]> git.ipfire.org Git - thirdparty/knot-resolver.git/log
thirdparty/knot-resolver.git
8 years agomemcached: rename module kmemcached -> memcached
Petr Špaček [Fri, 1 Dec 2017 12:46:11 +0000 (13:46 +0100)] 
memcached: rename module kmemcached -> memcached

The old name was confusing because module name backend name provided by
that modules were different.

8 years agoscripts: make scripts/inet_pton.py compatible with Python 2 and Python 3
Petr Špaček [Fri, 1 Dec 2017 12:43:19 +0000 (13:43 +0100)] 
scripts: make scripts/inet_pton.py compatible with Python 2 and Python 3

It is 2017 already, so Python 3 is the default.

8 years agoscripts: remove scripts/supervisor.py
Petr Špaček [Fri, 1 Dec 2017 12:41:35 +0000 (13:41 +0100)] 
scripts: remove scripts/supervisor.py

The script did not work under Python 3, and given broad support for
SO_REUSEPORT, it is easier to drop it than to debug it.

8 years agoMerge branch 'priming' into 'master'
Petr Špaček [Tue, 28 Nov 2017 14:38:19 +0000 (15:38 +0100)] 
Merge branch 'priming' into 'master'

priming: implement priming queries as module

Closes #220

See merge request knot/knot-resolver!391

8 years agopriming: enable module by default
Vítězslav Kříž [Mon, 27 Nov 2017 16:10:15 +0000 (17:10 +0100)] 
priming: enable module by default

Also removed priming query from trust anchor module.
Updated deckard repository.

8 years agopriming: implement priming queries as module
Vítězslav Kříž [Wed, 1 Nov 2017 17:26:54 +0000 (18:26 +0100)] 
priming: implement priming queries as module

8 years agoMerge branch 'code-coverage' into 'master'
Petr Špaček [Tue, 28 Nov 2017 12:21:33 +0000 (13:21 +0100)] 
Merge branch 'code-coverage' into 'master'

Updated config tests, fixing code coverage for C and Lua

See merge request knot/knot-resolver!403

8 years agobuild: detect presence of lcov and luacov
Petr Špaček [Tue, 28 Nov 2017 12:18:04 +0000 (13:18 +0100)] 
build: detect presence of lcov and luacov

8 years agoscripts: pack all submodules with make-archive.sh
Petr Špaček [Tue, 28 Nov 2017 10:27:43 +0000 (11:27 +0100)] 
scripts: pack all submodules with make-archive.sh

It does not make sense to create "distribution" tarball with dirty
working tree or missing submodules, so this is now forbidden.
make-archive.sh requires clean working tree and all submodules.

Packing all submodules prevents us from releasing incomplete tarball,
especially around release time.

Distributions like Fedora are gradually getting ability to run
integration test suites so it does not hurt to pack everything including
test.

8 years agoCI: Add coverage CI target to parse code coverage and display it in MRs
Marek Vavruša [Sun, 26 Nov 2017 01:05:08 +0000 (17:05 -0800)] 
CI: Add coverage CI target to parse code coverage and display it in MRs

8 years agofixed issues with newer luacheck
Marek Vavruša [Sun, 26 Nov 2017 04:15:35 +0000 (20:15 -0800)] 
fixed issues with newer luacheck

8 years agobuild: removed old check for libknot 2.3.0
Marek Vavruša [Sun, 26 Nov 2017 01:04:30 +0000 (17:04 -0800)] 
build: removed old check for libknot 2.3.0

The 2.4.0 is required as a minimum.

8 years agobuild: make lint depends on autogenerated Lua files
Marek Vavruša [Sun, 26 Nov 2017 01:03:48 +0000 (17:03 -0800)] 
build: make lint depends on autogenerated Lua files

Lint should always render the .lua from .lua.in files before checking.

8 years agobuild: added `make coverage` to compute C and Lua code coverage
Marek Vavruša [Sun, 26 Nov 2017 00:54:11 +0000 (16:54 -0800)] 
build: added `make coverage` to compute C and Lua code coverage

Currently it gathers gcov and luacov code coverage, and merges it
in a single lcov info file. It returns summary at the end which the
CI can parse and interpret. It can build a html report later using
the data.

