| Age | Commit message (Collapse) | Author | Files |
|---|
|
Signed-off-by: Lucian Musat <george.l.musat@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Signed-off-by: Lucian Musat <george.l.musat@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Signed-off-by: Lucian Musat <george.l.musat@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Signed-off-by: Lucian Musat <george.l.musat@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
TEST_SUITES="auto" is useful to run all suitable tests without
having to hard-code the list. However, it did not take test
dependencies into account, which can be an issue for tests
which really depend on some other test to run first.
To fix this, modules get loaded in the order determined by
TESTS_SUITES, but then get re-ordered based on dependencies
derived from @skipUnlessPassed before running them. The original
order is used to break ties when there are no dependencies, so
reordering only occurs when really necessary.
@skipUnlessPassed gets extended such that it makes the test name
a method depends on available for inspection by the test loader
in oetest.py.
Unfortunately Python's unittest offers no API to inspect tests
in a TestSuite, so the code has to rely on implementation details
to find all tests. The worst that can happen when the implementation
changes is that tests are not found and reordering does not happen.
Signed-off-by: Patrick Ohly <patrick.ohly@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The SystemdServiceTests assume that avahi-daemon is installed,
which is not necessarily the case depending on the image being
tested.
Better check this dependency before starting the tests and skip
them if the service is not installed. This has to be done for
each test instead of for the entire module, because other
tests in the module can run without avavi.
Signed-off-by: Patrick Ohly <patrick.ohly@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Plain print has two drawbacks:
- the output is handled by different processes, causing the message
about found tests to appear randomly after the result of the initial
tests
- the output is not part of the bitbake console output
Affects image testing with testimage.bbclass.
Signed-off-by: Patrick Ohly <patrick.ohly@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Use absolute path instead of relative path for meta-skeleton.
Otherwise the case would fail if the builddir outside the poky source
tree.
Signed-off-by: Yi Zhao <yi.zhao@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The logic tries to remove the -native suffix from pn to handle this (though it
doesn't succeed, as it doesn't assign the new pn to the variable), but we need
to do more for the swspec tasks than just not set the extrapath, we also need
to change from SSTATE_PKGSPEC to SSTATE_SWSPEC. Alter to correct the spec for
these cases, and also add preconfigure to align with the current logic in
sstate.bbclass, which includes that task as well in the list of tasks to
adjust to use swspec.
[YOCTO #7563]
Signed-off-by: Christopher Larson <kergoth@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
oe.utils.both_contain() should return the result as "checkvalue" or "",
but the latest implement returns as "set(['checkvalue'])" or "";
It causes that bitbake.conf generates the wrong result of COMBINED_FEATURES,
which contains the common components in both DISTRO_FEATURE and MACHINE_FEATURES.
For example, build in Dizzy branch, COMBINED_FEATURES is "alsa usbhost ...",
but recently, COMBINED_FEATURES is like "set(['alsa']) set(['usbhost']) ...".
Signed-off-by: Jun Zhu <R01007@freescale.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Rpm database in staging area is used only by createrepo.
createrepo fails with the error
"rpmdb: BDB0060 PANIC: fatal region error detected"
if rpm database is broken during previous run of createrepo.
Made createrepo to create rpm db in $WORKDIR/rpmdb/ from scratch
for every build and architecture. This should potentially fix the
failure as every run of createrepo will be using separate db.
[YOCTO #6571]
Signed-off-by: Ed Bartosh <ed.bartosh@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Since we have support of log checking in deb/ipk [1] rootfs generation
in non Debian-based hosts fails because apt-ftparchive generates a
Warning when not find /etc/apt/apt.conf.d/ (available in Debian-based
hosts).
package_manager.py: DpkgPMIndexer add support for generate and export apt
configuration.
[1] http://git.yoctoproject.org/cgit/cgit.cgi/poky/commit/?id=86aec93902af2e2d7d73ca9a643707fcca45055c
Signed-off-by: Andreas Oberritter <obi@opendreambox.org>
Signed-off-by: Aníbal Limón <anibal.limon@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
This brings back commit d41097a970e9bfa553cd4bc3c9fad4b9073d7bd5,
which got lost during transition from image.bbclass to rootfs.py.
