If you send this forked process a SIGTERM, it will execute all of the
parent's exit code leading to two sets of console/exit output which is
extremely confusing. Wrap the code in a try/finally to ensure we always
call os._exit() to avoid this.
I spent far too long trying to work out the crazy console output from this.
(From OE-Core rev: 652e40bfae24b8e23bbf7a7f35d900d2ab8d0f92)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Add "test_uboot_atf_tee_fit_image" test caste to check u-boot FIT image and
Image Tree Source(ITS) are built and the ITS has the correct fields.
Add "test_sign_standalone_uboot_atf_tee_fit_image" test case to check if u-boot
FIT image and Image Tree Source (ITS) are created and signed correctly for the
scenario where only the u-boot proper fitImage is being created and signed.
Currently, ATF and TEE(optee-os) recipes are placed in meta-arm layer.
OpenEmbedded-Core is a basic and core meta layer. To avoid OpenEmbedded-core
depends meta-arm, both test cases are used dummy images for testing.
(From OE-Core rev: 92e51452f8831f74e0907b960135eef8cecd012a)
Signed-off-by: Jamin Lin <jamin_lin@aspeedtech.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
A few tests are getting failed with x86 arch.The unsupported/failing tests
are added to the exclude list and ignore the failing unit tests.
Upstream-Status: Pending
(From OE-Core rev: b1340173be2a3a91fbb135eb0e24e50c3b996425)
Signed-off-by: Deepesh Varatharajan <Deepesh.Varatharajan@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
ifconfig is obsolete and being removed, convert to use ip instead.
(From OE-Core rev: 3b2fb477750606976359884b18c33a37832e5a78)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The tests don't need it. On top of that, this extra requirement
creates a dependency loop between systemd-systemctl-native and util-linux.
(From OE-Core rev: b5770d8c56036bdfef8d596d27433d8f408ee36f)
Signed-off-by: Vyacheslav Yurkov <uvv.mail@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This test checks for an IP address and then tests if interface aliases work. We
don't run it on any of our automated testing as it only applies for non-qemu.
The connectivity test is unrealted to connman and pretty pointless as it depends
on ssh being working, so networking is probably ok.
The alias interface test is unrelated to commman and a general networking test
but seems out of place.
The code uses obsolete ifconfig calls and overall, the value of the test we're
never using seems low. Delete it.
(From OE-Core rev: 5e40277d1ee9df8dbc612a39b575f9a50806cd62)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
With all the pieces in place, the self test can be re-enabled.
(From OE-Core rev: 46e8b94582ea9734117d20cd62c39fb4450c00c4)
Signed-off-by: Zoltán Böszörményi <zboszor@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
If networking fails, we can get useful informaiton over the serial connection. Add
this fallback code so that any issues can be more easily debugged by showing the
host and target networking states.
(From OE-Core rev: 3291f9d07ecfe7d3301dc914f5e6a80577cf1d5d)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
As well as commit counts, it is helpful to know when metadata dates from. Store
the unix timestamp for commits in a commit_time field alongside the commit count.
This is useful for performance graph analysis and saves having to recompute the
data.
(From OE-Core rev: 56d1bc3f8f45d2f9c8ca0319c429cec562a16384)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This reverts commit 0d3901b768.
We've still having CDN issues so go back to the double attempt to
see if this improves things.
(From OE-Core rev: 136f0edb75d9601a0ae56e025419d3cddfb90219)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Introduce a new variable "OEQA_REPRODUCIBLE_TEST_LEAF_TARGETS".
It can be used like OEQA_REPRODUCIBLE_TEST_TARGET but will try to use
sstate for the dependencies.
This can be used to "quickly" test the reproducibility of a single recipe.
[YOCTO #15701]
(From OE-Core rev: e2440618bd12d7288a7f3f7cfb92d377ae61a3b4)
Signed-off-by: Yoann Congal <yoann.congal@smile.fr>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The code line was moved in
commit 5a9a5ad6a7 ("reproducibility: continue testing in case of build failure")
(From OE-Core rev: b4b60941153ebeb5019b26eeed593d0a1cdb4012)
Signed-off-by: Yoann Congal <yoann.congal@smile.fr>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Rather than hiding the multilib logic in the package tests, move the
multilib setting to the init function of the class so the logic is
clearer.
(From OE-Core rev: 7699867c3ce979ab615eb677a417bb5b82faecb7)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The 'devtool' cases make a copy of 'poky', but before this patch, that
only included modifications to the 'meta/' subdirectory.
It's very frustrating to make changes to scripts/ and have them be
silently ignored by oe-selftest.
(From OE-Core rev: 35de7080c53808ade526b3b97cb54f528357deca)
Signed-off-by: Chris Laplante <chris.laplante@agilent.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The do_testsdk for lib32-core-image-sato aborts with below error:
configure: error: Package requirements (gtk+-3.0) were not met:
No package 'gtk+-3.0' found
Consider adjusting the PKG_CONFIG_PATH environment variable if you installed software in a non-standard prefix.
This causes due to an absolute path name in 'sdk_env', which is now stripped to have only the environment name.
(From OE-Core rev: 386e4132a9ced75599d92610cf5c0e8fe907c3b6)
Signed-off-by: Harish Sadineni <Harish.Sadineni@windriver.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Adjust test_recipetool_go_create test to a smaller repository.
Combine test_recipetool_go_create and test_recipetool_go_replace_modules
tests into test_recipetool_go_create, the new test performs both tests.
(From OE-Core rev: d1256f8ff4a94d4f32b2b9ca855359dd2d691beb)
Signed-off-by: Alexander Yurkov <alexander.v.yurkov@gmail.com>
Signed-off-by: Vyacheslav Yurkov <uvv.mail@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Looking at some of the autobuilder failures, it seems that somehow empty
reads might be possible despite not being EOF. Tweak the code to be a little
more robust in handling this.
In theory this shouldn't be possible but python does handle signals a bit
differently (e.g. transparrently retrying syscalls for EINTR) so adding this
check and a bit of code safety at least rules out this problem.
(From OE-Core rev: 9d91a5674c515a43ae76d8615f72e5e2dc16c961)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The test code in poison was flawed: as long as one CPP/CC/CXX has fatal
poisoning enabled then the test passes. However, at the moment due to
a bad rebase only CPP has fatal poisoning and CC/CXX do not.
Rewrite the do_compile() task to more carefully check the output so the
test harness itself just has to bitbake the recipe.
Note that this results in the test failing:
ERROR: poison-1.0-r0 do_compile: C Compiler is not poisoned.
Exit status 0, output: cc1: warning: include location "/usr/include" is unsafe for cross-compilation [-Wpoison-system-directories]
ERROR: poison-1.0-r0 do_compile: C++ Compiler is not poisoned.
Exit status 0, output: cc1plus: warning: include location "/usr/include" is unsafe for cross-compilation [-Wpoison-system-directories]
(From OE-Core rev: 5b413d1fdb4bdbaec86d630bb52c3ccf68aae789)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Printed message looks like this:
2025-01-12 00:40:08,326 - oe-selftest - ERROR - reproducibleA build failed. Trying to compute built packages differences but the test will fail.
2025-01-12 00:40:08,326 - oe-selftest - INFO - Failure log for reproducibleA will be copied to <snip>/oe-reproducible-20250112-mza9w940/bitbake-reproducibleA.log
Fixes [YOCTO #15655] (first half).
(From OE-Core rev: 1c2287849a2742445d80c81d6a1c6c3e3668765e)
Signed-off-by: Yoann Congal <yoann.congal@smile.fr>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
We want to ensure ARM nativesdk signatures match those for the 32 and 64 bit
x86 versions. Add a test to ensure this is the case.
