| Commit message (Collapse) | Author | Age | Files | Lines |
|\
| |
| | |
Use 'printf' instead of 'echo -e'
|
|/
|
|
|
|
| |
Should be less surprising
Signed-off-by: Laurent Bonnans <laurent.bonnans@here.com>
|
|\
| |
| | |
wic:plugins:otaimage.py: fix a potential issue
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
do_image_wic task is a standalone task that depending on
do_image_ota_ext4, so it's possible that do_image_wic runs
(taskhash contaminated) but do_image_ota_ext4 does not (taskhash not
contaminated), in which case, the otaimage would be in DEPLOY_DIR_IMAGE
instead of in IMGDEPLOYDIR, so we need check both of them.
Also, the logger.error is not supposed to raise a error, it just
prints out a error message, so we should use WicError.
And another typo is it checks IMGDEPLOYDIR but reports DEPLOY_DIR_IMAGE
not exist, also fix that.
Signed-off-by: Ming Liu <liu.ming50@gmail.com>
|
|\ \
| | |
| | | |
aktualizr: use echo -e when using escape sequences
|
|/ /
| |
| |
| |
| |
| |
| |
| |
| | |
Some shell require the -e argument when using escape sequences
in echo. This has been observed when building images on a
Fedora machine, where the configuration files ended up including
"\n" sequences instead of newlines.
Signed-off-by: Stefan Agner <stefan.agner@toradex.com>
|
|\ \
| | |
| | | |
fit-conf: se PACKAGE_ARCH to ${MACHINE_ARCH}
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
fit-conf should be a machine specific package instead of being a arch
specific package.
Signed-off-by: Ming Liu <liu.ming50@gmail.com>
|
|\ \ \
| | | |
| | | | |
Remove connman
|
| | | |
| | | |
| | | |
| | | | |
Signed-off-by: Laurent Bonnans <laurent.bonnans@here.com>
|
| | | |
| | | |
| | | |
| | | | |
Signed-off-by: Laurent Bonnans <laurent.bonnans@here.com>
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Now provided by connman and networkd-dhcp-conf, added to images in our
supported platforms
Signed-off-by: Laurent Bonnans <laurent.bonnans@here.com>
|
| | | |
| | | |
| | | |
| | | | |
Signed-off-by: Laurent Bonnans <laurent.bonnans@here.com>
|
| |/ /
| | |
| | |
| | |
| | |
| | |
| | | |
No sense to force that if poky does not, let's move it to
platform-specific for now.
Signed-off-by: Laurent Bonnans <laurent.bonnans@here.com>
|
|\ \ \
| |/ /
|/| | |
Simplify garage-sign fetching for aktualizr
|
| | |
| | |
| | |
| | | |
Signed-off-by: Laurent Bonnans <laurent.bonnans@here.com>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Two modes:
- `GARAGE_SIGN_AUTOVERSION=0`: use archive and version from aktualizr's recipe
- `GARAGE_SIGN_AUTOVERSION=1`: let aktualizr fetching it automatically
Signed-off-by: Laurent Bonnans <laurent.bonnans@here.com>
|
|/ /
| |
| |
| |
| |
| | |
Includes the new way of fetching garage-signa.
Signed-off-by: Laurent Bonnans <laurent.bonnans@here.com>
|
|\ \
| | |
| | | |
Move warning about insecure practise to the correct recipe
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
When SOTA_DEPLOY_CREDENTIALS got introduced deployment of the
provisioning credientials has been moved to
aktualizr-ca-implicit-prov-creds. Move the warning accordingly.
Signed-off-by: Stefan Agner <stefan.agner@toradex.com>
|
|\ \ \
| |/ /
|/| | |
README: document auto-reboot config feature.
|
|/ /
| |
| |
| |
| | |
Signed-off-by: Patrick Vacek <patrickvacek@gmail.com>
Suggested-by: Ben Luck <benjamin.luck@here.com>
|
|\ \
| | |
| | |
| | |
| | | |
advancedtelematic/fix/dont-send-mnfst-if-pending-updts
Switch to the aktualizr's master head
|
| | |
| | |
| | |
| | | |
Signed-off-by: Mike Sul <ext-mykhaylo.sul@here.com>
|
|\ \ \
| | | |
| | | | |
Remove add package test
|
| |/ /
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
It was a nice idea when we were less confident about bitbaking and
oe-selftest, but at this point, it is redundant and provides no useful
information while adding extraneous time to the already long test suite.
Signed-off-by: Patrick Vacek <patrickvacek@gmail.com>
|
|\ \ \
| |/ /
|/| | |
Add meta-python layer to all platforms
|
| | |
| | |
| | |
| | | |
Signed-off-by: Patrick Vacek <patrickvacek@gmail.com>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Put each item on its own line to make versioning and manual
modififcation easier going forward.
Signed-off-by: Patrick Vacek <patrickvacek@gmail.com>
|
|/ /
| |
| |
| |
| |
| |
| |
| | |
Needed for meta-oe since this change:
http://git.openembedded.org/meta-openembedded/commit/?id=987d28c27e70ba12edf2b99745c13e7525206f4a
Signed-off-by: Laurent Bonnans <laurent.bonnans@here.com>
|
|\ \
| | |
| | |
| | |
| | | |
advancedtelematic/fix/OTA-2211/simultaneous-bitbaking
image_types_ostree: Add a unique ref to fix simultaneous bitbaking.
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
To enable simultaneous bitbaking of two images with the same branch name, create
a new ref in the OSTree repo using the basename of the image.
As @OYTIS has pointed out, this probably won't solve every problem that might
come up with simultaneous bitbaking, such as if the two images use different
bootloaders. However, it does solve the immediate problem that comes up with our
primary-image + secondary-image example.
Signed-off-by: Patrick Vacek <patrickvacek@gmail.com>
|
|\ \ \
| | | |
| | | | |
aktualizr-auto-prov-creds: remove API Gateway URL if it exists.
|
| |/ /
| | |
| | |
| | |
| | |
| | | |
It will only be used by garage-sign.
Signed-off-by: Patrick Vacek <patrickvacek@gmail.com>
|
|\ \ \
| |/ /
|/| | |
OTA-2135: Aktualizr's config and recipe to auto reboot just after update
|
| |\ \
| |/ /
|/| | |
|
|\ \ \
| |_|/
|/| | |
Bump aktualizr (master)
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
To 2aa9d93ccea09ec327789eecf858de561ef632da, which changed the
installation result reporting format
Signed-off-by: Laurent Bonnans <laurent.bonnans@here.com>
|
| | |
| | |
| | |
| | |
| | |
| | | |
functionality
Signed-off-by: Mike Sul <ext-mykhaylo.sul@here.com>
|
| |/
|/|
| |
| | |
Signed-off-by: Mike Sul <mykhaylo.sul@innoteka.com>
|
|\ \
| |/
|/| |
sota.bbclass: fix a variable assignment regression
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
A regression was introduced by commit 9dcfcdb9:
[ classes, images: Use standard variables for initramfs ]
it replaced OSTREE_INITRAMFS_IMAGE with INITRAMFS_FSTYPES which is
fine, but the '??=' should be changed to "?=" as well, or else it will
not take effect since INITRAMFS_FSTYPES's already set in bitbake.conf
with '?='.
Signed-off-by: Ming Liu <liu.ming50@gmail.com>
|
|\ \
| | |
| | | |
Master aktualizr pkgconfig
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This patch mainly aims to fix the following issues:
- ${libdir}/sota/conf.d is being put into aktualizr package, but
${libdir}/sota/*.toml is being put into aktualizr-host-tools, this
does not make sense, if a end user only install aktualizr-host-tools
to a target, he will miss the config files in ${libdir}/sota/conf.d.
- A user should be able to install garage-deploy garage-push to a board
if he wants to do that by choosing building aktualizr with sota-tools,
but now it's not in that case, garage-deploy garage-push are only
available for native.
- It's not necessary to distinguish native/target when installing
systemd services, they will be dropped by sstate from native sysroot
any way.
- It would be better that all binaries to be put into their own
packages, so a end user can choose to install a package only with one
binary, rather than having to install the entire aktualizr-host-tools.
Introduce per-binary based packages for host tools, and gather all
config files to aktualizr-configs, it will be depended by binary
packages and main aktualizr package.
Signed-off-by: Ming Liu <liu.ming50@gmail.com>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
There are several flaws in current aktualizr recipe:
- It builds with 'BUILD_SYSTEMD=ON' by default but does not set DEPENDS
to systemd, this is not leading to compilation errors so far because
systemd is deployed into recipe sysroot by ostree if 'systemd' exists
in DISTRO_FEATURES, but in other cases, it will cause compilation
errors.
- GARAGE_SIGN_SHA256 is being passed to EXTRA_OECMAKE even it's not set
anythere, this will lead a wrong GARAGE_SIGN_SHA256_ARG to be set in
cmake files.
- pkgconfig is being called in aktualizr CMakeFile, so it needs inherit
pkgconfig.bbclass
To fix the above flaws, meanwhile to keep the code cleaner, we introduce
PACKAGECONFIG to handle the configurations.
Signed-off-by: Ming Liu <liu.ming50@gmail.com>
|
|/ /
| |
| |
| |
| |
| |
| | |
"${bindir}/aktualizr-check-discovery" is being set in both FILES_${PN}
and FILES_${PN}-tools, drop it from FILES_${PN}.
Signed-off-by: Ming Liu <liu.ming50@gmail.com>
|
|\ \
| | |
| | | |
Add support for device tree overlays in FIT images
|
| |/
| |
| |
| | |
Signed-off-by: Anton Gerasimov <anton.gerasimov@here.com>
|
|\ \
| |/
|/| |
sota.conf: use reproducible builds by default
|
|/
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Use reproducible builds by default. OSTree sets mtime to 0 at
commit time anyway. This makes sure that all mtime timestamps
are 0 already at build time. This is relevant for programs
which relies on mtime of deployed artifacts at runtime.
Python for instance writes the mtime of the source file into
bytecode files (pyc) at build time. When OSTree then clears
the mtime Python considers all bytecode files as stale.
Reproducible binaries also avoid unnecessary differences
between two OSTree commits. E.g. if the sstate cache is
cleared between two OSTree commits OpenEmbedded might create
slightly different binaries.
Note however that not all all recipe generate reproducible
binaries at this time.
Signed-off-by: Stefan Agner <stefan.agner@toradex.com>
|