Age | Commit message (Collapse) | Author | Files | Lines |
|
On 2020.02, qemu jobs fail to mount the rootfs.
This is due to LAVA ignore now compression for ramdisk and the AGL image
is in xz.
But AGL qemu jobs in fact does not use a ramdisk but a virtio disk.
Renaming the ramdisk definition entry to rootvd handle this issue.
Furthermore, let's introduce a new rootfs type rootvd.
Change-Id: I71ef6dee0d859c6aa0b29f1d47ac14dd4db900d8
Bug-AGL: SPEC-3210
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
With recent qemu, the ARM virt machine hit:
pci-host-generic 4010000000.pcie: can't claim ECAM area
[mem 0x10000000-0x1fffffff]: address conflict with pcie@10000000 [mem 0x10000000-0x3efeffff]
pci-host-generic: probe of 4010000000.pcie failed with error -16
Without the PCI working, the storage for tests cannot work.
So let's fix to the last version of the vmachine which work without
kernel hacking.
This patch should not be merged until LAVA 2020.02 since stretch qemu does not support virt-2.11
Change-Id: Ifbb352a491cf917c45d48e0f88af20c146ca5422
Bug-AGL: SPEC-3210
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
Specifying the kernel type in "actions: boot:" is deprecated.
It should now be set in "deploy: kernel:".
So this patch set unconditonaly the type: in the deploy section (before
it was only for ramdisk method) and remove the type: from the deprecated
place.
But due to a bug in LAVA, NBD jobs rely on type: to be in the deprecated
place until 2020.02 (2020.02 has the fix).
So this patch must be applied only after LAVA 2020.02 was deployed.
Change-Id: Ida30092d2abe14b05d5e37f651f91ab33876e4f5
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
{# ... #} is the proper comment style, otherwise we still assign.
Signed-off-by: Jan-Simon Moeller <jsmoeller@linuxfoundation.org>
Change-Id: I5b3f318d758825c30e6ec208d2eeaab51ba211ea
|
|
Use the plain h3ulcb dtb until the BSP for Kingfisher is back.
Signed-off-by: Jan-Simon Moeller <jsmoeller@linuxfoundation.org>
Change-Id: Ieedf385a28ab3371c6d0f9d5fbd3d92999db90e3
|
|
After upgraded our lab to 2020.02 and so upgraded to buster, the qemu
version has increased.
With this new version, the qemu x86 jobs fail with a kernel panic about
IO APIC timer.
This is a know issue and the simpliest fix is to add no_timer_check to
the kernel cmdline.
Change-Id: I75c2427e683690bfe961f482ad5267d1d957c36f
Bug-AGL: SPEC-3210
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
Change-Id: Ia6050d0ec9d35cae442dc0979b4ea75f95e6e70b
Signed-off-by: Jan-Simon Möller <jsmoeller@linuxfoundation.org>
|
|
CI fail with:
UnboundLocalError: local variable 'machine_branch' referenced before assignment
(See https://build.automotivelinux.org/job/ci-platform-meta-agl-verify-CIBT-qemuarm/568/console)
This patch add an initial declaration for it
Change-Id: I7fa7b836e1546f65cf6678c8527698208fc5581a
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
Signed-off-by: Jan-Simon Moeller <jsmoeller@linuxfoundation.org>
Change-Id: I9d6f5cb51501b1cdfd9a643c28af96c48323704f
|
|
The current job workload can take more than 15 minutes to boot on e.g. qemuarm
during firstboot.
Signed-off-by: Jan-Simon Moeller <jsmoeller@linuxfoundation.org>
Change-Id: If913db0a5984b1d4403cfb759797809a860bac6f
|
|
The raspberrypi4 should use booti for booting Image.
Bug-AGL: SPEC-3004
Change-Id: Ie241194eb4177f004b2f5c378a63db88e3293669
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
By default LAVA set the test device on IDE.
Stretch version of qemu doesnt care about this setting and was always
using virtio.
But buster qemu care now about this and fail with:
machine type does not support if=ide,bus=0,unit=0
This patch adds the necessary to restore the use of virtio.
Change-Id: If34194d82b6104b9b5be7118a5194bd3b61fc090
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
This is needed if we do not let systemd remount it.
