On 2/17/2026 10:31 PM, Mathieu Dubois-Briand wrote:
CAUTION: This email comes from a non Wind River email account!
Do not click links or open attachments unless you recognize the sender and know
the content is safe.
On Mon Feb 16, 2026 at 4:48 PM CET, Mathieu Dubois-Briand wrote:
On Mon Feb 16, 2026 at 2:09 PM CET, Harish Sadineni wrote:
Hi Yoann, Mathieu,
I tried few builds locally, but not able to reproduce this.
On the autobuilder, We ran following builds which were completed
successfully on ubuntu2504-vk-arm1:
https://autobuilder.yoctoproject.org/valkyrie/#/builders/23/builds/3326
https://autobuilder.yoctoproject.org/valkyrie/#/builders/23/builds/3331
Additionally, One of Mathieu's build also ran successfully on
oe-selftest-armhost ubuntu2504-vk-arm1:
https://autobuilder.yoctoproject.org/valkyrie/#/builders/23/builds/3323
<https://autobuilder.yoctoproject.org/valkyrie/#/builders/23/builds/3323>
From what I can see, build 3319
<https://autobuilder.yoctoproject.org/valkyrie/#/builders/23/builds/3319>was
triggered from a-full/3219
<https://autobuilder.yoctoproject.org/valkyrie/#/builders/29/builds/3219>:
However, when checking a-full/3219, it triggered build 3317
<https://autobuilder.yoctoproject.org/valkyrie/#/builders/23/builds/3317>,
where the rust kernel selftest passed.
I can see few other builds were successful on oe-selftest-armhost with
this patch series. I’m not sure why the failure occurred in this case.
Mathieu, could you please try running a few additional builds to see if
it reproduces?
Thanks,
Harish
Sure, I will launch a few builds and see how it goes.
AB maintenance will begin soon, so I will start either later today or
tomorrow morning.
I believe I got one:
https://autobuilder.yoctoproject.org/valkyrie/#/builders/23/builds/3346
I backed up the build folder, it should stay there until next Monday.
(/srv/pokybuild/yocto-worker/oe-selftest-armhost-3346-mathieu-2026-02-17
on ubuntu2404-vk-arm1)
Hi Mathieu,
I reviewed the build logs and observed that the generated rootfs does
not contain the lib/modules directory,
which should normally be created when kernel modules are installed into
the image.
Unfortunately, I was not able to find sufficient information in the logs
to explain why this is happening.
I attempted to reproduce the issue multiple times locally on an ARM host
and x86_64 host, but i am not able to reproduce the failure.
So far this intermittent failure seen only on arm host. So, we are
planning to run a few more tests on ab with some debug info added to the
test case.
If the issue still cannot be reproduced, we will temporarily skip this
test on arm64 for this release.
Will keep you posted by beginning of next week.
Thanks,
Harish
Please tell me if you need me to extract any build artifact.
Thanks,
Mathieu
--
Mathieu Dubois-Briand, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#232098):
https://lists.openembedded.org/g/openembedded-core/message/232098
Mute This Topic: https://lists.openembedded.org/mt/117779196/21656
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub
[[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-
