Include scarthgap and kirkstone into the push-triggered builds. While we
are at it, drop the honister, which is EOL
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Switch to the Debian container in order to get away from the static
shadow compilation issues.
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Enable building of the initramfs so that we test for possible dependency
issues between the kernel and the initramfs.
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Switch to the Debian container in order to get away from the static
shadow compilation issues.
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
The kirkstone branch wasn't migrated to qcom-armv7a-modem machine.
Continue using sdx55-mtp for the CI on that branch.
Fixes: 61f8d5b5af ("CI: switch from sdx55-mtp to qcom-armv7a-modem")
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
The leading '/' confuses bitbake which is unable to show the layer
name when showing the build configuration:
Build Configuration:
BB_VERSION = "2.2.0"
BUILD_SYS = "x86_64-linux"
NATIVELSBSTRING = "universal"
TARGET_SYS = "x86_64-poky-linux"
MACHINE = "qemux86-64"
DISTRO = "poky"
DISTRO_VERSION = "4.1"
TUNE_FEATURES = "m64 core2"
TARGET_FPU = ""
meta
meta-poky
meta-yocto-bsp = "HEAD:88385340d3ffea6071d18a8ef76138e485a2e6f6"
= "foo:d396ff38b9391f73ee98508628565fe084447623"
Without the leading slash, it is ok.
Signed-off-by: Nicolas Dechesne <nicolas.dechesne@linaro.org>
secrets: inherit can only be used with workflows that call reusable
workflows in the same organization. In this case since we reuse a
workflow from 96boards, we need to be more explicit.
Signed-off-by: Nicolas Dechesne <nicolas.dechesne@linaro.org>
Add an additional OE RPB build for each new meta-qcom PR, to ensure
that the PR is not breaking the OE RPB build.
Signed-off-by: Nicolas Dechesne <nicolas.dechesne@linaro.org>
artifacts now accepts a list of path relative to deploy_dir
to publish after the build. All the files from deploy_dir
are published if artifacts field is an empty list.
Signed-off-by: Vishal Bhoj <vishal.bhoj@linaro.org>
Tux had a convenient bug that allowed us to use a 'ref' type with a
SHA. This bug was fixed, so we need to make sure that we use a type
'sha' in the Tux plan when we use a SHA value.
Signed-off-by: Nicolas Dechesne <nicolas.dechesne@linaro.org>
As per Github Actions documentations, "Scheduled workflows run on the
latest commit on the default or base branch". So let's implement the
daily build in the default branch which triggers plans for all
supported branches.
Signed-off-by: Nicolas Dechesne <nicolas.dechesne@linaro.org>
For daily builds, using the 'schedule' trigger, the default branch is
being used only. So we need to be able to trigger builds by 'branch'
instead of by 'ref'. ref_type is a new input parameter, so that
calling jobs can decide to use a ref (SHA) or a branch.
Signed-off-by: Nicolas Dechesne <nicolas.dechesne@linaro.org>
When testing github workflows, it's convenient to use a secondary repo
fork, so let's make sure we don't hardcode the repo name in the
workflow, so that they work in a forked repo (such as meta-qcom-test).
Signed-off-by: Nicolas Dechesne <nicolas.dechesne@linaro.org>
Use TuxOE plans to submit multiple jobs at once, instead of matrix
build since it uses one Github Action job for each build.
Signed-off-by: Nicolas Dechesne <nicolas.dechesne@linaro.org>
In addition to testing pull_request events, also test push events. In
order to effectively reuse the template job, a few additional changes
were made, especially the addition of a couple of input params.
Signed-off-by: Nicolas Dechesne <nicolas.dechesne@linaro.org>
Add initramfs-rootfs-image and initramfs-firmware-image to test that occasional changes do not break them too.
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
