]> git.ipfire.org Git - thirdparty/knot-resolver.git/log
thirdparty/knot-resolver.git
5 years agolib/generic/array: improve the growth strategy hints-huge-tmp
Vladimír Čunát [Mon, 9 Mar 2020 14:39:15 +0000 (15:39 +0100)] 
lib/generic/array: improve the growth strategy

For large arrays it was linear, so huge amount of +1 would need
quadratic time together.  I've been hating that for a long time,
but now I finally have a use case where it makes a large difference.
The one from GCC looks good to me (theoretically) and it surely has
lots of practical deployment.

CI scan-build: I still have no idea about these array allocation
errors; I had just given up and believe they're false alarms.

5 years agokr_zonecut_add(): allow not checking for duplicates 60626
Vladimír Čunát [Mon, 9 Mar 2020 14:10:24 +0000 (15:10 +0100)] 
kr_zonecut_add(): allow not checking for duplicates

Large block-lists in hints will end up in keeping a huge reverse map
for some IP address(es), and our simple duplicate-checking algorithm
can't handle that efficiently.  The duplicate checking in hints seems
better removed, so that add-del pairs work well in that case.

5 years agoMerge branch 'compiler-warnings' into 'master'
Petr Špaček [Mon, 9 Mar 2020 13:22:05 +0000 (14:22 +0100)] 
Merge branch 'compiler-warnings' into 'master'

compiler warning improvements

See merge request knot/knot-resolver!952

5 years agoMerge branch 'ci-disable-epel8-check' into 'master'
Tomas Krizek [Mon, 9 Mar 2020 11:01:00 +0000 (12:01 +0100)] 
Merge branch 'ci-disable-epel8-check' into 'master'

ci: disable build check for EPEL8 (still unresolvable)

See merge request knot/knot-resolver!953

5 years agoci: disable build check for EPEL8 (still unresolvable)
Tomas Krizek [Mon, 9 Mar 2020 10:43:23 +0000 (11:43 +0100)] 
ci: disable build check for EPEL8 (still unresolvable)

5 years agomake bad <lua.h> a hard informative error
Vladimír Čunát [Thu, 27 Feb 2020 10:24:00 +0000 (11:24 +0100)] 
make bad <lua.h> a hard informative error

5 years agodaemon/main: -Wformat in info message
Vladimír Čunát [Sun, 22 Dec 2019 09:24:49 +0000 (10:24 +0100)] 
daemon/main: -Wformat in info message

Normally the types are equal, but that's not true e.g.
on 32-bit systems and macOS.

5 years agomeson: turn implicit-function warnings into hard errors
Vladimír Čunát [Tue, 25 Feb 2020 10:24:22 +0000 (11:24 +0100)] 
meson: turn implicit-function warnings into hard errors

I just saw a case with bad luajit includes that linked successfully even
with some of the macros getting this warning.

5 years agodaemon/bindings/net: avoid -Wunused-variable on non-Linux
Vladimír Čunát [Wed, 26 Feb 2020 14:50:07 +0000 (15:50 +0100)] 
daemon/bindings/net: avoid -Wunused-variable on non-Linux

5 years agoMerge !951: NEWS: I forgot to re-classify the last change obs-knot-dns-deve-jq0xxt/deployments/663 obs-knot-dns-deve-jq0xxt/deployments/665 obs-knot-dns-deve-jq0xxt/deployments/667 obs-knot-resolver-es11k1/deployments/664 obs-knot-resolver-es11k1/deployments/666 obs-knot-resolver-es11k1/deployments/668
Vladimír Čunát [Fri, 6 Mar 2020 08:15:48 +0000 (09:15 +0100)] 
Merge !951: NEWS: I forgot to re-classify the last change

5 years agoNEWS: I forgot to re-classify the last change
Vladimír Čunát [Fri, 6 Mar 2020 08:14:22 +0000 (09:14 +0100)] 
NEWS: I forgot to re-classify the last change

5 years agoMerge !949: FreeBSD-related improvements
Vladimír Čunát [Fri, 6 Mar 2020 08:10:14 +0000 (09:10 +0100)] 
Merge !949: FreeBSD-related improvements

Fixes #549

5 years agolib/cache: ignore EINVAL from space pre-allocation
Vladimír Čunát [Wed, 4 Mar 2020 12:55:15 +0000 (13:55 +0100)] 
lib/cache: ignore EINVAL from space pre-allocation

5 years agodaemon/io: extend logging for TCP Fast Open failure
Vladimír Čunát [Wed, 4 Mar 2020 12:47:02 +0000 (13:47 +0100)] 
daemon/io: extend logging for TCP Fast Open failure

In FreeBSD it's disabled by default, causing this error.
I tried disabling it on Linux via sysctl, but that didn't seem
to cause any errors.

5 years agoMerge !950: Fix spelling in module-prefill description obs-knot-dns-deve-jq0xxt/deployments/661 obs-knot-resolver-es11k1/deployments/662
Vladimír Čunát [Thu, 5 Mar 2020 17:47:32 +0000 (18:47 +0100)] 
Merge !950: Fix spelling in module-prefill description

5 years agoFix spelling error in module-prefill description
Peter Keresztes Schmidt [Thu, 5 Mar 2020 15:11:56 +0000 (16:11 +0100)] 
Fix spelling error in module-prefill description

5 years agoMerge !948: README.md: fix broken readthedocs link obs-knot-dns-deve-jq0xxt/deployments/655 obs-knot-dns-deve-jq0xxt/deployments/657 obs-knot-dns-deve-jq0xxt/deployments/659 obs-knot-resolver-es11k1/deployments/656 obs-knot-resolver-es11k1/deployments/658 obs-knot-resolver-es11k1/deployments/660
Vladimír Čunát [Mon, 2 Mar 2020 13:22:53 +0000 (14:22 +0100)] 
Merge !948: README.md: fix broken readthedocs link

5 years agoREADME.md: fix broken readthedocs link
Vladimír Čunát [Mon, 2 Mar 2020 13:10:32 +0000 (14:10 +0100)] 
README.md: fix broken readthedocs link

The nice restructuring had these side effects, unfortunately.

5 years agoMerge branch 'gc-cpu-usage' into 'master'
Tomas Krizek [Mon, 2 Mar 2020 11:02:12 +0000 (12:02 +0100)] 
Merge branch 'gc-cpu-usage' into 'master'

gc: don't re-open cache so often

See merge request knot/knot-resolver!946

5 years agogc: make the log more readable
Vladimír Čunát [Fri, 28 Feb 2020 14:37:23 +0000 (15:37 +0100)] 
gc: make the log more readable

One iteration of GC produces five lines of log (without -DDEBUG);
separating those chunks seems much nicer.

5 years agogc: make state in kr_cache_gc() explicit
Vladimír Čunát [Fri, 28 Feb 2020 14:02:27 +0000 (15:02 +0100)] 
gc: make state in kr_cache_gc() explicit

It really is nicer to work with.
Additionally re-open cache after any GC errors.

5 years agogc: don't re-open cache so often
Vladimír Čunát [Mon, 24 Feb 2020 17:32:26 +0000 (18:32 +0100)] 
gc: don't re-open cache so often

Re-opening the cache causes fsync - every interval (1s by default).
Normally that isn't noticeable, but on encrypted ZFS it causes
high CPU consumption (even when "idle").

grahamc first reported this and Mic92 tested the patch helps. Thanks!
https://gitter.im/CZ-NIC/knot-resolver?at=5e4ea2343ca8a67fb808e349

5 years agoMerge branch 'spdx-cleanup' into 'master' obs-knot-dns-deve-jq0xxt/deployments/645 obs-knot-dns-deve-jq0xxt/deployments/647 obs-knot-dns-deve-jq0xxt/deployments/649 obs-knot-dns-deve-jq0xxt/deployments/651 obs-knot-dns-deve-jq0xxt/deployments/653 obs-knot-resolver-es11k1/deployments/646 obs-knot-resolver-es11k1/deployments/648 obs-knot-resolver-es11k1/deployments/650 obs-knot-resolver-es11k1/deployments/652 obs-knot-resolver-es11k1/deployments/654
Tomas Krizek [Wed, 26 Feb 2020 13:51:51 +0000 (14:51 +0100)] 
Merge branch 'spdx-cleanup' into 'master'

spdx cleanup: fix installation and pkgtests

See merge request knot/knot-resolver!945

5 years agospdx cleanup: fix installation and pkgtests obs-knot-resolver-bs4hbr/deployments/644
Tomas Krizek [Wed, 26 Feb 2020 09:47:12 +0000 (10:47 +0100)] 
spdx cleanup: fix installation and pkgtests

