Signed-off-by: Emanuele Aina <emanuele.aina@collabora.com>

Signed-off-by: Emanuele Aina <emanuele.aina@collabora.com>

Signed-off-by: Emanuele Aina <emanuele.aina@collabora.com>

Walter Lozano authored 1 year ago and Emanuele Aina committed 1 year ago

Since v2024dev2 the file pkg-config.multiarch is not longer shipped, so
in order to detect the sysroot architecture check the arch specific
libraries folder present.

Signed-off-by: Walter Lozano <walter.lozano@collabora.com>

Walter Lozano authored 1 year ago and Emanuele Aina committed 1 year ago

Tool ade needs to find out release and distribution information for SDK and
sysroots. This was previously done reading /etc/image_version but this
custom file was dropped in favor of /etc/os-release.

In order to support both the old and the new approach fallback to
/etc/os-release if /etc/image_version is not found.

Signed-off-by: Walter Lozano <walter.lozano@collabora.com>

Walter Lozano authored 1 year ago and Emanuele Aina committed 1 year ago

Since same code is used in several places move it to new helper functions
to detect sysroots properties.

Signed-off-by: Walter Lozano <walter.lozano@collabora.com>

If no mount for / (root) at given mount point can be established before mounting a sub partition, a virtual mountpoint
with a small tmpfs (ramdisk) is set up for / (root), so also non hierarchical images without a defined root partition
can be set up. This is usable, if partitions of an image are not mountable at one hierarchical mount tree.
Unmounting the image automatically removes the ram disk.

Signed-off-by: Thomas Mittelstaedt <thomas.mittelstaedt@de.bosch.com>

Partprobe rereads disk structures to the kernel. After returning to caller the process is sometimes not completed.
So information about labels, part labels are sometimes are not available. A small sleep solves this issue so far, but
a better synchronization method would be more appopriate but too complex for such a small tool. And atm a reliable
algorithm is not clear.

Signed-off-by: Thomas Mittelstaedt <thomas.mittelstaedt@de.bosch.com>

Signed-off-by: Thomas Mittelstaedt <thomas.mittelstaedt@de.bosch.com>

Signed-off-by: Thomas Mittelstaedt <thomas.mittelstaedt@de.bosch.com>

Signed-off-by: Thomas Mittelstaedt <thomas.mittelstaedt@de.bosch.com>

When the soname changes, the name of the binary package changes
accordingly. In this case, ABI breakage is expected thus there is
no need to run abi-compliance-checker. apertis-abi-compare detects
this case and report the soname change without failing.

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

This flag requires a string explaining the reason of importing
a package. This reason will be written in debian/changelog and
will be useful for a future reconsideration of the value of this
package in Apertis.

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

When importing Debian packages it is possible to find cases where packages
already provided an upstream Gitlab CI configuration. In such cases,
when packages are imported Gitlab will try to run CI which has no sense.

For properly handle this scenario, create new repositories with CI
disabled, and let gitlab rulez to update the settings once the package
is fully imported.

Signed-off-by: Walter Lozano <walter.lozano@collabora.com>

Signed-off-by: mtt2hi <thomas.mittelstaedt@de.bosch.com>

Mountimage can be used to mount disks and system images at a given mount point.
One purpose is usage for devroot.

Signed-off-by: mtt2hi <thomas.mittelstaedt@de.bosch.com>

For consistency across all Apertis packages.

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

Due to a bug in abi-compliance-checker, std headers are occasionally
used for analyzing leading to wrongly detect an ABI breakage.
We globally exclude them to avoid this noise until it is fixed in ACC.

See: https://github.com/lvc/abi-compliance-checker/issues/85



Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

to be consistent with other arguments passed to run_abi_checker()

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

This summary is useful in case of multiple libraries to
easily and quickly identify which ones have issues.

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

The call of the abi-compliance-checker might fail due to
errors. Possible reasons are the failed compilation of the
header files. To ease the analysis of the job logs the
additional printout might help.
Correct 'apt update' command.

Signed-off-by: Tino Lippold <Tino.Lippold@de.bosch.com>

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

VER and PATH variables need to be dynamically replaced

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

Having these headers and libraries lists in a not reproducible order
can generate random errors with abi-compliance-checker

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

When an xml descriptor file is provided, its path is not returned.
Consequently, run_abi_checker() fails due to an empty value.

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

https://phabricator.apertis.org/T8901



Signed-off-by: Ryan Gonzalez <ryan.gonzalez@collabora.com>

The script is split in a apertis-pull-upstream-kernel and a apertis-rebase-upstream-kernel
script.

The first one only updates the upstream/linux-* branch and the second
one does the actual rebase of the patches on the new upstream.

Signed-off-by: Detlev Casanova <detlev.casanova@collabora.com>

Detlev Casanova authored 2 years ago and Ariel D'Alessandro committed 2 years ago

These scripts are separated in 2 parts:
 * Get the latest kernel and try to rebase the apertis kernel on it
 * Push the changes and create merge requests for the pristine-lfs and
   apertis/v.*-security branches

Signed-off-by: Detlev Casanova <detlev.casanova@collabora.com>

This value will be used to determine if the job should fail or succeed

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>