8 years agotests/config: added a TAP-based test environment for modules/configs
Marek Vavruša [Sun, 26 Nov 2017 00:23:46 +0000 (16:23 -0800)] 
tests/config: added a TAP-based test environment for modules/configs

I moved the test files to module directories because it allows
vendoring of whole modules including tests etc.

The test environment provides convenience functions and produces
test output in TAP format. Ideally all tests should use a common
format, so that CI can parse it provide better test output on PRs.
It seems like Gitlab CI doesn't support anything yet, but there
are two sort-of standards supported in CI tools - TAP and JUnit.
I chose TAP because it's easier to read for humans, cmocka supports it,
and it should be easier to adapt Deckard. There are also tools to
convert TAP into JUnit XML file.

Also added more tests for global functions and variables, and the
test tool now also tracks coverage (if `luacov` is installed).

8 years agoMerge branch 'docker-updates' into 'master'
Petr Špaček [Mon, 27 Nov 2017 10:14:51 +0000 (11:14 +0100)] 
Merge branch 'docker-updates' into 'master'

CI: unify Docker images for kresd and Deckard CI

See merge request knot/knot-resolver!404

8 years agoCI: unify Docker images for kresd and Deckard CI
Petr Špaček [Mon, 27 Nov 2017 10:10:09 +0000 (11:10 +0100)] 
CI: unify Docker images for kresd and Deckard CI

The Docker image now should work for both projects.
Also, the image was extended with tools to measure code coverage.

8 years agotemporarily enable packet caching for NSEC3 answers
Vladimír Čunát [Fri, 24 Nov 2017 13:45:01 +0000 (14:45 +0100)] 
temporarily enable packet caching for NSEC3 answers

Work up to this commit was squashed into cache-aggr 36302052.

8 years agoMerge branch 'lualint' into 'master'
Petr Špaček [Fri, 24 Nov 2017 14:56:51 +0000 (15:56 +0100)] 
Merge branch 'lualint' into 'master'

CI: add Lua lint and installcheck tests

See merge request knot/knot-resolver!402

8 years agofixup! tests: fixed config tests locking up on error, added test for predict
Petr Špaček [Fri, 24 Nov 2017 14:48:56 +0000 (15:48 +0100)] 
fixup! tests: fixed config tests locking up on error, added test for predict

/bin/sh in our image does not support "function" keyword

8 years agoCI: run installcheck tests with and without Valgrind
Petr Špaček [Fri, 24 Nov 2017 14:15:25 +0000 (15:15 +0100)] 
CI: run installcheck tests with and without Valgrind

8 years agoCI: disable Deckard under Valgrind
Petr Špaček [Fri, 24 Nov 2017 13:55:35 +0000 (14:55 +0100)] 
CI: disable Deckard under Valgrind

The test is useless in its current form because all the information goes
to logs, which are later discarded.

We need to fix issues detected by Valgrind first and then enable strict
mode which will not let the test pass if any problem is detected.

This might require recompiling LuaJIT in a way which is compatible with
Valgrind.

8 years agoCI: disallow Valgrind failures in unit tests
Petr Špaček [Fri, 24 Nov 2017 13:54:34 +0000 (14:54 +0100)] 
CI: disallow Valgrind failures in unit tests

8 years agoCI: add Lua lint
Petr Špaček [Fri, 24 Nov 2017 13:05:42 +0000 (14:05 +0100)] 
CI: add Lua lint

8 years agogenerate positive wildcard answers as well
Vladimír Čunát [Fri, 24 Nov 2017 12:43:39 +0000 (13:43 +0100)] 
generate positive wildcard answers as well

8 years agonitpicks
Vladimír Čunát [Fri, 24 Nov 2017 10:25:41 +0000 (11:25 +0100)] 
nitpicks

8 years agoMerge branch 'daemon-lua-linter-fixes' into 'master'
Petr Špaček [Fri, 24 Nov 2017 07:52:59 +0000 (08:52 +0100)] 
Merge branch 'daemon-lua-linter-fixes' into 'master'