5 years agoMerge !942: license cleanup obs-knot-dns-deve-jq0xxt/deployments/642 obs-knot-resolver-es11k1/deployments/643
Vladimír Čunát [Tue, 25 Feb 2020 08:42:17 +0000 (09:42 +0100)] 
Merge !942: license cleanup

5 years agoautogenerate AUTHORS file to get rid of outdated data
Petr Špaček [Mon, 17 Feb 2020 08:56:48 +0000 (09:56 +0100)] 
autogenerate AUTHORS file to get rid of outdated data

The file can be re-generated using scripts/update-authors.sh.
File .mailmap is used for name canonicalization and fresh list of
authors replaces the old one in AUTHORS file automatically.
Gitlab CI checks its content before release (on branches named release-*).

Marek Vavruša is listed in .mailmap file twice intentionally,
once as CZ.NIC employee and second time as external contributor.

5 years agohttp: document source of bootstrap/bootswatch theme
Petr Špaček [Fri, 21 Feb 2020 15:27:41 +0000 (16:27 +0100)] 
http: document source of bootstrap/bootswatch theme

Source is
git+https://github.com/twbs/bootstrap.git@a78dc3aed640a35914361b837ce24573a0515e19#yeti/bootstrap.min.css
but our version has minor modifications:
- removed link to font file
- minor font and line height adjustments
- added SPDX header

Diffing minimized CSS file is a mess so use something like
$ awk '{gsub(/{|}|;/,"&\n"); print}'
to get something more legible.

5 years agohttp: machine readable license for selectize.js
Petr Špaček [Thu, 20 Feb 2020 14:04:31 +0000 (15:04 +0100)] 
http: machine readable license for selectize.js

5 years agohttp: update selectize.js to match upstream version 0.12.6
Petr Špaček [Thu, 20 Feb 2020 13:51:04 +0000 (14:51 +0100)] 
http: update selectize.js to match upstream version 0.12.6

Originally we used version 0.12.0 but the files in our repo did not
exactly match files in the upstream repo and I could not confirm
if the changes were meaningful or just artifact of source minimization
process.

To remove uncertainity I've updated selectize.js to latest upstream version
0.12.6, did not do any extra CSS minimization, and removed unused file
selectize.min.css.

As far as I can tell it seems to work.

Source:
git+https://github.com/selectize/selectize.js.git@eb0fca364f9bd6864ae197ea58c853f2717052a9

5 years agomark bundled content with SPDX file
Petr Špaček [Tue, 18 Feb 2020 15:43:02 +0000 (16:43 +0100)] 
mark bundled content with SPDX file

The SPDX file contains hand-written metadata about source URL, version,
authors, and license.

There are two missing pieces in HTTP module:
- bootswatch theme - I could not find what theme we ship, maybe it is
  based on one of bootswatch themes but heavily modified
- selectize.js source does not match respective files in upstream but
  contains minor changes

I will deal with these in later commits.

5 years agolicense: add note that some files might have different license
Petr Špaček [Mon, 17 Feb 2020 14:00:32 +0000 (15:00 +0100)] 
license: add note that some files might have different license

5 years agotag license files according to Linux Kernel license identifier syntax
Petr Špaček [Mon, 17 Feb 2020 09:14:44 +0000 (10:14 +0100)] 
tag license files according to Linux Kernel license identifier syntax

https://www.kernel.org/doc/html/latest/process/license-rules.html#license-identifier-syntax

5 years agotreewide: machine readable license for all the rest
Petr Špaček [Fri, 14 Feb 2020 14:25:09 +0000 (15:25 +0100)] 
treewide: machine readable license for all the rest

These files did not have GNU GPL v3 boilderplate in them so
I've added machine readable tag with appropriate license.

5 years agotreewide: machine readable license for GNU GPLv3+ source code
Petr Špaček [Fri, 14 Feb 2020 10:08:59 +0000 (11:08 +0100)] 
treewide: machine readable license for GNU GPLv3+ source code

Long GNU GPLv3 boilderplate was automatically replaced
with machine readable tag.

