Re: [PATCH] rootfs: Restore previous naming scheme in INITRD_DEPLOY_FILE

["'Jan Kiszka' via isar-users" <isar-users@googlegroups.com>]

On 12.12.25 15:36, Zhihang Wei wrote:
> 
> 
> On 12/12/25 11:29, Jan Kiszka wrote:
>> On 11.12.25 17:39, 'Jan Kiszka' via isar-users wrote:
>>> On 11.12.25 17:09, Zhihang Wei wrote:
>>>>
>>>> On 12/11/25 14:53, 'Jan Kiszka' via isar-users wrote:
>>>>> On 09.12.25 11:15, Zhihang Wei wrote:
>>>>>> It cause a wired bug in full CI, when building the target
>>>>>> 'mc:qemuamd64-iso-bookworm:isar-image-ci' under the test case:
>>>>>> - citest.py:NoCrossTest.test_nocross
>>>>>> Error log from avocado:
>>>>>> ERROR: mc:qemuamd64-iso-bookworm:isar-image-ci-1.0-r0 do_image_wic:
>>>>>> ExecutionError('/work/build/tmp/work/debian-bookworm-amd64/isar-
>>>>>> image-
>>>>>> ci-qemuamd64-iso/1.0-r0/temp/run.do_image_wic.231803', 1, None, None)
>>>>>> ERROR: Logfile of failure stored in: /work/build/tmp/work/debian-
>>>>>> bookworm-amd64/isar-image-ci-qemuamd64-iso/1.0-r0/temp/
>>>>>> log.do_image_wic.231803
>>>>>> ERROR: Task (mc:qemuamd64-iso-bookworm:/work/meta-test/recipes-core/
>>>>>> images/isar-image-ci.bb:do_image_wic) failed with exit code '1'
>>>>>> Error log from the failed task:
>>>>>> DEBUG: Executing python function set_image_size
>>>>>> DEBUG: Python function set_image_size finished
>>>>>> DEBUG: Executing shell function do_image_wic
>>>>>> Creating /etc/schroot/chroot.d/isar-builder-50e1b98f-a270-4654-bbe8-
>>>>>> a40664e9bf72-231803
>>>>>> Started session: isar-builder-50e1b98f-a270-4654-bbe8-
>>>>>> a40664e9bf72-231803-f762d345-e215-4bb7-98ba-63ad30c6c51c
>>>>>> <snip>
>>>>>> Setting up grub-common (2.06-13+deb12u1) ...
>>>>>> Setting up grub-efi-amd64-bin (2.06-13+deb12u1) ...
>>>>>> Processing triggers for libc-bin (2.36-9+deb12u13) ...
>>>>>> Processing triggers for man-db (2.11.2-2) ...
>>>>>> INFO: Creating image(s)...
>>>>>> ERROR: _exec_cmd: cd /tmp/isar-image-ci-debian-bookworm-qemuamd64-
>>>>>> iso.wic/tmp.wic.0y6e0rtx/INITRD && find . | cpio -o -H newc -R
>>>>>> root:root
>>>>>>> /tmp/isar-image-ci-debian-bookworm-qemuamd64-iso.wic/
>>>>>>> tmp.wic.0y6e0rtx/
>>>>>> initrd.cpio returned '127' instead of 0
>>>>>> output: /bin/sh: 1: cpio: not found
>>>>>> To redo the test using avocado:
>>>>>> 1. Have a clean clone of isar, checkout to branch next and apply your
>>>>>> patches:
>>>>>> $ git clone -b next https://github.com/ilbers/isar.git
>>>>>> $ cd isar
>>>>>> $ git am /path-to/0001-my-contribution-to-isar.patch
>>>>>> 2. Disable unrelated targets to make error appears faster, by
>>>>>> applying
>>>>>> the following diff to "testsuite/citest.py":
>>>>>> diff --git a/testsuite/citest.py b/testsuite/citest.py
>>>>>> index 27cc9ff6..8dd40d62 100755
>>>>>> --- a/testsuite/citest.py
>>>>>> +++ b/testsuite/citest.py
>>>>>> @@ -494,36 +494,7 @@ class NoCrossTest(CIBaseTest):
>>>>>> def test_nocross(self):
>>>>>> targets = [
>>>>>> - 'mc:qemuarm-buster:isar-image-ci',
>>>>>> - 'mc:qemuarm-bullseye:isar-image-base',
