes"; then
])],
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_WINDOWS_THREADS, 1, [Using Windows threads])
,
,
AC_MSG_RESULT(no)
)
@@ -590,7 +674,7 @@ else
# check this first, so that the pthread lib does not get linked in via
# libssl or libpython, and thus distorts the tests, and we end up using
# the non-threadsafe C libraries.
AC_ARG_WITH(pthreads, AS_HELP_STRING([--with-pthreads],[use pthreads library, or --without-pthreads to disable threading support.]),
AC_ARG_WITH(pthreads, AS_HELP_STRING([--with-pthreads],[use pthreads library, or --without-pthreads to disable threading support.]),
[ ],[ withval="yes" ])
ub_have_pthreads=no
if test x_$withval != x_no; then
@@ -603,6 +687,8 @@ if test x_$withval != x_no; then
CC="$PTHREAD_CC"
ub_have_pthreads=yes
AC_CHECK_TYPES([pthread_spinlock_t, pthread_rwlock_t],,,[#include <pthread.h>])
AC_CHECK_SIZEOF([unsigned long])
AC_CHECK_SIZEOF(pthread_t)
if echo "$CFLAGS" | $GREP -e "-pthread" >/dev/null; then
AC_MSG_CHECKING([if -pthread unused during linking])
@@ -615,7 +701,7 @@ int main(void) {return 0;}
# first compile
echo "$CC $CFLAGS -c conftest.c -o conftest.o" >&AS_MESSAGE_LOG_FD
$CC $CFLAGS -c conftest.c -o conftest.o 2>&AS_MESSAGE_LOG_FD >&AS_MESSAGE_LOG_FD
if test $? = 0; then
if test $? = 0; then
# then link
echo "$CC $CFLAGS -Werror $LDFLAGS $LIBS -o conftest contest.o" >&AS_MESSAGE_LOG_FD
$CC $CFLAGS -Werror $LDFLAGS $LIBS -o conftest conftest.o 2>&AS_MESSAGE_LOG_FD >&AS_MESSAGE_LOG_FD
@@ -636,7 +722,7 @@ int main(void) {return 0;}
])
fi
# check solaris thread library
# check solaris thread library
AC_ARG_WITH(solaris-threads, AS_HELP_STRING([--with-solaris-threads],[use solaris native thread library.]), [ ],[ withval="no" ])
ub_have_sol_threads=no
if test x_$withval != x_no; then
@@ -650,8 +736,8 @@ if test x_$withval != x_no; then
ACX_CHECK_COMPILER_FLAG(mt, [CFLAGS="$CFLAGS -mt"],
[CFLAGS="$CFLAGS -D_REENTRANT"])
ub_have_sol_threads=yes
] , [
AC_MSG_ERROR([no solaris threads found.])
] , [
AC_MSG_ERROR([no solaris threads found.])
])
fi
fi
@@ -726,7 +812,14 @@ if test x_$ub_test_python != x_no; then
ac_save_LIBS="$LIBS" dnl otherwise AC_PYTHON_DEVEL thrashes $LIBS
AC_PYTHON_DEVEL
if test ! -z "$PYTHON_VERSION"; then
if test `$PYTHON -c "print('$PYTHON_VERSION' >= '2.4.0')"` = "False"; then
badversion="no"
if test "$PYTHON_VERSION_MAJOR" -lt 2; then
badversion="yes"
fi
if test "$PYTHON_VERSION_MAJOR" -eq 2 -a "$PYTHON_VERSION_MINOR" -lt 4; then
badversion="yes"
fi
if test "$badversion" = "yes"; then
AC_MSG_ERROR([Python version >= 2.4.0 is required])
fi
@@ -761,9 +854,9 @@ if test x_$ub_test_python != x_no; then
ub_have_swig=no
AC_ARG_ENABLE(swig-version-check, AS_HELP_STRING([--disable-swig-version-check],[Disable swig version check to build python modules with older swig even though that is unreliable]))
if test "$enable_swig_version_check" = "yes"; then
AC_PROG_SWIG(2.0.1)
AX_PKG_SWIG(2.0.1)
else
AC_PROG_SWIG
AX_PKG_SWIG
fi
AC_MSG_CHECKING(SWIG)
if test ! -x "$SWIG"; then
@@ -821,7 +914,8 @@ if test "`uname`" = "Linux"; then
GCC_DOCKER_LINTFLAGS='-syntax'
AC_SUBST(GCC_DOCKER_LINTFLAGS)
fi
CONFIG_DATE=`date +%Y%m%d`
AX_BUILD_DATE_EPOCH(CONFIG_DATE, [%Y%m%d])
AC_ARG_VAR(SOURCE_DATE_EPOCH, [If it is set, it uses the value of that variable instead of the current time as the build timestamp. The format is a unix timestamp. This enables reproducible build output.])
AC_SUBST(CONFIG_DATE)
# Checks for libraries.
@@ -905,7 +999,7 @@ else
AC_MSG_RESULT([no])
fi
AC_CHECK_HEADERS([openssl/conf.h openssl/engine.h openssl/bn.h openssl/dh.h openssl/dsa.h openssl/rsa.h openssl/core_names.h openssl/param_build.h],,, [AC_INCLUDES_DEFAULT])
AC_CHECK_FUNCS([OPENSSL_config EVP_sha1 EVP_sha256 EVP_sha512 FIPS_mode EVP_MD_CTX_new OpenSSL_add_all_digests OPENSSL_init_crypto EVP_cleanup ENGINE_cleanup ERR_load_crypto_strings CRYPTO_cleanup_all_ex_data ERR_free_strings RAND_cleanup DSA_SIG_set0 EVP_dss1 EVP_DigestVerify EVP_aes_256_cbc EVP_EncryptInit_ex HMAC_Init_ex CRYPTO_THREADID_set_callback EVP_MAC_CTX_set_params OSSL_PARAM_BLD_new BIO_set_callback_ex])
AC_CHECK_FUNCS([OPENSSL_config EVP_sha1 EVP_sha256 EVP_sha512 FIPS_mode EVP_default_properties_is_fips_enabled EVP_MD_CTX_new OpenSSL_add_all_digests OPENSSL_init_crypto EVP_cleanup ENGINE_cleanup ERR_load_crypto_strings CRYPTO_cleanup_all_ex_data ERR_free_strings RAND_cleanup DSA_SIG_set0 EVP_dss1 EVP_DigestVerify EVP_aes_256_cbc EVP_EncryptInit_ex HMAC_Init_ex CRYPTO_THREADID_set_callback EVP_MAC_CTX_set_params OSSL_PARAM_BLD_new BIO_set_callback_ex])
# these check_funcs need -lssl
BAKLIBS="$LIBS"
@@ -913,7 +1007,7 @@ LIBS="-lssl $LIBS"
AC_CHECK_FUNCS([OPENSSL_init_ssl SSL_CTX_set_security_level SSL_set1_host SSL_get0_peername X509_VERIFY_PARAM_set1_host SSL_CTX_set_ciphersuites SSL_CTX_set_tlsext_ticket_key_evp_cb SSL_CTX_set_alpn_select_cb SSL_get0_alpn_selected SSL_CTX_set_alpn_protos SSL_get1_peer_certificate])
LIBS="$BAKLIBS"
AC_CHECK_DECLS([SSL_COMP_get_compression_methods,sk_SSL_COMP_pop_free,SSL_CTX_set_ecdh_auto], [], [], [
AC_CHECK_DECLS([SSL_COMP_get_compression_methods,sk_SSL_COMP_pop_free,SSL_CTX_set_ecdh_auto,SSL_CTX_set_tmp_ecdh], [], [], [
AC_INCLUDES_DEFAULT
#ifdef HAVE_OPENSSL_ERR_H
#include <openssl/err.h>
@@ -1077,7 +1171,7 @@ int load_gost_id(void)
EVP_PKEY_asn1_get0_info(&gost_id, NULL, NULL, NULL, NULL, meth);
return gost_id;
}
int main(void) {
int main(void) {
EVP_MD_CTX* ctx;
const EVP_MD* md;
unsigned char digest[64]; /* its a 256-bit digest, so uses 32 bytes */
@@ -1108,7 +1202,8 @@ int main(void) {
return 6;
return 0;
}
]])] , [eval "ac_cv_c_gost_works=yes"], [eval "ac_cv_c_gost_works=no"])
]])] , [eval "ac_cv_c_gost_works=yes"], [eval "ac_cv_c_gost_works=no"],
[eval "ac_cv_c_gost_works=maybe"])
CFLAGS="$BAKCFLAGS"
else
eval "ac_cv_c_gost_works=maybe"
@@ -1141,8 +1236,17 @@ case "$enable_ecdsa" in
;;
*)
if test $USE_NSS = "no" -a $USE_NETTLE = "no"; then
AC_CHECK_FUNC(ECDSA_sign, [], [AC_MSG_ERROR([OpenSSL does not support ECDSA: please upgrade or rerun with --disable-ecdsa])])
AC_CHECK_FUNC(SHA384_Init, [], [AC_MSG_ERROR([OpenSSL does not support SHA384: please upgrade or rerun with --disable-ecdsa])])
AC_CHECK_FUNC(EVP_PKEY_fromdata, [
# with EVP_PKEY_fromdata, check if EC is not disabled
AC_CHECK_DECL([OPENSSL_NO_EC], [AC_MSG_ERROR([OpenSSL does not support ECDSA: please upgrade or rerun with --disable-ecdsa])
], [], [AC_INCLUDES_DEFAULT
#include <openssl/evp.h>
])
], [
# without EVP_PKEY_fromdata, older openssl, check for support
AC_CHECK_FUNC(ECDSA_sign, [], [AC_MSG_ERROR([OpenSSL does not support ECDSA: please upgrade or rerun with --disable-ecdsa])])
AC_CHECK_FUNC(SHA384_Init, [], [AC_MSG_ERROR([OpenSSL does not support SHA384: please upgrade or rerun with --disable-ecdsa])])
])
AC_CHECK_DECLS([NID_X9_62_prime256v1, NID_secp384r1], [], [AC_MSG_ERROR([OpenSSL does not support the ECDSA curves: please upgrade or rerun with --disable-ecdsa])], [AC_INCLUDES_DEFAULT
#include <openssl/evp.h>
])
@@ -1483,6 +1587,76 @@ if test x_$withval = x_yes -o x_$withval != x_no; then
])
fi
# ngtcp2
AC_ARG_WITH(libngtcp2, AS_HELP_STRING([--with-libngtcp2=path],[specify explicit path for libngtcp2, for QUIC.]),
[ ],[ withval="no" ])
found_libngtcp2="no"
if test x_$withval = x_yes -o x_$withval != x_no; then
AC_MSG_CHECKING(for libngtcp2)
if test x_$withval = x_ -o x_$withval = x_yes; then
withval="/usr/local /opt/local /usr/lib /usr/pkg /usr/sfw /usr"
fi
for dir in $withval ; do
if test -f "$dir/include/ngtcp2/ngtcp2.h"; then
found_libngtcp2="yes"
dnl assume /usr is in default path.
if test "$dir" != "/usr"; then
CPPFLAGS="$CPPFLAGS -I$dir/include"
LDFLAGS="$LDFLAGS -L$dir/lib"
fi
AC_MSG_RESULT(found in $dir)
AC_DEFINE([HAVE_NGTCP2], [1], [Define this to use ngtcp2.])
LIBS="$LIBS -lngtcp2"
break;
fi
done
if test x_$found_libngtcp2 != x_yes; then
AC_MSG_ERROR([Could not find libngtcp2, ngtcp2.h])
fi
AC_CHECK_HEADERS([ngtcp2/ngtcp2.h ngtcp2/ngtcp2_crypto_ossl.h ngtcp2/ngtcp2_crypto_openssl.h ngtcp2/ngtcp2_crypto_quictls.h],,, [AC_INCLUDES_DEFAULT])
AC_CHECK_DECLS([ngtcp2_conn_server_new], [], [], [AC_INCLUDES_DEFAULT
#include <ngtcp2/ngtcp2.h>
])
AC_CHECK_DECLS([ngtcp2_crypto_encrypt_cb], [], [], [AC_INCLUDES_DEFAULT
#include <ngtcp2/ngtcp2_crypto.h>
])
AC_CHECK_LIB([ngtcp2_crypto_ossl], [ngtcp2_crypto_encrypt_cb], [
LIBS="$LIBS -lngtcp2_crypto_ossl"
AC_DEFINE(USE_NGTCP2_CRYPTO_OSSL, 1, [Define this to use ngtcp2_crypto_ossl.])
], [
AC_CHECK_LIB([ngtcp2_crypto_openssl], [ngtcp2_crypto_encrypt_cb], [ LIBS="$LIBS -lngtcp2_crypto_openssl" ], [
AC_CHECK_LIB([ngtcp2_crypto_quictls], [ngtcp2_crypto_encrypt_cb], [ LIBS="$LIBS -lngtcp2_crypto_quictls" ])
])
])
AC_CHECK_FUNCS([ngtcp2_crypto_encrypt_cb ngtcp2_ccerr_default ngtcp2_conn_in_closing_period ngtcp2_conn_in_draining_period ngtcp2_conn_get_max_local_streams_uni ngtcp2_crypto_quictls_from_ossl_encryption_level ngtcp2_crypto_quictls_configure_server_context ngtcp2_crypto_quictls_configure_client_context ngtcp2_crypto_quictls_init ngtcp2_conn_get_num_scid ngtcp2_conn_tls_early_data_rejected ngtcp2_conn_encode_0rtt_transport_params])
# these check_funcs need -lssl
BAKLIBS="$LIBS"
LIBS="-lssl $LIBS"
AC_CHECK_FUNCS([SSL_is_quic], [], [AC_MSG_ERROR([No QUIC support detected in OpenSSL. Need OpenSSL version with QUIC support to enable DNS over QUIC with libngtcp2.])])
LIBS="$BAKLIBS"
AC_CHECK_TYPES([struct ngtcp2_version_cid, ngtcp2_encryption_level],,,[AC_INCLUDES_DEFAULT
#include <ngtcp2/ngtcp2.h>
])
AC_CHECK_MEMBERS([struct ngtcp2_pkt_hd.tokenlen, struct ngtcp2_settings.tokenlen, struct ngtcp2_settings.max_tx_udp_payload_size, struct ngtcp2_transport_params.original_dcid_present],,,[AC_INCLUDES_DEFAULT
#include <ngtcp2/ngtcp2.h>
])
AC_MSG_CHECKING([whether ngtcp2_conn_shutdown_stream has 4 arguments])
AC_COMPILE_IFELSE([AC_LANG_PROGRAM([AC_INCLUDES_DEFAULT
#include <ngtcp2/ngtcp2.h>
],[
(void)ngtcp2_conn_shutdown_stream(NULL, 0, 0, 0);
])],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_NGTCP2_CONN_SHUTDOWN_STREAM4, 1, [Define if ngtcp2_conn_shutdown_stream has 4 arguments.])
],[
AC_MSG_RESULT(no)
])
fi
# set static linking for uninstalled libraries if requested
AC_SUBST(staticexe)
staticexe=""
@@ -1493,13 +1667,17 @@ if test x_$enable_static_exe = x_yes; then
if test "$on_mingw" = yes; then
staticexe="-all-static"
# for static compile, include gdi32 and zlib here.
if echo $LIBS | grep 'lgdi32' >/dev/null; then
if echo "$LIBS" | grep 'lgdi32' >/dev/null; then
:
else
LIBS="$LIBS -lgdi32"
fi
LIBS="$LIBS -lz"
LIBS="$LIBS -l:libssp.a"
AC_CHECK_LIB([z], [compress], [ LIBS="$LIBS -lz" ])
if echo "$LIBS" | grep -e "libssp.a" -e "lssp" >/dev/null; then
:
else
LIBS="$LIBS -l:libssp.a"
fi
fi
fi
@@ -1515,13 +1693,17 @@ if test x_$enable_fully_static = x_yes; then
else
LIBS="$LIBS -lgdi32"
fi
LIBS="$LIBS -lz"
LIBS="$LIBS -l:libssp.a"
AC_CHECK_LIB([z], [compress], [ LIBS="$LIBS -lz" ])
if echo "$LIBS" | grep -e "libssp.a" -e "lssp" >/dev/null; then
:
else
LIBS="$LIBS -l:libssp.a"
fi
fi
fi
# set lock checking if requested
AC_ARG_ENABLE(lock_checks, AS_HELP_STRING([--enable-lock-checks],[ enable to check lock and unlock calls, for debug purposes ]),
AC_ARG_ENABLE(lock_checks, AS_HELP_STRING([--enable-lock-checks],[ enable to check lock and unlock calls, for debug purposes ]),
, )
if test x_$enable_lock_checks = x_yes; then
AC_DEFINE(ENABLE_LOCK_CHECKS, 1, [Define if you want to use debug lock checking (slow).])
@@ -1536,7 +1718,11 @@ if test "$USE_WINSOCK" = 1; then
#include <windows.h>
])
AC_CHECK_TOOL(WINDRES, windres)
LIBS="$LIBS -liphlpapi -lcrypt32"
if echo "$LIBS" | grep crypt32 >/dev/null; then
LIBS="$LIBS -liphlpapi"
else
LIBS="$LIBS -liphlpapi -lcrypt32"
fi
WINAPPS="unbound-service-install.exe unbound-service-remove.exe anchor-update.exe"
AC_SUBST(WINAPPS)
WIN_DAEMON_SRC="winrc/win_svc.c winrc/w_inst.c"
@@ -1553,6 +1739,7 @@ if test "$USE_WINSOCK" = 1; then
AC_SUBST(WIN_CONTROL_OBJ_LINK)
WIN_CHECKCONF_OBJ_LINK="rsrc_unbound_checkconf.o"
AC_SUBST(WIN_CHECKCONF_OBJ_LINK)
AC_DEFINE(__USE_MINGW_ANSI_STDIO, 1, [defined to use gcc ansi snprintf and sscanf that understands %lld when compiled for windows.])
fi
if test $ac_cv_func_getaddrinfo = no; then
AC_LIBOBJ([fake-rfc2553])
@@ -1639,7 +1826,7 @@ AC_LINK_IFELSE([AC_LANG_PROGRAM([
AC_MSG_RESULT(no))
AC_SEARCH_LIBS([setusercontext], [util])
AC_CHECK_FUNCS([tzset sigprocmask fcntl getpwnam endpwent getrlimit setrlimit setsid chroot kill chown sleep usleep random srandom recvmsg sendmsg writev socketpair glob initgroups strftime localtime_r setusercontext _beginthreadex endservent endprotoent fsync shmget accept4 getifaddrs if_nametoindex])
AC_CHECK_FUNCS([tzset sigprocmask fcntl getpwnam endpwent getrlimit setrlimit setsid chroot kill chown sleep usleep random srandom recvmsg sendmsg writev socketpair glob initgroups strftime localtime_r setusercontext _beginthreadex endservent endprotoent fsync shmget accept4 getifaddrs if_nametoindex poll gettid])
AC_CHECK_FUNCS([setresuid],,[AC_CHECK_FUNCS([setreuid])])
AC_CHECK_FUNCS([setresgid],,[AC_CHECK_FUNCS([setregid])])
@@ -1647,6 +1834,25 @@ AC_CHECK_FUNCS([setresgid],,[AC_CHECK_FUNCS([setregid])])
if echo $host_os | grep darwin8 > /dev/null; then
AC_DEFINE(DARWIN_BROKEN_SETREUID, 1, [Define this if on macOSX10.4-darwin8 and setreuid and setregid do not work])
fi
AC_MSG_CHECKING([for atomic_store])
AC_LINK_IFELSE([AC_LANG_PROGRAM(AC_INCLUDES_DEFAULT [[
#ifdef HAVE_STDATOMIC_H
#include <stdatomic.h>
#endif
]], [[
int newvar = 5, var = 0;
atomic_store((_Atomic int*)&var, newvar);
newvar = 0;
/* condition to use the variables. */
if(var == newvar) return 1;
]])], [
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_LINK_ATOMIC_STORE, 1, [If we have atomic_store])
], [
AC_MSG_RESULT([no])
])
AC_CHECK_DECLS([inet_pton,inet_ntop], [], [], [
AC_INCLUDES_DEFAULT
#ifdef HAVE_NETINET_IN_H
@@ -1684,7 +1890,7 @@ int main(void) { return !(snprintf(NULL, 0, "test") == 4); }
AC_MSG_RESULT(no)
AC_DEFINE([SNPRINTF_RET_BROKEN], [], [define if (v)snprintf does not return length needed, (but length used)])
AC_LIBOBJ(snprintf)
])
], [AC_MSG_RESULT(maybe)])
fi
fi
AC_REPLACE_FUNCS(strlcat)
@@ -1785,10 +1991,12 @@ if test x_$enable_lock_checks = x_yes; then
UBSYMS="-export-symbols clubsyms.def"
cp ${srcdir}/libunbound/ubsyms.def clubsyms.def
echo lock_protect >> clubsyms.def
echo lock_protect_place >> clubsyms.def
echo lock_unprotect >> clubsyms.def
echo lock_get_mem >> clubsyms.def
echo checklock_start >> clubsyms.def
echo checklock_stop >> clubsyms.def
echo checklock_set_output_name >> clubsyms.def
echo checklock_lock >> clubsyms.def
echo checklock_unlock >> clubsyms.def
echo checklock_init >> clubsyms.def
@@ -1869,30 +2077,42 @@ case "$enable_ipset" in
IPSET_OBJ="ipset.lo"
AC_SUBST(IPSET_OBJ)
# mnl
AC_ARG_WITH(libmnl, AS_HELP_STRING([--with-libmnl=path],[specify explicit path for libmnl.]),
# BSD's pf
AC_CHECK_HEADERS([net/pfvar.h], [], [
# mnl
AC_ARG_WITH(libmnl, AS_HELP_STRING([--with-libmnl=path],[specify explicit path for libmnl.]),
[ ],[ withval="yes" ])
found_libmnl="no"
AC_MSG_CHECKING(for libmnl)
if test x_$withval = x_ -o x_$withval = x_yes; then
found_libmnl="no"
AC_MSG_CHECKING(for libmnl)
if test x_$withval = x_ -o x_$withval = x_yes; then
withval="/usr/local /opt/local /usr/lib /usr/pkg /usr/sfw /usr"
fi
for dir in $withval ; do
if test -f "$dir/include/libmnl/libmnl.h"; then
fi
for dir in $withval ; do
if test -f "$dir/include/libmnl/libmnl.h" -o -f "$dir/include/libmnl/libmnl/libmnl.h"; then
found_libmnl="yes"
dnl assume /usr is in default path.
extralibmnl=""
if test -f "$dir/include/libmnl/libmnl/libmnl.h"; then
extralibmnl="/libmnl"
fi
if test "$dir" != "/usr" -o -n "$extralibmnl"; then
CPPFLAGS="$CPPFLAGS -I$dir/include$extralibmnl"
fi
if test "$dir" != "/usr"; then
CPPFLAGS="$CPPFLAGS -I$dir/include"
LDFLAGS="$LDFLAGS -L$dir/lib"
fi
AC_MSG_RESULT(found in $dir)
LIBS="$LIBS -lmnl"
break;
fi
done
if test x_$found_libmnl != x_yes; then
AC_MSG_ERROR([Could not find libmnl, libmnl.h])
fi
done
if test x_$found_libmnl != x_yes; then
AC_MSG_ERROR([Could not find libmnl, libmnl.h])
fi
], [
#include <netinet/in.h>
#include <net/if.h>
])
;;
no|*)
# nothing
@@ -1908,7 +2128,7 @@ case "$enable_explicit_port_randomisation" in
esac
if echo "$host" | $GREP -i -e linux >/dev/null; then
AC_ARG_ENABLE(linux-ip-local-port-range, AC_HELP_STRING([--enable-linux-ip-local-port-range], [Define this to enable use of /proc/sys/net/ipv4/ip_local_port_range as a default outgoing port range. This is only for the libunbound on Linux and does not affect unbound resolving daemon itself. This may severely limit the number of available outgoing ports and thus decrease randomness. Define this only when the target system restricts (e.g. some of SELinux enabled distributions) the use of non-ephemeral ports.]))
AC_ARG_ENABLE(linux-ip-local-port-range, AS_HELP_STRING([--enable-linux-ip-local-port-range], [Define this to enable use of /proc/sys/net/ipv4/ip_local_port_range as a default outgoing port range. This is only for the libunbound on Linux and does not affect unbound resolving daemon itself. This may severely limit the number of available outgoing ports and thus decrease randomness. Define this only when the target system restricts (e.g. some of SELinux enabled distributions) the use of non-ephemeral ports.]))
case "$enable_linux_ip_local_port_range" in
yes)
AC_DEFINE([USE_LINUX_IP_LOCAL_PORT_RANGE], [1], [Define this to enable use of /proc/sys/net/ipv4/ip_local_port_range as a default outgoing port range. This is only for the libunbound on Linux and does not affect unbound resolving daemon itself. This may severely limit the number of available outgoing ports and thus decrease randomness. Define this only when the target system restricts (e.g. some of SELinux enabled distributions) the use of non-ephemeral ports.])
@@ -1965,11 +2185,11 @@ AC_ARG_WITH(libunbound-only, AS_HELP_STRING([--with-libunbound-only],[do not bui
fi
])
if test $ALLTARGET = "alltargets"; then
if test $USE_NSS = "yes"; then
AC_MSG_ERROR([--with-nss can only be used in combination with --with-libunbound-only.])
if test $USE_NSS = "yes"; then
AC_MSG_ERROR([--with-nss can only be used in combination with --with-libunbound-only.])
fi
if test $USE_NETTLE = "yes"; then
AC_MSG_ERROR([--with-nettle can only be used in combination with --with-libunbound-only.])
AC_MSG_ERROR([--with-nettle can only be used in combination with --with-libunbound-only.])
fi
fi
@@ -1980,7 +2200,7 @@ ACX_STRIP_EXT_FLAGS
if test -n "$LATE_LDFLAGS"; then
LDFLAGS="$LATE_LDFLAGS $LDFLAGS"
fi
# remove start spaces
# remove start spaces
LDFLAGS=`echo "$LDFLAGS"|sed -e 's/^ *//'`
LIBS=`echo "$LIBS"|sed -e 's/^ *//'`
@@ -2060,7 +2280,7 @@ dnl includes
#include <ws2tcpip.h>
#endif
#ifndef USE_WINSOCK
#if !defined(USE_WINSOCK) || !defined(HAVE_SNPRINTF) || defined(SNPRINTF_RET_BROKEN) || defined(__USE_MINGW_ANSI_STDIO)
#define ARG_LL "%ll"
#else
#define ARG_LL "%I64"
@@ -2210,6 +2430,10 @@ struct sockaddr_storage;
# define calloc(n,s) unbound_stat_calloc_log(n, s, __FILE__, __LINE__, __func__)
# define free(p) unbound_stat_free_log(p, __FILE__, __LINE__, __func__)
# define realloc(p,s) unbound_stat_realloc_log(p, s, __FILE__, __LINE__, __func__)
# define strdup(s) unbound_stat_strdup_log(s, __FILE__, __LINE__, __func__)
#ifdef HAVE_REALLOCARRAY
# define reallocarray(p,n,s) unbound_stat_reallocarray_log(p, n, s, __FILE__, __LINE__, __func__)
#endif
void *unbound_stat_malloc(size_t size);
void *unbound_stat_calloc(size_t nmemb, size_t size);
void unbound_stat_free(void *ptr);
@@ -2222,6 +2446,10 @@ void unbound_stat_free_log(void *ptr, const char* file, int line,
const char* func);
void *unbound_stat_realloc_log(void *ptr, size_t size, const char* file,
int line, const char* func);
void *unbound_stat_reallocarray_log(void *ptr, size_t nmemb, size_t size,
const char* file, int line, const char* func);
char *unbound_stat_strdup_log(const char *s, const char* file, int line,
const char* func);
#elif defined(UNBOUND_ALLOC_LITE)
# include "util/alloc.h"
#endif /* UNBOUND_ALLOC_LITE and UNBOUND_ALLOC_STATS */
@@ -2232,6 +2460,8 @@ void *unbound_stat_realloc_log(void *ptr, size_t size, const char* file,
#define UNBOUND_DNS_OVER_TLS_PORT 853
/** default port for DNS over HTTPS traffic. */
#define UNBOUND_DNS_OVER_HTTPS_PORT 443
/** default port for DNS over QUIC traffic. */
#define UNBOUND_DNS_OVER_QUIC_PORT 853
/** default port for unbound control traffic, registered port with IANA,
ub-dns-control 8953/tcp unbound dns nameserver control */
#define UNBOUND_CONTROL_PORT 8953
@@ -2243,7 +2473,8 @@ void *unbound_stat_realloc_log(void *ptr, size_t size, const char* file,
dnl if we build from source tree, the man pages need @date@ and @version@
dnl if this is a distro tarball, that was already done by makedist.sh
AC_SUBST(version, [VERSION_MAJOR.VERSION_MINOR.VERSION_MICRO])
AC_SUBST(date, [`date +'%b %e, %Y'`])
AX_BUILD_DATE_EPOCH(date, [[%b %e, %Y]])
AC_SUBST(date)
AC_CONFIG_FILES([Makefile doc/example.conf doc/libunbound.3 doc/unbound.8 doc/unbound-anchor.8 doc/unbound-checkconf.8 doc/unbound.conf.5 doc/unbound-control.8 doc/unbound-host.1 smallapp/unbound-control-setup.sh dnstap/dnstap_config.h dnscrypt/dnscrypt_config.h contrib/libunbound.pc contrib/unbound.socket contrib/unbound.service contrib/unbound_portable.service])
AC_CONFIG_HEADERS([config.h])
+1 -3
View File
@@ -1,10 +1,8 @@
FROM gcc:latest
WORKDIR /usr/src/unbound
RUN apt-get update
# install semantic parser & lexical analyzer
RUN apt-get install -y bison flex
# install packages used in tests
RUN apt-get install -y ldnsutils dnsutils xxd splint doxygen netcat
RUN apt-get update && apt-get install -y bison flex ldnsutils dnsutils xxd splint doxygen netcat-openbsd
# accept short rsa keys, which are used in tests
RUN sed -i 's/SECLEVEL=2/SECLEVEL=1/g' /usr/lib/ssl/openssl.cnf
+3
View File
@@ -55,3 +55,6 @@ distribution but may be helpful.
contributed by Andreas Schulze.
* metrics.awk: awk script that can convert unbound-control stats to
Prometheus metrics format output.
* unbound.init_yocto: An init script to start and stop the server. Put it
in /etc/init.d/unbound to use it. It is for the Yocto Project, in
embedded systems, contributed by beni-sandu.
+52 -52
View File
@@ -1,10 +1,10 @@
diff --git a/doc/unbound.conf.5.in b/doc/unbound.conf.5.in
index 5a75e319..c6c6dbe2 100644
index 172eb26c..2921c87f 100644
--- a/doc/unbound.conf.5.in
+++ b/doc/unbound.conf.5.in
@@ -970,6 +970,13 @@ potentially broken nameservers. A lot of domains will not be resolvable when
this option in enabled. Only use if you know what you are doing.
This option only has effect when qname-minimisation is enabled. Default is no.
@@ -2146,6 +2146,13 @@ Default: no
.UNINDENT
.INDENT 0.0
.TP
+.B aaaa\-filter: \fI<yes or no>
+Activate behavior similar to BIND's AAAA-filter.
@@ -13,14 +13,14 @@ index 5a75e319..c6c6dbe2 100644
+This also causes an additional A query to be sent for each AAAA query.
+This breaks DNSSEC!
+.TP
.B aggressive\-nsec: \fI<yes or no>
Aggressive NSEC uses the DNSSEC NSEC chain to synthesize NXDOMAIN
and other denials, using information from previous NXDOMAINs answers.
.B aggressive\-nsec: \fI<yes or no>\fP
Aggressive NSEC uses the DNSSEC NSEC chain to synthesize NXDOMAIN and other
denials, using information from previous NXDOMAINs answers.
diff --git a/iterator/iter_scrub.c b/iterator/iter_scrub.c
index f093c1bf..e55a2246 100644
index 49a5f5da..fbe434fa 100644
--- a/iterator/iter_scrub.c
+++ b/iterator/iter_scrub.c
@@ -679,6 +679,32 @@ static int sanitize_nsec_is_overreach(sldns_buffer* pkt,
@@ -849,6 +849,32 @@ scrub_sanitize_rr_length(sldns_buffer* pkt, struct msg_parse* msg,
return 0;
}
@@ -53,15 +53,15 @@ index f093c1bf..e55a2246 100644
/**
* Given a response event, remove suspect RRsets from the response.
* "Suspect" rrsets are potentially poison. Note that this routine expects
@@ -698,6 +724,7 @@ scrub_sanitize(sldns_buffer* pkt, struct msg_parse* msg,
@@ -869,6 +895,7 @@ scrub_sanitize(sldns_buffer* pkt, struct msg_parse* msg,
struct query_info* qinfo, uint8_t* zonename, struct module_env* env,
struct iter_env* ie)
struct iter_env* ie, struct module_qstate* qstate)
{
+ int found_a_record = 0; /* ASN: do we have a A record? */
int del_addi = 0; /* if additional-holding rrsets are deleted, we
do not trust the normalized additional-A-AAAA any more */
struct rrset_parse* rrset, *prev;
@@ -733,6 +760,13 @@ scrub_sanitize(sldns_buffer* pkt, struct msg_parse* msg,
uint8_t* ns_rrset_dname = NULL;
@@ -906,6 +933,13 @@ scrub_sanitize(sldns_buffer* pkt, struct msg_parse* msg,
rrset = rrset->rrset_all_next;
}
@@ -75,9 +75,9 @@ index f093c1bf..e55a2246 100644
/* At this point, we brutally remove ALL rrsets that aren't
* children of the originating zone. The idea here is that,
* as far as we know, the server that we contacted is ONLY
@@ -744,6 +778,24 @@ scrub_sanitize(sldns_buffer* pkt, struct msg_parse* msg,
rrset = msg->rrset_first;
while(rrset) {
@@ -925,6 +959,24 @@ scrub_sanitize(sldns_buffer* pkt, struct msg_parse* msg,
continue;
}
+ /* ASN: For AAAA records only... */
+ if((ie->aaaa_filter) && (rrset->type == LDNS_RR_TYPE_AAAA)) {
@@ -101,24 +101,24 @@ index f093c1bf..e55a2246 100644
if( (rrset->type == LDNS_RR_TYPE_A ||
rrset->type == LDNS_RR_TYPE_AAAA)) {
diff --git a/iterator/iter_utils.c b/iterator/iter_utils.c
index 2482a1f4..bd5ba243 100644
index 1da21896..6583dd0e 100644
--- a/iterator/iter_utils.c
+++ b/iterator/iter_utils.c
@@ -177,6 +177,7 @@ iter_apply_cfg(struct iter_env* iter_env, struct config_file* cfg)
iter_env->supports_ipv6 = cfg->do_ip6;
iter_env->supports_ipv4 = cfg->do_ip4;
@@ -250,6 +250,7 @@ iter_apply_cfg(struct iter_env* iter_env, struct config_file* cfg)
iter_env->outbound_msg_retry = cfg->outbound_msg_retry;
iter_env->max_sent_count = cfg->max_sent_count;
iter_env->max_query_restarts = cfg->max_query_restarts;
+ iter_env->aaaa_filter = cfg->aaaa_filter;
return 1;
}
diff --git a/iterator/iterator.c b/iterator/iterator.c
index 54006940..768fe202 100644
index 71e64655..735f4ca0 100644
--- a/iterator/iterator.c
+++ b/iterator/iterator.c
@@ -2155,6 +2155,53 @@ processDSNSFind(struct module_qstate* qstate, struct iter_qstate* iq, int id)
return 0;
@@ -2412,6 +2412,53 @@ check_waiting_queries(struct iter_qstate* iq, struct module_qstate* qstate,
qstate->ext_state[id] = module_wait_reply;
}
}
+
+/**
@@ -170,8 +170,8 @@ index 54006940..768fe202 100644
/**
* This is the request event state where the request will be sent to one of
@@ -2216,6 +2263,13 @@ processQueryTargets(struct module_qstate* qstate, struct iter_qstate* iq,
return error_response(qstate, id, LDNS_RCODE_SERVFAIL);
@@ -2554,6 +2601,13 @@ processQueryTargets(struct module_qstate* qstate, struct iter_qstate* iq,
}
}
+ /* ASN: If we have a AAAA query, then also query for A records */
@@ -184,7 +184,7 @@ index 54006940..768fe202 100644
/* Make sure we have a delegation point, otherwise priming failed
* or another failure occurred */
if(!iq->dp) {
@@ -3648,6 +3702,61 @@ processFinished(struct module_qstate* qstate, struct iter_qstate* iq,
@@ -4178,6 +4232,61 @@ processFinished(struct module_qstate* qstate, struct iter_qstate* iq,
return 0;
}
@@ -236,7 +236,7 @@ index 54006940..768fe202 100644
+ /* see if the failure did get (parent-lame) info */
+ if(!cache_fill_missing(super->env,
+ super_iq->qchase.qclass, super->region,
+ super_iq->dp))
+ super_iq->dp, 0))
+ log_err("ASN-AAAA-filter: out of memory adding missing");
+ dpns->resolved = 1; /* mark as failed */
+ }
@@ -246,7 +246,7 @@ index 54006940..768fe202 100644
/*
* Return priming query results to interested super querystates.
*
@@ -3667,6 +3776,9 @@ iter_inform_super(struct module_qstate* qstate, int id,
@@ -4197,6 +4306,9 @@ iter_inform_super(struct module_qstate* qstate, int id,
else if(super->qinfo.qtype == LDNS_RR_TYPE_DS && ((struct iter_qstate*)
super->minfo[id])->state == DSNS_FIND_STATE)
processDSNSResponse(qstate, id, super);
@@ -256,7 +256,7 @@ index 54006940..768fe202 100644
else if(qstate->return_rcode != LDNS_RCODE_NOERROR)
error_supers(qstate, id, super);
else if(qstate->is_priming)
@@ -3704,6 +3816,9 @@ iter_handle(struct module_qstate* qstate, struct iter_qstate* iq,
@@ -4234,6 +4346,9 @@ iter_handle(struct module_qstate* qstate, struct iter_qstate* iq,
case INIT_REQUEST_3_STATE:
cont = processInitRequest3(qstate, iq, id);
break;
@@ -266,7 +266,7 @@ index 54006940..768fe202 100644
case QUERYTARGETS_STATE:
cont = processQueryTargets(qstate, iq, ie, id);
break;
@@ -4040,6 +4155,8 @@ iter_state_to_string(enum iter_state state)
@@ -4578,6 +4693,8 @@ iter_state_to_string(enum iter_state state)
return "INIT REQUEST STATE (stage 2)";
case INIT_REQUEST_3_STATE:
return "INIT REQUEST STATE (stage 3)";
@@ -275,7 +275,7 @@ index 54006940..768fe202 100644
case QUERYTARGETS_STATE :
return "QUERY TARGETS STATE";
case PRIME_RESP_STATE :
@@ -4064,6 +4181,7 @@ iter_state_is_responsestate(enum iter_state s)
@@ -4602,6 +4719,7 @@ iter_state_is_responsestate(enum iter_state s)
case INIT_REQUEST_STATE :
case INIT_REQUEST_2_STATE :
case INIT_REQUEST_3_STATE :
@@ -284,10 +284,10 @@ index 54006940..768fe202 100644
case COLLECT_CLASS_STATE :
return 0;
diff --git a/iterator/iterator.h b/iterator/iterator.h
index 8b840528..a61c4195 100644
index ae4b4e45..a44f9d27 100644
--- a/iterator/iterator.h
+++ b/iterator/iterator.h
@@ -133,6 +133,9 @@ struct iter_env {
@@ -157,6 +157,9 @@ struct iter_env {
*/
int* target_fetch_policy;
@@ -297,7 +297,7 @@ index 8b840528..a61c4195 100644
/** lock on ratelimit counter */
lock_basic_type queries_ratelimit_lock;
/** number of queries that have been ratelimited */
@@ -187,6 +190,14 @@ enum iter_state {
@@ -217,6 +220,14 @@ enum iter_state {
*/
INIT_REQUEST_3_STATE,
@@ -312,9 +312,9 @@ index 8b840528..a61c4195 100644
/**
* Each time a delegation point changes for a given query or a
* query times out and/or wakes up, this state is (re)visited.
@@ -376,6 +387,13 @@ struct iter_qstate {
*/
int refetch_glue;
@@ -434,6 +445,13 @@ struct iter_qstate {
* already so that it is accepted later. */
int empty_nodata_found;
+ /**
+ * ASN: This is a flag that, if true, means that this query is
@@ -327,10 +327,10 @@ index 8b840528..a61c4195 100644
struct outbound_list outlist;
diff --git a/pythonmod/interface.i b/pythonmod/interface.i
index 1ca8686a..d91b19ec 100644
index 2040fb9e..f073c3dc 100644
--- a/pythonmod/interface.i
+++ b/pythonmod/interface.i
@@ -995,6 +995,7 @@ struct config_file {
@@ -1013,6 +1013,7 @@ struct config_file {
int harden_dnssec_stripped;
int harden_referral_path;
int use_caps_bits_for_id;
@@ -339,23 +339,23 @@ index 1ca8686a..d91b19ec 100644
struct config_strlist* private_domain;
size_t unwanted_threshold;
diff --git a/util/config_file.c b/util/config_file.c
index 969d664b..8d94b008 100644
index b1e767b3..5eb3c099 100644
--- a/util/config_file.c
+++ b/util/config_file.c
@@ -231,6 +231,7 @@ config_create(void)
cfg->harden_referral_path = 0;
@@ -247,6 +247,7 @@ config_create(void)
cfg->harden_algo_downgrade = 0;
cfg->harden_unknown_additional = 0;
cfg->use_caps_bits_for_id = 0;
+ cfg->aaaa_filter = 0; /* ASN: default is disabled */
cfg->caps_whitelist = NULL;
cfg->private_address = NULL;
cfg->private_domain = NULL;
diff --git a/util/config_file.h b/util/config_file.h
index c7c9a0a4..e3aa15b0 100644
index 44ac036b..1e59ab07 100644
--- a/util/config_file.h
+++ b/util/config_file.h
@@ -285,6 +285,8 @@ struct config_file {
int harden_algo_downgrade;
@@ -311,6 +311,8 @@ struct config_file {
int harden_unknown_additional;
/** use 0x20 bits in query as random ID bits */
int use_caps_bits_for_id;
+ /** ASN: enable AAAA filter? */
@@ -364,10 +364,10 @@ index c7c9a0a4..e3aa15b0 100644
struct config_strlist* caps_whitelist;
/** strip away these private addrs from answers, no DNS Rebinding */
diff --git a/util/configlexer.lex b/util/configlexer.lex
index 34a0e5dd..c890be2a 100644
index bc258673..76aab170 100644
--- a/util/configlexer.lex
+++ b/util/configlexer.lex
@@ -317,6 +317,7 @@ use-caps-for-id{COLON} { YDVAR(1, VAR_USE_CAPS_FOR_ID) }
@@ -327,6 +327,7 @@ use-caps-for-id{COLON} { YDVAR(1, VAR_USE_CAPS_FOR_ID) }
caps-whitelist{COLON} { YDVAR(1, VAR_CAPS_WHITELIST) }
caps-exempt{COLON} { YDVAR(1, VAR_CAPS_WHITELIST) }
unwanted-reply-threshold{COLON} { YDVAR(1, VAR_UNWANTED_REPLY_THRESHOLD) }
@@ -376,10 +376,10 @@ index 34a0e5dd..c890be2a 100644
private-domain{COLON} { YDVAR(1, VAR_PRIVATE_DOMAIN) }
prefetch-key{COLON} { YDVAR(1, VAR_PREFETCH_KEY) }
diff --git a/util/configparser.y b/util/configparser.y
index d4f965f9..8cc237c6 100644
index 82e1d878..dc19bed5 100644
--- a/util/configparser.y
+++ b/util/configparser.y
@@ -97,6 +97,7 @@ extern struct config_parser_state* cfg_parser;
@@ -100,6 +100,7 @@ extern struct config_parser_state* cfg_parser;
%token VAR_STATISTICS_CUMULATIVE VAR_OUTGOING_PORT_PERMIT
%token VAR_OUTGOING_PORT_AVOID VAR_DLV_ANCHOR_FILE VAR_DLV_ANCHOR
%token VAR_NEG_CACHE_SIZE VAR_HARDEN_REFERRAL_PATH VAR_PRIVATE_ADDRESS
@@ -387,7 +387,7 @@ index d4f965f9..8cc237c6 100644
%token VAR_PRIVATE_DOMAIN VAR_REMOTE_CONTROL VAR_CONTROL_ENABLE
%token VAR_CONTROL_INTERFACE VAR_CONTROL_PORT VAR_SERVER_KEY_FILE
%token VAR_SERVER_CERT_FILE VAR_CONTROL_KEY_FILE VAR_CONTROL_CERT_FILE
@@ -247,6 +248,7 @@ content_server: server_num_threads | server_verbosity | server_port |
@@ -276,6 +277,7 @@ content_server: server_num_threads | server_verbosity | server_port |
server_dlv_anchor_file | server_dlv_anchor | server_neg_cache_size |
server_harden_referral_path | server_private_address |
server_private_domain | server_extended_statistics |
@@ -395,7 +395,7 @@ index d4f965f9..8cc237c6 100644
server_local_data_ptr | server_jostle_timeout |
server_unwanted_reply_threshold | server_log_time_ascii |
server_domain_insecure | server_val_sig_skew_min |
@@ -1754,6 +1756,15 @@ server_caps_whitelist: VAR_CAPS_WHITELIST STRING_ARG
@@ -1932,6 +1934,15 @@ server_caps_whitelist: VAR_CAPS_WHITELIST STRING_ARG
yyerror("out of memory");
}
;
+7 -4
View File
@@ -1,21 +1,24 @@
#!/usr/bin/env bash
LIBEXPAT_FNAME=expat-2.7.0
LIBEXPAT_VERSION_DIR=R_2_7_0
echo "Downloading Expat"
if ! curl -L -k -s -o expat-2.2.9.tar.gz https://github.com/libexpat/libexpat/releases/download/R_2_2_9/expat-2.2.9.tar.gz;
if ! curl -L -k -s -o $LIBEXPAT_FNAME.tar.gz https://github.com/libexpat/libexpat/releases/download/$LIBEXPAT_VERSION_DIR/$LIBEXPAT_FNAME.tar.gz;
then
echo "Failed to download Expat"
exit 1
fi
echo "Unpacking Expat"
rm -rf ./expat-2.2.9
if ! tar -xf expat-2.2.9.tar.gz;
rm -rf ./$LIBEXPAT_FNAME
if ! tar -xf $LIBEXPAT_FNAME.tar.gz;
then
echo "Failed to unpack Expat"
exit 1
fi
cd expat-2.2.9 || exit 1
cd $LIBEXPAT_FNAME || exit 1
echo "Configuring Expat"
if ! ./configure --build="$AUTOTOOLS_BUILD" --host="$AUTOTOOLS_HOST" --prefix="$ANDROID_PREFIX"; then
+11 -7
View File
@@ -1,28 +1,32 @@
#!/usr/bin/env bash
LIBEXPAT_FNAME=expat-2.7.0
LIBEXPAT_VERSION_DIR=R_2_7_0
echo "Downloading Expat"
if ! curl -L -k -s -o expat-2.2.9.tar.gz https://github.com/libexpat/libexpat/releases/download/R_2_2_9/expat-2.2.9.tar.gz;
if ! curl -L -k -s -o $LIBEXPAT_FNAME.tar.gz https://github.com/libexpat/libexpat/releases/download/$LIBEXPAT_VERSION_DIR/$LIBEXPAT_FNAME.tar.gz;
then
echo "Failed to download Expat"
exit 1
fi
echo "Unpacking Expat"
rm -rf ./expat-2.2.9
if ! tar -xf expat-2.2.9.tar.gz;
rm -rf ./$LIBEXPAT_FNAME
if ! tar -xf $LIBEXPAT_FNAME.tar.gz;
then
echo "Failed to unpack Expat"
exit 1
fi
cd expat-2.2.9 || exit 1
cd $LIBEXPAT_FNAME || exit 1
export PKG_CONFIG_PATH="$IOS_PREFIX/lib/pkgconfig"
echo "Configuring Expat"
if ! ./configure \
--build="$AUTOTOOLS_BUILD" --host="$AUTOTOOLS_HOST" \
--prefix="$IOS_PREFIX" ; then
if ! ./configure --without-tests \
--build="$AUTOTOOLS_BUILD" --host="$AUTOTOOLS_HOST" \
--prefix="$IOS_PREFIX" ;
then
echo "Error: Failed to configure Expat"
cat config.log
exit 1
+1
View File
@@ -28,6 +28,7 @@ END {
print "unbound_hits_queries{type=\"total.num.prefetch\"} " val["total.num.prefetch"];
print "unbound_hits_queries{type=\"num.query.tcp\"} " val["num.query.tcp"];
print "unbound_hits_queries{type=\"num.query.tcpout\"} " val["num.query.tcpout"];
print "unbound_hits_queries{type=\"num.query.udpout\"} " val["num.query.udpout"];
print "unbound_hits_queries{type=\"num.query.tls\"} " val["num.query.tls"];
print "unbound_hits_queries{type=\"num.query.tls.resume\"} " val["num.query.tls.resume"];
print "unbound_hits_queries{type=\"num.query.ipv6\"} " val["num.query.ipv6"];
+9
View File
@@ -22,4 +22,13 @@ pidfile=${unbound_pidfile:-"/usr/local/etc/unbound/unbound.pid"}
command_args=${unbound_flags:-"-c /usr/local/etc/unbound/unbound.conf"}
extra_commands="reload"
if test "$1" = "stop" ; then
run_rc_command "$1"
ret=$?
if test $ret -eq 0; then
rm -f "$pidfile"
fi
exit $ret
fi
run_rc_command "$1"
+1
View File
@@ -75,6 +75,7 @@ stop() {
retval=$?
echo
[ $retval -eq 0 ] && rm -f $lockfile
[ $retval -eq 0 ] && rm -f $pidfile
if egrep -q '^/[^[:space:]]+[[:space:]]+'${rootdir}'/dev/log' /proc/mounts; then
umount ${rootdir}/dev/log >/dev/null 2>&1
fi;
+1
View File
@@ -58,6 +58,7 @@ stop() {
killproc -p $pidfile unbound
retval=$?
[ $retval -eq 0 ] && rm -f $lockfile
[ $retval -eq 0 ] && rm -f $pidfile
for mountfile in /dev/log /dev/urandom /etc/localtime /etc/resolv.conf /var/run/unbound
do
if egrep -q '^/[^[:space:]]+[[:space:]]+'${rootdir}''${mountfile}'' /proc/mounts; then
+140
View File
@@ -0,0 +1,140 @@
#!/bin/sh
#
# unbound This shell script takes care of starting and stopping
# unbound (DNS server).
#
# chkconfig: - 14 86
# description: unbound is a Domain Name Server (DNS) \
# that is used to resolve host names to IP addresses.
### BEGIN INIT INFO
# Provides: $named unbound
# Required-Start: $network $local_fs
# Required-Stop: $network $local_fs
# Should-Start: $syslog
# Should-Stop: $syslog
# Short-Description: unbound recursive Domain Name Server.
# Description: unbound is a Domain Name Server (DNS)
# that is used to resolve host names to IP addresses.
### END INIT INFO
# Source function library.
. /etc/init.d/functions
exec="/usr/sbin/unbound"
prog="unbound"
config="/etc/unbound/unbound.conf"
pidfile="/var/unbound/unbound.pid"
rootdir="/var/unbound"
[ -e /etc/sysconfig/$prog ] && . /etc/sysconfig/$prog
lockfile=/var/lock/subsys/$prog
start() {
[ -x $exec ] || exit 5
[ -f $config ] || exit 6
echo -n $"Starting $prog: "
# setup root jail
if [ -s /etc/localtime ]; then
[ -d ${rootdir}/etc ] || mkdir -p ${rootdir}/etc ;
if [ ! -e ${rootdir}/etc/localtime ] || ! /usr/bin/cmp -s /etc/localtime ${rootdir}/etc/localtime; then
cp -fp /etc/localtime ${rootdir}/etc/localtime
fi;
fi;
if [ -s /etc/resolv.conf ]; then
[ -d ${rootdir}/etc ] || mkdir -p ${rootdir}/etc ;
if [ ! -e ${rootdir}/etc/resolv.conf ] || ! /usr/bin/cmp -s /etc/resolv.conf ${rootdir}/etc/resolv.conf; then
cp -fp /etc/resolv.conf ${rootdir}/etc/resolv.conf
fi;
fi;
if ! egrep -q '^/[^[:space:]]+[[:space:]]+'${rootdir}'/dev/log' /proc/mounts; then
[ -d ${rootdir}/dev ] || mkdir -p ${rootdir}/dev ;
[ -e ${rootdir}/dev/log ] || touch ${rootdir}/dev/log
mount --bind -n /dev/log ${rootdir}/dev/log >/dev/null 2>&1;
fi;
if ! egrep -q '^/[^[:space:]]+[[:space:]]+'${rootdir}'/dev/random' /proc/mounts; then
[ -d ${rootdir}/dev ] || mkdir -p ${rootdir}/dev ;
[ -e ${rootdir}/dev/random ] || touch ${rootdir}/dev/random
mount --bind -n /dev/random ${rootdir}/dev/random >/dev/null 2>&1;
fi;
# if not running, start it up here
daemonize $exec
retval=$?
echo
[ $retval -eq 0 ] && touch $lockfile
return $retval
}
stop() {
echo -n $"Stopping $prog: "
# stop it here, often "killproc $prog"
killproc $prog
retval=$?
echo
[ $retval -eq 0 ] && rm -f $lockfile
[ $retval -eq 0 ] && rm -f $pidfile
if egrep -q '^/[^[:space:]]+[[:space:]]+'${rootdir}'/dev/log' /proc/mounts; then
umount ${rootdir}/dev/log >/dev/null 2>&1
fi;
if egrep -q '^/[^[:space:]]+[[:space:]]+'${rootdir}'/dev/random' /proc/mounts; then
umount ${rootdir}/dev/random >/dev/null 2>&1
fi;
return $retval
}
restart() {
stop
start
}
reload() {
kill -HUP `cat $pidfile`
}
force_reload() {
restart
}
rh_status() {
# run checks to determine if the service is running or use generic status
status $prog
}
rh_status_q() {
rh_status -p $pidfile >/dev/null 2>&1
}
case "$1" in
start)
rh_status_q && exit 0
$1
;;
stop)
rh_status_q || exit 0
$1
;;
restart)
$1
;;
reload)
rh_status_q || exit 7
$1
;;
force-reload)
force_reload
;;
status)
rh_status
;;
condrestart|try-restart)
rh_status_q || exit 0
restart
;;
*)
echo $"Usage: $0 {start|stop|status|restart|condrestart|try-restart|reload|force-reload}"
exit 2
esac
exit $?
+6 -1
View File
@@ -38,13 +38,18 @@
; - `LockPersonality=yes` locks down the personality system call so that the
; kernel execution domain may not be changed from the default.