Signed-off-by: Jan-Simon Moeller <jsmoeller@linuxfoundation.org>
Change-Id: I6f3ddbbc45844bff2a4adb423c7ecb8b57c0c0fd
|
|
to avoid systemd fstab generator to mount
partitions that do not exist in the netboot case.
Signed-off-by: Jan-Simon Moeller <jsmoeller@linuxfoundation.org>
Change-Id: I4c08bd50690062202590d7aae8f6508798f29997
|
|
This patch add the template for raspberrypi4.
Like rpi3, the job need some custom parameters like console_device and
extra_kernel_args.
Unlike rpi3, we use directly extra_kernel_args as the device_type will
permit to override it.
Bug-AGL: SPEC-3004
Change-Id: Iac88f04e0954f77bc2e3a3daf0baf43aec3fa7c7
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
Asking dtb for upsquare lead to utils/job-prereq.py crash since x86
machines does not have dtb.
This patch made upsquare to use FILE_MAP_X86
Bug-AGL: SPEC-3132
Change-Id: I4d33b61719c0d5e3afd83e5d8bb8dc468e26e259
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
This reverts commit b0eeec805136c097873dac29a1dfdbc5de545e98.
Reason for revert: This patch should be reverted because the changes related to this patch not accepted yet in KernleCI backend, so there will be no job accepted in the backend with this new metadata field.
Change-Id: If6045e437275d7d10ae354f39aeb2e2caaec9f47
Signed-off-by: Khouloud Touil <ktouil@baylibre.com>
|
|
It's a field that is going to be used later in the email report
to provied the link of each faild test case of a test job.
This metadata will provid the link of the lava web interface.
AGL-bug: SPEC-3020
Signed-off-by: Khouloud Touil <ktouil@baylibre.com>
Change-Id: If4d90e0c992dc3c07972cdb053f2ef25e934450f
|
|
The plan is to have: H3 with KF, M3 without KF.
So let's revert to the right DTB for single M3.
If we need later M3+KF the right releng templates are already present
for it.
This reverts commit 8a349e865441c277498a008ac4fd9f13893b5527.
Change-Id: I2d2bcf0f3658804d98e06ca7f094aff8efaf2adf
Bug-AGL: SPEC-2921
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
The prompt for h3ulcb-kf is h3ulcb, so yocto_machine need to be set to
it.
Bug-AGL: SPEC-2921
Change-Id: I73b9606b7c9de5142619f531d6cced402763bb85
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
The build name for h3ulcb-kf is h3ulcb-nogfx.
This patch adds the same hack than for m3ulcb.
Bug-AGL: SPEC-2921
Change-Id: Ib687449e911ec0533cac811b648dd02cc8ae286f
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
The initrd path for h3ulcb-kf is the same as h3ulcb.
Bug-AGL: SPEC-2921
Change-Id: Ieb6812bcb121d2334c94027e9049c45fdb536733
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
This patch adds M3ULCB+KF and H3ULCB+KF support to releng-scripts
Change-Id: I97ef1d63f330f4aaf669ad33f770d4527c00a271
Bug-AGL: SPEC-2921
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
Since AGL use only M3 with KingFisher, let's use the right DTB.
Change-Id: I89bfe021b13f69f786d091b630f44f77ed20a3a0
Bug-AGL: SPEC-2921
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
Split the can tests in two, one for basic tests that doesn't need the
two can interfaces to be related, and the other one is for a board with
two related interfaces.
Added a tag for Baylibre board "can_if_hw_connection" to notice the difference.
AGL-bug: SPEC-2878
Signed-off-by: Khouloud Touil <ktouil@baylibre.com>
Change-Id: I1350098709de05ba50faf719bbb13f9a59f7eac0
|
|
To fix the xmlsec ptest in
https://gerrit.automotivelinux.org/gerrit/#/c/AGL/releng-scripts/+/22510/
we added the virtio-rng-pci to the qemuarm and qemuarm64 targets.
Do this also to the x86-64 target.
Change-Id: I3ab0a08371656137621120d7eb5356ba993c8b2b
Bug-AGL: SPEC-2739
Signed-off-by: Jan-Simon Möller <jsmoeller@linuxfoundation.org>
|
|
ptest-runner fail on xmlsec1 on ARM/ARM64.
This is due to lack of randomness as the first test in doing a RSA
signature.
This patch adds a virtio-rng-pci device on qemuarm and qemuarm64 which
adds enough randomness to permit this test to pass.
Bug-AGL: SPEC-2739
Change-Id: I98dddd421572cecf2027793aad62a12853e07c59
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
build-tags: is not a valid LAVA keyword/section so is causing LAVA
jobs to be refused. Only "tags:" is a valid keyword/section.
However, build-tags are not needed for selecting a correct LAVA
device. This is what 'device-tags' are for. The build-tags are only
propogated to the test-scripts.
So, to fix, just remove the entire build-tags section.
Bug-AGL: SPEC-2751
Change-Id: Ic24bf9c0caf1f4bc2dacd8065eb0edbace7937b1
Signed-off-by: Kevin Hilman <khilman@baylibre.com>
|
|
Add new argument --build-tags to be used in the job template and to
be passed to the tests to be run along with using --device-tags.
Bug-AGL: SPEC-2721
Signed-off-by: Khouloud Touil <ktouil@baylibre.com>
Change-Id: I2fb4c3d0cb8cd1d0e3c36c1df55185cb4dde43c3
|
|
This silences a lot of useless systemd messages.
Change-Id: Icacd3bb09125ff4a16b57764490e570a1fc9258b
Signed-off-by: Jan-Simon Möller <jsmoeller@linuxfoundation.org>
|
|
On my gentoo, using yaml.load now give:
Traceback (most recent call last):
File "./utils/job-prereq.py", line 81, in <module>
main()
File "./utils/job-prereq.py", line 66, in main
job_yaml = yaml.load(job)
File "/usr/lib64/python2.7/site-packages/yaml/__init__.py", line 109, in load
raise RuntimeError("Unsafe load() call disabled by Gentoo. See bug #659348")
RuntimeError: Unsafe load() call disabled by Gentoo. See bug #659348
Note that on recent ubuntu, a warning appears also. (like on AGL jenkins)
/w/workspace/ci-platform-meta-agl-demo-verify-CIBT-qemuarm64/repoclone/releng-scripts/utils/job-prereq.py:66: YAMLLoadWarning: calling yaml.load() without Loader=... is deprecated, as the default Loader is unsafe. Please read https://msg.pyyaml.org/load for full details.
job_yaml = yaml.load(job)
This is due to a security risk of using yaml.load()
Since releng-script does not rely on any behavour provided by load(), let's convert the call to safe_load().
This will also clean/reduce the output of jenkins logs
Change-Id: I6158dcc21c0f7ec4645fb7b16acfaf1a6963cb26
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
This patch take care of the snapshot/daily/weekly build-types if they
are requested without mentioning the branch and version args.
Bug-AGL: SPEC-2471
Change-Id: I86e5465162cc20be82b33feedcbed7f78300053d
Signed-off-by: Khouloud Touil <ktouil@baylibre.com>
|
|
This patch adds a template for qemuarm
Bug-AGL: SPEC-2376
Change-Id: Ib0b9f44d96fc7f365f7da17ce3ce3461d0a3c6c0
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
When adding support for upsquare the URL generation for m3ulcb was broken.
This patchs fix the generation by adding a new variable for the build name.
Fixes: 32b88dfd726b ("SPEC-2075: fix URL generation for upsquare non-CI build")
Bug-AGL: SPEC-2470
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
Change-Id: I84bc2d8bbcbc49b048bbd95f84a531bde9ac6aba
|
|
This patchs permits utils/job-prereq.py to work with qemuarm/qemuarm64
Bug-AGL: SPEC-2376
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
Change-Id: Ie8615a9ab4d1f0053afc098db508814e6aad1bf7
|
|
This mitigates slower downloads.
Change-Id: I4e18f133097e8a3c32889fcc30a076ad24b882f9
Signed-off-by: build.automotivelinux.org <jenkins@automotivelinux.org>
|
|
Whenever --device-tags are passed, a "tags:" section is added to the
job with all of the tags passed on the command-line.