5 years agocontrib/licenses: machine readable license
Petr Špaček [Fri, 14 Feb 2020 08:15:09 +0000 (09:15 +0100)] 
contrib/licenses: machine readable license

5 years agocontrib/ucw: machine readable license
Petr Špaček [Fri, 14 Feb 2020 08:42:01 +0000 (09:42 +0100)] 
contrib/ucw: machine readable license

5 years agocontrib/murmurhash3: machine readable license
Petr Špaček [Fri, 14 Feb 2020 08:08:02 +0000 (09:08 +0100)] 
contrib/murmurhash3: machine readable license

5 years agocontrib/json: remove dead tests
Petr Špaček [Fri, 14 Feb 2020 08:05:50 +0000 (09:05 +0100)] 
contrib/json: remove dead tests

5 years agocontrib/json: machine readable license
Petr Špaček [Fri, 14 Feb 2020 08:05:27 +0000 (09:05 +0100)] 
contrib/json: machine readable license

5 years agocontrib/compiler: remove dead tests
Petr Špaček [Fri, 14 Feb 2020 08:03:28 +0000 (09:03 +0100)] 
contrib/compiler: remove dead tests

5 years agocontrib/compiler: machine readable license
Petr Špaček [Fri, 14 Feb 2020 08:02:53 +0000 (09:02 +0100)] 
contrib/compiler: machine readable license

5 years agocontrib/asprintf: machine readable license
Petr Špaček [Fri, 14 Feb 2020 07:56:58 +0000 (08:56 +0100)] 
contrib/asprintf: machine readable license

5 years agocontrib: remove unused ccan/ilog
Petr Špaček [Fri, 14 Feb 2020 07:53:09 +0000 (08:53 +0100)] 
contrib: remove unused ccan/ilog

5 years agocontrib: remove unsed _info files from CCAN
Petr Špaček [Thu, 13 Feb 2020 09:05:07 +0000 (10:05 +0100)] 
contrib: remove unsed _info files from CCAN

5 years agoMerge branch 'tls-log-pins' into 'master'
Petr Špaček [Tue, 25 Feb 2020 07:32:26 +0000 (08:32 +0100)] 
Merge branch 'tls-log-pins' into 'master'

daemon/tls: downgrade a log line to --verbose mode

See merge request knot/knot-resolver!944

5 years agodaemon/tls: downgrade a log line to --verbose mode
Vladimír Čunát [Mon, 24 Feb 2020 12:31:11 +0000 (13:31 +0100)] 
daemon/tls: downgrade a log line to --verbose mode

It just reports a failure to show the key-pin hash in the log.

5 years agoMerge branch 'asprintf-cleanup' into 'master'
Petr Špaček [Tue, 25 Feb 2020 07:24:54 +0000 (08:24 +0100)] 
Merge branch 'asprintf-cleanup' into 'master'

asprintf cleanup

See merge request knot/knot-resolver!941

5 years agoMerge branch 'travis-homebrew-update' into 'master' obs-knot-dns-deve-jq0xxt/deployments/640 obs-knot-resolver-es11k1/deployments/641
Vladimír Čunát [Mon, 24 Feb 2020 09:24:12 +0000 (10:24 +0100)] 
Merge branch 'travis-homebrew-update' into 'master'

travis: force homebrew update to fix travis issue

See merge request knot/knot-resolver!943

5 years agotravis: force homebrew update to fix travis issue
Tomas Krizek [Mon, 24 Feb 2020 09:08:29 +0000 (10:08 +0100)] 
travis: force homebrew update to fix travis issue

Build fails in Travis CI, because package download fails. Forcing update
of homebrew should fix it.

https://travis-ci.community/t/macos-build-fails-because-of-homebrew-bundle-unknown-command/7296

5 years agomeson: actually #define HAVE_ASPRINTF
Vladimír Čunát [Tue, 18 Feb 2020 19:03:14 +0000 (20:03 +0100)] 
meson: actually #define HAVE_ASPRINTF

I believe this should now correctly enable the code iff it's needed.