;
; - With /etc/systemd/network/*.network a setting to make sure the network
; is not considered online too early, can reduce network unreachable
; errors on server start:
; [Link]
; RequiredForOnline=routable
;
[Unit]
Description=Validating, recursive, and caching DNS resolver
Documentation=man:unbound(8)
After=network-online.target
Wants=network-online.target
Before=nss-lookup.target
Wants=network-online.target nss-lookup.target
[Install]
WantedBy=multi-user.target
+2 -1
View File
@@ -253,6 +253,7 @@ if test "$1" = "config" ; then
p_config "total.num.prefetch" "cache prefetch" "ABSOLUTE"
p_config "num.query.tcp" "TCP queries" "ABSOLUTE"
p_config "num.query.tcpout" "TCP out queries" "ABSOLUTE"
p_config "num.query.udpout" "UDP out queries" "ABSOLUTE"
p_config "num.query.tls" "TLS queries" "ABSOLUTE"
p_config "num.query.tls.resume" "TLS resumes" "ABSOLUTE"
p_config "num.query.ipv6" "IPv6 queries" "ABSOLUTE"
@@ -452,7 +453,7 @@ hits)
for x in `grep "^thread[0-9][0-9]*\.num\.queries=" $state |
sed -e 's/=.*//'` total.num.queries \
total.num.cachehits total.num.prefetch num.query.tcp \
num.query.tcpout num.query.tls num.query.tls.resume \
num.query.tcpout num.query.udpout num.query.tls num.query.tls.resume \
num.query.ipv6 unwanted.queries \
unwanted.replies; do
if grep "^"$x"=" $state >/dev/null 2>&1; then
+2 -2
View File
@@ -14,8 +14,8 @@
[Unit]
Description=Validating, recursive, and caching DNS resolver
Documentation=man:unbound(8)
After=network.target
Before=network-online.target nss-lookup.target
After=network-online.target
Before=nss-lookup.target
Wants=nss-lookup.target
[Install]
+428 -88
View File
@@ -46,9 +46,10 @@
#include "util/config_file.h"
#include "util/net_help.h"
#include "services/localzone.h"
#include "services/listen_dnsport.h"
#include "sldns/str2wire.h"
struct acl_list*
struct acl_list*
acl_list_create(void)
{
struct acl_list* acl = (struct acl_list*)calloc(1,
@@ -63,10 +64,10 @@ acl_list_create(void)
return acl;
}
void
void
acl_list_delete(struct acl_list* acl)
{
if(!acl)
if(!acl)
return;
regional_destroy(acl->region);
free(acl);
@@ -74,8 +75,8 @@ acl_list_delete(struct acl_list* acl)
/** insert new address into acl_list structure */
static struct acl_addr*
acl_list_insert(struct acl_list* acl, struct sockaddr_storage* addr,
socklen_t addrlen, int net, enum acl_access control,
acl_list_insert(struct acl_list* acl, struct sockaddr_storage* addr,
socklen_t addrlen, int net, enum acl_access control,
int complain_duplicates)
{
struct acl_addr* node = regional_alloc_zero(acl->region,
@@ -90,6 +91,33 @@ acl_list_insert(struct acl_list* acl, struct sockaddr_storage* addr,
return node;
}
/** parse str to acl_access enum */
static int
parse_acl_access(const char* str, enum acl_access* control)
{
if(strcmp(str, "allow") == 0)
*control = acl_allow;
else if(strcmp(str, "deny") == 0)
*control = acl_deny;
else if(strcmp(str, "refuse") == 0)
*control = acl_refuse;
else if(strcmp(str, "deny_non_local") == 0)
*control = acl_deny_non_local;
else if(strcmp(str, "refuse_non_local") == 0)
*control = acl_refuse_non_local;
else if(strcmp(str, "allow_snoop") == 0)
*control = acl_allow_snoop;
else if(strcmp(str, "allow_setrd") == 0)
*control = acl_allow_setrd;
else if (strcmp(str, "allow_cookie") == 0)
*control = acl_allow_cookie;
else {
log_err("access control type %s unknown", str);
return 0;
}
return 1;
}
/** apply acl_list string */
static int
acl_list_str_cfg(struct acl_list* acl, const char* str, const char* s2,
@@ -99,29 +127,14 @@ acl_list_str_cfg(struct acl_list* acl, const char* str, const char* s2,
int net;
socklen_t addrlen;
enum acl_access control;
if(strcmp(s2, "allow") == 0)
control = acl_allow;
else if(strcmp(s2, "deny") == 0)
control = acl_deny;
else if(strcmp(s2, "refuse") == 0)
control = acl_refuse;
else if(strcmp(s2, "deny_non_local") == 0)
control = acl_deny_non_local;
else if(strcmp(s2, "refuse_non_local") == 0)
control = acl_refuse_non_local;
else if(strcmp(s2, "allow_snoop") == 0)
control = acl_allow_snoop;
else if(strcmp(s2, "allow_setrd") == 0)
control = acl_allow_setrd;
else {
log_err("access control type %s unknown", str);
if(!parse_acl_access(s2, &control)) {
return 0;
}
if(!netblockstrtoaddr(str, UNBOUND_DNS_PORT, &addr, &addrlen, &net)) {
log_err("cannot parse access control: %s %s", str, s2);
return 0;
}
if(!acl_list_insert(acl, &addr, addrlen, net, control,
if(!acl_list_insert(acl, &addr, addrlen, net, control,
complain_duplicates)) {
log_err("out of memory");
return 0;
@@ -131,19 +144,27 @@ acl_list_str_cfg(struct acl_list* acl, const char* str, const char* s2,
/** find or create node (NULL on parse or error) */
static struct acl_addr*
acl_find_or_create(struct acl_list* acl, const char* str)
acl_find_or_create_str2addr(struct acl_list* acl, const char* str,
int is_interface, int port)
{
struct acl_addr* node;
struct sockaddr_storage addr;
int net;
socklen_t addrlen;
if(!netblockstrtoaddr(str, UNBOUND_DNS_PORT, &addr, &addrlen, &net)) {
log_err("cannot parse netblock: %s", str);
return NULL;
int net = (str_is_ip6(str)?128:32);
if(is_interface) {
if(!extstrtoaddr(str, &addr, &addrlen, port)) {
log_err("cannot parse interface: %s", str);
return NULL;
}
} else {
if(!netblockstrtoaddr(str, UNBOUND_DNS_PORT, &addr, &addrlen, &net)) {
log_err("cannot parse netblock: %s", str);
return NULL;
}
}
/* find or create node */
if(!(node=(struct acl_addr*)addr_tree_find(&acl->tree, &addr,
addrlen, net))) {
addrlen, net)) && !is_interface) {
/* create node, type 'allow' since otherwise tags are
* pointless, can override with specific access-control: cfg */
if(!(node=(struct acl_addr*)acl_list_insert(acl, &addr,
@@ -155,14 +176,67 @@ acl_find_or_create(struct acl_list* acl, const char* str)
return node;
}
/** find or create node (NULL on error) */
static struct acl_addr*
acl_find_or_create(struct acl_list* acl, struct sockaddr_storage* addr,
socklen_t addrlen, enum acl_access control)
{
struct acl_addr* node;
int net = (addr_is_ip6(addr, addrlen)?128:32);
/* find or create node */
if(!(node=(struct acl_addr*)addr_tree_find(&acl->tree, addr,
addrlen, net))) {
/* create node;
* can override with specific access-control: cfg */
if(!(node=(struct acl_addr*)acl_list_insert(acl, addr,
addrlen, net, control, 1))) {
log_err("out of memory");
return NULL;
}
}
return node;
}
/** apply acl_interface string */
static int
acl_interface_str_cfg(struct acl_list* acl_interface, const char* iface,
const char* s2, int port)
{
struct acl_addr* node;
enum acl_access control;
if(!parse_acl_access(s2, &control)) {
return 0;
}
if(!(node=acl_find_or_create_str2addr(acl_interface, iface, 1, port))) {
log_err("cannot update ACL on non-configured interface: %s %d",
iface, port);
return 0;
}
node->control = control;
return 1;
}
struct acl_addr*
acl_interface_insert(struct acl_list* acl_interface,
struct sockaddr_storage* addr, socklen_t addrlen,
enum acl_access control)
{
struct acl_addr* node = acl_find_or_create(acl_interface, addr, addrlen, control);
node->is_interface = 1;
return node;
}
/** apply acl_tag string */
static int
acl_list_tags_cfg(struct acl_list* acl, const char* str, uint8_t* bitmap,
size_t bitmaplen)
size_t bitmaplen, int is_interface, int port)
{
struct acl_addr* node;
if(!(node=acl_find_or_create(acl, str)))
if(!(node=acl_find_or_create_str2addr(acl, str, is_interface, port))) {
if(is_interface)
log_err("non-configured interface: %s", str);
return 0;
}
node->taglen = bitmaplen;
node->taglist = regional_alloc_init(acl->region, bitmap, bitmaplen);
if(!node->taglist) {
@@ -175,11 +249,14 @@ acl_list_tags_cfg(struct acl_list* acl, const char* str, uint8_t* bitmap,
/** apply acl_view string */
static int
acl_list_view_cfg(struct acl_list* acl, const char* str, const char* str2,
struct views* vs)
struct views* vs, int is_interface, int port)
{
struct acl_addr* node;
if(!(node=acl_find_or_create(acl, str)))
if(!(node=acl_find_or_create_str2addr(acl, str, is_interface, port))) {
if(is_interface)
log_err("non-configured interface: %s", str);
return 0;
}
node->view = views_find_view(vs, str2, 0 /* get read lock*/);
if(!node->view) {
log_err("no view with name: %s", str2);
@@ -192,13 +269,17 @@ acl_list_view_cfg(struct acl_list* acl, const char* str, const char* str2,
/** apply acl_tag_action string */
static int
acl_list_tag_action_cfg(struct acl_list* acl, struct config_file* cfg,
const char* str, const char* tag, const char* action)
const char* str, const char* tag, const char* action,
int is_interface, int port)
{
struct acl_addr* node;
int tagid;
enum localzone_type t;
if(!(node=acl_find_or_create(acl, str)))
if(!(node=acl_find_or_create_str2addr(acl, str, is_interface, port))) {
if(is_interface)
log_err("non-configured interface: %s", str);
return 0;
}
/* allocate array if not yet */
if(!node->tag_actions) {
node->tag_actions = (uint8_t*)regional_alloc_zero(acl->region,
@@ -281,13 +362,17 @@ check_data(const char* data, const struct config_strlist* head)
/** apply acl_tag_data string */
static int
acl_list_tag_data_cfg(struct acl_list* acl, struct config_file* cfg,
const char* str, const char* tag, const char* data)
const char* str, const char* tag, const char* data,
int is_interface, int port)
{
struct acl_addr* node;
int tagid;
char* dupdata;
if(!(node=acl_find_or_create(acl, str)))
if(!(node=acl_find_or_create_str2addr(acl, str, is_interface, port))) {
if(is_interface)
log_err("non-configured interface: %s", str);
return 0;
}
/* allocate array if not yet */
if(!node->tag_datas) {
node->tag_datas = (struct config_strlist**)regional_alloc_zero(
@@ -329,11 +414,11 @@ acl_list_tag_data_cfg(struct acl_list* acl, struct config_file* cfg,
}
/** read acl_list config */
static int
read_acl_list(struct acl_list* acl, struct config_file* cfg)
static int
read_acl_list(struct acl_list* acl, struct config_str2list* acls)
{
struct config_str2list* p;
for(p = cfg->acls; p; p = p->next) {
for(p = acls; p; p = p->next) {
log_assert(p->str && p->str2);
if(!acl_list_str_cfg(acl, p->str, p->str2, 1))
return 0;
@@ -341,15 +426,38 @@ read_acl_list(struct acl_list* acl, struct config_file* cfg)
return 1;
}
/** read acl tags config */
static int
read_acl_tags(struct acl_list* acl, struct config_file* cfg)
/** read acl view config */
static int
read_acl_view(struct acl_list* acl, struct config_str2list** acl_view,
struct views* v)
{
struct config_strbytelist* np, *p = cfg->acl_tags;
cfg->acl_tags = NULL;
struct config_str2list* np, *p = *acl_view;
*acl_view = NULL;
while(p) {
log_assert(p->str && p->str2);
if(!acl_list_tags_cfg(acl, p->str, p->str2, p->str2len)) {
if(!acl_list_view_cfg(acl, p->str, p->str2, v, 0, 0)) {
config_deldblstrlist(p);
return 0;
}
/* free the items as we go to free up memory */
np = p->next;
free(p->str);
free(p->str2);
free(p);
p = np;
}
return 1;
}
/** read acl tags config */
static int
read_acl_tags(struct acl_list* acl, struct config_strbytelist** acl_tags)
{
struct config_strbytelist* np, *p = *acl_tags;
*acl_tags = NULL;
while(p) {
log_assert(p->str && p->str2);
if(!acl_list_tags_cfg(acl, p->str, p->str2, p->str2len, 0, 0)) {
config_del_strbytelist(p);
return 0;
}
@@ -363,38 +471,18 @@ read_acl_tags(struct acl_list* acl, struct config_file* cfg)
return 1;
}
/** read acl view config */
static int
read_acl_view(struct acl_list* acl, struct config_file* cfg, struct views* v)
{
struct config_str2list* np, *p = cfg->acl_view;
cfg->acl_view = NULL;
while(p) {
log_assert(p->str && p->str2);
if(!acl_list_view_cfg(acl, p->str, p->str2, v)) {
return 0;
}
/* free the items as we go to free up memory */
np = p->next;
free(p->str);
free(p->str2);
free(p);
p = np;
}
return 1;
}
/** read acl tag actions config */
static int
read_acl_tag_actions(struct acl_list* acl, struct config_file* cfg)
static int
read_acl_tag_actions(struct acl_list* acl, struct config_file* cfg,
struct config_str3list** acl_tag_actions)
{
struct config_str3list* p, *np;
p = cfg->acl_tag_actions;
cfg->acl_tag_actions = NULL;
p = *acl_tag_actions;
*acl_tag_actions = NULL;
while(p) {
log_assert(p->str && p->str2 && p->str3);
if(!acl_list_tag_action_cfg(acl, cfg, p->str, p->str2,
p->str3)) {
p->str3, 0, 0)) {
config_deltrplstrlist(p);
return 0;
}
@@ -410,15 +498,17 @@ read_acl_tag_actions(struct acl_list* acl, struct config_file* cfg)
}
/** read acl tag datas config */
static int
read_acl_tag_datas(struct acl_list* acl, struct config_file* cfg)
static int
read_acl_tag_datas(struct acl_list* acl, struct config_file* cfg,
struct config_str3list** acl_tag_datas)
{
struct config_str3list* p, *np;
p = cfg->acl_tag_datas;
cfg->acl_tag_datas = NULL;
p = *acl_tag_datas;
*acl_tag_datas = NULL;
while(p) {
log_assert(p->str && p->str2 && p->str3);
if(!acl_list_tag_data_cfg(acl, cfg, p->str, p->str2, p->str3)) {
if(!acl_list_tag_data_cfg(acl, cfg, p->str, p->str2, p->str3,
0, 0)) {
config_deltrplstrlist(p);
return 0;
}
@@ -433,30 +523,27 @@ read_acl_tag_datas(struct acl_list* acl, struct config_file* cfg)
return 1;
}
int
int
acl_list_apply_cfg(struct acl_list* acl, struct config_file* cfg,
struct views* v)
{
regional_free_all(acl->region);
addr_tree_init(&acl->tree);
if(!read_acl_list(acl, cfg))
if(!read_acl_list(acl, cfg->acls))
return 0;
if(!read_acl_view(acl, cfg, v))
if(!read_acl_view(acl, &cfg->acl_view, v))
return 0;
if(!read_acl_tags(acl, cfg))
if(!read_acl_tags(acl, &cfg->acl_tags))
return 0;
if(!read_acl_tag_actions(acl, cfg))
if(!read_acl_tag_actions(acl, cfg, &cfg->acl_tag_actions))
return 0;
if(!read_acl_tag_datas(acl, cfg))
if(!read_acl_tag_datas(acl, cfg, &cfg->acl_tag_datas))
return 0;
/* insert defaults, with '0' to ignore them if they are duplicates */
if(!acl_list_str_cfg(acl, "0.0.0.0/0", "refuse", 0))
return 0;
/* the 'refuse' defaults for /0 are now done per interface instead */
if(!acl_list_str_cfg(acl, "127.0.0.0/8", "allow", 0))
return 0;
if(cfg->do_ip6) {
if(!acl_list_str_cfg(acl, "::0/0", "refuse", 0))
return 0;
if(!acl_list_str_cfg(acl, "::1", "allow", 0))
return 0;
if(!acl_list_str_cfg(acl, "::ffff:127.0.0.1", "allow", 0))
@@ -466,7 +553,212 @@ acl_list_apply_cfg(struct acl_list* acl, struct config_file* cfg,
return 1;
}
enum acl_access
void
acl_interface_init(struct acl_list* acl_interface)
{
regional_free_all(acl_interface->region);
/* We want comparison in the tree to include only address and port.
* We don't care about comparing node->net. All addresses in the
* acl_interface->tree should have either 32 (ipv4) or 128 (ipv6).
* Initialise with the appropriate compare function but keep treating
* it as an addr_tree. */
addr_tree_addrport_init(&acl_interface->tree);
}
static int
read_acl_interface_action(struct acl_list* acl_interface,
struct config_str2list* acls, int port)
{
struct config_str2list* p;
for(p = acls; p; p = p->next) {
char** resif = NULL;
int num_resif = 0;
int i;
log_assert(p->str && p->str2);
if(!resolve_interface_names(&p->str, 1, NULL, &resif, &num_resif))
return 0;
for(i = 0; i<num_resif; i++) {
if(!acl_interface_str_cfg(acl_interface, resif[i], p->str2, port)){
config_del_strarray(resif, num_resif);
return 0;
}
}
config_del_strarray(resif, num_resif);
}
return 1;
}
/** read acl view config for interface */
static int
read_acl_interface_view(struct acl_list* acl_interface,
struct config_str2list** acl_view,
struct views* v, int port)
{
struct config_str2list* np, *p = *acl_view;
*acl_view = NULL;
while(p) {
char** resif = NULL;
int num_resif = 0;
int i;
log_assert(p->str && p->str2);
if(!resolve_interface_names(&p->str, 1, NULL, &resif, &num_resif)) {
config_deldblstrlist(p);
return 0;
}
for(i = 0; i<num_resif; i++) {
if(!acl_list_view_cfg(acl_interface, resif[i], p->str2,
v, 1, port)) {
config_del_strarray(resif, num_resif);
config_deldblstrlist(p);
return 0;
}
}
config_del_strarray(resif, num_resif);
/* free the items as we go to free up memory */
np = p->next;
free(p->str);
free(p->str2);
free(p);
p = np;
}
return 1;
}
/** read acl tags config for interface */
static int
read_acl_interface_tags(struct acl_list* acl_interface,
struct config_strbytelist** acl_tags, int port)
{
struct config_strbytelist* np, *p = *acl_tags;
*acl_tags = NULL;
while(p) {
char** resif = NULL;
int num_resif = 0;
int i;
log_assert(p->str && p->str2);
if(!resolve_interface_names(&p->str, 1, NULL, &resif, &num_resif)) {
config_del_strbytelist(p);
return 0;
}
for(i = 0; i<num_resif; i++) {
if(!acl_list_tags_cfg(acl_interface, resif[i], p->str2,
p->str2len, 1, port)) {
config_del_strbytelist(p);
config_del_strarray(resif, num_resif);
return 0;
}
}
config_del_strarray(resif, num_resif);
/* free the items as we go to free up memory */
np = p->next;
free(p->str);
free(p->str2);
free(p);
p = np;
}
return 1;
}
/** read acl tag actions config for interface*/
static int
read_acl_interface_tag_actions(struct acl_list* acl_interface,
struct config_file* cfg,
struct config_str3list** acl_tag_actions, int port)
{
struct config_str3list* p, *np;
p = *acl_tag_actions;
*acl_tag_actions = NULL;
while(p) {
char** resif = NULL;
int num_resif = 0;
int i;
log_assert(p->str && p->str2 && p->str3);
if(!resolve_interface_names(&p->str, 1, NULL, &resif, &num_resif)) {
config_deltrplstrlist(p);
return 0;
}
for(i = 0; i<num_resif; i++) {
if(!acl_list_tag_action_cfg(acl_interface, cfg,
resif[i], p->str2, p->str3, 1, port)) {
config_deltrplstrlist(p);
config_del_strarray(resif, num_resif);
return 0;
}
}
config_del_strarray(resif, num_resif);
/* free the items as we go to free up memory */
np = p->next;
free(p->str);
free(p->str2);
free(p->str3);
free(p);
p = np;
}
return 1;
}
/** read acl tag datas config for interface */
static int
read_acl_interface_tag_datas(struct acl_list* acl_interface,
struct config_file* cfg,
struct config_str3list** acl_tag_datas, int port)
{
struct config_str3list* p, *np;
p = *acl_tag_datas;
*acl_tag_datas = NULL;
while(p) {
char** resif = NULL;
int num_resif = 0;
int i;
log_assert(p->str && p->str2 && p->str3);
if(!resolve_interface_names(&p->str, 1, NULL, &resif, &num_resif)) {
config_deltrplstrlist(p);
return 0;
}
for(i = 0; i<num_resif; i++) {
if(!acl_list_tag_data_cfg(acl_interface, cfg,
resif[i], p->str2, p->str3, 1, port)) {
config_deltrplstrlist(p);
config_del_strarray(resif, num_resif);
return 0;
}
}
config_del_strarray(resif, num_resif);
/* free the items as we go to free up memory */
np = p->next;
free(p->str);
free(p->str2);
free(p->str3);
free(p);
p = np;
}
return 1;
}
int
acl_interface_apply_cfg(struct acl_list* acl_interface, struct config_file* cfg,
struct views* v)
{
if(!read_acl_interface_action(acl_interface, cfg->interface_actions,
cfg->port))
return 0;
if(!read_acl_interface_view(acl_interface, &cfg->interface_view, v,
cfg->port))
return 0;
if(!read_acl_interface_tags(acl_interface, &cfg->interface_tags,
cfg->port))
return 0;
if(!read_acl_interface_tag_actions(acl_interface, cfg,
&cfg->interface_tag_actions, cfg->port))
return 0;
if(!read_acl_interface_tag_datas(acl_interface, cfg,
&cfg->interface_tag_datas, cfg->port))
return 0;
addr_tree_init_parents(&acl_interface->tree);
return 1;
}
enum acl_access
acl_get_control(struct acl_addr* acl)
{
if(acl) return acl->control;
@@ -481,9 +773,57 @@ acl_addr_lookup(struct acl_list* acl, struct sockaddr_storage* addr,
addr, addrlen);
}
size_t
size_t
acl_list_get_mem(struct acl_list* acl)
{
if(!acl) return 0;
return sizeof(*acl) + regional_get_mem(acl->region);
}
const char* acl_access_to_str(enum acl_access acl)
{
switch(acl) {
case acl_deny: return "deny";
case acl_refuse: return "refuse";
case acl_deny_non_local: return "deny_non_local";
case acl_refuse_non_local: return "refuse_non_local";
case acl_allow: return "allow";
case acl_allow_snoop: return "allow_snoop";
case acl_allow_setrd: return "allow_setrd";
default: break;
}
return "unknown";
}
void
log_acl_action(const char* action, struct sockaddr_storage* addr,
socklen_t addrlen, enum acl_access acl, struct acl_addr* acladdr)
{
char a[128], n[128];
uint16_t port;
addr_to_str(addr, addrlen, a, sizeof(a));
port = ntohs(((struct sockaddr_in*)addr)->sin_port);
if(acladdr) {
addr_to_str(&acladdr->node.addr, acladdr->node.addrlen,
n, sizeof(n));
verbose(VERB_ALGO, "%s query from %s port %d because of "
"%s/%d %s%s", action, a, (int)port, n,
acladdr->node.net,
acladdr->is_interface?"(ACL on interface IP) ":"",
acl_access_to_str(acl));
} else {
verbose(VERB_ALGO, "%s query from %s port %d", action, a,
(int)port);
}
}
void acl_list_swap_tree(struct acl_list* acl, struct acl_list* data)
{
/* swap tree and region */
rbtree_type oldtree = acl->tree;
struct regional* oldregion = acl->region;
acl->tree = data->tree;
acl->region = data->region;
data->tree = oldtree;
data->region = oldregion;
}
+60 -5
View File
@@ -36,7 +36,7 @@
/**
* \file
*
* This file keeps track of the list of clients that are allowed to
* This file keeps track of the list of clients that are allowed to
* access the server.
*/
@@ -64,8 +64,12 @@ enum acl_access {
acl_allow,
/** allow full access for all queries, recursion and cache snooping */
acl_allow_snoop,
/** allow full access for recursion queries and set RD flag regardless of request */
acl_allow_setrd
/** allow full access for recursion queries and set RD flag regardless
* of request */
acl_allow_setrd,
/** allow full access for recursion (+RD) queries if valid cookie
* present or stateful transport */
acl_allow_cookie
};
/**
@@ -74,7 +78,7 @@ enum acl_access {
struct acl_list {
/** regional for allocation */
struct regional* region;
/**
/**
* Tree of the addresses that are allowed/blocked.
* contents of type acl_addr.
*/
@@ -103,12 +107,14 @@ struct acl_addr {
struct config_strlist** tag_datas;
/** size of the tag_datas array */
size_t tag_datas_size;
/* If the acl node is for an interface */
int is_interface;
/* view element, NULL if none */
struct view* view;
};
/**
* Create acl structure
* Create acl structure
* @return new structure or NULL on error.
*/
struct acl_list* acl_list_create(void);
@@ -119,6 +125,20 @@ struct acl_list* acl_list_create(void);
*/
void acl_list_delete(struct acl_list* acl);
/**
* Insert interface in the acl_list. This should happen when the listening
* interface is setup.
* @param acl_interface: acl_list to insert to.
* @param addr: interface IP.
* @param addrlen: length of the interface IP.
* @param control: acl_access.
* @return new structure or NULL on error.
*/
struct acl_addr*
acl_interface_insert(struct acl_list* acl_interface,
struct sockaddr_storage* addr, socklen_t addrlen,
enum acl_access control);
/**
* Process access control config.
* @param acl: where to store.
@@ -129,6 +149,22 @@ void acl_list_delete(struct acl_list* acl);
int acl_list_apply_cfg(struct acl_list* acl, struct config_file* cfg,
struct views* v);
/**
* Initialise (also clean) the acl_interface struct.
* @param acl_interface: where to store.
*/
void acl_interface_init(struct acl_list* acl_interface);
/**
* Process interface control config.
* @param acl_interface: where to store.
* @param cfg: config options.
* @param v: views structure
* @return 0 on error.
*/
int acl_interface_apply_cfg(struct acl_list* acl_interface, struct config_file* cfg,
struct views* v);
/**
* Lookup access control status for acl structure.
* @param acl: structure for acl storage.
@@ -154,4 +190,23 @@ acl_addr_lookup(struct acl_list* acl, struct sockaddr_storage* addr,
*/
size_t acl_list_get_mem(struct acl_list* acl);
/*
* Get string for acl access specification
* @param acl: access type value
* @return string
*/
const char* acl_access_to_str(enum acl_access acl);
/* log acl and addr for action */
void log_acl_action(const char* action, struct sockaddr_storage* addr,
socklen_t addrlen, enum acl_access acl, struct acl_addr* acladdr);
/**
* Swap internal tree with preallocated entries.
* @param acl: the acl structure.
* @param data: the data structure used to take elements from. This contains
* the old elements on return.
*/
void acl_list_swap_tree(struct acl_list* acl, struct acl_list* data);
#endif /* DAEMON_ACL_LIST_H */
+277 -121
View File
@@ -47,10 +47,12 @@
#include "services/cache/rrset.h"
#include "services/cache/dns.h"
#include "services/cache/infra.h"
#include "services/outside_network.h"
#include "util/data/msgreply.h"
#include "util/regional.h"
#include "util/net_help.h"
#include "util/data/dname.h"
#include "util/config_file.h"
#include "iterator/iterator.h"
#include "iterator/iter_delegpt.h"
#include "iterator/iter_utils.h"
@@ -60,84 +62,231 @@
#include "sldns/wire2str.h"
#include "sldns/str2wire.h"
static void spool_txt_printf(struct config_strlist_head* txt,
const char* format, ...) ATTR_FORMAT(printf, 2, 3);
/** Append to strlist at end, and log error if out of memory. */
static void
spool_txt_string(struct config_strlist_head* txt, char* str)
{
if(!cfg_strlist_append(txt, strdup(str))) {
log_err("out of memory in spool text");
}
}
/** Spool txt to spool list. */
static void
spool_txt_vmsg(struct config_strlist_head* txt, const char* format,
va_list args)
{
char msg[65535];
vsnprintf(msg, sizeof(msg), format, args);
spool_txt_string(txt, msg);
}
/** Print item to spool list. On alloc failure the list is as before. */
static void
spool_txt_printf(struct config_strlist_head* txt, const char* format, ...)
{
va_list args;
va_start(args, format);
spool_txt_vmsg(txt, format, args);
va_end(args);
}
/** dump one rrset zonefile line */
static int
dump_rrset_line(RES* ssl, struct ub_packed_rrset_key* k, time_t now, size_t i)
static void
dump_rrset_line(struct config_strlist_head* txt, struct ub_packed_rrset_key* k,
time_t now, size_t i)
{
char s[65535];
if(!packed_rr_to_string(k, i, now, s, sizeof(s))) {
return ssl_printf(ssl, "BADRR\n");
spool_txt_string(txt, "BADRR\n");
return;
}
return ssl_printf(ssl, "%s", s);
spool_txt_string(txt, s);
}
/** dump rrset key and data info */
static int
dump_rrset(RES* ssl, struct ub_packed_rrset_key* k,
static void
dump_rrset(struct config_strlist_head* txt, struct ub_packed_rrset_key* k,
struct packed_rrset_data* d, time_t now)
{
size_t i;
/* rd lock held by caller */
if(!k || !d) return 1;
if(k->id == 0) return 1; /* deleted */
if(d->ttl < now) return 1; /* expired */
if(!k || !d) return;
if(k->id == 0) return; /* deleted */
if(d->ttl < now) return; /* expired */
/* meta line */
if(!ssl_printf(ssl, ";rrset%s " ARG_LL "d %u %u %d %d\n",
spool_txt_printf(txt, ";rrset%s " ARG_LL "d %u %u %d %d\n",
(k->rk.flags & PACKED_RRSET_NSEC_AT_APEX)?" nsec_apex":"",
(long long)(d->ttl - now),
(unsigned)d->count, (unsigned)d->rrsig_count,
(int)d->trust, (int)d->security
))
return 0;
);
for(i=0; i<d->count + d->rrsig_count; i++) {
if(!dump_rrset_line(ssl, k, now, i))
dump_rrset_line(txt, k, now, i);
}
}
/** Spool strlist to the output. */
static int
spool_strlist(RES* ssl, struct config_strlist* list)
{
struct config_strlist* s;
for(s=list; s; s=s->next) {
if(!ssl_printf(ssl, "%s", s->str))
return 0;
}
return 1;
}
/** dump lruhash rrset cache */
/** dump lruhash cache and call callback for every item. */
static int
dump_rrset_lruhash(RES* ssl, struct lruhash* h, time_t now)
dump_lruhash(struct lruhash* table,
void (*func)(struct lruhash_entry*, struct config_strlist_head*, void*),
RES* ssl, void* arg)
{
struct lruhash_entry* e;
/* lruhash already locked by caller */
/* walk in order of lru; best first */
for(e=h->lru_start; e; e = e->lru_next) {
lock_rw_rdlock(&e->lock);
if(!dump_rrset(ssl, (struct ub_packed_rrset_key*)e->key,
(struct packed_rrset_data*)e->data, now)) {
lock_rw_unlock(&e->lock);
int just_started = 1;
int not_done = 1;
hashvalue_type hash;
size_t num = 0; /* number of entries processed. */
size_t max = 2; /* number of entries after which it unlocks. */
struct config_strlist_head txt; /* Text strings spooled. */
memset(&txt, 0, sizeof(txt));
while(not_done) {
size_t i; /* hash bin. */
/* Process a number of items. */
num = 0;
lock_quick_lock(&table->lock);
if(just_started) {
i = 0;
} else {
i = hash&table->size_mask;
}
while(num < max) {
/* Process bin. */
int found = 0;
size_t num_bin = 0;
struct lruhash_bin* bin = &table->array[i];
struct lruhash_entry* e;
lock_quick_lock(&bin->lock);
for(e = bin->overflow_list; e; e = e->overflow_next) {
/* Entry e is locked by the func. */
func(e, &txt, arg);
num_bin++;
}
lock_quick_unlock(&bin->lock);
/* This addition of bin number of entries may take
* it over the max. */
num += num_bin;
/* Move to next bin. */
/* Find one with an entry, with a hash value, so we
* can continue from the hash value. The hash value
* can be indexed also if the array changes size. */
i++;
while(i < table->size) {
bin = &table->array[i];
lock_quick_lock(&bin->lock);
if(bin->overflow_list) {
hash = bin->overflow_list->hash;
lock_quick_unlock(&bin->lock);
found = 1;
just_started = 0;
break;
}
lock_quick_unlock(&bin->lock);
i++;
}
if(!found) {
not_done = 0;
break;
}
}
lock_quick_unlock(&table->lock);
/* Print the spooled items, that are collected while the
* locks are locked. The print happens while they are not
* locked. */
if(txt.first) {
if(!spool_strlist(ssl, txt.first)) {
config_delstrlist(txt.first);
return 0;
}
config_delstrlist(txt.first);
memset(&txt, 0, sizeof(txt));
}
}
/* Print the final spooled items. */
if(txt.first) {
if(!spool_strlist(ssl, txt.first)) {
config_delstrlist(txt.first);
return 0;
}
lock_rw_unlock(&e->lock);
config_delstrlist(txt.first);
}
return 1;
}
/** dump slabhash cache and call callback for every item. */
static int
dump_slabhash(struct slabhash* sh,
void (*func)(struct lruhash_entry*, struct config_strlist_head*, void*),
RES* ssl, void* arg)
{
/* Process a number of items at a time, then unlock the cache,
* so that ordinary processing can continue. Keep an iteration marker
* to continue the loop. That means the cache can change, items
* could be inserted and deleted. And, for example, the hash table
* can grow. */
size_t slab;
for(slab=0; slab<sh->size; slab++) {
if(!dump_lruhash(sh->array[slab], func, ssl, arg))
return 0;
}
return 1;
}
/** Struct for dump information. */
struct dump_info {
/** The worker. */
struct worker* worker;
/** The printout connection. */
RES* ssl;
};
/** Dump the rrset cache entry */
static void
dump_rrset_entry(struct lruhash_entry* e, struct config_strlist_head* txt,
void* arg)
{
struct dump_info* dump_info = (struct dump_info*)arg;
lock_rw_rdlock(&e->lock);
dump_rrset(txt, (struct ub_packed_rrset_key*)e->key,
(struct packed_rrset_data*)e->data,
*dump_info->worker->env.now);
lock_rw_unlock(&e->lock);
}
/** dump rrset cache */
static int
dump_rrset_cache(RES* ssl, struct worker* worker)
{
struct rrset_cache* r = worker->env.rrset_cache;
size_t slab;
struct dump_info dump_info;
dump_info.worker = worker;
dump_info.ssl = ssl;
if(!ssl_printf(ssl, "START_RRSET_CACHE\n")) return 0;
for(slab=0; slab<r->table.size; slab++) {
lock_quick_lock(&r->table.array[slab]->lock);
if(!dump_rrset_lruhash(ssl, r->table.array[slab],
*worker->env.now)) {
lock_quick_unlock(&r->table.array[slab]->lock);
return 0;
}
lock_quick_unlock(&r->table.array[slab]->lock);
}
if(!dump_slabhash(&r->table, &dump_rrset_entry, ssl, &dump_info))
return 0;
return ssl_printf(ssl, "END_RRSET_CACHE\n");
}
/** dump message to rrset reference */
static int
dump_msg_ref(RES* ssl, struct ub_packed_rrset_key* k)
static void
dump_msg_ref(struct config_strlist_head* txt, struct ub_packed_rrset_key* k)
{
char* nm, *tp, *cl;
nm = sldns_wire2str_dname(k->rk.dname, k->rk.dname_len);
@@ -147,31 +296,25 @@ dump_msg_ref(RES* ssl, struct ub_packed_rrset_key* k)
free(nm);
free(tp);
free(cl);
return ssl_printf(ssl, "BADREF\n");
}
if(!ssl_printf(ssl, "%s %s %s %d\n", nm, cl, tp, (int)k->rk.flags)) {
free(nm);
free(tp);
free(cl);
return 0;
spool_txt_string(txt, "BADREF\n");
return;
}
spool_txt_printf(txt, "%s %s %s %d\n", nm, cl, tp, (int)k->rk.flags);
free(nm);
free(tp);
free(cl);
return 1;
}
/** dump message entry */
static int
dump_msg(RES* ssl, struct query_info* k, struct reply_info* d,
time_t now)
static void
dump_msg(struct config_strlist_head* txt, struct query_info* k,
struct reply_info* d, time_t now)
{
size_t i;
char* nm, *tp, *cl;
if(!k || !d) return 1;
if(d->ttl < now) return 1; /* expired */
if(!k || !d) return;
if(d->ttl < now) return; /* expired */
nm = sldns_wire2str_dname(k->qname, k->qname_len);
tp = sldns_wire2str_type(k->qtype);
cl = sldns_wire2str_class(k->qclass);
@@ -179,43 +322,35 @@ dump_msg(RES* ssl, struct query_info* k, struct reply_info* d,
free(nm);
free(tp);
free(cl);
return 1; /* skip this entry */
return; /* skip this entry */
}
if(!rrset_array_lock(d->ref, d->rrset_count, now)) {
/* rrsets have timed out or do not exist */
free(nm);
free(tp);
free(cl);
return 1; /* skip this entry */
return; /* skip this entry */
}
/* meta line */
if(!ssl_printf(ssl, "msg %s %s %s %d %d " ARG_LL "d %d %u %u %u\n",
nm, cl, tp,
(int)d->flags, (int)d->qdcount,
(long long)(d->ttl-now), (int)d->security,
(unsigned)d->an_numrrsets,
(unsigned)d->ns_numrrsets,
(unsigned)d->ar_numrrsets)) {
free(nm);
free(tp);
free(cl);
rrset_array_unlock(d->ref, d->rrset_count);
return 0;
}
spool_txt_printf(txt,
"msg %s %s %s %d %d " ARG_LL "d %d %u %u %u %d %s\n",
nm, cl, tp,
(int)d->flags, (int)d->qdcount,
(long long)(d->ttl-now), (int)d->security,
(unsigned)d->an_numrrsets,
(unsigned)d->ns_numrrsets,
(unsigned)d->ar_numrrsets,
(int)d->reason_bogus,
d->reason_bogus_str?d->reason_bogus_str:"");
free(nm);
free(tp);
free(cl);
for(i=0; i<d->rrset_count; i++) {
if(!dump_msg_ref(ssl, d->rrsets[i])) {
rrset_array_unlock(d->ref, d->rrset_count);
return 0;
}
dump_msg_ref(txt, d->rrsets[i]);
}
rrset_array_unlock(d->ref, d->rrset_count);
return 1;
}
/** copy msg to worker pad */
@@ -244,49 +379,40 @@ copy_msg(struct regional* region, struct lruhash_entry* e,
return (*k)->qname != NULL;
}
/** dump lruhash msg cache */
static int
dump_msg_lruhash(RES* ssl, struct worker* worker, struct lruhash* h)
/** Dump the msg entry. */
static void
dump_msg_entry(struct lruhash_entry* e, struct config_strlist_head* txt,
void* arg)
{
struct lruhash_entry* e;
struct dump_info* dump_info = (struct dump_info*)arg;
struct query_info* k;
struct reply_info* d;
/* lruhash already locked by caller */
/* walk in order of lru; best first */
for(e=h->lru_start; e; e = e->lru_next) {
regional_free_all(worker->scratchpad);
lock_rw_rdlock(&e->lock);
/* make copy of rrset in worker buffer */
if(!copy_msg(worker->scratchpad, e, &k, &d)) {
lock_rw_unlock(&e->lock);
return 0;
}
regional_free_all(dump_info->worker->scratchpad);
/* Make copy of rrset in worker buffer. */
lock_rw_rdlock(&e->lock);
if(!copy_msg(dump_info->worker->scratchpad, e, &k, &d)) {
lock_rw_unlock(&e->lock);
/* release lock so we can lookup the rrset references
* in the rrset cache */
if(!dump_msg(ssl, k, d, *worker->env.now)) {
return 0;
}
log_err("out of memory in dump_msg_entry");
return;
}
return 1;
lock_rw_unlock(&e->lock);
/* Release lock so we can lookup the rrset references
* in the rrset cache. */
dump_msg(txt, k, d, *dump_info->worker->env.now);
}
/** dump msg cache */
static int
dump_msg_cache(RES* ssl, struct worker* worker)
{
struct slabhash* sh = worker->env.msg_cache;
size_t slab;
struct dump_info dump_info;
dump_info.worker = worker;
dump_info.ssl = ssl;
if(!ssl_printf(ssl, "START_MSG_CACHE\n")) return 0;
for(slab=0; slab<sh->size; slab++) {
lock_quick_lock(&sh->array[slab]->lock);
if(!dump_msg_lruhash(ssl, worker, sh->array[slab])) {
lock_quick_unlock(&sh->array[slab]->lock);
return 0;
}
lock_quick_unlock(&sh->array[slab]->lock);
}
if(!dump_slabhash(worker->env.msg_cache, &dump_msg_entry, ssl,
&dump_info))
return 0;
return ssl_printf(ssl, "END_MSG_CACHE\n");
}
@@ -385,7 +511,7 @@ move_into_cache(struct ub_packed_rrset_key* k,
struct rrset_ref ref;
uint8_t* p;
ak = alloc_special_obtain(&worker->alloc);
ak = alloc_special_obtain(worker->alloc);
if(!ak) {
log_warn("error out of memory");
return 0;
@@ -396,7 +522,7 @@ move_into_cache(struct ub_packed_rrset_key* k,
ak->rk.dname = (uint8_t*)memdup(k->rk.dname, k->rk.dname_len);
if(!ak->rk.dname) {
log_warn("error out of memory");
ub_packed_rrset_parsedelete(ak, &worker->alloc);
ub_packed_rrset_parsedelete(ak, worker->alloc);
return 0;
}
s = sizeof(*ad) + (sizeof(size_t) + sizeof(uint8_t*) +
@@ -406,7 +532,7 @@ move_into_cache(struct ub_packed_rrset_key* k,
ad = (struct packed_rrset_data*)malloc(s);
if(!ad) {
log_warn("error out of memory");
ub_packed_rrset_parsedelete(ak, &worker->alloc);
ub_packed_rrset_parsedelete(ak, worker->alloc);
return 0;
}
p = (uint8_t*)ad;
@@ -429,7 +555,8 @@ move_into_cache(struct ub_packed_rrset_key* k,
ref.key = ak;
ref.id = ak->id;
(void)rrset_cache_update(worker->env.rrset_cache, &ref,
&worker->alloc, *worker->env.now);
worker->alloc, *worker->env.now);
return 1;
}
@@ -630,6 +757,9 @@ load_msg(RES* ssl, sldns_buffer* buf, struct worker* worker)
long long ttl;
size_t i;
int go_on = 1;
int ede;
int consumed = 0;
char* ede_str = NULL;
regional_free_all(region);
@@ -644,11 +774,16 @@ load_msg(RES* ssl, sldns_buffer* buf, struct worker* worker)
}
/* read remainder of line */
if(sscanf(s, " %u %u " ARG_LL "d %u %u %u %u", &flags, &qdcount, &ttl,
&security, &an, &ns, &ar) != 7) {
/* note the last space before any possible EDE text */
if(sscanf(s, " %u %u " ARG_LL "d %u %u %u %u %d %n", &flags, &qdcount, &ttl,
&security, &an, &ns, &ar, &ede, &consumed) != 8) {
log_warn("error cannot parse numbers: %s", s);
return 0;
}
/* there may be EDE text after the numbers */
if(consumed > 0 && (size_t)consumed < strlen(s))
ede_str = s + consumed;
memset(&rep, 0, sizeof(rep));
rep.flags = (uint16_t)flags;
rep.qdcount = (uint16_t)qdcount;
rep.ttl = (time_t)ttl;
@@ -663,6 +798,8 @@ load_msg(RES* ssl, sldns_buffer* buf, struct worker* worker)
rep.ns_numrrsets = (size_t)ns;
rep.ar_numrrsets = (size_t)ar;
rep.rrset_count = (size_t)an+(size_t)ns+(size_t)ar;
rep.reason_bogus = (sldns_ede_code)ede;
rep.reason_bogus_str = ede_str?(char*)regional_strdup(region, ede_str):NULL;
rep.rrsets = (struct ub_packed_rrset_key**)regional_alloc_zero(
region, sizeof(struct ub_packed_rrset_key*)*rep.rrset_count);
@@ -677,7 +814,8 @@ load_msg(RES* ssl, sldns_buffer* buf, struct worker* worker)
if(!go_on)
return 1; /* skip this one, not all references satisfied */
if(!dns_cache_store(&worker->env, &qinf, &rep, 0, 0, 0, NULL, flags)) {
if(!dns_cache_store(&worker->env, &qinf, &rep, 0, 0, 0, NULL, flags,
*worker->env.now, 1)) {
log_warn("error out of memory");
return 0;
}
@@ -796,12 +934,18 @@ print_dp_main(RES* ssl, struct delegpt* dp, struct dns_msg* msg)
struct ub_packed_rrset_key* k = msg->rep->rrsets[i];
struct packed_rrset_data* d =
(struct packed_rrset_data*)k->entry.data;
struct config_strlist_head txt;
memset(&txt, 0, sizeof(txt));
if(d->security == sec_status_bogus) {
if(!ssl_printf(ssl, "Address is BOGUS:\n"))
return;
}
if(!dump_rrset(ssl, k, d, 0))
dump_rrset(&txt, k, d, 0);
if(!spool_strlist(ssl, txt.first)) {
config_delstrlist(txt.first);
return;
}
config_delstrlist(txt.first);
}
delegpt_count_ns(dp, &n_ns, &n_miss);
delegpt_count_addr(dp, &n_addr, &n_res, &n_avail);
@@ -821,9 +965,10 @@ int print_deleg_lookup(RES* ssl, struct worker* worker, uint8_t* nm,
struct delegpt* dp;
struct dns_msg* msg;
struct regional* region = worker->scratchpad;
char b[260];
char b[LDNS_MAX_DOMAINLEN];
struct query_info qinfo;
struct iter_hints_stub* stub;
int nolock = 0;
regional_free_all(region);
qinfo.qname = nm;
qinfo.qname_len = nmlen;
@@ -835,26 +980,32 @@ int print_deleg_lookup(RES* ssl, struct worker* worker, uint8_t* nm,
if(!ssl_printf(ssl, "The following name servers are used for lookup "
"of %s\n", b))
return 0;
dp = forwards_lookup(worker->env.fwds, nm, qinfo.qclass);
dp = forwards_lookup(worker->env.fwds, nm, qinfo.qclass, nolock);
if(dp) {
if(!ssl_printf(ssl, "forwarding request:\n"))
if(!ssl_printf(ssl, "forwarding request:\n")) {
lock_rw_unlock(&worker->env.fwds->lock);
return 0;
}
print_dp_main(ssl, dp, NULL);
print_dp_details(ssl, worker, dp);
lock_rw_unlock(&worker->env.fwds->lock);
return 1;
}
while(1) {
dp = dns_cache_find_delegation(&worker->env, nm, nmlen,
qinfo.qtype, qinfo.qclass, region, &msg,
*worker->env.now);
*worker->env.now, 0, NULL, 0);
if(!dp) {
return ssl_printf(ssl, "no delegation from "
"cache; goes to configured roots\n");
}
/* go up? */
if(iter_dp_is_useless(&qinfo, BIT_RD, dp)) {
if(iter_dp_is_useless(&qinfo, BIT_RD, dp,
(worker->env.cfg->do_ip4 && worker->back->num_ip4 != 0),
(worker->env.cfg->do_ip6 && worker->back->num_ip6 != 0),
worker->env.cfg->do_nat64)) {
print_dp_main(ssl, dp, msg);
print_dp_details(ssl, worker, dp);
if(!ssl_printf(ssl, "cache delegation was "
@@ -874,21 +1025,26 @@ int print_deleg_lookup(RES* ssl, struct worker* worker, uint8_t* nm,
return 0;
continue;
}
}
}
stub = hints_lookup_stub(worker->env.hints, nm, qinfo.qclass,
dp);
dp, nolock);
if(stub) {
if(stub->noprime) {
if(!ssl_printf(ssl, "The noprime stub servers "
"are used:\n"))
"are used:\n")) {
lock_rw_unlock(&worker->env.hints->lock);
return 0;
}
} else {
if(!ssl_printf(ssl, "The stub is primed "
"with servers:\n"))
"with servers:\n")) {
lock_rw_unlock(&worker->env.hints->lock);
return 0;
}
}
print_dp_main(ssl, stub->dp, NULL);
print_dp_details(ssl, worker, stub->dp);
lock_rw_unlock(&worker->env.hints->lock);
} else {
print_dp_main(ssl, dp, msg);
print_dp_details(ssl, worker, dp);
+236 -27
View File
@@ -91,11 +91,19 @@
#include "util/net_help.h"
#include "sldns/keyraw.h"
#include "respip/respip.h"
#include "iterator/iter_fwd.h"
#include "iterator/iter_hints.h"
#include <signal.h>
#ifdef HAVE_SYSTEMD
#include <systemd/sd-daemon.h>
#endif
#ifdef HAVE_NETDB_H
#include <netdb.h>
#endif
#ifdef USE_CACHEDB
#include "cachedb/cachedb.h"
#endif
/** How many quit requests happened. */
static int sig_record_quit = 0;
@@ -257,6 +265,7 @@ daemon_init(void)
free(daemon);
return NULL;
}
daemon->env->modstack = &daemon->mods;
/* init edns_known_options */
if(!edns_known_options_init(daemon->env)) {
free(daemon->env);
@@ -271,8 +280,17 @@ daemon_init(void)
free(daemon);
return NULL;
}
daemon->acl_interface = acl_list_create();
if(!daemon->acl_interface) {
acl_list_delete(daemon->acl);
edns_known_options_delete(daemon->env);
free(daemon->env);
free(daemon);
return NULL;
}
daemon->tcl = tcl_list_create();
if(!daemon->tcl) {
acl_list_delete(daemon->acl_interface);
acl_list_delete(daemon->acl);
edns_known_options_delete(daemon->env);
free(daemon->env);
@@ -284,6 +302,7 @@ daemon_init(void)
log_err("gettimeofday: %s", strerror(errno));
daemon->time_last_stat = daemon->time_boot;
if((daemon->env->auth_zones = auth_zones_create()) == 0) {
acl_list_delete(daemon->acl_interface);
acl_list_delete(daemon->acl);
tcl_list_delete(daemon->tcl);
edns_known_options_delete(daemon->env);
@@ -293,6 +312,7 @@ daemon_init(void)
}
if(!(daemon->env->edns_strings = edns_strings_create())) {
auth_zones_delete(daemon->env->auth_zones);
acl_list_delete(daemon->acl_interface);
acl_list_delete(daemon->acl);
tcl_list_delete(daemon->tcl);
edns_known_options_delete(daemon->env);
@@ -303,7 +323,27 @@ daemon_init(void)
return daemon;
}
int
int setup_acl_for_ports(struct acl_list* list, struct listen_port* port_list)
{
struct acl_addr* acl_node;
for(; port_list; port_list=port_list->next) {
if(!port_list->socket) {
/* This is mainly for testbound where port_list is
* empty. */
continue;
}
if(!(acl_node = acl_interface_insert(list,
(struct sockaddr_storage*)port_list->socket->addr,
port_list->socket->addrlen,
acl_refuse))) {
return 0;
}
port_list->socket->acl = acl_node;
}
return 1;
}
int
daemon_open_shared_ports(struct daemon* daemon)
{
log_assert(daemon);
@@ -320,6 +360,8 @@ daemon_open_shared_ports(struct daemon* daemon)
free(daemon->ports);
daemon->ports = NULL;
}
/* clean acl_interface */
acl_interface_init(daemon->acl_interface);
if(!resolve_interface_names(daemon->cfg->ifs,
daemon->cfg->num_ifs, NULL, &resif, &num_resif))
return 0;
@@ -329,7 +371,8 @@ daemon_open_shared_ports(struct daemon* daemon)
daemon->reuseport = 1;
#endif
/* try to use reuseport */
p0 = listening_ports_open(daemon->cfg, resif, num_resif, &daemon->reuseport);
p0 = listening_ports_open(daemon->cfg, resif, num_resif,
&daemon->reuseport);
if(!p0) {
listening_ports_free(p0);
config_del_strarray(resif, num_resif);
@@ -350,6 +393,12 @@ daemon_open_shared_ports(struct daemon* daemon)
return 0;
}
daemon->ports[0] = p0;
if(!setup_acl_for_ports(daemon->acl_interface,
daemon->ports[0])) {
listening_ports_free(p0);
config_del_strarray(resif, num_resif);
return 0;
}
if(daemon->reuseport) {
/* continue to use reuseport */
for(i=1; i<daemon->num_ports; i++) {
@@ -365,6 +414,15 @@ daemon_open_shared_ports(struct daemon* daemon)
config_del_strarray(resif, num_resif);
return 0;
}
if(!setup_acl_for_ports(daemon->acl_interface,
daemon->ports[i])) {
for(i=0; i<daemon->num_ports; i++)
listening_ports_free(daemon->ports[i]);
free(daemon->ports);
daemon->ports = NULL;
config_del_strarray(resif, num_resif);
return 0;
}
}
}
config_del_strarray(resif, num_resif);
@@ -385,6 +443,19 @@ daemon_open_shared_ports(struct daemon* daemon)
return 1;
}
int
daemon_privileged(struct daemon* daemon)
{
daemon->env->cfg = daemon->cfg;
daemon->env->alloc = &daemon->superalloc;
daemon->env->worker = NULL;
if(!modstack_call_startup(&daemon->mods, daemon->cfg->module_conf,
daemon->env)) {
fatal_exit("failed to startup modules");
}
return 1;
}
/**
* Setup modules. setup module stack.
* @param daemon: the daemon
@@ -394,11 +465,15 @@ static void daemon_setup_modules(struct daemon* daemon)
daemon->env->cfg = daemon->cfg;
daemon->env->alloc = &daemon->superalloc;
daemon->env->worker = NULL;
daemon->env->need_to_validate = 0; /* set by module init below */
if(!modstack_setup(&daemon->mods, daemon->cfg->module_conf,
daemon->env)) {
fatal_exit("failed to setup modules");
if(daemon->mods_inited) {
modstack_call_deinit(&daemon->mods, daemon->env);
}
daemon->env->need_to_validate = 0; /* set by module init below */
if(!modstack_call_init(&daemon->mods, daemon->cfg->module_conf,
daemon->env)) {
fatal_exit("failed to init modules");
}
daemon->mods_inited = 1;
log_edns_known_options(VERB_ALGO, daemon->env);
}
@@ -433,6 +508,30 @@ static int daemon_get_shufport(struct daemon* daemon, int* shufport)
return avail;
}
/**
* Clear and delete per-worker alloc caches, and free memory maintained in
* superalloc.
* The rrset and message caches must be empty at the time of call.
* @param daemon: the daemon that maintains the alloc caches to be cleared.
*/
static void
daemon_clear_allocs(struct daemon* daemon)
{
int i;
/* daemon->num may be different during reloads (after configuration
* read). Use old_num which has the correct value used to setup the
* worker_allocs */
for(i=0; i<daemon->old_num; i++) {
alloc_clear(daemon->worker_allocs[i]);
free(daemon->worker_allocs[i]);
}
free(daemon->worker_allocs);
daemon->worker_allocs = NULL;
alloc_clear_special(&daemon->superalloc);
}
/**
* Allocate empty worker structures. With backptr and thread-number,
* from 0..numthread initialised. Used as user arguments to new threads.
@@ -457,6 +556,12 @@ daemon_create_workers(struct daemon* daemon)
fatal_exit("out of memory during daemon init");
numport = daemon_get_shufport(daemon, shufport);
verbose(VERB_ALGO, "total of %d outgoing ports available", numport);
#ifdef HAVE_NGTCP2
daemon->doq_table = doq_table_create(daemon->cfg, daemon->rand);
if(!daemon->doq_table)
fatal_exit("could not create doq_table: out of memory");
#endif
daemon->num = (daemon->cfg->num_threads?daemon->cfg->num_threads:1);
if(daemon->reuseport && (int)daemon->num < (int)daemon->num_ports) {
@@ -485,6 +590,21 @@ daemon_create_workers(struct daemon* daemon)
/* the above is not ports/numthr, due to rounding */
fatal_exit("could not create worker");
}
/* create per-worker alloc caches if not reusing existing ones. */
if(!daemon->worker_allocs) {
daemon->worker_allocs = (struct alloc_cache**)calloc(
(size_t)daemon->num, sizeof(struct alloc_cache*));
if(!daemon->worker_allocs)
fatal_exit("could not allocate worker allocs");
for(i=0; i<daemon->num; i++) {
struct alloc_cache* alloc = calloc(1,
sizeof(struct alloc_cache));
if (!alloc)
fatal_exit("could not allocate worker alloc");
alloc_init(alloc, &daemon->superalloc, i);
daemon->worker_allocs[i] = alloc;
}
}
free(shufport);
}
@@ -596,14 +716,17 @@ daemon_fork(struct daemon* daemon)
#endif
log_assert(daemon);
if(!(daemon->views = views_create()))
if(!(daemon->env->views = views_create()))
fatal_exit("Could not create views: out of memory");
/* create individual views and their localzone/data trees */
if(!views_apply_cfg(daemon->views, daemon->cfg))
if(!views_apply_cfg(daemon->env->views, daemon->cfg))
fatal_exit("Could not set up views");
if(!acl_list_apply_cfg(daemon->acl, daemon->cfg, daemon->views))
if(!acl_list_apply_cfg(daemon->acl, daemon->cfg, daemon->env->views))
fatal_exit("Could not setup access control list");
if(!acl_interface_apply_cfg(daemon->acl_interface, daemon->cfg,
daemon->env->views))
fatal_exit("Could not setup interface control list");
if(!tcl_list_apply_cfg(daemon->tcl, daemon->cfg))
fatal_exit("Could not setup TCP connection limits");
if(daemon->cfg->dnscrypt) {
@@ -617,22 +740,36 @@ daemon_fork(struct daemon* daemon)
"dnscrypt support");
#endif
}
if(daemon->cfg->cookie_secret_file &&
daemon->cfg->cookie_secret_file[0]) {
if(!(daemon->cookie_secrets = cookie_secrets_create()))
fatal_exit("Could not create cookie_secrets: out of memory");
if(!cookie_secrets_apply_cfg(daemon->cookie_secrets,
daemon->cfg->cookie_secret_file))
fatal_exit("Could not setup cookie_secrets");
}
/* create global local_zones */
if(!(daemon->local_zones = local_zones_create()))
fatal_exit("Could not create local zones: out of memory");
if(!local_zones_apply_cfg(daemon->local_zones, daemon->cfg))
fatal_exit("Could not set up local zones");
if(!(daemon->env->fwds = forwards_create()) ||
!forwards_apply_cfg(daemon->env->fwds, daemon->cfg))
fatal_exit("Could not set forward zones");
if(!(daemon->env->hints = hints_create()) ||
!hints_apply_cfg(daemon->env->hints, daemon->cfg))
fatal_exit("Could not set root or stub hints");
/* process raw response-ip configuration data */
if(!(daemon->respip_set = respip_set_create()))
if(!(daemon->env->respip_set = respip_set_create()))
fatal_exit("Could not create response IP set");
if(!respip_global_apply_cfg(daemon->respip_set, daemon->cfg))
if(!respip_global_apply_cfg(daemon->env->respip_set, daemon->cfg))
fatal_exit("Could not set up response IP set");
if(!respip_views_apply_cfg(daemon->views, daemon->cfg,
if(!respip_views_apply_cfg(daemon->env->views, daemon->cfg,
&have_view_respip_cfg))
fatal_exit("Could not set up per-view response IP sets");
daemon->use_response_ip = !respip_set_is_empty(daemon->respip_set) ||
have_view_respip_cfg;
daemon->use_response_ip = !respip_set_is_empty(
daemon->env->respip_set) || have_view_respip_cfg;
/* setup modules */
daemon_setup_modules(daemon);
@@ -646,6 +783,10 @@ daemon_fork(struct daemon* daemon)
if(!edns_strings_apply_cfg(daemon->env->edns_strings, daemon->cfg))
fatal_exit("Could not set up EDNS strings");
#ifdef USE_CACHEDB
daemon->env->cachedb_enabled = cachedb_is_enabled(&daemon->mods,
daemon->env);
#endif
/* response-ip-xxx options don't work as expected without the respip
* module. To avoid run-time operational surprise we reject such
* configuration. */
@@ -713,6 +854,7 @@ daemon_fork(struct daemon* daemon)
/* Shutdown SHM */
shm_main_shutdown(daemon);
daemon->reuse_cache = daemon->workers[0]->reuse_cache;
daemon->need_to_exit = daemon->workers[0]->need_to_exit;
}
@@ -727,25 +869,45 @@ daemon_cleanup(struct daemon* daemon)
log_thread_set(NULL);
/* clean up caches because
* a) RRset IDs will be recycled after a reload, causing collisions
* b) validation config can change, thus rrset, msg, keycache clear */
slabhash_clear(&daemon->env->rrset_cache->table);
slabhash_clear(daemon->env->msg_cache);
* b) validation config can change, thus rrset, msg, keycache clear
*
* If we are trying to keep the cache as long as possible, we should
* defer the cleanup until we know whether the new configuration allows
* the reuse. (If we're exiting, cleanup should be done here). */
if(!daemon->reuse_cache || daemon->need_to_exit) {
slabhash_clear(&daemon->env->rrset_cache->table);
slabhash_clear(daemon->env->msg_cache);
}
daemon->old_num = daemon->num; /* save the current num */
forwards_delete(daemon->env->fwds);
daemon->env->fwds = NULL;
hints_delete(daemon->env->hints);
daemon->env->hints = NULL;
local_zones_delete(daemon->local_zones);
daemon->local_zones = NULL;
respip_set_delete(daemon->respip_set);
daemon->respip_set = NULL;
views_delete(daemon->views);
daemon->views = NULL;
respip_set_delete(daemon->env->respip_set);
daemon->env->respip_set = NULL;
views_delete(daemon->env->views);
daemon->env->views = NULL;
if(daemon->env->auth_zones)
auth_zones_cleanup(daemon->env->auth_zones);
/* key cache is cleared by module desetup during next daemon_fork() */
/* key cache is cleared by module deinit during next daemon_fork() */
daemon_remote_clear(daemon->rc);
if(daemon->fast_reload_thread)
fast_reload_thread_stop(daemon->fast_reload_thread);
if(daemon->fast_reload_printq_list)
fast_reload_printq_list_delete(daemon->fast_reload_printq_list);
for(i=0; i<daemon->num; i++)
worker_delete(daemon->workers[i]);
free(daemon->workers);
daemon->workers = NULL;
/* Unless we're trying to keep the cache, worker alloc_caches should be
* cleared and freed here. We do this after deleting workers to
* guarantee that the alloc caches are valid throughout the lifetime
* of workers. */
if(!daemon->reuse_cache || daemon->need_to_exit)
daemon_clear_allocs(daemon);
daemon->num = 0;
alloc_clear_special(&daemon->superalloc);
#ifdef USE_DNSTAP
dt_delete(daemon->dtenv);
daemon->dtenv = NULL;
@@ -753,6 +915,10 @@ daemon_cleanup(struct daemon* daemon)
#ifdef USE_DNSCRYPT
dnsc_delete(daemon->dnscenv);
daemon->dnscenv = NULL;
#endif
#ifdef HAVE_NGTCP2
doq_table_delete(daemon->doq_table);
daemon->doq_table = NULL;
#endif
daemon->cfg = NULL;
}
@@ -763,7 +929,9 @@ daemon_delete(struct daemon* daemon)
size_t i;
if(!daemon)
return;
modstack_desetup(&daemon->mods, daemon->env);
modstack_call_deinit(&daemon->mods, daemon->env);
modstack_call_destartup(&daemon->mods, daemon->env);
modstack_free(&daemon->mods);
daemon_remote_delete(daemon->rc);
for(i = 0; i < daemon->num_ports; i++)
listening_ports_free(daemon->ports[i]);
@@ -780,22 +948,29 @@ daemon_delete(struct daemon* daemon)
ub_randfree(daemon->rand);
alloc_clear(&daemon->superalloc);
acl_list_delete(daemon->acl);
acl_list_delete(daemon->acl_interface);
tcl_list_delete(daemon->tcl);
cookie_secrets_delete(daemon->cookie_secrets);
listen_desetup_locks();
free(daemon->chroot);
free(daemon->pidfile);
free(daemon->cfgfile);
free(daemon->env);
#ifdef HAVE_SSL
listen_sslctx_delete_ticket_keys();
SSL_CTX_free((SSL_CTX*)daemon->listen_sslctx);
SSL_CTX_free((SSL_CTX*)daemon->connect_sslctx);
SSL_CTX_free((SSL_CTX*)daemon->listen_dot_sslctx);
SSL_CTX_free((SSL_CTX*)daemon->listen_doh_sslctx);
SSL_CTX_free((SSL_CTX*)daemon->connect_dot_sslctx);
#endif
#ifdef HAVE_NGTCP2
SSL_CTX_free((SSL_CTX*)daemon->listen_quic_sslctx);
#endif
free(daemon);
/* lex cleanup */
ub_c_lex_destroy();
/* libcrypto cleanup */
#ifdef HAVE_SSL
# if defined(USE_GOST) && defined(HAVE_LDNS_KEY_EVP_UNLOAD_GOST)
# if defined(USE_GOST)
sldns_key_EVP_unload_gost();
# endif
# if HAVE_DECL_SSL_COMP_GET_COMPRESSION_METHODS && HAVE_DECL_SK_SSL_COMP_POP_FREE
@@ -841,8 +1016,42 @@ daemon_delete(struct daemon* daemon)
void daemon_apply_cfg(struct daemon* daemon, struct config_file* cfg)
{
int new_num = cfg->num_threads?cfg->num_threads:1;
daemon->cfg = cfg;
config_apply(cfg);
/* If this is a reload and we deferred the decision on whether to
* reuse the alloc, RRset, and message caches, then check to see if
* it's safe to keep the caches:
* - changing the number of threads is obviously incompatible with
* keeping the per-thread alloc caches. It also means we have to
* clear RRset and message caches. (note that 'new_num' may be
* adjusted in daemon_create_workers, but for our purpose we can
* simply compare it with 'old_num'; if they are equal here,
* 'new_num' won't be adjusted to a different value than 'old_num').
* - changing RRset cache size effectively clears any remaining cache
* entries. We could keep their keys in alloc caches, but it would
* be more consistent with the sense of the change to clear allocs
* and free memory. To do so we also have to clear message cache.