Fixed linter issues in `daemon/lua`, added more config tests

See merge request knot/knot-resolver!400

8 years agoMerge branch 'fix-obvious-linter-problems' into 'daemon-lua-linter-fixes'
Petr Špaček [Fri, 24 Nov 2017 07:46:59 +0000 (08:46 +0100)] 
Merge branch 'fix-obvious-linter-problems' into 'daemon-lua-linter-fixes'

fixed obvious linter problems

See merge request knot/knot-resolver!401

8 years agofixed obvious linter problems
Marek Vavruša [Fri, 24 Nov 2017 05:07:31 +0000 (21:07 -0800)] 
fixed obvious linter problems

there are two modules that couldn't work before:

* graphite
* ketcd

8 years agodaemon/lua: fixed unused variables, whitespace, cleanup
Marek Vavruša [Thu, 23 Nov 2017 09:02:15 +0000 (01:02 -0800)] 
daemon/lua: fixed unused variables, whitespace, cleanup

Also fixed improper promotion of `ffi` to global variable.

```
$ luacheck --codes daemon/lua/
Checking daemon/lua/config.lua                    OK
Checking daemon/lua/kres-gen.lua                  OK
Checking daemon/lua/kres.lua                      OK
Checking daemon/lua/sandbox.lua                   OK
Checking daemon/lua/trust_anchors.lua             OK
Checking daemon/lua/zonefile.lua                  OK

Total: 0 warnings / 0 errors in 6 files
```

8 years agoadded basic config test for consts, depend on knot >= 2.4, cleanup
Marek Vavruša [Thu, 23 Nov 2017 08:59:31 +0000 (00:59 -0800)] 
added basic config test for consts, depend on knot >= 2.4, cleanup

The library now depends on libknot >= 2.4.0 (in Debian stable),
this allows us to remove a dead code and sed-ing of the kres.lua

Added a basic config tests to check that constants still work,
and basic interface to rrsets still works after the change.

8 years agoMerge branch 'fix-predict-module' into 'master'
Petr Špaček [Fri, 24 Nov 2017 07:18:12 +0000 (08:18 +0100)] 
Merge branch 'fix-predict-module' into 'master'

Fixed predict module, config tests, cleanup

See merge request knot/knot-resolver!399

8 years agomodules/predict: added test for prediction process
Marek Vavruša [Fri, 24 Nov 2017 04:32:01 +0000 (20:32 -0800)] 
modules/predict: added test for prediction process

this tests that:
* sampling frequent queries works
* the code to find periodic appearences of the same name and type
* resolving predicted queries works

it doesn't test pessimistic cases or failure modes

8 years agotests/config: added basic assert support (compatible with busted)
Marek Vavruša [Fri, 24 Nov 2017 04:30:00 +0000 (20:30 -0800)] 
tests/config: added basic assert support (compatible with busted)

There is no dependency on a testing library yet, so I added a
basic interface for mocking and asserting test values to get
something to start with. I'll probably replace it with busted
or telescope later on to get nicer testing output.

8 years agorenamed TEST env variable to KRESD_NO_LISTEN
Marek Vavruša [Fri, 24 Nov 2017 03:31:55 +0000 (19:31 -0800)] 
renamed TEST env variable to KRESD_NO_LISTEN

8 years agoconverted constant tables, support kres.type.TYPE1234
Marek Vavruša [Thu, 23 Nov 2017 07:50:58 +0000 (23:50 -0800)] 
converted constant tables, support kres.type.TYPE1234

The difficulty with using structs as constant tables is that access
to non-existent fields throws an error. This is difficult to handle
without wrapping every access in a pcall, for example in predict module:

```
error: /usr/local/lib/kdns_modules/predict.lua:34: 'struct rr_type' has no member named 'TYPE65535'
```

So I converted the constant tables into regular Lua tables,
and added a metatable for RR types to allow looking up unnamed types,
in the TYPE%d format. Looking up non-existent fields will now
return nil instead of throwing an error.

8 years agomodules/predict: cleaned up unused function, variables, trailing whitespace
Marek Vavruša [Thu, 23 Nov 2017 07:55:31 +0000 (23:55 -0800)] 
modules/predict: cleaned up unused function, variables, trailing whitespace

The `stats` module is now also included if not exists.

```
$ luacheck modules/predict
Checking modules/predict/predict.lua              OK

Total: 0 warnings / 0 errors in 1 file
```

8 years agotests: fixed config tests locking up on error, added test for predict
Marek Vavruša [Thu, 23 Nov 2017 07:40:28 +0000 (23:40 -0800)] 
tests: fixed config tests locking up on error, added test for predict

The config tests locked up on error as if error was raised from the
event callback, it would never reach the `quit()` statement, so
server would never close on error.