(From OE-Core rev: dced735e2cc9663527ff8b386bf20a5e1fb9ab07)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The cve_check functionality to parse CVE IDs from the patch filename and
patch contents have been reworked to improve parsing and also utilize
tests. This ensures that the parsing works as intended.
Additionally, the new patched_cves dict has a few issues I tried to fix
as well. If multiple patch files exist for a single CVE ID, only the
last one will show up with the "resource" key. The value for the
"resource" key has been updated to hold a list and return all patch
files associated with a given CVE ID. Also, at the end of
get_patch_cves, CVE_STATUS can overwrite an existing entry in the dict.
This could cause an issue, for example, if a CVE has been addressed via
a patch, but a CVE_STATUS line also exists that ignores the given CVE
ID. A warning has been added if this ever happens.
(From OE-Core rev: 87c6da681609b4f8e048eca2a27ae8e068c724e1)
Signed-off-by: Colin McAllister <colinmca242@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This allows fine-tuning local configurations with pre-frabricated
configuration snippets in a structured, controlled way. It's also
an important building block for bitbake-setup.
The tool requires that each fragment contains a one-line summary, and one or more
lines of description, as BB_CONF_FRAGMENT_SUMMARY style metadata.
There are three (and a half) operations (list/enable/disable/disable all), and here's the 'list' output:
alex@Zen2:/srv/storage/alex/yocto/build-64$ bitbake-config-build list-fragments
NOTE: Starting bitbake server...
Available fragments in selftest layer located in /srv/work/alex/poky/meta-selftest:
Enabled fragments:
selftest/test-fragment This is a configuration fragment intended for testing in oe-selftest context
Unused fragments:
selftest/more-fragments-here/test-another-fragment This is a second configuration fragment intended for testing in oe-selftest context
(From OE-Core rev: fdb611e13bd7aa00360d3a68e4818ef5f05c8944)
Signed-off-by: Alexander Kanavin <alex@linutronix.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
rpm2cpio has been deprecated upstream, so this prepares for its
eventual removal.
rpm2archive produces a tar archive which can be uncompressed
with tar executable from the host.
(From OE-Core rev: ed824d3fb23f0c89d8dfdacb2c4ef0b7c21a5144)
Signed-off-by: Alexander Kanavin <alex@linutronix.de>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
With this release, as expected, deprecated openpgp support
has been fully removed and replaced with sequoia:
https://github.com/rpm-software-management/rpm-sequoia/
Alas, it's written in rust, and has no recipe, so for now
signing rpm packages has to be disabled.
Remove package name parameter to %prep in source archiving, as it's
been dropped in the new rpm, and wasn't needed to begin with.
Drop 0001-perl-disable-auto-reqs.patch
(files removed upstream).
(From OE-Core rev: 8c15b4577d5e554cc2dd5adfb88b816894b05a9a)
Signed-off-by: Alexander Kanavin <alex@linutronix.de>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When used with `--newbuilddir` option, it replaces relative paths in
BBLAYERS variable with absolute paths by evaluating the final value of
the variable, converting paths and saving result by assignment
(`BBLAYERS =`) at the end of bblayers.conf.
This breaks tests when bblayers.conf contains BBLAYERS:append because
:append statements are evaluated after all assignments and we end up
with the appended layer added twice - first by evaluating final value
of the variable, then again by bitbake evaluating BBLAYERS:append.
The error is as follows:
AssertionError: Command 'bitbake -e' returned non-zero exit status 1:
ERROR: Found duplicated BBFILE_COLLECTIONS 'layer-name', check bblayers.conf or layer.conf to fix it.
(From OE-Core rev: 516c8e8f722ed0eb85a715a23b5e68d05abe9275)
Signed-off-by: Oleksandr Hnatiuk <ohnatiuk@cisco.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The SPDX code needs to be able to look up an Element by its SPDX ID,
locating the file that (should) contain the SPDX ID and opening it for
parsing. Previously, the code would do this be hashing each Element
SPDX ID and Alias, and the creating a symbolic link to the file that
contains the element with a name of the hash.