This commits and the parent will probably be no-op in practice.
I believe distros commonly do use fortification by default,
and I'm not aware of any platform with kresd and without asprintf().

I considered using `conf_data.set()` in meson, but that would require
us modifying contrib/ccan/asprintf/asprintf.h to include kresconfig.h,
and that seemed weird.

5 years agocontrib/ccan/asprintf: remove fortification guard
Vladimír Čunát [Tue, 18 Feb 2020 19:01:27 +0000 (20:01 +0100)] 
contrib/ccan/asprintf: remove fortification guard

Perhaps it served some purpose in 28b230823, but it seems incorrect.

5 years agoMerge branch 'ci-docker-job' into 'master' obs-knot-dns-deve-jq0xxt/deployments/626 obs-knot-dns-deve-jq0xxt/deployments/628 obs-knot-dns-deve-jq0xxt/deployments/630 obs-knot-dns-deve-jq0xxt/deployments/632 obs-knot-dns-deve-jq0xxt/deployments/634 obs-knot-dns-deve-jq0xxt/deployments/636 obs-knot-dns-deve-jq0xxt/deployments/638 obs-knot-resolver-es11k1/deployments/627 obs-knot-resolver-es11k1/deployments/629 obs-knot-resolver-es11k1/deployments/631 obs-knot-resolver-es11k1/deployments/633 obs-knot-resolver-es11k1/deployments/635 obs-knot-resolver-es11k1/deployments/637 obs-knot-resolver-es11k1/deployments/639
Tomas Krizek [Mon, 17 Feb 2020 11:38:03 +0000 (12:38 +0100)] 
Merge branch 'ci-docker-job' into 'master'

ci: run docker job only for our repos

See merge request knot/knot-resolver!940

5 years agoci: run docker job only for our repos
Tomas Krizek [Mon, 17 Feb 2020 11:31:24 +0000 (12:31 +0100)] 
ci: run docker job only for our repos

5 years agoMerge branch 'fixup_mr_937' into 'master'
Tomas Krizek [Mon, 17 Feb 2020 10:33:04 +0000 (11:33 +0100)] 
Merge branch 'fixup_mr_937' into 'master'

fixup! use js scheme detection

See merge request knot/knot-resolver!939

5 years agoci: fix job conditions for build:darwin
Tomas Krizek [Mon, 17 Feb 2020 09:30:05 +0000 (10:30 +0100)] 
ci: fix job conditions for build:darwin

5 years agoci: run tests on contributor master branches
Tomas Krizek [Mon, 17 Feb 2020 09:03:17 +0000 (10:03 +0100)] 
ci: run tests on contributor master branches

5 years agofixup! use js scheme detection
Petr Špaček [Fri, 14 Feb 2020 16:44:18 +0000 (17:44 +0100)] 
fixup! use js scheme detection

5 years agoMerge branch 'master' into 'master' obs-knot-dns-deve-jq0xxt/deployments/612 obs-knot-dns-deve-jq0xxt/deployments/614 obs-knot-dns-deve-jq0xxt/deployments/616 obs-knot-dns-deve-jq0xxt/deployments/618 obs-knot-dns-deve-jq0xxt/deployments/620 obs-knot-dns-deve-jq0xxt/deployments/622 obs-knot-dns-deve-jq0xxt/deployments/624 obs-knot-resolver-es11k1/deployments/613 obs-knot-resolver-es11k1/deployments/615 obs-knot-resolver-es11k1/deployments/617 obs-knot-resolver-es11k1/deployments/619 obs-knot-resolver-es11k1/deployments/621 obs-knot-resolver-es11k1/deployments/623 obs-knot-resolver-es11k1/deployments/625
Tomas Krizek [Mon, 10 Feb 2020 11:38:27 +0000 (12:38 +0100)] 
Merge branch 'master' into 'master'

use js scheme detection at webmanagement page

Closes #546

See merge request knot/knot-resolver!937

5 years agoMerge branch 'base64-namespace' into 'master'
Tomas Krizek [Mon, 10 Feb 2020 10:53:01 +0000 (11:53 +0100)] 
Merge branch 'base64-namespace' into 'master'

base64_* -> kr_base64_*

See merge request knot/knot-resolver!938

5 years agobase64_* -> kr_base64_*
Vladimír Čunát [Mon, 10 Feb 2020 08:54:48 +0000 (09:54 +0100)] 
base64_* -> kr_base64_*

GnuTLS 3.3 contained these as public symbols.  This doesn't seem
important, but why not use more collision-resistant names anyway?
Details/discussion: https://github.com/CZ-NIC/knot/pull/14
Basically noone should notice, so I didn't include any NEWS entry.