Added a script to make running these types of tests a little bit nicer
and to allow concurrent execution of config tests.

Added a test for the predict module, that fails on prediction
of unknown types:

```
error: /usr/local/lib/kdns_modules/predict.lua:34: 'struct rr_type' has no member named 'TYPE65535'
```

8 years agoMerge branch 'add-lua-linter' into 'master'
Petr Špaček [Thu, 23 Nov 2017 16:10:00 +0000 (17:10 +0100)] 
Merge branch 'add-lua-linter' into 'master'

Added luacheck for linting Lua files and static analysis

See merge request knot/knot-resolver!398

8 years agodo cache_sync at the end of cache_peek
Vladimír Čunát [Thu, 23 Nov 2017 07:22:55 +0000 (08:22 +0100)] 
do cache_sync at the end of cache_peek

We do no writing in that phase (yet), but let's do it anyway.

8 years agolmdb write fix
Vladimír Čunát [Thu, 23 Nov 2017 07:18:19 +0000 (08:18 +0100)] 
lmdb write fix

Fix incorrectly computed bound, leading to writes beyond the buffer
returned by lmdb, resulting in all kinds of weird errors later.

8 years agoAdded luacheck for linting Lua files and static analysis
Marek Vavruša [Thu, 23 Nov 2017 02:57:39 +0000 (18:57 -0800)] 
Added luacheck for linting Lua files and static analysis

This is super useful for checking things like misusing undefined
variables or modifying globals, especially in modules when it's
not immediately visible which variables are in the global
namespace and which are not.

I added several exceptions for files in daemon/lua and tests,
as for example sandbox module needs to legitimately modify
global namespace.

There's a lot of things failing, so I didn't make it part of the
standard `make check`, but we should eventually enable it to
improve code quality and spot problems with CI.

8 years agoset ranks of RRSIGs to _OMIT
Vladimír Čunát [Wed, 22 Nov 2017 14:03:51 +0000 (15:03 +0100)] 
set ranks of RRSIGs to _OMIT

Otherwise the validator may get confused, trying to "validate" them.

8 years agocache: stash packets also if there's no TA at all
Vladimír Čunát [Wed, 22 Nov 2017 12:48:16 +0000 (13:48 +0100)] 
cache: stash packets also if there's no TA at all

8 years agoreview TTL limit handling
Vladimír Čunát [Wed, 22 Nov 2017 12:08:43 +0000 (13:08 +0100)] 
review TTL limit handling

- aggressive cache forgot to apply the maximum TTL bound
- packet cache had a separate ad-hoc limit of 15 minutes;
  this got removed
- cleanups, etc.

8 years agocommit the regenerated lua stuff, finally
Vladimír Čunát [Tue, 21 Nov 2017 18:12:25 +0000 (19:12 +0100)] 
commit the regenerated lua stuff, finally

8 years agomove NSEC1-related stuff into a separate file
Vladimír Čunát [Tue, 21 Nov 2017 18:11:28 +0000 (19:11 +0100)] 
move NSEC1-related stuff into a separate file

Also some minor changes, as usual.

8 years agoWIP: wildcards
Vladimír Čunát [Tue, 21 Nov 2017 14:55:47 +0000 (15:55 +0100)] 
WIP: wildcards

- minor fix
- improved verbose logging
- other nitpicks

Negative NSEC wildcards should be hopefully OK now.
No generation of positive answers from wildcards yet.

8 years agoWIP: wildcards
Vladimír Čunát [Mon, 20 Nov 2017 18:52:06 +0000 (19:52 +0100)] 
WIP: wildcards

8 years agoMerge branch 'deckard_fail_artifacts' into 'master'
Petr Špaček [Mon, 20 Nov 2017 16:45:28 +0000 (17:45 +0100)] 
Merge branch 'deckard_fail_artifacts' into 'master'

CI optimization

Closes #270

See merge request knot/knot-resolver!397

8 years agofix rcode for some negative answers
Vladimír Čunát [Mon, 20 Nov 2017 15:50:02 +0000 (16:50 +0100)] 
fix rcode for some negative answers

Empty non-terminals don't need to have a matching NSEC record.