Signed-off-by: Andreas Oberritter <obi@opendreambox.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Moved parsing of /etc/os-release before parsing of
/etc/SuSE-release as /etc/SuSE-release is deprecated in
recent releases of OpenSuSE.
Here is the quote from /etc/SuSE-release:
/etc/SuSE-release is deprecated and will be removed in the future,
use /etc/os-release instead
Signed-off-by: Ed Bartosh <ed.bartosh@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Extract the common codes of log checking and add the ability of log
checking for deb and ipk package backend.
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
This patch involves two changes.
1. Extend the regular expression to also catch '^WARNING:' in _log_check_warn.
Warnings from bb.note or bbnote begin with 'WARNING:'. So if we decide to
catch warnings at rootfs time, we should not ignore those produced by
the build system itself.
2. Delay _log_check in rootfs process so that more warnings are likely to be
catched. Note that we should at least delay the _log_check after the
execution of ROOTFS_POSTPROCESS_COMMANDS, because we want to catch warnings
there.
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The test tries to query the rpm list with a non-root user
Signed-off-by: Lucian Musat <george.l.musat@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Indenting-only change.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
This patch creates a variable for the install_dir name so it can be
something other than /install, also by doing this we can correctly
clean up the empty directories (/install/tmp) during the clean-up
phase. The new default is /oe_install so as to not conflict with other
possible packages that might use /install to place files.
[YOCTO #7353]
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
The test_nonmatching_checksum doesn't need the default dependencies,
and they would just take extra time to generate.
Signed-off-by: Randy Witt <randy.e.witt@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
test_foo is renamed to test_nonmatching_checksum. It was a mistake to
submit it with the old name.
There was also some extra whitespace removed.
Signed-off-by: Randy Witt <randy.e.witt@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
libmatchbox links to libjpeg if it is present so just explicitly build
it and then check it appears in DEPENDS.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
With the change to externalsrc we can now handle these, so add a test
to ensure they can be built.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Whilst this test would seemingly be better placed as a runtime test,
unfortunately the runtime tests run under bitbake and you can't run
devtool within bitbake (since devtool needs to run bitbake itself).
Additionally we are testing build-time functionality as well, so
really this has to be done as an oe-selftest test.
This test does have a few perhaps unusual requirements in order to run:
* pexpect is installed
* MACHINE is set to one of the qemu machines
* runqemu tap devices have been set up
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Allow get_bb_var() to work with unexported variable values such as
MACHINE - the workaround is a little crude but should suffice for now.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Consolidate code for checking compatible recipes and consider meta and
packagegroup recipes as well as package-index and gcc-source to be
incompatible.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Add a -a/--all option to allow you to quickly reset all recipes in your
workspace.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
This fix solves the problem with the ownership of files in packages.
The do_install task was producing correct and expected output but when
the files were being put in, e.g. a rpm package, the ownership could
be different than that in the do_install task.
[YOCTO #7428]
Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com>
Signed-off-by: Fabrice Coulon <fabrice.coulon@axis.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
common errors.
And some minor tweaks like moving some errors from qemu to common.
Signed-off-by: Lucian Musat <george.l.musat@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
* Add a test for add-layer and remove-layer
* Correct message for test_bitbakelayers_showoverlayed()
* Improve test_bitbakelayers_flatten() to use a more unique name for the
temp output directory and clean it up using track_for_cleanup()
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
The following commit caused log checking for warnings not working for RPM
rootfs.
rootfs.py: ignore "NOTE:" when catching warnings
The problem is that checking for warnings is always skipped because the
following statement is always true.
if 'log_check' or 'NOTE:' in line:
This patch fixes the above problem so that warning checking in RPM rootfs
can work again.
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
The test_systemd_journal() test will fail if journalctl exits with a
non-zero exit status. The exit status is non-zero even if there are no
journal files, which is what drove the creation of this test.
[Yocto #7388]
Signed-off-by: Randy Witt <randy.e.witt@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
These occur when running images under virtualisation on machines with
high load which sometimes trigger timeouts in the kernel DMA code.
They're harmless to ignore in these cases.
Adding this since "failing" the build due to this is more annoying
thank useful.
[YOCTO #7387]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Fixes bug 6933
Signed-off-by: Lucian Musat <george.l.musat@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
The test added verifies that when a file with an absolute path in
LIC_FILES_CHKSUM changes without a corresponding change to the md5sum,
it triggers the license qa test again with a failure.
[Yocto #6450]
Signed-off-by: Randy Witt <randy.e.witt@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
oe.utils.both_contain() just does a find() on the value
rather than splitting the value and then looking in the
list of split items. The result is that if you add a
feature to MACHINE_FEATURES that itself has a substring
that matches one of the values looked for when building
COMBINED_FEATURES, you end up with an incomprehensible
error (here with "ext2i" in MACHINE_FEATURES):
ERROR: Nothing RPROVIDES 'packagegroup-base-ext2'
(but /home/balister/src/oe-core/oe-core/meta/recipes-core/
/packagegroups/packagegroup-base.bb RDEPENDS on or otherwise requires it)
Fix [YOCTO #6888].
Signed-off-by: Cristian Iorga <cristian.iorga@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
It seems that the --nofork option genuinely stops konsole from going
into the background now; I'm not sure when this changed but it does seem
to be working so we can use it. (Tested with Konsole 2.10 and 2.14.2).
Fixes [YOCTO #4934].
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Fix up some issues introduced by OE-Core commit
818c94f5b9882c2028ef9f056714a0a3c9045551:
* If we want to support versions with more than two parts, versions with
only one part, or versions with non-integer parts, then we have to
stay with strings. We can use distutils.version.LooseVersion() to help
with comparisons.
* We don't want a warning when launching gnome-terminal 3.10+ and
logger.warn() doesn't take a first integer parameter anyway
(logger.debug() does).
* Also clean up tabs.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
--disable-factory has been disabled in earlier versions of gnome-terminal
but from version 3.10 it raises an error and quits. This makes devshell
unusable with gnome-terminal >= 3.10. This patch checks for the version and
removes --disable-factory if you have the terminal version 3.10 or higher.
Signed-off-by: Sven Ebenfeld <sven.ebenfeld@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Added the helper functions necessary to copy the sstate from the
current build, and generate the file to "lock" it.
Signed-off-by: Randy Witt <randy.e.witt@linux.intel.com>
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
This file provides a way to take bitbake and the layers in the
current build and copy them to a target specified.
Signed-off-by: Randy Witt <randy.e.witt@linux.intel.com>
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Because we move the workdir when extracting source, then move the source
and delete the temporary workdir, you lose the indirection symlink
pointed to by the alternates file (which is created when the fetcher
clones it from DL_DIR with -s) and the resulting repository is broken.
In any case, for a source repo that the user may put their own changes
into, we can't really rely on a clone made with -s in case the
original goes away - because of cleanall, DL_DIR disappearing, etc. So
repack the repository so that it is a complete, non-shared clone after
unpacking.
(While I'm at it, add a test for devtool modify with a git recipe which
verifies that this works.)
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
When updating git-based recipes, in a lot of cases what you want is to
push the changes to the repository and update SRCREV rather than to
apply patches within the recipe. Updating SRCREV is now the default
behaviour for recipes that fetch from git, but this can be overridden
in both directions using a new -m/--mode option.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
* Ensure it knows where to put SRCREV and S
* Handle prepend/append and functions in general
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
If you want to be able to make use of libraries in conjunction with
devtool then we need to install them into the sysroot for other recipes
to use. Make it a configuration option in case it needs to be changed at
runtime.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
If you've added a new recipe, you want the output cleaned when you do
devtool reset, otherwise cruft from building the recipe may remain which
could interfere with future builds.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
For recipes that have their actual source in a subdirectory of what is
fetched (e.g. mkelfimage), we need to find the root of the repository
within the GitApplyTree code that attempts to set up the required git
hooks and use that, rather than expecting the root to be the same as
${S}.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
There is no good reason not to use ext4 at this point, it has advantages
and few drawbacks. Therefore switch the qemu machines over (and the default
runqemu script options).
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
bitbake-layers subcommands with underscores are the old syntax; the
dashed form has been supported (and displayed in the help text) for
quite a while now, and the old syntax is about to be unsupported, so use
the dashed form in the tests.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
