]>
git.ipfire.org Git - thirdparty/knot-resolver.git/log
Vladimír Čunát [Thu, 3 Aug 2017 11:55:04 +0000 (13:55 +0200)]
dnssec: guard against out-of-bailiwick signatures
Vladimír Čunát [Thu, 3 Aug 2017 13:23:43 +0000 (15:23 +0200)]
Merge !346: kr_rand_uint: nitpicks
Vladimír Čunát [Thu, 3 Aug 2017 12:59:48 +0000 (14:59 +0200)]
kr_rand_uint: nitpicks
- the returned value is up to max-1 and not max (tiny bias)
- improve efficiency slightly
- unsigned -> uint32_t, as that's the range it supports (mostly equal)
Vladimír Čunát [Thu, 3 Aug 2017 08:41:35 +0000 (10:41 +0200)]
Merge !343: iterate: remove counter-productive validation
Grigorii Demidov [Wed, 2 Aug 2017 16:13:49 +0000 (18:13 +0200)]
layer/iterate: remove counter-productive validation
... functionality from iterator: don't fail immediately if actual number
of labels in owner name exceeds number in label field of RRSIG rrset
Petr Špaček [Wed, 2 Aug 2017 18:45:00 +0000 (20:45 +0200)]
Merge branch 'ci-respdiff' into 'master'
CI: integrate respdiff v2 into gitlab-ci
See merge request !345
Stepan Kotek [Wed, 2 Aug 2017 10:42:01 +0000 (12:42 +0200)]
CI: migrate to respdiff v2
All resolvers are configured as recursors (no forwarding).
We should extend testing so we can test recursion and forwarding at the
same time.
List of queries is taken dynamically from
https://gitlab.labs.nic.cz/knot/knot-resolver/snippets/69
This will likely change in future.
The test will fail if mismatch rate >= 1 %.
This is error margin for network instability and few possibly broken upstreams
in the test list.
Petr Špaček [Wed, 2 Aug 2017 17:25:22 +0000 (19:25 +0200)]
Merge branch 'ci-deckard' into 'master'
CI: update Deckard to use newest resolver test suite
See merge request !344
Petr Špaček [Wed, 2 Aug 2017 17:17:42 +0000 (19:17 +0200)]
CI: update Deckard to use newest resolver test suite
Vladimír Čunát [Wed, 2 Aug 2017 15:41:51 +0000 (17:41 +0200)]
Merge !325: policy: more/better special-use domain names
Vladimír Čunát [Wed, 2 Aug 2017 14:59:14 +0000 (16:59 +0200)]
NEWS: changes in this branch
Vladimír Čunát [Wed, 2 Aug 2017 14:48:21 +0000 (16:48 +0200)]
Merge branch 'master' into policy-reserved-domains
Vladimír Čunát [Wed, 2 Aug 2017 14:29:21 +0000 (16:29 +0200)]
Merge !342: make: revert adding -rpath by default
Vladimír Čunát [Wed, 2 Aug 2017 14:23:19 +0000 (16:23 +0200)]
make: revert adding -rpath by default
This reverts commits
89ebf5293 and
1d8b5706cd .
I didn't realize that setting rpath is considered bad practice by most
distributions. I'm really "spoiled" by nix(pkgs/os).
Vladimír Čunát [Wed, 2 Aug 2017 09:31:12 +0000 (11:31 +0200)]
policy docs: rework it all
- greatly reduce duplication - mainly actions and filters that were
described on two separate places
- try to improve readability etc.
Vladimír Čunát [Wed, 2 Aug 2017 08:31:04 +0000 (10:31 +0200)]
hints docs: clean a bit, note interaction with policies
Vladimír Čunát [Tue, 1 Aug 2017 14:50:29 +0000 (16:50 +0200)]
Merge !340: make: add -rpath only on non-Darwin POSIX
Vladimír Čunát [Tue, 1 Aug 2017 11:47:10 +0000 (13:47 +0200)]
make: add -rpath only on non-Darwin POSIX
I didn't recall this doesn't work on Darwin.
Vladimír Čunát [Tue, 1 Aug 2017 11:28:28 +0000 (13:28 +0200)]
Merge !338: make: add -rpath when linking to find libkres
Vladimír Čunát [Tue, 1 Aug 2017 11:23:38 +0000 (13:23 +0200)]
make: add -rpath when linking to find libkres
... even if LIBDIR isn't on a standard place.