8 years agoCI: enable QTRACE logging for kresd
Petr Špaček [Mon, 20 Nov 2017 15:35:06 +0000 (16:35 +0100)] 
CI: enable QTRACE logging for kresd

8 years agoCI: run non-interactive kresd to avoid false positives
Petr Špaček [Mon, 20 Nov 2017 15:33:38 +0000 (16:33 +0100)] 
CI: run non-interactive kresd to avoid false positives

See Deckard MR!85 for details.

8 years agoCI: archive Deckard artifacts on failure
Petr Špaček [Mon, 20 Nov 2017 15:31:54 +0000 (16:31 +0100)] 
CI: archive Deckard artifacts on failure

8 years agoCI: optimize Deckard-under-Valgrind
Petr Špaček [Thu, 16 Nov 2017 13:23:56 +0000 (14:23 +0100)] 
CI: optimize Deckard-under-Valgrind

Packages are already present in the image so we do not need to spend
time on package maintenance.

8 years agofix another bug for non-root zones
Vladimír Čunát [Mon, 20 Nov 2017 15:14:22 +0000 (16:14 +0100)] 
fix another bug for non-root zones

Sigh.

8 years agoMerge !396: lua: fixup making the bindings bogus
Vladimír Čunát [Mon, 20 Nov 2017 13:41:15 +0000 (14:41 +0100)] 
Merge !396: lua: fixup making the bindings bogus

8 years agolua: fixup making the bindings bogus
Vladimír Čunát [Mon, 20 Nov 2017 13:23:24 +0000 (14:23 +0100)] 
lua: fixup making the bindings bogus

For now I didn't touch the generator to be correct on this line,
as we will probably want to change it after bumping minimal libknot
version.

8 years agoMerge !395: minor fixes (see commits)
Vladimír Čunát [Mon, 20 Nov 2017 13:12:48 +0000 (14:12 +0100)] 
Merge !395: minor fixes (see commits)

8 years agolua: fix mistakes in bindings (forgotten regeneration)
Vladimír Čunát [Mon, 20 Nov 2017 13:07:46 +0000 (14:07 +0100)] 
lua: fix mistakes in bindings (forgotten regeneration)

8 years agodaemon: fix a typo in SIGPIPE blocking
Vladimír Čunát [Tue, 14 Nov 2017 13:18:43 +0000 (14:18 +0100)] 
daemon: fix a typo in SIGPIPE blocking

I'm sorry. Buggy commit: 2ba2a5e8

8 years agoNSEC: better verbose logging + nitpicks
Vladimír Čunát [Thu, 16 Nov 2017 17:25:19 +0000 (18:25 +0100)] 
NSEC: better verbose logging + nitpicks

8 years agoMerge !394: util: fix policy.QTRACE for QDCOUNT != 1
Vladimír Čunát [Thu, 16 Nov 2017 16:54:28 +0000 (17:54 +0100)] 
Merge !394: util: fix policy.QTRACE for QDCOUNT != 1

8 years agoutil: fix policy.QTRACE for QDCOUNT != 1
Petr Špaček [Thu, 16 Nov 2017 16:41:32 +0000 (17:41 +0100)] 
util: fix policy.QTRACE for QDCOUNT != 1

8 years agoMerge !393: lua resolve(): make DNSSEC_WANT imply +dnssec
Vladimír Čunát [Thu, 16 Nov 2017 16:04:13 +0000 (17:04 +0100)] 
Merge !393: lua resolve(): make DNSSEC_WANT imply +dnssec

8 years agolua resolve(): make DNSSEC_WANT imply +dnssec
Vladimír Čunát [Thu, 16 Nov 2017 15:32:12 +0000 (16:32 +0100)] 
lua resolve(): make DNSSEC_WANT imply +dnssec

8 years agorename module 'cache_lmdb' -> 'cache'
Vladimír Čunát [Thu, 16 Nov 2017 14:09:59 +0000 (15:09 +0100)] 
rename module 'cache_lmdb' -> 'cache'

The name was weird.  There's no other implementation ATM and the module
will always be only really suitable for *local* caches, but otherwise
it's not neccessarily bound to LMDB.

8 years agofix a bug for non-root zones
Vladimír Čunát [Wed, 15 Nov 2017 10:21:47 +0000 (11:21 +0100)] 
fix a bug for non-root zones

Owner name reconstruction didn't work.