>>>>>> - 'mc:qemuarm64-bullseye:isar-image-base',
>>>>>> - 'mc:qemuarm64-bookworm:isar-image-ci',
>>>>>> - 'mc:qemui386-buster:isar-image-base',
>>>>>> - 'mc:qemui386-bullseye:isar-image-base',
>>>>>> - 'mc:qemuamd64-buster:isar-image-ci',
>>>>>> - 'mc:qemuamd64-bullseye:isar-initramfs',
>>>>>> - 'mc:qemumipsel-bullseye:isar-image-base',
>>>>>> - 'mc:imx6-sabrelite-bullseye:isar-image-base',
>>>>>> - 'mc:phyboard-mira-bullseye:isar-image-base',
>>>>>> - 'mc:hikey-bullseye:isar-image-base',
>>>>>> - 'mc:virtualbox-bullseye:isar-image-base',
>>>>>> - 'mc:virtualbox-bookworm:isar-image-base',
>>>>>> - 'mc:bananapi-bullseye:isar-image-base',
>>>>>> - 'mc:bananapi-bookworm:isar-image-base',
>>>>>> - 'mc:nanopi-neo-bullseye:isar-image-base',
>>>>>> - 'mc:nanopi-neo-bookworm:isar-image-base',
>>>>>> - 'mc:qemuamd64-focal:isar-image-ci',
>>>>>> - 'mc:qemuamd64-bookworm:isar-image-ci',
>>>>>> 'mc:qemuamd64-iso-bookworm:isar-image-ci',
>>>>>> - 'mc:qemui386-bookworm:isar-image-base',
>>>>>> - 'mc:qemumipsel-bookworm:isar-image-ci',
>>>>>> - 'mc:hikey-bookworm:isar-image-base',
>>>>>> - 'mc:beagleplay-bookworm:isar-image-base',
>>>>>> - 'mc:qemuarm64-noble:isar-image-base',
>>>>>> - 'mc:qemuamd64-noble:isar-image-base',
>>>>>> - 'mc:qemuamd64-jammy:isar-image-base',
>>>>>> - 'mc:qemuarm64-jammy:isar-image-base',
>>>>>> - 'mc:x86-pc-bookworm:isar-image-base',
>>>>>> ]
>>>>>> self.init()
>>>>>> 3.Run kas shell, setup CI prerequisites (avocado, qemu) and cleanup:
>>>>>> $ ./kas/kas-container shell kas/isar.yaml --command \
>>>>>> "rm -rf /work/build/conf && /work/scripts/ci_setup.sh"
>>>>>> 4.Run the failed test in fast:
>>>>>> $ cd /work/testsuite
>>>>>> $ avocado run citest.py:NoCrossTest.test_nocross$
>>>>>> Zhihang
>>>>>> On 11/29/25 16:13, Jan Kiszka wrote:
>>>>>>> From: Jan Kiszka <jan.kiszka@siemens.com>
>>>>>>>
>>>>>>> Might have been accidentally changed, specifically as
>>>>>>> initramfs.bbclass
>>>>>>> was not updated as well, and that broke downstream users of
>>>>>>> INITRD_DEPLOY_FILE via wic.
>>>>>>>
>>>>>>> Fixes: c3c4e72cbfc4 ("image: introduce IMAGE_INITRD, deprecate
>>>>>>> INITRD_IMAGE")
>>>>>>> Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
>>>>>>> ---
>>>>>>>     meta/classes/image.bbclass | 2 +-
>>>>>>>     1 file changed, 1 insertion(+), 1 deletion(-)
>>>>>>>
>>>>>>> diff --git a/meta/classes/image.bbclass b/meta/classes/image.bbclass
>>>>>>> index b030024f..d367d8c7 100644
>>>>>>> --- a/meta/classes/image.bbclass
>>>>>>> +++ b/meta/classes/image.bbclass
>>>>>>> @@ -29,7 +29,7 @@ INITRD_IMAGE ?= ""
>>>>>>>     IMAGE_INITRD ?= ""
>>>>>>>       # Name of the deployed initrd image
>>>>>>> -INITRD_DEPLOY_FILE = "${@ d.getVar('IMAGE_INITRD') or '${PN}' }-
>>>>>>> ${DISTRO}-${MACHINE}-initrd.img"
>>>>>>> +INITRD_DEPLOY_FILE = "${@ d.getVar('IMAGE_INITRD') or '${PN}' }-
>>>>>>> ${DISTRO}-${MACHINE}.initrd.img"
>>>>>>>       # Make sure dependent initramfs recipe is built
>>>>>>>     do_image[depends] += "${@ '${IMAGE_INITRD}:do_build' if