* - only changing message cache size does not necessarily affect
* RRset or alloc cache. But almost all new subsequent queries will
* require recursive resolution anyway, so it doesn't help much to
* just keep RRset and alloc caches. For simplicity we clear/free
* the other two, too. */
if(daemon->worker_allocs &&
(new_num != daemon->old_num ||
!slabhash_is_size(daemon->env->msg_cache, cfg->msg_cache_size,
cfg->msg_cache_slabs) ||
!slabhash_is_size(&daemon->env->rrset_cache->table,
cfg->rrset_cache_size, cfg->rrset_cache_slabs)))
{
log_warn("cannot reuse caches due to critical config change");
slabhash_clear(&daemon->env->rrset_cache->table);
slabhash_clear(daemon->env->msg_cache);
daemon_clear_allocs(daemon);
}
if(!slabhash_is_size(daemon->env->msg_cache, cfg->msg_cache_size,
cfg->msg_cache_slabs)) {
slabhash_delete(daemon->env->msg_cache);
+54 -6
View File
@@ -58,6 +58,10 @@ struct ub_randstate;
struct daemon_remote;
struct respip_set;
struct shm_main_info;
struct doq_table;
struct cookie_secrets;
struct fast_reload_thread;
struct fast_reload_printq;
#include "dnstap/dnstap_config.h"
#ifdef USE_DNSTAP
@@ -95,12 +99,22 @@ struct daemon {
struct listen_port* rc_ports;
/** remote control connections management (for first worker) */
struct daemon_remote* rc;
/** ssl context for listening to dnstcp over ssl, and connecting ssl */
void* listen_sslctx, *connect_sslctx;
/** ssl context for listening to dnstcp over ssl */
void* listen_dot_sslctx;
/** ssl context for connecting to dnstcp over ssl */
void* connect_dot_sslctx;
/** ssl context for listening to DoH */
void* listen_doh_sslctx;
/** ssl context for listening to quic */
void* listen_quic_sslctx;
/** num threads allocated */
int num;
/** num threads allocated in the previous config or 0 at first */
int old_num;
/** the worker entries */
struct worker** workers;
/** per-worker allocation cache */
struct alloc_cache **worker_allocs;
/** do we need to exit unbound (or is it only a reload?) */
int need_to_exit;
/** master random table ; used for port div between threads on reload*/
@@ -111,8 +125,12 @@ struct daemon {
struct module_env* env;
/** stack of module callbacks */
struct module_stack mods;
/** The module stack has been inited */
int mods_inited;
/** access control, which client IPs are allowed to connect */
struct acl_list* acl;
/** access control, which interfaces are allowed to connect */
struct acl_list* acl_interface;
/** TCP connection limit, limit connections from client IPs */
struct tcl_list* tcl;
/** local authority zones */
@@ -121,15 +139,11 @@ struct daemon {
struct timeval time_last_stat;
/** time when daemon started */
struct timeval time_boot;
/** views structure containing view tree */
struct views* views;
#ifdef USE_DNSTAP
/** the dnstap environment master value, copied and changed by threads*/
struct dt_env* dtenv;
#endif
struct shm_main_info* shm_info;
/** response-ip set with associated actions and tags. */
struct respip_set* respip_set;
/** some response-ip tags or actions are configured if true */
int use_response_ip;
/** some RPZ policies are configured */
@@ -138,6 +152,23 @@ struct daemon {
/** the dnscrypt environment */
struct dnsc_env* dnscenv;
#endif
/** the doq connection table */
struct doq_table* doq_table;
/** reuse existing cache on reload if other conditions allow it. */
int reuse_cache;
/** the EDNS cookie secrets from the cookie-secret-file */
struct cookie_secrets* cookie_secrets;
/** the fast reload thread, or NULL */
struct fast_reload_thread* fast_reload_thread;
/** the fast reload printq list */
struct fast_reload_printq* fast_reload_printq_list;
/** the fast reload option to drop mesh queries, true if so. */
int fast_reload_drop_mesh;
/** for fast reload, if the tcl, tcp connection limits, has
* changes for workers */
int fast_reload_tcl_has_changes;
/** config file name */
char* cfgfile;
};
/**
@@ -154,6 +185,15 @@ struct daemon* daemon_init(void);
*/
int daemon_open_shared_ports(struct daemon* daemon);
/**
* Do daemon setup that needs privileges
* like opening privileged ports or opening device files.
* The cfg member pointer must have been set for the daemon.
* @param daemon: the daemon.
* @return: false on error.
*/
int daemon_privileged(struct daemon* daemon);
/**
* Fork workers and start service.
* When the routine exits, it is no longer forked.
@@ -181,4 +221,12 @@ void daemon_delete(struct daemon* daemon);
*/
void daemon_apply_cfg(struct daemon* daemon, struct config_file* cfg);
/**
* Setup acl list to have entries for the port list.
* @param list: the acl interface
* @param port_list: list of open ports, or none.
* @return false on failure
*/
int setup_acl_for_ports(struct acl_list* list, struct listen_port* port_list);
#endif /* DAEMON_H */
+5191 -310
View File
File diff suppressed because it is too large Load Diff
+168 -1
View File
@@ -46,8 +46,9 @@
#ifndef DAEMON_REMOTE_H
#define DAEMON_REMOTE_H
#ifdef HAVE_OPENSSL_SSL_H
#include "openssl/ssl.h"
#include <openssl/ssl.h>
#endif
#include "util/locks.h"
struct config_file;
struct listen_list;
struct listen_port;
@@ -55,6 +56,7 @@ struct worker;
struct comm_reply;
struct comm_point;
struct daemon_remote;
struct config_strlist_head;
/** number of milliseconds timeout on incoming remote control handshake */
#define REMOTE_CONTROL_TCP_TIMEOUT 120000
@@ -118,6 +120,137 @@ struct remote_stream {
};
typedef struct remote_stream RES;
/**
* Notification status. This is exchanged between the fast reload thread
* and the server thread, over the commpair sockets.
*/
enum fast_reload_notification {
/** nothing, not used */
fast_reload_notification_none = 0,
/** the fast reload thread is done */
fast_reload_notification_done = 1,
/** the fast reload thread is done but with an error, it failed */
fast_reload_notification_done_error = 2,
/** the fast reload thread is told to exit by the server thread.
* Sent on server quit while the reload is running. */
fast_reload_notification_exit = 3,
/** the fast reload thread has exited, after being told to exit */
fast_reload_notification_exited = 4,
/** the fast reload thread has information to print out */
fast_reload_notification_printout = 5,
/** stop as part of the reload the thread and other threads */
fast_reload_notification_reload_stop = 6,
/** ack the stop as part of the reload, and also ack start */
fast_reload_notification_reload_ack = 7,
/** resume from stop as part of the reload */
fast_reload_notification_reload_start = 8,
/** the fast reload thread wants the mainthread to poll workers,
* after the reload, sent when nopause is used */
fast_reload_notification_reload_nopause_poll = 9
};
/**
* Fast reload printout queue. Contains a list of strings, that need to be
* printed over the file descriptor.
*/
struct fast_reload_printq {
/** if this item is in a list, the previous and next */
struct fast_reload_printq *prev, *next;
/** if this item is in a list, it is true. */
int in_list;
/** list of strings to printout */
struct config_strlist_head* to_print;
/** the current item to print. It is malloced. NULL if none. */
char* client_item;
/** The length, strlen, of the client_item, that has to be sent. */
int client_len;
/** The number of bytes sent of client_item. */
int client_byte_count;
/** the comm point for the client connection, the remote control
* client. */
struct comm_point* client_cp;
/** the remote control connection to print output to. */
struct remote_stream remote;
/** the worker that the event is added in */
struct worker* worker;
};
/**
* Fast reload auth zone change. Keeps track if an auth zone was removed,
* added or changed. This is needed because workers can have events for
* dealing with auth zones, like transfers, and those have to be removed
* too, not just the auth zone structure from the tree. */
struct fast_reload_auth_change {
/** next in the list of auth zone changes. */
struct fast_reload_auth_change* next;
/** the zone in the old config */
struct auth_zone* old_z;
/** the zone in the new config */
struct auth_zone* new_z;
/** if the zone was deleted */
int is_deleted;
/** if the zone was added */
int is_added;
/** if the zone has been changed */
int is_changed;
};
/**
* Fast reload thread structure
*/
struct fast_reload_thread {
/** the thread number for the dtio thread,
* must be first to cast thread arg to int* in checklock code. */
int threadnum;
/** communication socket pair, that sends commands */
int commpair[2];
/** thread id, of the io thread */
ub_thread_type tid;
/** if the io processing has started */
int started;
/** if the thread has to quit */
int need_to_quit;
/** verbosity of the fast_reload command, the number of +v options */
int fr_verb;
/** option to not pause threads during reload */
int fr_nopause;
/** option to drop mesh queries */
int fr_drop_mesh;
/** the event that listens on the remote service worker to the
* commpair, it receives content from the fast reload thread. */
void* service_event;
/** if the event that listens on the remote service worker has
* been added to the comm base. */
int service_event_is_added;
/** the service event can read a cmd, nonblocking, so it can
* save the partial read cmd here */
uint32_t service_read_cmd;
/** the number of bytes in service_read_cmd */
int service_read_cmd_count;
/** the worker that the service_event is added in */
struct worker* worker;
/** the printout of output to the remote client. */
struct fast_reload_printq *printq;
/** lock on fr_output, to stop race when both remote control thread
* and fast reload thread use fr_output list. */
lock_basic_type fr_output_lock;
/** list of strings, that the fast reload thread produces that have
* to be printed. The remote control thread can pick them up with
* the lock. */
struct config_strlist_head* fr_output;
/** communication socket pair, to respond to the reload request */
int commreload[2];
/** the list of auth zone changes. */
struct fast_reload_auth_change* auth_zone_change_list;
/** the old tree of auth zones, to lookup. */
struct auth_zones* old_auth_zones;
};
/**
* Create new remote control state for the daemon.
* @param cfg: config file with key file settings.
@@ -203,4 +336,38 @@ int ssl_printf(RES* ssl, const char* format, ...)
int ssl_read_line(RES* ssl, char* buf, size_t max);
#endif /* HAVE_SSL */
/**
* Start fast reload thread
* @param ssl: the RES connection to print to.
* @param worker: the remote servicing worker.
* @param s: the rc_state that is servicing the remote control connection to
* the remote control client. It needs to be moved away to stay connected
* while the fast reload is running.
* @param fr_verb: verbosity to print output at. 0 is nothing, 1 is some
* and 2 is more detail.
* @param fr_nopause: option to not pause threads during reload.
* @param fr_drop_mesh: option to drop mesh queries.
*/
void fast_reload_thread_start(RES* ssl, struct worker* worker,
struct rc_state* s, int fr_verb, int fr_nopause, int fr_drop_mesh);
/**
* Stop fast reload thread
* @param fast_reload_thread: the thread struct.
*/
void fast_reload_thread_stop(struct fast_reload_thread* fast_reload_thread);
/** fast reload thread commands to remote service thread event callback */
void fast_reload_service_cb(int fd, short bits, void* arg);
/** fast reload callback for the remote control client connection */
int fast_reload_client_callback(struct comm_point* c, void* arg, int err,
struct comm_reply* rep);
/** fast reload printq delete list */
void fast_reload_printq_list_delete(struct fast_reload_printq* list);
/** Pick up per worker changes after a fast reload. */
void fast_reload_worker_pickup_changes(struct worker* worker);
#endif /* DAEMON_REMOTE_H */
+107 -30
View File
@@ -4,22 +4,22 @@
* Copyright (c) 2007, NLnet Labs. All rights reserved.
*
* This software is open source.
*
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
*
*
* Redistributions of source code must retain the above copyright notice,
* this list of conditions and the following disclaimer.
*
*
* Redistributions in binary form must reproduce the above copyright notice,
* this list of conditions and the following disclaimer in the documentation
* and/or other materials provided with the distribution.
*
*
* Neither the name of the NLNET LABS nor the names of its contributors may
* be used to endorse or promote products derived from this software without
* specific prior written permission.
*
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
* "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
* LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
@@ -70,6 +70,9 @@
#include <openssl/ssl.h>
#endif
/** How long to wait for threads to transmit statistics, in msec. */
#define STATS_THREAD_WAIT 60000
/** add timers and the values do not overflow or become negative */
static void
stats_timeval_add(long long* d_sec, long long* d_usec, long long add_sec, long long add_usec)
@@ -113,8 +116,8 @@ void server_stats_log(struct ub_server_stats* stats, struct worker* worker,
log_info("server stats for thread %d: %u queries, "
"%u answers from cache, %u recursions, %u prefetch, %u rejected by "
"ip ratelimiting",
threadnum, (unsigned)stats->num_queries,
(unsigned)(stats->num_queries -
threadnum, (unsigned)stats->num_queries,
(unsigned)(stats->num_queries -
stats->num_queries_missed_cache),
(unsigned)stats->num_queries_missed_cache,
(unsigned)stats->num_queries_prefetch,
@@ -270,17 +273,25 @@ server_stats_compile(struct worker* worker, struct ub_stats_info* s, int reset)
/* add in the values from the mesh */
s->svr.ans_secure += (long long)worker->env.mesh->ans_secure;
s->svr.ans_bogus += (long long)worker->env.mesh->ans_bogus;
s->svr.val_ops += (long long)worker->env.mesh->val_ops;
s->svr.ans_rcode_nodata += (long long)worker->env.mesh->ans_nodata;
s->svr.ans_expired += (long long)worker->env.mesh->ans_expired;
for(i=0; i<UB_STATS_RCODE_NUM; i++)
s->svr.ans_rcode[i] += (long long)worker->env.mesh->ans_rcode[i];
for(i=0; i<UB_STATS_RPZ_ACTION_NUM; i++)
s->svr.rpz_action[i] += (long long)worker->env.mesh->rpz_action[i];
timehist_export(worker->env.mesh->histogram, s->svr.hist,
timehist_export(worker->env.mesh->histogram, s->svr.hist,
NUM_BUCKETS_HIST);
s->svr.num_queries_discard_timeout +=
(long long)worker->env.mesh->num_queries_discard_timeout;
s->svr.num_queries_wait_limit +=
(long long)worker->env.mesh->num_queries_wait_limit;
s->svr.num_dns_error_reports +=
(long long)worker->env.mesh->num_dns_error_reports;
/* values from outside network */
s->svr.unwanted_replies = (long long)worker->back->unwanted_replies;
s->svr.qtcp_outgoing = (long long)worker->back->num_tcp_outgoing;
s->svr.qudp_outgoing = (long long)worker->back->num_udp_outgoing;
/* get and reset validator rrset bogus number */
s->svr.rrset_bogus = (long long)get_rrset_bogus(worker, reset);
@@ -289,8 +300,10 @@ server_stats_compile(struct worker* worker, struct ub_stats_info* s, int reset)
s->svr.queries_ratelimited = (long long)get_queries_ratelimit(worker, reset);
/* get cache sizes */
s->svr.msg_cache_count = (long long)count_slabhash_entries(worker->env.msg_cache);
s->svr.rrset_cache_count = (long long)count_slabhash_entries(&worker->env.rrset_cache->table);
get_slabhash_stats(worker->env.msg_cache,
&s->svr.msg_cache_count, &s->svr.msg_cache_max_collisions);
get_slabhash_stats(&worker->env.rrset_cache->table,
&s->svr.rrset_cache_count, &s->svr.rrset_cache_max_collisions);
s->svr.infra_cache_count = (long long)count_slabhash_entries(worker->env.infra_cache->hosts);
if(worker->env.key_cache)
s->svr.key_cache_count = (long long)count_slabhash_entries(worker->env.key_cache->slab);
@@ -319,20 +332,8 @@ server_stats_compile(struct worker* worker, struct ub_stats_info* s, int reset)
s->svr.num_query_dnscrypt_replay = 0;
#endif /* USE_DNSCRYPT */
if(worker->env.auth_zones) {
if(reset && !worker->env.cfg->stat_cumulative) {
lock_rw_wrlock(&worker->env.auth_zones->lock);
} else {
lock_rw_rdlock(&worker->env.auth_zones->lock);
}
s->svr.num_query_authzone_up = (long long)worker->env.
auth_zones->num_query_up;
s->svr.num_query_authzone_down = (long long)worker->env.
auth_zones->num_query_down;
if(reset && !worker->env.cfg->stat_cumulative) {
worker->env.auth_zones->num_query_up = 0;
worker->env.auth_zones->num_query_down = 0;
}
lock_rw_unlock(&worker->env.auth_zones->lock);
s->svr.num_query_authzone_up += (long long)worker->env.mesh->num_query_authzone_up;
s->svr.num_query_authzone_down += (long long)worker->env.mesh->num_query_authzone_down;
}
s->svr.mem_stream_wait =
(long long)tcp_req_info_get_stream_buffer_size();
@@ -340,6 +341,12 @@ server_stats_compile(struct worker* worker, struct ub_stats_info* s, int reset)
(long long)http2_get_query_buffer_size();
s->svr.mem_http2_response_buffer =
(long long)http2_get_response_buffer_size();
#ifdef HAVE_NGTCP2
s->svr.mem_quic = (long long)doq_table_quic_size_get(
worker->daemon->doq_table);
#else
s->svr.mem_quic = 0;
#endif /* HAVE_NGTCP2 */
/* Set neg cache usage numbers */
set_neg_cache_stats(worker, &s->svr, reset);
@@ -350,6 +357,11 @@ server_stats_compile(struct worker* worker, struct ub_stats_info* s, int reset)
s->svr.num_query_subnet = 0;
s->svr.num_query_subnet_cache = 0;
#endif
#ifdef USE_CACHEDB
s->svr.num_query_cachedb = (long long)worker->env.mesh->ans_cachedb;
#else
s->svr.num_query_cachedb = 0;
#endif
/* get tcp accept usage */
s->svr.tcp_accept_usage = 0;
@@ -379,6 +391,35 @@ void server_stats_obtain(struct worker* worker, struct worker* who,
worker_send_cmd(who, worker_cmd_stats);
else worker_send_cmd(who, worker_cmd_stats_noreset);
verbose(VERB_ALGO, "wait for stats reply");
if(tube_wait_timeout(worker->cmd, STATS_THREAD_WAIT) == 0) {
#if defined(HAVE_PTHREAD) && defined(SIZEOF_PTHREAD_T) && defined(SIZEOF_UNSIGNED_LONG)
# if SIZEOF_PTHREAD_T == SIZEOF_UNSIGNED_LONG
unsigned long pthid = 0;
if(verbosity >= VERB_OPS)
memcpy(&pthid, &who->thr_id, sizeof(unsigned long));
# endif
#endif
verbose(VERB_OPS, "no response from thread %d"
#ifdef HAVE_GETTID
" LWP %u"
#endif
#if defined(HAVE_PTHREAD) && defined(SIZEOF_PTHREAD_T) && defined(SIZEOF_UNSIGNED_LONG)
# if SIZEOF_PTHREAD_T == SIZEOF_UNSIGNED_LONG
" pthread 0x%lx"
# endif
#endif
,
who->thread_num
#ifdef HAVE_GETTID
, (unsigned)who->thread_tid
#endif
#if defined(HAVE_PTHREAD) && defined(SIZEOF_PTHREAD_T) && defined(SIZEOF_UNSIGNED_LONG)
# if SIZEOF_PTHREAD_T == SIZEOF_UNSIGNED_LONG
, pthid
# endif
#endif
);
}
if(!tube_read_msg(worker->cmd, &reply, &len, 0))
fatal_exit("failed to read stats over cmd channel");
if(len != (uint32_t)sizeof(*s))
@@ -393,7 +434,7 @@ void server_stats_reply(struct worker* worker, int reset)
struct ub_stats_info s;
server_stats_compile(worker, &s, reset);
verbose(VERB_ALGO, "write stats replymsg");
if(!tube_write_msg(worker->daemon->workers[0]->cmd,
if(!tube_write_msg(worker->daemon->workers[0]->cmd,
(uint8_t*)&s, sizeof(s), 0))
fatal_exit("could not write stat values over cmd channel");
}
@@ -402,16 +443,28 @@ void server_stats_add(struct ub_stats_info* total, struct ub_stats_info* a)
{
total->svr.num_queries += a->svr.num_queries;
total->svr.num_queries_ip_ratelimited += a->svr.num_queries_ip_ratelimited;
total->svr.num_queries_cookie_valid += a->svr.num_queries_cookie_valid;
total->svr.num_queries_cookie_client += a->svr.num_queries_cookie_client;
total->svr.num_queries_cookie_invalid += a->svr.num_queries_cookie_invalid;
total->svr.num_queries_discard_timeout +=
a->svr.num_queries_discard_timeout;
total->svr.num_queries_wait_limit += a->svr.num_queries_wait_limit;
total->svr.num_dns_error_reports += a->svr.num_dns_error_reports;
total->svr.num_queries_missed_cache += a->svr.num_queries_missed_cache;
total->svr.num_queries_prefetch += a->svr.num_queries_prefetch;
total->svr.num_queries_timed_out += a->svr.num_queries_timed_out;
total->svr.num_query_authzone_up += a->svr.num_query_authzone_up;
total->svr.num_query_authzone_down += a->svr.num_query_authzone_down;
if (total->svr.max_query_time_us < a->svr.max_query_time_us)
total->svr.max_query_time_us = a->svr.max_query_time_us;
total->svr.sum_query_list_size += a->svr.sum_query_list_size;
total->svr.ans_expired += a->svr.ans_expired;
#ifdef USE_DNSCRYPT
total->svr.num_query_dnscrypt_crypted += a->svr.num_query_dnscrypt_crypted;
total->svr.num_query_dnscrypt_cert += a->svr.num_query_dnscrypt_cert;
total->svr.num_query_dnscrypt_cleartext += \
total->svr.num_query_dnscrypt_cleartext +=
a->svr.num_query_dnscrypt_cleartext;
total->svr.num_query_dnscrypt_crypted_malformed += \
total->svr.num_query_dnscrypt_crypted_malformed +=
a->svr.num_query_dnscrypt_crypted_malformed;
#endif /* USE_DNSCRYPT */
/* the max size reached is upped to higher of both */
@@ -424,9 +477,11 @@ void server_stats_add(struct ub_stats_info* total, struct ub_stats_info* a)
total->svr.qclass_big += a->svr.qclass_big;
total->svr.qtcp += a->svr.qtcp;
total->svr.qtcp_outgoing += a->svr.qtcp_outgoing;
total->svr.qudp_outgoing += a->svr.qudp_outgoing;
total->svr.qtls += a->svr.qtls;
total->svr.qtls_resume += a->svr.qtls_resume;
total->svr.qhttps += a->svr.qhttps;
total->svr.qquic += a->svr.qquic;
total->svr.qipv6 += a->svr.qipv6;
total->svr.qbit_QR += a->svr.qbit_QR;
total->svr.qbit_AA += a->svr.qbit_AA;
@@ -441,9 +496,13 @@ void server_stats_add(struct ub_stats_info* total, struct ub_stats_info* a)
total->svr.ans_rcode_nodata += a->svr.ans_rcode_nodata;
total->svr.ans_secure += a->svr.ans_secure;
total->svr.ans_bogus += a->svr.ans_bogus;
total->svr.val_ops += a->svr.val_ops;
total->svr.unwanted_replies += a->svr.unwanted_replies;
total->svr.unwanted_queries += a->svr.unwanted_queries;
total->svr.tcp_accept_usage += a->svr.tcp_accept_usage;
#ifdef USE_CACHEDB
total->svr.num_query_cachedb += a->svr.num_query_cachedb;
#endif
for(i=0; i<UB_STATS_QTYPE_NUM; i++)
total->svr.qtype[i] += a->svr.qtype[i];
for(i=0; i<UB_STATS_QCLASS_NUM; i++)
@@ -483,18 +542,23 @@ void server_stats_insquery(struct ub_server_stats* stats, struct comm_point* c,
else stats->qclass_big++;
stats->qopcode[ LDNS_OPCODE_WIRE(sldns_buffer_begin(c->buffer)) ]++;
if(c->type != comm_udp) {
stats->qtcp++;
if(c->type != comm_doq)
stats->qtcp++;
if(c->ssl != NULL) {
stats->qtls++;
#ifdef HAVE_SSL
if(SSL_session_reused(c->ssl))
if(SSL_session_reused(c->ssl))
stats->qtls_resume++;
#endif
if(c->type == comm_http)
stats->qhttps++;
#ifdef HAVE_NGTCP2
else if(c->type == comm_doq)
stats->qquic++;
#endif
}
}
if(repinfo && addr_is_ip6(&repinfo->addr, repinfo->addrlen))
if(repinfo && addr_is_ip6(&repinfo->remote_addr, repinfo->remote_addrlen))
stats->qipv6++;
if( (flags&BIT_QR) )
stats->qbit_QR++;
@@ -528,3 +592,16 @@ void server_stats_insrcode(struct ub_server_stats* stats, sldns_buffer* buf)
stats->ans_rcode_nodata ++;
}
}
void server_stats_downstream_cookie(struct ub_server_stats* stats,
struct edns_data* edns)
{
if(!(edns->edns_present && edns->cookie_present)) return;
if(edns->cookie_valid) {
stats->num_queries_cookie_valid++;
} else if(edns->cookie_client) {
stats->num_queries_cookie_client++;
} else {
stats->num_queries_cookie_invalid++;
}
}
+7
View File
@@ -126,4 +126,11 @@ void server_stats_insquery(struct ub_server_stats* stats, struct comm_point* c,
*/
void server_stats_insrcode(struct ub_server_stats* stats, struct sldns_buffer* buf);
/**
* Add DNS Cookie stats for this query
* @param stats: the stats
* @param edns: edns record
*/
void server_stats_downstream_cookie(struct ub_server_stats* stats,
struct edns_data* edns);
#endif /* DAEMON_STATS_H */
+85 -51
View File
@@ -174,7 +174,7 @@ static void
checkrlimits(struct config_file* cfg)
{
#ifndef S_SPLINT_S
#ifdef HAVE_GETRLIMIT
#if defined(HAVE_GETRLIMIT) && !defined(unbound_testbound)
/* list has number of ports to listen to, ifs number addresses */
int list = ((cfg->do_udp?1:0) + (cfg->do_tcp?1 +
(int)cfg->incoming_num_tcp:0));
@@ -366,9 +366,8 @@ readpid (const char* file)
/** write pid to file.
* @param pidfile: file name of pid file.
* @param pid: pid to write to file.
* @return false on failure
*/
static int
static void
writepid (const char* pidfile, pid_t pid)
{
int fd;
@@ -383,7 +382,7 @@ writepid (const char* pidfile, pid_t pid)
, 0644)) == -1) {
log_err("cannot open pidfile %s: %s",
pidfile, strerror(errno));
return 0;
return;
}
while(count < strlen(pidbuf)) {
ssize_t r = write(fd, pidbuf+count, strlen(pidbuf)-count);
@@ -393,17 +392,16 @@ writepid (const char* pidfile, pid_t pid)
log_err("cannot write to pidfile %s: %s",
pidfile, strerror(errno));
close(fd);
return 0;
return;
} else if(r == 0) {
log_err("cannot write any bytes to pidfile %s: "
"write returns 0 bytes written", pidfile);
close(fd);
return 0;
return;
}
count += r;
}
close(fd);
return 1;
}
/**
@@ -465,6 +463,62 @@ detach(void)
#endif /* HAVE_DAEMON */
}
#ifdef HAVE_SSL
/* setup a listening ssl context, fatal_exit() on any failure */
static void
setup_listen_sslctx(void** ctx, int is_dot, int is_doh, struct config_file* cfg)
{
if(!(*ctx = listen_sslctx_create(
cfg->ssl_service_key, cfg->ssl_service_pem, NULL,
cfg->tls_ciphers, cfg->tls_ciphersuites,
(cfg->tls_session_ticket_keys.first &&
cfg->tls_session_ticket_keys.first->str[0] != 0),
is_dot, is_doh))) {
fatal_exit("could not set up listen SSL_CTX");
}
}
#endif /* HAVE_SSL */
/* setups the needed ssl contexts, fatal_exit() on any failure */
static void
setup_sslctxs(struct daemon* daemon, struct config_file* cfg)
{
#ifdef HAVE_SSL
if(!(daemon->rc = daemon_remote_create(cfg)))
fatal_exit("could not set up remote-control");
if(cfg->ssl_service_key && cfg->ssl_service_key[0]) {
/* setup the session keys; the callback to use them will be
* attached to each sslctx separately */
if(cfg->tls_session_ticket_keys.first &&
cfg->tls_session_ticket_keys.first->str[0] != 0) {
if(!listen_sslctx_setup_ticket_keys(
cfg->tls_session_ticket_keys.first)) {
fatal_exit("could not set session ticket SSL_CTX");
}
}
(void)setup_listen_sslctx(&daemon->listen_dot_sslctx, 1, 0, cfg);
#ifdef HAVE_NGHTTP2_NGHTTP2_H
if(cfg_has_https(cfg)) {
(void)setup_listen_sslctx(&daemon->listen_doh_sslctx, 0, 1, cfg);
}
#endif
#ifdef HAVE_NGTCP2
if(cfg_has_quic(cfg)) {
if(!(daemon->listen_quic_sslctx = quic_sslctx_create(
cfg->ssl_service_key, cfg->ssl_service_pem, NULL))) {
fatal_exit("could not set up quic SSL_CTX");
}
}
#endif /* HAVE_NGTCP2 */
}
if(!(daemon->connect_dot_sslctx = connect_sslctx_create(NULL, NULL,
cfg->tls_cert_bundle, cfg->tls_win_cert)))
fatal_exit("could not set up connect SSL_CTX");
#else /* HAVE_SSL */
(void)daemon;(void)cfg;
#endif /* HAVE_SSL */
}
/** daemonize, drop user privileges and chroot if needed */
static void
perform_setup(struct daemon* daemon, struct config_file* cfg, int debug_mode,
@@ -475,7 +529,11 @@ perform_setup(struct daemon* daemon, struct config_file* cfg, int debug_mode,
#endif
#ifdef HAVE_GETPWNAM
struct passwd *pwd = NULL;
#endif
if(!daemon_privileged(daemon))
fatal_exit("could not do privileged setup");
#ifdef HAVE_GETPWNAM
if(cfg->username && cfg->username[0]) {
if((pwd = getpwnam(cfg->username)) == NULL)
fatal_exit("user '%s' does not exist.", cfg->username);
@@ -487,36 +545,7 @@ perform_setup(struct daemon* daemon, struct config_file* cfg, int debug_mode,
#endif
/* read ssl keys while superuser and outside chroot */
#ifdef HAVE_SSL
if(!(daemon->rc = daemon_remote_create(cfg)))
fatal_exit("could not set up remote-control");
if(cfg->ssl_service_key && cfg->ssl_service_key[0]) {
if(!(daemon->listen_sslctx = listen_sslctx_create(
cfg->ssl_service_key, cfg->ssl_service_pem, NULL)))
fatal_exit("could not set up listen SSL_CTX");
if(cfg->tls_ciphers && cfg->tls_ciphers[0]) {
if (!SSL_CTX_set_cipher_list(daemon->listen_sslctx, cfg->tls_ciphers)) {
fatal_exit("failed to set tls-cipher %s", cfg->tls_ciphers);
}
}
#ifdef HAVE_SSL_CTX_SET_CIPHERSUITES
if(cfg->tls_ciphersuites && cfg->tls_ciphersuites[0]) {
if (!SSL_CTX_set_ciphersuites(daemon->listen_sslctx, cfg->tls_ciphersuites)) {
fatal_exit("failed to set tls-ciphersuites %s", cfg->tls_ciphersuites);
}
}
#endif
if(cfg->tls_session_ticket_keys.first &&
cfg->tls_session_ticket_keys.first->str[0] != 0) {
if(!listen_sslctx_setup_ticket_keys(daemon->listen_sslctx, cfg->tls_session_ticket_keys.first)) {
fatal_exit("could not set session ticket SSL_CTX");
}
}
}
if(!(daemon->connect_sslctx = connect_sslctx_create(NULL, NULL,
cfg->tls_cert_bundle, cfg->tls_win_cert)))
fatal_exit("could not set up connect SSL_CTX");
#endif
(void)setup_sslctxs(daemon, cfg);
/* init syslog (as root) if needed, before daemonize, otherwise
* a fork error could not be printed since daemonize closed stderr.*/
@@ -545,7 +574,15 @@ perform_setup(struct daemon* daemon, struct config_file* cfg, int debug_mode,
cfg, 1);
if(!daemon->pidfile)
fatal_exit("pidfile alloc: out of memory");
checkoldpid(daemon->pidfile, pidinchroot);
/* Check old pid if there is no username configured.
* With a username, the assumption is that the privilege
* drop makes a pidfile not removed when the server stopped
* last time. The server does not chown the pidfile for it,
* because that creates privilege escape problems, with the
* pidfile writable by unprivileged users, but used by
* privileged users. */
if(!(cfg->username && cfg->username[0]))
checkoldpid(daemon->pidfile, pidinchroot);
}
#endif
@@ -557,18 +594,7 @@ perform_setup(struct daemon* daemon, struct config_file* cfg, int debug_mode,
/* write new pidfile (while still root, so can be outside chroot) */
#ifdef HAVE_KILL
if(cfg->pidfile && cfg->pidfile[0] && need_pidfile) {
if(writepid(daemon->pidfile, getpid())) {
if(cfg->username && cfg->username[0] && cfg_uid != (uid_t)-1 &&
pidinchroot) {
# ifdef HAVE_CHOWN
if(chown(daemon->pidfile, cfg_uid, cfg_gid) == -1) {
verbose(VERB_QUERY, "cannot chown %u.%u %s: %s",
(unsigned)cfg_uid, (unsigned)cfg_gid,
daemon->pidfile, strerror(errno));
}
# endif /* HAVE_CHOWN */
}
}
writepid(daemon->pidfile, getpid());
}
#else
(void)daemon;
@@ -682,6 +708,9 @@ perform_setup(struct daemon* daemon, struct config_file* cfg, int debug_mode,
* it would succeed on SIGHUP as well */
if(!cfg->use_syslog)
log_init(cfg->logfile, cfg->use_syslog, cfg->chrootdir);
daemon->cfgfile = strdup(*cfgfile);
if(!daemon->cfgfile)
fatal_exit("out of memory in daemon cfgfile strdup");
}
/**
@@ -716,6 +745,7 @@ run_daemon(const char* cfgfile, int cmdline_verbose, int debug_mode, int need_pi
"the commandline to see more errors, "
"or unbound-checkconf", cfgfile);
log_warn("Continuing with default config settings");
config_auto_slab_values(cfg);
}
apply_settings(daemon, cfg, cmdline_verbose, debug_mode);
if(!done_setup)
@@ -746,7 +776,11 @@ run_daemon(const char* cfgfile, int cmdline_verbose, int debug_mode, int need_pi
if(daemon->pidfile) {
int fd;
/* truncate pidfile */
fd = open(daemon->pidfile, O_WRONLY | O_TRUNC, 0644);
fd = open(daemon->pidfile, O_WRONLY | O_TRUNC
#ifdef O_NOFOLLOW
| O_NOFOLLOW
#endif
, 0644);
if(fd != -1)
close(fd);
/* delete pidfile */
+773 -270
View File
File diff suppressed because it is too large Load Diff
+14 -2
View File
@@ -72,7 +72,13 @@ enum worker_commands {
/** obtain statistics without statsclear */
worker_cmd_stats_noreset,
/** execute remote control command */
worker_cmd_remote
worker_cmd_remote,
/** for fast-reload, perform stop */
worker_cmd_reload_stop,
/** for fast-reload, start again */
worker_cmd_reload_start,
/** for fast-reload, poll to make sure worker has released data */
worker_cmd_reload_poll
};
/**
@@ -86,6 +92,10 @@ struct worker {
struct daemon* daemon;
/** thread id */
ub_thread_type thr_id;
#ifdef HAVE_GETTID
/** thread tid, the LWP id. */
pid_t thread_tid;
#endif
/** pipe, for commands for this worker */
struct tube* cmd;
/** the event base this worker works with */
@@ -114,7 +124,7 @@ struct worker {
/** do we need to restart or quit (on signal) */
int need_to_exit;
/** allocation cache for this thread */
struct alloc_cache alloc;
struct alloc_cache *alloc;
/** per thread statistics */
struct ub_server_stats stats;
/** thread scratch regional */
@@ -127,6 +137,8 @@ struct worker {
/** dnstap environment, changed for this thread */
struct dt_env dtenv;
#endif
/** reuse existing cache on reload if other conditions allow it. */
int reuse_cache;
};
/**
+65 -46
View File
@@ -59,7 +59,7 @@
******************************************************************************/
/**
* This is the default DNS64 prefix that is used whent he dns64 module is listed
* This is the default DNS64 prefix that is used when the dns64 module is listed
* in module-config but when the dns64-prefix variable is not present.
*/
static const char DEFAULT_DNS64_PREFIX[] = "64:ff9b::/96";
@@ -573,28 +573,29 @@ static enum module_ext_state
handle_event_pass(struct module_qstate* qstate, int id)
{
struct dns64_qstate* iq = (struct dns64_qstate*)qstate->minfo[id];
if (iq && iq->state == DNS64_NEW_QUERY
&& qstate->qinfo.qtype == LDNS_RR_TYPE_PTR
&& qstate->qinfo.qname_len == 74
&& !strcmp((char*)&qstate->qinfo.qname[64], "\03ip6\04arpa"))
/* Handle PTR queries for IPv6 addresses. */
return handle_ipv6_ptr(qstate, id);
int synth_all_cfg = qstate->env->cfg->dns64_synthall;
int synth_qname = 0;
if (qstate->env->cfg->dns64_synthall &&
iq && iq->state == DNS64_NEW_QUERY
&& qstate->qinfo.qtype == LDNS_RR_TYPE_AAAA)
return generate_type_A_query(qstate, id);
if(iq && iq->state == DNS64_NEW_QUERY
&& qstate->qinfo.qtype == LDNS_RR_TYPE_PTR
&& qstate->qinfo.qname_len == 74
&& !strcmp((char*)&qstate->qinfo.qname[64], "\03ip6\04arpa")) {
/* Handle PTR queries for IPv6 addresses. */
return handle_ipv6_ptr(qstate, id);
}
if(dns64_always_synth_for_qname(qstate, id) &&
iq && iq->state == DNS64_NEW_QUERY
&& !(qstate->query_flags & BIT_CD)
&& qstate->qinfo.qtype == LDNS_RR_TYPE_AAAA) {
verbose(VERB_ALGO, "dns64: ignore-aaaa and synthesize anyway");
if(iq && iq->state == DNS64_NEW_QUERY &&
qstate->qinfo.qtype == LDNS_RR_TYPE_AAAA &&
(synth_all_cfg ||
(synth_qname=(dns64_always_synth_for_qname(qstate, id)
&& !(qstate->query_flags & BIT_CD))))) {
if(synth_qname)
verbose(VERB_ALGO, "dns64: ignore-aaaa and synthesize anyway");
return generate_type_A_query(qstate, id);
}
/* We are finished when our sub-query is finished. */
if (iq && iq->state == DNS64_SUBQUERY_FINISHED)
if(iq && iq->state == DNS64_SUBQUERY_FINISHED)
return module_finished;
/* Otherwise, pass request to next module. */
@@ -627,32 +628,38 @@ handle_event_moddone(struct module_qstate* qstate, int id)
* synthesize in (sec 5.1.2 of RFC6147).
* - A successful AAAA query with an answer.
*/
if((!iq || iq->state != DNS64_INTERNAL_QUERY)
&& qstate->qinfo.qtype == LDNS_RR_TYPE_AAAA
&& !(qstate->query_flags & BIT_CD)
&& !(qstate->return_msg &&
qstate->return_msg->rep &&
reply_find_answer_rrset(&qstate->qinfo,
qstate->return_msg->rep)))
/* not internal, type AAAA, not CD, and no answer RRset,
* So, this is a AAAA noerror/nodata answer */
return generate_type_A_query(qstate, id);
if((!iq || iq->state != DNS64_INTERNAL_QUERY)
&& qstate->qinfo.qtype == LDNS_RR_TYPE_AAAA
&& !(qstate->query_flags & BIT_CD)
&& dns64_always_synth_for_qname(qstate, id)) {
/* if it is not internal, AAAA, not CD and listed domain,
* generate from A record and ignore AAAA */
verbose(VERB_ALGO, "dns64: ignore-aaaa and synthesize anyway");
/* When an AAAA query completes check if we want to perform DNS64
* synthesis. We skip queries with DNSSEC enabled (!CD) and
* ones generated by us to retrieve the A/PTR record to use for
* synth. */
int could_synth =
qstate->qinfo.qtype == LDNS_RR_TYPE_AAAA &&
(!iq || iq->state != DNS64_INTERNAL_QUERY) &&
!(qstate->query_flags & BIT_CD);
int has_data = /* whether query returned non-empty rrset */
qstate->return_msg &&
qstate->return_msg->rep &&
reply_find_answer_rrset(&qstate->qinfo, qstate->return_msg->rep);
int synth_qname = 0;
if(could_synth &&
(!has_data ||
(synth_qname=dns64_always_synth_for_qname(qstate, id)))) {
if(synth_qname)
verbose(VERB_ALGO, "dns64: ignore-aaaa and synthesize anyway");
return generate_type_A_query(qstate, id);
}
/* Store the response in cache. */
if ( (!iq || !iq->started_no_cache_store) &&
qstate->return_msg && qstate->return_msg->rep &&
!dns_cache_store(qstate->env, &qstate->qinfo, qstate->return_msg->rep,
0, 0, 0, NULL, qstate->query_flags))
if( (!iq || !iq->started_no_cache_store) &&
qstate->return_msg &&
qstate->return_msg->rep &&
!dns_cache_store(
qstate->env, &qstate->qinfo, qstate->return_msg->rep,
0, qstate->prefetch_leeway, 0, NULL,
qstate->query_flags, qstate->qstarttime,
qstate->is_valrec))
log_err("out of memory");
/* do nothing */
@@ -695,6 +702,7 @@ dns64_operate(struct module_qstate* qstate, enum module_ev event, int id,
iq->state = DNS64_NEW_QUERY;
iq->started_no_cache_store = qstate->no_cache_store;
qstate->no_cache_store = 1;
ATTR_FALLTHROUGH
/* fallthrough */
case module_event_pass:
qstate->ext_state[id] = handle_event_pass(qstate, id);
@@ -840,8 +848,9 @@ dns64_adjust_a(int id, struct module_qstate* super, struct module_qstate* qstate
*/
cp = construct_reply_info_base(super->region, rep->flags, rep->qdcount,
rep->ttl, rep->prefetch_ttl, rep->serve_expired_ttl,
rep->serve_expired_norec_ttl,
rep->an_numrrsets, rep->ns_numrrsets, rep->ar_numrrsets,
rep->rrset_count, rep->security);
rep->rrset_count, rep->security, LDNS_EDE_NONE);
if(!cp)
return;
@@ -969,10 +978,19 @@ dns64_inform_super(struct module_qstate* qstate, int id,
}
super_dq->state = DNS64_SUBQUERY_FINISHED;
/* If there is no successful answer, we're done. */
if (qstate->return_rcode != LDNS_RCODE_NOERROR
|| !qstate->return_msg
|| !qstate->return_msg->rep) {
/* If there is no successful answer, we're done.
* Guarantee that we have at least a NOERROR reply further on. */
if(qstate->return_rcode != LDNS_RCODE_NOERROR
|| !qstate->return_msg
|| !qstate->return_msg->rep) {
return;
}
/* When no A record is found for synthesis fall back to AAAA again. */
if(qstate->qinfo.qtype == LDNS_RR_TYPE_A &&
!reply_find_answer_rrset(&qstate->qinfo,
qstate->return_msg->rep)) {
super_dq->state = DNS64_INTERNAL_QUERY;
return;
}
@@ -991,7 +1009,8 @@ dns64_inform_super(struct module_qstate* qstate, int id,
/* Store the generated response in cache. */
if ( (!super_dq || !super_dq->started_no_cache_store) &&
!dns_cache_store(super->env, &super->qinfo, super->return_msg->rep,
0, 0, 0, NULL, super->query_flags))
0, super->prefetch_leeway, 0, NULL, super->query_flags,
qstate->qstarttime, qstate->is_valrec))
log_err("out of memory");
}
@@ -1029,8 +1048,8 @@ dns64_get_mem(struct module_env* env, int id)
*/
static struct module_func_block dns64_block = {
"dns64",
&dns64_init, &dns64_deinit, &dns64_operate, &dns64_inform_super,
&dns64_clear, &dns64_get_mem
NULL, NULL, &dns64_init, &dns64_deinit, &dns64_operate,
&dns64_inform_super, &dns64_clear, &dns64_get_mem
};
/**
+95 -21
View File
@@ -86,6 +86,31 @@ dt_pack(const Dnstap__Dnstap *d, void **buf, size_t *sz)
return 1;
}
/** See if the message is sent due to dnstap sample rate */
static int
dt_sample_rate_limited(struct dt_env* env)
{
lock_basic_lock(&env->sample_lock);
/* Sampling is every [n] packets. Where n==1, every packet is sent */
if(env->sample_rate > 1) {
int submit = 0;
/* if sampling is engaged... */
if (env->sample_rate_count > env->sample_rate) {
/* once the count passes the limit */
/* submit the message */
submit = 1;
/* and reset the count */
env->sample_rate_count = 0;
}
/* increment count regardless */
env->sample_rate_count++;
lock_basic_unlock(&env->sample_lock);
return !submit;
}
lock_basic_unlock(&env->sample_lock);
return 0;
}
static void
dt_send(const struct dt_env *env, void *buf, size_t len_buf)
{
@@ -146,6 +171,7 @@ dt_create(struct config_file* cfg)
env = (struct dt_env *) calloc(1, sizeof(struct dt_env));
if (!env)
return NULL;
lock_basic_init(&env->sample_lock);
env->dtio = dt_io_thread_create();
if(!env->dtio) {
@@ -166,8 +192,11 @@ static void
dt_apply_identity(struct dt_env *env, struct config_file *cfg)
{
char buf[MAXHOSTNAMELEN+1];
if (!cfg->dnstap_send_identity)
if (!cfg->dnstap_send_identity) {
free(env->identity);
env->identity = NULL;
return;
}
free(env->identity);
if (cfg->dnstap_identity == NULL || cfg->dnstap_identity[0] == 0) {
if (gethostname(buf, MAXHOSTNAMELEN) == 0) {
@@ -189,8 +218,11 @@ dt_apply_identity(struct dt_env *env, struct config_file *cfg)
static void
dt_apply_version(struct dt_env *env, struct config_file *cfg)
{
if (!cfg->dnstap_send_version)
if (!cfg->dnstap_send_version) {
free(env->version);
env->version = NULL;
return;
}
free(env->version);
if (cfg->dnstap_version == NULL || cfg->dnstap_version[0] == 0)
env->version = strdup(PACKAGE_STRING);
@@ -204,13 +236,8 @@ dt_apply_version(struct dt_env *env, struct config_file *cfg)
}
void
dt_apply_cfg(struct dt_env *env, struct config_file *cfg)
dt_apply_logcfg(struct dt_env *env, struct config_file *cfg)
{
if (!cfg->dnstap)
return;
dt_apply_identity(env, cfg);
dt_apply_version(env, cfg);
if ((env->log_resolver_query_messages = (unsigned int)
cfg->dnstap_log_resolver_query_messages))
{
@@ -241,6 +268,23 @@ dt_apply_cfg(struct dt_env *env, struct config_file *cfg)
{
verbose(VERB_OPS, "dnstap Message/FORWARDER_RESPONSE enabled");
}
lock_basic_lock(&env->sample_lock);
if((env->sample_rate = (unsigned int)cfg->dnstap_sample_rate))
{
verbose(VERB_OPS, "dnstap SAMPLE_RATE enabled and set to \"%d\"", (int)env->sample_rate);
}
lock_basic_unlock(&env->sample_lock);
}
void
dt_apply_cfg(struct dt_env *env, struct config_file *cfg)
{
if (!cfg->dnstap)
return;
dt_apply_identity(env, cfg);
dt_apply_version(env, cfg);
dt_apply_logcfg(env, cfg);
}
int
@@ -273,6 +317,7 @@ dt_delete(struct dt_env *env)
if (!env)
return;
dt_io_thread_delete(env->dtio);
lock_basic_destroy(&env->sample_lock);
free(env->identity);
free(env->version);
free(env);
@@ -305,6 +350,7 @@ dt_msg_fill_net(struct dt_msg *dm,
struct sockaddr_storage *qs,
struct sockaddr_storage *rs,
enum comm_point_type cptype,
void *cpssl,
ProtobufCBinaryData *qaddr, protobuf_c_boolean *has_qaddr,
uint32_t *qport, protobuf_c_boolean *has_qport,
ProtobufCBinaryData *raddr, protobuf_c_boolean *has_raddr,
@@ -371,13 +417,26 @@ dt_msg_fill_net(struct dt_msg *dm,
*has_rport = 1;
}
log_assert(cptype == comm_udp || cptype == comm_tcp);
if (cptype == comm_udp) {
/* socket_protocol */
dm->m.socket_protocol = DNSTAP__SOCKET_PROTOCOL__UDP;
dm->m.has_socket_protocol = 1;
} else if (cptype == comm_tcp) {
if (cpssl == NULL) {
/* socket_protocol */
dm->m.socket_protocol = DNSTAP__SOCKET_PROTOCOL__TCP;
dm->m.has_socket_protocol = 1;
} else {
/* socket_protocol */
dm->m.socket_protocol = DNSTAP__SOCKET_PROTOCOL__DOT;
dm->m.has_socket_protocol = 1;
}
} else if (cptype == comm_http) {
/* socket_protocol */
dm->m.socket_protocol = DNSTAP__SOCKET_PROTOCOL__DOH;
dm->m.has_socket_protocol = 1;
} else {
/* other socket protocol */
dm->m.socket_protocol = DNSTAP__SOCKET_PROTOCOL__TCP;
dm->m.has_socket_protocol = 1;
}
@@ -388,12 +447,19 @@ dt_msg_send_client_query(struct dt_env *env,
struct sockaddr_storage *qsock,
struct sockaddr_storage *rsock,
enum comm_point_type cptype,
sldns_buffer *qmsg)
void *cpssl,
sldns_buffer *qmsg,
struct timeval* tstamp)
{
struct dt_msg dm;
struct timeval qtime;
gettimeofday(&qtime, NULL);
if(dt_sample_rate_limited(env))
return;
if(tstamp)
memcpy(&qtime, tstamp, sizeof(qtime));
else gettimeofday(&qtime, NULL);
/* type */
dt_msg_init(env, &dm, DNSTAP__MESSAGE__TYPE__CLIENT_QUERY);
@@ -407,8 +473,7 @@ dt_msg_send_client_query(struct dt_env *env,
dt_fill_buffer(qmsg, &dm.m.query_message, &dm.m.has_query_message);
/* socket_family, socket_protocol, query_address, query_port, response_address, response_port */
log_assert(cptype == comm_udp || cptype == comm_tcp);
dt_msg_fill_net(&dm, qsock, rsock, cptype,
dt_msg_fill_net(&dm, qsock, rsock, cptype, cpssl,
&dm.m.query_address, &dm.m.has_query_address,
&dm.m.query_port, &dm.m.has_query_port,
&dm.m.response_address, &dm.m.has_response_address,
@@ -424,11 +489,15 @@ dt_msg_send_client_response(struct dt_env *env,
struct sockaddr_storage *qsock,
struct sockaddr_storage *rsock,
enum comm_point_type cptype,
void *cpssl,
sldns_buffer *rmsg)
{
struct dt_msg dm;
struct timeval rtime;
if(dt_sample_rate_limited(env))
return;
gettimeofday(&rtime, NULL);
/* type */
@@ -443,8 +512,7 @@ dt_msg_send_client_response(struct dt_env *env,
dt_fill_buffer(rmsg, &dm.m.response_message, &dm.m.has_response_message);
/* socket_family, socket_protocol, query_address, query_port, response_address, response_port */
log_assert(cptype == comm_udp || cptype == comm_tcp);
dt_msg_fill_net(&dm, qsock, rsock, cptype,
dt_msg_fill_net(&dm, qsock, rsock, cptype, cpssl,
&dm.m.query_address, &dm.m.has_query_address,
&dm.m.query_port, &dm.m.has_query_port,
&dm.m.response_address, &dm.m.has_response_address,
@@ -459,6 +527,7 @@ dt_msg_send_outside_query(struct dt_env *env,
struct sockaddr_storage *rsock,
struct sockaddr_storage *qsock,
enum comm_point_type cptype,
void *cpssl,
uint8_t *zone, size_t zone_len,
sldns_buffer *qmsg)
{
@@ -466,11 +535,14 @@ dt_msg_send_outside_query(struct dt_env *env,
struct timeval qtime;
uint16_t qflags;
if(dt_sample_rate_limited(env))
return;
gettimeofday(&qtime, NULL);
qflags = sldns_buffer_read_u16_at(qmsg, 2);
/* type */
if (qflags & BIT_RD) {
if ((qflags & BIT_RD)) {
if (!env->log_forwarder_query_messages)
return;
dt_msg_init(env, &dm, DNSTAP__MESSAGE__TYPE__FORWARDER_QUERY);
@@ -494,8 +566,7 @@ dt_msg_send_outside_query(struct dt_env *env,
dt_fill_buffer(qmsg, &dm.m.query_message, &dm.m.has_query_message);
/* socket_family, socket_protocol, response_address, response_port, query_address, query_port */
log_assert(cptype == comm_udp || cptype == comm_tcp);
dt_msg_fill_net(&dm, rsock, qsock, cptype,
dt_msg_fill_net(&dm, rsock, qsock, cptype, cpssl,
&dm.m.response_address, &dm.m.has_response_address,
&dm.m.response_port, &dm.m.has_response_port,
&dm.m.query_address, &dm.m.has_query_address,
@@ -510,6 +581,7 @@ dt_msg_send_outside_response(struct dt_env *env,
struct sockaddr_storage *rsock,
struct sockaddr_storage *qsock,
enum comm_point_type cptype,
void *cpssl,
uint8_t *zone, size_t zone_len,
uint8_t *qbuf, size_t qbuf_len,
const struct timeval *qtime,
@@ -519,12 +591,15 @@ dt_msg_send_outside_response(struct dt_env *env,
struct dt_msg dm;
uint16_t qflags;
if(dt_sample_rate_limited(env))
return;
(void)qbuf_len; log_assert(qbuf_len >= sizeof(qflags));
memcpy(&qflags, qbuf, sizeof(qflags));
qflags = ntohs(qflags);
/* type */
if (qflags & BIT_RD) {
if ((qflags & BIT_RD)) {
if (!env->log_forwarder_response_messages)
return;
dt_msg_init(env, &dm, DNSTAP__MESSAGE__TYPE__FORWARDER_RESPONSE);
@@ -553,8 +628,7 @@ dt_msg_send_outside_response(struct dt_env *env,
dt_fill_buffer(rmsg, &dm.m.response_message, &dm.m.has_response_message);
/* socket_family, socket_protocol, response_address, response_port, query_address, query_port */
log_assert(cptype == comm_udp || cptype == comm_tcp);
dt_msg_fill_net(&dm, rsock, qsock, cptype,
dt_msg_fill_net(&dm, rsock, qsock, cptype, cpssl,
&dm.m.response_address, &dm.m.has_response_address,
&dm.m.response_port, &dm.m.has_response_port,
&dm.m.query_address, &dm.m.has_query_address,
+22 -1
View File
@@ -39,6 +39,7 @@
#ifdef USE_DNSTAP
#include "util/locks.h"
struct config_file;
struct sldns_buffer;
struct dt_msg_queue;
@@ -75,6 +76,13 @@ struct dt_env {
unsigned log_forwarder_query_messages : 1;
/** whether to log Message/FORWARDER_RESPONSE */
unsigned log_forwarder_response_messages : 1;
/** lock on sample count */
lock_basic_type sample_lock;
/** rate limit value from config, samples 1/N messages */
unsigned int sample_rate;
/** rate limit counter */
unsigned int sample_rate_count;
};
/**
@@ -98,6 +106,13 @@ dt_create(struct config_file* cfg);
void
dt_apply_cfg(struct dt_env *env, struct config_file *cfg);
/**
* Apply config settings for log enable for message types.
* @param env: dnstap environment object.
* @param cfg: new config settings.
*/
void dt_apply_logcfg(struct dt_env *env, struct config_file *cfg);
/**
* Initialize per-worker state in dnstap environment object.
* @param env: dnstap environment object to initialize, created with dt_create().
@@ -126,13 +141,16 @@ dt_delete(struct dt_env *env);
* @param rsock: local (service) address/port.
* @param cptype: comm_udp or comm_tcp.
* @param qmsg: query message.
* @param tstamp: timestamp or NULL if none provided.
*/
void
dt_msg_send_client_query(struct dt_env *env,
struct sockaddr_storage *qsock,
struct sockaddr_storage *rsock,
enum comm_point_type cptype,
struct sldns_buffer *qmsg);
void *cpssl,
struct sldns_buffer *qmsg,
struct timeval* tstamp);
/**
* Create and send a new dnstap "Message" event of type CLIENT_RESPONSE.
@@ -147,6 +165,7 @@ dt_msg_send_client_response(struct dt_env *env,
struct sockaddr_storage *qsock,
struct sockaddr_storage *rsock,
enum comm_point_type cptype,
void *cpssl,
struct sldns_buffer *rmsg);
/**
@@ -166,6 +185,7 @@ dt_msg_send_outside_query(struct dt_env *env,
struct sockaddr_storage *rsock,
struct sockaddr_storage *qsock,
enum comm_point_type cptype,
void *cpssl,
uint8_t *zone, size_t zone_len,
struct sldns_buffer *qmsg);
@@ -190,6 +210,7 @@ dt_msg_send_outside_response(struct dt_env *env,
struct sockaddr_storage *rsock,
struct sockaddr_storage *qsock,
enum comm_point_type cptype,
void *cpssl,
uint8_t *zone, size_t zone_len,
uint8_t *qbuf, size_t qbuf_len,
const struct timeval *qtime,
+97 -38
View File
@@ -5,45 +5,104 @@
# Check for required dnstap libraries and add dnstap configure args.
AC_DEFUN([dt_DNSTAP],
[
AC_ARG_ENABLE([dnstap],
AS_HELP_STRING([--enable-dnstap],
[Enable dnstap support (requires protobuf-c)]),
[opt_dnstap=$enableval], [opt_dnstap=no])
AC_ARG_ENABLE([dnstap],
AS_HELP_STRING([--enable-dnstap],
[Enable dnstap support (requires protobuf-c)]),
[opt_dnstap=$enableval],
[opt_dnstap=no])
AC_ARG_WITH([dnstap-socket-path],
AS_HELP_STRING([--with-dnstap-socket-path=pathname],
[set default dnstap socket path]),
[opt_dnstap_socket_path=$withval], [opt_dnstap_socket_path="$1"])
AC_ARG_WITH([dnstap-socket-path],
AS_HELP_STRING([--with-dnstap-socket-path=pathname],
[set default dnstap socket path]),
[opt_dnstap_socket_path=$withval],
[opt_dnstap_socket_path="$1"])
if test "x$opt_dnstap" != "xno"; then
AC_PATH_PROG([PROTOC_C], [protoc-c])
if test -z "$PROTOC_C"; then
AC_MSG_ERROR([The protoc-c program was not found. Please install protobuf-c!])
if test "x$opt_dnstap" != "xno"; then
AC_PATH_PROG([PROTOC], [protoc])
# 'protoc-c' is deprecated. We use 'protoc' instead. If it can not be
# found, try 'protoc-c'.
if test -z "$PROTOC"; then
AC_PATH_PROG([PROTOC_C], [protoc-c])
else
PROTOC_C="$PROTOC"
fi
if test -z "$PROTOC_C"; then
AC_MSG_ERROR([[The protoc or protoc-c program was not found. It is needed for dnstap, use --disable-dnstap, or install protobuf-c to provide protoc or protoc-c]])
fi
# Check for protoc-gen-c plugin
AC_PATH_PROG([PROTOC_GEN_C], [protoc-gen-c])
if test -z "$PROTOC_GEN_C"; then
AC_MSG_ERROR([[The protoc-gen-c plugin was not found. It is needed for dnstap, use --disable-dnstap, or install protobuf-c-compiler to provide protoc-gen-c]])
fi
# Test that protoc-gen-c actually works
AC_MSG_CHECKING([if protoc-gen-c plugin works])
cat > conftest.proto << EOF
syntax = "proto2";
message TestMessage {
optional string test_field = 1;
}
EOF
if $PROTOC_C --c_out=. conftest.proto >/dev/null 2>&1; then
AC_MSG_RESULT([yes])
rm -f conftest.proto conftest.pb-c.c conftest.pb-c.h
else
AC_MSG_RESULT([no])
rm -f conftest.proto conftest.pb-c.c conftest.pb-c.h
AC_MSG_ERROR([[The protoc-gen-c plugin is not working properly. Please ensure protobuf-c-compiler is properly installed]])
fi
AC_ARG_WITH([protobuf-c],
AS_HELP_STRING([--with-protobuf-c=path], [Path where protobuf-c is installed, for dnstap]),
[
# workaround for protobuf-c includes at old dir before protobuf-c-1.0.0
if test -f $withval/include/google/protobuf-c/protobuf-c.h; then
CFLAGS="$CFLAGS -I$withval/include/google"
else
CFLAGS="$CFLAGS -I$withval/include"
fi
LDFLAGS="$LDFLAGS -L$withval/lib"
],
[
if test -n "$PKG_CONFIG"; then
PKG_CHECK_MODULES([PROTOBUFC], [libprotobuf-c],
[
CFLAGS="$CFLAGS $PROTOBUFC_CFLAGS"
LIBS="$LIBS $PROTOBUFC_LIBS"
],
[
# pkg-config failed; try falling back to known values
# workaround for protobuf-c includes at old dir before protobuf-c-1.0.0
if test -f /usr/include/google/protobuf-c/protobuf-c.h; then
CFLAGS="$CFLAGS -I/usr/include/google"
else
if test -f /usr/local/include/google/protobuf-c/protobuf-c.h; then
CFLAGS="$CFLAGS -I/usr/local/include/google"
LDFLAGS="$LDFLAGS -L/usr/local/lib"
else
AC_MSG_ERROR([The protobuf-c package was not found with pkg-config. Please install protobuf-c!])
fi
fi
]
)
else
# workaround for protobuf-c includes at old dir before protobuf-c-1.0.0
if test -f /usr/include/google/protobuf-c/protobuf-c.h; then
CFLAGS="$CFLAGS -I/usr/include/google"
else
if test -f /usr/local/include/google/protobuf-c/protobuf-c.h; then
CFLAGS="$CFLAGS -I/usr/local/include/google"
LDFLAGS="$LDFLAGS -L/usr/local/lib"
fi
fi
fi
]
)
AC_SEARCH_LIBS([protobuf_c_message_pack], [protobuf-c], [],
AC_MSG_ERROR([The protobuf-c library was not found. Please install the development libraries for protobuf-c!]))