8 years agoMerge !390: utils kr_strcatdup: deal with overflow
Vladimír Čunát [Wed, 15 Nov 2017 09:07:44 +0000 (10:07 +0100)] 
Merge !390: utils kr_strcatdup: deal with overflow

8 years agoutils kr_strcatdup: deal with overflowing size_t
Vladimír Čunát [Wed, 15 Nov 2017 08:48:29 +0000 (09:48 +0100)] 
utils kr_strcatdup: deal with overflowing size_t

It's very unlikely to happen - sum of string lengths overflowing -
even on a 32-bit platform, and the input seems not controllable by
adversaries, but let's fix it anyway.

8 years agominor: cleanups, logging, fixes, comments/docs
Vladimír Čunát [Tue, 14 Nov 2017 17:05:27 +0000 (18:05 +0100)] 
minor: cleanups, logging, fixes, comments/docs

8 years agoMerge branch 'fix-check-config' into 'master'
Petr Špaček [Mon, 13 Nov 2017 13:12:26 +0000 (14:12 +0100)] 
Merge branch 'fix-check-config' into 'master'

Fix check target (check-config requires installed version of kresd)

Closes #272

See merge request knot/knot-resolver!389

8 years agotests: check-config and check-integration require installed version
Petr Špaček [Mon, 13 Nov 2017 12:55:29 +0000 (13:55 +0100)] 
tests: check-config and check-integration require installed version

Path to module directory gets hardcoded into binaries so check-config
and check-integration cannot work prior installation.

To indicate this depedency, the target check-config was moved from
"check" target to "installcheck".

Deckard requires additional libraries so it is not included in
installcheck.

fixes: #272

8 years agounit tests: always use libraries from build directory
Petr Špaček [Mon, 13 Nov 2017 12:43:29 +0000 (13:43 +0100)] 
unit tests: always use libraries from build directory

Previously it might happen that installed libraries
had higher priority over libraries in build directory.

8 years agoMerge !388: minor fixes
Vladimír Čunát [Mon, 13 Nov 2017 09:50:46 +0000 (10:50 +0100)] 
Merge !388: minor fixes

8 years agotests_cache: give it more space
Vladimír Čunát [Mon, 13 Nov 2017 09:41:25 +0000 (10:41 +0100)] 
tests_cache: give it more space

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=878976
In general, platforms with page size other than 4k might better compile
with -DCPU_PAGE_SIZE=nnnnn but real impact should only be this test and
better alignment in the mempool allocator.

8 years agodaemon: block SIGPIPE
Vladimír Čunát [Mon, 13 Nov 2017 09:24:53 +0000 (10:24 +0100)] 
daemon: block SIGPIPE

We can ignore the signal, as the affected libuv calls report error
by returning EPIPE anyway.
Fixes https://gitlab.labs.nic.cz/knot/knot-resolver/issues/271

8 years agominor cleanups, reduce verbosity
Vladimír Čunát [Fri, 10 Nov 2017 16:52:17 +0000 (17:52 +0100)] 
minor cleanups, reduce verbosity

8 years agoremove some old code
Vladimír Čunát [Fri, 10 Nov 2017 16:31:35 +0000 (17:31 +0100)] 
remove some old code

8 years agomemcached, redis: disable as they won't work ATM
Vladimír Čunát [Fri, 10 Nov 2017 15:38:22 +0000 (16:38 +0100)] 
memcached, redis: disable as they won't work ATM

8 years agoRR-stashing nitpicks
Vladimír Čunát [Fri, 10 Nov 2017 15:20:25 +0000 (16:20 +0100)] 
RR-stashing nitpicks

8 years agopreparing knot_pkt_t: move to a separate file
Vladimír Čunát [Fri, 10 Nov 2017 15:04:38 +0000 (16:04 +0100)] 
preparing knot_pkt_t: move to a separate file

8 years agoRR (de)serialization: move to a separate file
Vladimír Čunát [Fri, 10 Nov 2017 14:41:39 +0000 (15:41 +0100)] 
RR (de)serialization: move to a separate file

8 years agocache entry lists: move to a separate file
Vladimír Čunát [Fri, 10 Nov 2017 14:05:16 +0000 (15:05 +0100)] 
cache entry lists: move to a separate file

