Petr Špaček [Fri, 1 Dec 2017 20:30:58 +0000 (21:30 +0100)]
fixup! client: move sources into separate directory
Previous cleanup attempt broke installation on systems without libedit.
client-install has to be defined at least as .PHONY target.
(I want to avoid if spagethi outside of client.mk.)
Petr Špaček [Fri, 1 Dec 2017 13:00:36 +0000 (14:00 +0100)]
client: move sources into separate directory
Let's not mix daemon and client files as client might grow to multiple
files in future. This will also help with upcomming changes to packaging
scripts.
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.
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.
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).
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
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
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.
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.
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'
```
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.
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.
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.
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
This allows the socket to be bound before the network interface comes
up, which is preferable to having a race between the interface coming
up and the socket being bound.
systemd: drop RefuseManualStart=true, disable kresd.service by default
I introduced RefuseManualStart=true last year when writing the initial
systemd unit file for kresd. I did this because i didn't want to
start running the daemon if no one was going to talk to it.
I'm no longer convinced that this directive has a good justification.
While i think that there is no need to activate the service without
socket activation, i think preventing manual invocation is actively
harmful (admins will reflexively do "systemctl restart kresd" just
like they do "systemctl restart apache2" and it's jarring to have it
rejected).
A distributor who wants to (by default) delay kresd startup until it
is accessed can ship the preset file included here, but leave the
*.socket units enabled by default.