config: pipeline: Update bookworm-rt rootfs

[musamaanjum]

New package has been added to bookworm-rt rootfs. Hence update the URL to newly updated rootfs.

Related PR: kernelci/kernelci-core#2614

[musamaanjum]

I've set the rootfs URL to staging. After verification I'll put URL of production before merging on this PR.

[a-wai]

I've set the rootfs URL to staging. After verification I'll put URL of production before merging on this PR.

Sounds good, can you please set this PR as draft so it can't get merged as-is by accident?

[nuclearcat]

LAVA job referenced in link is from production, which is not updated yet

[musamaanjum]

The test failed in reality as rtla isn't found. The test result reporting is wrong. Let's debug why failure happend:

kernelmsg (rt-tests-rtla-osnoise)
KernelCI node: [66bf60458ac59a075e697815](https://staging.kernelci.org:9000/viewer?node_id=66bf60458ac59a075e697815)
Result: PASS
Date: 2024-08-16T14:20:53.251000
Tree: stable-rt
Branch: v6.1-rt
Kernel version: v6.1.105-rt38
Commit: 8fe1d58eff313241a0c3187db014fbccaad0b680 (https://git.kernel.org/pub/scm/linux/kernel/git/rt/linux-stable-rt.git)
Arch : arm
Platform : imx6q-sabrelite
Config: multi_v7_defconfig+preempt_rt
Compiler: gcc-12
Runtime: lava-collabora
Logs:
[lava_log](https://kciapistagingstorage1.file.core.windows.net/staging/rt-tests-rtla-osnoise-66bf5f908ac59a075e697801/log.txt.gz?sv=2022-11-02&ss=f&srt=sco&sp=r&se=2024-10-17T19:19:12Z&st=2023-10-17T11:19:12Z&spr=https&sig=sLmFlvZHXRrZsSGubsDUIvTiv%2BtzgDq6vALfkrtWnv8%3D)

From above node link, I've found the grand parent node (https://staging.kernelci.org:9000/viewer?node_id=66bf5f908ac59a075e697801) to find lava log.
https://kciapistagingstorage1.file.core.windows.net/staging/rt-tests-rtla-osnoise-66bf5f908ac59a075e697801/lava_callback.json.gz?sv=2022-11-02&ss=f&srt=sco&sp=r&se=2024-10-17T19:19:12Z&st=2023-10-17T11:19:12Z&spr=https&sig=sLmFlvZHXRrZsSGubsDUIvTiv%2BtzgDq6vALfkrtWnv8%3D

This log shows that older and wrong kselftest rootfs is being used:

name: rt-tests-rtla-osnoise
storage.kernelci.org/images/rootfs/debian/bookworm-kselftest/20240221.0

Why the bookworm-rt isn't being used here? @nuclearcat @pawiecz please can you help debug this?

[pawiecz]

@musamaanjum It looks like only depthcharge boot method currently supports nfsroot substitution. Boot methods for other bootloaders have hardcoded images in their templates, in this case u-boot template needs to be updated.

[musamaanjum]

@musamaanjum It looks like only depthcharge boot method currently supports nfsroot substitution. Boot methods for other bootloaders have hardcoded images in their templates, in this case u-boot template needs to be updated.

Many thanks @pawiecz. I was scratching my head over it from some time.

@nuclearcat @a-wai can we schedule the work to update the templates? You may have pre-planned it already. We should prioritize fixing rt-tests work as the work started on the rt maintainer's request and we don't want to lose his interest. (cc: @helen-fornazier as she is leading the community engagement work and can help in deciding the priority of this work.)

[a-wai]

@musamaanjum this was trivial enough that I pushed kernelci/kernelci-core#2641

[a-wai]

Setting staging-skip to allow testing #766

[pawiecz]

#766 has been merged, dropping staging-skip label to bring this PR back to staging

[a-wai]

Note: with #766 merged it now conflicts, so this also needs a rebase to be deployed.

[musamaanjum]

@a-wai Thanks for fixing. Rebased the PR.

[pawiecz]

LGTM

Most recently triggered Maestro Jobs (bottom) showed that the LAVA TestJob definition might still need some adjustments but the rootfs now contains necessary tools which should be enough to merge this PR as well as kernelci/kernelci-core#2614

[musamaanjum]

I've collected the results on kernelci/kernelci-project#439 and fix issues under this ticket. Overall, 3/11 tests have issues in their executor script in https://github.com/Linaro/test-definitions.