From: Anton Mikanovich <amikan@ilbers.de>
To: Badrikesh Prusty <badrikesh.prusty%siemens.com@gtempaccount.com>,
isar-users <isar-users@googlegroups.com>
Cc: Baurzhan Ismagulov <ibr@ilbers.de>,
"Kiszka, Jan" <jan.kiszka@siemens.com>,
Arulpandiyan Vadivel <arulcse2008@gmail.com>
Subject: Re: [PATCH] image: copy DTB_FILES to DEPLOYDIR/IMAGE_FULLNAME/KERNEL_NAME/
Date: Wed, 8 Oct 2025 12:45:50 +0300 [thread overview]
Message-ID: <1c93851a-7b56-4088-bd19-83ed1b1d8ecc@ilbers.de> (raw)
In-Reply-To: <1a2a38b0-49aa-4167-a95e-16f4c5ba385dn@googlegroups.com>
08/10/2025 09:20, 'Badrikesh Prusty' via isar-users wrote:
> Hello all,
>
> Apologies for the delayed response. I somehow missed the message.
>
> > This sounds familiar, and I think we discussed that multiple times
> > already. Did you study previous threads/patches, e.g.
> > https://patchwork.isar-build.org/project/isar/list/?series=1241?
> >
> > Yes, there were few patches that tried to solve the issue.
> >
> > Using subdirectory (variant similar to TS's)
> > - https://patchwork.isar-build.org/project/isar/list/?series=1149
> > Allows to overlap files
> > - https://patchwork.isar-build.org/project/isar/list/?series=1235
> > Make kernel deploy the files, instead of image recipes
> > - https://patchwork.isar-build.org/project/isar/list/?series=1241
> >
>
> I reviewed the previously shared patch, but not in full — thanks for
> sharing it here.
>
> The patch I sent is very similar to @nicusor
> patch: https://patchwork.isar-build.org/project/isar/list/?series=1149
>
> I just reviewed the other patches and believe that many may encounter
> the same issue. For instance, when attempting to build multiple
> kernels for a single image—such as `linux-image-amd64` and
> `linux-image-amd64-rt` kernels—this overwrite problem is likely to occur.
>
> This patch also resolves the issue with image builds that include
> multiple kernels by storing the DTB files in the directory
> DEPLOYDIR/IMAGE_FULLNAME/KERNEL_NAME. This will also fix issues when
> building with different DISTROs as IMAGE_FULLNAME includes target
> DISTRO. Additionally, it fixes problems that arise when building with
> different DISTROs, since IMAGE_FULLNAME includes the target DISTRO.
>
> > This would affect downstream as well. You are missing a
> > RECIPE-API-CHANGELOG.md entry.
> >
>
> Yes, it will affect downstream layers. E.g. I checked with the
> isar-cip-core repo, changes will be required in `efibootguard-boot`
> wic plugin. I will work on fixing it, once this is approved.
>
> Thanks and Regards,
> Badrikesh
Hello Badrikesh,
Files like scripts/lib/wic/plugins/source/bootimg-efi.py are borrowed
from the
upstream and supposed to have as minimal changes as possible. Even having
changes in custom WIC plugins can results in additional side-effects.
Also, in general, DTBs should be part of a kernel, but not an image.
I've just released the new version of DTB deployment fix proposal which
I hope
will cover all the cases. Please have a look:
https://patchwork.isar-build.org/project/isar/cover/20251008085507.1017500-1-amikan@ilbers.de/
There are also some observations regarding other implementations of the same
fix were shared on the list previously:
https://patchwork.isar-build.org/project/isar/patch/20240416100746.5681-1-nicusor_huhulea@mentor.com/
- We don't want to modify WIC plugins if solving outside of them is
possible.
https://patchwork.isar-build.org/project/isar/patch/20240705071856.904-1-cedric.hombourger@siemens.com/
- Manipulating of sstate sounds tricky and errors might be costly.
- Allowing overlapping will cause deploy file loosing on multiple target
setup.
--
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/1c93851a-7b56-4088-bd19-83ed1b1d8ecc%40ilbers.de.
next prev parent reply other threads:[~2025-10-08 9:46 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-04-03 17:07 'Badrikesh Prusty' via isar-users
2025-04-04 9:55 ` 'Jan Kiszka' via isar-users
2025-04-04 14:31 ` Uladzimir Bely
2025-09-04 9:36 ` Arulpandiyan Vadivel
2025-09-04 14:11 ` 'Jan Kiszka' via isar-users
2025-10-08 6:20 ` 'Badrikesh Prusty' via isar-users
2025-10-08 6:33 ` 'Badrikesh Prusty' via isar-users
2025-10-08 9:45 ` Anton Mikanovich [this message]
2025-10-13 4:38 ` 'Badrikesh Prusty' via isar-users
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1c93851a-7b56-4088-bd19-83ed1b1d8ecc@ilbers.de \
--to=amikan@ilbers.de \
--cc=arulcse2008@gmail.com \
--cc=badrikesh.prusty%siemens.com@gtempaccount.com \
--cc=ibr@ilbers.de \
--cc=isar-users@googlegroups.com \
--cc=jan.kiszka@siemens.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox