Re: [RFC v2 19/19] use copy of sbom-chroot for sbom creation

["'MOESSBAUER, Felix' via isar-users" <isar-users@googlegroups.com>]

On Tue, 2026-02-24 at 14:59 +0300, Uladzimir Bely wrote:
> On Tue, 2026-02-24 at 11:36 +0000, MOESSBAUER, Felix wrote:
> > On Tue, 2026-02-24 at 13:33 +0300, Uladzimir Bely wrote:
> > > Hello Felix.
> > > 
> > > On Fri, 2026-02-20 at 18:16 +0100, 'Felix Moessbauer' via isar-
> > > users
> > > wrote:
> > > > We previously used the same sbom-chroot for generating the sbom
> > > > of
> > > > different root filesystems. This required to have a live copy of
> > > > the
> > > > sbom-chroot in the deploy dir, on which also was operated on.
> > > > Further,
> > > > this copy was left behind in the deploy dir.
> > > > 
> > > > We improve this by just storing a minimized tarball of the sbom-
> > > > chroot
> > > > in the deploy dir and extract that into the workdir of the
> > > > rootfs.
> > > > 
> > > > Signed-off-by: Felix Moessbauer <felix.moessbauer@siemens.com>
> > > > ---
> > > >  meta/classes/sbom.bbclass                     | 29
> > > > ++++++++++++++++-
> > > > --
> > > >  .../sbom-chroot/sbom-chroot.bb                | 11 ++++++-
> > > >  2 files changed, 35 insertions(+), 5 deletions(-)
> > > > 
> > > > diff --git a/meta/classes/sbom.bbclass
> > > > b/meta/classes/sbom.bbclass
> > > > index e3d0e702..69c5d1b0 100644
> > > > --- a/meta/classes/sbom.bbclass
> > > > +++ b/meta/classes/sbom.bbclass
> > > > @@ -23,7 +23,8 @@ SBOM_SPDX_NAMESPACE_PREFIX ?=
> > > > "https://spdx.org/spdxdocs"
> > > >  DEPLOY_DIR_SBOM = "${DEPLOY_DIR_IMAGE}"
> > > >  
> > > >  SBOM_DIR = "${DEPLOY_DIR}/sbom"
> > > > -SBOM_CHROOT = "${SBOM_DIR}/sbom-chroot"
> > > > +SBOM_CHROOT = "${SBOM_DIR}/sbom-chroot.tar.zst"
> > > > +SBOM_CHROOT_LOCAL = "${WORKDIR}/sbom-chroot"
> > > 
> > > This change also requires appropriate changes in at least image-
> > > tools-
> > > extension.bbclass and imagetype_wic.bbclass, e.g.:
> > 
> > Yes, I also found this later on and already fixed it. Will be part of
> > the v3. Are you considering this patch relevant for the release? In
> > this case I can just send the fixed version as a standalone patch on
> > next. Just let me know.
> > 
> > 
> > Not sure it's ready enough for release.

I was just referring to the single patch "use copy of sbom-chroot for
sbom creation".

> > 
> > I managed to build some targets using KAS_CONTAINER_ENGINE=podman,
> > but it doesn't work with default "docker" value. Initial issue comes
> > from option "--userns=keep-id" podman has since docker doesn't have.
> > But even when bypassing it, I faced some more issues under docker.

I did not yet prepare the kas-container for that. The whole thing is
only tested in podman (yet), but hopefully I can also make it work
under docker.

However, I'm currently hunting a very strange bug on
citest.py:CrossTest.test_cross_kselftest where the
rootfs_install_sstate_finalize task fails under schroot (rootfull) and
sbuild-chroot-host for arm64.

The bug reproduces reliably in CI, but not locally.

Felix

-- 
Siemens AG
Linux Expert Center
Friedrich-Ludwig-Bauer-Str. 3
85748 Garching, Germany

-- 
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/36a2763f8052c28fed91e81f093809b257e7b71e.camel%40siemens.com.