5 years agouse js scheme detection
SH [Thu, 6 Feb 2020 15:24:38 +0000 (16:24 +0100)] 
use js scheme detection

5 years agoMerge branch 'release-5-0-1' into 'master' obs-knot-dns-deve-jq0xxt/deployments/602 obs-knot-dns-deve-jq0xxt/deployments/604 obs-knot-dns-deve-jq0xxt/deployments/606 obs-knot-dns-deve-jq0xxt/deployments/608 obs-knot-dns-deve-jq0xxt/deployments/610 obs-knot-resolver-8xyvhu/deployments/601 obs-knot-resolver-es11k1/deployments/603 obs-knot-resolver-es11k1/deployments/605 obs-knot-resolver-es11k1/deployments/607 obs-knot-resolver-es11k1/deployments/609 obs-knot-resolver-es11k1/deployments/611 obs-knot-resolver-kv62s6/deployments/600 v5.0.1
Tomas Krizek [Wed, 5 Feb 2020 15:33:05 +0000 (16:33 +0100)] 
Merge branch 'release-5-0-1' into 'master'

release 5.0.1

See merge request knot/knot-resolver!936

5 years agorelease 5.0.1
Tomas Krizek [Wed, 5 Feb 2020 15:16:19 +0000 (16:16 +0100)] 
release 5.0.1

5 years agoMerge branch 'deb-restart-kres-gc' into 'master' obs-knot-dns-deve-jq0xxt/deployments/596 obs-knot-dns-deve-jq0xxt/deployments/598 obs-knot-resolver-es11k1/deployments/597 obs-knot-resolver-es11k1/deployments/599
Tomas Krizek [Mon, 3 Feb 2020 18:19:00 +0000 (19:19 +0100)] 
Merge branch 'deb-restart-kres-gc' into 'master'

packaging: ensure GC is restarted on pkg update

See merge request knot/knot-resolver!935

5 years agodistro/rpm: remove useless %systemd_post obs-knot-resolver-bs4hbr/deployments/595
Tomas Krizek [Mon, 3 Feb 2020 12:54:53 +0000 (13:54 +0100)] 
distro/rpm: remove useless %systemd_post

%system_post creates a preset, which we don't use (but could perhaps
replace the symlink for kresd.target).

5 years agodistro/rpm: fix typo in description
rickhg12hs [Sun, 2 Feb 2020 18:37:14 +0000 (19:37 +0100)] 
distro/rpm: fix typo in description

5 years agodistro/rpm: ensure GC is restarted on upgrade
Tomas Krizek [Fri, 31 Jan 2020 10:40:12 +0000 (11:40 +0100)] 
distro/rpm: ensure GC is restarted on upgrade

5 years agodistro/deb: ensure GC is restarted on pkg configure obs-knot-resolver-bs4hbr/deployments/583
Tomas Krizek [Fri, 31 Jan 2020 09:21:34 +0000 (10:21 +0100)] 
distro/deb: ensure GC is restarted on pkg configure

5 years agoMerge branch 'nitpicks' into 'master' obs-knot-dns-deve-jq0xxt/deployments/581 obs-knot-dns-deve-jq0xxt/deployments/585 obs-knot-dns-deve-jq0xxt/deployments/587 obs-knot-dns-deve-jq0xxt/deployments/589 obs-knot-resolver-bs4hbr/deployments/580 obs-knot-resolver-es11k1/deployments/582 obs-knot-resolver-es11k1/deployments/586 obs-knot-resolver-es11k1/deployments/588 obs-knot-resolver-es11k1/deployments/590
Tomas Krizek [Thu, 30 Jan 2020 11:28:53 +0000 (12:28 +0100)] 
Merge branch 'nitpicks' into 'master'

modules/prefill docs: fix dependencies

See merge request knot/knot-resolver!934

5 years agomodules/prefill docs: fix dependencies
Vladimír Čunát [Thu, 30 Jan 2020 10:42:51 +0000 (11:42 +0100)] 
modules/prefill docs: fix dependencies

LFS isn't used since !912 (6d209ef99).

5 years agoMerge !932: add cache.fssize() obs-knot-dns-deve-jq0xxt/deployments/578 obs-knot-resolver-es11k1/deployments/579
Vladimír Čunát [Wed, 29 Jan 2020 16:07:41 +0000 (17:07 +0100)] 
Merge !932: add cache.fssize()

Closes #524.

5 years agodoc/cache: remove SIGBUS note, since cache is preallocated
Tomas Krizek [Tue, 28 Jan 2020 14:39:29 +0000 (15:39 +0100)] 
doc/cache: remove SIGBUS note, since cache is preallocated

5 years agodoc/cache: add note explaining cache size choice
Tomas Krizek [Tue, 28 Jan 2020 14:38:37 +0000 (15:38 +0100)] 
doc/cache: add note explaining cache size choice

5 years agocache: Add cache.fssize() - filesystem size where the cache resides
Vladimír Čunát [Wed, 8 Jan 2020 16:39:53 +0000 (17:39 +0100)] 
cache: Add cache.fssize() - filesystem size where the cache resides

5 years agoMerge !931: systemd/kres-cache-gc: use correct cache directory
Vladimír Čunát [Wed, 29 Jan 2020 10:29:37 +0000 (11:29 +0100)] 
Merge !931: systemd/kres-cache-gc: use correct cache directory

Fixes #543.

5 years agosystemd/kres-cache-gc: use correct cache directory obs-knot-resolver-bs4hbr/deployments/575
Tomas Krizek [Tue, 28 Jan 2020 10:36:32 +0000 (11:36 +0100)] 
systemd/kres-cache-gc: use correct cache directory

In 5.0.0, the cache directory has been moved and the previous path is
incorrect.

Fixes #543

5 years agoMerge branch 'ci-minor-fixes' into 'master' obs-knot-dns-deve-jq0xxt/deployments/576 obs-knot-resolver-es11k1/deployments/577
Tomas Krizek [Tue, 28 Jan 2020 11:02:12 +0000 (12:02 +0100)] 
Merge branch 'ci-minor-fixes' into 'master'

ci: minor fixes

See merge request knot/knot-resolver!929

5 years agoci: remove uneeded dependencies from deploytests
Tomas Krizek [Mon, 27 Jan 2020 16:28:54 +0000 (17:28 +0100)] 
ci: remove uneeded dependencies from deploytests

5 years agoci: increase doc artifacts expiration time
Tomas Krizek [Mon, 27 Jan 2020 16:28:29 +0000 (17:28 +0100)] 
ci: increase doc artifacts expiration time

5 years agoMerge !930: systemd: explicitly add kresd@ to system-kresd.slice obs-knot-dns-deve-jq0xxt/deployments/573 obs-knot-resolver-es11k1/deployments/574
Vladimír Čunát [Mon, 27 Jan 2020 17:39:49 +0000 (18:39 +0100)] 
Merge !930: systemd: explicitly add kresd@ to system-kresd.slice

5 years agosystemd: explicitly add kresd@ to system-kresd.slice
Vladimír Čunát [Mon, 27 Jan 2020 16:42:18 +0000 (17:42 +0100)] 
systemd: explicitly add kresd@ to system-kresd.slice

By some magic (unknown so far) the service was ending up in the slice
anyway, but it seems much better to put it there explicitly.
Closely related to !925.

5 years agoMerge branch 'release-5-0-0' into 'master' obs-knot-resolver-8xyvhu/deployments/572 obs-knot-resolver-kv62s6/deployments/571 v5.0.0
Tomas Krizek [Mon, 27 Jan 2020 12:37:09 +0000 (13:37 +0100)] 
Merge branch 'release-5-0-0' into 'master'

release 5.0.0

See merge request knot/knot-resolver!928

5 years agorelease 5.0.0
Tomas Krizek [Mon, 27 Jan 2020 11:18:01 +0000 (12:18 +0100)] 
release 5.0.0

5 years agoMerge branch '5.0-upgrade' into 'master'
Tomas Krizek [Mon, 27 Jan 2020 11:07:53 +0000 (12:07 +0100)] 
Merge branch '5.0-upgrade' into 'master'

upgrade script for 5.0

See merge request knot/knot-resolver!927

5 years agodaemon/engine: prevent re-definition of the_args obs-knot-resolver-bs4hbr/deployments/564
Tomas Krizek [Thu, 23 Jan 2020 11:13:19 +0000 (12:13 +0100)] 
daemon/engine: prevent re-definition of the_args

5 years agodistro/deb: ensure tmpfiles.d is created before service restart
Tomas Krizek [Thu, 23 Jan 2020 11:08:25 +0000 (12:08 +0100)] 
distro/deb: ensure tmpfiles.d is created before service restart

5 years agodaemon/ffimodule: add missing extern to kr_layer_t_static
Tomas Krizek [Wed, 22 Jan 2020 12:55:48 +0000 (13:55 +0100)] 
daemon/ffimodule: add missing extern to kr_layer_t_static

This also fixes a build error with GCC 10 on fedora.

5 years agodistro/rpm: correctly package and create tmpfiles dirs
Tomas Krizek [Wed, 22 Jan 2020 12:28:32 +0000 (13:28 +0100)] 
distro/rpm: correctly package and create tmpfiles dirs

5 years agodistro/rpm: make sure the rundir is writable by knot-resolver
Tomas Krizek [Wed, 22 Jan 2020 11:24:53 +0000 (12:24 +0100)] 
distro/rpm: make sure the rundir is writable by knot-resolver

5 years agodistro/rpm: don't disable non-existent kresd@* on uninstall
Tomas Krizek [Wed, 22 Jan 2020 11:23:39 +0000 (12:23 +0100)] 
distro/rpm: don't disable non-existent kresd@* on uninstall

The macro calls systemctl disable, which doesn't accept wildcards.
Disabling kresd.target should be sufficient

5 years agodistro/deb: save systemd socket conf before http module uninstall
Tomas Krizek [Wed, 22 Jan 2020 11:23:30 +0000 (12:23 +0100)] 
distro/deb: save systemd socket conf before http module uninstall

5 years agodistro/deb: use daemon-reload to update units before restart
Tomas Krizek [Wed, 22 Jan 2020 09:20:11 +0000 (10:20 +0100)] 
distro/deb: use daemon-reload to update units before restart

5 years agodoc/upgrading: mention path to generated config with network interfaces
Tomas Krizek [Tue, 21 Jan 2020 16:11:43 +0000 (17:11 +0100)] 
doc/upgrading: mention path to generated config with network interfaces

5 years agoutils/upgrade: minor changes to upgrade script
Tomas Krizek [Tue, 21 Jan 2020 14:22:15 +0000 (15:22 +0100)] 
utils/upgrade: minor changes to upgrade script

5 years agoupgrade: use /var/lib path for config instead of /etc
Tomas Krizek [Tue, 21 Jan 2020 14:00:01 +0000 (15:00 +0100)] 
upgrade: use /var/lib path for config instead of /etc

5 years agodoc: add net.listen() example with multiple IPs
Tomas Krizek [Tue, 21 Jan 2020 13:58:32 +0000 (14:58 +0100)] 
doc: add net.listen() example with multiple IPs

5 years agodaemon/main: set non-buffering stdout/stderr early
Tomas Krizek [Mon, 20 Jan 2020 16:49:28 +0000 (17:49 +0100)] 
daemon/main: set non-buffering stdout/stderr early

5 years agodistro: generate separate config from upgrade script
Tomas Krizek [Tue, 21 Jan 2020 12:45:30 +0000 (13:45 +0100)] 
distro: generate separate config from upgrade script

5 years agodaemon/main: set non-buffering stdout/stderr early
Tomas Krizek [Mon, 20 Jan 2020 16:49:28 +0000 (17:49 +0100)] 
daemon/main: set non-buffering stdout/stderr early

Journald was missing some error messages when kresd failed to start.

5 years agoluacheck: allow globals for upgrade script
Tomas Krizek [Mon, 20 Jan 2020 14:48:04 +0000 (15:48 +0100)] 
luacheck: allow globals for upgrade script