This worked well as it was possible to look up any arbitrary SPDX ID or
alias by simply hashing it and following the symbolic link to get the
file. However, the down side of this approach is that it creates a lot
of symbolic links, since it will make one or two per Element in the
document. This can be a problem when using SPDX_INCLUDE_SOURCES, for
example.
This change reworks this strategy so that the only Element that gets a
symbolic link based on the hash is the singular SpdxDocument that is
create for each file. All other Elements are assigned an alias with a
special prefix that encodes the hash of SpdxDocument alias. Thus, when
attempting to look up an arbitrary alias, the code sees the special
prefix, extract the hash, opens the file based on the symlink with that
hash name, then finds the matching Element in the file. This drastically
reduces the number of symbolic links by making only one per file.
This also means that the custom link extension can be removed since it
is now superfluous.
(From OE-Core rev: 838d64c09657ac53175737fc4e7fd6f01f3dcf47)
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Adds a test for several of the extra options provided by the SPDX
classes. In particular, these are the options that can produce
non-reproducible results, so are not enabled by default in OE core. This
test takes care to configure the build so that the tests do run in a
reproducible manner so that pre-built test objects can be pulled from
sstate
(From OE-Core rev: 14f534f15f7fe6362723d7f064d39783c5bd758f)
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Tweak the gitarchive exclude handling not to error if excluded files
don't match.
Also return the tagname created so that other code can then use it.
(From OE-Core rev: 1adba3430faffdf6217b6a00533a3b48a9388abc)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Using 4 space indentation in resulted in hundreds of megabytes of extra file size
in general use. Reduce this to make filesizes more managable and reduce the processing
cost. Some level of indentation and spacing does make the files more readable and allows
use of git diff so we need to retain some of it.
(From OE-Core rev: a274cdcaf852cca9497f0358f44dda99c06aacbe)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
There are several problems with these paths. Firstly they contain full
system paths which depend upon where the test was run. These are pretty
pointless and just take up a lot of space making the results files large.
Secondly, they contain the same path twice. The reference and target path
will always be the same thing in two different locations.
Strip off the prefix and remove the duplication. This does change the output
data but that can't really be avoided. It does shrink the results data and makes
it more readable.
(From OE-Core rev: 81a44de36e864b08687451fd85aeba7c529fd7f7)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The "rawlogs" data consists of a long string of results data which is
already in a structured data format. I can't see this is adding much
value in duplciating the data but it does create a huge string with a
lot of long problematic pathnames and inflates the results data size.
I suggest we drop this data as obsolete and not necessary.
(From OE-Core rev: 5b2c70fab2ffa409b861d83f048b65d458d03a90)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
We've seen cases where a ptest (strace) has a 43GB sparse file in the test
directory. busybox tar doesn't work well with this. The resulting 1.4GB archive
takes hours to extract too.
Ensure tar is added to our full images and use the sparse option to collect
files for debugging. This stops crazy build hangs.
Since tar is GPLv3, we have to exclude it from that test code. We don't boot
any of those images so the debug collection code is safe there, at least for now.
(From OE-Core rev: fefeb919696b6ac76f0997acfb0f612203ef7f1b)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The current code is not fit for purpose when handling large files via ssh. In the strace
ptest case, we can end up with a 1.4GB archive being transferred for which every
byte is printed into the task logfile twice over. This is then sent over bitbake IPC
which compounds the problems.
Make the following improvements:
* when the output is large (over 64kb), don't print it
* use a bytearray for better concat performance since strings are slow for this
* when there is no ssh output, say that
* print periodic size status output rather than the data itself since this could be binary and/or large
* fix the killed process message logic which appeared broken
(From OE-Core rev: e7dd009a17dc902852983a82bce41bf78bb1e242)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
If SLIRP is being used instead of TAP for networking to the guest then
the target IP will be localhost. There's no point in pinging localhost
to see if the target is up but whilst you'd think it is harmless, in
some containers ping doesn't actually have enough rights to work:
ping: socktype: SOCK_RAW
ping: socket: Operation not permitted
ping: => missing cap_net_raw+p capability or setuid?
Look at the target address and if it's localhost or 127.0.0.* return
immediately.
(From OE-Core rev: a06ef43d2a50e16c32bd6edbdc7b32c3528687d5)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
We had previously excluded this test case due to failure in earlier versions.
However, with the latest version of Rust, this test case is passing.
As a result, we have removed it from the exclude list and added back into
the test suite for execution.
The rust-1.79 has been successfully tested with the current test inclusion.
(From OE-Core rev: a33edee3efe67f81229496e03a098454dc7df051)
Signed-off-by: Deepesh Varatharajan <Deepesh.Varatharajan@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
If Unified Kernel Image was enabled via IMAGE_CLASSES, then
target should also boot the same uki at runtime.
(From OE-Core rev: 2ae651c869a19fd24b5a915451c1f99e1d7b32f0)
Signed-off-by: Mikko Rapeli <mikko.rapeli@linaro.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Assigning IMAGE_FEATURES="" in local.conf doesn't actually do anything
useful, as bitbake.conf has IMAGE_FEATURES += EXTRA_IMAGE_FEATURES after
the include of local.conf.
In this test case, this results in IMAGE_FEATURES still using
EXTRA_IMAGE_FEATURES, so the image contains the files installed by the
post-install-logging feature.
As the intention here is that container-test-image is as minimal as
possible, move the IMAGE_FEATURES assignment into the image itself so
it actually works, and update the expected file list to remove the
post-install-logging files.
(From OE-Core rev: c65349e1238ddc6634dfa4759c57e6168a355200)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Since [0], the reproducibility test tries to save logs of failed builds.
Using the memory persistent bitbake prevent this by concatenating logs in the
single persistent instance (See [YOCTO #15641]).
So, force a BitBake server restart for each build to ensure a new log
file is created.
This fixes this error (seen on AB [1]):
2024-11-02 08:16:00,816 - oe-selftest - ERROR - reproducibleA build failed. Trying to compute built packages differences but the test will fail.
2024-11-02 08:16:00,819 - oe-selftest - INFO - ... ERROR
2024-11-02 08:16:00,819 - oe-selftest - INFO - Traceback (most recent call last):
File "/srv/pokybuild/yocto-worker/reproducible-meta-oe/build/meta/lib/oeqa/selftest/cases/reproducible.py", line 321, in test_reproducible_builds
self.copy_file(variables["BB_CONSOLELOG"], os.path.join(save_dir, "bitbake-%s.log" % name))
File "/srv/pokybuild/yocto-worker/reproducible-meta-oe/build/meta/lib/oeqa/selftest/cases/reproducible.py", line 216, in copy_file
shutil.copyfile(source, dest)
File "/usr/lib/python3.11/shutil.py", line 256, in copyfile
with open(src, 'rb') as fsrc:
^^^^^^^^^^^^^^^
FileNotFoundError: [Errno 2] No such file or directory: '/srv/pokybuild/yocto-worker/reproducible-meta-oe/build/build/build-st-meta-python/build-st/tmp/log/cooker/qemux86-64/20241102071141.log'
[0]: https://git.openembedded.org/openembedded-core/commit/?id=c78cc753843d4199443052e8deb0c9c3b7e4b580
[1]: https://valkyrie.yoctoproject.org/#/builders/87/builds/17/steps/40/logs/stdio
(From OE-Core rev: e89bbc00ba16574d719b199c01ffbf37646f4f54)
Signed-off-by: Yoann Congal <yoann.congal@smile.fr>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