8 years agofix clang warnings
Vladimír Čunát [Fri, 10 Nov 2017 10:14:53 +0000 (11:14 +0100)] 
fix clang warnings

8 years agopacket cache: move into a separate file, describe API
Vladimír Čunát [Fri, 10 Nov 2017 10:12:01 +0000 (11:12 +0100)] 
packet cache: move into a separate file, describe API

8 years agominor improvements, readability, clang compatibility
Vladimír Čunát [Fri, 10 Nov 2017 09:02:42 +0000 (10:02 +0100)] 
minor improvements, readability, clang compatibility

It doesn't accept `abs(10)` as a constant for `case` labels.

8 years agoadd packet cache - only direct NAME+TYPE hit ATM
Vladimír Čunát [Thu, 9 Nov 2017 16:32:40 +0000 (17:32 +0100)] 
add packet cache - only direct NAME+TYPE hit ATM

Used for queries that are BOGUS or
(negative and either insecure or unvalidated).

FIXME: review, opt-out, etc.

8 years agocache API: for `read/write` the key is const, really
Vladimír Čunát [Thu, 9 Nov 2017 09:29:56 +0000 (10:29 +0100)] 
cache API: for `read/write` the key is const, really

It might better be passed by value, but let me keep the way for now.

8 years agoMerge tag 'v1.5.0' into cache-aggr-wip
Vladimír Čunát [Wed, 8 Nov 2017 09:10:07 +0000 (10:10 +0100)] 
Merge tag 'v1.5.0' into cache-aggr-wip

8 years agoMerge branch 'build-docs-submodules' into 'master'
Petr Špaček [Mon, 6 Nov 2017 15:08:13 +0000 (16:08 +0100)] 
Merge branch 'build-docs-submodules' into 'master'

Add note about initializing submodules needed to build successfully.

See merge request knot/knot-resolver!387

8 years agoAdd note about initializing submodules needed to build successfully.
Stepan Balazik [Mon, 6 Nov 2017 14:55:08 +0000 (15:55 +0100)] 
Add note about initializing submodules needed to build successfully.

8 years agoMerge !385: release 1.5.0 v1.5.0
Vladimír Čunát [Thu, 2 Nov 2017 11:37:43 +0000 (12:37 +0100)] 
Merge !385: release 1.5.0

8 years agorelease 1.5.0
Vladimír Čunát [Thu, 2 Nov 2017 11:22:16 +0000 (12:22 +0100)] 
release 1.5.0

8 years agoMerge !384: add KR_RANK_TRY
Vladimír Čunát [Thu, 2 Nov 2017 11:19:16 +0000 (12:19 +0100)] 
Merge !384: add KR_RANK_TRY

attempt validation for more records but require it for fewer of them
(e.g. avoids SERVFAIL when server adds extra records but omits RRSIGs)

8 years agoadd KR_RANK_TRY
Vladimír Čunát [Wed, 1 Nov 2017 15:36:34 +0000 (16:36 +0100)] 
add KR_RANK_TRY

attempt validation for more records but require it for fewer of them
(e.g. avoids SERVFAIL when server adds extra records but omits RRSIGs)

8 years agoMerge !383: rfc8145: DNSSEC Key Tag Signaling (by default)
Vladimír Čunát [Thu, 2 Nov 2017 11:07:57 +0000 (12:07 +0100)] 
Merge !383: rfc8145: DNSSEC Key Tag Signaling (by default)

8 years agoSignaling trust anchor is enabled by default
Vítězslav Kříž [Wed, 1 Nov 2017 17:52:40 +0000 (18:52 +0100)] 
Signaling trust anchor is enabled by default

8 years agoSignaling Trust Anchor Knowledge in DNSSEC using Key Tag Query
Vítězslav Kříž [Mon, 30 Oct 2017 15:59:37 +0000 (16:59 +0100)] 
Signaling Trust Anchor Knowledge in DNSSEC using Key Tag Query

Implementation of RFC 8145 section 5 as module.

Fixes: #383
fixup! Signaling Trust Anchor Knowledge in DNSSEC using Key Tag Query

8 years agodaemon: add ability to load modules with _ in name
Petr Špaček [Wed, 1 Nov 2017 16:52:12 +0000 (17:52 +0100)] 
daemon: add ability to load modules with _ in name