$2
else
$3
fi
AC_ARG_WITH([protobuf-c], AS_HELP_STRING([--with-protobuf-c=path],
[Path where protobuf-c is installed, for dnstap]), [
# workaround for protobuf-c includes at old dir before protobuf-c-1.0.0
if test -f $withval/include/google/protobuf-c/protobuf-c.h; then
CFLAGS="$CFLAGS -I$withval/include/google"
else
CFLAGS="$CFLAGS -I$withval/include"
fi
LDFLAGS="$LDFLAGS -L$withval/lib"
], [
# workaround for protobuf-c includes at old dir before protobuf-c-1.0.0
if test -f /usr/include/google/protobuf-c/protobuf-c.h; then
CFLAGS="$CFLAGS -I/usr/include/google"
else
if test -f /usr/local/include/google/protobuf-c/protobuf-c.h; then
CFLAGS="$CFLAGS -I/usr/local/include/google"
LDFLAGS="$LDFLAGS -L/usr/local/lib"
fi
fi
])
AC_SEARCH_LIBS([protobuf_c_message_pack], [protobuf-c], [],
AC_MSG_ERROR([The protobuf-c library was not found. Please install protobuf-c!]))
$2
else
$3
fi
])
+75 -7
View File
@@ -13,8 +13,8 @@
// with this file. If not, see:
//
// <http://creativecommons.org/publicdomain/zero/1.0/>.
syntax = "proto2";
syntax = "proto2";
package dnstap;
// "Dnstap": this is the top-level dnstap type, which is a "union" type that
@@ -56,11 +56,60 @@ enum SocketFamily {
INET6 = 2; // IPv6 (RFC 2460)
}
// SocketProtocol: the transport protocol of a socket. This specifies how to
// interpret "transport port" fields.
// SocketProtocol: the protocol used to transport a DNS message.
enum SocketProtocol {
UDP = 1; // User Datagram Protocol (RFC 768)
TCP = 2; // Transmission Control Protocol (RFC 793)
UDP = 1; // DNS over UDP transport (RFC 1035 section 4.2.1)
TCP = 2; // DNS over TCP transport (RFC 1035 section 4.2.2)
DOT = 3; // DNS over TLS (RFC 7858)
DOH = 4; // DNS over HTTPS (RFC 8484)
DNSCryptUDP = 5; // DNSCrypt over UDP (https://dnscrypt.info/protocol)
DNSCryptTCP = 6; // DNSCrypt over TCP (https://dnscrypt.info/protocol)
DOQ = 7; // DNS over QUIC (RFC 9250)
}
// Policy: information about any name server operator policy
// applied to the processing of a DNS message.
message Policy {
// Match: what aspect of the message or message exchange
// triggered the application of the Policy.
enum Match {
QNAME = 1; // Name in question section of query
CLIENT_IP = 2; // Client IP address
RESPONSE_IP = 3; // Address in A/AAAA RRSet
NS_NAME = 4; // Authoritative name server, by name
NS_IP = 5; // Authoritative name server, by IP address
}
// The Action taken to implement the Policy.
enum Action {
NXDOMAIN = 1; // Respond with NXDOMAIN
NODATA = 2; // Respond with empty answer section
PASS = 3; // Do not alter the response (passthrough)
DROP = 4; // Do not respond.
TRUNCATE = 5; // Truncate UDP response, forcing TCP retry
LOCAL_DATA = 6; // Respond with local data from policy
}
// type: the type of policy applied, e.g. "RPZ" for a
// policy from a Response Policy Zone.
optional string type = 1;
// rule: the rule matched by the message.
//
// In a RPZ context, this is the owner name of the rule in
// the Response Policy Zone in wire format.
optional bytes rule = 2;
// action: the policy action taken in response to the
// rule match.
optional Action action = 3;
// match: the feature of the message exchange which matched the rule.
optional Match match = 4;
// The matched value. Format depends on the matched feature .
optional bytes value = 5;
}
// Message: a wire-format (RFC 1035 section 4) DNS message and associated
@@ -158,6 +207,24 @@ message Message {
// STUB_RESPONSE is a DNS response message sent from a DNS server to a
// stub resolver, from the perspective of the stub resolver.
STUB_RESPONSE = 10;
// TOOL_QUERY is a DNS query message sent from a DNS software tool to a
// DNS server, from the perspective of the tool.
TOOL_QUERY = 11;
// TOOL_RESPONSE is a DNS response message received by a DNS software
// tool from a DNS server, from the perspective of the tool.
TOOL_RESPONSE = 12;
// UPDATE_QUERY is a Dynamic DNS Update request (RFC 2136) received
// by an authoritative name server, from the perspective of the
// authoritative name server.
UPDATE_QUERY = 13;
// UPDATE_RESPONSE is a Dynamic DNS Update response (RFC 2136) sent
// from an authoritative name server, from the perspective of the
// authoritative name server.
UPDATE_RESPONSE = 14;
}
// One of the Type values described above.
@@ -215,6 +282,9 @@ message Message {
// The responder's original wire-format DNS response message, verbatim.
optional bytes response_message = 14;
// Operator policy applied to the processing of this message, if any.
optional Policy policy = 15;
}
// All fields except for 'type' in the Message schema are optional.
@@ -236,7 +306,6 @@ message Message {
// RESOLVER_QUERY:
// socket_family, socket_protocol
// query_name, query_type, query_class
// query_message
// query_time_sec, query_time_nsec
// query_zone
@@ -244,7 +313,6 @@ message Message {
// RESOLVER_RESPONSE:
// socket_family, socket_protocol
// query_name, query_type, query_class
// query_time_sec, query_time_nsec
// query_zone
// response_address, response_port
+52 -15
View File
@@ -176,10 +176,7 @@ void
mq_wakeup_cb(void* arg)
{
struct dt_msg_queue* mq = (struct dt_msg_queue*)arg;
/* even if the dtio is already active, because perhaps much
* traffic suddenly, we leave the timer running to save on
* managing it, the once a second timer is less work then
* starting and stopping the timer frequently */
lock_basic_lock(&mq->dtio->wakeup_timer_lock);
mq->dtio->wakeup_timer_enabled = 0;
lock_basic_unlock(&mq->dtio->wakeup_timer_lock);
@@ -210,6 +207,8 @@ dt_msg_queue_start_timer(struct dt_msg_queue* mq, int wakeupnow)
lock_basic_lock(&mq->dtio->wakeup_timer_lock);
if(mq->dtio->wakeup_timer_enabled) {
if(wakeupnow) {
tv.tv_sec = 0;
tv.tv_usec = 0;
comm_timer_set(mq->wakeup_timer, &tv);
}
lock_basic_unlock(&mq->dtio->wakeup_timer_lock);
@@ -221,8 +220,14 @@ dt_msg_queue_start_timer(struct dt_msg_queue* mq, int wakeupnow)
if(!wakeupnow) {
tv.tv_sec = 1;
tv.tv_usec = 0;
/* If it is already set, keep it running. */
if(!comm_timer_is_set(mq->wakeup_timer))
comm_timer_set(mq->wakeup_timer, &tv);
} else {
tv.tv_sec = 0;
tv.tv_usec = 0;
comm_timer_set(mq->wakeup_timer, &tv);
}
comm_timer_set(mq->wakeup_timer, &tv);
lock_basic_unlock(&mq->dtio->wakeup_timer_lock);
}
@@ -260,8 +265,9 @@ dt_msg_queue_submit(struct dt_msg_queue* mq, void* buf, size_t len)
/* acquire lock */
lock_basic_lock(&mq->lock);
/* if list was empty, start timer for (eventual) wakeup */
if(mq->first == NULL)
/* if list was empty, start timer for (eventual) wakeup,
* or if dtio is not writing now an eventual wakeup is needed. */
if(mq->first == NULL || !mq->dtio->event_added_is_write)
wakeupstarttimer = 1;
/* if list contains more than wakeupnum elements, wakeup now,
* or if list is (going to be) almost full */
@@ -788,7 +794,7 @@ static int dtio_write_ssl(struct dt_io_thread* dtio, uint8_t* buf,
}
return -1;
}
log_crypto_err("dnstap io, could not SSL_write");
log_crypto_err_io("dnstap io, could not SSL_write", want);
return -1;
}
return r;
@@ -954,7 +960,7 @@ static int dtio_write_more(struct dt_io_thread* dtio)
* -1: continue, >0: number of bytes read into buffer */
static ssize_t receive_bytes(struct dt_io_thread* dtio, void* buf, size_t len) {
ssize_t r;
r = recv(dtio->fd, (void*)buf, len, 0);
r = recv(dtio->fd, (void*)buf, len, MSG_DONTWAIT);
if(r == -1) {
char* to = dtio->socket_path;
if(!to) to = dtio->ip_str;
@@ -1029,7 +1035,7 @@ static int ssl_read_bytes(struct dt_io_thread* dtio, void* buf, size_t len)
"other side");
return 0;
}
log_crypto_err("could not SSL_read");
log_crypto_err_io("could not SSL_read", want);
verbose(VERB_DETAIL, "dnstap io: output closed by the "
"other side");
return 0;
@@ -1259,6 +1265,13 @@ static void dtio_sleep(struct dt_io_thread* dtio)
/* unregister the event polling for write, because there is
* nothing to be written */
(void)dtio_add_output_event_read(dtio);
/* Set wakeuptimer enabled off; so that the next worker thread that
* wants to log starts a timer if needed, since the writer thread
* has gone to sleep. */
lock_basic_lock(&dtio->wakeup_timer_lock);
dtio->wakeup_timer_enabled = 0;
lock_basic_unlock(&dtio->wakeup_timer_lock);
}
#ifdef HAVE_SSL
@@ -1322,7 +1335,11 @@ static int dtio_ssl_check_peer(struct dt_io_thread* dtio)
if((SSL_get_verify_mode(dtio->ssl)&SSL_VERIFY_PEER)) {
/* verification */
if(SSL_get_verify_result(dtio->ssl) == X509_V_OK) {
#ifdef HAVE_SSL_GET1_PEER_CERTIFICATE
X509* x = SSL_get1_peer_certificate(dtio->ssl);
#else
X509* x = SSL_get_peer_certificate(dtio->ssl);
#endif
if(!x) {
verbose(VERB_ALGO, "dnstap io, %s, SSL "
"connection failed no certificate",
@@ -1347,7 +1364,11 @@ static int dtio_ssl_check_peer(struct dt_io_thread* dtio)
#endif
X509_free(x);
} else {
#ifdef HAVE_SSL_GET1_PEER_CERTIFICATE
X509* x = SSL_get1_peer_certificate(dtio->ssl);
#else
X509* x = SSL_get_peer_certificate(dtio->ssl);
#endif
if(x) {
log_cert(VERB_ALGO, "dnstap io, peer "
"certificate", x);
@@ -1431,8 +1452,8 @@ static int dtio_ssl_handshake(struct dt_io_thread* dtio,
} else {
unsigned long err = ERR_get_error();
if(!squelch_err_ssl_handshake(err)) {
log_crypto_err_code("dnstap io, ssl handshake failed",
err);
log_crypto_err_io_code("dnstap io, ssl handshake failed",
want, err);
verbose(VERB_OPS, "dnstap io, ssl handshake failed "
"from %s", dtio->ip_str);
}
@@ -1488,9 +1509,11 @@ void dtio_output_cb(int ATTR_UNUSED(fd), short bits, void* arg)
}
#endif
if((bits&UB_EV_READ || dtio->ssl_brief_write)) {
if((bits&UB_EV_READ) || dtio->ssl_brief_write) {
#ifdef HAVE_SSL
if(dtio->ssl_brief_write)
(void)dtio_disable_brief_write(dtio);
#endif
if(dtio->ready_frame_sent && !dtio->accept_frame_received) {
if(dtio_read_accept_frame(dtio) <= 0)
return;
@@ -1513,8 +1536,22 @@ void dtio_output_cb(int ATTR_UNUSED(fd), short bits, void* arg)
/* no messages on the first iteration,
* the queues are all empty */
dtio_sleep(dtio);
/* After putting to sleep, see if
* a message is in a message queue,
* if so, resume service. Stops a
* race condition where a thread could
* have one message but the dtio
* also just went to sleep. With the
* message queued between the
* dtio_find_msg and dtio_sleep
* calls. */
if(dtio_find_msg(dtio)) {
if(!dtio_add_output_event_write(dtio))
return;
}
}
return; /* nothing to do */
if(!dtio->cur_msg)
return; /* nothing to do */
}
}
@@ -1960,7 +1997,7 @@ static int dtio_open_output_tcp(struct dt_io_thread* dtio)
memset(&addr, 0, sizeof(addr));
addrlen = (socklen_t)sizeof(addr);
if(!extstrtoaddr(dtio->ip_str, &addr, &addrlen)) {
if(!extstrtoaddr(dtio->ip_str, &addr, &addrlen, UNBOUND_DNS_PORT)) {
log_err("could not parse IP '%s'", dtio->ip_str);
return 0;
}
+319 -58
View File
@@ -61,6 +61,7 @@
#include "services/listen_dnsport.h"
#include "sldns/sbuffer.h"
#include "sldns/wire2str.h"
#include "sldns/pkthdr.h"
#ifdef USE_DNSTAP
#include <protobuf-c/protobuf-c.h>
#include "dnstap/dnstap.pb-c.h"
@@ -74,17 +75,18 @@
static void usage(char* argv[])
{
printf("usage: %s [options]\n", argv[0]);
printf(" Listen to dnstap messages\n");
printf(" Listen to dnstap messages\n");
printf("stdout has dnstap log, stderr has verbose server log\n");
printf("-u <socketpath> listen to unix socket with this file name\n");
printf("-s <serverip[@port]> listen for TCP on the IP and port\n");
printf("-t <serverip[@port]> listen for TLS on IP and port\n");
printf("-x <server.key> server key file for TLS service\n");
printf("-y <server.pem> server cert file for TLS service\n");
printf("-z <verify.pem> cert file to verify client connections\n");
printf("-l long format for DNS printout\n");
printf("-v more verbose log output\n");
printf("-h this help text\n");
printf("-u <socketpath> listen to unix socket with this file name\n");
printf("-s <serverip[@port]> listen for TCP on the IP and port\n");
printf("-t <serverip[@port]> listen for TLS on IP and port\n");
printf("-x <server.key> server key file for TLS service\n");
printf("-y <server.pem> server cert file for TLS service\n");
printf("-z <verify.pem> cert file to verify client connections\n");
printf("-l long format for DNS printout\n");
printf("-v more verbose log output\n");
printf("-c internal unit test and exit\n");
printf("-h this help text\n");
exit(1);
}
@@ -101,6 +103,14 @@ struct main_tap_data {
struct tap_socket_list* acceptlist;
};
/* list of data */
struct tap_data_list {
/** next in list */
struct tap_data_list* next;
/** the data */
struct tap_data* d;
};
/** tap callback variables */
struct tap_data {
/** the fd */
@@ -127,6 +137,10 @@ struct tap_data {
uint8_t* frame;
/** length of this frame */
size_t len;
/** back pointer to the tap_data_list entry;
* used to NULL the forward pointer to this data
* when this data is freed. */
struct tap_data_list* data_list;
};
/** list of sockets */
@@ -155,8 +169,89 @@ struct tap_socket {
char* ip;
/** for a TLS socket, the tls context */
SSL_CTX* sslctx;
/** dumb way to deal with memory leaks:
* tap_data was only freed on errors and not during exit leading to
* false positives when testing for memory leaks. */
struct tap_data_list* data_list;
};
/** try to delete tail entries from the list if all of them have no data */
static void tap_data_list_try_to_free_tail(struct tap_data_list* list)
{
struct tap_data_list* current = list;
log_assert(!list->d);
if(!list->next) /* we are the last, we can't remove ourselves */
return;
list = list->next;
while(list) {
if(list->d) /* a tail entry still has data; return */
return;
list = list->next;
}
/* keep the next */
list = current->next;
/* the tail will be removed; but not ourselves */
current->next = NULL;
while(list) {
current = list;
list = list->next;
free(current);
}
}
/** delete the tap structure */
static void tap_data_free(struct tap_data* data, int free_tail)
{
if(!data)
return;
if(data->ev) {
ub_event_del(data->ev);
ub_event_free(data->ev);
}
#ifdef HAVE_SSL
SSL_free(data->ssl);
#endif
sock_close(data->fd);
free(data->id);
free(data->frame);
if(data->data_list) {
data->data_list->d = NULL;
if(free_tail)
tap_data_list_try_to_free_tail(data->data_list);
}
free(data);
}
/** insert tap_data in the tap_data_list */
static int tap_data_list_insert(struct tap_data_list** liststart,
struct tap_data* d)
{
struct tap_data_list* entry = (struct tap_data_list*)
malloc(sizeof(*entry));
if(!entry)
return 0;
entry->next = *liststart;
entry->d = d;
d->data_list = entry;
*liststart = entry;
return 1;
}
/** delete the tap_data_list and free any remaining tap_data */
static void tap_data_list_delete(struct tap_data_list* list)
{
struct tap_data_list* e = list, *next;
while(e) {
next = e->next;
if(e->d) {
tap_data_free(e->d, 0);
e->d = NULL;
}
free(e);
e = next;
}
}
/** del the tap event */
static void tap_socket_delev(struct tap_socket* s)
{
@@ -172,7 +267,7 @@ static void tap_socket_close(struct tap_socket* s)
{
if(!s) return;
if(s->fd == -1) return;
close(s->fd);
sock_close(s->fd);
s->fd = -1;
}
@@ -183,6 +278,7 @@ static void tap_socket_delete(struct tap_socket* s)
#ifdef HAVE_SSL
SSL_CTX_free(s->sslctx);
#endif
tap_data_list_delete(s->data_list);
ub_event_free(s->ev);
free(s->socketpath);
free(s->ip);
@@ -250,7 +346,8 @@ static struct tap_socket* tap_socket_new_tlsaccept(char* ip,
s->fd = -1;
s->ev_cb = ev_cb;
s->data = data;
s->sslctx = listen_sslctx_create(server_key, server_cert, verifypem);
s->sslctx = listen_sslctx_create(server_key, server_cert, verifypem,
NULL, NULL, 0, 0, 0);
if(!s->sslctx) {
log_err("could not create ssl context");
free(s->ip);
@@ -272,7 +369,7 @@ static int make_tcp_accept(char* ip)
memset(&addr, 0, sizeof(addr));
len = (socklen_t)sizeof(addr);
if(!extstrtoaddr(ip, &addr, &len)) {
if(!extstrtoaddr(ip, &addr, &len, UNBOUND_DNS_PORT)) {
log_err("could not parse IP '%s'", ip);
return -1;
}
@@ -448,6 +545,7 @@ static char* q_of_msg(ProtobufCBinaryData message)
char buf[300];
/* header, name, type, class minimum to get the query tuple */
if(message.len < 12 + 1 + 4 + 4) return NULL;
if(LDNS_QDCOUNT(message.data) < 1) return NULL;
if(sldns_wire2str_rrquestion_buf(message.data+12, message.len-12,
buf, sizeof(buf)) != 0) {
/* remove trailing newline, tabs to spaces */
@@ -502,7 +600,7 @@ static char* tv_to_str(protobuf_c_boolean has_time_sec, uint64_t time_sec,
time_t time_t_sec;
memset(&tv, 0, sizeof(tv));
if(has_time_sec) tv.tv_sec = time_sec;
if(has_time_nsec) tv.tv_usec = time_nsec;
if(has_time_nsec) tv.tv_usec = time_nsec/1000;
buf[0]=0;
time_t_sec = tv.tv_sec;
@@ -617,7 +715,7 @@ static void log_data_frame(uint8_t* pkt, size_t len)
static ssize_t receive_bytes(struct tap_data* data, int fd, void* buf,
size_t len)
{
ssize_t ret = recv(fd, buf, len, 0);
ssize_t ret = recv(fd, buf, len, MSG_DONTWAIT);
if(ret == 0) {
/* closed */
if(verbosity) log_info("dnstap client stream closed from %s",
@@ -706,7 +804,7 @@ static ssize_t ssl_read_bytes(struct tap_data* data, void* buf, size_t len)
(data->id?data->id:""));
return 0;
}
log_crypto_err("could not SSL_read");
log_crypto_err_io("could not SSL_read", want);
if(verbosity) log_info("dnstap client stream closed from %s",
(data->id?data->id:""));
return 0;
@@ -726,27 +824,12 @@ static ssize_t tap_receive(struct tap_data* data, void* buf, size_t len)
return receive_bytes(data, data->fd, buf, len);
}
/** delete the tap structure */
static void tap_data_free(struct tap_data* data)
{
ub_event_del(data->ev);
ub_event_free(data->ev);
#ifdef HAVE_SSL
SSL_free(data->ssl);
#endif
close(data->fd);
free(data->id);
free(data->frame);
free(data);
}
/** reply with ACCEPT control frame to bidirectional client,
* returns 0 on error */
static int reply_with_accept(struct tap_data* data)
{
#ifdef USE_DNSTAP
/* len includes the escape and framelength */
int r;
size_t len = 0;
void* acceptframe = fstrm_create_control_frame_accept(
DNSTAP_CONTENT_TYPE, &len);
@@ -757,15 +840,19 @@ static int reply_with_accept(struct tap_data* data)
fd_set_block(data->fd);
if(data->ssl) {
#ifdef HAVE_SSL
int r;
if((r=SSL_write(data->ssl, acceptframe, len)) <= 0) {
if(SSL_get_error(data->ssl, r) == SSL_ERROR_ZERO_RETURN)
int r2;
if((r2=SSL_get_error(data->ssl, r)) == SSL_ERROR_ZERO_RETURN)
log_err("SSL_write, peer closed connection");
else
log_err("could not SSL_write");
log_crypto_err_io("could not SSL_write", r2);
fd_set_nonblock(data->fd);
free(acceptframe);
return 0;
}
#endif
} else {
if(send(data->fd, acceptframe, len, 0) == -1) {
log_err("send failed: %s", sock_strerror(errno));
@@ -789,7 +876,7 @@ static int reply_with_accept(struct tap_data* data)
/** reply with FINISH control frame to bidirectional client,
* returns 0 on error */
static int reply_with_finish(int fd)
static int reply_with_finish(struct tap_data* data)
{
#ifdef USE_DNSTAP
size_t len = 0;
@@ -799,21 +886,37 @@ static int reply_with_finish(int fd)
return 0;
}
fd_set_block(fd);
if(send(fd, finishframe, len, 0) == -1) {
log_err("send failed: %s", sock_strerror(errno));
fd_set_nonblock(fd);
free(finishframe);
return 0;
fd_set_block(data->fd);
if(data->ssl) {
#ifdef HAVE_SSL
int r;
if((r=SSL_write(data->ssl, finishframe, len)) <= 0) {
int r2;
if((r2=SSL_get_error(data->ssl, r)) == SSL_ERROR_ZERO_RETURN)
log_err("SSL_write, peer closed connection");
else
log_crypto_err_io("could not SSL_write", r2);
fd_set_nonblock(data->fd);
free(finishframe);
return 0;
}
#endif
} else {
if(send(data->fd, finishframe, len, 0) == -1) {
log_err("send failed: %s", sock_strerror(errno));
fd_set_nonblock(data->fd);
free(finishframe);
return 0;
}
}
if(verbosity) log_info("sent control frame(finish)");
fd_set_nonblock(fd);
fd_set_nonblock(data->fd);
free(finishframe);
return 1;
#else
log_err("no dnstap compiled, no reply");
(void)fd;
(void)data;
return 0;
#endif
}
@@ -825,7 +928,11 @@ static int tap_check_peer(struct tap_data* data)
if((SSL_get_verify_mode(data->ssl)&SSL_VERIFY_PEER)) {
/* verification */
if(SSL_get_verify_result(data->ssl) == X509_V_OK) {
#ifdef HAVE_SSL_GET1_PEER_CERTIFICATE
X509* x = SSL_get1_peer_certificate(data->ssl);
#else
X509* x = SSL_get_peer_certificate(data->ssl);
#endif
if(!x) {
if(verbosity) log_info("SSL connection %s"
" failed no certificate", data->id);
@@ -847,7 +954,11 @@ static int tap_check_peer(struct tap_data* data)
#endif
X509_free(x);
} else {
#ifdef HAVE_SSL_GET1_PEER_CERTIFICATE
X509* x = SSL_get1_peer_certificate(data->ssl);
#else
X509* x = SSL_get_peer_certificate(data->ssl);
#endif
if(x) {
if(verbosity)
log_cert(VERB_ALGO, "peer certificate", x);
@@ -889,7 +1000,7 @@ static int tap_handshake(struct tap_data* data)
return 0;
} else if(r == 0) {
/* closed */
tap_data_free(data);
tap_data_free(data, 1);
return 0;
} else if(want == SSL_ERROR_SYSCALL) {
/* SYSCALL and errno==0 means closed uncleanly */
@@ -907,7 +1018,7 @@ static int tap_handshake(struct tap_data* data)
if(!silent)
log_err("SSL_handshake syscall: %s",
strerror(errno));
tap_data_free(data);
tap_data_free(data, 1);
return 0;
} else {
unsigned long err = ERR_get_error();
@@ -917,7 +1028,7 @@ static int tap_handshake(struct tap_data* data)
verbose(VERB_OPS, "ssl handshake failed "
"from %s", data->id);
}
tap_data_free(data);
tap_data_free(data, 1);
return 0;
}
}
@@ -925,7 +1036,7 @@ static int tap_handshake(struct tap_data* data)
data->ssl_handshake_done = 1;
if(!tap_check_peer(data)) {
/* closed */
tap_data_free(data);
tap_data_free(data, 1);
return 0;
}
return 1;
@@ -933,7 +1044,7 @@ static int tap_handshake(struct tap_data* data)
#endif /* HAVE_SSL */
/** callback for dnstap listener */
void dtio_tap_callback(int fd, short ATTR_UNUSED(bits), void* arg)
void dtio_tap_callback(int ATTR_UNUSED(fd), short ATTR_UNUSED(bits), void* arg)
{
struct tap_data* data = (struct tap_data*)arg;
if(verbosity>=3) log_info("tap callback");
@@ -951,7 +1062,7 @@ void dtio_tap_callback(int fd, short ATTR_UNUSED(bits), void* arg)
if(verbosity>=4) log_info("s recv %d", (int)ret);
if(ret == 0) {
/* closed or error */
tap_data_free(data);
tap_data_free(data, 1);
return;
} else if(ret == -1) {
/* continue later */
@@ -973,7 +1084,7 @@ void dtio_tap_callback(int fd, short ATTR_UNUSED(bits), void* arg)
data->frame = calloc(1, data->len);
if(!data->frame) {
log_err("out of memory");
tap_data_free(data);
tap_data_free(data, 1);
return;
}
}
@@ -986,7 +1097,7 @@ void dtio_tap_callback(int fd, short ATTR_UNUSED(bits), void* arg)
if(verbosity>=4) log_info("f recv %d", (int)r);
if(r == 0) {
/* closed or error */
tap_data_free(data);
tap_data_free(data, 1);
return;
} else if(r == -1) {
/* continue later */
@@ -1011,13 +1122,13 @@ void dtio_tap_callback(int fd, short ATTR_UNUSED(bits), void* arg)
data->is_bidirectional = 1;
if(verbosity) log_info("bidirectional stream");
if(!reply_with_accept(data)) {
tap_data_free(data);
tap_data_free(data, 1);
return;
}
} else if(data->len >= 4 && sldns_read_uint32(data->frame) ==
FSTRM_CONTROL_FRAME_STOP && data->is_bidirectional) {
if(!reply_with_finish(fd)) {
tap_data_free(data);
if(!reply_with_finish(data)) {
tap_data_free(data, 1);
return;
}
}
@@ -1029,7 +1140,6 @@ void dtio_tap_callback(int fd, short ATTR_UNUSED(bits), void* arg)
data->len = 0;
data->len_done = 0;
data->data_done = 0;
}
/** callback for main listening file descriptor */
@@ -1042,7 +1152,9 @@ void dtio_mainfdcallback(int fd, short ATTR_UNUSED(bits), void* arg)
char* id = NULL;
struct sockaddr_storage addr;
socklen_t addrlen = (socklen_t)sizeof(addr);
int s = accept(fd, (struct sockaddr*)&addr, &addrlen);
int s;
memset(&addr, 0, sizeof(addr));
s = accept(fd, (struct sockaddr*)&addr, &addrlen);
if(s == -1) {
#ifndef USE_WINSOCK
/* EINTR is signal interrupt. others are closed connection. */
@@ -1112,6 +1224,8 @@ void dtio_mainfdcallback(int fd, short ATTR_UNUSED(bits), void* arg)
&dtio_tap_callback, data);
if(!data->ev) fatal_exit("could not ub_event_new");
if(ub_event_add(data->ev, NULL) != 0) fatal_exit("could not ub_event_add");
if(!tap_data_list_insert(&tap_sock->data_list, data))
fatal_exit("could not tap_data_list_insert");
}
/** setup local accept sockets */
@@ -1226,6 +1340,114 @@ setup_and_run(struct config_strlist_head* local_list,
free(maindata);
}
/* internal unit tests */
static int internal_unittest()
{
/* unit test tap_data_list_try_to_free_tail() */
#define unit_tap_datas_max 5
struct tap_data* datas[unit_tap_datas_max];
struct tap_data_list* list;
struct tap_socket* socket = calloc(1, sizeof(*socket));
size_t i = 0;
log_assert(socket);
log_assert(unit_tap_datas_max>2); /* needed for the test */
for(i=0; i<unit_tap_datas_max; i++) {
datas[i] = calloc(1, sizeof(struct tap_data));
log_assert(datas[i]);
log_assert(tap_data_list_insert(&socket->data_list, datas[i]));
}
/* sanity base check */
list = socket->data_list;
for(i=0; list; i++) list = list->next;
log_assert(i==unit_tap_datas_max);
/* Free the last data, tail cannot be erased */
list = socket->data_list;
while(list->next) list = list->next;
free(list->d);
list->d = NULL;
tap_data_list_try_to_free_tail(list);
list = socket->data_list;
for(i=0; list; i++) list = list->next;
log_assert(i==unit_tap_datas_max);
/* Free the third to last data, tail cannot be erased */
list = socket->data_list;
for(i=0; i<unit_tap_datas_max-3; i++) list = list->next;
free(list->d);
list->d = NULL;
tap_data_list_try_to_free_tail(list);
list = socket->data_list;
for(i=0; list; i++) list = list->next;
log_assert(i==unit_tap_datas_max);
/* Free the second to last data, try to remove tail from the third
* again, tail (last 2) should be removed */
list = socket->data_list;
for(i=0; i<unit_tap_datas_max-2; i++) list = list->next;
free(list->d);
list->d = NULL;
list = socket->data_list;
while(list->d) list = list->next;
tap_data_list_try_to_free_tail(list);
list = socket->data_list;
for(i=0; list; i++) list = list->next;
log_assert(i==unit_tap_datas_max-2);
/* Free all the remaining data, try to remove tail from the start,
* only the start should remain */
list = socket->data_list;
while(list) {
free(list->d);
list->d = NULL;
list = list->next;
}
tap_data_list_try_to_free_tail(socket->data_list);
list = socket->data_list;
for(i=0; list; i++) list = list->next;
log_assert(i==1);
/* clean up */
tap_data_list_delete(socket->data_list);
free(socket);
/* Start again. Add two elements */
socket = calloc(1, sizeof(*socket));
log_assert(socket);
for(i=0; i<2; i++) {
datas[i] = calloc(1, sizeof(struct tap_data));
log_assert(datas[i]);
log_assert(tap_data_list_insert(&socket->data_list, datas[i]));
}
/* sanity base check */
list = socket->data_list;
for(i=0; list; i++) list = list->next;
log_assert(i==2);
/* Free the last data, tail cannot be erased */
list = socket->data_list;
while(list->next) list = list->next;
free(list->d);
list->d = NULL;
tap_data_list_try_to_free_tail(list);
list = socket->data_list;
for(i=0; list; i++) list = list->next;
log_assert(i==2);
/* clean up */
tap_data_list_delete(socket->data_list);
free(socket);
if(log_get_lock()) {
lock_basic_destroy((lock_basic_type*)log_get_lock());
}
checklock_stop();
#ifdef USE_WINSOCK
WSACleanup();
#endif
return 0;
}
/** getopt global, in case header files fail to declare it. */
extern int optind;
/** getopt global, in case header files fail to declare it. */
@@ -1276,7 +1498,7 @@ int main(int argc, char** argv)
#endif
/* command line options */
while( (c=getopt(argc, argv, "hls:t:u:vx:y:z:")) != -1) {
while( (c=getopt(argc, argv, "hcls:t:u:vx:y:z:")) != -1) {
switch(c) {
case 'u':
if(!cfg_strlist_append(&local_list,
@@ -1312,14 +1534,20 @@ int main(int argc, char** argv)
case 'v':
verbosity++;
break;
case 'c':
#ifndef UNBOUND_DEBUG
fatal_exit("-c option needs compilation with "
"--enable-debug");
#endif
return internal_unittest();
case 'h':
case '?':
default:
usage(argv);
}
}
argc -= optind;
argv += optind;
/* argc -= optind; not using further arguments */
/* argv += optind; not using further arguments */
if(usessl) {
#ifdef HAVE_SSL
@@ -1348,6 +1576,9 @@ int main(int argc, char** argv)
config_delstrlist(tcp_list.first);
config_delstrlist(tls_list.first);
if(log_get_lock()) {
lock_basic_destroy((lock_basic_type*)log_get_lock());
}
checklock_stop();
#ifdef USE_WINSOCK
WSACleanup();
@@ -1555,3 +1786,33 @@ void remote_get_opt_ssl(char* ATTR_UNUSED(str), void* ATTR_UNUSED(arg))
{
log_assert(0);
}
void fast_reload_service_cb(int ATTR_UNUSED(fd), short ATTR_UNUSED(ev),
void* ATTR_UNUSED(arg))
{
log_assert(0);
}
int fast_reload_client_callback(struct comm_point* ATTR_UNUSED(c),
void* ATTR_UNUSED(arg), int ATTR_UNUSED(error),
struct comm_reply* ATTR_UNUSED(repinfo))
{
log_assert(0);
return 0;
}
#ifdef HAVE_NGTCP2
void doq_client_event_cb(int ATTR_UNUSED(fd), short ATTR_UNUSED(ev),
void* ATTR_UNUSED(arg))
{
log_assert(0);
}
#endif
#ifdef HAVE_NGTCP2
void doq_client_timer_cb(int ATTR_UNUSED(fd), short ATTR_UNUSED(ev),
void* ATTR_UNUSED(arg))
{
log_assert(0);
}
#endif
+2195 -19
View File
File diff suppressed because it is too large Load Diff
+4 -5
View File
@@ -19,11 +19,10 @@ The DNSTAP code has BSD license in dnstap/dnstap.c.
* Make and install: ./configure; make; make install
* --with-libevent=/path/to/libevent
Can be set to either the system install or the build directory.
--with-libevent=no (default) gives a builtin alternative
implementation. libevent is useful when having many (thousands)
of outgoing ports. This improves randomization and spoof
resistance. For the default of 16 ports the builtin alternative
works well and is a little faster.
--with-libevent=no gives a builtin alternative implementation.
Libevent is enabled by default, it is useful when having many
(thousands) of outgoing ports. This improves randomization and spoof
resistance. It also allows a higher number of outgoing queries.
* --with-libexpat=/path/to/libexpat
Can be set to the install directory of libexpat.
* --without-pthreads
+20
View File
@@ -28,3 +28,23 @@ prefix. For example:
;; ANSWER SECTION:
jazz-v4.viagenie.ca. 86400 IN AAAA 64:ff9b::ce7b:1f02
NAT64 support was added by David Lamparter in 2022; license(s) of the
surrounding code apply. Note that NAT64 is closely related but functionally
orthogonal to DNS64; it allows Unbound to send outgoing queries to IPv4-only
servers over IPv6 through the configured NAT64 prefix. This allows running
an Unbound instance on an IPv6-only host without breaking every single domain
that only has IPv4 servers. Whether that Unbound instance also does DNS64 is
an independent choice.
To enable NAT64 in Unbound, add to unbound.conf's "server" section:
do-nat64: yes
The NAT64 prefix defaults to the DNS64 prefix, which in turn defaults to the
standard 64:FF9B::/96 prefix. You can reconfigure it with:
nat64-prefix: 64:FF9B::/96
To test NAT64 operation, pick a domain that only has IPv4 reachability for its
nameservers and try resolving any names in that domain.
+16
View File
@@ -0,0 +1,16 @@
After Unbound 1.23.0, the source of the man pages is in reStructuredText format.
This helps with the online documentation at https://unbound.docs.nlnetlabs.nl
and makes it easier to maintain and contribute to the documentation.
The templated man pages (*.in) are still part of the code repository as to not
alter current procedures that could be in place by users/packagers.
The templated man pages (*.in) are generated by Sphinx (used for the online
documentation).
The online documentation has its own repository at
https://github.com/NLnetLabs/unbound-manual.
In the README.md there (branch test-auto for now), there are further simple
instructions on how to generate the templated man pages there and update them
in this repository.
+258 -26
View File
@@ -17,7 +17,7 @@ server:
# whitespace is not necessary, but looks cleaner.
# verbosity number, 0 is least verbose. 1 is default.
verbosity: 1
# verbosity: 1
# print statistics to the log (for every thread) every N seconds.
# Set to "" or 0 to disable. Default is disabled.
@@ -35,9 +35,14 @@ server:
# statistics-cumulative: no
# enable extended statistics (query types, answer codes, status)
# printed from unbound-control. default off, because of speed.
# printed from unbound-control. Default off, because of speed.
# extended-statistics: no
# Inhibits selected extended statistics (qtype, qclass, qopcode, rcode,
# rpz-actions) from printing if their value is 0.
# Default on.
# statistics-inhibit-zero: yes
# number of threads to create. 1 disables threading.
# num-threads: 1
@@ -50,11 +55,16 @@ server:
# interface: 192.0.2.154
# interface: 192.0.2.154@5003
# interface: 2001:DB8::5
# interface: eth0@5003
# enable this feature to copy the source address of queries to reply.
# Socket options are not supported on all platforms. experimental.
# interface-automatic: no
# instead of the default port, open additional ports separated by
# spaces when interface-automatic is enabled, by listing them here.
# interface-automatic-ports: ""
# port to answer queries from
# port: 53
@@ -106,8 +116,8 @@ server:
# so-rcvbuf: 0
# buffer size for UDP port 53 outgoing (SO_SNDBUF socket option).
# 0 is system default. Use 4m to handle spikes on very busy servers.
# so-sndbuf: 0
# 0 is system default. Set larger to handle spikes on very busy servers.
# so-sndbuf: 4m
# use SO_REUSEPORT to distribute queries over threads.
# at extreme load it could be better to turn it off to distribute even.
@@ -133,8 +143,8 @@ server:
# edns-buffer-size: 1232
# Maximum UDP response size (not applied to TCP response).
# Suggested values are 512 to 4096. Default is 4096. 65536 disables it.
# max-udp-size: 4096
# Suggested values are 512 to 4096. Default is 1232. 65536 disables it.
# max-udp-size: 1232
# max memory to use for stream(tcp and tls) waiting result buffers.
# stream-wait-size: 4m
@@ -153,7 +163,7 @@ server:
# msg-cache-slabs: 4
# the number of queries that a thread gets to service.
# num-queries-per-thread: 1024
# num-queries-per-thread: 2048
# if very busy, 50% queries run to completion, 50% get timeout in msec
# jostle-timeout: 200
@@ -164,13 +174,57 @@ server:
# perform connect for UDP sockets to mitigate ICMP side channel.
# udp-connect: yes
# The number of retries when a non-positive response is received.
# The number of retries, per upstream nameserver in a delegation, when
# a throwaway response (also timeouts) is received.
# outbound-msg-retry: 5
# Hard limit on the number of outgoing queries Unbound will make while
# resolving a name, making sure large NS sets do not loop.
# It resets on query restarts (e.g., CNAME) and referrals.
# max-sent-count: 32
# Hard limit on the number of times Unbound is allowed to restart a
# query upon encountering a CNAME record.
# max-query-restarts: 11
# Limit on number of NS records in NS RRset for incoming packets.
# iter-scrub-ns: 20
# Limit on number of CNAME, DNAME records for incoming packets.
# iter-scrub-cname: 11
# Limit on upstream queries for an incoming query and its recursion.
# max-global-quota: 200
# Should the scrubber remove promiscuous NS from positive answers,
# protects against poison attempts.
# iter-scrub-promiscuous: yes
# msec for waiting for an unknown server to reply. Increase if you
# are behind a slow satellite link, to eg. 1128.
# unknown-server-time-limit: 376
# msec before recursion replies are dropped. The work item continues.
# discard-timeout: 1900
# Max number of replies waiting for recursion per IP address.
# wait-limit: 1000
# Max replies waiting for recursion for IP address with cookie.
# wait-limit-cookie: 10000
# Apart from the default, the wait limit can be set for a netblock.
# wait-limit-netblock: 192.0.2.0/24 50000
# Apart from the default, the wait limit with cookie can be adjusted.
# wait-limit-cookie-netblock: 192.0.2.0/24 50000
# Defaults for loopback, it has no wait limit.
# wait-limit-netblock: 127.0.0.0/8 -1
# wait-limit-netblock: ::1/128 -1
# wait-limit-cookie-netblock: 127.0.0.0/8 -1
# wait-limit-cookie-netblock: ::1/128 -1
# the amount of memory to use for the RRset cache.
# plain value in bytes or you can append k, m or G. default is "4Mb".
# rrset-cache-size: 4m
@@ -191,6 +245,11 @@ server:
# the time to live (TTL) value cap for negative responses in the cache
# cache-max-negative-ttl: 3600
# the time to live (TTL) value lower bound, in seconds. Default 0.
# For negative responses in the cache. If disabled, default,
# cache-min-ttl applies if configured.
# cache-min-negative-ttl: 0
# the time to live (TTL) value for cached roundtrip times, lameness and
# EDNS version information for hosts. In seconds.
# infra-host-ttl: 900
@@ -198,6 +257,9 @@ server:
# minimum wait time for responses, increase if uplink is long. In msec.
# infra-cache-min-rtt: 50
# maximum wait time for responses. In msec.
# infra-cache-max-rtt: 120000
# enable to make server probe down hosts more frequently.
# infra-keep-probing: no
@@ -209,7 +271,8 @@ server:
# the maximum number of hosts that are cached (roundtrip, EDNS, lame).
# infra-cache-numhosts: 10000
# define a number of tags here, use with local-zone, access-control.
# define a number of tags here, use with local-zone, access-control,
# interface-*.
# repeat the define-tag statement to add additional tags.
# define-tag: "tag1 tag2 tag3"
@@ -219,6 +282,18 @@ server:
# Enable IPv6, "yes" or "no".
# do-ip6: yes
# If running unbound on an IPv6-only host, domains that only have
# IPv4 servers would become unresolvable. If NAT64 is available in
# the network, unbound can use NAT64 to reach these servers with
# the following option. This is NOT needed for enabling DNS64 on a
# system that has IPv4 connectivity.
# Consider also enabling prefer-ip6 to prefer native IPv6 connections
# to nameservers.
# do-nat64: no
# NAT64 prefix. Defaults to using dns64-prefix value.
# nat64-prefix: 64:ff9b::0/96
# Enable UDP, "yes" or "no".
# do-udp: yes
@@ -247,9 +322,14 @@ server:
# Enable EDNS TCP keepalive option.
# edns-tcp-keepalive: no
# Timeout for EDNS TCP keepalive, in msec.
# Timeout for EDNS TCP keepalive, in msec. Overrides tcp-idle-timeout
# if edns-tcp-keepalive is set.
# edns-tcp-keepalive-timeout: 120000
# UDP queries that have waited in the socket buffer for a long time
# can be dropped. Default is 0, disabled. In seconds, such as 3.
# sock-queue-timeout: 0
# Use systemd socket activation for UDP, TCP, and control sockets.
# use-systemd: no
@@ -263,11 +343,10 @@ server:
# Choose deny (drop message), refuse (polite error reply),
# allow (recursive ok), allow_setrd (recursive ok, rd bit is forced on),
# allow_snoop (recursive and nonrecursive ok)
# allow_cookie (allow UDP with valid cookie or stateful transport)
# deny_non_local (drop queries unless can be answered from local-data)
# refuse_non_local (like deny_non_local but polite error reply).
# access-control: 0.0.0.0/0 refuse
# access-control: 127.0.0.0/8 allow
# access-control: ::0/0 refuse
# access-control: ::1 allow
# access-control: ::ffff:127.0.0.1 allow
@@ -276,7 +355,7 @@ server:
# are tagged with one of these tags.
# access-control-tag: 192.0.2.0/24 "tag2 tag3"
# set action for particular tag for given access control element
# set action for particular tag for given access control element.
# if you have multiple tag values, the tag used to lookup the action
# is the first tag match between access-control-tag and local-zone-tag
# where "first" comes from the order of the define-tag values.
@@ -288,6 +367,58 @@ server:
# Set view for access control element
# access-control-view: 192.0.2.0/24 viewname
# Similar to 'access-control:' but for interfaces.
# Control which listening interfaces are allowed to accept (recursive)
# queries for this server.
# The specified interfaces should be the same as the ones specified in
# 'interface:' followed by the action.
# The actions are the same as 'access-control:' above.
# By default all the interfaces configured are refused.
# Note: any 'access-control*:' setting overrides all 'interface-*:'
# settings for targeted clients.
# interface-action: 192.0.2.153 allow
# interface-action: 192.0.2.154 allow
# interface-action: 192.0.2.154@5003 allow
# interface-action: 2001:DB8::5 allow
# interface-action: eth0@5003 allow
# Similar to 'access-control-tag:' but for interfaces.
# Tag interfaces with a list of tags (in "" with spaces between).
# Interfaces using these tags use localzones that are tagged with one
# of these tags.
# The specified interfaces should be the same as the ones specified in
# 'interface:' followed by the list of tags.
# Note: any 'access-control*:' setting overrides all 'interface-*:'
# settings for targeted clients.
# interface-tag: eth0@5003 "tag2 tag3"
# Similar to 'access-control-tag-action:' but for interfaces.
# Set action for particular tag for a given interface element.
# If you have multiple tag values, the tag used to lookup the action
# is the first tag match between interface-tag and local-zone-tag
# where "first" comes from the order of the define-tag values.
# The specified interfaces should be the same as the ones specified in
# 'interface:' followed by the tag and action.
# Note: any 'access-control*:' setting overrides all 'interface-*:'
# settings for targeted clients.
# interface-tag-action: eth0@5003 tag3 refuse
# Similar to 'access-control-tag-data:' but for interfaces.
# Set redirect data for a particular tag for an interface element.
# The specified interfaces should be the same as the ones specified in
# 'interface:' followed by the tag and the redirect data.
# Note: any 'access-control*:' setting overrides all 'interface-*:'
# settings for targeted clients.
# interface-tag-data: eth0@5003 tag2 "A 127.0.0.1"
# Similar to 'access-control-view:' but for interfaces.
# Set view for an interface element.
# The specified interfaces should be the same as the ones specified in
# 'interface:' followed by the view name.
# Note: any 'access-control*:' setting overrides all 'interface-*:'
# settings for targeted clients.
# interface-view: eth0@5003 viewname
# if given, a chroot(2) is done to the given directory.
# i.e. you can chroot to the working directory, for example,
# for extra security, but make sure all files are in that directory.
@@ -340,6 +471,10 @@ server:
# print UTC timestamp in ascii to logfile, default is epoch in seconds.
# log-time-ascii: no
# log timestamp in ISO8601 format if also log-time-ascii is enabled.
# (y-m-dTh:m:s.msec[+-]tzhours:tzminutes)
# log-time-iso: no
# print one line with time, IP, name, type, class for every query.
# log-queries: no
@@ -351,6 +486,9 @@ server:
# filtering log-queries and log-replies from the log.
# log-tag-queryreply: no
# log with destination address, port and type for log-replies.
# log-destaddr: no
# log the local-zone actions, like local-zone type inform is enabled
# also for the other local zone types.
# log-local-actions: no
@@ -409,6 +547,9 @@ server:
# Harden against out of zone rrsets, to avoid spoofing attempts.
# harden-glue: yes
# Harden against unverified (outside-zone, including sibling zone) glue rrsets
# harden-unverified-glue: no
# Harden against receiving dnssec-stripped data. If you turn it
# off, failing to validate dnskey data for a trustanchor will
# trigger insecure mode for that zone (like without a trustanchor).
@@ -425,10 +566,15 @@ server:
# harden-referral-path: no
# Harden against algorithm downgrade when multiple algorithms are
# advertised in the DS record. If no, allows the weakest algorithm
# to validate the zone.
# advertised in the DS record. If no, allows any algorithm
# to validate the zone which is the standard behavior for validators.
# Check the manpage for detailed information.
# harden-algo-downgrade: no
# Harden against unknown records in the authority section and the
# additional section.
# harden-unknown-additional: no
# Sent minimum amount of information to upstream servers to enhance
# privacy. Only sent minimum required labels of the QNAME and set QTYPE
# to A when possible.
@@ -589,13 +735,19 @@ server:
# that set CD but cannot validate themselves.
# ignore-cd-flag: no
# Disable the DO flag in outgoing requests. It is helpful for upstream
# devices that cannot handle DNSSEC information. But do not enable it
# otherwise, because it would stop DNSSEC validation.
# disable-edns-do: no
# Serve expired responses from cache, with serve-expired-reply-ttl in
# the response, and then attempt to fetch the data afresh.
# the response. By default it first tries to refresh an expired answer.
# Can be configured with serve-expired-client-timeout.
# serve-expired: no
#
# Limit serving of expired responses to configured seconds after
# expiration. 0 disables the limit.
# serve-expired-ttl: 0
# serve-expired-ttl: 86400
#
# Set the TTL of expired records to the serve-expired-ttl value after a
# failed attempt to retrieve the record from upstream. This makes sure
@@ -608,10 +760,9 @@ server:
#
# Time in milliseconds before replying to the client with expired data.
# This essentially enables the serve-stale behavior as specified in
# RFC 8767 that first tries to resolve before
# immediately responding with expired data. 0 disables this behavior.
# A recommended value is 1800.
# serve-expired-client-timeout: 0
# RFC 8767 that first tries to resolve before immediately responding
# with expired data. 0 disables this behavior.
# serve-expired-client-timeout: 1800
# Return the original TTL as received from the upstream name server rather
# than the decrementing TTL as stored in the cache. Enabling this feature
@@ -670,6 +821,8 @@ server:
# local-zone: "127.in-addr.arpa." nodefault
# local-zone: "1.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.ip6.arpa." nodefault
# local-zone: "home.arpa." nodefault
# local-zone: "resolver.arpa." nodefault
# local-zone: "service.arpa." nodefault
# local-zone: "onion." nodefault
# local-zone: "test." nodefault
# local-zone: "invalid." nodefault
@@ -736,6 +889,8 @@ server:
# o always_transparent, always_refuse, always_nxdomain, always_nodata,
# always_deny resolve in that way but ignore local data for
# that name
# o block_a resolves all records normally but returns
# NODATA for A queries and ignores local data for that name
# o always_null returns 0.0.0.0 or ::0 for any name in the zone.
# o noview breaks out of that view towards global local-zones.
#
@@ -778,6 +933,7 @@ server:
# tls-service-pem: "path/to/publiccertfile.pem"
# tls-port: 853
# https-port: 443
# quic-port: 853
# cipher setting for TLSv1.2
# tls-ciphers: "DHE-RSA-AES256-GCM-SHA384:DHE-RSA-AES128-GCM-SHA256:ECDHE-RSA-AES256-GCM-SHA384:ECDHE-RSA-AES128-GCM-SHA256:DHE-RSA-AES256-SHA256:DHE-RSA-AES128-SHA256:ECDHE-RSA-AES256-SHA384:ECDHE-RSA-AES128-SHA256"
@@ -811,6 +967,8 @@ server:
# Add system certs to the cert bundle, from the Windows Cert Store
# tls-win-cert: no
# and on other systems, the default openssl certificates
# tls-system-cert: no
# Pad queries over TLS upstreams
# pad-queries: yes
@@ -840,6 +998,13 @@ server:
# Disable TLS for DNS-over-HTTP downstream service.
# http-notls-downstream: no
# Maximum number of bytes used for QUIC buffers.
# quic-size: 8m
# The interfaces that use these listed port numbers will support and
# expect PROXYv2. For UDP and TCP/TLS interfaces.
# proxy-protocol-port: portno for each of the port numbers.
# DNS64 prefix. Must be specified when DNS64 is use.
# Enable dns64 in module-config. Used to synthesize IPv6 from IPv4.
# dns64-prefix: 64:ff9b::0/96
@@ -876,6 +1041,13 @@ server:
# if 0(default) it is disabled, otherwise states qps allowed per ip address
# ip-ratelimit: 0
# global query ratelimit for all ip addresses with a valid DNS Cookie.
# feature is experimental.
# if 0(default) it is disabled, otherwise states qps allowed per ip address
# useful in combination with 'allow_cookie'.
# If used, suggested to be higher than ip-ratelimit, tenfold.
# ip-ratelimit-cookie: 0
# ip ratelimits are tracked in a cache, size in bytes of cache (or k,m).
# ip-ratelimit-size: 4m
# ip ratelimit cache slabs, reduces lock contention if equal to cpucount.
@@ -897,6 +1069,32 @@ server:
# the number of servers that will be used in the fast server selection.
# fast-server-num: 3
# reply to requests containing DNS Cookies as specified in RFC 7873 and RFC 9018.
# answer-cookie: no
# secret for DNS Cookie generation.
# useful for anycast deployments.
# example value "000102030405060708090a0b0c0d0e0f".
# cookie-secret: <128 bit random hex string>
# File with cookie secrets, the 'cookie-secret:' option is ignored
# and the file can be managed to have staging and active secrets
# with remote control commands. Disabled with "". Default is "".
# cookie-secret-file: "/usr/local/etc/unbound_cookiesecrets.txt"
# Enable to attach Extended DNS Error codes (RFC8914) to responses.
# ede: no
# Enable to attach an Extended DNS Error (RFC8914) Code 3 - Stale
# Answer as EDNS0 option to expired responses.
# Note that the ede option above needs to be enabled for this to work.
# ede-serve-expired: no
# Enable DNS Error Reporting (RFC9567).
# qname-minimisation is advised to be turned on as well to increase
# privacy on the outgoing reports.
# dns-error-reporting: no
# Specific options for ipsecmod. Unbound needs to be configured with
# --enable-ipsecmod for these to take effect.
#
@@ -1031,11 +1229,11 @@ remote-control:
# has a copy of the root for local usage. The second serves example.org
# authoritatively. zonefile: reads from file (and writes to it if you also
# download it), primary: fetches with AXFR and IXFR, or url to zonefile.
# With allow-notify: you can give additional (apart from primaries) sources of
# notifies.
# With allow-notify: you can give additional (apart from primaries and urls)
# sources of notifies.
# auth-zone:
# name: "."
# primary: 199.9.14.201 # b.root-servers.net
# primary: 170.247.170.2 # b.root-servers.net
# primary: 192.33.4.12 # c.root-servers.net
# primary: 199.7.91.13 # d.root-servers.net
# primary: 192.5.5.241 # f.root-servers.net
@@ -1043,7 +1241,7 @@ remote-control:
# primary: 193.0.14.129 # k.root-servers.net
# primary: 192.0.47.132 # xfr.cjr.dns.icann.org
# primary: 192.0.32.132 # xfr.lax.dns.icann.org
# primary: 2001:500:200::b # b.root-servers.net
# primary: 2801:1b8:10::b # b.root-servers.net
# primary: 2001:500:2::c # c.root-servers.net
# primary: 2001:500:2d::d # d.root-servers.net
# primary: 2001:500:2f::f # f.root-servers.net
@@ -1111,6 +1309,11 @@ remote-control:
# backend: "testframe"
# # secret seed string to calculate hashed keys
# secret-seed: "default"
# # if the backend should be read from, but not written to.
# cachedb-no-store: no
# # if the cachedb should be checked before a serve-expired response is
# # given, when serve-expired is enabled.
# cachedb-check-when-serve-expired: yes
#
# # For "redis" backend:
# # (to enable, use --with-libhiredis to configure before compiling)
@@ -1118,10 +1321,36 @@ remote-control:
# redis-server-host: 127.0.0.1
# # redis server's TCP port
# redis-server-port: 6379
# # if the server uses a unix socket, set its path, or "" when not used.
# redis-server-path: "/var/lib/redis/redis-server.sock"
# # if the server uses an AUTH password, specify here, or "" when not used.
# redis-server-password: ""
# # timeout (in ms) for communication with the redis server
# redis-timeout: 100
# # timeout (in ms) for commands, if 0, uses redis-timeout.
# redis-command-timeout: 0
# # timeout (in ms) for connection set up, if 0, uses redis-timeout.
# redis-connect-timeout: 0
# # set timeout on redis records based on DNS response TTL
# redis-expire-records: no
# # redis logical database to use, 0 is the default database.
# redis-logical-db: 0
# # redis replica server's IP address or host name
# redis-replica-server-host: 127.0.0.1
# # redis replica server's TCP port
# redis-replica-server-port: 6379
# # if the replica server uses a unix socket, set its path, or "" when not used.
# redis-replica-server-path: "/var/lib/redis/redis-server.sock"
# # if the replica server uses an AUTH password, specify here, or "" when not used.
# redis-replica-server-password: ""
# # timeout (in ms) for communication with the redis replica server
# redis-replica-timeout: 100
# # timeout (in ms) for redis replica commands, if 0, uses redis-replica-timeout.
# redis-replica-command-timeout: 0
# # timeout (in ms) for redis replica connection set up, if 0, uses redis-replica-timeout.
# redis-replica-connect-timeout: 0
# # redis logical database to use for the replica server, 0 is the default database.
# redis-replica-logical-db: 0
# IPSet
# Add specify domain into set via ipset.
@@ -1163,6 +1392,8 @@ remote-control:
# dnstap-identity: ""
# # if "" it uses the package version.
# dnstap-version: ""
# # log only 1/N messages, if 0 it is disabled. default 0.
# dnstap-sample-rate: 0
# dnstap-log-resolver-query-messages: no
# dnstap-log-resolver-response-messages: no
# dnstap-log-client-query-messages: no
@@ -1171,7 +1402,8 @@ remote-control:
# dnstap-log-forwarder-response-messages: no
# Response Policy Zones
# RPZ policies. Applied in order of configuration. QNAME, Response IP
# RPZ policies. Applied in order of configuration. Any match from an earlier
# RPZ zone will terminate the RPZ lookup. QNAME, Response IP
# Address, nsdname, nsip and clientip triggers are supported. Supported
# actions are: NXDOMAIN, NODATA, PASSTHRU, DROP, Local Data, tcp-only
# and drop. Policies can be loaded from a file, or using zone
+330 -343
View File
@@ -1,335 +1,306 @@
.TH "libunbound" "3" "@date@" "NLnet Labs" "unbound @version@"
.\"
.\" libunbound.3 -- unbound library functions manual
.\"
.\" Copyright (c) 2007, NLnet Labs. All rights reserved.
.\"
.\" See LICENSE for the license.
.\"
.\"
.SH "NAME"
.B libunbound,
.B unbound.h,
.B ub_ctx,
.B ub_result,
.B ub_callback_type,
.B ub_ctx_create,
.B ub_ctx_delete,
.B ub_ctx_set_option,
.B ub_ctx_get_option,
.B ub_ctx_config,
.B ub_ctx_set_fwd,
.B ub_ctx_set_stub,
.B ub_ctx_set_tls,
.B ub_ctx_resolvconf,
.B ub_ctx_hosts,
.B ub_ctx_add_ta,
.B ub_ctx_add_ta_autr,
.B ub_ctx_add_ta_file,
.B ub_ctx_trustedkeys,
.B ub_ctx_debugout,
.B ub_ctx_debuglevel,
.B ub_ctx_async,
.B ub_poll,
.B ub_wait,
.B ub_fd,
.B ub_process,
.B ub_resolve,
.B ub_resolve_async,
.B ub_cancel,
.B ub_resolve_free,
.B ub_strerror,
.B ub_ctx_print_local_zones,
.B ub_ctx_zone_add,
.B ub_ctx_zone_remove,
.B ub_ctx_data_add,
.B ub_ctx_data_remove
\- Unbound DNS validating resolver @version@ functions.
.SH "SYNOPSIS"
.B #include <unbound.h>
.LP
\fIstruct ub_ctx *\fR
\fBub_ctx_create\fR(\fIvoid\fR);
.LP
\fIvoid\fR
\fBub_ctx_delete\fR(\fIstruct ub_ctx*\fR ctx);
.LP
\fIint\fR
\fBub_ctx_set_option\fR(\fIstruct ub_ctx*\fR ctx, \fIchar*\fR opt, \fIchar*\fR val);
.LP
\fIint\fR
\fBub_ctx_get_option\fR(\fIstruct ub_ctx*\fR ctx, \fIchar*\fR opt, \fIchar**\fR val);
.LP
\fIint\fR
\fBub_ctx_config\fR(\fIstruct ub_ctx*\fR ctx, \fIchar*\fR fname);
.LP
\fIint\fR
\fBub_ctx_set_fwd\fR(\fIstruct ub_ctx*\fR ctx, \fIchar*\fR addr);
.LP
\fIint\fR
\fBub_ctx_set_stub\fR(\fIstruct ub_ctx*\fR ctx, \fIchar*\fR zone,
\fIchar*\fR addr,
.br
\fIint\fR isprime);
.LP
\fIint\fR
\fBub_ctx_set_tls\fR(\fIstruct ub_ctx*\fR ctx, \fIint\fR tls);
.LP
\fIint\fR
\fBub_ctx_resolvconf\fR(\fIstruct ub_ctx*\fR ctx, \fIchar*\fR fname);
.LP
\fIint\fR
\fBub_ctx_hosts\fR(\fIstruct ub_ctx*\fR ctx, \fIchar*\fR fname);
.LP
\fIint\fR
\fBub_ctx_add_ta\fR(\fIstruct ub_ctx*\fR ctx, \fIchar*\fR ta);
.LP
\fIint\fR
\fBub_ctx_add_ta_autr\fR(\fIstruct ub_ctx*\fR ctx, \fIchar*\fR fname);
.LP
\fIint\fR
\fBub_ctx_add_ta_file\fR(\fIstruct ub_ctx*\fR ctx, \fIchar*\fR fname);
.LP
\fIint\fR
\fBub_ctx_trustedkeys\fR(\fIstruct ub_ctx*\fR ctx, \fIchar*\fR fname);
.LP
\fIint\fR
\fBub_ctx_debugout\fR(\fIstruct ub_ctx*\fR ctx, \fIFILE*\fR out);
.LP
\fIint\fR
\fBub_ctx_debuglevel\fR(\fIstruct ub_ctx*\fR ctx, \fIint\fR d);
.LP
\fIint\fR
\fBub_ctx_async\fR(\fIstruct ub_ctx*\fR ctx, \fIint\fR dothread);
.LP
\fIint\fR
\fBub_poll\fR(\fIstruct ub_ctx*\fR ctx);
.LP
\fIint\fR
\fBub_wait\fR(\fIstruct ub_ctx*\fR ctx);
.LP
\fIint\fR
\fBub_fd\fR(\fIstruct ub_ctx*\fR ctx);
.LP
\fIint\fR
\fBub_process\fR(\fIstruct ub_ctx*\fR ctx);
.LP
\fIint\fR
\fBub_resolve\fR(\fIstruct ub_ctx*\fR ctx, \fIchar*\fR name,
.br
\fIint\fR rrtype, \fIint\fR rrclass, \fIstruct ub_result**\fR result);
.LP
\fIint\fR
\fBub_resolve_async\fR(\fIstruct ub_ctx*\fR ctx, \fIchar*\fR name,
.br
\fIint\fR rrtype, \fIint\fR rrclass, \fIvoid*\fR mydata,
.br
\fIub_callback_type\fR callback, \fIint*\fR async_id);
.LP
\fIint\fR
\fBub_cancel\fR(\fIstruct ub_ctx*\fR ctx, \fIint\fR async_id);
.LP
\fIvoid\fR
\fBub_resolve_free\fR(\fIstruct ub_result*\fR result);
.LP
\fIconst char *\fR
\fBub_strerror\fR(\fIint\fR err);
.LP
\fIint\fR
\fBub_ctx_print_local_zones\fR(\fIstruct ub_ctx*\fR ctx);
.LP
\fIint\fR
\fBub_ctx_zone_add\fR(\fIstruct ub_ctx*\fR ctx, \fIchar*\fR zone_name, \fIchar*\fR zone_type);
.LP
\fIint\fR
\fBub_ctx_zone_remove\fR(\fIstruct ub_ctx*\fR ctx, \fIchar*\fR zone_name);
.LP
\fIint\fR
\fBub_ctx_data_add\fR(\fIstruct ub_ctx*\fR ctx, \fIchar*\fR data);
.LP
\fIint\fR
\fBub_ctx_data_remove\fR(\fIstruct ub_ctx*\fR ctx, \fIchar*\fR data);
.SH "DESCRIPTION"
.B Unbound
is an implementation of a DNS resolver, that does caching and
DNSSEC validation. This is the library API, for using the \-lunbound library.
The server daemon is described in \fIunbound\fR(8).
The library works independent from a running unbound server, and
can be used to convert hostnames to ip addresses, and back,
and obtain other information from the DNS. The library performs public\-key
validation of results with DNSSEC.
.P
The library uses a variable of type \fIstruct ub_ctx\fR to keep context
between calls. The user must maintain it, creating it with
.B ub_ctx_create
and deleting it with
.B ub_ctx_delete\fR.
It can be created and deleted at any time. Creating it anew removes any
previous configuration (such as trusted keys) and clears any cached results.
.P
The functions are thread\-safe, and a context can be used in a threaded (as
well as in a non\-threaded) environment. Also resolution (and validation)
can be performed blocking and non\-blocking (also called asynchronous).
The async method returns from the call immediately, so that processing
can go on, while the results become available later.
.P
.\" Man page generated from reStructuredText.
.
.
.nr rst2man-indent-level 0
.
.de1 rstReportMargin
\\$1 \\n[an-margin]
level \\n[rst2man-indent-level]
level margin: \\n[rst2man-indent\\n[rst2man-indent-level]]
-
\\n[rst2man-indent0]
\\n[rst2man-indent1]
\\n[rst2man-indent2]
..
.de1 INDENT
.\" .rstReportMargin pre:
. RS \\$1
. nr rst2man-indent\\n[rst2man-indent-level] \\n[an-margin]
. nr rst2man-indent-level +1
.\" .rstReportMargin post:
..
.de UNINDENT
. RE
.\" indent \\n[an-margin]
.\" old: \\n[rst2man-indent\\n[rst2man-indent-level]]
.nr rst2man-indent-level -1
.\" new: \\n[rst2man-indent\\n[rst2man-indent-level]]
.in \\n[rst2man-indent\\n[rst2man-indent-level]]u
..
.TH "LIBUNBOUND" "3" "@date@" "@version@" "Unbound"
.SH NAME
libunbound \- Unbound DNS validating resolver @version@ functions.
.SH SYNOPSIS
.sp
\fB#include <unbound.h>\fP
.sp
struct ub_ctx * \fBub_ctx_create\fP(void);
.sp
void \fBub_ctx_delete\fP(struct ub_ctx* ctx);
.sp
int \fBub_ctx_set_option\fP(struct ub_ctx* ctx, char* opt, char* val);
.sp
int \fBub_ctx_get_option\fP(struct ub_ctx* ctx, char* opt, char** val);
.sp
int \fBub_ctx_config\fP(struct ub_ctx* ctx, char* fname);
.sp
int \fBub_ctx_set_fwd\fP(struct ub_ctx* ctx, char* addr);
.INDENT 0.0
.TP
int \fBub_ctx_set_stub\fP(struct ub_ctx* ctx, char* zone, char* addr,
int isprime);
.UNINDENT
.sp
int \fBub_ctx_set_tls\fP(struct ub_ctx* ctx, int tls);
.sp
int \fBub_ctx_resolvconf\fP(struct ub_ctx* ctx, char* fname);
.sp
int \fBub_ctx_hosts\fP(struct ub_ctx* ctx, char* fname);
.sp
int \fBub_ctx_add_ta\fP(struct ub_ctx* ctx, char* ta);
.sp
int \fBub_ctx_add_ta_autr\fP(struct ub_ctx* ctx, char* fname);
.sp
int \fBub_ctx_add_ta_file\fP(struct ub_ctx* ctx, char* fname);
.sp
int \fBub_ctx_trustedkeys\fP(struct ub_ctx* ctx, char* fname);
.sp
int \fBub_ctx_debugout\fP(struct ub_ctx* ctx, FILE* out);
.sp
int \fBub_ctx_debuglevel\fP(struct ub_ctx* ctx, int d);
.sp
int \fBub_ctx_async\fP(struct ub_ctx* ctx, int dothread);
.sp
int \fBub_poll\fP(struct ub_ctx* ctx);
.sp
int \fBub_wait\fP(struct ub_ctx* ctx);
.sp
int \fBub_fd\fP(struct ub_ctx* ctx);
.sp
int \fBub_process\fP(struct ub_ctx* ctx);
.INDENT 0.0
.TP
int \fBub_resolve\fP(struct ub_ctx* ctx, char* name,
int rrtype, int rrclass, struct ub_result** result);
.TP
int \fBub_resolve_async\fP(struct ub_ctx* ctx, char* name,
int rrtype, int rrclass, void* mydata,
ub_callback_type* callback, int* async_id);
.UNINDENT
.sp
int \fBub_cancel\fP(struct ub_ctx* ctx, int async_id);
.sp
void \fBub_resolve_free\fP(struct ub_result* result);
.sp
const char * \fBub_strerror\fP(int err);
.sp
int \fBub_ctx_print_local_zones\fP(struct ub_ctx* ctx);
.sp
int \fBub_ctx_zone_add\fP(struct ub_ctx* ctx, char* zone_name, char* zone_type);
.sp
int \fBub_ctx_zone_remove\fP(struct ub_ctx* ctx, char* zone_name);
.sp
int \fBub_ctx_data_add\fP(struct ub_ctx* ctx, char* data);
.sp
int \fBub_ctx_data_remove\fP(struct ub_ctx* ctx, char* data);
.SH DESCRIPTION
.sp
Unbound is an implementation of a DNS resolver, that does caching and DNSSEC
validation.
This is the library API, for using the \fB\-lunbound\fP library.
The server daemon is described in \fI\%unbound(8)\fP\&.
The library works independent from a running unbound server, and can be used to
convert hostnames to ip addresses, and back, and obtain other information from
the DNS.
The library performs public\-key validation of results with DNSSEC.
.sp
The library uses a variable of type \fIstruct ub_ctx\fP to keep context between
calls.
The user must maintain it, creating it with \fBub_ctx_create\fP and deleting it
with \fBub_ctx_delete\fP\&.
It can be created and deleted at any time.
Creating it anew removes any previous configuration (such as trusted keys) and
clears any cached results.
.sp
The functions are thread\-safe, and a context can be used in a threaded (as well
as in a non\-threaded) environment.
Also resolution (and validation) can be performed blocking and non\-blocking
(also called asynchronous).
The async method returns from the call immediately, so that processing can go
on, while the results become available later.
.sp
The functions are discussed in turn below.
.SH "FUNCTIONS"
.TP
.SH FUNCTIONS
.INDENT 0.0
.TP
.B ub_ctx_create
Create a new context, initialised with defaults.
The information from /etc/resolv.conf and /etc/hosts is not utilised
by default. Use
.B ub_ctx_resolvconf
and
.B ub_ctx_hosts
to read them.
Before you call this, use the openssl functions CRYPTO_set_id_callback and
CRYPTO_set_locking_callback to set up asynchronous operation if you use
lib openssl (the application calls these functions once for initialisation).
Openssl 1.0.0 or later uses the CRYPTO_THREADID_set_callback function.
The information from \fB/etc/resolv.conf\fP and \fB/etc/hosts\fP is
not utilised by default.
Use \fBub_ctx_resolvconf\fP and \fBub_ctx_hosts\fP to read them.
Before you call this, use the openssl functions
\fBCRYPTO_set_id_callback\fP and \fBCRYPTO_set_locking_callback\fP to set
up asynchronous operation if you use lib openssl (the application calls
these functions once for initialisation).
Openssl 1.0.0 or later uses the \fBCRYPTO_THREADID_set_callback\fP
function.
.TP
.B ub_ctx_delete
Delete validation context and free associated resources.
Outstanding async queries are killed and callbacks are not called for them.
Outstanding async queries are killed and callbacks are not called for
them.
.TP
.B ub_ctx_set_option
A power\-user interface that lets you specify one of the options from the
config file format, see \fIunbound.conf\fR(5). Not all options are
relevant. For some specific options, such as adding trust anchors, special
routines exist. Pass the option name with the trailing ':'.
A power\-user interface that lets you specify one of the options from
the config file format, see \fI\%unbound.conf(5)\fP\&.
Not all options are relevant.
For some specific options, such as adding trust anchors, special
routines exist.
Pass the option name with the trailing \fB\(aq:\(aq\fP\&.
.TP
.B ub_ctx_get_option
A power\-user interface that gets an option value. Some options cannot be
gotten, and others return a newline separated list. Pass the option name
without trailing ':'. The returned value must be free(2)d by the caller.
A power\-user interface that gets an option value.
Some options cannot be gotten, and others return a newline separated
list.
Pass the option name without trailing \fB\(aq:\(aq\fP\&.
The returned value must be free(2)d by the caller.
.TP
.B ub_ctx_config
A power\-user interface that lets you specify an unbound config file, see
\fIunbound.conf\fR(5), which is read for configuration. Not all options are
relevant. For some specific options, such as adding trust anchors, special
routines exist. This function is thread\-safe only if a single instance of
ub_ctx* exists in the application. If several instances exist the
application has to ensure that ub_ctx_config is not called in parallel by
the different instances.
A power\-user interface that lets you specify an unbound config file,
see \fI\%unbound.conf(5)\fP, which is read for
configuration.
Not all options are relevant.
For some specific options, such as adding trust anchors, special
routines exist.
This function is thread\-safe only if a single instance of \fBub_ctx\fP*
exists in the application.
If several instances exist the application has to ensure that
\fBub_ctx_config\fP is not called in parallel by the different instances.
.TP
.B ub_ctx_set_fwd
Set machine to forward DNS queries to, the caching resolver to use.
IP4 or IP6 address. Forwards all DNS requests to that machine, which
is expected to run a recursive resolver. If the proxy is not
DNSSEC capable, validation may fail. Can be called several times, in
that case the addresses are used as backup servers.
At this time it is only possible to set configuration before the
first resolve is done.
Set machine to forward DNS queries to, the caching resolver to use.
IP4 or IP6 address.
Forwards all DNS requests to that machine, which is expected to run a
recursive resolver.
If the proxy is not DNSSEC capable, validation may fail.
Can be called several times, in that case the addresses are used as
backup servers.
At this time it is only possible to set configuration before the first
resolve is done.
.TP
.B ub_ctx_set_stub
Set a stub zone, authoritative dns servers to use for a particular zone.
IP4 or IP6 address. If the address is NULL the stub entry is removed.
Set isprime true if you configure root hints with it. Otherwise similar to
the stub zone item from unbound's config file. Can be called several times,
for different zones, or to add multiple addresses for a particular zone.
At this time it is only possible to set configuration before the
first resolve is done.
Set a stub zone, authoritative dns servers to use for a particular
zone.
IP4 or IP6 address.
If the address is NULL the stub entry is removed.
Set isprime true if you configure root hints with it.
Otherwise similar to the stub zone item from unbound\(aqs config file.
Can be called several times, for different zones, or to add multiple
addresses for a particular zone.
At this time it is only possible to set configuration before the first
resolve is done.
.TP
.B ub_ctx_set_tls
Enable DNS over TLS (DoT) for machines set with
.B ub_ctx_set_fwd.
At this time it is only possible to set configuration before the
first resolve is done.
Enable DNS over TLS (DoT) for machines set with \fBub_ctx_set_fwd\fP\&.
At this time it is only possible to set configuration before the first
resolve is done.
.TP
.B ub_ctx_resolvconf
By default the root servers are queried and full resolver mode is used, but
you can use this call to read the list of nameservers to use from the
filename given.
Usually "/etc/resolv.conf". Uses those nameservers as caching proxies.
By default the root servers are queried and full resolver mode is used,
but you can use this call to read the list of nameservers to use from
the filename given.
Usually \fB\(dq/etc/resolv.conf\(dq\fP\&.
Uses those nameservers as caching proxies.
If they do not support DNSSEC, validation may fail.
Only nameservers are picked up, the searchdomain, ndots and other
settings from \fIresolv.conf\fR(5) are ignored.
If fname NULL is passed, "/etc/resolv.conf" is used (if on Windows,
the system\-wide configured nameserver is picked instead).
At this time it is only possible to set configuration before the
first resolve is done.
settings from \fIresolv.conf(5)\fP are ignored.
If fname NULL is passed, \fB\(dq/etc/resolv.conf\(dq\fP is used (if on
Windows, the system\-wide configured nameserver is picked instead).
At this time it is only possible to set configuration before the first
resolve is done.
.TP
.B ub_ctx_hosts
Read list of hosts from the filename given.
Usually "/etc/hosts". When queried for, these addresses are not marked
DNSSEC secure. If fname NULL is passed, "/etc/hosts" is used
(if on Windows, etc/hosts from WINDIR is picked instead).
At this time it is only possible to set configuration before the
first resolve is done.
Usually \fB\(dq/etc/hosts\(dq\fP\&.
When queried for, these addresses are not marked DNSSEC secure.
If fname NULL is passed, \fB\(dq/etc/hosts\(dq\fP is used (if on Windows,
\fBetc/hosts\fP from WINDIR is picked instead).
At this time it is only possible to set configuration before the first
resolve is done.
.TP
.B
ub_ctx_add_ta
.B ub_ctx_add_ta
Add a trust anchor to the given context.
At this time it is only possible to add trusted keys before the
first resolve is done.
At this time it is only possible to add trusted keys before the first
resolve is done.
The format is a string, similar to the zone\-file format,
[domainname] [type] [rdata contents]. Both DS and DNSKEY records are accepted.
\fB[domainname]\fP \fB[type]\fP \fB[rdata contents]\fP\&.
Both DS and DNSKEY records are accepted.
.TP
.B ub_ctx_add_ta_autr
Add filename with automatically tracked trust anchor to the given context.
Pass name of a file with the managed trust anchor. You can create this
file with \fIunbound\-anchor\fR(8) for the root anchor. You can also
create it with an initial file with one line with a DNSKEY or DS record.
Add filename with automatically tracked trust anchor to the given
context.
Pass name of a file with the managed trust anchor.
You can create this file with
\fI\%unbound\-anchor(8)\fP for the root anchor.
You can also create it with an initial file with one line with a DNSKEY
or DS record.
If the file is writable, it is updated when the trust anchor changes.
At this time it is only possible to add trusted keys before the
first resolve is done.
At this time it is only possible to add trusted keys before the first
resolve is done.
.TP
.B ub_ctx_add_ta_file
Add trust anchors to the given context.
Pass name of a file with DS and DNSKEY records in zone file format.
At this time it is only possible to add trusted keys before the
first resolve is done.
At this time it is only possible to add trusted keys before the first
resolve is done.
.TP
.B ub_ctx_trustedkeys
Add trust anchors to the given context.
Pass the name of a bind\-style config file with trusted\-keys{}.
At this time it is only possible to add trusted keys before the
first resolve is done.
Pass the name of a bind\-style config file with \fBtrusted\-keys{}\fP\&.
At this time it is only possible to add trusted keys before the first
resolve is done.
.TP
.B ub_ctx_debugout
Set debug and error log output to the given stream. Pass NULL to disable
output. Default is stderr. File\-names or using syslog can be enabled
using config options, this routine is for using your own stream.
Set debug and error log output to the given stream.
Pass NULL to disable output.
Default is stderr.
File\-names or using syslog can be enabled using config options, this
routine is for using your own stream.
.TP
.B ub_ctx_debuglevel
Set debug verbosity for the context. Output is directed to stderr.
Set debug verbosity for the context.
Output is directed to stderr.
Higher debug level gives more output.
.TP
.B ub_ctx_async
Set a context behaviour for asynchronous action.
if set to true, enables threading and a call to
.B ub_resolve_async
if set to true, enables threading and a call to \fBub_resolve_async\fP
creates a thread to handle work in the background.
If false, a process is forked to handle work in the background.
Changes to this setting after
.B ub_resolve_async
calls have been made have no effect (delete and re\-create the context
to change).
Changes to this setting after \fBub_resolve_async\fP calls have been made
have no effect (delete and re\-create the context to change).
.TP
.B ub_poll
Poll a context to see if it has any new results.
Do not poll in a loop, instead extract the fd below to poll for readiness,
and then check, or wait using the wait routine.
Do not poll in a loop, instead extract the \fBfd\fP below to poll for
readiness, and then check, or wait using the wait routine.
Returns 0 if nothing to read, or nonzero if a result is available.
If nonzero, call
.B ub_process
to do callbacks.
If nonzero, call \fBub_process\fP to do callbacks.
.TP
.B ub_wait
Wait for a context to finish with results. Calls
.B ub_process
after the wait for you. After the wait, there are no more outstanding
asynchronous queries.
Wait for a context to finish with results.
Calls \fBub_process\fP after the wait for you.
After the wait, there are no more outstanding asynchronous queries.
.TP
.B ub_fd
Get file descriptor. Wait for it to become readable, at this point
answers are returned from the asynchronous validating resolver.
Then call the \fBub_process\fR to continue processing.
Get file descriptor.
Wait for it to become readable, at this point answers are returned from
the asynchronous validating resolver.
Then call the \fBub_process\fP to continue processing.
.TP
.B ub_process
Call this routine to continue processing results from the validating
resolver (when the fd becomes readable).
resolver (when the \fBfd\fP becomes readable).
Will perform necessary callbacks.
.TP
.B ub_resolve
@@ -340,95 +311,111 @@ The result structure is newly allocated with the resulting data.
.TP
.B ub_resolve_async
Perform asynchronous resolution and validation of the target name.
Arguments mean the same as for \fBub_resolve\fR except no
data is returned immediately, instead a callback is called later.
The callback receives a copy of the mydata pointer, that you can use to pass
information to the callback. The callback type is a function pointer to
a function declared as
.IP
void my_callback_function(void* my_arg, int err,
.br
struct ub_result* result);
.IP
The async_id is returned so you can (at your option) decide to track it
and cancel the request if needed. If you pass a NULL pointer the async_id
is not returned.
Arguments mean the same as for \fBub_resolve\fP except no data is
returned immediately, instead a callback is called later.
The callback receives a copy of the mydata pointer, that you can use to
pass information to the callback.
The callback type is a function pointer to a function declared as:
.INDENT 7.0
.INDENT 3.5
.sp
.nf
.ft C
void my_callback_function(void* my_arg, int err,
struct ub_result* result);
.ft P
.fi
.UNINDENT
.UNINDENT
.sp
The \fBasync_id\fP is returned so you can (at your option) decide to
track it and cancel the request if needed.
If you pass a NULL pointer the \fBasync_id\fP is not returned.
.TP
.B ub_cancel
Cancel an async query in progress. This may return an error if the query
does not exist, or the query is already being delivered, in that case you
may still get a callback for the query.
Cancel an async query in progress.
This may return an error if the query does not exist, or the query is
already being delivered, in that case you may still get a callback for
the query.
.TP
.B ub_resolve_free
Free struct ub_result contents after use.
Free struct \fBub_result\fP contents after use.
.TP
.B ub_strerror
Convert error value from one of the unbound library functions
to a human readable string.
Convert error value from one of the unbound library functions to a
human readable string.
.TP
.B ub_ctx_print_local_zones
Debug printout the local authority information to debug output.
.TP
.B ub_ctx_zone_add
Add new zone to local authority info, like local\-zone \fIunbound.conf\fR(5)
statement.
Add new zone to local authority info, like local\-zone
\fI\%unbound.conf(5)\fP statement.
.TP
.B ub_ctx_zone_remove
Delete zone from local authority info.
.TP
.B ub_ctx_data_add
Add resource record data to local authority info, like local\-data
\fIunbound.conf\fR(5) statement.
\fI\%unbound.conf(5)\fP statement.
.TP
.B ub_ctx_data_remove
Delete local authority data from the name given.
.SH "RESULT DATA STRUCTURE"
The result of the DNS resolution and validation is returned as
\fIstruct ub_result\fR. The result structure contains the following entries.
.P
.UNINDENT
.SH RESULT DATA STRUCTURE
.sp
The result of the DNS resolution and validation is returned as \fIstruct
ub_result\fP\&.
The result structure contains the following entries:
.INDENT 0.0
.INDENT 3.5
.sp
.nf
struct ub_result {
char* qname; /* text string, original question */
int qtype; /* type code asked for */
int qclass; /* class code asked for */
char** data; /* array of rdata items, NULL terminated*/
int* len; /* array with lengths of rdata items */
char* canonname; /* canonical name of result */
int rcode; /* additional error code in case of no data */
void* answer_packet; /* full network format answer packet */
int answer_len; /* length of packet in octets */
int havedata; /* true if there is data */
int nxdomain; /* true if nodata because name does not exist */
int secure; /* true if result is secure */
int bogus; /* true if a security failure happened */
char* why_bogus; /* string with error if bogus */
int was_ratelimited; /* true if the query was ratelimited (SERVFAIL) by unbound */
int ttl; /* number of seconds the result is valid */
};
.ft C
struct ub_result {
char* qname; /* text string, original question */
int qtype; /* type code asked for */
int qclass; /* class code asked for */
char** data; /* array of rdata items, NULL terminated*/
int* len; /* array with lengths of rdata items */
char* canonname; /* canonical name of result */
int rcode; /* additional error code in case of no data */
void* answer_packet; /* full network format answer packet */
int answer_len; /* length of packet in octets */
int havedata; /* true if there is data */
int nxdomain; /* true if nodata because name does not exist */
int secure; /* true if result is secure */
int bogus; /* true if a security failure happened */
char* why_bogus; /* string with error if bogus */
int was_ratelimited; /* true if the query was ratelimited (SERVFAIL) by unbound */
int ttl; /* number of seconds the result is valid */
};
.ft P
.fi
.P
If both secure and bogus are false, security was not enabled for the
domain of the query. Else, they are not both true, one of them is true.
.SH "RETURN VALUES"
Many routines return an error code. The value 0 (zero) denotes no error
happened. Other values can be passed to
.B ub_strerror
to obtain a readable error string.
.B ub_strerror
returns a zero terminated string.
.B ub_ctx_create
returns NULL on an error (a malloc failure).
.B ub_poll
returns true if some information may be available, false otherwise.
.B ub_fd
returns a file descriptor or \-1 on error.
.B ub_ctx_config
and
.B ub_ctx_resolvconf
attempt to leave errno informative on a function return with file read failure.
.SH "SEE ALSO"
\fIunbound.conf\fR(5),
\fIunbound\fR(8).
.SH "AUTHORS"
.B Unbound
developers are mentioned in the CREDITS file in the distribution.
.UNINDENT
.UNINDENT
.sp
If both secure and bogus are false, security was not enabled for the domain of
the query.
Else, they are not both true, one of them is true.
.SH RETURN VALUES
.sp
Many routines return an error code.
The value 0 (zero) denotes no error happened.
Other values can be passed to \fBub_strerror\fP to obtain a readable error
string.
\fBub_strerror\fP returns a zero terminated string.
\fBub_ctx_create\fP returns NULL on an error (a malloc failure).
\fBub_poll\fP returns true if some information may be available, false otherwise.
\fBub_fd\fP returns a file descriptor or \-1 on error.
\fBub_ctx_config\fP and \fBub_ctx_resolvconf\fP attempt to leave errno informative
on a function return with file read failure.
.SH SEE ALSO
.sp
\fI\%unbound.conf(5)\fP, \fI\%unbound(8)\fP\&.
.SH AUTHOR
Unbound developers are mentioned in the CREDITS file in the distribution.
.SH COPYRIGHT
1999-2025, NLnet Labs
.\" Generated by docutils manpage writer.
.
+491
View File
@@ -0,0 +1,491 @@
..
WHEN EDITING MAKE SURE EACH SENTENCE STARTS ON A NEW LINE
..
IT HELPS RENDERERS TO DO THE RIGHT THING WRT SPACE
..
IT HELPS PEOPLE DIFFING THE CHANGES
..
WHEN EDITING MAKE SURE EACH SENTENCE STARTS ON A NEW LINE
..
IT HELPS RENDERERS TO DO THE RIGHT THING WRT SPACE
..
IT HELPS PEOPLE DIFFING THE CHANGES
..
WHEN EDITING MAKE SURE EACH SENTENCE STARTS ON A NEW LINE
..
IT HELPS RENDERERS TO DO THE RIGHT THING WRT SPACE
..
IT HELPS PEOPLE DIFFING THE CHANGES
..
WHEN EDITING MAKE SURE EACH SENTENCE STARTS ON A NEW LINE
..
IT HELPS RENDERERS TO DO THE RIGHT THING WRT SPACE
..
IT HELPS PEOPLE DIFFING THE CHANGES
libunbound(3)
=============
Synopsis
--------
.. only:: html
.. code-block:: c
#include <unbound.h>
struct ub_ctx * ub_ctx_create(void);
void ub_ctx_delete(struct ub_ctx* ctx);
int ub_ctx_set_option(struct ub_ctx* ctx, char* opt, char* val);
int ub_ctx_get_option(struct ub_ctx* ctx, char* opt, char** val);
int ub_ctx_config(struct ub_ctx* ctx, char* fname);
int ub_ctx_set_fwd(struct ub_ctx* ctx, char* addr);
int ub_ctx_set_stub(struct ub_ctx* ctx, char* zone, char* addr,
int isprime);
int ub_ctx_set_tls(struct ub_ctx* ctx, int tls);
int ub_ctx_resolvconf(struct ub_ctx* ctx, char* fname);
int ub_ctx_hosts(struct ub_ctx* ctx, char* fname);
int ub_ctx_add_ta(struct ub_ctx* ctx, char* ta);
int ub_ctx_add_ta_autr(struct ub_ctx* ctx, char* fname);
int ub_ctx_add_ta_file(struct ub_ctx* ctx, char* fname);
int ub_ctx_trustedkeys(struct ub_ctx* ctx, char* fname);
int ub_ctx_debugout(struct ub_ctx* ctx, FILE* out);
int ub_ctx_debuglevel(struct ub_ctx* ctx, int d);
int ub_ctx_async(struct ub_ctx* ctx, int dothread);
int ub_poll(struct ub_ctx* ctx);
int ub_wait(struct ub_ctx* ctx);
int ub_fd(struct ub_ctx* ctx);
int ub_process(struct ub_ctx* ctx);
int ub_resolve(struct ub_ctx* ctx, char* name, int rrtype,
int rrclass, struct ub_result** result);
int ub_resolve_async(struct ub_ctx* ctx, char* name, int rrtype,
int rrclass, void* mydata, ub_callback_type callback,
int* async_id);
int ub_cancel(struct ub_ctx* ctx, int async_id);
void ub_resolve_free(struct ub_result* result);
const char * ub_strerror(int err);
int ub_ctx_print_local_zones(struct ub_ctx* ctx);
int ub_ctx_zone_add(struct ub_ctx* ctx, char* zone_name, char* zone_type);
int ub_ctx_zone_remove(struct ub_ctx* ctx, char* zone_name);
int ub_ctx_data_add(struct ub_ctx* ctx, char* data);
int ub_ctx_data_remove(struct ub_ctx* ctx, char* data);
.. only:: man
**#include <unbound.h>**
struct ub_ctx \* **ub_ctx_create**\ (void);
void **ub_ctx_delete**\ (struct ub_ctx\* ctx);
int **ub_ctx_set_option**\ (struct ub_ctx\* ctx, char\* opt, char\* val);
int **ub_ctx_get_option**\ (struct ub_ctx\* ctx, char\* opt, char\*\* val);
int **ub_ctx_config**\ (struct ub_ctx\* ctx, char* fname);
int **ub_ctx_set_fwd**\ (struct ub_ctx\* ctx, char\* addr);
int **ub_ctx_set_stub**\ (struct ub_ctx\* ctx, char\* zone, char\* addr,
int isprime);
int **ub_ctx_set_tls**\ (struct ub_ctx\* ctx, int tls);
int **ub_ctx_resolvconf**\ (struct ub_ctx\* ctx, char\* fname);
int **ub_ctx_hosts**\ (struct ub_ctx\* ctx, char\* fname);
int **ub_ctx_add_ta**\ (struct ub_ctx\* ctx, char\* ta);
int **ub_ctx_add_ta_autr**\ (struct ub_ctx\* ctx, char\* fname);
int **ub_ctx_add_ta_file**\ (struct ub_ctx\* ctx, char\* fname);
int **ub_ctx_trustedkeys**\ (struct ub_ctx\* ctx, char\* fname);
int **ub_ctx_debugout**\ (struct ub_ctx\* ctx, FILE\* out);
int **ub_ctx_debuglevel**\ (struct ub_ctx\* ctx, int d);
int **ub_ctx_async**\ (struct ub_ctx\* ctx, int dothread);
int **ub_poll**\ (struct ub_ctx\* ctx);
int **ub_wait**\ (struct ub_ctx\* ctx);
int **ub_fd**\ (struct ub_ctx\* ctx);
int **ub_process**\ (struct ub_ctx\* ctx);
int **ub_resolve**\ (struct ub_ctx\* ctx, char\* name,
int rrtype, int rrclass, struct ub_result\*\* result);
int **ub_resolve_async**\ (struct ub_ctx\* ctx, char\* name,
int rrtype, int rrclass, void\* mydata,
ub_callback_type\* callback, int\* async_id);
int **ub_cancel**\ (struct ub_ctx\* ctx, int async_id);
void **ub_resolve_free**\ (struct ub_result\* result);
const char \* **ub_strerror**\ (int err);
int **ub_ctx_print_local_zones**\ (struct ub_ctx\* ctx);
int **ub_ctx_zone_add**\ (struct ub_ctx\* ctx, char\* zone_name, char\* zone_type);
int **ub_ctx_zone_remove**\ (struct ub_ctx\* ctx, char\* zone_name);
int **ub_ctx_data_add**\ (struct ub_ctx\* ctx, char\* data);
int **ub_ctx_data_remove**\ (struct ub_ctx\* ctx, char\* data);
Description
-----------
Unbound is an implementation of a DNS resolver, that does caching and DNSSEC
validation.
This is the library API, for using the ``-lunbound`` library.
The server daemon is described in :doc:`unbound(8)</manpages/unbound>`.
The library works independent from a running unbound server, and can be used to
convert hostnames to ip addresses, and back, and obtain other information from
the DNS.
The library performs public-key validation of results with DNSSEC.
The library uses a variable of type *struct ub_ctx* to keep context between
calls.
The user must maintain it, creating it with **ub_ctx_create** and deleting it
with **ub_ctx_delete**.
It can be created and deleted at any time.
Creating it anew removes any previous configuration (such as trusted keys) and
clears any cached results.
The functions are thread-safe, and a context can be used in a threaded (as well
as in a non-threaded) environment.
Also resolution (and validation) can be performed blocking and non-blocking
(also called asynchronous).
The async method returns from the call immediately, so that processing can go
on, while the results become available later.
The functions are discussed in turn below.
Functions
---------
.. glossary::
ub_ctx_create
Create a new context, initialised with defaults.
The information from :file:`/etc/resolv.conf` and :file:`/etc/hosts` is
not utilised by default.
Use **ub_ctx_resolvconf** and **ub_ctx_hosts** to read them.
Before you call this, use the openssl functions
**CRYPTO_set_id_callback** and **CRYPTO_set_locking_callback** to set
up asynchronous operation if you use lib openssl (the application calls
these functions once for initialisation).
Openssl 1.0.0 or later uses the **CRYPTO_THREADID_set_callback**
function.
ub_ctx_delete
Delete validation context and free associated resources.
Outstanding async queries are killed and callbacks are not called for
them.
ub_ctx_set_option
A power-user interface that lets you specify one of the options from
the config file format, see :doc:`unbound.conf(5)</manpages/unbound.conf>`.
Not all options are relevant.
For some specific options, such as adding trust anchors, special
routines exist.
Pass the option name with the trailing ``':'``.
ub_ctx_get_option
A power-user interface that gets an option value.
Some options cannot be gotten, and others return a newline separated
list.
Pass the option name without trailing ``':'``.
The returned value must be free(2)d by the caller.
ub_ctx_config
A power-user interface that lets you specify an unbound config file,
see :doc:`unbound.conf(5)</manpages/unbound.conf>`, which is read for
configuration.
Not all options are relevant.
For some specific options, such as adding trust anchors, special
routines exist.
This function is thread-safe only if a single instance of **ub_ctx**\*
exists in the application.
If several instances exist the application has to ensure that
**ub_ctx_config** is not called in parallel by the different instances.
ub_ctx_set_fwd
Set machine to forward DNS queries to, the caching resolver to use.
IP4 or IP6 address.
Forwards all DNS requests to that machine, which is expected to run a
recursive resolver.
If the proxy is not DNSSEC capable, validation may fail.
Can be called several times, in that case the addresses are used as
backup servers.
At this time it is only possible to set configuration before the first
resolve is done.
ub_ctx_set_stub
Set a stub zone, authoritative dns servers to use for a particular
zone.
IP4 or IP6 address.
If the address is NULL the stub entry is removed.
Set isprime true if you configure root hints with it.
Otherwise similar to the stub zone item from unbound's config file.
Can be called several times, for different zones, or to add multiple
addresses for a particular zone.
At this time it is only possible to set configuration before the first
resolve is done.
ub_ctx_set_tls
Enable DNS over TLS (DoT) for machines set with **ub_ctx_set_fwd**.
At this time it is only possible to set configuration before the first
resolve is done.
ub_ctx_resolvconf
By default the root servers are queried and full resolver mode is used,
but you can use this call to read the list of nameservers to use from
the filename given.
Usually :file:`"/etc/resolv.conf"`.
Uses those nameservers as caching proxies.
If they do not support DNSSEC, validation may fail.
Only nameservers are picked up, the searchdomain, ndots and other
settings from *resolv.conf(5)* are ignored.
If fname NULL is passed, :file:`"/etc/resolv.conf"` is used (if on
Windows, the system-wide configured nameserver is picked instead).
At this time it is only possible to set configuration before the first
resolve is done.
ub_ctx_hosts
Read list of hosts from the filename given.
Usually :file:`"/etc/hosts"`.
When queried for, these addresses are not marked DNSSEC secure.
If fname NULL is passed, :file:`"/etc/hosts"` is used (if on Windows,
:file:`etc/hosts` from WINDIR is picked instead).
At this time it is only possible to set configuration before the first
resolve is done.
ub_ctx_add_ta
Add a trust anchor to the given context.
At this time it is only possible to add trusted keys before the first
resolve is done.
The format is a string, similar to the zone-file format,
**[domainname]** **[type]** **[rdata contents]**.
Both DS and DNSKEY records are accepted.
ub_ctx_add_ta_autr
Add filename with automatically tracked trust anchor to the given
context.
Pass name of a file with the managed trust anchor.
You can create this file with
:doc:`unbound-anchor(8)</manpages/unbound-anchor>` for the root anchor.
You can also create it with an initial file with one line with a DNSKEY
or DS record.
If the file is writable, it is updated when the trust anchor changes.
At this time it is only possible to add trusted keys before the first
resolve is done.
ub_ctx_add_ta_file
Add trust anchors to the given context.
Pass name of a file with DS and DNSKEY records in zone file format.
At this time it is only possible to add trusted keys before the first
resolve is done.
ub_ctx_trustedkeys
Add trust anchors to the given context.
Pass the name of a bind-style config file with ``trusted-keys{}``.
At this time it is only possible to add trusted keys before the first
resolve is done.
ub_ctx_debugout
Set debug and error log output to the given stream.
Pass NULL to disable output.
Default is stderr.
File-names or using syslog can be enabled using config options, this
routine is for using your own stream.
ub_ctx_debuglevel
Set debug verbosity for the context.
Output is directed to stderr.
Higher debug level gives more output.
ub_ctx_async
Set a context behaviour for asynchronous action.
if set to true, enables threading and a call to **ub_resolve_async**
creates a thread to handle work in the background.
If false, a process is forked to handle work in the background.
Changes to this setting after **ub_resolve_async** calls have been made
have no effect (delete and re-create the context to change).
ub_poll
Poll a context to see if it has any new results.
Do not poll in a loop, instead extract the **fd** below to poll for
readiness, and then check, or wait using the wait routine.
Returns 0 if nothing to read, or nonzero if a result is available.
If nonzero, call **ub_process** to do callbacks.
ub_wait
Wait for a context to finish with results.
Calls **ub_process** after the wait for you.
After the wait, there are no more outstanding asynchronous queries.
ub_fd
Get file descriptor.
Wait for it to become readable, at this point answers are returned from
the asynchronous validating resolver.
Then call the **ub_process** to continue processing.
ub_process
Call this routine to continue processing results from the validating
resolver (when the **fd** becomes readable).
Will perform necessary callbacks.
ub_resolve
Perform resolution and validation of the target name.
The name is a domain name in a zero terminated text string.
The rrtype and rrclass are DNS type and class codes.
The result structure is newly allocated with the resulting data.
ub_resolve_async
Perform asynchronous resolution and validation of the target name.
Arguments mean the same as for **ub_resolve** except no data is
returned immediately, instead a callback is called later.
The callback receives a copy of the mydata pointer, that you can use to
pass information to the callback.
The callback type is a function pointer to a function declared as:
.. code-block:: c
void my_callback_function(void* my_arg, int err,
struct ub_result* result);
The **async_id** is returned so you can (at your option) decide to
track it and cancel the request if needed.
If you pass a NULL pointer the **async_id** is not returned.
ub_cancel
Cancel an async query in progress.
This may return an error if the query does not exist, or the query is
already being delivered, in that case you may still get a callback for
the query.
ub_resolve_free
Free struct **ub_result** contents after use.
ub_strerror
Convert error value from one of the unbound library functions to a
human readable string.
ub_ctx_print_local_zones
Debug printout the local authority information to debug output.
ub_ctx_zone_add
Add new zone to local authority info, like local-zone
:doc:`unbound.conf(5)</manpages/unbound.conf>` statement.
ub_ctx_zone_remove
Delete zone from local authority info.
ub_ctx_data_add
Add resource record data to local authority info, like local-data
:doc:`unbound.conf(5)</manpages/unbound.conf>` statement.
ub_ctx_data_remove
Delete local authority data from the name given.
Result Data structure
---------------------
The result of the DNS resolution and validation is returned as *struct
ub_result*.
The result structure contains the following entries:
.. code-block:: c
struct ub_result {
char* qname; /* text string, original question */
int qtype; /* type code asked for */
int qclass; /* class code asked for */
char** data; /* array of rdata items, NULL terminated*/
int* len; /* array with lengths of rdata items */
char* canonname; /* canonical name of result */
int rcode; /* additional error code in case of no data */
void* answer_packet; /* full network format answer packet */
int answer_len; /* length of packet in octets */
int havedata; /* true if there is data */
int nxdomain; /* true if nodata because name does not exist */
int secure; /* true if result is secure */
int bogus; /* true if a security failure happened */
char* why_bogus; /* string with error if bogus */
int was_ratelimited; /* true if the query was ratelimited (SERVFAIL) by unbound */
int ttl; /* number of seconds the result is valid */
};
If both secure and bogus are false, security was not enabled for the domain of
the query.
Else, they are not both true, one of them is true.
Return Values
-------------
Many routines return an error code.
The value 0 (zero) denotes no error happened.
Other values can be passed to **ub_strerror** to obtain a readable error
string.
**ub_strerror** returns a zero terminated string.
**ub_ctx_create** returns NULL on an error (a malloc failure).
**ub_poll** returns true if some information may be available, false otherwise.
**ub_fd** returns a file descriptor or -1 on error.
**ub_ctx_config** and **ub_ctx_resolvconf** attempt to leave errno informative
on a function return with file read failure.
See Also
--------
:doc:`unbound.conf(5)</manpages/unbound.conf>`, :doc:`unbound(8)</manpages/unbound>`.
+245 -134
View File
@@ -1,189 +1,300 @@
.TH "unbound-anchor" "8" "@date@" "NLnet Labs" "unbound @version@"
.\"
.\" unbound-anchor.8 -- unbound anchor maintenance utility manual
.\"
.\" Copyright (c) 2008, NLnet Labs. All rights reserved.
.\"
.\" See LICENSE for the license.
.\"
.\"
.SH "NAME"
.B unbound\-anchor
\- Unbound anchor utility.
.SH "SYNOPSIS"
.B unbound\-anchor
.RB [ opts ]
.SH "DESCRIPTION"
.B Unbound\-anchor
performs setup or update of the root trust anchor for DNSSEC validation.
The program fetches the trust anchor with the method from RFC7958 when
regular RFC5011 update fails to bring it up to date.
.\" Man page generated from reStructuredText.
.
.
.nr rst2man-indent-level 0
.
.de1 rstReportMargin
\\$1 \\n[an-margin]
level \\n[rst2man-indent-level]
level margin: \\n[rst2man-indent\\n[rst2man-indent-level]]
-
\\n[rst2man-indent0]
\\n[rst2man-indent1]
\\n[rst2man-indent2]
..
.de1 INDENT
.\" .rstReportMargin pre:
. RS \\$1
. nr rst2man-indent\\n[rst2man-indent-level] \\n[an-margin]
. nr rst2man-indent-level +1
.\" .rstReportMargin post:
..
.de UNINDENT
. RE
.\" indent \\n[an-margin]
.\" old: \\n[rst2man-indent\\n[rst2man-indent-level]]
.nr rst2man-indent-level -1
.\" new: \\n[rst2man-indent\\n[rst2man-indent-level]]
.in \\n[rst2man-indent\\n[rst2man-indent-level]]u
..
.TH "UNBOUND-ANCHOR" "8" "@date@" "@version@" "Unbound"
.SH NAME
unbound-anchor \- Unbound @version@ anchor utility.
.SH SYNOPSIS
.sp
\fBunbound\-anchor\fP [\fBopts\fP]
.SH DESCRIPTION
.sp
\fBunbound\-anchor\fP performs setup or update of the root trust anchor for DNSSEC
validation.
The program fetches the trust anchor with the method from \fI\%RFC 7958\fP when
regular \fI\%RFC 5011\fP update fails to bring it up to date.
It can be run (as root) from the commandline, or run as part of startup
scripts. Before you start the \fIunbound\fR(8) DNS server.
.P
scripts.
Before you start the \fI\%unbound(8)\fP DNS server.
.sp
Suggested usage:
.P
.INDENT 0.0
.INDENT 3.5
.sp
.nf
# in the init scripts.
# provide or update the root anchor (if necessary)
unbound-anchor \-a "@UNBOUND_ROOTKEY_FILE@"
# Please note usage of this root anchor is at your own risk
# and under the terms of our LICENSE (see source).
#
# start validating resolver
# the unbound.conf contains:
# auto-trust-anchor-file: "@UNBOUND_ROOTKEY_FILE@"
unbound \-c unbound.conf
.ft C
# in the init scripts.
# provide or update the root anchor (if necessary)
unbound\-anchor \-a \(dq@UNBOUND_ROOTKEY_FILE@\(dq
# Please note usage of this root anchor is at your own risk
# and under the terms of our LICENSE (see source).
#
# start validating resolver
# the unbound.conf contains:
# auto\-trust\-anchor\-file: \(dq@UNBOUND_ROOTKEY_FILE@\(dq
unbound \-c unbound.conf
.ft P
.fi
.P
This tool provides builtin default contents for the root anchor and root
update certificate files.
.P
.UNINDENT
.UNINDENT
.sp
This tool provides builtin default contents for the root anchor and root update
certificate files.
.sp
It tests if the root anchor file works, and if not, and an update is possible,
attempts to update the root anchor using the root update certificate.
It performs a https fetch of root-anchors.xml and checks the results (RFC7958),
if all checks are successful, it updates the root anchor file. Otherwise
the root anchor file is unchanged. It performs RFC5011 tracking if the
DNSSEC information available via the DNS makes that possible.
.P
It does not perform an update if the certificate is expired, if the network
is down or other errors occur.
.P
It performs a https fetch of
\fI\%root\-anchors.xml\fP
and checks the results (\fI\%RFC 7958\fP); if all checks are successful, it updates
the root anchor file.
Otherwise the root anchor file is unchanged.
It performs \fI\%RFC 5011\fP tracking if the DNSSEC information available via the
DNS makes that possible.
.sp
It does not perform an update if the certificate is expired, if the network is
down or other errors occur.
.sp
The available options are:
.INDENT 0.0
.TP
.B \-a \fIfile
.B \-a <file>
The root anchor key file, that is read in and written out.
Default is @UNBOUND_ROOTKEY_FILE@.
If the file does not exist, or is empty, a builtin root key is written to it.
Default is \fB@UNBOUND_ROOTKEY_FILE@\fP\&.
If the file does not exist, or is empty, a builtin root key is written
to it.
.UNINDENT
.INDENT 0.0
.TP
.B \-c \fIfile
.B \-c <file>
The root update certificate file, that is read in.
Default is @UNBOUND_ROOTCERT_FILE@.
Default is \fB@UNBOUND_ROOTCERT_FILE@\fP\&.
If the file does not exist, or is empty, a builtin certificate is used.
.UNINDENT
.INDENT 0.0
.TP
.B \-l
List the builtin root key and builtin root update certificate on stdout.
.UNINDENT
.INDENT 0.0
.TP
.B \-u \fIname
The server name, it connects to https://name. Specify without https:// prefix.
The default is "data.iana.org". It connects to the port specified with \-P.
.B \-u <name>
The server name, it connects to \fBhttps://name\fP\&.
Specify without \fBhttps://\fP prefix.
The default is \fB\(dqdata.iana.org\(dq\fP\&.
It connects to the port specified with \fI\%\-P\fP\&.
You can pass an IPv4 address or IPv6 address (no brackets) if you want.
.UNINDENT
.INDENT 0.0
.TP
.B \-S
Do not use SNI for the HTTPS connection. Default is to use SNI.
Do not use SNI for the HTTPS connection.
Default is to use SNI.
.UNINDENT
.INDENT 0.0
.TP
.B \-b \fIaddress
The source address to bind to for domain resolution and contacting the server
on https. May be either an IPv4 address or IPv6 address (no brackets).
.B \-b <address>
The source address to bind to for domain resolution and contacting the
server on https.
May be either an IPv4 address or IPv6 address (no brackets).
.UNINDENT
.INDENT 0.0
.TP
.B \-x \fIpath
The pathname to the root\-anchors.xml file on the server. (forms URL with \-u).
The default is /root\-anchors/root\-anchors.xml.
.B \-x <path>
The pathname to the root\-anchors.xml file on the server.
(forms URL with \fI\%\-u\fP).
The default is \fB/root\-anchors/root\-anchors.xml\fP\&.
.UNINDENT
.INDENT 0.0
.TP
.B \-s \fIpath
The pathname to the root\-anchors.p7s file on the server. (forms URL with \-u).
The default is /root\-anchors/root\-anchors.p7s. This file has to be a PKCS7
signature over the xml file, using the pem file (\-c) as trust anchor.
.B \-s <path>
The pathname to the root\-anchors.p7s file on the server.
(forms URL with \fI\%\-u\fP).
The default is \fB/root\-anchors/root\-anchors.p7s\fP\&.
This file has to be a PKCS7 signature over the xml file, using the pem
file (\fI\%\-c\fP) as trust anchor.
.UNINDENT
.INDENT 0.0
.TP
.B \-n \fIname
The emailAddress for the Subject of the signer's certificate from the p7s
signature file. Only signatures from this name are allowed. default is
dnssec@iana.org. If you pass "" then the emailAddress is not checked.
.B \-n <name>
The emailAddress for the Subject of the signer\(aqs certificate from the
p7s signature file.
Only signatures from this name are allowed.
The default is \fBdnssec@iana.org\fP\&.
If you pass \fB\(dq\(dq\fP then the emailAddress is not checked.
.UNINDENT
.INDENT 0.0
.TP
.B \-4
Use IPv4 for domain resolution and contacting the server on https. Default is
to use IPv4 and IPv6 where appropriate.
Use IPv4 for domain resolution and contacting the server on
https.
Default is to use IPv4 and IPv6 where appropriate.
.UNINDENT
.INDENT 0.0
.TP
.B \-6
Use IPv6 for domain resolution and contacting the server on https. Default is
to use IPv4 and IPv6 where appropriate.
Use IPv6 for domain resolution and contacting the server on https.
Default is to use IPv4 and IPv6 where appropriate.
.UNINDENT
.INDENT 0.0
.TP
.B \-f \fIresolv.conf
Use the given resolv.conf file. Not enabled by default, but you could try to
pass /etc/resolv.conf on some systems. It contains the IP addresses of the
recursive nameservers to use. However, since this tool could be used to
bootstrap that very recursive nameserver, it would not be useful (since
that server is not up yet, since we are bootstrapping it). It could be
useful in a situation where you know an upstream cache is deployed (and
running) and in captive portal situations.
.B \-f <resolv.conf>
Use the given resolv.conf file.
Not enabled by default, but you could try to pass
\fB/etc/resolv.conf\fP on some systems.
It contains the IP addresses of the recursive nameservers to use.
However, since this tool could be used to bootstrap that very recursive
nameserver, it would not be useful (since that server is not up yet,
since we are bootstrapping it).
It could be useful in a situation where you know an upstream cache is
deployed (and running) and in captive portal situations.
.UNINDENT
.INDENT 0.0
.TP
.B \-r \fIroot.hints
Use the given root.hints file (same syntax as the BIND and Unbound root hints
file) to bootstrap domain resolution. By default a list of builtin root
hints is used. Unbound\-anchor goes to the network itself for these roots,
to resolve the server (\-u option) and to check the root DNSKEY records.
.B \-r <root.hints>
Use the given root.hints file (same syntax as the BIND and Unbound root
hints file) to bootstrap domain resolution.
By default a list of builtin root hints is used.
unbound\-anchor goes to the network itself for these roots, to resolve
the server (\fI\%\-u\fP option) and to check the root DNSKEY records.
It does so, because the tool when used for bootstrapping the recursive
resolver, cannot use that recursive resolver itself because it is bootstrapping
that server.
resolver, cannot use that recursive resolver itself because it is
bootstrapping that server.
.UNINDENT
.INDENT 0.0
.TP
.B \-R
Allow fallback from \-f resolv.conf file to direct root servers query.
It allows you to prefer local resolvers, but fallback automatically
to direct root query if they do not respond or do not support DNSSEC.
Allow fallback from \fI\%\-f\fP \fB<resolv.conf>\fP file to direct root
servers query.
It allows you to prefer local resolvers, but fallback automatically to
direct root query if they do not respond or do not support DNSSEC.
.UNINDENT
.INDENT 0.0
.TP
.B \-v
More verbose. Once prints informational messages, multiple times may enable
large debug amounts (such as full certificates or byte\-dumps of downloaded
files). By default it prints almost nothing. It also prints nothing on
errors by default; in that case the original root anchor file is simply
left undisturbed, so that a recursive server can start right after it.
More verbose.
Once prints informational messages, multiple times may enable large
debug amounts (such as full certificates or byte\-dumps of downloaded
files).
By default it prints almost nothing.
It also prints nothing on errors by default; in that case the original
root anchor file is simply left undisturbed, so that a recursive server
can start right after it.
.UNINDENT
.INDENT 0.0
.TP
.B \-C \fIunbound.conf
Debug option to read unbound.conf into the resolver process used.
.B \-C <unbound.conf>
Debug option to read \fB<unbound.conf>\fP into the resolver process
used.
.UNINDENT
.INDENT 0.0
.TP
.B \-P \fIport
Set the port number to use for the https connection. The default is 443.
.B \-P <port>
Set the port number to use for the https connection.
The default is 443.
.UNINDENT
.INDENT 0.0
.TP
.B \-F
Debug option to force update of the root anchor through downloading the xml
file and verifying it with the certificate. By default it first tries to
update by contacting the DNS, which uses much less bandwidth, is much
faster (200 msec not 2 sec), and is nicer to the deployed infrastructure.
With this option, it still attempts to do so (and may verbosely tell you),
but then ignores the result and goes on to use the xml fallback method.
Debug option to force update of the root anchor through downloading the
xml file and verifying it with the certificate.
By default it first tries to update by contacting the DNS, which uses
much less bandwidth, is much faster (200 msec not 2 sec), and is nicer
to the deployed infrastructure.
With this option, it still attempts to do so (and may verbosely tell
you), but then ignores the result and goes on to use the xml fallback
method.
.UNINDENT
.INDENT 0.0
.TP
.B \-h
Show the version and commandline option help.
.SH "EXIT CODE"
.UNINDENT
.SH EXIT CODE
.sp
This tool exits with value 1 if the root anchor was updated using the
certificate or if the builtin root-anchor was used. It exits with code
0 if no update was necessary, if the update was possible with RFC5011
tracking, or if an error occurred.
.P
certificate or if the builtin root\-anchor was used.
It exits with code 0 if no update was necessary, if the update was possible
with \fI\%RFC 5011\fP tracking, or if an error occurred.
.sp
You can check the exit value in this manner:
.INDENT 0.0
.INDENT 3.5
.sp
.nf
unbound-anchor \-a "root.key" || logger "Please check root.key"
.ft C
unbound\-anchor \-a \(dqroot.key\(dq || logger \(dqPlease check root.key\(dq
.ft P
.fi
.UNINDENT
.UNINDENT
.sp
Or something more suitable for your operational environment.
.SH "TRUST"
The root keys and update certificate included in this tool
are provided for convenience and under the terms of our
license (see the LICENSE file in the source distribution or
https://github.com/NLnetLabs/unbound/blob/master/LICENSE) and might be stale or
not suitable to your purpose.
.P
By running "unbound\-anchor \-l" the keys and certificate that are
.SH TRUST
.sp
The root keys and update certificate included in this tool are provided for
convenience and under the terms of our license (see the LICENSE file in the
source distribution or \fI\%https://github.com/NLnetLabs/unbound/blob/master/LICENSE\fP
and might be stale or not suitable to your purpose.
.sp
By running \fI\%unbound\-anchor \-l\fP the keys and certificate that are
configured in the code are printed for your convenience.
.P
The build\-in configuration can be overridden by providing a root\-cert
file and a rootkey file.
.SH "FILES"
.sp
The built\-in configuration can be overridden by providing a root\-cert file and
a rootkey file.
.SH FILES
.INDENT 0.0
.TP
.I @UNBOUND_ROOTKEY_FILE@
The root anchor file, updated with 5011 tracking, and read and written to.
.B @UNBOUND_ROOTKEY_FILE@
The root anchor file, updated with 5011 tracking, and read and written
to.
The file is created if it does not exist.
.TP
.I @UNBOUND_ROOTCERT_FILE@
The trusted self\-signed certificate that is used to verify the downloaded
DNSSEC root trust anchor. You can update it by fetching it from
https://data.iana.org/root\-anchors/icannbundle.pem (and validate it).
.B @UNBOUND_ROOTCERT_FILE@
The trusted self\-signed certificate that is used to verify the
downloaded DNSSEC root trust anchor.
You can update it by fetching it from
\fI\%https://data.iana.org/root\-anchors/icannbundle.pem\fP (and validate it).
If the file does not exist or is empty, a builtin version is used.
.TP
.I https://data.iana.org/root\-anchors/root\-anchors.xml
.B \fI\%https://data.iana.org/root\-anchors/root\-anchors.xml\fP
Source for the root key information.
.TP
.I https://data.iana.org/root\-anchors/root\-anchors.p7s
.B \fI\%https://data.iana.org/root\-anchors/root\-anchors.p7s\fP
Signature on the root key information.
.SH "SEE ALSO"
\fIunbound.conf\fR(5),
\fIunbound\fR(8).
.UNINDENT
.SH SEE ALSO
.sp
\fI\%unbound.conf(5)\fP,
\fI\%unbound(8)\fP\&.
.SH AUTHOR
Unbound developers are mentioned in the CREDITS file in the distribution.
.SH COPYRIGHT
1999-2025, NLnet Labs
.\" Generated by docutils manpage writer.
.
+281
View File
@@ -0,0 +1,281 @@
..
WHEN EDITING MAKE SURE EACH SENTENCE STARTS ON A NEW LINE
..
IT HELPS RENDERERS TO DO THE RIGHT THING WRT SPACE
..
IT HELPS PEOPLE DIFFING THE CHANGES
..
WHEN EDITING MAKE SURE EACH SENTENCE STARTS ON A NEW LINE
..
IT HELPS RENDERERS TO DO THE RIGHT THING WRT SPACE
..
IT HELPS PEOPLE DIFFING THE CHANGES
..
WHEN EDITING MAKE SURE EACH SENTENCE STARTS ON A NEW LINE
..
IT HELPS RENDERERS TO DO THE RIGHT THING WRT SPACE
..
IT HELPS PEOPLE DIFFING THE CHANGES
..
WHEN EDITING MAKE SURE EACH SENTENCE STARTS ON A NEW LINE
..
IT HELPS RENDERERS TO DO THE RIGHT THING WRT SPACE
..
IT HELPS PEOPLE DIFFING THE CHANGES
.. program:: unbound-anchor
unbound-anchor(8)
=================
Synopsis
--------
**unbound-anchor** [``opts``]
Description
-----------
``unbound-anchor`` performs setup or update of the root trust anchor for DNSSEC
validation.
The program fetches the trust anchor with the method from :rfc:`7958` when
regular :rfc:`5011` update fails to bring it up to date.
It can be run (as root) from the commandline, or run as part of startup
scripts.
Before you start the :doc:`unbound(8)</manpages/unbound>` DNS server.
Suggested usage:
.. code-block:: text
# in the init scripts.
# provide or update the root anchor (if necessary)
unbound-anchor -a "@UNBOUND_ROOTKEY_FILE@"
# Please note usage of this root anchor is at your own risk
# and under the terms of our LICENSE (see source).
#
# start validating resolver
# the unbound.conf contains:
# auto-trust-anchor-file: "@UNBOUND_ROOTKEY_FILE@"
unbound -c unbound.conf
This tool provides builtin default contents for the root anchor and root update
certificate files.
It tests if the root anchor file works, and if not, and an update is possible,
attempts to update the root anchor using the root update certificate.
It performs a https fetch of
`root-anchors.xml <http://data.iana.org/root-anchors/root-anchors.xml>`__
and checks the results (:rfc:`7958`); if all checks are successful, it updates
the root anchor file.
Otherwise the root anchor file is unchanged.
It performs :rfc:`5011` tracking if the DNSSEC information available via the
DNS makes that possible.
It does not perform an update if the certificate is expired, if the network is
down or other errors occur.
The available options are:
.. option:: -a <file>
The root anchor key file, that is read in and written out.
Default is :file:`@UNBOUND_ROOTKEY_FILE@`.
If the file does not exist, or is empty, a builtin root key is written
to it.
.. option:: -c <file>
The root update certificate file, that is read in.
Default is :file:`@UNBOUND_ROOTCERT_FILE@`.
If the file does not exist, or is empty, a builtin certificate is used.
.. option:: -l
List the builtin root key and builtin root update certificate on stdout.
.. option:: -u <name>
The server name, it connects to ``https://name``.
Specify without ``https://`` prefix.
The default is ``"data.iana.org"``.
It connects to the port specified with :option:`-P`.