Vladimír Čunát [Fri, 28 Jul 2017 09:03:01 +0000 (11:03 +0200)]
Merge !336: release 1.3.2
Vladimír Čunát [Fri, 28 Jul 2017 08:59:56 +0000 (10:59 +0200)]
release 1.3.2
Vladimír Čunát [Thu, 27 Jul 2017 11:51:12 +0000 (13:51 +0200)]
Merge !335: predict: fixes for refreshing expiring RRs
Vladimír Čunát [Thu, 27 Jul 2017 11:38:03 +0000 (13:38 +0200)]
predict docs: fix the description of expiring records
The description didn't match the reality. Change docs for now,
even though 1% seems a bit too little to me personally.
Vladimír Čunát [Thu, 27 Jul 2017 09:25:59 +0000 (11:25 +0200)]
predict: refresh expiring records immediately
It seemed a bit strange to have 'stats' module collect expiring records
in an LRU, then once a few minutes convert that via JSON to a lua table,
and put those records into prefetching queue. Apart from complexity,
it wouldn't work well with short TTLs typical for today's CDNs, e.g. 30
or 60 seconds.
Vladimír Čunát [Wed, 26 Jul 2017 16:21:01 +0000 (18:21 +0200)]
rrcache: fix prefetching of expiring insecure RRs
Our security policy prevented those from reaching cache,
crippling the feature.
Vladimír Čunát [Thu, 27 Jul 2017 10:54:03 +0000 (12:54 +0200)]
Merge !334: bootstrap-depends.sh: fixup travis
Vladimír Čunát [Thu, 27 Jul 2017 08:36:53 +0000 (10:36 +0200)]
bootstrap-depends.sh: try to fix travis
Vladimír Čunát [Wed, 26 Jul 2017 10:29:44 +0000 (12:29 +0200)]
Merge !330: predict: fix enqueuing from predict log
Closes #154.
Vladimír Čunát [Wed, 26 Jul 2017 10:21:44 +0000 (12:21 +0200)]
update NEWS
Fixes https://gitlab.labs.nic.cz/knot/knot-resolver/issues/154
I'm sorry I broke the module in
06b0d3d4bf6c . Thaks Vita!
Vladimír Čunát [Wed, 26 Jul 2017 10:18:57 +0000 (12:18 +0200)]
Merge branch 'master' into predict-fix
... to allow NEWS update
Vladimír Čunát [Wed, 26 Jul 2017 10:15:06 +0000 (12:15 +0200)]
predict.generate: fix modular arithmetic
The epoch indices are from [1..period].
Vladimír Čunát [Mon, 24 Jul 2017 14:20:14 +0000 (16:20 +0200)]
Merge !333: dnssec: handle unknown DNSKEY/DS algorithms
Vladimír Čunát [Mon, 26 Jun 2017 09:49:49 +0000 (11:49 +0200)]
layer/validate: handle unknown algorithms
i.e. downgrade a zone to insecure when *all* DNSKEYs of the apex are
unverifiable due to unimplemented DNSKEY or DS algorithms.
Fixes https://gitlab.labs.nic.cz/knot/resolver/issues/210
Vladimír Čunát [Thu, 29 Jun 2017 09:55:29 +0000 (11:55 +0200)]
Makefile refactoring: abstract knot version
Vladimír Čunát [Mon, 24 Jul 2017 09:13:22 +0000 (11:13 +0200)]
Merge !309: hints: improve interpretation of hosts-like files
Closes #204.
Vladimír Čunát [Mon, 24 Jul 2017 09:09:15 +0000 (11:09 +0200)]
Merge branch 'master' into hints-name-order
... to fix conflict in NEWS.
Vladimír Čunát [Mon, 24 Jul 2017 09:06:17 +0000 (11:06 +0200)]
Merge !332: fixup old gitlab URLs
Vladimír Čunát [Mon, 17 Jul 2017 16:39:33 +0000 (18:39 +0200)]
hints.add_hosts: error out if a bad entry is encountered
Vladimír Čunát [Mon, 17 Jul 2017 16:21:44 +0000 (18:21 +0200)]
hints: add (shadowed) PTR records for non-canonical names
Which semantically reverses part of
fbe88ccd2 .
I would personally prefer not to, but I guess it would be less consistent.