To facilitate this, the device_tags list is passed directly from
python to the jinja templates, and iteration is done in jinja
instead of a string created in python.
Also, while here, ensure that tags are always passed lower-cased.
IMPORTANT: this implies that any tags in LAVA labs that are upper-case
will *never* be used.
Change-Id: I8f274098a69e34c32c5ed2d1c430eca0d364bcc5
Signed-off-by: Kevin Hilman <khilman@baylibre.com>
|
|
The kernelci-backend is now more strict about enforcing valid and
required keys in the metadata. The field `vcs.url` was already a
duplicate of `git.url` but now `vcs.url` is no longer considered
valid, and using it causes callbacks to be refused by the backend.
Fix it by using git.url instead.
Change-Id: I768229f3fabc8d127f68f4c33bd6ae22c1f654c4
Signed-off-by: Kevin Hilman <khilman@baylibre.com>
|
|
This patch adds a template for qemuarm64
Bug-AGL: SPEC-2376
Change-Id: I2f8a5fbd1625df8fcb03ba4798c2d8329889408b
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
LAVA expect machine, memory and cpu to be passed as argument in context and not
via extra args.
This patchs introduce qemu_cpu, qemu_memory and qemu_machine as jinja
variables.
Change-Id: I8893943947a046f90be388febccfc29c7923d5b4
Bug-AGL: SPEC-2376
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
See the qemu case.
Bug-AGL: SPEC-2449
Signed-off-by: Jan-Simon Möller <jsmoeller@linuxfoundation.org>
Change-Id: Ib23bd4897e89d3eca1eaba33a37c256c932e2d36
|
|
When adding support for upsquare, I have added a new machine_frag_url
which contains the machine to use in url.
It works for CI build type, but for other I miss to convert one machine
to machine_frag_url
Bug-AGL: SPEC-2075
Change-Id: Ieaaf71b62425cf73fb9582b84666640c4e9d6cae
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
I have assumed that only master is used for appfw but it is false.
Some app review are done against non-master branch.
This patch adds a way to change the branch used via --app-branch.
Bug-AGL: SPEC-1850
Change-Id: I10c8ad0e5a08158f9c2295692e4b3c266b97b14e
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
We should not include application-lifecycle tests based on APPURL, but
on the presence of the --test application-lifecycle parameter.
Bug-AGL: SPEC-1850
Change-Id: Idda5d9011f54cd7eb1710e15254ab358ef7b99b1
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
The generated APPURL was invalid:
- it needed to be always using ci
- it missed branch and arches path
This patch move the generation of the application url in the jinja file
for accessing the new sdk_arch
Bug-AGL: SPEC-1850
Change-Id: Ie3ffff5f6181fdc89c3ebbd41078e4d60a06992b
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
As asked by JSM on review 21232, this patch removes the porter machine.
Bug-AGL: SPEC-2375
Change-Id: Iaa8083a271bb61ee1f9208f671d617fc15a00028
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
Change-Id: Ieadfc20e9abbc01a272faf16f59d88ff0f1a3ff5
Signed-off-by: Jan-Simon Möller <jsmoeller@linuxfoundation.org>
|
|
This patch adds suppport for the upsquare.
Upsquare use a GRUB boot method, so we needed to add a new
generic-grub-tftp boot template.
Since all x86_64 use the same yocto generated build name, we add a new
upsquare machine, but tweaked a bit utils/agljobtemplate.py to keep
intel-corei7-64 everywhere (except for the LAVA device-type).
Bug-AGL: SPEC-2075
Change-Id: I15c850fada53efbcda4fd58527d7b7414865f89d
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
This patch takes the branch and machine arguments and decide which is
the dtb name to be used.
Change-Id: I09bd67de51920c81780a77be396906a15d8247f5
Signed-off-by: Khouloud Touil <ktouil@baylibre.com>
|
|
In order to take care with the old dtb names and to keep it as flexible
as possible we create a file for each board and branch with the name
that matches with each branch.
Change-Id: I223588aa794e4af6428b2ddcd13a1f8d396152a8
Signed-off-by: Khouloud Touil <ktouil@baylibre.com>
|