public inbox for isar-users@googlegroups.com
 help / color / mirror / Atom feed
From: "'Niedermayr, BENEDIKT' via isar-users" <isar-users@googlegroups.com>
To: "isar-users@googlegroups.com" <isar-users@googlegroups.com>
Subject: Re: Cross-building for older releases on trixie broken?
Date: Wed, 5 Nov 2025 10:27:09 +0000	[thread overview]
Message-ID: <745cf04f-a281-4aea-a253-8147278992c7@siemens.com> (raw)
In-Reply-To: <57f1b43c-678b-4b1b-890e-6b4321ee5cc3@siemens.com>

On 27.10.25 10:43, 'Niedermayr, BENEDIKT' via isar-users wrote:
> On 03.10.25 15:50, Uladzimir Bely wrote:
>> On Fri, 2025-09-19 at 11:32 +0200, Jan Kiszka wrote:
>>> On 19.09.25 10:23, Uladzimir Bely wrote:
>>>> On Thu, 2025-09-18 at 07:49 +0200, 'