aboutsummaryrefslogtreecommitdiff
path: root/.github
Commit message (Collapse)AuthorAge
* ci: Use openwrt/gh-action-sdk@v5Jeffery To2022-05-31
| | | | | | | The previous build errors with v5 have been fixed. This version builds packages as a normal user instead of as root. Signed-off-by: Jeffery To <jeffery.to@gmail.com>
* Revert "ci: Use openwrt/gh-action-sdk@v5"Rosen Penev2022-05-18
| | | | This reverts commit 7b7265293f953c71a65099ecbdbfbaf5f14bf4f8.
* ci: Use openwrt/gh-action-sdk@v5Jeffery To2022-05-18
| | | | | | This version builds packages as a normal user instead of as root. Signed-off-by: Jeffery To <jeffery.to@gmail.com>
* ci: Look for changed packages in the PR branch onlyJeffery To2022-05-14
| | | | | | Fixes https://github.com/openwrt/packages/issues/15267. Signed-off-by: Jeffery To <jeffery.to@gmail.com>
* github: fix typo in issue templateSebastian Kemper2022-05-07
| | | | Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
* ci: use openwrt/gh-action-sdk@v4Petr Štetiar2022-03-24
| | | | | | | In order to use feeds from GH mirror for GH actions, thus saving a lot of resources being wasted. While at it fix whitespace issue. Signed-off-by: Petr Štetiar <ynezz@true.cz>
* CI: fix runtime testing for non master branchPaul Spooren2022-01-14
| | | | | | | The runtime testing always ran on master branch aka snapshots since the branch wasn't passed over to the container execution! Signed-off-by: Paul Spooren <mail@aparcar.org>
* CI: do not crash during PKG-INFO generation if there are no packagesPeter van Dijk2021-11-09
| | | | Signed-off-by: Peter van Dijk <peter.van.dijk@powerdns.com>
* CI: add PKG-INFO metadata filePeter van Dijk2021-11-08
| | | | Signed-off-by: Peter van Dijk <peter.van.dijk@powerdns.com>
* artifacts: use git commit sha in namePeter van Dijk2021-11-05
| | | | Signed-off-by: Peter van Dijk <peter.van.dijk@powerdns.com>
* github-ci: also build mipsel_24kc / mt7621Kevin David2021-10-09
| | | | | | Signed-off-by: Kevin David <kevin.p.david@gmail.com> This seems like a fairly popular configuration and is at least handy for me for temporary testing.
* CI: update SDK action to v3Paul Spooren2021-10-09
| | | | | | No also check mirror hashes and badly formated init files Signed-off-by: Paul Spooren <mail@aparcar.org>
* github-ci: fix small typoMichael Heimpold2021-09-28
| | | | Signed-off-by: Michael Heimpold <mhei@heimpold.de>
* CI: revert adding of *depending* packagesPaul Spooren2021-03-25
| | | | | | | | | | | | | | | | | | While the idea may make sense the current implementation is faulty. Problem is that OpenWrt uses the folder name of packages within the build system while `opkg` spits out the actual packages names. An example, compiling the packages of folder `vim` (`make package/vim/compile`) creates a package called `xxd`, where `make package/xxd/compile` would fail. The current implementation uses `opkg` to figure out dependent packages, but the resulting names do not match the above mentioned folders. Revert this for now until we come up with a better implementation to avoid false positive CI failures. Signed-off-by: Paul Spooren <mail@aparcar.org>
* CI: test build dependent packagesPaul Spooren2021-03-23
| | | | | | | | | | | | | | By using OPKGs `whatdepends` all packages dependent on a library are printed. Use that feature to obtain packages which a version change may break and build them as well. Change "default" packages to contain a lib on which other packages depend, instead of compiling `tmux` compile `attendedsysupgrade-common` on which other packages depend. Ignore any LuCI packages which only contains translations. Signed-off-by: Paul Spooren <mail@aparcar.org>
* CI: reformat extra long linePaul Spooren2021-03-23
| | | | | | Better readability on split screens. Signed-off-by: Paul Spooren <mail@aparcar.org>
* CI: define matrix individually and add targetsPaul Spooren2021-03-23
| | | | | | | This allows more specific definition per test, like on which target packages should be compiled. Signed-off-by: Paul Spooren <mail@aparcar.org>
* CI: use new `openwrt` Docker usernamePaul Spooren2021-03-17
| | | | | | | We now own `openwrtorg` and `openwrt`, where the latter replaces the former. Slowly migrate over. Signed-off-by: Paul Spooren <mail@aparcar.org>
* CI: checkout HEAD commit rather than merge commitPaul Spooren2021-03-04
| | | | | | | GitHub CI actions/checkout uses a merge commit which isn't compatible with our formality checks. Instead checkout the pull request HEAD. Signed-off-by: Paul Spooren <mail@aparcar.org>
* CI: migrate formal checks from CircleCI to GitHubPaul Spooren2021-03-04
| | | | | | Run the formal checks like SoB message via the GitHub CI. Signed-off-by: Paul Spooren <mail@aparcar.org>
* CI: fix handling of release branchesPaul Spooren2021-03-01
| | | | | | | | | | The "changed package" logic triggers all packages changed since `origin/master` while for releases branches all changes since e.g. `origin/openwrt-21.02` should be considered. First figure out the active branch, then find changed packages. Signed-off-by: Paul Spooren <mail@aparcar.org>
* CI: enhance package detection logicIlya Lipnitskiy2021-02-23
| | | | Signed-off-by: Ilya Lipnitskiy <ilya.lipnitskiy@gmail.com>
* CI: switch back to ARCHSRosen Penev2020-12-30
| | | | | | | Upstream wants to remove ARC700. That and currently, ARC700 is not compiling with glibc. Signed-off-by: Rosen Penev <rosenp@gmail.com>
* CI: use SDK depending on destination branchPaul Spooren2020-12-28
| | | | | | | Not only use snapshot SDKs for master test but whatever destination branch is used. Signed-off-by: Paul Spooren <mail@aparcar.org>
* CI: test aarch64_cortex-a53 instead of _genericPaul Spooren2020-12-15
| | | | | | | The aarch64_cortex-a53 architecture is used by more targets and should therefore be tested rather than the relatively rare _generic one. Signed-off-by: Paul Spooren <mail@aparcar.org>
* github: test i386_pentium-mmx instead of pentium4Rosen Penev2020-12-10
| | | | | | | It seems the former causes more compilation failures due to not having SSE. Signed-off-by: Rosen Penev <rosenp@gmail.com>
* Revert "CI: add super-linter for pull requests"Aaron Goodman2020-11-06
| | | | | | | | | | | | | Initial motivation for this commit was 'Just merge and see how people like it?' https://github.com/openwrt/packages/pull/13358#issuecomment-692309158 It seems that many contributors don't like it. https://github.com/openwrt/packages/issues/13395 So this should be reverted until issues with the linter can be resolved. This reverts commit 2b5de22202dafe849bfbce3adf1488d542ba8b24.
* CI: fix CVE-2020-15228Josef Schlehofer2020-10-16
| | | | | | | More details: https://github.blog/changelog/2020-10-01-github-actions-deprecating-set-env-and-add-path-commands/ Signed-off-by: Josef Schlehofer <pepe.schlehofer@gmail.com>
* CI: fix grammar typo when test is succesfulJosef Schlehofer2020-10-16
| | | | Signed-off-by: Josef Schlehofer <pepe.schlehofer@gmail.com>
* CI: Export few helpful variables for CI testsMichal Hrusecky2020-10-10
| | | | | | | Export PKG_NAME and PKG_VERSION to make test scripts better readable and also export a path to the helper script providing colorful output. Signed-off-by: Michal Hrusecky <michal.hrusecky@turris.com>
* CI: Add helper functionsMichal Hrusecky2020-10-10
| | | | | | | Add a shell script that can produce colorfull output to help to identify problems during CI runs. Signed-off-by: Michal Hrusecky <michal.hrusecky@turris.com>
* actions: remove arm_cortex-a9_neonRosen Penev2020-10-09
| | | | | | | This is the second ARM NEON target that is tested. It's unlikely that one will fail and the other succeed. Signed-off-by: Rosen Penev <rosenp@gmail.com>
* CI: use PKG_SOURCE instead of PKG_NAMEPaul Spooren2020-10-01
| | | | | | | | | | | | | | The PKG_NAME is the installable name of a package while PKG_SOURCE is the folder containig both `Makefile` and possibliy `test.sh` This approach previously worked for packages where both NAME and SOURCE are the same, e.g. `vim`, however fore more complex packages like `mariadb` (SOURCE) the NAMES are partly *mariadb-server-plugin-handlersocket*, which is no existing folder. With this commit the `PKG_SOURCE` is used to find the `test.sh` script. Signed-off-by: Paul Spooren <mail@aparcar.org>
* CI: get rid of the arc_archs buildbotRosen Penev2020-09-29
| | | | | | Both ARC platforms are not useful. One is enough for uClibc-ng. Signed-off-by: Rosen Penev <rosenp@gmail.com>
* github-ci: add --force-removal-of-dependent-packagesRosen Penev2020-09-19
| | | | | | | | Without this option, a package gets installed with its dependencies but those do not get removed, causing issues later on with other packages. Signed-off-by: Rosen Penev <rosenp@gmail.com>
* CI: pass only VERSION to test.shPaul Spooren2020-09-17
| | | | | | | | Currently the passed VERSION includes the release, which is usually not part of the compiled binary. Removing it simplifies the `grep` command to check for correct package output during runtime tests. Signed-off-by: Paul Spooren <mail@aparcar.org>
* CI: run also on test.sh changesPaul Spooren2020-09-17
| | | | | | | | | | | | | The CI only tests packages if a Makefile changed, e.g. containing a version or release bump. This covers package related files as at least the package release must change whenever a file was touched. The `test.sh` file is a runtime test used to verify working packages within OpenWrt containers. This file can independently change and will never be included in the package ipk files, therefore trigger the CI on its changes as well. Signed-off-by: Paul Spooren <mail@aparcar.org>
* CI: add runtime tests for packagesPaul Spooren2020-09-16
| | | | | | | | | | | | | | | | | | | | | Additional to manual runtime tests this CI addition runs a custom test script per package. Ideally this lowers the errors of package bumps, something which is time consuming when done manually for multiple architectures. This CI uses the official OpenWrt containers and tries to install and run compiled packages. The run depends on the content of `test.sh`, which is an `ash` script. It's called with the *packge name* and *package version* as arguments. This allows different behaviour if a single package generates multiple IPK files. The version is usable for the most trivial runtime check, e.g. `tmux -V | grep "$2"`. The current approach uses the qus project[1] which contains multiple QEMU binaries to run various architectures. [1]: https://github.com/dbhi/qus Signed-off-by: Paul Spooren <mail@aparcar.org>
* CI: add super-linter for pull requestsPaul Spooren2020-09-14
| | | | | | | | This commit adds a linter which checks most common filetypes, including Shell and Lua. Ideally this improves the quality of especially `init` scripts written in Shell. Signed-off-by: Paul Spooren <mail@aparcar.org>
* CI: add compile test for arc_archsPaul Spooren2020-09-12
| | | | | | | A package recently failed building exclusively for this architecture, therefore test it as well. Signed-off-by: Paul Spooren <mail@aparcar.org>
* CI: fix compilation of all packagesPaul Spooren2020-09-08
| | | | | | | | By using `feeds install -d y` the CI tries to install e.g. `libcxx` which fails, unrelated to the tested packages. Now follow the approach of the current CircleCI implementation. Signed-off-by: Paul Spooren <mail@aparcar.org>
* CI: fix artifact folder and add more test SDKsPaul Spooren2020-09-05
| | | | | | | | | | | | | | | | | The CI adds the `packages.git` repository to the `feeds.conf`, which makes the repository redundant. Once called `packages` including the upstream status, once `packages_ci` (previously `action`) including the PRs changes. This commit changes the binary artifact folder from `packages` to `packages_ci`, as the SDK choses packages from the modified PR branch over the `packages` branch. Also add additional targets to test, as each target only takes a few minutes to test: aarch64_cortex-a53, arm_cortex-a15_neon-vfpv4 and i386_pentium4. Signed-off-by: Paul Spooren <mail@aparcar.org>
* CI: use GitHub action for Pull RequestsPaul Spooren2020-09-04
| | | | | | | | Test all GitHub pull requests for multiple architectures and store build logs and created packages. This is the first commit of a series of patches to simplify the life of maintainers. Signed-off-by: Paul Spooren <mail@aparcar.org>
* repo: Add more information to the issue templateDaniel Engberg2019-04-07
| | | | | | | Rephrase the template a bit and add information about older releases and forks. Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
* github: adapt templatesYousong Zhou2018-01-27
| | | | | | | | | | | | | Remove refs to LEDE and use "OpenWrt" instead of "OpenWRT" Remove instructions on self-checking pull request content for the following considerations - The checks are now enforced by travis autocheck scripts - Github now prompts users to refer to the contributing guide on submitting new issue and pull request Signed-off-by: Yousong Zhou <yszhou4tech@gmail.com>
* Add an issue templateYousong Zhou2016-08-25
| | | | | | | So that maintainers can receive at least one notification at the issue creation time ;) Signed-off-by: Yousong Zhou <yszhou4tech@gmail.com>
* Improve pull_request_templateEtienne CHAMPETIER2016-07-03
| | | | | | | commit message is copied at the end of the template when creating PR, so swap instructions and template part Signed-off-by: Etienne CHAMPETIER <champetier.etienne@gmail.com>
* Improve pull_request_templateEtienne CHAMPETIER2016-07-02
| | | | Signed-off-by: Etienne CHAMPETIER <champetier.etienne@gmail.com>
* Rename pull_request_template (it's text not markdown)Etienne CHAMPETIER2016-07-02
| | | | Signed-off-by: Etienne CHAMPETIER <champetier.etienne@gmail.com>
* Add a pull request templateEtienne CHAMPETIER2016-06-17
I hope this will help triage PR and reduce commits not following guidelines Signed-off-by: Etienne CHAMPETIER <champetier.etienne@gmail.com>
Powered by cgit, Themed by Ted Yin.