You can pass an IPv4 address or IPv6 address (no brackets) if you want.
.. option:: -S
Do not use SNI for the HTTPS connection.
Default is to use SNI.
.. option:: -b <address>
The source address to bind to for domain resolution and contacting the
server on https.
May be either an IPv4 address or IPv6 address (no brackets).
.. option:: -x <path>
The pathname to the root-anchors.xml file on the server.
(forms URL with :option:`-u`).
The default is :file:`/root-anchors/root-anchors.xml`.
.. option:: -s <path>
The pathname to the root-anchors.p7s file on the server.
(forms URL with :option:`-u`).
The default is :file:`/root-anchors/root-anchors.p7s`.
This file has to be a PKCS7 signature over the xml file, using the pem
file (:option:`-c`) as trust anchor.
.. option:: -n <name>
The emailAddress for the Subject of the signer's certificate from the
p7s signature file.
Only signatures from this name are allowed.
The default is ``dnssec@iana.org``.
If you pass ``""`` then the emailAddress is not checked.
.. option:: -4
Use IPv4 for domain resolution and contacting the server on
https.
Default is to use IPv4 and IPv6 where appropriate.
.. option:: -6
Use IPv6 for domain resolution and contacting the server on https.
Default is to use IPv4 and IPv6 where appropriate.
.. option:: -f <resolv.conf>
Use the given resolv.conf file.
Not enabled by default, but you could try to pass
:file:`/etc/resolv.conf` on some systems.
It contains the IP addresses of the recursive nameservers to use.
However, since this tool could be used to bootstrap that very recursive
nameserver, it would not be useful (since that server is not up yet,
since we are bootstrapping it).
It could be useful in a situation where you know an upstream cache is
deployed (and running) and in captive portal situations.
.. option:: -r <root.hints>
Use the given root.hints file (same syntax as the BIND and Unbound root
hints file) to bootstrap domain resolution.
By default a list of builtin root hints is used.
unbound-anchor goes to the network itself for these roots, to resolve
the server (:option:`-u` option) and to check the root DNSKEY records.
It does so, because the tool when used for bootstrapping the recursive
resolver, cannot use that recursive resolver itself because it is
bootstrapping that server.
.. option:: -R
Allow fallback from :option:`-f` ``<resolv.conf>`` file to direct root
servers query.
It allows you to prefer local resolvers, but fallback automatically to
direct root query if they do not respond or do not support DNSSEC.
.. option:: -v
More verbose.
Once prints informational messages, multiple times may enable large
debug amounts (such as full certificates or byte-dumps of downloaded
files).
By default it prints almost nothing.
It also prints nothing on errors by default; in that case the original
root anchor file is simply left undisturbed, so that a recursive server
can start right after it.
.. option:: -C <unbound.conf>
Debug option to read :file:`<unbound.conf>` into the resolver process
used.
.. option:: -P <port>
Set the port number to use for the https connection.
The default is 443.
.. option:: -F
Debug option to force update of the root anchor through downloading the
xml file and verifying it with the certificate.
By default it first tries to update by contacting the DNS, which uses
much less bandwidth, is much faster (200 msec not 2 sec), and is nicer
to the deployed infrastructure.
With this option, it still attempts to do so (and may verbosely tell
you), but then ignores the result and goes on to use the xml fallback
method.
.. option:: -h
Show the version and commandline option help.
Exit Code
---------
This tool exits with value 1 if the root anchor was updated using the
certificate or if the builtin root-anchor was used.
It exits with code 0 if no update was necessary, if the update was possible
with :rfc:`5011` tracking, or if an error occurred.
You can check the exit value in this manner:
.. code-block:: text
unbound-anchor -a "root.key" || logger "Please check root.key"
Or something more suitable for your operational environment.
Trust
-----
The root keys and update certificate included in this tool are provided for
convenience and under the terms of our license (see the LICENSE file in the
source distribution or https://github.com/NLnetLabs/unbound/blob/master/LICENSE
and might be stale or not suitable to your purpose.
By running :option:`unbound-anchor -l` the keys and certificate that are
configured in the code are printed for your convenience.
The built-in configuration can be overridden by providing a root-cert file and
a rootkey file.
Files
-----
@UNBOUND_ROOTKEY_FILE@
The root anchor file, updated with 5011 tracking, and read and written
to.
The file is created if it does not exist.
@UNBOUND_ROOTCERT_FILE@
The trusted self-signed certificate that is used to verify the
downloaded DNSSEC root trust anchor.
You can update it by fetching it from
https://data.iana.org/root-anchors/icannbundle.pem (and validate it).
If the file does not exist or is empty, a builtin version is used.
https://data.iana.org/root-anchors/root-anchors.xml
Source for the root key information.
https://data.iana.org/root-anchors/root-anchors.p7s
Signature on the root key information.
See Also
--------
:doc:`unbound.conf(5)</manpages/unbound.conf>`,
:doc:`unbound(8)</manpages/unbound>`.
+80 -39
View File
@@ -1,52 +1,93 @@
.TH "unbound-checkconf" "8" "@date@" "NLnet Labs" "unbound @version@"
.\"
.\" unbound-checkconf.8 -- unbound configuration checker manual
.\"
.\" Copyright (c) 2007, NLnet Labs. All rights reserved.
.\"
.\" See LICENSE for the license.
.\"
.\"
.SH "NAME"
unbound\-checkconf
\- Check Unbound configuration file for errors.
.SH "SYNOPSIS"
.B unbound\-checkconf
.RB [ \-h ]
.RB [ \-f ]
.RB [ \-o
.IR option ]
.RI [ cfgfile ]
.SH "DESCRIPTION"
.B Unbound\-checkconf
checks the configuration file for the
\fIunbound\fR(8)
DNS resolver for syntax and other errors.
.\" Man page generated from reStructuredText.
.
.
.nr rst2man-indent-level 0
.
.de1 rstReportMargin
\\$1 \\n[an-margin]
level \\n[rst2man-indent-level]
level margin: \\n[rst2man-indent\\n[rst2man-indent-level]]
-
\\n[rst2man-indent0]
\\n[rst2man-indent1]
\\n[rst2man-indent2]
..
.de1 INDENT
.\" .rstReportMargin pre:
. RS \\$1
. nr rst2man-indent\\n[rst2man-indent-level] \\n[an-margin]
. nr rst2man-indent-level +1
.\" .rstReportMargin post:
..
.de UNINDENT
. RE
.\" indent \\n[an-margin]
.\" old: \\n[rst2man-indent\\n[rst2man-indent-level]]
.nr rst2man-indent-level -1
.\" new: \\n[rst2man-indent\\n[rst2man-indent-level]]
.in \\n[rst2man-indent\\n[rst2man-indent-level]]u
..
.TH "UNBOUND-CHECKCONF" "8" "@date@" "@version@" "Unbound"
.SH NAME
unbound-checkconf \- Check Unbound @version@ configuration file for errors.
.SH SYNOPSIS
.sp
\fBunbound\-checkconf\fP [\fB\-hf\fP] [\fB\-o option\fP] [cfgfile]
.SH DESCRIPTION
.sp
\fBunbound\-checkconf\fP checks the configuration file for the
\fI\%unbound(8)\fP DNS resolver for syntax and other errors.
The config file syntax is described in
\fIunbound.conf\fR(5).
.P
\fI\%unbound.conf(5)\fP\&.
.sp
The available options are:
.INDENT 0.0
.TP
.B \-h
Show the version and commandline option help.
.UNINDENT
.INDENT 0.0
.TP
.B \-f
Print full pathname, with chroot applied to it. Use with the \-o option.
Print full pathname, with chroot applied to it.
Use with the \fI\%\-o\fP option.
.UNINDENT
.INDENT 0.0
.TP
.B \-o\fI option
.B \-q
Make the operation quiet, suppress output on success.
.UNINDENT
.INDENT 0.0
.TP
.B \-o <option>
If given, after checking the config file the value of this option is
printed to stdout. For "" (disabled) options an empty line is printed.
printed to stdout.
For \fB\(dq\(dq\fP (disabled) options an empty line is printed.
.UNINDENT
.INDENT 0.0
.TP
.I cfgfile
The config file to read with settings for Unbound. It is checked.
.B cfgfile
The config file to read with settings for Unbound.
It is checked.
If omitted, the config file at the default location is checked.
.SH "EXIT CODE"
The unbound\-checkconf program exits with status code 1 on error,
0 for a correct config file.
.SH "FILES"
.UNINDENT
.SH EXIT CODE
.sp
The \fBunbound\-checkconf\fP program exits with status code 1 on error, 0 for a
correct config file.
.SH FILES
.INDENT 0.0
.TP
.I @ub_conf_file@
.B @ub_conf_file@
Unbound configuration file.
.SH "SEE ALSO"
\fIunbound.conf\fR(5),
\fIunbound\fR(8).
.UNINDENT
.SH SEE ALSO
.sp
\fI\%unbound.conf(5)\fP,
\fI\%unbound(8)\fP\&.
.SH AUTHOR
Unbound developers are mentioned in the CREDITS file in the distribution.
.SH COPYRIGHT
1999-2025, NLnet Labs
.\" Generated by docutils manpage writer.
.
+98
View File
@@ -0,0 +1,98 @@
..
WHEN EDITING MAKE SURE EACH SENTENCE STARTS ON A NEW LINE
..
IT HELPS RENDERERS TO DO THE RIGHT THING WRT SPACE
..
IT HELPS PEOPLE DIFFING THE CHANGES
..
WHEN EDITING MAKE SURE EACH SENTENCE STARTS ON A NEW LINE
..
IT HELPS RENDERERS TO DO THE RIGHT THING WRT SPACE
..
IT HELPS PEOPLE DIFFING THE CHANGES
..
WHEN EDITING MAKE SURE EACH SENTENCE STARTS ON A NEW LINE
..
IT HELPS RENDERERS TO DO THE RIGHT THING WRT SPACE
..
IT HELPS PEOPLE DIFFING THE CHANGES
..
WHEN EDITING MAKE SURE EACH SENTENCE STARTS ON A NEW LINE
..
IT HELPS RENDERERS TO DO THE RIGHT THING WRT SPACE
..
IT HELPS PEOPLE DIFFING THE CHANGES
.. program:: unbound-checkconf
unbound-checkconf(8)
====================
Synopsis
--------
**unbound-checkconf** [``-hf``] [``-o option``] [cfgfile]
Description
-----------
``unbound-checkconf`` checks the configuration file for the
:doc:`unbound(8)</manpages/unbound>` DNS resolver for syntax and other errors.
The config file syntax is described in
:doc:`unbound.conf(5)</manpages/unbound.conf>`.
The available options are:
.. option:: -h
Show the version and commandline option help.
.. option:: -f
Print full pathname, with chroot applied to it.
Use with the :option:`-o` option.
.. option:: -q
Make the operation quiet, suppress output on success.
.. option:: -o <option>
If given, after checking the config file the value of this option is
printed to stdout.
For ``""`` (disabled) options an empty line is printed.
.. option:: cfgfile
The config file to read with settings for Unbound.
It is checked.
If omitted, the config file at the default location is checked.
Exit Code
---------
The ``unbound-checkconf`` program exits with status code 1 on error, 0 for a
correct config file.
Files
-----
@ub_conf_file@
Unbound configuration file.
See Also
--------
:doc:`unbound.conf(5)</manpages/unbound.conf>`,
:doc:`unbound(8)</manpages/unbound>`.
+1286 -437
View File
File diff suppressed because it is too large Load Diff
File diff suppressed because it is too large Load Diff
+157 -85
View File
@@ -1,118 +1,190 @@
.TH "unbound\-host" "1" "@date@" "NLnet Labs" "unbound @version@"
.\"
.\" unbound-host.1 -- unbound DNS lookup utility
.\"
.\" Copyright (c) 2007, NLnet Labs. All rights reserved.
.\"
.\" See LICENSE for the license.
.\"
.\"
.SH "NAME"
.B unbound\-host
\- unbound DNS lookup utility
.SH "SYNOPSIS"
.B unbound\-host
.RB [ \-C
.IR configfile ]
.RB [ \-vdhr46D ]
.RB [ \-c
.IR class ]
.RB [ \-t
.IR type ]
.RB [ \-y
.IR key ]
.RB [ \-f
.IR keyfile ]
.RB [ \-F
.IR namedkeyfile ]
.I hostname
.SH "DESCRIPTION"
.B Unbound\-host
uses the Unbound validating resolver to query for the hostname and display
results. With the \fB\-v\fR option it displays validation
status: secure, insecure, bogus (security failure).
.P
By default it reads no configuration file whatsoever. It attempts to reach
the internet root servers. With \fB\-C\fR an Unbound config file and with
\fB\-r\fR resolv.conf can be read.
.P
.\" Man page generated from reStructuredText.
.
.
.nr rst2man-indent-level 0
.
.de1 rstReportMargin
\\$1 \\n[an-margin]
level \\n[rst2man-indent-level]
level margin: \\n[rst2man-indent\\n[rst2man-indent-level]]
-
\\n[rst2man-indent0]
\\n[rst2man-indent1]
\\n[rst2man-indent2]
..
.de1 INDENT
.\" .rstReportMargin pre:
. RS \\$1
. nr rst2man-indent\\n[rst2man-indent-level] \\n[an-margin]
. nr rst2man-indent-level +1
.\" .rstReportMargin post:
..
.de UNINDENT
. RE
.\" indent \\n[an-margin]
.\" old: \\n[rst2man-indent\\n[rst2man-indent-level]]
.nr rst2man-indent-level -1
.\" new: \\n[rst2man-indent\\n[rst2man-indent-level]]
.in \\n[rst2man-indent\\n[rst2man-indent-level]]u
..
.TH "UNBOUND-HOST" "1" "@date@" "@version@" "Unbound"
.SH NAME
unbound-host \- Unbound @version@ DNS lookup utility.
.SH SYNOPSIS
.sp
\fBunbound\-host\fP [\fB\-C configfile\fP] [\fB\-vdhr46D\fP] [\fB\-c class\fP]
[\fB\-t type\fP] [\fB\-y key\fP] [\fB\-f keyfile\fP] [\fB\-F namedkeyfile\fP] hostname
.SH DESCRIPTION
.sp
\fBunbound\-host\fP uses the Unbound validating resolver to query for the hostname
and display results.
With the \fI\%\-v\fP option it displays validation status: secure, insecure,
bogus (security failure).
.sp
By default it reads no configuration file whatsoever.
It attempts to reach the internet root servers.
With \fI\%\-C\fP an unbound config file and with \fI\%\-r\fP \fBresolv.conf\fP
can be read.
.sp
The available options are:
.INDENT 0.0
.TP
.I hostname
.B hostname
This name is resolved (looked up in the DNS).
If a IPv4 or IPv6 address is given, a reverse lookup is performed.
.UNINDENT
.INDENT 0.0
.TP
.B \-h
Show the version and commandline option help.
.UNINDENT
.INDENT 0.0
.TP
.B \-v
Enable verbose output and it shows validation results, on every line.
Secure means that the NXDOMAIN (no such domain name), nodata (no such data)
or positive data response validated correctly with one of the keys.
Secure means that the NXDOMAIN (no such domain name), nodata (no such
data) or positive data response validated correctly with one of the
keys.
Insecure means that that domain name has no security set up for it.
Bogus (security failure) means that the response failed one or more checks,
it is likely wrong, outdated, tampered with, or broken.
Bogus (security failure) means that the response failed one or more
checks, it is likely wrong, outdated, tampered with, or broken.
.UNINDENT
.INDENT 0.0
.TP
.B \-d
Enable debug output to stderr. One \-d shows what the resolver and validator
are doing and may tell you what is going on. More times, \-d \-d, gives a
lot of output, with every packet sent and received.
Enable debug output to stderr.
One \fI\%\-d\fP shows what the resolver and validator are doing and may
tell you what is going on.
More times, \fI\%\-d\fP \fI\%\-d\fP, gives a lot of output, with every
packet sent and received.
.UNINDENT
.INDENT 0.0
.TP
.B \-c \fIclass
Specify the class to lookup for, the default is IN the internet class.
.B \-c <class>
Specify the class to lookup for, the default is IN the internet
class.
.UNINDENT
.INDENT 0.0
.TP
.B \-t \fItype
Specify the type of data to lookup. The default looks for IPv4, IPv6 and
mail handler data, or domain name pointers for reverse queries.
.B \-t <type>
Specify the type of data to lookup.
The default looks for IPv4, IPv6 and mail handler data, or domain name
pointers for reverse queries.
.UNINDENT
.INDENT 0.0
.TP
.B \-y \fIkey
Specify a public key to use as trust anchor. This is the base for a chain
of trust that is built up from the trust anchor to the response, in order
to validate the response message. Can be given as a DS or DNSKEY record.
For example \-y "example.com DS 31560 5 1 1CFED84787E6E19CCF9372C1187325972FE546CD".
.B \-y <key>
Specify a public key to use as trust anchor.
This is the base for a chain of trust that is built up from the trust
anchor to the response, in order to validate the response message.
Can be given as a DS or DNSKEY record.
For example:
.INDENT 7.0
.INDENT 3.5
.sp
.nf
.ft C
\-y \(dqexample.com DS 31560 5 1 1CFED84787E6E19CCF9372C1187325972FE546CD\(dq
.ft P
.fi
.UNINDENT
.UNINDENT
.UNINDENT
.INDENT 0.0
.TP
.B \-D
Enables DNSSEC validation. Reads the root anchor from the default configured
root anchor at the default location, \fI@UNBOUND_ROOTKEY_FILE@\fR.
Enables DNSSEC validation.
Reads the root anchor from the default configured root anchor at the
default location, \fB@UNBOUND_ROOTKEY_FILE@\fP\&.
.UNINDENT
.INDENT 0.0
.TP
.B \-f \fIkeyfile
Reads keys from a file. Every line has a DS or DNSKEY record, in the format
as for \-y. The zone file format, the same as dig and drill produce.
.B \-f <keyfile>
Reads keys from a file.
Every line has a DS or DNSKEY record, in the format as for \fI\%\-y\fP\&.
The zone file format, the same as \fBdig\fP and \fBdrill\fP produce.
.UNINDENT
.INDENT 0.0
.TP
.B \-F \fInamedkeyfile
Reads keys from a BIND\-style named.conf file. Only the trusted\-key {}; entries
are read.
.B \-F <namedkeyfile>
Reads keys from a BIND\-style \fBnamed.conf\fP file.
Only the \fBtrusted\-key {};\fP entries are read.
.UNINDENT
.INDENT 0.0
.TP
.B \-C \fIconfigfile
Uses the specified unbound.conf to prime
.IR libunbound (3).
.B \-C <configfile>
Uses the specified unbound.conf to prime \fI\%libunbound(3)\fP\&.
Pass it as first argument if you want to override some options from the
config file with further arguments on the commandline.
.UNINDENT
.INDENT 0.0
.TP
.B \-r
Read /etc/resolv.conf, and use the forward DNS servers from there (those could
have been set by DHCP). More info in
.IR resolv.conf (5).
Read \fB/etc/resolv.conf\fP, and use the forward DNS servers from
there (those could have been set by DHCP).
More info in \fIresolv.conf(5)\fP\&.
Breaks validation if those servers do not support DNSSEC.
.UNINDENT
.INDENT 0.0
.TP
.B \-4
Use solely the IPv4 network for sending packets.
.UNINDENT
.INDENT 0.0
.TP
.B \-6
Use solely the IPv6 network for sending packets.
.SH "EXAMPLES"
Some examples of use. The keys shown below are fakes, thus a security failure
is encountered.
.P
.UNINDENT
.SH EXAMPLES
.sp
Some examples of use.
The keys shown below are fakes, thus a security failure is encountered.
.INDENT 0.0
.INDENT 3.5
.sp
.nf
.ft C
$ unbound\-host www.example.com
.P
$ unbound\-host \-v \-y "example.com DS 31560 5 1 1CFED84787E6E19CCF9372C1187325972FE546CD" www.example.com
.P
$ unbound\-host \-v \-y "example.com DS 31560 5 1 1CFED84787E6E19CCF9372C1187325972FE546CD" 192.0.2.153
.SH "EXIT CODE"
The unbound\-host program exits with status code 1 on error,
0 on no error. The data may not be available on exit code 0, exit code 1
means the lookup encountered a fatal error.
.SH "SEE ALSO"
\fIunbound.conf\fR(5),
\fIunbound\fR(8).
$ unbound\-host \-v \-y \(dqexample.com DS 31560 5 1 1CFED84787E6E19CCF9372C1187325972FE546CD\(dq www.example.com
$ unbound\-host \-v \-y \(dqexample.com DS 31560 5 1 1CFED84787E6E19CCF9372C1187325972FE546CD\(dq 192.0.2.153
.ft P
.fi
.UNINDENT
.UNINDENT
.SH EXIT CODE
.sp
The \fBunbound\-host\fP program exits with status code 1 on error, 0 on no error.
The data may not be available on exit code 0, exit code 1 means the lookup
encountered a fatal error.
.SH SEE ALSO
.sp
\fI\%unbound.conf(5)\fP,
\fI\%unbound(8)\fP\&.
.SH AUTHOR
Unbound developers are mentioned in the CREDITS file in the distribution.
.SH COPYRIGHT
1999-2025, NLnet Labs
.\" Generated by docutils manpage writer.
.
+176
View File
@@ -0,0 +1,176 @@
..
WHEN EDITING MAKE SURE EACH SENTENCE STARTS ON A NEW LINE
..
IT HELPS RENDERERS TO DO THE RIGHT THING WRT SPACE
..
IT HELPS PEOPLE DIFFING THE CHANGES
..
WHEN EDITING MAKE SURE EACH SENTENCE STARTS ON A NEW LINE
..
IT HELPS RENDERERS TO DO THE RIGHT THING WRT SPACE
..
IT HELPS PEOPLE DIFFING THE CHANGES
..
WHEN EDITING MAKE SURE EACH SENTENCE STARTS ON A NEW LINE
..
IT HELPS RENDERERS TO DO THE RIGHT THING WRT SPACE
..
IT HELPS PEOPLE DIFFING THE CHANGES
..
WHEN EDITING MAKE SURE EACH SENTENCE STARTS ON A NEW LINE
..
IT HELPS RENDERERS TO DO THE RIGHT THING WRT SPACE
..
IT HELPS PEOPLE DIFFING THE CHANGES
.. program:: unbound-host
unbound-host(1)
===============
Synopsis
--------
**unbound-host** [``-C configfile``] [``-vdhr46D``] [``-c class``]
[``-t type``] [``-y key``] [``-f keyfile``] [``-F namedkeyfile``] hostname
Description
-----------
``unbound-host`` uses the Unbound validating resolver to query for the hostname
and display results.
With the :option:`-v` option it displays validation status: secure, insecure,
bogus (security failure).
By default it reads no configuration file whatsoever.
It attempts to reach the internet root servers.
With :option:`-C` an unbound config file and with :option:`-r` ``resolv.conf``
can be read.
The available options are:
.. option:: hostname
This name is resolved (looked up in the DNS).
If a IPv4 or IPv6 address is given, a reverse lookup is performed.
.. option:: -h
Show the version and commandline option help.
.. option:: -v
Enable verbose output and it shows validation results, on every line.
Secure means that the NXDOMAIN (no such domain name), nodata (no such
data) or positive data response validated correctly with one of the
keys.
Insecure means that that domain name has no security set up for it.
Bogus (security failure) means that the response failed one or more
checks, it is likely wrong, outdated, tampered with, or broken.
.. option:: -d
Enable debug output to stderr.
One :option:`-d` shows what the resolver and validator are doing and may
tell you what is going on.
More times, :option:`-d` :option:`-d`, gives a lot of output, with every
packet sent and received.
.. option:: -c <class>
Specify the class to lookup for, the default is IN the internet
class.
.. option:: -t <type>
Specify the type of data to lookup.
The default looks for IPv4, IPv6 and mail handler data, or domain name
pointers for reverse queries.
.. option:: -y <key>
Specify a public key to use as trust anchor.
This is the base for a chain of trust that is built up from the trust
anchor to the response, in order to validate the response message.
Can be given as a DS or DNSKEY record.
For example:
.. code-block:: text
-y "example.com DS 31560 5 1 1CFED84787E6E19CCF9372C1187325972FE546CD"
.. option:: -D
Enables DNSSEC validation.
Reads the root anchor from the default configured root anchor at the
default location, :file:`@UNBOUND_ROOTKEY_FILE@`.
.. option:: -f <keyfile>
Reads keys from a file.
Every line has a DS or DNSKEY record, in the format as for :option:`-y`.
The zone file format, the same as ``dig`` and ``drill`` produce.
.. option:: -F <namedkeyfile>
Reads keys from a BIND-style :file:`named.conf` file.
Only the ``trusted-key {};`` entries are read.
.. option:: -C <configfile>
Uses the specified unbound.conf to prime :doc:`libunbound(3)</manpages/libunbound>`.
Pass it as first argument if you want to override some options from the
config file with further arguments on the commandline.
.. option:: -r
Read :file:`/etc/resolv.conf`, and use the forward DNS servers from
there (those could have been set by DHCP).
More info in *resolv.conf(5)*.
Breaks validation if those servers do not support DNSSEC.
.. option:: -4
Use solely the IPv4 network for sending packets.
.. option:: -6
Use solely the IPv6 network for sending packets.
Examples
--------
Some examples of use.
The keys shown below are fakes, thus a security failure is encountered.
.. code-block:: text
$ unbound-host www.example.com
$ unbound-host -v -y "example.com DS 31560 5 1 1CFED84787E6E19CCF9372C1187325972FE546CD" www.example.com
$ unbound-host -v -y "example.com DS 31560 5 1 1CFED84787E6E19CCF9372C1187325972FE546CD" 192.0.2.153
Exit Code
---------
The ``unbound-host`` program exits with status code 1 on error, 0 on no error.
The data may not be available on exit code 0, exit code 1 means the lookup
encountered a fatal error.
See Also
--------
:doc:`unbound.conf(5)</manpages/unbound.conf>`,
:doc:`unbound(8)</manpages/unbound>`.
+101 -66
View File
@@ -1,88 +1,123 @@
.TH "unbound" "8" "@date@" "NLnet Labs" "unbound @version@"
.\"
.\" unbound.8 -- unbound manual
.\"
.\" Copyright (c) 2007, NLnet Labs. All rights reserved.
.\"
.\" See LICENSE for the license.
.\"
.\"
.SH "NAME"
.B unbound
\- Unbound DNS validating resolver @version@.
.SH "SYNOPSIS"
.B unbound
.RB [ \-h ]
.RB [ \-d ]
.RB [ \-p ]
.RB [ \-v ]
.RB [ \-c
.IR cfgfile ]
.SH "DESCRIPTION"
.B Unbound
is a caching DNS resolver.
.P
It uses a built in list of authoritative nameservers for the root zone (.),
.\" Man page generated from reStructuredText.
.
.
.nr rst2man-indent-level 0
.
.de1 rstReportMargin
\\$1 \\n[an-margin]
level \\n[rst2man-indent-level]
level margin: \\n[rst2man-indent\\n[rst2man-indent-level]]
-
\\n[rst2man-indent0]
\\n[rst2man-indent1]
\\n[rst2man-indent2]
..
.de1 INDENT
.\" .rstReportMargin pre:
. RS \\$1
. nr rst2man-indent\\n[rst2man-indent-level] \\n[an-margin]
. nr rst2man-indent-level +1
.\" .rstReportMargin post:
..
.de UNINDENT
. RE
.\" indent \\n[an-margin]
.\" old: \\n[rst2man-indent\\n[rst2man-indent-level]]
.nr rst2man-indent-level -1
.\" new: \\n[rst2man-indent\\n[rst2man-indent-level]]
.in \\n[rst2man-indent\\n[rst2man-indent-level]]u
..
.TH "UNBOUND" "8" "@date@" "@version@" "Unbound"
.SH NAME
unbound \- Unbound DNS validating resolver @version@.
.SH SYNOPSIS
.sp
\fBunbound\fP [\fB\-hdpv\fP] [\fB\-c <cfgfile>\fP]
.SH DESCRIPTION
.sp
\fBunbound\fP is a caching DNS resolver.
.sp
It uses a built in list of authoritative nameservers for the root zone (\fB\&.\fP),
the so called root hints.
On receiving a DNS query it will ask the root nameservers for
an answer and will in almost all cases receive a delegation to a top level
domain (TLD) authoritative nameserver.
On receiving a DNS query it will ask the root nameservers for an answer and
will in almost all cases receive a delegation to a top level domain (TLD)
authoritative nameserver.
It will then ask that nameserver for an answer.
It will recursively continue until an answer is found or no answer is
available (NXDOMAIN).
For performance and efficiency reasons that answer is cached for a
certain time (the answer's time\-to\-live or TTL).
It will recursively continue until an answer is found or no answer is available
(NXDOMAIN).
For performance and efficiency reasons that answer is cached for a certain time
(the answer\(aqs time\-to\-live or TTL).
A second query for the same name will then be answered from the cache.
Unbound can also do DNSSEC validation.
.P
To use a locally running
.B Unbound
for resolving put
.sp
.RS 6n
To use a locally running Unbound for resolving put:
.INDENT 0.0
.INDENT 3.5
.sp
.nf
.ft C
nameserver 127.0.0.1
.RE
.ft P
.fi
.UNINDENT
.UNINDENT
.sp
into \fIresolv.conf(5)\fP\&.
.sp
If authoritative DNS is needed as well using \fI\%nsd(8)\fP,
careful setup is required because authoritative nameservers and resolvers are
using the same port number (53).
.sp
into
.IR resolv.conf (5).
.P
If authoritative DNS is needed as well using
.IR nsd (8),
careful setup is required because authoritative nameservers and
resolvers are using the same port number (53).
.P
The available options are:
.INDENT 0.0
.TP
.B \-h
Show the version number and commandline option help, and exit.
.UNINDENT
.INDENT 0.0
.TP
.B \-c\fI cfgfile
Set the config file with settings for Unbound to read instead of reading the
file at the default location, @ub_conf_file@. The syntax is
described in \fIunbound.conf\fR(5).
.B \-c <cfgfile>
Set the config file with settings for unbound to read instead of reading the
file at the default location, \fB@ub_conf_file@\fP\&.
The syntax is described in \fI\%unbound.conf(5)\fP\&.
.UNINDENT
.INDENT 0.0
.TP
.B \-d
Debug flag: do not fork into the background, but stay attached to
the console. This flag will also delay writing to the log file until
the thread\-spawn time, so that most config and setup errors appear on
stderr. If given twice or more, logging does not switch to the log file
or to syslog, but the log messages are printed to stderr all the time.
Debug flag: do not fork into the background, but stay attached to the
console.
This flag will also delay writing to the log file until the thread\-spawn
time, so that most config and setup errors appear on stderr.
If given twice or more, logging does not switch to the log file or to
syslog, but the log messages are printed to stderr all the time.
.UNINDENT
.INDENT 0.0
.TP
.B \-p
Don't use a pidfile. This argument should only be used by supervision
systems which can ensure that only one instance of Unbound will run
concurrently.
Don\(aqt use a pidfile.
This argument should only be used by supervision systems which can ensure
that only one instance of Unbound will run concurrently.
.UNINDENT
.INDENT 0.0
.TP
.B \-v
Increase verbosity. If given multiple times, more information is logged.
Increase verbosity.
If given multiple times, more information is logged.
This is in addition to the verbosity (if any) from the config file.
.UNINDENT
.INDENT 0.0
.TP
.B \-V
Show the version number and build options, and exit.
.SH "SEE ALSO"
\fIunbound.conf\fR(5),
\fIunbound\-checkconf\fR(8),
\fInsd\fR(8).
.SH "AUTHORS"
.B Unbound
developers are mentioned in the CREDITS file in the distribution.
.UNINDENT
.SH SEE ALSO
.sp
\fI\%unbound.conf(5)\fP,
\fI\%unbound\-checkconf(8)\fP,
\fI\%nsd(8)\fP\&.
.SH AUTHOR
Unbound developers are mentioned in the CREDITS file in the distribution.
.SH COPYRIGHT
1999-2025, NLnet Labs
.\" Generated by docutils manpage writer.
.
+5191 -2196
View File
File diff suppressed because it is too large Load Diff
+5005
View File
File diff suppressed because it is too large Load Diff
+2007 -886
View File
File diff suppressed because it is too large Load Diff
+119
View File
@@ -0,0 +1,119 @@
..
WHEN EDITING MAKE SURE EACH SENTENCE STARTS ON A NEW LINE
..
IT HELPS RENDERERS TO DO THE RIGHT THING WRT SPACE
..
IT HELPS PEOPLE DIFFING THE CHANGES
..
WHEN EDITING MAKE SURE EACH SENTENCE STARTS ON A NEW LINE
..
IT HELPS RENDERERS TO DO THE RIGHT THING WRT SPACE
..
IT HELPS PEOPLE DIFFING THE CHANGES
..
WHEN EDITING MAKE SURE EACH SENTENCE STARTS ON A NEW LINE
..
IT HELPS RENDERERS TO DO THE RIGHT THING WRT SPACE
..
IT HELPS PEOPLE DIFFING THE CHANGES
..
WHEN EDITING MAKE SURE EACH SENTENCE STARTS ON A NEW LINE
..
IT HELPS RENDERERS TO DO THE RIGHT THING WRT SPACE
..
IT HELPS PEOPLE DIFFING THE CHANGES
.. program:: unbound
unbound(8)
==========
Synopsis
--------
**unbound** [``-hdpv``] [``-c <cfgfile>``]
Description
-----------
``unbound`` is a caching DNS resolver.
It uses a built in list of authoritative nameservers for the root zone (``.``),
the so called root hints.
On receiving a DNS query it will ask the root nameservers for an answer and
will in almost all cases receive a delegation to a top level domain (TLD)
authoritative nameserver.
It will then ask that nameserver for an answer.
It will recursively continue until an answer is found or no answer is available
(NXDOMAIN).
For performance and efficiency reasons that answer is cached for a certain time
(the answer's time-to-live or TTL).
A second query for the same name will then be answered from the cache.
Unbound can also do DNSSEC validation.
To use a locally running Unbound for resolving put:
.. code-block:: text
nameserver 127.0.0.1
into *resolv.conf(5)*.
If authoritative DNS is needed as well using :external+nsd:doc:`manpages/nsd`,
careful setup is required because authoritative nameservers and resolvers are
using the same port number (53).
The available options are:
.. option:: -h
Show the version number and commandline option help, and exit.
.. option:: -c <cfgfile>
Set the config file with settings for unbound to read instead of reading the
file at the default location, :file:`@ub_conf_file@`.
The syntax is described in :doc:`unbound.conf(5)</manpages/unbound.conf>`.
.. option:: -d
Debug flag: do not fork into the background, but stay attached to the
console.
This flag will also delay writing to the log file until the thread-spawn
time, so that most config and setup errors appear on stderr.
If given twice or more, logging does not switch to the log file or to
syslog, but the log messages are printed to stderr all the time.
.. option:: -p
Don't use a pidfile.
This argument should only be used by supervision systems which can ensure
that only one instance of Unbound will run concurrently.
.. option:: -v
Increase verbosity.
If given multiple times, more information is logged.
This is in addition to the verbosity (if any) from the config file.
.. option:: -V
Show the version number and build options, and exit.
See Also
--------
:doc:`unbound.conf(5)</manpages/unbound.conf>`,
:doc:`unbound-checkconf(8)</manpages/unbound-checkconf>`,
:external+nsd:doc:`manpages/nsd`.
+4 -3
View File
@@ -75,6 +75,7 @@ int dynlibmod_init(struct module_env* env, int id) {
struct config_strlist* cfg_item = env->cfg->dynlib_file;
struct dynlibmod_env* de = (struct dynlibmod_env*)calloc(1, sizeof(struct dynlibmod_env));
__DYNMOD dynamic_library;
int i;
if (!de)
{
log_err("dynlibmod[%d]: malloc failure", dynlib_mod_idx);
@@ -84,7 +85,7 @@ int dynlibmod_init(struct module_env* env, int id) {
env->modinfo[id] = (void*) de;
de->fname = NULL;
for(int i = dynlib_mod_idx;
for(i = dynlib_mod_idx;
i != 0 && cfg_item != NULL;
i--, cfg_item = cfg_item->next) {}
@@ -296,8 +297,8 @@ inplace_cb_delete_wrapped(struct module_env* env, enum inplace_cb_list_type type
*/
static struct module_func_block dynlibmod_block = {
"dynlib",
&dynlibmod_init, &dynlibmod_deinit, &dynlibmod_operate, &dynlibmod_inform_super,
&dynlibmod_clear, &dynlibmod_get_mem
NULL, NULL, &dynlibmod_init, &dynlibmod_deinit, &dynlibmod_operate,
&dynlibmod_inform_super, &dynlibmod_clear, &dynlibmod_get_mem
};
struct module_func_block* dynlibmod_get_funcblock(void)
+7 -2
View File
@@ -97,6 +97,7 @@ node_create(struct addrtree *tree, void *elem, addrlen_t scope,
tree->node_count++;
node->scope = scope;
node->ttl = ttl;
node->only_match_scope_zero = 0;
node->edge[0] = NULL;
node->edge[1] = NULL;
node->parent_edge = NULL;
@@ -155,6 +156,7 @@ clean_node(struct addrtree *tree, struct addrnode *node)
if (!node->elem) return;
tree->size_bytes -= tree->sizefunc(node->elem);
tree->delfunc(tree->env, node->elem);
node->only_match_scope_zero = 0;
node->elem = NULL;
}
@@ -358,7 +360,7 @@ issub(const addrkey_t *s1, addrlen_t l1,
void
addrtree_insert(struct addrtree *tree, const addrkey_t *addr,
addrlen_t sourcemask, addrlen_t scope, void *elem, time_t ttl,
time_t now)
time_t now, int only_match_scope_zero)
{
struct addrnode *newnode, *node;
struct addredge *edge;
@@ -381,6 +383,7 @@ addrtree_insert(struct addrtree *tree, const addrkey_t *addr,
/* update this node's scope and data */
clean_node(tree, node);
node->ttl = ttl;
node->only_match_scope_zero = only_match_scope_zero;
node->elem = elem;
node->scope = scope;
tree->size_bytes += tree->sizefunc(elem);
@@ -447,6 +450,7 @@ addrtree_insert(struct addrtree *tree, const addrkey_t *addr,
newnode->elem = elem;
newnode->scope = scope;
newnode->ttl = ttl;
newnode->only_match_scope_zero = only_match_scope_zero;
}
tree->size_bytes += node_size(tree, newnode);
@@ -483,7 +487,8 @@ addrtree_find(struct addrtree *tree, const addrkey_t *addr,
/* Current node more specific then question. */
log_assert(depth <= sourcemask);
/* does this node have data? if yes, see if we have a match */
if (node->elem && node->ttl >= now) {
if (node->elem && node->ttl >= now &&
!(sourcemask != 0 && node->only_match_scope_zero)) {
/* saved at wrong depth */;
log_assert(node->scope >= depth);
if (depth == node->scope ||
+7 -2
View File
@@ -95,6 +95,10 @@ struct addrnode {
time_t ttl;
/** Number of significant bits in address. */
addrlen_t scope;
/** Only use the element for queries for subnet/0. Set if the query
* for /0 was answered with scope 0. For query /x answer scope 0,
* they can match anything and this is false. */
int only_match_scope_zero;
/** A node can have 0-2 edges, set to NULL for unused */
struct addredge *edge[2];
/** edge between this node and parent */
@@ -112,7 +116,7 @@ struct addredge {
addrlen_t len;
/** child node this edge is connected to */
struct addrnode *node;
/** Parent node this ege is connected to */
/** Parent node this edge is connected to */
struct addrnode *parent_node;
/** Index of this edge in parent_node */
int parent_index;
@@ -157,11 +161,12 @@ void addrtree_delete(struct addrtree *tree);
* @param scope: Number of significant bits in addr.
* @param elem: data to store in the tree.
* @param ttl: elem is valid up to this time, seconds.
* @param only_match_scope_zero: set for when query /0 has scope /0 answer.
* @param now: Current time in seconds.
*/
void addrtree_insert(struct addrtree *tree, const addrkey_t *addr,
addrlen_t sourcemask, addrlen_t scope, void *elem, time_t ttl,
time_t now);
time_t now, int only_match_scope_zero);
/**
* Find a node containing an element in the tree.
+347 -59
View File
@@ -51,11 +51,16 @@
#include "services/cache/dns.h"
#include "util/module.h"
#include "util/regional.h"
#include "util/fptr_wlist.h"
#include "util/storage/slabhash.h"
#include "util/config_file.h"
#include "util/data/msgreply.h"
#include "sldns/sbuffer.h"
#include "sldns/wire2str.h"
#include "iterator/iter_utils.h"
#ifdef USE_CACHEDB
#include "cachedb/cachedb.h"
#endif
/** externally called */
void
@@ -93,13 +98,14 @@ subnet_new_qstate(struct module_qstate *qstate, int id)
qstate->minfo[id] = sq;
memset(sq, 0, sizeof(*sq));
sq->started_no_cache_store = qstate->no_cache_store;
sq->started_no_cache_lookup = qstate->no_cache_lookup;
return 1;
}
/** Add ecs struct to edns list, after parsing it to wire format. */
static void
ecs_opt_list_append(struct ecs_data* ecs, struct edns_option** list,
struct module_qstate *qstate)
void
subnet_ecs_opt_list_append(struct ecs_data* ecs, struct edns_option** list,
struct module_qstate *qstate, struct regional *region)
{
size_t sn_octs, sn_octs_remainder;
sldns_buffer* buf = qstate->env->scratch_buffer;
@@ -131,7 +137,7 @@ ecs_opt_list_append(struct ecs_data* ecs, struct edns_option** list,
edns_opt_list_append(list,
qstate->env->cfg->client_subnet_opcode,
sn_octs + sn_octs_remainder + 4,
sldns_buffer_begin(buf), qstate->region);
sldns_buffer_begin(buf), region);
}
}
@@ -139,7 +145,7 @@ int ecs_whitelist_check(struct query_info* qinfo,
uint16_t ATTR_UNUSED(flags), struct module_qstate* qstate,
struct sockaddr_storage* addr, socklen_t addrlen,
uint8_t* ATTR_UNUSED(zone), size_t ATTR_UNUSED(zonelen),
struct regional* ATTR_UNUSED(region), int id, void* ATTR_UNUSED(cbargs))
struct regional *region, int id, void* ATTR_UNUSED(cbargs))
{
struct subnet_qstate *sq;
struct subnet_env *sn_env;
@@ -148,12 +154,29 @@ int ecs_whitelist_check(struct query_info* qinfo,
return 1;
sn_env = (struct subnet_env*)qstate->env->modinfo[id];
if(sq->is_subquery_nonsubnet) {
if(sq->is_subquery_scopezero) {
/* Check if the result can be stored in the global cache,
* this is okay if the address and name are not configured
* as subnet address and subnet zone. */
if(!ecs_is_whitelisted(sn_env->whitelist,
addr, addrlen, qinfo->qname, qinfo->qname_len,
qinfo->qclass)) {
verbose(VERB_ALGO, "subnet store subquery global, name and addr have no subnet treatment.");
qstate->no_cache_store = 0;
}
}
return 1;
}
/* Cache by default, might be disabled after parsing EDNS option
* received from nameserver. */
if(!iter_stub_fwd_no_cache(qstate, &qstate->qinfo, NULL, NULL)) {
if(!iter_stub_fwd_no_cache(qstate, &qstate->qinfo, NULL, NULL, NULL, 0)
&& sq->ecs_client_in.subnet_validdata) {
qstate->no_cache_store = 0;
}
sq->subnet_sent_no_subnet = 0;
if(sq->ecs_server_out.subnet_validdata && ((sq->subnet_downstream &&
qstate->env->cfg->client_subnet_always_forward) ||
ecs_is_whitelisted(sn_env->whitelist,
@@ -162,17 +185,29 @@ int ecs_whitelist_check(struct query_info* qinfo,
/* Address on whitelist or client query contains ECS option, we
* want to sent out ECS. Only add option if it is not already
* set. */
if(!(sq->subnet_sent)) {
ecs_opt_list_append(&sq->ecs_server_out,
&qstate->edns_opts_back_out, qstate);
sq->subnet_sent = 1;
if(!edns_opt_list_find(qstate->edns_opts_back_out,
qstate->env->cfg->client_subnet_opcode)) {
/* if the client is not wanting an EDNS subnet option,
* omit it and store that we omitted it but actually
* are doing EDNS subnet to the server. */
if(sq->ecs_server_out.subnet_source_mask == 0) {
sq->subnet_sent_no_subnet = 1;
sq->subnet_sent = 0;
return 1;
}
subnet_ecs_opt_list_append(&sq->ecs_server_out,
&qstate->edns_opts_back_out, qstate, region);
}
sq->subnet_sent = 1;
}
else if(sq->subnet_sent) {
else {
/* Outgoing ECS option is set, but we don't want to sent it to
* this address, remove option. */
edns_opt_list_remove(&qstate->edns_opts_back_out,
qstate->env->cfg->client_subnet_opcode);
if(edns_opt_list_find(qstate->edns_opts_back_out,
qstate->env->cfg->client_subnet_opcode)) {
edns_opt_list_remove(&qstate->edns_opts_back_out,
qstate->env->cfg->client_subnet_opcode);
}
sq->subnet_sent = 0;
}
return 1;
@@ -198,18 +233,29 @@ subnetmod_init(struct module_env *env, int id)
}
alloc_init(&sn_env->alloc, NULL, 0);
env->modinfo[id] = (void*)sn_env;
/* Warn that serve-expired and prefetch do not work with the subnet
* module cache. */
if(env->cfg->serve_expired)
log_warn(
"subnetcache: serve-expired is set but not working "
"for data originating from the subnet module cache.");
if(env->cfg->prefetch)
log_warn(
"subnetcache: prefetch is set but not working "
"for data originating from the subnet module cache.");
/* Copy msg_cache settings */
sn_env->subnet_msg_cache = slabhash_create(env->cfg->msg_cache_slabs,
HASH_DEFAULT_STARTARRAY, env->cfg->msg_cache_size,
msg_cache_sizefunc, query_info_compare, query_entry_delete,
subnet_data_delete, NULL);
slabhash_setmarkdel(sn_env->subnet_msg_cache, &subnet_markdel);
if(!sn_env->subnet_msg_cache) {
log_err("subnetcache: could not create cache");
free(sn_env);
env->modinfo[id] = NULL;
return 0;
}
slabhash_setmarkdel(sn_env->subnet_msg_cache, &subnet_markdel);
/* whitelist for edns subnet capable servers */
sn_env->whitelist = ecs_whitelist_create();
if(!sn_env->whitelist ||
@@ -227,7 +273,7 @@ subnetmod_init(struct module_env *env, int id)
env->unique_mesh = 1;
if(!edns_register_option(env->cfg->client_subnet_opcode,
env->cfg->client_subnet_always_forward /* bypass cache */,
0 /* no aggregation */, env)) {
1 /* no aggregation */, env)) {
log_err("subnetcache: could not register opcode");
ecs_whitelist_delete(sn_env->whitelist);
slabhash_delete(sn_env->subnet_msg_cache);
@@ -284,9 +330,18 @@ delfunc(void *envptr, void *elemptr) {
static size_t
sizefunc(void *elemptr) {
struct reply_info *elem = (struct reply_info *)elemptr;
return sizeof (struct reply_info) - sizeof (struct rrset_ref)
size_t s = sizeof (struct reply_info) - sizeof (struct rrset_ref)
+ elem->rrset_count * sizeof (struct rrset_ref)
+ elem->rrset_count * sizeof (struct ub_packed_rrset_key *);
size_t i;
for (i = 0; i < elem->rrset_count; i++) {
struct ub_packed_rrset_key *key = elem->rrsets[i];
struct packed_rrset_data *data = key->entry.data;
s += ub_rrset_sizefunc(key, data);
}
if(elem->reason_bogus_str)
s += strlen(elem->reason_bogus_str)+1;
return s;
}
/**
@@ -326,13 +381,16 @@ update_cache(struct module_qstate *qstate, int id)
struct slabhash *subnet_msg_cache = sne->subnet_msg_cache;
struct ecs_data *edns = &sq->ecs_client_in;
size_t i;
int only_match_scope_zero, diff_size;
/* We already calculated hash upon lookup */
hashvalue_type h = qstate->minfo[id] ?
((struct subnet_qstate*)qstate->minfo[id])->qinfo_hash :
/* We already calculated hash upon lookup (lookup_and_reply) if we were
* allowed to look in the ECS cache */
hashvalue_type h = qstate->minfo[id] &&
((struct subnet_qstate*)qstate->minfo[id])->qinfo_hash_calculated?
((struct subnet_qstate*)qstate->minfo[id])->qinfo_hash :
query_info_hash(&qstate->qinfo, qstate->query_flags);
/* Step 1, general qinfo lookup */
struct lruhash_entry *lru_entry = slabhash_lookup(subnet_msg_cache, h,
struct lruhash_entry* lru_entry = slabhash_lookup(subnet_msg_cache, h,
&qstate->qinfo, 1);
int need_to_insert = (lru_entry == NULL);
if (!lru_entry) {
@@ -376,29 +434,38 @@ update_cache(struct module_qstate *qstate, int id)
log_err("subnetcache: cache insertion failed");
return;
}
/* store RRsets */
for(i=0; i<rep->rrset_count; i++) {
rep->ref[i].key = rep->rrsets[i];
rep->ref[i].id = rep->rrsets[i]->id;
}
reply_info_set_ttls(rep, *qstate->env->now);
reply_info_sortref(rep);
rep->flags |= (BIT_RA | BIT_QR); /* fix flags to be sensible for */
rep->flags &= ~(BIT_AA | BIT_CD);/* a reply based on the cache */
if(edns->subnet_source_mask == 0 && edns->subnet_scope_mask == 0)
only_match_scope_zero = 1;
else only_match_scope_zero = 0;
diff_size = (int)tree->size_bytes;
addrtree_insert(tree, (addrkey_t*)edns->subnet_addr,
edns->subnet_source_mask, sq->max_scope, rep,
rep->ttl, *qstate->env->now);
rep->ttl, *qstate->env->now, only_match_scope_zero);
diff_size = (int)tree->size_bytes - diff_size;
lock_rw_unlock(&lru_entry->lock);
if (need_to_insert) {
slabhash_insert(subnet_msg_cache, h, lru_entry, lru_entry->data,
NULL);
} else {
slabhash_update_space_used(subnet_msg_cache, h, NULL,
diff_size);
}
}
/** Lookup in cache and reply true iff reply is sent. */
static int
lookup_and_reply(struct module_qstate *qstate, int id, struct subnet_qstate *sq)
lookup_and_reply(struct module_qstate *qstate, int id, struct subnet_qstate *sq, int prefetch)
{
struct lruhash_entry *e;
struct module_env *env = qstate->env;
@@ -412,7 +479,10 @@ lookup_and_reply(struct module_qstate *qstate, int id, struct subnet_qstate *sq)
memset(&sq->ecs_client_out, 0, sizeof(sq->ecs_client_out));
if (sq) sq->qinfo_hash = h; /* Might be useful on cache miss */
if (sq) {
sq->qinfo_hash = h; /* Might be useful on cache miss */
sq->qinfo_hash_calculated = 1;
}
e = slabhash_lookup(sne->subnet_msg_cache, h, &qstate->qinfo, 1);
if (!e) return 0; /* qinfo not in cache */
data = e->data;
@@ -447,6 +517,10 @@ lookup_and_reply(struct module_qstate *qstate, int id, struct subnet_qstate *sq)
INET6_SIZE);
sq->ecs_client_out.subnet_validdata = 1;
}
if (prefetch && *qstate->env->now >= ((struct reply_info *)node->elem)->prefetch_ttl) {
qstate->need_refetch = 1;
}
return 1;
}
@@ -465,6 +539,83 @@ common_prefix(uint8_t *a, uint8_t *b, uint8_t net)
return !memcmp(a, b, n) && ((net % 8) == 0 || a[n] == b[n]);
}
/**
* Create sub request that looks up the query.
* @param qstate: query state
* @param id: module id.
* @param sq: subnet qstate
* @return false on failure.
*/
static int
generate_sub_request(struct module_qstate *qstate, int id, struct subnet_qstate* sq)
{
struct module_qstate* subq = NULL;
uint16_t qflags = 0; /* OPCODE QUERY, no flags */
int prime = 0;
int valrec = 0;
struct query_info qinf;
qinf.qname = qstate->qinfo.qname;
qinf.qname_len = qstate->qinfo.qname_len;
qinf.qtype = qstate->qinfo.qtype;
qinf.qclass = qstate->qinfo.qclass;
qinf.local_alias = NULL;
qflags |= BIT_RD;
if((qstate->query_flags & BIT_CD)!=0) {
qflags |= BIT_CD;
valrec = 1;
}
fptr_ok(fptr_whitelist_modenv_attach_sub(qstate->env->attach_sub));
if(!(*qstate->env->attach_sub)(qstate, &qinf, qflags, prime, valrec,
&subq)) {
return 0;
}
if(subq) {
/* It is possible to access the subquery module state. */
struct subnet_qstate* subsq;
if(!subnet_new_qstate(subq, id)) {
verbose(VERB_ALGO, "Could not allocate new subnet qstate");
return 0;
}
subsq = (struct subnet_qstate*)subq->minfo[id];
subsq->is_subquery_nonsubnet = 1;
/* When the client asks 0.0.0.0/0 and the name is not treated
* as subnet, it is to be stored in the global cache.
* Store that the client asked for that, if so. */
if(sq->ecs_client_in.subnet_source_mask == 0 &&
edns_opt_list_find(qstate->edns_opts_front_in,
qstate->env->cfg->client_subnet_opcode)) {
subq->no_cache_store = 1;
subsq->is_subquery_scopezero = 1;
}
}
return 1;
}
/**
* Perform the query without subnet
* @param qstate: query state
* @param id: module id.
* @param sq: subnet qstate
* @return module state
*/
static enum module_ext_state
generate_lookup_without_subnet(struct module_qstate *qstate, int id,
struct subnet_qstate* sq)
{
verbose(VERB_ALGO, "subnetcache: make subquery to look up without subnet");
if(!generate_sub_request(qstate, id, sq)) {
verbose(VERB_ALGO, "Could not generate sub query");
qstate->return_rcode = LDNS_RCODE_SERVFAIL;
qstate->return_msg = NULL;
return module_finished;
}
sq->wait_subquery = 1;
return module_wait_subquery;
}
static enum module_ext_state
eval_response(struct module_qstate *qstate, int id, struct subnet_qstate *sq)
{
@@ -483,50 +634,56 @@ eval_response(struct module_qstate *qstate, int id, struct subnet_qstate *sq)
* module_finished */
return module_finished;
}
/* We have not asked for subnet data */
if (!sq->subnet_sent) {
if (!sq->subnet_sent && !sq->subnet_sent_no_subnet) {
if (s_in->subnet_validdata)
verbose(VERB_QUERY, "subnetcache: received spurious data");
if (sq->subnet_downstream) /* Copy back to client */
cp_edns_bad_response(c_out, c_in);
return module_finished;
}
/* subnet sent but nothing came back */
if (!s_in->subnet_validdata) {
if (!s_in->subnet_validdata && !sq->subnet_sent_no_subnet) {
/* The authority indicated no support for edns subnet. As a
* consequence the answer ended up in the regular cache. It
* is still useful to put it in the edns subnet cache for
* when a client explicitly asks for subnet specific answer. */
verbose(VERB_QUERY, "subnetcache: Authority indicates no support");
if(!sq->started_no_cache_store) {
lock_rw_wrlock(&sne->biglock);
update_cache(qstate, id);
lock_rw_unlock(&sne->biglock);
}
if (sq->subnet_downstream)
cp_edns_bad_response(c_out, c_in);
return module_finished;
return generate_lookup_without_subnet(qstate, id, sq);
}
/* Purposefully there was no sent subnet, and there is consequently
* no subnet in the answer. If there was, use the subnet in the answer
* anyway. But if there is not, treat it as a prefix 0 answer. */
if(sq->subnet_sent_no_subnet && !s_in->subnet_validdata) {
/* Fill in 0.0.0.0/0 scope 0, or ::0/0 scope 0, for caching. */
s_in->subnet_addr_fam = s_out->subnet_addr_fam;
s_in->subnet_source_mask = 0;
s_in->subnet_scope_mask = 0;
memset(s_in->subnet_addr, 0, INET6_SIZE);
s_in->subnet_validdata = 1;
}
/* Being here means we have asked for and got a subnet specific
* answer. Also, the answer from the authority is not yet cached
* anywhere. */
/* can we accept response? */
if(s_out->subnet_addr_fam != s_in->subnet_addr_fam ||
s_out->subnet_source_mask != s_in->subnet_source_mask ||
!common_prefix(s_out->subnet_addr, s_in->subnet_addr,
s_out->subnet_source_mask))
{
/* we can not accept, restart query without option */
/* we can not accept, perform query without option */
verbose(VERB_QUERY, "subnetcache: forged data");
s_out->subnet_validdata = 0;
(void)edns_opt_list_remove(&qstate->edns_opts_back_out,
qstate->env->cfg->client_subnet_opcode);
sq->subnet_sent = 0;
return module_restart_next;
sq->subnet_sent_no_subnet = 0;
return generate_lookup_without_subnet(qstate, id, sq);
}
lock_rw_wrlock(&sne->biglock);
@@ -535,7 +692,21 @@ eval_response(struct module_qstate *qstate, int id, struct subnet_qstate *sq)
}
sne->num_msg_nocache++;
lock_rw_unlock(&sne->biglock);
/* If there is an expired answer in the global cache, remove that,
* because expired answers would otherwise resurface once the ecs data
* expires, giving once in a while global data responses for ecs
* domains, with serve expired enabled. */
if(qstate->env->cfg->serve_expired) {
msg_cache_remove(qstate->env, qstate->qinfo.qname,
qstate->qinfo.qname_len, qstate->qinfo.qtype,
qstate->qinfo.qclass, 0);
#ifdef USE_CACHEDB
if(qstate->env->cachedb_enabled)
cachedb_msg_remove(qstate);
#endif
}
if (sq->subnet_downstream) {
/* Client wants to see the answer, echo option back
* and adjust the scope. */
@@ -598,7 +769,7 @@ parse_subnet_option(struct edns_option* ecs_option, struct ecs_data* ecs)
return 1;
}
static void
void
subnet_option_from_ss(struct sockaddr_storage *ss, struct ecs_data* ecs,
struct config_file* cfg)
{
@@ -646,6 +817,7 @@ ecs_query_response(struct module_qstate* qstate, struct dns_msg* response,
edns_opt_list_remove(&qstate->edns_opts_back_out,
qstate->env->cfg->client_subnet_opcode);
sq->subnet_sent = 0;
sq->subnet_sent_no_subnet = 0;
memset(&sq->ecs_server_out, 0, sizeof(sq->ecs_server_out));
} else if (!sq->track_max_scope &&
FLAGS_GET_RCODE(response->rep->flags) == LDNS_RCODE_NOERROR &&
@@ -664,6 +836,24 @@ ecs_query_response(struct module_qstate* qstate, struct dns_msg* response,
return 1;
}
/** verbose print edns subnet option in pretty print */
static void
subnet_log_print(const char* s, struct edns_option* ecs_opt)
{
if(verbosity >= VERB_ALGO) {
char buf[256];
char* str = buf;
size_t str_len = sizeof(buf);
if(!ecs_opt) {
verbose(VERB_ALGO, "%s (null)", s);
return;
}
(void)sldns_wire2str_edns_subnet_print(&str, &str_len,
ecs_opt->opt_data, ecs_opt->opt_len);
verbose(VERB_ALGO, "%s %s", s, buf);
}
}
int
ecs_edns_back_parsed(struct module_qstate* qstate, int id,
void* ATTR_UNUSED(cbargs))
@@ -678,6 +868,7 @@ ecs_edns_back_parsed(struct module_qstate* qstate, int id,
qstate->env->cfg->client_subnet_opcode)) &&
parse_subnet_option(ecs_opt, &sq->ecs_server_in) &&
sq->subnet_sent && sq->ecs_server_in.subnet_validdata) {
subnet_log_print("answer has edns subnet", ecs_opt);
/* Only skip global cache store if we sent an ECS option
* and received one back. Answers from non-whitelisted
* servers will end up in global cache. Answers for
@@ -688,6 +879,12 @@ ecs_edns_back_parsed(struct module_qstate* qstate, int id,
sq->ecs_server_in.subnet_scope_mask >
sq->max_scope))
sq->max_scope = sq->ecs_server_in.subnet_scope_mask;
} else if(sq->subnet_sent_no_subnet) {
/* The answer can be stored as scope 0, not in global cache. */
qstate->no_cache_store = 1;
} else if(sq->subnet_sent) {
/* Need another query to be able to store in global cache. */
qstate->no_cache_store = 1;
}
return 1;
@@ -705,6 +902,32 @@ subnetmod_operate(struct module_qstate *qstate, enum module_ev event,
strmodulevent(event));
log_query_info(VERB_QUERY, "subnetcache operate: query", &qstate->qinfo);
if(sq && sq->wait_subquery_done) {
/* The subquery lookup returned. */
if(sq->ecs_client_in.subnet_source_mask == 0 &&
edns_opt_list_find(qstate->edns_opts_front_in,
qstate->env->cfg->client_subnet_opcode)) {
if(!sq->started_no_cache_store &&
qstate->return_msg) {
lock_rw_wrlock(&sne->biglock);
update_cache(qstate, id);
lock_rw_unlock(&sne->biglock);
}
if (sq->subnet_downstream)
cp_edns_bad_response(&sq->ecs_client_out,
&sq->ecs_client_in);
/* It is a scope zero lookup, append edns subnet
* option to the querier. */
subnet_ecs_opt_list_append(&sq->ecs_client_out,
&qstate->edns_opts_front_out, qstate,
qstate->region);
}
sq->wait_subquery_done = 0;
qstate->ext_state[id] = module_finished;
qstate->no_cache_store = sq->started_no_cache_store;
qstate->no_cache_lookup = sq->started_no_cache_lookup;
return;
}
if((event == module_event_new || event == module_event_pass) &&
sq == NULL) {
struct edns_option* ecs_opt;
@@ -715,6 +938,8 @@ subnetmod_operate(struct module_qstate *qstate, enum module_ev event,
}
sq = (struct subnet_qstate*)qstate->minfo[id];
if(sq->wait_subquery)
return; /* Wait for that subquery to return */
if((ecs_opt = edns_opt_list_find(
qstate->edns_opts_front_in,
@@ -726,11 +951,17 @@ subnetmod_operate(struct module_qstate *qstate, enum module_ev event,
qstate->ext_state[id] = module_finished;
return;
}
subnet_log_print("query has edns subnet", ecs_opt);
sq->subnet_downstream = 1;
}
else if(qstate->mesh_info->reply_list) {
subnet_option_from_ss(
&qstate->mesh_info->reply_list->query_reply.addr,
&qstate->mesh_info->reply_list->query_reply.client_addr,
&sq->ecs_client_in, qstate->env->cfg);
}
else if(qstate->client_addr.ss_family != AF_UNSPEC) {
subnet_option_from_ss(
&qstate->client_addr,
&sq->ecs_client_in, qstate->env->cfg);
}
@@ -738,6 +969,14 @@ subnetmod_operate(struct module_qstate *qstate, enum module_ev event,
/* No clients are interested in result or we could not
* parse it, we don't do client subnet */
sq->ecs_server_out.subnet_validdata = 0;
if(edns_opt_list_find(qstate->edns_opts_front_in,
qstate->env->cfg->client_subnet_opcode)) {
/* aggregated this deaggregated state */
qstate->ext_state[id] =
generate_lookup_without_subnet(
qstate, id, sq);
return;
}
verbose(VERB_ALGO, "subnetcache: pass to next module");
qstate->ext_state[id] = module_wait_module;
return;
@@ -754,18 +993,38 @@ subnetmod_operate(struct module_qstate *qstate, enum module_ev event,
return;
}
lock_rw_wrlock(&sne->biglock);
if (lookup_and_reply(qstate, id, sq)) {
sne->num_msg_cache++;
lock_rw_unlock(&sne->biglock);
verbose(VERB_QUERY, "subnetcache: answered from cache");
qstate->ext_state[id] = module_finished;
if(!sq->started_no_cache_lookup && !qstate->blacklist) {
lock_rw_wrlock(&sne->biglock);
if(qstate->mesh_info->reply_list &&
lookup_and_reply(qstate, id, sq,
qstate->env->cfg->prefetch)) {
sne->num_msg_cache++;
lock_rw_unlock(&sne->biglock);
verbose(VERB_QUERY, "subnetcache: answered from cache");
qstate->ext_state[id] = module_finished;
ecs_opt_list_append(&sq->ecs_client_out,
&qstate->edns_opts_front_out, qstate);
subnet_ecs_opt_list_append(&sq->ecs_client_out,
&qstate->edns_opts_front_out, qstate,
qstate->region);
if(verbosity >= VERB_ALGO) {
subnet_log_print("reply has edns subnet",
edns_opt_list_find(
qstate->edns_opts_front_out,
qstate->env->cfg->
client_subnet_opcode));
}
return;
}
lock_rw_unlock(&sne->biglock);
}
if(sq->ecs_client_in.subnet_source_mask == 0 &&
edns_opt_list_find(qstate->edns_opts_front_in,
qstate->env->cfg->client_subnet_opcode)) {
/* client asked for resolution without edns subnet */
qstate->ext_state[id] = generate_lookup_without_subnet(
qstate, id, sq);
return;
}
lock_rw_unlock(&sne->biglock);
sq->ecs_server_out.subnet_addr_fam =
sq->ecs_client_in.subnet_addr_fam;
@@ -783,7 +1042,7 @@ subnetmod_operate(struct module_qstate *qstate, enum module_ev event,
sq->ecs_server_out.subnet_source_mask =
qstate->env->cfg->max_client_subnet_ipv6;
/* Safe to copy completely, even if the source is limited by the
* configuration. ecs_opt_list_append() will limit the address.