>>>>>>> '${IMAGE_INITRD}' else '' }"
>>>>> Did you understand that issue by now? The patch itself is still a
>>>>> must-have to fix initrd generation in downstream layers.
>>>>>
>>>>> Jan
>>>>>
>>>> No. can you re-create the same issue on your side?
>>>>
>>> Let me try that tomorrow, wanted to "play" with the testsuite once again
>>> anyway.
>>>
>> Still no fun (slow startup of the build env, need to patch the setup to
>> run the test etc. etc.). Quickly translating the testsuite target into a
>> proper kas configuration is way more enjoyable:
>>
>> kas-container build kas/isar.yaml:kas/machine/qemuamd64-iso.yaml:kas/
>> distro/debian-bookworm.yaml:kas/image/isar-image-ci.yaml
>>
>> At this chance I noticed that isar's kas configuration is missing all
>> the multiconfig targets you have, thus can't be fed with mc: targets
>> directly. Needs fixing!
>>
>>
>> Now to the actual bug: As expected, it was most likely a pre-existing
>> one. I do not get yet why it was working before c3c4e72cbfc4, but maybe
>> we are not entering a questionable code path in meta/scripts/lib/wic/
>> plugins/source/isoimage-isohybrid-isar.py. Fixing that, we get things
>> working again:
>>
>> diff --git a/meta/scripts/lib/wic/plugins/source/isoimage-isohybrid-
>> isar.py b/meta/scripts/lib/wic/plugins/source/isoimage-isohybrid-isar.py
>> index fb6c6b39..7e981a78 100644
>> --- a/meta/scripts/lib/wic/plugins/source/isoimage-isohybrid-isar.py
>> +++ b/meta/scripts/lib/wic/plugins/source/isoimage-isohybrid-isar.py
>> @@ -194,7 +194,7 @@ class IsoImagePlugin(SourcePlugin):
>>               machine = os.path.basename(initrd_dir)
>>                 pattern = '%s/%s*%s.%s' % (initrd_dir, image_name,
>> machine, image_type)
>> -            pattern = '%s/%s-%s-initrd.%s' % (initrd_dir, image_name,
>> machine, image_type)
>> +            pattern = '%s/%s-%s.initrd.%s' % (initrd_dir, image_name,
>> machine, image_type)
>>               files = glob.glob(pattern)
>>               if files:
>>                   initrd = files[0]
>>
>> But that raises the question why we are duplicating INITRD_DEPLOY_FILE
>> here, rather than using it. Maybe it was used before via setting the
>> INITRD bitbar var, I still need to understand that. In any case, this
>> pattern duplication in the wic plugin should be avoided, possibly even
>> the whole code path. Felix, you introduced that.
>>
>> Jan
>>
> Just FYI, I'll be on vacation since 13.12.2025 until 11.01.2026. Anton
> will take the maintenance work during this time. Please add Anton to the
> relevant email threads if it's needed. I'll be back on 12.01.2025.
> 

Sure - enjoy the time off!

Jan

-- 
Siemens AG, Foundational Technologies
Linux Expert Center

-- 
You received this message because you are subscribed to the Google Groups "isar-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to isar-users+unsubscribe@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/isar-users/346185f5-b41d-4f13-80f5-02855b380b36%40siemens.com.