Also make the error messages a little better.
Vladimír Čunát [Mon, 17 Jul 2017 15:57:57 +0000 (17:57 +0200)]
fixup old gitlab URLs
Almost all of them were getting redirected correctly, but why keep the
old ones. Also update a couple nitpicks directly around.
Vladimír Čunát [Mon, 17 Jul 2017 15:46:06 +0000 (17:46 +0200)]
Merge !331: Dockerfile: fix repo URL
Vladimír Čunát [Mon, 17 Jul 2017 15:44:42 +0000 (17:44 +0200)]
Dockerfile: fix repo URL
Vladimír Čunát [Mon, 17 Jul 2017 12:26:23 +0000 (14:26 +0200)]
Merge !329: rrcache: mangle TTL in first answer already
Vladimír Čunát [Mon, 17 Jul 2017 10:27:23 +0000 (12:27 +0200)]
hints: fix removal of PTR records
I broke this in
3da1535ece when changing stuff around addr2reverse;
now it should be OK, after partially reversing that commit.
I'm sorry. Thanks to Vita for noticing the breakage.
Vladimír Čunát [Mon, 17 Jul 2017 09:23:52 +0000 (11:23 +0200)]
Merge !328: fix dns64 with policy.FORWARD
Vladimír Čunát [Mon, 17 Jul 2017 09:20:44 +0000 (11:20 +0200)]
dns64 docs: document that policy.FORWARD works OK
At least it seems so...
Vladimír Čunát [Fri, 14 Jul 2017 11:32:34 +0000 (13:32 +0200)]
update NEWS, docs
Fixes https://gitlab.labs.nic.cz/knot/resolver/issues/204
Vladimír Čunát [Fri, 14 Jul 2017 11:14:41 +0000 (13:14 +0200)]
Merge branch 'master' into hints-name-order
(so that NEWS can get updated)
Vladimír Čunát [Thu, 15 Jun 2017 14:33:33 +0000 (16:33 +0200)]
hints: change the handling of repeated hints (again)
In reverse (PTR) records, now the last definition wins,
and non-canonical names don't get (shadowed) records.
vitas [Thu, 13 Jul 2017 08:38:15 +0000 (10:38 +0200)]
predict: fix enquing from predict log
* changed stype to type
* fix enquening queries from predict log
* keep heuristic in function generate to predict upcoming epoch
Vladimír Čunát [Wed, 12 Jul 2017 14:14:23 +0000 (16:14 +0200)]
Merge !327: zonecut: fix possible opportunities to use insecure
... data from cache as keys for validation
Vladimír Čunát [Tue, 11 Jul 2017 11:35:59 +0000 (13:35 +0200)]
zonecut: fix possible opportunities to use insecure
... data from cache as keys for validation
Grigorii Demidov [Wed, 12 Jul 2017 13:15:17 +0000 (15:15 +0200)]
lib/resolve: make the dns64 module working correctly in forwarding mode
Vladimír Čunát [Tue, 11 Jul 2017 14:18:18 +0000 (16:18 +0200)]
policy: change the localhost. domain
The RFC seems to read that resolvers should reply with address records
even for foo.bar.localhost. Note: Unbound-1.6.4 does NOT do that.
https://tools.ietf.org/html/rfc6761#section-6.3
Vladimír Čunát [Tue, 11 Jul 2017 13:46:03 +0000 (15:46 +0200)]
policy: implement NXDOMAIN for onion.
Vladimír Čunát [Tue, 11 Jul 2017 12:50:14 +0000 (14:50 +0200)]
Merge branch 'master' into policy-reserved-domains
... to fix conflict in a generated lua file
Vladimír Čunát [Tue, 11 Jul 2017 12:03:49 +0000 (14:03 +0200)]
policy: follow RFC6303 more closely
In particular, try to make the locally-served zones valid,
including SOA and NS in apex, empty non-terminal vs. NXDOMAIN, etc.
I might've missed something, but it should certainly be closer to ideal.
vitas [Tue, 11 Jul 2017 08:43:50 +0000 (10:43 +0200)]
rrcache: mangle TTL in first answer when using cache.min_ttl
Issue #127.