* configuration. subnet_ecs_opt_list_append() will limit the address.
* */
memcpy(&sq->ecs_server_out.subnet_addr,
sq->ecs_client_in.subnet_addr, INET6_SIZE);
@@ -802,15 +1061,26 @@ subnetmod_operate(struct module_qstate *qstate, enum module_ev event,
qstate->ext_state[id] = module_wait_module;
return;
}
if(sq && sq->wait_subquery)
return; /* Wait for that subquery to return */
/* Query handed back by next module, we have a 'final' answer */
if(sq && event == module_event_moddone) {
qstate->ext_state[id] = eval_response(qstate, id, sq);
if(qstate->ext_state[id] == module_finished &&
qstate->return_msg) {
ecs_opt_list_append(&sq->ecs_client_out,
&qstate->edns_opts_front_out, qstate);
subnet_ecs_opt_list_append(&sq->ecs_client_out,
&qstate->edns_opts_front_out, qstate,
qstate->region);
if(verbosity >= VERB_ALGO) {
subnet_log_print("reply has edns subnet",
edns_opt_list_find(
qstate->edns_opts_front_out,
qstate->env->cfg->
client_subnet_opcode));
}
}
qstate->no_cache_store = sq->started_no_cache_store;
qstate->no_cache_lookup = sq->started_no_cache_lookup;
return;
}
if(sq && outbound) {
@@ -841,10 +1111,27 @@ subnetmod_clear(struct module_qstate *ATTR_UNUSED(qstate),
}
void
subnetmod_inform_super(struct module_qstate *ATTR_UNUSED(qstate),
int ATTR_UNUSED(id), struct module_qstate *ATTR_UNUSED(super))
subnetmod_inform_super(struct module_qstate *qstate, int id,
struct module_qstate *super)
{
/* Not used */
struct subnet_qstate* super_sq =
(struct subnet_qstate*)super->minfo[id];
log_query_info(VERB_ALGO, "subnetcache inform_super: query",
&super->qinfo);
super_sq->wait_subquery = 0;
super_sq->wait_subquery_done = 1;
if(qstate->return_rcode != LDNS_RCODE_NOERROR ||
!qstate->return_msg) {
super->return_msg = NULL;
super->return_rcode = LDNS_RCODE_SERVFAIL;
return;
}
super->return_rcode = LDNS_RCODE_NOERROR;
super->return_msg = dns_copy_msg(qstate->return_msg, super->region);
if(!super->return_msg) {
log_err("subnetcache: copy response, out of memory");
super->return_rcode = LDNS_RCODE_SERVFAIL;
}
}
size_t
@@ -861,7 +1148,8 @@ subnetmod_get_mem(struct module_env *env, int id)
* The module function block
*/
static struct module_func_block subnetmod_block = {
"subnetcache", &subnetmod_init, &subnetmod_deinit, &subnetmod_operate,
"subnetcache",
NULL, NULL, &subnetmod_init, &subnetmod_deinit, &subnetmod_operate,
&subnetmod_inform_super, &subnetmod_clear, &subnetmod_get_mem
};
+25
View File
@@ -76,6 +76,7 @@ struct subnet_msg_cache_data {
struct subnet_qstate {
/** We need the hash for both cache lookup and insert */
hashvalue_type qinfo_hash;
int qinfo_hash_calculated;
/** ecs_data for client communication */
struct ecs_data ecs_client_in;
struct ecs_data ecs_client_out;
@@ -84,6 +85,13 @@ struct subnet_qstate {
struct ecs_data ecs_server_out;
int subnet_downstream;
int subnet_sent;
/**
* If there was no subnet sent because the client used source prefix
* length 0 for omitting the information. Then the answer is cached
* like subnet was a /0 scope. Like the subnet_sent flag, but when
* the EDNS subnet option is omitted because the client asked.
*/
int subnet_sent_no_subnet;
/** keep track of longest received scope, set after receiving CNAME for
* incoming QNAME. */
int track_max_scope;
@@ -92,6 +100,16 @@ struct subnet_qstate {
uint8_t max_scope;
/** has the subnet module been started with no_cache_store? */
int started_no_cache_store;
/** has the subnet module been started with no_cache_lookup? */
int started_no_cache_lookup;
/** Wait for subquery that has been started for nonsubnet lookup. */
int wait_subquery;
/** The subquery waited for is done. */
int wait_subquery_done;
/** The subnet state is a subquery state for nonsubnet lookup. */
int is_subquery_nonsubnet;
/** This is a subquery, and it is made due to a scope zero request. */
int is_subquery_scopezero;
};
void subnet_data_delete(void* d, void* ATTR_UNUSED(arg));
@@ -143,4 +161,11 @@ int ecs_query_response(struct module_qstate* qstate, struct dns_msg* response,
/** mark subnet msg to be deleted */
void subnet_markdel(void* key);
/** Add ecs struct to edns list, after parsing it to wire format. */
void subnet_ecs_opt_list_append(struct ecs_data* ecs, struct edns_option** list,
struct module_qstate *qstate, struct regional *region);
/** Create ecs_data from the sockaddr_storage information. */
void subnet_option_from_ss(struct sockaddr_storage *ss, struct ecs_data* ecs,
struct config_file* cfg);
#endif /* SUBNETMOD_H */
+3 -2
View File
@@ -456,7 +456,8 @@ ipsecmod_handle_query(struct module_qstate* qstate,
/* Store A/AAAA in cache. */
if(!dns_cache_store(qstate->env, &qstate->qinfo,
qstate->return_msg->rep, 0, qstate->prefetch_leeway,
0, qstate->region, qstate->query_flags)) {
0, qstate->region, qstate->query_flags, qstate->qstarttime,
qstate->is_valrec)) {
log_err("ipsecmod: out of memory caching record");
}
qstate->ext_state[id] = module_finished;
@@ -615,7 +616,7 @@ ipsecmod_get_mem(struct module_env* env, int id)
*/
static struct module_func_block ipsecmod_block = {
"ipsecmod",
&ipsecmod_init, &ipsecmod_deinit, &ipsecmod_operate,
NULL, NULL, &ipsecmod_init, &ipsecmod_deinit, &ipsecmod_operate,
&ipsecmod_inform_super, &ipsecmod_clear, &ipsecmod_get_mem
};
+191 -73
View File
@@ -17,9 +17,19 @@
#include "sldns/wire2str.h"
#include "sldns/parseutil.h"
#ifdef HAVE_NET_PFVAR_H
#include <fcntl.h>
#include <sys/ioctl.h>
#include <netinet/in.h>
#include <net/if.h>
#include <net/pfvar.h>
typedef intptr_t filter_dev;
#else
#include <libmnl/libmnl.h>
#include <linux/netfilter/nfnetlink.h>
#include <linux/netfilter/ipset/ip_set.h>
typedef struct mnl_socket * filter_dev;
#endif
#define BUFF_LEN 256
@@ -41,24 +51,95 @@ static int error_response(struct module_qstate* qstate, int id, int rcode) {
return 0;
}
static struct mnl_socket * open_mnl_socket() {
struct mnl_socket *mnl;
#ifdef HAVE_NET_PFVAR_H
static void * open_filter() {
filter_dev dev;
mnl = mnl_socket_open(NETLINK_NETFILTER);
if (!mnl) {
dev = open("/dev/pf", O_RDWR);
if (dev == -1) {
log_err("open(\"/dev/pf\") failed: %s", strerror(errno));
return NULL;
}
else
return (void *)dev;
}
#else
static void * open_filter() {
filter_dev dev;
dev = mnl_socket_open(NETLINK_NETFILTER);
if (!dev) {
log_err("ipset: could not open netfilter.");
return NULL;
}
if (mnl_socket_bind(mnl, 0, MNL_SOCKET_AUTOPID) < 0) {
mnl_socket_close(mnl);
if (mnl_socket_bind(dev, 0, MNL_SOCKET_AUTOPID) < 0) {
mnl_socket_close(dev);
log_err("ipset: could not bind netfilter.");
return NULL;
}
return mnl;
return (void *)dev;
}
#endif
static int add_to_ipset(struct mnl_socket *mnl, const char *setname, const void *ipaddr, int af) {
#ifdef HAVE_NET_PFVAR_H
static int add_to_ipset(filter_dev dev, const char *setname, const void *ipaddr, int af) {
struct pfioc_table io;
struct pfr_addr addr;
const char *p;
int i;
bzero(&io, sizeof(io));
bzero(&addr, sizeof(addr));
p = strrchr(setname, '/');
if (p) {
i = p - setname;
if (i >= PATH_MAX) {
errno = ENAMETOOLONG;
return -1;
}
memcpy(io.pfrio_table.pfrt_anchor, setname, i);
if (i < PATH_MAX)
io.pfrio_table.pfrt_anchor[i] = '\0';
p++;
}
else
p = setname;
if (strlen(p) >= PF_TABLE_NAME_SIZE) {
errno = ENAMETOOLONG;
return -1;
}
strlcpy(io.pfrio_table.pfrt_name, p, PF_TABLE_NAME_SIZE);
io.pfrio_buffer = &addr;
io.pfrio_size = 1;
io.pfrio_esize = sizeof(addr);
switch (af) {
case AF_INET:
addr.pfra_ip4addr = *(struct in_addr *)ipaddr;
addr.pfra_net = 32;
break;
case AF_INET6:
addr.pfra_ip6addr = *(struct in6_addr *)ipaddr;
addr.pfra_net = 128;
break;
default:
errno = EAFNOSUPPORT;
return -1;
}
addr.pfra_af = af;
if (ioctl(dev, DIOCRADDADDRS, &io) == -1) {
log_err("ioctl failed: %s", strerror(errno));
return -1;
}
return 0;
}
#else
static int add_to_ipset(filter_dev dev, const char *setname, const void *ipaddr, int af) {
struct nlmsghdr *nlh;
struct nfgenmsg *nfg;
struct nlattr *nested[2];
@@ -91,14 +172,15 @@ static int add_to_ipset(struct mnl_socket *mnl, const char *setname, const void
mnl_attr_nest_end(nlh, nested[1]);
mnl_attr_nest_end(nlh, nested[0]);
if (mnl_socket_sendto(mnl, nlh, nlh->nlmsg_len) < 0) {
if (mnl_socket_sendto(dev, nlh, nlh->nlmsg_len) < 0) {
return -1;
}
return 0;
}
#endif
static void
ipset_add_rrset_data(struct ipset_env *ie, struct mnl_socket *mnl,
ipset_add_rrset_data(struct ipset_env *ie,
struct packed_rrset_data *d, const char* setname, int af,
const char* dname)
{
@@ -123,12 +205,16 @@ ipset_add_rrset_data(struct ipset_env *ie, struct mnl_socket *mnl,
snprintf(ip, sizeof(ip), "(inet_ntop_error)");
verbose(VERB_QUERY, "ipset: add %s to %s for %s", ip, setname, dname);
}
ret = add_to_ipset(mnl, setname, rr_data + 2, af);
ret = add_to_ipset((filter_dev)ie->dev, setname, rr_data + 2, af);
if (ret < 0) {
log_err("ipset: could not add %s into %s", dname, setname);
mnl_socket_close(mnl);
ie->mnl = NULL;
#if HAVE_NET_PFVAR_H
/* don't close as we might not be able to open again due to dropped privs */
#else
mnl_socket_close((filter_dev)ie->dev);
ie->dev = NULL;
#endif
break;
}
}
@@ -137,11 +223,11 @@ ipset_add_rrset_data(struct ipset_env *ie, struct mnl_socket *mnl,
static int
ipset_check_zones_for_rrset(struct module_env *env, struct ipset_env *ie,
struct mnl_socket *mnl, struct ub_packed_rrset_key *rrset,
struct ub_packed_rrset_key *rrset, const char *qname, int qlen,
const char *setname, int af)
{
static char dname[BUFF_LEN];
const char *s;
const char *ds, *qs;
int dlen, plen;
struct config_strlist *p;
@@ -155,67 +241,79 @@ ipset_check_zones_for_rrset(struct module_env *env, struct ipset_env *ie,
if (dname[dlen - 1] == '.') {
dlen--;
}
if (qname[qlen - 1] == '.') {
qlen--;
}
for (p = env->cfg->local_zones_ipset; p; p = p->next) {
ds = NULL;
qs = NULL;
plen = strlen(p->str);
if (p->str[plen - 1] == '.') {
plen--;
}
if (dlen >= plen) {
s = dname + (dlen - plen);
if (strncasecmp(p->str, s, plen) == 0) {
d = (struct packed_rrset_data*)rrset->entry.data;
ipset_add_rrset_data(ie, mnl, d, setname,
af, dname);
break;
}
if (dlen == plen || (dlen > plen && dname[dlen - plen - 1] == '.' )) {
ds = dname + (dlen - plen);
}
if (qlen == plen || (qlen > plen && qname[qlen - plen - 1] == '.' )) {
qs = qname + (qlen - plen);
}
if ((ds && strncasecmp(p->str, ds, plen) == 0)
|| (qs && strncasecmp(p->str, qs, plen) == 0)) {
d = (struct packed_rrset_data*)rrset->entry.data;
ipset_add_rrset_data(ie, d, setname, af, dname);
break;
}
}
return 0;
}
static int ipset_update(struct module_env *env, struct dns_msg *return_msg, struct ipset_env *ie) {
struct mnl_socket *mnl;
static int ipset_update(struct module_env *env, struct dns_msg *return_msg,
struct query_info qinfo, struct ipset_env *ie)
{
size_t i;
const char *setname;
struct ub_packed_rrset_key *rrset;
int af;
static char qname[BUFF_LEN];
int qlen;
mnl = (struct mnl_socket *)ie->mnl;
if (!mnl) {
// retry to create mnl socket
mnl = open_mnl_socket();
if (!mnl) {
#ifdef HAVE_NET_PFVAR_H
#else
if (!ie->dev) {
/* retry to create mnl socket */
ie->dev = open_filter();
if (!ie->dev) {
log_warn("ipset open_filter failed");
return -1;
}
}
#endif
ie->mnl = mnl;
qlen = sldns_wire2str_dname_buf(qinfo.qname, qinfo.qname_len,
qname, BUFF_LEN);
if(qlen == 0) {
log_err("bad domain name");
return -1;
}
for (i = 0; i < return_msg->rep->rrset_count; ++i) {
for(i = 0; i < return_msg->rep->rrset_count; i++) {
setname = NULL;
rrset = return_msg->rep->rrsets[i];
if (rrset->rk.type == htons(LDNS_RR_TYPE_A)) {
if(ntohs(rrset->rk.type) == LDNS_RR_TYPE_A &&
ie->v4_enabled == 1) {
af = AF_INET;
if ((ie->v4_enabled == 1)) {
setname = ie->name_v4;
}
} else {
setname = ie->name_v4;
} else if(ntohs(rrset->rk.type) == LDNS_RR_TYPE_AAAA &&
ie->v6_enabled == 1) {
af = AF_INET6;
if ((ie->v6_enabled == 1)) {
setname = ie->name_v6;
}
setname = ie->name_v6;
}
if (setname) {
if(ipset_check_zones_for_rrset(env, ie, mnl, rrset,
setname, af) == -1)
if(ipset_check_zones_for_rrset(env, ie, rrset, qname,
qlen, setname, af) == -1)
return -1;
}
}
@@ -223,7 +321,7 @@ static int ipset_update(struct module_env *env, struct dns_msg *return_msg, stru
return 0;
}
int ipset_init(struct module_env* env, int id) {
int ipset_startup(struct module_env* env, int id) {
struct ipset_env *ipset_env;
ipset_env = (struct ipset_env *)calloc(1, sizeof(struct ipset_env));
@@ -234,7 +332,43 @@ int ipset_init(struct module_env* env, int id) {
env->modinfo[id] = (void *)ipset_env;
ipset_env->mnl = NULL;
#ifdef HAVE_NET_PFVAR_H
ipset_env->dev = open_filter();
if (!ipset_env->dev) {
log_err("ipset open_filter failed");
return 0;
}
#else
ipset_env->dev = NULL;
#endif
return 1;
}
void ipset_destartup(struct module_env* env, int id) {
filter_dev dev;
struct ipset_env *ipset_env;
if (!env || !env->modinfo[id]) {
return;
}
ipset_env = (struct ipset_env*)env->modinfo[id];
dev = (filter_dev)ipset_env->dev;
if (dev) {
#if HAVE_NET_PFVAR_H
close(dev);
#else
mnl_socket_close(dev);
#endif
ipset_env->dev = NULL;
}
free(ipset_env);
env->modinfo[id] = NULL;
}
int ipset_init(struct module_env* env, int id) {
struct ipset_env *ipset_env = env->modinfo[id];
ipset_env->name_v4 = env->cfg->ipset_name_v4;
ipset_env->name_v6 = env->cfg->ipset_name_v6;
@@ -250,24 +384,8 @@ int ipset_init(struct module_env* env, int id) {
return 1;
}
void ipset_deinit(struct module_env *env, int id) {
struct mnl_socket *mnl;
struct ipset_env *ipset_env;
if (!env || !env->modinfo[id]) {
return;
}
ipset_env = (struct ipset_env *)env->modinfo[id];
mnl = (struct mnl_socket *)ipset_env->mnl;
if (mnl) {
mnl_socket_close(mnl);
ipset_env->mnl = NULL;
}
free(ipset_env);
env->modinfo[id] = NULL;
void ipset_deinit(struct module_env *ATTR_UNUSED(env), int ATTR_UNUSED(id)) {
/* nothing */
}
static int ipset_new(struct module_qstate* qstate, int id) {
@@ -311,7 +429,7 @@ void ipset_operate(struct module_qstate *qstate, enum module_ev event, int id,
if (iq && (event == module_event_moddone)) {
if (qstate->return_msg && qstate->return_msg->rep) {
ipset_update(qstate->env, qstate->return_msg, ie);
ipset_update(qstate->env, qstate->return_msg, qstate->qinfo, ie);
}
qstate->ext_state[id] = module_finished;
return;
@@ -373,8 +491,8 @@ size_t ipset_get_mem(struct module_env *env, int id) {
*/
static struct module_func_block ipset_block = {
"ipset",
&ipset_init, &ipset_deinit, &ipset_operate,
&ipset_inform_super, &ipset_clear, &ipset_get_mem
&ipset_startup, &ipset_destartup, &ipset_init, &ipset_deinit,
&ipset_operate, &ipset_inform_super, &ipset_clear, &ipset_get_mem
};
struct module_func_block * ipset_get_funcblock(void) {
+5 -1
View File
@@ -37,7 +37,7 @@ extern "C" {
#endif
struct ipset_env {
void* mnl;
void* dev;
int v4_enabled;
int v6_enabled;
@@ -50,6 +50,10 @@ struct ipset_qstate {
int dummy;
};
/** Startup the ipset module */
int ipset_startup(struct module_env* env, int id);
/** Destartup the ipset module */
void ipset_destartup(struct module_env* env, int id);
/** Init the ipset module */
int ipset_init(struct module_env* env, int id);
/** Deinit the ipset module */
+59 -6
View File
@@ -78,6 +78,7 @@ struct delegpt* delegpt_copy(struct delegpt* dp, struct regional* region)
if(!delegpt_add_ns(copy, region, ns->name, ns->lame,
ns->tls_auth_name, ns->port))
return NULL;
copy->nslist->cache_lookup_count = ns->cache_lookup_count;
copy->nslist->resolved = ns->resolved;
copy->nslist->got4 = ns->got4;
copy->nslist->got6 = ns->got6;
@@ -121,6 +122,7 @@ delegpt_add_ns(struct delegpt* dp, struct regional* region, uint8_t* name,
ns->namelen = len;
dp->nslist = ns;
ns->name = regional_alloc_init(region, name, ns->namelen);
ns->cache_lookup_count = 0;
ns->resolved = 0;
ns->got4 = 0;
ns->got6 = 0;
@@ -185,6 +187,10 @@ delegpt_add_target(struct delegpt* dp, struct regional* region,
else ns->got4 = 1;
if(ns->got4 && ns->got6)
ns->resolved = 1;
} else {
if(addr_is_ip6(addr, addrlen))
ns->done_pside6 = 1;
else ns->done_pside4 = 1;
}
log_assert(ns->port>0);
return delegpt_add_addr(dp, region, addr, addrlen, bogus, lame,
@@ -272,7 +278,7 @@ delegpt_count_addr(struct delegpt* dp, size_t* numaddr, size_t* numres,
void delegpt_log(enum verbosity_value v, struct delegpt* dp)
{
char buf[LDNS_MAX_DOMAINLEN+1];
char buf[LDNS_MAX_DOMAINLEN];
struct delegpt_ns* ns;
struct delegpt_addr* a;
size_t missing=0, numns=0, numaddr=0, numres=0, numavail=0;
@@ -315,6 +321,45 @@ void delegpt_log(enum verbosity_value v, struct delegpt* dp)
}
}
int
delegpt_addr_on_result_list(struct delegpt* dp, struct delegpt_addr* find)
{
struct delegpt_addr* a = dp->result_list;
while(a) {
if(a == find)
return 1;
a = a->next_result;
}
return 0;
}
void
delegpt_usable_list_remove_addr(struct delegpt* dp, struct delegpt_addr* del)
{
struct delegpt_addr* usa = dp->usable_list, *prev = NULL;
while(usa) {
if(usa == del) {
/* snip off the usable list */
if(prev)
prev->next_usable = usa->next_usable;
else dp->usable_list = usa->next_usable;
return;
}
prev = usa;
usa = usa->next_usable;
}
}
void
delegpt_add_to_result_list(struct delegpt* dp, struct delegpt_addr* a)
{
if(delegpt_addr_on_result_list(dp, a))
return;
delegpt_usable_list_remove_addr(dp, a);
a->next_result = dp->result_list;
dp->result_list = a;
}
void
delegpt_add_unused_targets(struct delegpt* dp)
{
@@ -338,13 +383,16 @@ delegpt_count_targets(struct delegpt* dp)
}
size_t
delegpt_count_missing_targets(struct delegpt* dp)
delegpt_count_missing_targets(struct delegpt* dp, int* alllame)
{
struct delegpt_ns* ns;
size_t n = 0;
for(ns = dp->nslist; ns; ns = ns->next)
if(!ns->resolved)
n++;
size_t n = 0, nlame = 0;
for(ns = dp->nslist; ns; ns = ns->next) {
if(ns->resolved) continue;
n++;
if(ns->lame) nlame++;
}
if(alllame && n == nlame) *alllame = 1;
return n;
}
@@ -613,6 +661,7 @@ int delegpt_add_ns_mlc(struct delegpt* dp, uint8_t* name, uint8_t lame,
}
ns->next = dp->nslist;
dp->nslist = ns;
ns->cache_lookup_count = 0;
ns->resolved = 0;
ns->got4 = 0;
ns->got6 = 0;
@@ -694,6 +743,10 @@ int delegpt_add_target_mlc(struct delegpt* dp, uint8_t* name, size_t namelen,
else ns->got4 = 1;
if(ns->got4 && ns->got6)
ns->resolved = 1;
} else {
if(addr_is_ip6(addr, addrlen))
ns->done_pside6 = 1;
else ns->done_pside4 = 1;
}
log_assert(ns->port>0);
return delegpt_add_addr_mlc(dp, addr, addrlen, bogus, lame,
+40 -2
View File
@@ -79,6 +79,16 @@ struct delegpt {
* Also true if the delegationpoint was created from a delegation
* message and thus contains the parent-side-info already. */
uint8_t has_parent_side_NS;
/** if true, the delegation point has reached last resort processing
* and the parent side information has been possibly added to the
* delegation point.
* For now this signals that further target lookups will ignore
* the configured target-fetch-policy and only resolve on
* demand to try and avoid triggering limits at this stage (.i.e, it
* is very likely that the A/AAAA queries for the newly added name
* servers will not yield new IP addresses and trigger NXNS
* countermeasures. */
uint8_t fallback_to_parent_side_NS;
/** for assertions on type of delegpt */
uint8_t dp_type_mlc;
/** use SSL for upstream query */
@@ -101,6 +111,8 @@ struct delegpt_ns {
uint8_t* name;
/** length of name */
size_t namelen;
/** number of cache lookups for the name */
int cache_lookup_count;
/**
* If the name has been resolved. false if not queried for yet.
* true if the A, AAAA queries have been generated.
@@ -128,7 +140,7 @@ struct delegpt_ns {
uint8_t done_pside6;
/** the TLS authentication name, (if not NULL) to use. */
char* tls_auth_name;
/** the port to use; it should mosty be the default 53 but configured
/** the port to use; it should mostly be the default 53 but configured
* upstreams can provide nondefault ports. */
int port;
};
@@ -330,9 +342,10 @@ void delegpt_add_unused_targets(struct delegpt* dp);
/**
* Count number of missing targets. These are ns names with no resolved flag.
* @param dp: delegation point.
* @param alllame: if set, check if all the missing targets are lame.
* @return number of missing targets (or 0).
*/
size_t delegpt_count_missing_targets(struct delegpt* dp);
size_t delegpt_count_missing_targets(struct delegpt* dp, int* alllame);
/** count total number of targets in dp */
size_t delegpt_count_targets(struct delegpt* dp);
@@ -454,4 +467,29 @@ int delegpt_add_target_mlc(struct delegpt* dp, uint8_t* name, size_t namelen,
/** get memory in use by dp */
size_t delegpt_get_mem(struct delegpt* dp);
/**
* See if the addr is on the result list.
* @param dp: delegation point.
* @param find: the pointer is searched for on the result list.
* @return 1 if found, 0 if not found.
*/
int delegpt_addr_on_result_list(struct delegpt* dp, struct delegpt_addr* find);
/**
* Remove the addr from the usable list.
* @param dp: the delegation point.
* @param del: the addr to remove from the list, the pointer is searched for.
*/
void delegpt_usable_list_remove_addr(struct delegpt* dp,
struct delegpt_addr* del);
/**
* Add the delegpt_addr back to the result list, if it is not already on
* the result list. Also removes it from the usable list.
* @param dp: delegation point.
* @param a: addr to add, nothing happens if it is already on the result list.
* It is removed from the usable list.
*/
void delegpt_add_to_result_list(struct delegpt* dp, struct delegpt_addr* a);
#endif /* ITERATOR_ITER_DELEGPT_H */
+164 -38
View File
@@ -71,6 +71,7 @@ forwards_create(void)
sizeof(struct iter_forwards));
if(!fwd)
return NULL;
lock_rw_init(&fwd->lock);
return fwd;
}
@@ -100,6 +101,7 @@ forwards_delete(struct iter_forwards* fwd)
{
if(!fwd)
return;
lock_rw_destroy(&fwd->lock);
fwd_del_tree(fwd);
free(fwd);
}
@@ -127,7 +129,7 @@ forwards_insert_data(struct iter_forwards* fwd, uint16_t c, uint8_t* nm,
node->namelabs = nmlabs;
node->dp = dp;
if(!rbtree_insert(fwd->tree, &node->node)) {
char buf[257];
char buf[LDNS_MAX_DOMAINLEN];
dname_str(nm, buf);
log_err("duplicate forward zone %s ignored.", buf);
delegpt_free_mlc(dp);
@@ -137,6 +139,17 @@ forwards_insert_data(struct iter_forwards* fwd, uint16_t c, uint8_t* nm,
return 1;
}
static struct iter_forward_zone*
fwd_zone_find(struct iter_forwards* fwd, uint16_t c, uint8_t* nm)
{
struct iter_forward_zone key;
key.node.key = &key;
key.dclass = c;
key.name = nm;
key.namelabs = dname_count_size_labels(nm, &key.namelen);
return (struct iter_forward_zone*)rbtree_search(fwd->tree, &key);
}
/** insert new info into forward structure given dp */
static int
forwards_insert(struct iter_forwards* fwd, uint16_t c, struct delegpt* dp)
@@ -319,6 +332,40 @@ make_stub_holes(struct iter_forwards* fwd, struct config_file* cfg)
log_err("cannot parse stub name '%s'", s->name);
return 0;
}
if(fwd_zone_find(fwd, LDNS_RR_CLASS_IN, dname) != NULL) {
/* Already a forward zone there. */
free(dname);
continue;
}
if(!fwd_add_stub_hole(fwd, LDNS_RR_CLASS_IN, dname)) {
free(dname);
log_err("out of memory");
return 0;
}
free(dname);
}
return 1;
}
/** make NULL entries for auths */
static int
make_auth_holes(struct iter_forwards* fwd, struct config_file* cfg)
{
struct config_auth* a;
uint8_t* dname;
size_t dname_len;
for(a = cfg->auths; a; a = a->next) {
if(!a->name) continue;
dname = sldns_str2wire_dname(a->name, &dname_len);
if(!dname) {
log_err("cannot parse auth name '%s'", a->name);
return 0;
}
if(fwd_zone_find(fwd, LDNS_RR_CLASS_IN, dname) != NULL) {
/* Already a forward zone there. */
free(dname);
continue;
}
if(!fwd_add_stub_hole(fwd, LDNS_RR_CLASS_IN, dname)) {
free(dname);
log_err("out of memory");
@@ -332,45 +379,74 @@ make_stub_holes(struct iter_forwards* fwd, struct config_file* cfg)
int
forwards_apply_cfg(struct iter_forwards* fwd, struct config_file* cfg)
{
if(fwd->tree) {
lock_unprotect(&fwd->lock, fwd->tree);
}
fwd_del_tree(fwd);
fwd->tree = rbtree_create(fwd_cmp);
if(!fwd->tree)
return 0;
lock_protect(&fwd->lock, fwd->tree, sizeof(*fwd->tree));
lock_rw_wrlock(&fwd->lock);
/* read forward zones */
if(!read_forwards(fwd, cfg))
if(!read_forwards(fwd, cfg)) {
lock_rw_unlock(&fwd->lock);
return 0;
if(!make_stub_holes(fwd, cfg))
}
if(!make_stub_holes(fwd, cfg)) {
lock_rw_unlock(&fwd->lock);
return 0;
}
/* TODO: Now we punch holes for auth zones as well so that in
* iterator:forward_request() we see the configured
* delegation point, but code flow/naming is hard to follow.
* Consider having a single tree with configured
* delegation points for all categories
* (stubs, forwards, auths). */
if(!make_auth_holes(fwd, cfg)) {
lock_rw_unlock(&fwd->lock);
return 0;
}
fwd_init_parents(fwd);
lock_rw_unlock(&fwd->lock);
return 1;
}
struct delegpt*
forwards_find(struct iter_forwards* fwd, uint8_t* qname, uint16_t qclass)
forwards_find(struct iter_forwards* fwd, uint8_t* qname, uint16_t qclass,
int nolock)
{
rbnode_type* res = NULL;
struct iter_forward_zone* res;
struct iter_forward_zone key;
int has_dp;
key.node.key = &key;
key.dclass = qclass;
key.name = qname;
key.namelabs = dname_count_size_labels(qname, &key.namelen);
res = rbtree_search(fwd->tree, &key);
if(res) return ((struct iter_forward_zone*)res)->dp;
return NULL;
/* lock_() calls are macros that could be nothing, surround in {} */
if(!nolock) { lock_rw_rdlock(&fwd->lock); }
res = (struct iter_forward_zone*)rbtree_search(fwd->tree, &key);
has_dp = res && res->dp;
if(!has_dp && !nolock) { lock_rw_unlock(&fwd->lock); }
return has_dp?res->dp:NULL;
}
struct delegpt*
forwards_lookup(struct iter_forwards* fwd, uint8_t* qname, uint16_t qclass)
forwards_lookup(struct iter_forwards* fwd, uint8_t* qname, uint16_t qclass,
int nolock)
{
/* lookup the forward zone in the tree */
rbnode_type* res = NULL;
struct iter_forward_zone *result;
struct iter_forward_zone key;
int has_dp;
key.node.key = &key;
key.dclass = qclass;
key.name = qname;
key.namelabs = dname_count_size_labels(qname, &key.namelen);
/* lock_() calls are macros that could be nothing, surround in {} */
if(!nolock) { lock_rw_rdlock(&fwd->lock); }
if(rbtree_find_less_equal(fwd->tree, &key, &res)) {
/* exact */
result = (struct iter_forward_zone*)res;
@@ -378,8 +454,10 @@ forwards_lookup(struct iter_forwards* fwd, uint8_t* qname, uint16_t qclass)
/* smaller element (or no element) */
int m;
result = (struct iter_forward_zone*)res;
if(!result || result->dclass != qclass)
if(!result || result->dclass != qclass) {
if(!nolock) { lock_rw_unlock(&fwd->lock); }
return NULL;
}
/* count number of labels matched */
(void)dname_lab_cmp(result->name, result->namelabs, key.name,
key.namelabs, &m);
@@ -389,20 +467,22 @@ forwards_lookup(struct iter_forwards* fwd, uint8_t* qname, uint16_t qclass)
result = result->parent;
}
}
if(result)
return result->dp;
return NULL;
has_dp = result && result->dp;
if(!has_dp && !nolock) { lock_rw_unlock(&fwd->lock); }
return has_dp?result->dp:NULL;
}
struct delegpt*
forwards_lookup_root(struct iter_forwards* fwd, uint16_t qclass)
forwards_lookup_root(struct iter_forwards* fwd, uint16_t qclass, int nolock)
{
uint8_t root = 0;
return forwards_lookup(fwd, &root, qclass);
return forwards_lookup(fwd, &root, qclass, nolock);
}
int
forwards_next_root(struct iter_forwards* fwd, uint16_t* dclass)
/* Finds next root item in forwards lookup tree.
* Caller needs to handle locking of the forwards structure. */
static int
next_root_locked(struct iter_forwards* fwd, uint16_t* dclass)
{
struct iter_forward_zone key;
rbnode_type* n;
@@ -419,7 +499,7 @@ forwards_next_root(struct iter_forwards* fwd, uint16_t* dclass)
}
/* root not first item? search for higher items */
*dclass = p->dclass + 1;
return forwards_next_root(fwd, dclass);
return next_root_locked(fwd, dclass);
}
/* find class n in tree, we may get a direct hit, or if we don't
* this is the last item of the previous class so rbtree_next() takes
@@ -447,10 +527,21 @@ forwards_next_root(struct iter_forwards* fwd, uint16_t* dclass)
}
/* not a root node, return next higher item */
*dclass = p->dclass+1;
return forwards_next_root(fwd, dclass);
return next_root_locked(fwd, dclass);
}
}
int
forwards_next_root(struct iter_forwards* fwd, uint16_t* dclass, int nolock)
{
int ret;
/* lock_() calls are macros that could be nothing, surround in {} */
if(!nolock) { lock_rw_rdlock(&fwd->lock); }
ret = next_root_locked(fwd, dclass);
if(!nolock) { lock_rw_unlock(&fwd->lock); }
return ret;
}
size_t
forwards_get_mem(struct iter_forwards* fwd)
{
@@ -458,69 +549,104 @@ forwards_get_mem(struct iter_forwards* fwd)
size_t s;
if(!fwd)
return 0;
lock_rw_rdlock(&fwd->lock);
s = sizeof(*fwd) + sizeof(*fwd->tree);
RBTREE_FOR(p, struct iter_forward_zone*, fwd->tree) {
s += sizeof(*p) + p->namelen + delegpt_get_mem(p->dp);
}
lock_rw_unlock(&fwd->lock);
return s;
}
static struct iter_forward_zone*
fwd_zone_find(struct iter_forwards* fwd, uint16_t c, uint8_t* nm)
{
struct iter_forward_zone key;
key.node.key = &key;
key.dclass = c;
key.name = nm;
key.namelabs = dname_count_size_labels(nm, &key.namelen);
return (struct iter_forward_zone*)rbtree_search(fwd->tree, &key);
}
int
forwards_add_zone(struct iter_forwards* fwd, uint16_t c, struct delegpt* dp)
forwards_add_zone(struct iter_forwards* fwd, uint16_t c, struct delegpt* dp,
int nolock)
{
struct iter_forward_zone *z;
/* lock_() calls are macros that could be nothing, surround in {} */
if(!nolock) { lock_rw_wrlock(&fwd->lock); }
if((z=fwd_zone_find(fwd, c, dp->name)) != NULL) {
(void)rbtree_delete(fwd->tree, &z->node);
fwd_zone_free(z);
}
if(!forwards_insert(fwd, c, dp))
if(!forwards_insert(fwd, c, dp)) {
if(!nolock) { lock_rw_unlock(&fwd->lock); }
return 0;
}
fwd_init_parents(fwd);
if(!nolock) { lock_rw_unlock(&fwd->lock); }
return 1;
}
void
forwards_delete_zone(struct iter_forwards* fwd, uint16_t c, uint8_t* nm)
forwards_delete_zone(struct iter_forwards* fwd, uint16_t c, uint8_t* nm,
int nolock)
{
struct iter_forward_zone *z;
if(!(z=fwd_zone_find(fwd, c, nm)))
/* lock_() calls are macros that could be nothing, surround in {} */
if(!nolock) { lock_rw_wrlock(&fwd->lock); }
if(!(z=fwd_zone_find(fwd, c, nm))) {
if(!nolock) { lock_rw_unlock(&fwd->lock); }
return; /* nothing to do */
}
(void)rbtree_delete(fwd->tree, &z->node);
fwd_zone_free(z);
fwd_init_parents(fwd);
if(!nolock) { lock_rw_unlock(&fwd->lock); }
}
int
forwards_add_stub_hole(struct iter_forwards* fwd, uint16_t c, uint8_t* nm)
forwards_add_stub_hole(struct iter_forwards* fwd, uint16_t c, uint8_t* nm,
int nolock)
{
/* lock_() calls are macros that could be nothing, surround in {} */
if(!nolock) { lock_rw_wrlock(&fwd->lock); }
if(fwd_zone_find(fwd, c, nm) != NULL) {
if(!nolock) { lock_rw_unlock(&fwd->lock); }
return 1; /* already a stub zone there */
}
if(!fwd_add_stub_hole(fwd, c, nm)) {
if(!nolock) { lock_rw_unlock(&fwd->lock); }
return 0;
}
fwd_init_parents(fwd);
if(!nolock) { lock_rw_unlock(&fwd->lock); }
return 1;
}
void
forwards_delete_stub_hole(struct iter_forwards* fwd, uint16_t c, uint8_t* nm)
forwards_delete_stub_hole(struct iter_forwards* fwd, uint16_t c,
uint8_t* nm, int nolock)
{
struct iter_forward_zone *z;
if(!(z=fwd_zone_find(fwd, c, nm)))
/* lock_() calls are macros that could be nothing, surround in {} */
if(!nolock) { lock_rw_wrlock(&fwd->lock); }
if(!(z=fwd_zone_find(fwd, c, nm))) {
if(!nolock) { lock_rw_unlock(&fwd->lock); }
return; /* nothing to do */
if(z->dp != NULL)
}
if(z->dp != NULL) {
if(!nolock) { lock_rw_unlock(&fwd->lock); }
return; /* not a stub hole */
}
(void)rbtree_delete(fwd->tree, &z->node);
fwd_zone_free(z);
fwd_init_parents(fwd);
if(!nolock) { lock_rw_unlock(&fwd->lock); }
}
void
forwards_swap_tree(struct iter_forwards* fwd, struct iter_forwards* data)
{
rbtree_type* oldtree = fwd->tree;
if(oldtree) {
lock_unprotect(&fwd->lock, oldtree);
}
if(data->tree) {
lock_unprotect(&data->lock, data->tree);
}
fwd->tree = data->tree;
data->tree = oldtree;
lock_protect(&fwd->lock, fwd->tree, sizeof(*fwd->tree));
lock_protect(&data->lock, data->tree, sizeof(*data->tree));
}
+58 -11
View File
@@ -43,6 +43,7 @@
#ifndef ITERATOR_ITER_FWD_H
#define ITERATOR_ITER_FWD_H
#include "util/rbtree.h"
#include "util/locks.h"
struct config_file;
struct delegpt;
@@ -50,6 +51,11 @@ struct delegpt;
* Iterator forward zones structure
*/
struct iter_forwards {
/** lock on the forwards tree.
* When grabbing both this lock and the anchors.lock, this lock
* is grabbed first. When grabbing both this lock and the hints.lock
* this lock is grabbed first. */
lock_rw_type lock;
/**
* Zones are stored in this tree. Sort order is specially chosen.
* first sorted on qclass. Then on dname in nsec-like order, so that
@@ -106,47 +112,65 @@ int forwards_apply_cfg(struct iter_forwards* fwd, struct config_file* cfg);
/**
* Find forward zone exactly by name
* The return value is contents of the forwards structure.
* Caller should lock and unlock a readlock on the forwards structure if nolock
* is set.
* Otherwise caller should unlock the readlock on the forwards structure if a
* value was returned.
* @param fwd: forward storage.
* @param qname: The qname of the query.
* @param qclass: The qclass of the query.
* @param nolock: Skip locking, locking is handled by the caller.
* @return: A delegation point or null.
*/
struct delegpt* forwards_find(struct iter_forwards* fwd, uint8_t* qname,
uint16_t qclass);
uint16_t qclass, int nolock);
/**
* Find forward zone information
* For this qname/qclass find forward zone information, returns delegation
* point with server names and addresses, or NULL if no forwarding is needed.
* The return value is contents of the forwards structure.
* Caller should lock and unlock a readlock on the forwards structure if nolock
* is set.
* Otherwise caller should unlock the readlock on the forwards structure if a
* value was returned.
*
* @param fwd: forward storage.
* @param qname: The qname of the query.
* @param qclass: The qclass of the query.
* @param nolock: Skip locking, locking is handled by the caller.
* @return: A delegation point if the query has to be forwarded to that list,
* otherwise null.
*/
struct delegpt* forwards_lookup(struct iter_forwards* fwd,
uint8_t* qname, uint16_t qclass);
struct delegpt* forwards_lookup(struct iter_forwards* fwd,
uint8_t* qname, uint16_t qclass, int nolock);
/**
* Same as forwards_lookup, but for the root only
* @param fwd: forward storage.
* @param qclass: The qclass of the query.
* @param nolock: Skip locking, locking is handled by the caller.
* @return: A delegation point if root forward exists, otherwise null.
*/
struct delegpt* forwards_lookup_root(struct iter_forwards* fwd,
uint16_t qclass);
struct delegpt* forwards_lookup_root(struct iter_forwards* fwd,
uint16_t qclass, int nolock);
/**
* Find next root item in forwards lookup tree.
* Handles its own locking unless nolock is set. In that case the caller
* should lock and unlock a readlock on the forwards structure.
* @param fwd: the forward storage
* @param qclass: class to look at next, or higher.
* @param nolock: Skip locking, locking is handled by the caller.
* @return false if none found, or if true stored in qclass.
*/
int forwards_next_root(struct iter_forwards* fwd, uint16_t* qclass);
int forwards_next_root(struct iter_forwards* fwd, uint16_t* qclass,
int nolock);
/**
* Get memory in use by forward storage
* Locks and unlocks the structure.
* @param fwd: forward storage.
* @return bytes in use
*/
@@ -158,42 +182,65 @@ int fwd_cmp(const void* k1, const void* k2);
/**
* Add zone to forward structure. For external use since it recalcs
* the tree parents.
* Handles its own locking unless nolock is set. In that case the caller
* should lock and unlock a writelock on the forwards structure.
* @param fwd: the forward data structure
* @param c: class of zone
* @param dp: delegation point with name and target nameservers for new
* forward zone. malloced.
* @param nolock: Skip locking, locking is handled by the caller.
* @return false on failure (out of memory);
*/
int forwards_add_zone(struct iter_forwards* fwd, uint16_t c,
struct delegpt* dp);
int forwards_add_zone(struct iter_forwards* fwd, uint16_t c,
struct delegpt* dp, int nolock);
/**
* Remove zone from forward structure. For external use since it
* recalcs the tree parents.
* Handles its own locking unless nolock is set. In that case the caller
* should lock and unlock a writelock on the forwards structure.
* @param fwd: the forward data structure
* @param c: class of zone
* @param nm: name of zone (in uncompressed wireformat).
* @param nolock: Skip locking, locking is handled by the caller.
*/
void forwards_delete_zone(struct iter_forwards* fwd, uint16_t c, uint8_t* nm);
void forwards_delete_zone(struct iter_forwards* fwd, uint16_t c,
uint8_t* nm, int nolock);
/**
* Add stub hole (empty entry in forward table, that makes resolution skip
* a forward-zone because the stub zone should override the forward zone).
* Does not add one if not necessary.
* Handles its own locking unless nolock is set. In that case the caller
* should lock and unlock a writelock on the forwards structure.
* @param fwd: the forward data structure
* @param c: class of zone
* @param nm: name of zone (in uncompressed wireformat).
* @param nolock: Skip locking, locking is handled by the caller.
* @return false on failure (out of memory);
*/
int forwards_add_stub_hole(struct iter_forwards* fwd, uint16_t c, uint8_t* nm);
int forwards_add_stub_hole(struct iter_forwards* fwd, uint16_t c,
uint8_t* nm, int nolock);
/**
* Remove stub hole, if one exists.
* Handles its own locking unless nolock is set. In that case the caller
* should lock and unlock a writelock on the forwards structure.
* @param fwd: the forward data structure
* @param c: class of zone
* @param nm: name of zone (in uncompressed wireformat).
* @param nolock: Skip locking, locking is handled by the caller.
*/
void forwards_delete_stub_hole(struct iter_forwards* fwd, uint16_t c,
uint8_t* nm);
uint8_t* nm, int nolock);
/**
* Swap internal tree with preallocated entries. Caller should manage
* the locks.
* @param fwd: the forward data structure.
* @param data: the data structure used to take elements from. This contains
* the old elements on return.
*/
void forwards_swap_tree(struct iter_forwards* fwd, struct iter_forwards* data);
#endif /* ITERATOR_ITER_FWD_H */
+88 -27
View File
@@ -57,6 +57,8 @@ hints_create(void)
sizeof(struct iter_hints));
if(!hints)
return NULL;
lock_rw_init(&hints->lock);
lock_protect(&hints->lock, &hints->tree, sizeof(hints->tree));
return hints;
}
@@ -83,6 +85,7 @@ hints_delete(struct iter_hints* hints)
{
if(!hints)
return;
lock_rw_destroy(&hints->lock);
hints_del_tree(hints);
free(hints);
}
@@ -100,7 +103,7 @@ ah(struct delegpt* dp, const char* sv, const char* ip)
return 0;
}
if(!delegpt_add_ns_mlc(dp, dname, 0, NULL, UNBOUND_DNS_PORT) ||
!extstrtoaddr(ip, &addr, &addrlen) ||
!extstrtoaddr(ip, &addr, &addrlen, UNBOUND_DNS_PORT) ||
!delegpt_add_target_mlc(dp, dname, dname_len,
&addr, addrlen, 0, 0)) {
free(dname);
@@ -129,7 +132,7 @@ compile_time_root_prime(int do_ip4, int do_ip6)
dp->has_parent_side_NS = 1;
if(do_ip4) {
if(!ah(dp, "A.ROOT-SERVERS.NET.", "198.41.0.4")) goto failed;
if(!ah(dp, "B.ROOT-SERVERS.NET.", "199.9.14.201")) goto failed;
if(!ah(dp, "B.ROOT-SERVERS.NET.", "170.247.170.2")) goto failed;
if(!ah(dp, "C.ROOT-SERVERS.NET.", "192.33.4.12")) goto failed;
if(!ah(dp, "D.ROOT-SERVERS.NET.", "199.7.91.13")) goto failed;
if(!ah(dp, "E.ROOT-SERVERS.NET.", "192.203.230.10")) goto failed;
@@ -144,7 +147,7 @@ compile_time_root_prime(int do_ip4, int do_ip6)
}
if(do_ip6) {
if(!ah(dp, "A.ROOT-SERVERS.NET.", "2001:503:ba3e::2:30")) goto failed;
if(!ah(dp, "B.ROOT-SERVERS.NET.", "2001:500:200::b")) goto failed;
if(!ah(dp, "B.ROOT-SERVERS.NET.", "2801:1b8:10::b")) goto failed;
if(!ah(dp, "C.ROOT-SERVERS.NET.", "2001:500:2::c")) goto failed;
if(!ah(dp, "D.ROOT-SERVERS.NET.", "2001:500:2d::d")) goto failed;
if(!ah(dp, "E.ROOT-SERVERS.NET.", "2001:500:a8::e")) goto failed;
@@ -178,7 +181,7 @@ hints_insert(struct iter_hints* hints, uint16_t c, struct delegpt* dp,
node->noprime = (uint8_t)noprime;
if(!name_tree_insert(&hints->tree, &node->node, dp->name, dp->namelen,
dp->namelabs, c)) {
char buf[257];
char buf[LDNS_MAX_DOMAINLEN];
dname_str(dp->name, buf);
log_err("second hints for zone %s ignored.", buf);
delegpt_free_mlc(dp);
@@ -438,47 +441,70 @@ read_root_hints_list(struct iter_hints* hints, struct config_file* cfg)
int
hints_apply_cfg(struct iter_hints* hints, struct config_file* cfg)
{
int nolock = 1;
lock_rw_wrlock(&hints->lock);
hints_del_tree(hints);
name_tree_init(&hints->tree);
/* read root hints */
if(!read_root_hints_list(hints, cfg))
if(!read_root_hints_list(hints, cfg)) {
lock_rw_unlock(&hints->lock);
return 0;
}
/* read stub hints */
if(!read_stubs(hints, cfg))
if(!read_stubs(hints, cfg)) {
lock_rw_unlock(&hints->lock);
return 0;
}
/* use fallback compiletime root hints */
if(!hints_lookup_root(hints, LDNS_RR_CLASS_IN)) {
if(!hints_find_root(hints, LDNS_RR_CLASS_IN, nolock)) {
struct delegpt* dp = compile_time_root_prime(cfg->do_ip4,
cfg->do_ip6);
verbose(VERB_ALGO, "no config, using builtin root hints.");
if(!dp)
if(!dp) {
lock_rw_unlock(&hints->lock);
return 0;
if(!hints_insert(hints, LDNS_RR_CLASS_IN, dp, 0))
}
if(!hints_insert(hints, LDNS_RR_CLASS_IN, dp, 0)) {
lock_rw_unlock(&hints->lock);
return 0;
}
}
name_tree_init_parents(&hints->tree);
lock_rw_unlock(&hints->lock);
return 1;
}
struct delegpt*
hints_lookup_root(struct iter_hints* hints, uint16_t qclass)
struct delegpt*
hints_find(struct iter_hints* hints, uint8_t* qname, uint16_t qclass,
int nolock)
{
struct iter_hints_stub *stub;
size_t len;
int has_dp;
int labs = dname_count_size_labels(qname, &len);
/* lock_() calls are macros that could be nothing, surround in {} */
if(!nolock) { lock_rw_rdlock(&hints->lock); }
stub = (struct iter_hints_stub*)name_tree_find(&hints->tree,
qname, len, labs, qclass);
has_dp = stub && stub->dp;
if(!has_dp && !nolock) { lock_rw_unlock(&hints->lock); }
return has_dp?stub->dp:NULL;
}
struct delegpt*
hints_find_root(struct iter_hints* hints, uint16_t qclass, int nolock)
{
uint8_t rootlab = 0;
struct iter_hints_stub *stub;
stub = (struct iter_hints_stub*)name_tree_find(&hints->tree,
&rootlab, 1, 1, qclass);
if(!stub)
return NULL;
return stub->dp;
return hints_find(hints, &rootlab, qclass, nolock);
}
struct iter_hints_stub*
hints_lookup_stub(struct iter_hints* hints, uint8_t* qname,
uint16_t qclass, struct delegpt* cache_dp)
hints_lookup_stub(struct iter_hints* hints, uint8_t* qname,
uint16_t qclass, struct delegpt* cache_dp, int nolock)
{
size_t len;
int labs;
@@ -486,14 +512,20 @@ hints_lookup_stub(struct iter_hints* hints, uint8_t* qname,
/* first lookup the stub */
labs = dname_count_size_labels(qname, &len);
/* lock_() calls are macros that could be nothing, surround in {} */
if(!nolock) { lock_rw_rdlock(&hints->lock); }
r = (struct iter_hints_stub*)name_tree_lookup(&hints->tree, qname,
len, labs, qclass);
if(!r) return NULL;
if(!r) {
if(!nolock) { lock_rw_unlock(&hints->lock); }
return NULL;
}
/* If there is no cache (root prime situation) */
if(cache_dp == NULL) {
if(r->dp->namelabs != 1)
return r; /* no cache dp, use any non-root stub */
if(!nolock) { lock_rw_unlock(&hints->lock); }
return NULL;
}
@@ -510,12 +542,18 @@ hints_lookup_stub(struct iter_hints* hints, uint8_t* qname,
if(dname_strict_subdomain(r->dp->name, r->dp->namelabs,
cache_dp->name, cache_dp->namelabs))
return r; /* need to prime this stub */
if(!nolock) { lock_rw_unlock(&hints->lock); }
return NULL;
}
int hints_next_root(struct iter_hints* hints, uint16_t* qclass)
int hints_next_root(struct iter_hints* hints, uint16_t* qclass, int nolock)
{
return name_tree_next_root(&hints->tree, qclass);
int ret;
/* lock_() calls are macros that could be nothing, surround in {} */
if(!nolock) { lock_rw_rdlock(&hints->lock); }
ret = name_tree_next_root(&hints->tree, qclass);
if(!nolock) { lock_rw_unlock(&hints->lock); }
return ret;
}
size_t
@@ -524,40 +562,63 @@ hints_get_mem(struct iter_hints* hints)
size_t s;
struct iter_hints_stub* p;
if(!hints) return 0;
lock_rw_rdlock(&hints->lock);
s = sizeof(*hints);
RBTREE_FOR(p, struct iter_hints_stub*, &hints->tree) {
s += sizeof(*p) + delegpt_get_mem(p->dp);
}
lock_rw_unlock(&hints->lock);
return s;
}
int
hints_add_stub(struct iter_hints* hints, uint16_t c, struct delegpt* dp,
int noprime)
int noprime, int nolock)
{
struct iter_hints_stub *z;
/* lock_() calls are macros that could be nothing, surround in {} */
if(!nolock) { lock_rw_wrlock(&hints->lock); }
if((z=(struct iter_hints_stub*)name_tree_find(&hints->tree,
dp->name, dp->namelen, dp->namelabs, c)) != NULL) {
(void)rbtree_delete(&hints->tree, &z->node);
hints_stub_free(z);
}
if(!hints_insert(hints, c, dp, noprime))
if(!hints_insert(hints, c, dp, noprime)) {
if(!nolock) { lock_rw_unlock(&hints->lock); }
return 0;
}
name_tree_init_parents(&hints->tree);
if(!nolock) { lock_rw_unlock(&hints->lock); }
return 1;
}
void
hints_delete_stub(struct iter_hints* hints, uint16_t c, uint8_t* nm)
hints_delete_stub(struct iter_hints* hints, uint16_t c, uint8_t* nm,
int nolock)
{
struct iter_hints_stub *z;
size_t len;
int labs = dname_count_size_labels(nm, &len);
/* lock_() calls are macros that could be nothing, surround in {} */
if(!nolock) { lock_rw_wrlock(&hints->lock); }
if(!(z=(struct iter_hints_stub*)name_tree_find(&hints->tree,
nm, len, labs, c)))
nm, len, labs, c))) {
if(!nolock) { lock_rw_unlock(&hints->lock); }
return; /* nothing to do */
}
(void)rbtree_delete(&hints->tree, &z->node);
hints_stub_free(z);
name_tree_init_parents(&hints->tree);
if(!nolock) { lock_rw_unlock(&hints->lock); }
}
void
hints_swap_tree(struct iter_hints* hints, struct iter_hints* data)
{
rbnode_type* oldroot = hints->tree.root;
size_t oldcount = hints->tree.count;
hints->tree.root = data->tree.root;
hints->tree.count = data->tree.count;
data->tree.root = oldroot;
data->tree.count = oldcount;
}
+56 -7
View File
@@ -43,6 +43,7 @@
#ifndef ITERATOR_ITER_HINTS_H
#define ITERATOR_ITER_HINTS_H
#include "util/storage/dnstree.h"
#include "util/locks.h"
struct iter_env;
struct config_file;
struct delegpt;
@@ -51,6 +52,10 @@ struct delegpt;
* Iterator hints structure
*/
struct iter_hints {
/** lock on the forwards tree.
* When grabbing both this lock and the anchors.lock, this lock
* is grabbed first. */
lock_rw_type lock;
/**
* Hints are stored in this tree. Sort order is specially chosen.
* first sorted on qclass. Then on dname in nsec-like order, so that
@@ -95,42 +100,70 @@ void hints_delete(struct iter_hints* hints);
int hints_apply_cfg(struct iter_hints* hints, struct config_file* cfg);
/**
* Find root hints for the given class.
* Find hints for the given class.
* The return value is contents of the hints structure.
* Caller should lock and unlock a readlock on the hints structure if nolock
* is set.
* Otherwise caller should unlock the readlock on the hints structure if a
* value was returned.
* @param hints: hint storage.
* @param qname: the qname that generated the delegation point.
* @param qclass: class for which root hints are requested. host order.
* @param nolock: Skip locking, locking is handled by the caller.
* @return: NULL if no hints, or a ptr to stored hints.
*/
struct delegpt* hints_lookup_root(struct iter_hints* hints, uint16_t qclass);
struct delegpt* hints_find(struct iter_hints* hints, uint8_t* qname,
uint16_t qclass, int nolock);
/**
* Same as hints_lookup, but for the root only.
* @param hints: hint storage.
* @param qclass: class for which root hints are requested. host order.
* @param nolock: Skip locking, locking is handled by the caller.
* @return: NULL if no hints, or a ptr to stored hints.
*/
struct delegpt* hints_find_root(struct iter_hints* hints,
uint16_t qclass, int nolock);
/**
* Find next root hints (to cycle through all root hints).
* Handles its own locking unless nolock is set. In that case the caller
* should lock and unlock a readlock on the hints structure.
* @param hints: hint storage
* @param qclass: class for which root hints are sought.
* 0 means give the first available root hints class.
* x means, give class x or a higher class if any.
* returns the found class in this variable.
* @param nolock: Skip locking, locking is handled by the caller.
* @return true if a root hint class is found.
* false if not root hint class is found (qclass may have been changed).
*/
int hints_next_root(struct iter_hints* hints, uint16_t* qclass);
int hints_next_root(struct iter_hints* hints, uint16_t* qclass, int nolock);
/**
* Given a qname/qclass combination, and the delegation point from the cache
* for this qname/qclass, determine if this combination indicates that a
* stub hint exists and must be primed.
* The return value is contents of the hints structure.
* Caller should lock and unlock a readlock on the hints structure if nolock
* is set.
* Otherwise caller should unlock the readlock on the hints structure if a
* value was returned.
*
* @param hints: hint storage.
* @param qname: The qname that generated the delegation point.
* @param qclass: The qclass that generated the delegation point.
* @param dp: The cache generated delegation point.
* @param nolock: Skip locking, locking is handled by the caller.
* @return: A priming delegation point if there is a stub hint that must
* be primed, otherwise null.
*/
struct iter_hints_stub* hints_lookup_stub(struct iter_hints* hints,
uint8_t* qname, uint16_t qclass, struct delegpt* dp);
struct iter_hints_stub* hints_lookup_stub(struct iter_hints* hints,
uint8_t* qname, uint16_t qclass, struct delegpt* dp, int nolock);
/**
* Get memory in use by hints
* Locks and unlocks the structure.
* @param hints: hint storage.
* @return bytes in use
*/
@@ -139,23 +172,39 @@ size_t hints_get_mem(struct iter_hints* hints);
/**
* Add stub to hints structure. For external use since it recalcs
* the tree parents.
* Handles its own locking unless nolock is set. In that case the caller
* should lock and unlock a writelock on the hints structure.
* @param hints: the hints data structure
* @param c: class of zone
* @param dp: delegation point with name and target nameservers for new
* hints stub. malloced.
* @param noprime: set noprime option to true or false on new hint stub.
* @param nolock: Skip locking, locking is handled by the caller.
* @return false on failure (out of memory);
*/
int hints_add_stub(struct iter_hints* hints, uint16_t c, struct delegpt* dp,
int noprime);
int noprime, int nolock);
/**
* Remove stub from hints structure. For external use since it
* recalcs the tree parents.
* Handles its own locking unless nolock is set. In that case the caller
* should lock and unlock a writelock on the hints structure.
* @param hints: the hints data structure
* @param c: class of stub zone
* @param nm: name of stub zone (in uncompressed wireformat).
* @param nolock: Skip locking, locking is handled by the caller.
*/
void hints_delete_stub(struct iter_hints* hints, uint16_t c, uint8_t* nm);
void hints_delete_stub(struct iter_hints* hints, uint16_t c,
uint8_t* nm, int nolock);
/**
* Swap internal tree with preallocated entries. Caller should manage
* the locks.
* @param hints: the hints data structure.
* @param data: the data structure used to take elements from. This contains
* the old elements on return.
*/
void hints_swap_tree(struct iter_hints* hints, struct iter_hints* data);
#endif /* ITERATOR_ITER_HINTS_H */
+2 -24
View File
@@ -207,28 +207,6 @@ size_t priv_get_mem(struct iter_priv* priv)
return sizeof(*priv) + regional_get_mem(priv->region);
}
/** remove RR from msgparse RRset, return true if rrset is entirely bad */
static int
remove_rr(const char* str, sldns_buffer* pkt, struct rrset_parse* rrset,
struct rr_parse* prev, struct rr_parse** rr, struct sockaddr_storage* addr, socklen_t addrlen)
{
if(verbosity >= VERB_QUERY && rrset->dname_len <= LDNS_MAX_DOMAINLEN && str) {
uint8_t buf[LDNS_MAX_DOMAINLEN+1];
dname_pkt_copy(pkt, buf, rrset->dname);
log_name_addr(VERB_QUERY, str, buf, addr, addrlen);
}
if(prev)
prev->next = (*rr)->next;
else rrset->rr_first = (*rr)->next;
if(rrset->rr_last == *rr)
rrset->rr_last = prev;
rrset->rr_count --;
rrset->size -= (*rr)->size;
/* rr struct still exists, but is unlinked, so that in the for loop
* the rr->next works fine to continue. */
return rrset->rr_count == 0;
}
int priv_rrset_bad(struct iter_priv* priv, sldns_buffer* pkt,
struct rrset_parse* rrset)
{
@@ -261,7 +239,7 @@ int priv_rrset_bad(struct iter_priv* priv, sldns_buffer* pkt,
INET_SIZE);
memmove(&addr, &sa, len);
if(priv_lookup_addr(priv, &addr, len)) {
if(remove_rr("sanitize: removing public name with private address", pkt, rrset, prev, &rr, &addr, len))
if(msgparse_rrset_remove_rr("sanitize: removing public name with private address", pkt, rrset, prev, rr, &addr, len))
return 1;
continue;
}
@@ -284,7 +262,7 @@ int priv_rrset_bad(struct iter_priv* priv, sldns_buffer* pkt,
INET6_SIZE);
memmove(&addr, &sa, len);
if(priv_lookup_addr(priv, &addr, len)) {
if(remove_rr("sanitize: removing public name with private address", pkt, rrset, prev, &rr, &addr, len))
if(msgparse_rrset_remove_rr("sanitize: removing public name with private address", pkt, rrset, prev, rr, &addr, len))
return 1;
continue;
}
+24 -2
View File
@@ -42,6 +42,7 @@
#include "config.h"
#include "iterator/iter_resptype.h"
#include "iterator/iter_delegpt.h"
#include "iterator/iterator.h"
#include "services/cache/dns.h"
#include "util/net_help.h"
#include "util/data/dname.h"
@@ -105,7 +106,8 @@ response_type_from_cache(struct dns_msg* msg,
enum response_type
response_type_from_server(int rdset,
struct dns_msg* msg, struct query_info* request, struct delegpt* dp)
struct dns_msg* msg, struct query_info* request, struct delegpt* dp,
int* empty_nodata_found)
{
uint8_t* origzone = (uint8_t*)"\000"; /* the default */
struct ub_packed_rrset_key* s;
@@ -113,7 +115,11 @@ response_type_from_server(int rdset,
if(!msg || !request)
return RESPONSE_TYPE_THROWAWAY;
/* If the TC flag is set, the response is incomplete. Too large to
* fit even in TCP or so. Discard it, it cannot be retrieved here. */
if((msg->rep->flags & BIT_TC))
return RESPONSE_TYPE_THROWAWAY;
/* If the message is NXDOMAIN, then it answers the question. */
if(FLAGS_GET_RCODE(msg->rep->flags) == LDNS_RCODE_NXDOMAIN) {
/* make sure its not recursive when we don't want it to */
@@ -280,6 +286,22 @@ response_type_from_server(int rdset,
/* If we've gotten this far, this is NOERROR/NODATA (which could
* be an entirely empty message) */
/* For entirely empty messages, try again, at first, then accept
* it it happens more. A regular noerror/nodata response has a soa
* negative ttl value in the authority section. This makes it try
* again at another authority. And decides between storing a 5 second
* empty message or a 5 second servfail response. */
if(msg->rep->an_numrrsets == 0 && msg->rep->ns_numrrsets == 0 &&
msg->rep->ar_numrrsets == 0) {
if(empty_nodata_found) {
/* detect as throwaway at first, but accept later. */
(*empty_nodata_found)++;
if(*empty_nodata_found < EMPTY_NODATA_RETRY_COUNT)
return RESPONSE_TYPE_THROWAWAY;
return RESPONSE_TYPE_ANSWER;
}
return RESPONSE_TYPE_ANSWER;
}
/* check if recursive answer; saying it has empty cache */
if( (msg->rep->flags&BIT_RA) && !(msg->rep->flags&BIT_AA) && !rdset)
return RESPONSE_TYPE_REC_LAME;
+3 -1
View File
@@ -119,9 +119,11 @@ enum response_type response_type_from_cache(struct dns_msg* msg,
* @param request: the request that generated the response.
* @param dp: The delegation point that was being queried
* when the response was returned.
* @param empty_nodata_found: flag to keep track of empty nodata detection.
* @return the response type (CNAME or ANSWER).
*/
enum response_type response_type_from_server(int rdset,
struct dns_msg* msg, struct query_info* request, struct delegpt* dp);
struct dns_msg* msg, struct query_info* request, struct delegpt* dp,
int* empty_nodata_found);
#endif /* ITERATOR_ITER_RESPTYPE_H */
+268 -9
View File
@@ -283,7 +283,8 @@ synth_cname_rrset(uint8_t** sname, size_t* snamelen, uint8_t* alias,
sizeof(uint32_t)+sizeof(uint16_t)+aliaslen);
if(!cn->rr_first->ttl_data)
return NULL;
sldns_write_uint32(cn->rr_first->ttl_data, 0); /* TTL = 0 */
memmove(cn->rr_first->ttl_data, rrset->rr_first->ttl_data,
sizeof(uint32_t)); /* RFC6672: synth CNAME TTL == DNAME TTL */
sldns_write_uint16(cn->rr_first->ttl_data+4, aliaslen);
memmove(cn->rr_first->ttl_data+6, alias, aliaslen);
cn->rr_first->size = sizeof(uint16_t)+aliaslen;
@@ -346,6 +347,67 @@ soa_in_auth(struct msg_parse* msg)
return 0;
}
/** Check if type is allowed in the authority section */
static int
type_allowed_in_authority_section(uint16_t tp)
{
if(tp == LDNS_RR_TYPE_SOA || tp == LDNS_RR_TYPE_NS ||
tp == LDNS_RR_TYPE_DS || tp == LDNS_RR_TYPE_NSEC ||
tp == LDNS_RR_TYPE_NSEC3)
return 1;
return 0;
}
/** Check if type is allowed in the additional section */
static int
type_allowed_in_additional_section(uint16_t tp)
{
if(tp == LDNS_RR_TYPE_A || tp == LDNS_RR_TYPE_AAAA)
return 1;
return 0;
}
/** Shorten RRset */
static void
shorten_rrset(sldns_buffer* pkt, struct rrset_parse* rrset, int count)
{
/* The too large NS RRset is shortened. This is so that too large
* content does not overwhelm the cache. It may make the rrset
* bogus if it was signed, and then the domain is not resolved any
* more, that is okay, the NS RRset was too large. During a referral
* it can be shortened and then the first part of the list could
* be used to resolve. The scrub continues to disallow glue for the
* removed nameserver RRs and removes that too. Because the glue
* is not marked as okay, since the RRs have been removed here. */
int i;
struct rr_parse* rr = rrset->rr_first, *prev = NULL;
if(!rr)
return;
for(i=0; i<count; i++) {
prev = rr;
rr = rr->next;
if(!rr)
return; /* The RRset is already short. */
}
if(verbosity >= VERB_QUERY
&& rrset->dname_len <= LDNS_MAX_DOMAINLEN) {
uint8_t buf[LDNS_MAX_DOMAINLEN+1];
dname_pkt_copy(pkt, buf, rrset->dname);
log_nametypeclass(VERB_QUERY, "normalize: shorten RRset:", buf,
rrset->type, ntohs(rrset->rrset_class));
}
/* remove further rrs */
rrset->rr_last = prev;
rrset->rr_count = count;
while(rr) {
rrset->size -= rr->size;
rr = rr->next;
}
if(rrset->rr_last)
rrset->rr_last->next = NULL;
else rrset->rr_first = NULL;
}
/**
* This routine normalizes a response. This includes removing "irrelevant"
* records from the answer and additional sections and (re)synthesizing
@@ -355,18 +417,23 @@ soa_in_auth(struct msg_parse* msg)
* @param msg: msg to normalize.
* @param qinfo: original query.
* @param region: where to allocate synthesized CNAMEs.
* @param env: module env with config options.
* @param zonename: name of server zone.
* @return 0 on error.
*/
static int
scrub_normalize(sldns_buffer* pkt, struct msg_parse* msg,
struct query_info* qinfo, struct regional* region)
struct query_info* qinfo, struct regional* region,
struct module_env* env, uint8_t* zonename)
{
uint8_t* sname = qinfo->qname;
size_t snamelen = qinfo->qname_len;
struct rrset_parse* rrset, *prev, *nsset=NULL;
int cname_length = 0; /* number of CNAMEs, or DNAMEs */
if(FLAGS_GET_RCODE(msg->flags) != LDNS_RCODE_NOERROR &&
FLAGS_GET_RCODE(msg->flags) != LDNS_RCODE_NXDOMAIN)
FLAGS_GET_RCODE(msg->flags) != LDNS_RCODE_NXDOMAIN &&
FLAGS_GET_RCODE(msg->flags) != LDNS_RCODE_YXDOMAIN)
return 1;
/* For the ANSWER section, remove all "irrelevant" records and add
@@ -378,6 +445,16 @@ scrub_normalize(sldns_buffer* pkt, struct msg_parse* msg,
prev = NULL;
rrset = msg->rrset_first;
while(rrset && rrset->section == LDNS_SECTION_ANSWER) {
if(cname_length > env->cfg->iter_scrub_cname) {
/* Too many CNAMEs, or DNAMEs, from the authority
* server, scrub down the length to something
* shorter. This deletes everything after the limit
* is reached. The iterator is going to look up
* the content one by one anyway. */
remove_rrset("normalize: removing because too many cnames:",
pkt, msg, prev, &rrset);
continue;
}
if(rrset->type == LDNS_RR_TYPE_DNAME &&
pkt_strict_sub(pkt, sname, rrset->dname)) {
/* check if next rrset is correct CNAME. else,
@@ -395,8 +472,14 @@ scrub_normalize(sldns_buffer* pkt, struct msg_parse* msg,
&aliaslen, pkt)) {
verbose(VERB_ALGO, "synthesized CNAME "
"too long");
if(FLAGS_GET_RCODE(msg->flags) == LDNS_RCODE_YXDOMAIN) {
prev = rrset;
rrset = rrset->rrset_all_next;
continue;
}
return 0;
}
cname_length++;
if(nx && nx->type == LDNS_RR_TYPE_CNAME &&
dname_pkt_compare(pkt, sname, nx->dname) == 0) {
/* check next cname */
@@ -437,6 +520,7 @@ scrub_normalize(sldns_buffer* pkt, struct msg_parse* msg,
if(rrset->type == LDNS_RR_TYPE_CNAME) {
struct rrset_parse* nx = rrset->rrset_all_next;
uint8_t* oldsname = sname;
cname_length++;
/* see if the next one is a DNAME, if so, swap them */
if(nx && nx->section == LDNS_SECTION_ANSWER &&
nx->type == LDNS_RR_TYPE_DNAME &&
@@ -484,6 +568,10 @@ scrub_normalize(sldns_buffer* pkt, struct msg_parse* msg,
LDNS_SECTION_ANSWER &&
dname_pkt_compare(pkt, oldsname,
rrset->dname) == 0) {
if(rrset->type == LDNS_RR_TYPE_NS &&
rrset->rr_count > env->cfg->iter_scrub_ns) {
shorten_rrset(pkt, rrset, env->cfg->iter_scrub_ns);
}
prev = rrset;
rrset = rrset->rrset_all_next;
}
@@ -499,6 +587,11 @@ scrub_normalize(sldns_buffer* pkt, struct msg_parse* msg,
continue;
}
if(rrset->type == LDNS_RR_TYPE_NS &&
rrset->rr_count > env->cfg->iter_scrub_ns) {
shorten_rrset(pkt, rrset, env->cfg->iter_scrub_ns);
}
/* Mark the additional names from relevant rrset as OK. */
/* only for RRsets that match the query name, other ones
* will be removed by sanitize, so no additional for them */
@@ -511,6 +604,7 @@ scrub_normalize(sldns_buffer* pkt, struct msg_parse* msg,
/* Mark additional names from AUTHORITY */
while(rrset && rrset->section == LDNS_SECTION_AUTHORITY) {
/* protect internals of recursor by making sure to del these */
if(rrset->type==LDNS_RR_TYPE_DNAME ||
rrset->type==LDNS_RR_TYPE_CNAME ||
rrset->type==LDNS_RR_TYPE_A ||
@@ -519,6 +613,13 @@ scrub_normalize(sldns_buffer* pkt, struct msg_parse* msg,
"RRset:", pkt, msg, prev, &rrset);
continue;
}
/* Allowed list of types in the authority section */
if(env->cfg->harden_unknown_additional &&
!type_allowed_in_authority_section(rrset->type)) {
remove_rrset("normalize: removing irrelevant "
"RRset:", pkt, msg, prev, &rrset);
continue;
}
/* only one NS set allowed in authority section */
if(rrset->type==LDNS_RR_TYPE_NS) {
/* NS set must be pertinent to the query */
@@ -540,6 +641,45 @@ scrub_normalize(sldns_buffer* pkt, struct msg_parse* msg,
"RRset:", pkt, msg, prev, &rrset);
continue;
}
/* If the NS set is a promiscuous NS set, scrub that
* to remove potential for poisonous contents that
* affects other names in the same zone. Remove
* promiscuous NS sets in positive answers, that
* thus have records in the answer section. Nodata
* and nxdomain promiscuous NS sets have been removed
* already. Since the NS rrset is scrubbed, its
* address records are also not marked to be allowed
* and are removed later. */
if(FLAGS_GET_RCODE(msg->flags) == LDNS_RCODE_NOERROR &&
msg->an_rrsets != 0 &&
env->cfg->iter_scrub_promiscuous) {
remove_rrset("normalize: removing promiscuous "
"RRset:", pkt, msg, prev, &rrset);
continue;
}
/* Also delete promiscuous NS for other RCODEs */
if(FLAGS_GET_RCODE(msg->flags) != LDNS_RCODE_NOERROR
&& env->cfg->iter_scrub_promiscuous) {
remove_rrset("normalize: removing promiscuous "
"RRset:", pkt, msg, prev, &rrset);
continue;
}
/* Also delete promiscuous NS for NOERROR with nodata
* for authoritative answers, not for delegations.
* NOERROR with an_rrsets!=0 already handled.
* Also NOERROR and soa_in_auth already handled.
* NOERROR with an_rrsets==0, and not a referral.
* referral is (NS not the zonename, noSOA).
*/
if(FLAGS_GET_RCODE(msg->flags) == LDNS_RCODE_NOERROR
&& msg->an_rrsets == 0
&& !(dname_pkt_compare(pkt, rrset->dname,
zonename) != 0 && !soa_in_auth(msg))
&& env->cfg->iter_scrub_promiscuous) {
remove_rrset("normalize: removing promiscuous "
"RRset:", pkt, msg, prev, &rrset);
continue;
}
if(nsset == NULL) {
nsset = rrset;
} else {
@@ -547,6 +687,25 @@ scrub_normalize(sldns_buffer* pkt, struct msg_parse* msg,
"RRset:", pkt, msg, prev, &rrset);
continue;
}
if(rrset->rr_count > env->cfg->iter_scrub_ns) {
/* If this is not a referral, and the NS RRset
* is signed, then remove it entirely, so
* that when it becomes bogus it does not
* make the message that is otherwise fine
* into a bogus message. */
if(!(msg->an_rrsets == 0 &&
FLAGS_GET_RCODE(msg->flags) ==
LDNS_RCODE_NOERROR &&
!soa_in_auth(msg) &&
!(msg->flags & BIT_AA)) &&
rrset->rrsig_count != 0) {
remove_rrset("normalize: removing too large NS "
"RRset:", pkt, msg, prev, &rrset);
continue;
} else {
shorten_rrset(pkt, rrset, env->cfg->iter_scrub_ns);
}
}
}
/* if this is type DS and we query for type DS we just got
* a referral answer for our type DS query, fix packet */
@@ -576,7 +735,6 @@ scrub_normalize(sldns_buffer* pkt, struct msg_parse* msg,
* found in ANSWER and AUTHORITY. */
/* These records have not been marked OK previously */
while(rrset && rrset->section == LDNS_SECTION_ADDITIONAL) {
/* FIXME: what about other types? */
if(rrset->type==LDNS_RR_TYPE_A ||
rrset->type==LDNS_RR_TYPE_AAAA)
{
@@ -589,6 +747,7 @@ scrub_normalize(sldns_buffer* pkt, struct msg_parse* msg,
continue;
}
}
/* protect internals of recursor by making sure to del these */
if(rrset->type==LDNS_RR_TYPE_DNAME ||
rrset->type==LDNS_RR_TYPE_CNAME ||
rrset->type==LDNS_RR_TYPE_NS) {
@@ -596,6 +755,13 @@ scrub_normalize(sldns_buffer* pkt, struct msg_parse* msg,
"RRset:", pkt, msg, prev, &rrset);
continue;
}
/* Allowed list of types in the additional section */
if(env->cfg->harden_unknown_additional &&
!type_allowed_in_additional_section(rrset->type)) {
remove_rrset("normalize: removing irrelevant "
"RRset:", pkt, msg, prev, &rrset);
continue;
}
prev = rrset;
rrset = rrset->rrset_all_next;
}
@@ -679,6 +845,56 @@ static int sanitize_nsec_is_overreach(sldns_buffer* pkt,
return 0;
}
/** Remove individual RRs, if the length is wrong. Returns true if the RRset
* has been removed. */
static int
scrub_sanitize_rr_length(sldns_buffer* pkt, struct msg_parse* msg,
struct rrset_parse* prev, struct rrset_parse** rrset, int* added_ede,
struct module_qstate* qstate)
{
struct rr_parse* rr, *rr_prev = NULL;
for(rr = (*rrset)->rr_first; rr; rr = rr->next) {
/* Sanity check for length of records
* An A record should be 6 bytes only
* (2 bytes for length and 4 for IPv4 addr)*/
if((*rrset)->type == LDNS_RR_TYPE_A && rr->size != 6 ) {
if(!*added_ede) {
*added_ede = 1;
errinf_ede(qstate, "sanitize: records of inappropriate length have been removed.",
LDNS_EDE_OTHER);
}
if(msgparse_rrset_remove_rr("sanitize: removing type A RR of inappropriate length:",
pkt, *rrset, rr_prev, rr, NULL, 0)) {
remove_rrset("sanitize: removing type A RRset of inappropriate length:",
pkt, msg, prev, rrset);
return 1;
}
continue;
}
/* Sanity check for length of records
* An AAAA record should be 18 bytes only
* (2 bytes for length and 16 for IPv6 addr)*/
if((*rrset)->type == LDNS_RR_TYPE_AAAA && rr->size != 18 ) {
if(!*added_ede) {
*added_ede = 1;
errinf_ede(qstate, "sanitize: records of inappropriate length have been removed.",
LDNS_EDE_OTHER);
}
if(msgparse_rrset_remove_rr("sanitize: removing type AAAA RR of inappropriate length:",
pkt, *rrset, rr_prev, rr, NULL, 0)) {
remove_rrset("sanitize: removing type AAAA RRset of inappropriate length:",
pkt, msg, prev, rrset);
return 1;
}
continue;
}
rr_prev = rr;
}
return 0;
}
/**
* Given a response event, remove suspect RRsets from the response.
* "Suspect" rrsets are potentially poison. Note that this routine expects
@@ -691,15 +907,18 @@ static int sanitize_nsec_is_overreach(sldns_buffer* pkt,
* @param zonename: name of server zone.
* @param env: module environment with config and cache.
* @param ie: iterator environment with private address data.
* @param qstate: for setting errinf for EDE error messages.
* @return 0 on error.
*/
static int
scrub_sanitize(sldns_buffer* pkt, struct msg_parse* msg,
struct query_info* qinfo, uint8_t* zonename, struct module_env* env,
struct iter_env* ie)
struct iter_env* ie, struct module_qstate* qstate)
{
int del_addi = 0; /* if additional-holding rrsets are deleted, we
do not trust the normalized additional-A-AAAA any more */
uint8_t* ns_rrset_dname = NULL;
int added_rrlen_ede = 0;
struct rrset_parse* rrset, *prev;
prev = NULL;
rrset = msg->rrset_first;
@@ -744,6 +963,14 @@ scrub_sanitize(sldns_buffer* pkt, struct msg_parse* msg,
rrset = msg->rrset_first;
while(rrset) {
/* Sanity check for length of records */
if(rrset->type == LDNS_RR_TYPE_A ||
rrset->type == LDNS_RR_TYPE_AAAA) {
if(scrub_sanitize_rr_length(pkt, msg, prev, &rrset,
&added_rrlen_ede, qstate))
continue;
}
/* remove private addresses */
if( (rrset->type == LDNS_RR_TYPE_A ||
rrset->type == LDNS_RR_TYPE_AAAA)) {
@@ -796,6 +1023,16 @@ scrub_sanitize(sldns_buffer* pkt, struct msg_parse* msg,
continue;
}
}
if(rrset->type == LDNS_RR_TYPE_NS &&
(rrset->section == LDNS_SECTION_AUTHORITY ||
rrset->section == LDNS_SECTION_ANSWER)) {
/* If the type is NS, and we're in the
* answer or authority section, then
* store the dname so we can check
* against the glue records
* further down */
ns_rrset_dname = rrset->dname;
}
if(del_addi && rrset->section == LDNS_SECTION_ADDITIONAL) {
remove_rrset("sanitize: removing potential "
"poison reference RRset:", pkt, msg, prev, &rrset);
@@ -808,6 +1045,26 @@ scrub_sanitize(sldns_buffer* pkt, struct msg_parse* msg,
"RRset:", pkt, msg, prev, &rrset);
continue;
}
if(env->cfg->harden_unverified_glue && ns_rrset_dname &&
rrset->section == LDNS_SECTION_ADDITIONAL &&
(rrset->type == LDNS_RR_TYPE_A || rrset->type == LDNS_RR_TYPE_AAAA) &&
!pkt_strict_sub(pkt, rrset->dname, ns_rrset_dname)) {
/* We're in the additional section, looking
* at an A/AAAA rrset, have a previous
* delegation point and we notice that
* the glue records are NOT for strict
* subdomains of the delegation. So set a
* flag, recompute the hash for the rrset
* and write the A/AAAA record to cache.
* It'll be retrieved if we can't separately
* resolve the glue */
rrset->flags = PACKED_RRSET_UNVERIFIED_GLUE;
rrset->hash = pkt_hash_rrset(pkt, rrset->dname, rrset->type, rrset->rrset_class, rrset->flags);
store_rrset(pkt, msg, env, rrset);
remove_rrset("sanitize: storing potential "
"unverified glue reference RRset:", pkt, msg, prev, &rrset);
continue;
}
prev = rrset;
rrset = rrset->rrset_all_next;
}
@@ -817,7 +1074,8 @@ scrub_sanitize(sldns_buffer* pkt, struct msg_parse* msg,
int
scrub_message(sldns_buffer* pkt, struct msg_parse* msg,
struct query_info* qinfo, uint8_t* zonename, struct regional* region,
struct module_env* env, struct iter_env* ie)
struct module_env* env, struct module_qstate* qstate,
struct iter_env* ie)
{
/* basic sanity checks */
log_nametypeclass(VERB_ALGO, "scrub for", zonename, LDNS_RR_TYPE_NS,
@@ -832,7 +1090,8 @@ scrub_message(sldns_buffer* pkt, struct msg_parse* msg,
/* this is not required for basic operation but is a forgery
* resistance (security) feature */
if((FLAGS_GET_RCODE(msg->flags) == LDNS_RCODE_NOERROR ||
FLAGS_GET_RCODE(msg->flags) == LDNS_RCODE_NXDOMAIN) &&
FLAGS_GET_RCODE(msg->flags) == LDNS_RCODE_NXDOMAIN ||
FLAGS_GET_RCODE(msg->flags) == LDNS_RCODE_YXDOMAIN) &&
msg->qdcount == 0)
return 0;
@@ -846,10 +1105,10 @@ scrub_message(sldns_buffer* pkt, struct msg_parse* msg,
}
/* normalize the response, this cleans up the additional. */
if(!scrub_normalize(pkt, msg, qinfo, region))
if(!scrub_normalize(pkt, msg, qinfo, region, env, zonename))
return 0;
/* delete all out-of-zone information */
if(!scrub_sanitize(pkt, msg, qinfo, zonename, env, ie))
if(!scrub_sanitize(pkt, msg, qinfo, zonename, env, ie, qstate))
return 0;
return 1;
}
+4 -1
View File
@@ -48,6 +48,7 @@ struct query_info;
struct regional;
struct module_env;
struct iter_env;
struct module_qstate;
/**
* Cleanup the passed dns message.
@@ -59,11 +60,13 @@ struct iter_env;
* Used to determine out of bailiwick information.
* @param regional: where to allocate (new) parts of the message.
* @param env: module environment with config settings and cache.
* @param qstate: for setting errinf for EDE error messages.
* @param ie: iterator module environment data.
* @return: false if the message is total waste. true if scrubbed with success.
*/
int scrub_message(struct sldns_buffer* pkt, struct msg_parse* msg,
struct query_info* qinfo, uint8_t* zonename, struct regional* regional,
struct module_env* env, struct iter_env* ie);
struct module_env* env, struct module_qstate* qstate,
struct iter_env* ie);
#endif /* ITERATOR_ITER_SCRUB_H */
+241 -49
View File
@@ -70,45 +70,80 @@
/** time when nameserver glue is said to be 'recent' */
#define SUSPICION_RECENT_EXPIRY 86400
/** penalty to validation failed blacklisted IPs */
#define BLACKLIST_PENALTY (USEFUL_SERVER_TOP_TIMEOUT*4)
/** if NAT64 is enabled and no NAT64 prefix is configured, first fall back to
* DNS64 prefix. If that is not configured, fall back to this default value.
*/
static const char DEFAULT_NAT64_PREFIX[] = "64:ff9b::/96";
/** fillup fetch policy array */
static void
fetch_fill(struct iter_env* ie, const char* str)
static int
fetch_fill(int* target_fetch_policy, int max_dependency_depth, const char* str)
{
char* s = (char*)str, *e;
int i;
for(i=0; i<ie->max_dependency_depth+1; i++) {
ie->target_fetch_policy[i] = strtol(s, &e, 10);
if(s == e)
fatal_exit("cannot parse fetch policy number %s", s);
for(i=0; i<max_dependency_depth+1; i++) {
target_fetch_policy[i] = strtol(s, &e, 10);
if(s == e) {
log_err("cannot parse fetch policy number %s", s);
return 0;
}
s = e;
}
return 1;
}
/** Read config string that represents the target fetch policy */
static int
read_fetch_policy(struct iter_env* ie, const char* str)
int
read_fetch_policy(int** target_fetch_policy, int* max_dependency_depth,
const char* str)
{
int count = cfg_count_numbers(str);
if(count < 1) {
log_err("Cannot parse target fetch policy: \"%s\"", str);
return 0;
}
ie->max_dependency_depth = count - 1;
ie->target_fetch_policy = (int*)calloc(
(size_t)ie->max_dependency_depth+1, sizeof(int));
if(!ie->target_fetch_policy) {
*max_dependency_depth = count - 1;
*target_fetch_policy = (int*)calloc(
(size_t)(*max_dependency_depth)+1, sizeof(int));
if(!*target_fetch_policy) {
log_err("alloc fetch policy: out of memory");
return 0;
}
fetch_fill(ie, str);
if(!fetch_fill(*target_fetch_policy, *max_dependency_depth, str))
return 0;
return 1;
}
/** apply config caps whitelist items to name tree */
static int
struct rbtree_type*
caps_white_create(void)
{
struct rbtree_type* caps_white = rbtree_create(name_tree_compare);
if(!caps_white)
log_err("out of memory");
return caps_white;
}
/** delete caps_whitelist element */
static void
caps_free(struct rbnode_type* n, void* ATTR_UNUSED(d))
{
if(n) {
free(((struct name_tree_node*)n)->name);
free(n);
}
}
void
caps_white_delete(struct rbtree_type* caps_white)
{
if(!caps_white)
return;
traverse_postorder(caps_white, caps_free, NULL);
free(caps_white);
}
int
caps_white_apply_cfg(rbtree_type* ntree, struct config_file* cfg)
{
struct config_strlist* p;
@@ -141,12 +176,42 @@ caps_white_apply_cfg(rbtree_type* ntree, struct config_file* cfg)
return 1;
}
int
nat64_apply_cfg(struct iter_nat64* nat64, struct config_file* cfg)
{
const char *nat64_prefix;
nat64_prefix = cfg->nat64_prefix;
if(!nat64_prefix)
nat64_prefix = cfg->dns64_prefix;
if(!nat64_prefix)
nat64_prefix = DEFAULT_NAT64_PREFIX;
if(!netblockstrtoaddr(nat64_prefix, 0, &nat64->nat64_prefix_addr,
&nat64->nat64_prefix_addrlen, &nat64->nat64_prefix_net)) {
log_err("cannot parse nat64-prefix netblock: %s", nat64_prefix);
return 0;
}
if(!addr_is_ip6(&nat64->nat64_prefix_addr,
nat64->nat64_prefix_addrlen)) {
log_err("nat64-prefix is not IPv6: %s", cfg->nat64_prefix);
return 0;
}
if(!prefixnet_is_nat64(nat64->nat64_prefix_net)) {
log_err("nat64-prefix length it not 32, 40, 48, 56, 64 or 96: %s",
nat64_prefix);
return 0;
}
nat64->use_nat64 = cfg->do_nat64;
return 1;
}
int
iter_apply_cfg(struct iter_env* iter_env, struct config_file* cfg)
{
int i;
/* target fetch policy */
if(!read_fetch_policy(iter_env, cfg->target_fetch_policy))
if(!read_fetch_policy(&iter_env->target_fetch_policy,
&iter_env->max_dependency_depth, cfg->target_fetch_policy))
return 0;
for(i=0; i<iter_env->max_dependency_depth+1; i++)
verbose(VERB_QUERY, "target fetch policy for level %d is %d",
@@ -166,7 +231,7 @@ iter_apply_cfg(struct iter_env* iter_env, struct config_file* cfg)
}
if(cfg->caps_whitelist) {
if(!iter_env->caps_white)
iter_env->caps_white = rbtree_create(name_tree_compare);
iter_env->caps_white = caps_white_create();
if(!iter_env->caps_white || !caps_white_apply_cfg(
iter_env->caps_white, cfg)) {
log_err("Could not set capsforid whitelist");
@@ -174,9 +239,17 @@ iter_apply_cfg(struct iter_env* iter_env, struct config_file* cfg)
}
}
if(!nat64_apply_cfg(&iter_env->nat64, cfg)) {
log_err("Could not setup nat64");
return 0;
}
iter_env->supports_ipv6 = cfg->do_ip6;
iter_env->supports_ipv4 = cfg->do_ip4;
iter_env->outbound_msg_retry = cfg->outbound_msg_retry;
iter_env->max_sent_count = cfg->max_sent_count;
iter_env->max_query_restarts = cfg->max_query_restarts;
return 1;
}
@@ -240,7 +313,8 @@ iter_filter_unsuitable(struct iter_env* iter_env, struct module_env* env,
if(!iter_env->supports_ipv6 && addr_is_ip6(&a->addr, a->addrlen)) {
return -1; /* there is no ip6 available */
}
if(!iter_env->supports_ipv4 && !addr_is_ip6(&a->addr, a->addrlen)) {
if(!iter_env->supports_ipv4 && !iter_env->nat64.use_nat64 &&
!addr_is_ip6(&a->addr, a->addrlen)) {
return -1; /* there is no ip4 available */
}
/* check lameness - need zone , class info */
@@ -248,9 +322,10 @@ iter_filter_unsuitable(struct iter_env* iter_env, struct module_env* env,
name, namelen, qtype, &lame, &dnsseclame, &reclame,
&rtt, now)) {
log_addr(VERB_ALGO, "servselect", &a->addr, a->addrlen);
verbose(VERB_ALGO, " rtt=%d%s%s%s%s", rtt,
verbose(VERB_ALGO, " rtt=%d%s%s%s%s%s", rtt,
lame?" LAME":"",
dnsseclame?" DNSSEC_LAME":"",
a->dnsseclame?" ADDR_DNSSEC_LAME":"",
reclame?" REC_LAME":"",
a->lame?" ADDR_LAME":"");
if(lame)
@@ -367,6 +442,7 @@ iter_filter_order(struct iter_env* iter_env, struct module_env* env,
struct sock_list* blacklist, time_t prefetch)
{
int got_num = 0, low_rtt = 0, swap_to_front, rtt_band = RTT_BAND, nth;
int alllame = 0;
size_t num_results;
struct delegpt_addr* a, *n, *prev=NULL;
@@ -376,7 +452,10 @@ iter_filter_order(struct iter_env* iter_env, struct module_env* env,
if(got_num == 0)
return 0;
if(low_rtt >= USEFUL_SERVER_TOP_TIMEOUT &&
(delegpt_count_missing_targets(dp) > 0 || open_target > 0)) {
/* If all missing (or not fully resolved) targets are lame,
* then use the remaining lame address. */
((delegpt_count_missing_targets(dp, &alllame) > 0 && !alllame) ||
open_target > 0)) {
verbose(VERB_ALGO, "Bad choices, trying to get more choice");
return 0; /* we want more choice. The best choice is a bad one.
return 0 to force the caller to fetch more */
@@ -657,10 +736,11 @@ dns_copy_msg(struct dns_msg* from, struct regional* region)
void
iter_dns_store(struct module_env* env, struct query_info* msgqinf,
struct reply_info* msgrep, int is_referral, time_t leeway, int pside,
struct regional* region, uint16_t flags)
struct regional* region, uint16_t flags, time_t qstarttime,
int is_valrec)
{
if(!dns_cache_store(env, msgqinf, msgrep, is_referral, leeway,
pside, region, flags))
pside, region, flags, qstarttime, is_valrec))
log_err("out of memory: cannot store data in cache");
}
@@ -743,9 +823,15 @@ iter_mark_pside_cycle_targets(struct module_qstate* qstate, struct delegpt* dp)
int
iter_dp_is_useless(struct query_info* qinfo, uint16_t qflags,
struct delegpt* dp)
struct delegpt* dp, int supports_ipv4, int supports_ipv6,
int use_nat64)
{
struct delegpt_ns* ns;
struct delegpt_addr* a;
if(supports_ipv6 && use_nat64)
supports_ipv4 = 1;
/* check:
* o RD qflag is on.
* o no addresses are provided.
@@ -758,13 +844,24 @@ iter_dp_is_useless(struct query_info* qinfo, uint16_t qflags,
*/
if(!(qflags&BIT_RD))
return 0;
/* either available or unused targets */
if(dp->usable_list || dp->result_list)
return 0;
/* either available or unused targets,
* if they exist, the dp is not useless. */
for(a = dp->usable_list; a; a = a->next_usable) {
if(!addr_is_ip6(&a->addr, a->addrlen) && supports_ipv4)
return 0;
else if(addr_is_ip6(&a->addr, a->addrlen) && supports_ipv6)
return 0;
}
for(a = dp->result_list; a; a = a->next_result) {
if(!addr_is_ip6(&a->addr, a->addrlen) && supports_ipv4)
return 0;
else if(addr_is_ip6(&a->addr, a->addrlen) && supports_ipv6)
return 0;
}
/* see if query is for one of the nameservers, which is glue */
if( (qinfo->qtype == LDNS_RR_TYPE_A ||
qinfo->qtype == LDNS_RR_TYPE_AAAA) &&
if( ((qinfo->qtype == LDNS_RR_TYPE_A && supports_ipv4) ||
(qinfo->qtype == LDNS_RR_TYPE_AAAA && supports_ipv6)) &&
dname_subdomain_c(qinfo->qname, dp->name) &&
delegpt_find_ns(dp, qinfo->qname, qinfo->qname_len))
return 1;
@@ -1195,6 +1292,9 @@ int iter_lookup_parent_glue_from_cache(struct module_env* env,
struct delegpt_ns* ns;
size_t num = delegpt_count_targets(dp);
for(ns = dp->nslist; ns; ns = ns->next) {
if(ns->cache_lookup_count > ITERATOR_NAME_CACHELOOKUP_MAX_PSIDE)
continue;
ns->cache_lookup_count++;
/* get cached parentside A */
akey = rrset_cache_lookup(env->rrset_cache, ns->name,
ns->namelen, LDNS_RR_TYPE_A, qinfo->qclass,
@@ -1229,8 +1329,17 @@ iter_get_next_root(struct iter_hints* hints, struct iter_forwards* fwd,
uint16_t* c)
{
uint16_t c1 = *c, c2 = *c;
int r1 = hints_next_root(hints, &c1);
int r2 = forwards_next_root(fwd, &c2);
int r1, r2;
int nolock = 1;
/* prelock both forwards and hints for atomic read. */
lock_rw_rdlock(&fwd->lock);
lock_rw_rdlock(&hints->lock);
r1 = hints_next_root(hints, &c1, nolock);
r2 = forwards_next_root(fwd, &c2, nolock);
lock_rw_unlock(&fwd->lock);
lock_rw_unlock(&hints->lock);
if(!r1 && !r2) /* got none, end of list */
return 0;
else if(!r1) /* got one, return that */
@@ -1291,8 +1400,7 @@ void iter_dec_attempts(struct delegpt* dp, int d, int outbound_msg_retry)
for(a=dp->target_list; a; a = a->next_target) {
if(a->attempts >= outbound_msg_retry) {
/* add back to result list */
a->next_result = dp->result_list;
dp->result_list = a;
delegpt_add_to_result_list(dp, a);
}
if(a->attempts > d)
a->attempts -= d;
@@ -1396,15 +1504,21 @@ int iter_dp_cangodown(struct query_info* qinfo, struct delegpt* dp)
int
iter_stub_fwd_no_cache(struct module_qstate *qstate, struct query_info *qinf,
uint8_t** retdpname, size_t* retdpnamelen)
uint8_t** retdpname, size_t* retdpnamelen, uint8_t* dpname_storage,
size_t dpname_storage_len)
{
struct iter_hints_stub *stub;
struct delegpt *dp;
int nolock = 1;
/* Check for stub. */
/* Lock both forwards and hints for atomic read. */
lock_rw_rdlock(&qstate->env->fwds->lock);
lock_rw_rdlock(&qstate->env->hints->lock);
stub = hints_lookup_stub(qstate->env->hints, qinf->qname,
qinf->qclass, NULL);
dp = forwards_lookup(qstate->env->fwds, qinf->qname, qinf->qclass);
qinf->qclass, NULL, nolock);
dp = forwards_lookup(qstate->env->fwds, qinf->qname, qinf->qclass,
nolock);
/* see if forward or stub is more pertinent */
if(stub && stub->dp && dp) {
@@ -1418,35 +1532,62 @@ iter_stub_fwd_no_cache(struct module_qstate *qstate, struct query_info *qinf,
/* check stub */
if (stub != NULL && stub->dp != NULL) {
if(stub->dp->no_cache) {
char qname[255+1];
char dpname[255+1];
enum verbosity_value level = VERB_ALGO;
int stub_no_cache = stub->dp->no_cache;
lock_rw_unlock(&qstate->env->fwds->lock);
if(verbosity >= level && stub_no_cache) {
char qname[LDNS_MAX_DOMAINLEN];
char dpname[LDNS_MAX_DOMAINLEN];
dname_str(qinf->qname, qname);
dname_str(stub->dp->name, dpname);
verbose(VERB_ALGO, "stub for %s %s has no_cache", qname, dpname);
verbose(level, "stub for %s %s has no_cache", qname, dpname);
}
if(retdpname) {
*retdpname = stub->dp->name;
if(stub->dp->namelen > dpname_storage_len) {
verbose(VERB_ALGO, "no cache stub dpname too long");
lock_rw_unlock(&qstate->env->hints->lock);
*retdpname = NULL;
*retdpnamelen = 0;
return stub_no_cache;
}
memmove(dpname_storage, stub->dp->name,
stub->dp->namelen);
*retdpname = dpname_storage;
*retdpnamelen = stub->dp->namelen;
}
return (stub->dp->no_cache);
lock_rw_unlock(&qstate->env->hints->lock);
return stub_no_cache;
}
/* Check for forward. */
if (dp) {
if(dp->no_cache) {
char qname[255+1];
char dpname[255+1];
enum verbosity_value level = VERB_ALGO;
int dp_no_cache = dp->no_cache;
lock_rw_unlock(&qstate->env->hints->lock);
if(verbosity >= level && dp_no_cache) {
char qname[LDNS_MAX_DOMAINLEN];
char dpname[LDNS_MAX_DOMAINLEN];
dname_str(qinf->qname, qname);
dname_str(dp->name, dpname);
verbose(VERB_ALGO, "forward for %s %s has no_cache", qname, dpname);
verbose(level, "forward for %s %s has no_cache", qname, dpname);
}
if(retdpname) {
*retdpname = dp->name;
if(dp->namelen > dpname_storage_len) {
verbose(VERB_ALGO, "no cache dpname too long");
lock_rw_unlock(&qstate->env->fwds->lock);
*retdpname = NULL;
*retdpnamelen = 0;
return dp_no_cache;
}
memmove(dpname_storage, dp->name, dp->namelen);
*retdpname = dpname_storage;
*retdpnamelen = dp->namelen;
}
return (dp->no_cache);
lock_rw_unlock(&qstate->env->fwds->lock);
return dp_no_cache;
}
lock_rw_unlock(&qstate->env->fwds->lock);
lock_rw_unlock(&qstate->env->hints->lock);
if(retdpname) {
*retdpname = NULL;
*retdpnamelen = 0;
@@ -1469,3 +1610,54 @@ void iterator_set_ip46_support(struct module_stack* mods,
if(outnet->num_ip6 == 0)
ie->supports_ipv6 = 0;
}
void
limit_nsec_ttl(struct dns_msg* msg)
{
/* Limit NSEC and NSEC3 TTL in response, RFC9077 */
size_t i;
int found = 0;
time_t soa_ttl = 0;
/* Limit the NSEC and NSEC3 TTL values to the SOA TTL and SOA minimum
* TTL. That has already been applied to the SOA record ttl. */
for(i=0; i<msg->rep->rrset_count; i++) {
struct ub_packed_rrset_key* s = msg->rep->rrsets[i];
if(ntohs(s->rk.type) == LDNS_RR_TYPE_SOA) {
struct packed_rrset_data* soadata = (struct packed_rrset_data*)s->entry.data;
found = 1;
soa_ttl = soadata->ttl;
break;
}
}
if(!found)
return;
for(i=0; i<msg->rep->rrset_count; i++) {
struct ub_packed_rrset_key* s = msg->rep->rrsets[i];
if(ntohs(s->rk.type) == LDNS_RR_TYPE_NSEC ||
ntohs(s->rk.type) == LDNS_RR_TYPE_NSEC3) {
struct packed_rrset_data* data = (struct packed_rrset_data*)s->entry.data;
/* Limit the negative TTL. */
if(data->ttl > soa_ttl) {
if(verbosity >= VERB_ALGO) {
char buf[256];
snprintf(buf, sizeof(buf),
"limiting TTL %d of %s record to the SOA TTL of %d for",
(int)data->ttl, ((ntohs(s->rk.type) == LDNS_RR_TYPE_NSEC)?"NSEC":"NSEC3"), (int)soa_ttl);
log_nametypeclass(VERB_ALGO, buf,
s->rk.dname, ntohs(s->rk.type),
ntohs(s->rk.rrset_class));
}
data->ttl = soa_ttl;
}
}
}
}
void
iter_make_minimal(struct reply_info* rep)
{
size_t rem = rep->ns_numrrsets + rep->ar_numrrsets;
rep->ns_numrrsets = 0;
rep->ar_numrrsets = 0;
rep->rrset_count -= rem;
}
+79 -4
View File
@@ -61,6 +61,16 @@ struct sock_list;
struct ub_packed_rrset_key;
struct module_stack;
struct outside_network;
struct iter_nat64;
/* max number of lookups in the cache for target nameserver names.
* This stops, for large delegations, N*N lookups in the cache. */
#define ITERATOR_NAME_CACHELOOKUP_MAX 3
/* max number of lookups in the cache for parentside glue for nameserver names
* This stops, for larger delegations, N*N lookups in the cache.
* It is a little larger than the nonpside max, so it allows a couple extra
* lookups of parent side glue. */
#define ITERATOR_NAME_CACHELOOKUP_MAX_PSIDE 5
/**
* Process config options and set iterator module state.
@@ -132,6 +142,8 @@ struct dns_msg* dns_copy_msg(struct dns_msg* from, struct regional* regional);
* can be prefetch-updates.
* @param region: to copy modified (cache is better) rrs back to.
* @param flags: with BIT_CD for dns64 AAAA translated queries.
* @param qstarttime: time of query start.
* @param is_valrec: if the query is validation recursion and does not get
* return void, because we are not interested in alloc errors,
* the iterator and validator can operate on the results in their
* scratch space (the qstate.region) and are not dependent on the cache.
@@ -140,7 +152,8 @@ struct dns_msg* dns_copy_msg(struct dns_msg* from, struct regional* regional);
*/
void iter_dns_store(struct module_env* env, struct query_info* qinf,
struct reply_info* rep, int is_referral, time_t leeway, int pside,
struct regional* region, uint16_t flags);
struct regional* region, uint16_t flags, time_t qstarttime,
int is_valrec);
/**
* Select randomly with n/m probability.
@@ -175,10 +188,17 @@ void iter_mark_pside_cycle_targets(struct module_qstate* qstate,
* @param qinfo: query name and type
* @param qflags: query flags with RD flag
* @param dp: delegpt to check.
* @param supports_ipv4: if we support ipv4 for lookups to the target.
* if not, then the IPv4 addresses are useless.
* @param supports_ipv6: if we support ipv6 for lookups to the target.
* if not, then the IPv6 addresses are useless.
* @param use_nat64: if we support NAT64 for lookups to the target.
* if yes, IPv4 addresses are useful even if we don't support IPv4.
* @return true if dp is useless.
*/
int iter_dp_is_useless(struct query_info* qinfo, uint16_t qflags,
struct delegpt* dp);
int iter_dp_is_useless(struct query_info* qinfo, uint16_t qflags,
struct delegpt* dp, int supports_ipv4, int supports_ipv6,
int use_nat64);
/**
* See if qname has DNSSEC needs. This is true if there is a trust anchor above
@@ -390,10 +410,14 @@ int iter_dp_cangodown(struct query_info* qinfo, struct delegpt* dp);
* Used for NXDOMAIN checks, above that it is an nxdomain from a
* different server and zone. You can pass NULL to not get it.
* @param retdpnamelen: returns the length of the dpname.
* @param dpname_storage: this is where the dpname buf is stored, if any.
* So that caller can manage the buffer.
* @param dpname_storage_len: size of dpname_storage buffer.
* @return true if no_cache is set in stub or fwd.
*/
int iter_stub_fwd_no_cache(struct module_qstate *qstate,
struct query_info *qinf, uint8_t** retdpname, size_t* retdpnamelen);
struct query_info *qinf, uint8_t** retdpname, size_t* retdpnamelen,
uint8_t* dpname_storage, size_t dpname_storage_len);
/**
* Set support for IP4 and IP6 depending on outgoing interfaces
@@ -407,4 +431,55 @@ int iter_stub_fwd_no_cache(struct module_qstate *qstate,
void iterator_set_ip46_support(struct module_stack* mods,
struct module_env* env, struct outside_network* outnet);
/**
* Read config string that represents the target fetch policy.
* @param target_fetch_policy: alloced on return.
* @param max_dependency_depth: set on return.
* @param str: the config string
* @return false on failure.
*/
int read_fetch_policy(int** target_fetch_policy, int* max_dependency_depth,
const char* str);
/**
* Create caps exempt data structure.
* @return NULL on failure.
*/
struct rbtree_type* caps_white_create(void);
/**
* Delete caps exempt data structure.
* @param caps_white: caps exempt tree.
*/
void caps_white_delete(struct rbtree_type* caps_white);
/**
* Apply config caps whitelist items to name tree
* @param ntree: caps exempt tree.
* @param cfg: config with options.
*/
int caps_white_apply_cfg(struct rbtree_type* ntree, struct config_file* cfg);
/**
* Apply config for nat64
* @param nat64: the nat64 state.
* @param cfg: config with options.
* @return false on failure.
*/
int nat64_apply_cfg(struct iter_nat64* nat64, struct config_file* cfg);
/**
* Limit NSEC and NSEC3 TTL in response, RFC9077
* @param msg: dns message, the SOA record ttl is used to restrict ttls
* of NSEC and NSEC3 RRsets. If no SOA record, nothing happens.
*/
void limit_nsec_ttl(struct dns_msg* msg);
/**
* Make the response minimal. Removed authority and additional section,
* that works when there is an answer in the answer section.
* @param rep: reply to modify.
*/
void iter_make_minimal(struct reply_info* rep);
#endif /* ITERATOR_ITER_UTILS_H */
+848 -312
View File
File diff suppressed because it is too large Load Diff
+82 -17
View File
@@ -46,8 +46,6 @@
#include "util/data/msgreply.h"
#include "util/module.h"
struct delegpt;
struct iter_hints;
struct iter_forwards;
struct iter_donotq;
struct iter_prep_list;
struct iter_priv;
@@ -55,17 +53,19 @@ struct rbtree_type;
/** max number of targets spawned for a query and its subqueries */
#define MAX_TARGET_COUNT 64
/** max number of upstream queries for a query and its subqueries, it is
* never reset. */
extern int MAX_GLOBAL_QUOTA;
/** max number of target lookups per qstate, per delegation point */
#define MAX_DP_TARGET_COUNT 16
/** max number of nxdomains allowed for target lookups for a query and
* its subqueries */
#define MAX_TARGET_NX 5
/** max number of query restarts. Determines max number of CNAME chain. */
#define MAX_RESTART_COUNT 11
/** max number of nxdomains allowed for target lookups for a query and
* its subqueries when fallback has kicked in */
#define MAX_TARGET_NX_FALLBACK (MAX_TARGET_NX*2)
/** max number of referrals. Makes sure resolver does not run away */
#define MAX_REFERRAL_COUNT 130
/** max number of queries-sent-out. Make sure large NS set does not loop */
#define MAX_SENT_COUNT 32
/** max number of queries for which to perform dnsseclameness detection,
* (rrsigs missing detection) after that, just pick up that response */
#define DNSSEC_LAME_DETECT_COUNT 4
@@ -91,18 +91,39 @@ struct rbtree_type;
extern int UNKNOWN_SERVER_NICENESS;
/** maximum timeout before a host is deemed unsuitable, in msec.
* After host_ttl this will be timed out and the host will be tried again.
* Equals RTT_MAX_TIMEOUT
*/
#define USEFUL_SERVER_TOP_TIMEOUT 120000
* Equals RTT_MAX_TIMEOUT, and thus when RTT_MAX_TIMEOUT is overwritten by
* config infra_cache_max_rtt, it will be overwritten as well. */
extern int USEFUL_SERVER_TOP_TIMEOUT;
/** penalty to validation failed blacklisted IPs
* Equals USEFUL_SERVER_TOP_TIMEOUT*4, and thus when RTT_MAX_TIMEOUT is
* overwritten by config infra_cache_max_rtt, it will be overwritten as well. */
extern int BLACKLIST_PENALTY;
/** RTT band, within this amount from the best, servers are chosen randomly.
* Chosen so that the UNKNOWN_SERVER_NICENESS falls within the band of a
* fast server, this causes server exploration as a side benefit. msec. */
#define RTT_BAND 400
/** Start value for blacklisting a host, 2*USEFUL_SERVER_TOP_TIMEOUT in sec */
#define INFRA_BACKOFF_INITIAL 240
/** Number of retries for empty nodata packets before it is accepted. */
#define EMPTY_NODATA_RETRY_COUNT 2
/**
* Global state for the iterator.
* Iterator global state for nat64.
*/
struct iter_nat64 {
/** A flag to locally apply NAT64 to make IPv4 addrs into IPv6 */
int use_nat64;
/** NAT64 prefix address, cf. dns64_env->prefix_addr */
struct sockaddr_storage nat64_prefix_addr;
/** sizeof(sockaddr_in6) */
socklen_t nat64_prefix_addrlen;
/** CIDR mask length of NAT64 prefix */
int nat64_prefix_net;
};
/**
* Global state for the iterator.
*/
struct iter_env {
/** A flag to indicate whether or not we have an IPv6 route */
@@ -111,6 +132,9 @@ struct iter_env {
/** A flag to indicate whether or not we have an IPv4 route */
int supports_ipv4;
/** State for nat64 */
struct iter_nat64 nat64;
/** A set of inetaddrs that should never be queried. */
struct iter_donotq* donotq;
@@ -140,6 +164,12 @@ struct iter_env {
/** number of retries on outgoing queries */
int outbound_msg_retry;
/** number of queries_sent */
int max_sent_count;
/** max number of query restarts to limit length of CNAME chain */
int max_query_restarts;
};
/**
@@ -217,6 +247,24 @@ enum iter_state {
FINISHED_STATE
};
/**
* Shared counters for queries.
*/
enum target_count_variables {
/** Reference count for the shared iter_qstate->target_count. */
TARGET_COUNT_REF = 0,
/** Number of target queries spawned for the query and subqueries. */
TARGET_COUNT_QUERIES,
/** Number of nxdomain responses encountered. */
TARGET_COUNT_NX,
/** Global quota on number of queries to upstream servers per
* client request, that is never reset. */
TARGET_COUNT_GLOBAL_QUOTA,
/** This should stay last here, it is used for the allocation */
TARGET_COUNT_MAX,
};
/**
* Per query state for the iterator module.
*/
@@ -304,21 +352,26 @@ struct iter_qstate {
/** the number of times this query has been restarted. */
int query_restart_count;
/** the number of times this query as followed a referral. */
/** the number of times this query has followed a referral. */
int referral_count;
/** number of queries fired off */
int sent_count;
/** number of target queries spawned in [1], for this query and its
* subqueries, the malloced-array is shared, [0] refcount.
* in [2] the number of nxdomains is counted. */
/** malloced-array shared with this query and its subqueries. It keeps
* track of the defined enum target_count_variables counters. */
int* target_count;
/** number of target lookups per delegation point. Reset to 0 after
* receiving referral answer. Not shared with subqueries. */
int dp_target_count;
/** Delegation point that triggered the NXNS fallback; shared with
* this query and its subqueries, count-referenced by the reference
* counter in target_count.
* This also marks the fallback activation. */
uint8_t** nxns_dp;
/** if true, already tested for ratelimiting and passed the test */
int ratelimit_ok;
@@ -376,6 +429,11 @@ struct iter_qstate {
*/
int refetch_glue;
/**
* This flag detects that a completely empty nodata was received,
* already so that it is accepted later. */
int empty_nodata_found;
/** list of pending queries to authoritative servers. */
struct outbound_list outlist;
@@ -412,7 +470,14 @@ struct iter_qstate {
/** true if there have been parse failures of reply packets */
int parse_failures;
/** a failure printout address for last received answer */
struct comm_reply* fail_reply;
union {
struct in_addr in;
#ifdef AF_INET6
struct in6_addr in6;
#endif
} fail_addr;
/** which fail_addr, 0 is nothing, 4 or 6 */
int fail_addr_type;
};
/**
+13 -2
View File
@@ -53,6 +53,8 @@
#include "util/storage/slabhash.h"
#include "util/edns.h"
#include "sldns/sbuffer.h"
#include "iterator/iter_fwd.h"
#include "iterator/iter_hints.h"
int
context_finalize(struct ub_ctx* ctx)
@@ -70,9 +72,12 @@ context_finalize(struct ub_ctx* ctx)
} else {
log_init(cfg->logfile, cfg->use_syslog, NULL);
}
ctx->pipe_pid = getpid();
cfg_apply_local_port_policy(cfg, 65536);
config_apply(cfg);
if(!modstack_setup(&ctx->mods, cfg->module_conf, ctx->env))
if(!modstack_call_startup(&ctx->mods, cfg->module_conf, ctx->env))
return UB_INITFAIL;
if(!modstack_call_init(&ctx->mods, cfg->module_conf, ctx->env))
return UB_INITFAIL;
listen_setup_locks();
log_edns_known_options(VERB_ALGO, ctx->env);
@@ -84,6 +89,12 @@ context_finalize(struct ub_ctx* ctx)
if(!auth_zones_apply_cfg(ctx->env->auth_zones, cfg, 1, &is_rpz,
ctx->env, &ctx->mods))
return UB_INITFAIL;
if(!(ctx->env->fwds = forwards_create()) ||
!forwards_apply_cfg(ctx->env->fwds, cfg))
return UB_INITFAIL;
if(!(ctx->env->hints = hints_create()) ||
!hints_apply_cfg(ctx->env->hints, cfg))
return UB_INITFAIL;
if(!edns_strings_apply_cfg(ctx->env->edns_strings, cfg))
return UB_INITFAIL;
if(!slabhash_is_size(ctx->env->msg_cache, cfg->msg_cache_size,
@@ -384,7 +395,7 @@ context_serialize_cancel(struct ctx_query* q, uint32_t* len)
/* format of cancel:
* o uint32 cmd
* o uint32 async-id */
uint8_t* p = (uint8_t*)reallocarray(NULL, sizeof(uint32_t), 2);
uint8_t* p = (uint8_t*)reallocarray(NULL, 2, sizeof(uint32_t));
if(!p) return NULL;
*len = 2*sizeof(uint32_t);
sldns_write_uint32(p, UB_LIBCMD_CANCEL);
+6
View File
@@ -89,6 +89,12 @@ struct ub_ctx {
pid_t bg_pid;
/** tid of bg worker thread */
ub_thread_type bg_tid;
/** pid when pipes are created. This was the process when the
* setup was called. Helps with clean up, so we can tell after a fork
* which side of the fork the delete is on. */
pid_t pipe_pid;
/** when threaded, the worker that exists in the created thread. */
struct libworker* thread_worker;
/** do threading (instead of forking) for async resolution */
int dothread;

Some files were not shown because too many files have changed in this diff Show More