Answer is mangled first and then copied to cache.
vitas [Mon, 10 Jul 2017 14:31:05 +0000 (16:31 +0200)]
policy: answer to reversed localhost query
Vladimír Čunát [Fri, 7 Jul 2017 15:51:25 +0000 (17:51 +0200)]
Merge !326: various nitpicks
Vladimír Čunát [Fri, 7 Jul 2017 15:47:12 +0000 (17:47 +0200)]
dns64 nitpick: make the query flag "official"
Vladimír Čunát [Fri, 7 Jul 2017 12:01:00 +0000 (14:01 +0200)]
dnstap tests: ignore+clean additional directories
Vladimír Čunát [Fri, 7 Jul 2017 11:16:41 +0000 (13:16 +0200)]
utils: remove unused kr_rrarray_add(...)
Noticed after removal of its lua binding in commit
3be76db7610 .
Vladimír Čunát [Fri, 7 Jul 2017 15:31:30 +0000 (17:31 +0200)]
Merge !324: dns64: correct order of CNAME and AAAA in answer
Vladimír Čunát [Fri, 7 Jul 2017 13:14:05 +0000 (15:14 +0200)]
NEWS, docs: update dns64 info
Vladimír Čunát [Fri, 7 Jul 2017 15:13:26 +0000 (17:13 +0200)]
dns64: improve with policy.FORWARD or .STUB
Clearing the query flags lead to kresd iterating over authoritative
servers. Related: https://gitlab.labs.nic.cz/knot/resolver/issues/217
Note that validation of synthesised AAAA behind a CNAME doesn't work
when policy.FORWARD + dns64. Example: www.regiojet.cz.
policy.STUB + dns64 seems not to work at all.
Vladimír Čunát [Fri, 7 Jul 2017 12:59:13 +0000 (14:59 +0200)]
lua nitpicks: memory management
Vítězslav Kříž [Mon, 19 Jun 2017 11:17:57 +0000 (13:17 +0200)]
dns64-cname: correct order of CNAME and AAAA in answer
If CNAME chain occurs CNAME is placed before AAAA.
Vladimír Čunát [Mon, 3 Jul 2017 10:21:43 +0000 (12:21 +0200)]
Merge !323: contrib/lmdb: 0.9.18 -> 0.9.21
Vladimír Čunát [Mon, 3 Jul 2017 08:39:35 +0000 (10:39 +0200)]
contrib/lmdb: 0.9.18 -> 0.9.21
Maybe some coverity issues will get fixed, even though
they had seemed not to be affecting our use cases in kresd.
Tests+Deckard look OK and I've been using my system and kresd with
system-wide 0.9.21 for some time already.
Vladimír Čunát [Fri, 30 Jun 2017 15:17:40 +0000 (17:17 +0200)]
Merge !318: lib/resolve: use RTT tracking to choose forwarders
Closes #125 and #208.
Vladimír Čunát [Fri, 30 Jun 2017 11:52:21 +0000 (13:52 +0200)]
lib/resolve: even better support for forwarder choice
- sort the list instead of just picking the best one
- prefer unknown RTTs to probe them
- verbose output of the choice
Fixes https://gitlab.labs.nic.cz/knot/resolver/issues/125
Fixes https://gitlab.labs.nic.cz/knot/resolver/issues/208
Vladimír Čunát [Fri, 30 Jun 2017 12:41:00 +0000 (14:41 +0200)]
nsrep: eradicate kr_nsrep_inaddr*
They're just useless when we have kr_inaddr*.
Grigorii Demidov [Mon, 26 Jun 2017 10:20:39 +0000 (12:20 +0200)]
lib/resolve: better support for forwarder choice
Vladimír Čunát [Fri, 30 Jun 2017 07:13:02 +0000 (09:13 +0200)]
Merge !322: coverity fixes (nothing important)
Vladimír Čunát [Fri, 30 Jun 2017 06:57:06 +0000 (08:57 +0200)]
utils: shut up a coverity warning (CID 155456)
Vladimír Čunát [Fri, 30 Jun 2017 06:48:59 +0000 (08:48 +0200)]
rplan: fix a coverity warning (CID 164834)
There would be no real resource leak, as the memory is on a mempool,
so the lifetime is limited by kr_request finishing. Also, it's only
on an unlikely error branch, but still...
Vladimír Čunát [Fri, 30 Jun 2017 06:27:54 +0000 (08:27 +0200)]
Merge !321: bootstrap-depends nitpick: warn about ~/.local
Vladimír Čunát [Mon, 19 Jun 2017 16:01:46 +0000 (18:01 +0200)]
bootstrap-depends nitpick: warn about rm -rf ~/.local
Some people might get surprised, as it's the default location for quite
some data nowadays ($XDG_DATA_HOME).
Paul Hoffman [Wed, 28 Jun 2017 17:12:19 +0000 (10:12 -0700)]
Fix doc for building and running
Vladimír Čunát [Wed, 28 Jun 2017 08:07:26 +0000 (10:07 +0200)]
Merge !320: dnssec/nsec: add a precaution for DS nodata
Vladimír Čunát [Wed, 28 Jun 2017 06:48:40 +0000 (08:48 +0200)]
dnssec/nsec: add a precaution when checking DS nodata
I just found it by accident when studying RFCs again.
Hmm, I'd like an operator for logical implication; I'm spoiled by Maths.
vitas [Tue, 27 Jun 2017 15:54:22 +0000 (17:54 +0200)]
policy: handle localhost domain properly
According to RFC6761 query to localhost domain should generate
immediate response with loopback ip address.
vitas [Tue, 27 Jun 2017 13:43:07 +0000 (15:43 +0200)]
policy: added test. and invalid. to special domain list
Vladimír Čunát [Tue, 27 Jun 2017 13:33:17 +0000 (15:33 +0200)]
Merge branch 'check-config' into 'master'
daemon: check existence of config file
See merge request !317
vitas [Tue, 27 Jun 2017 11:53:37 +0000 (13:53 +0200)]
policy: private reverse zone handled separately
This allows to disable specific deny rules on zone-by-zone basis
as it should be according to RFC6303 sec.3. Disable can be done by
policy rule PASS.
Also any FORWARD rule for such zone will be evaluated before DENY
that query.
Vladimír Čunát [Tue, 27 Jun 2017 12:48:38 +0000 (14:48 +0200)]
Merge !319: Dockerfile: add missing curl dependency
Vladimír Čunát [Tue, 27 Jun 2017 12:43:24 +0000 (14:43 +0200)]
Dockerfile: add missing curl dependency
Without it we were failing to download fstrm tarball,
and the shown errors were very confusing.
Tested building an image locally.
Vladimír Čunát [Mon, 26 Jun 2017 08:26:33 +0000 (10:26 +0200)]
daemon: check existence of config file
... even if rundir isn't specified. No other changes in semantics.
Before this a typo in config path would pass silently.
Vladimír Čunát [Fri, 23 Jun 2017 15:37:21 +0000 (17:37 +0200)]
Merge !315: doc/build.rst updates
Vladimír Čunát [Fri, 23 Jun 2017 12:10:44 +0000 (14:10 +0200)]
Merge !316: release 1.3.1
Vladimír Čunát [Wed, 21 Jun 2017 07:46:51 +0000 (09:46 +0200)]
release 1.3.1
Grigorii Demidov [Thu, 22 Jun 2017 12:56:52 +0000 (14:56 +0200)]
lib/resolve: bugfixes for forwarding mode
unecessary queries in some circumstances; some minor bugfixes
Grigorii Demidov [Thu, 22 Jun 2017 09:21:01 +0000 (11:21 +0200)]
layer/pktcache: don't cache CNAME'ed negative answer when forwarding
it can contain records which have not been validated by validator
Grigorii Demidov [Wed, 21 Jun 2017 08:02:13 +0000 (10:02 +0200)]
layer/validate: forwarding mode, another CNAME fix
Don't check proof of nonexistance in NODATA or NAME ERROR answers
which contains a CNAME.
Vladimír Čunát [Mon, 19 Jun 2017 15:11:01 +0000 (17:11 +0200)]
doc/build.rst: quick pass to fix/update the info
In particular, don't mention things that probably don't work (well):
Windows, amalgamation, very old OpenSUSE package.
Vladimír Čunát [Mon, 19 Jun 2017 14:21:49 +0000 (16:21 +0200)]
Merge !314: policy.FORWARD: fix some CNAME problems
Vladimír Čunát [Mon, 19 Jun 2017 14:15:44 +0000 (16:15 +0200)]
NEWS: update with for this branch
Grigorii Demidov [Thu, 15 Jun 2017 11:11:30 +0000 (13:11 +0200)]
layer/iterate: forwarding mode - treat CNAME'ed NS&DS answers as proof of zonecut nonexistance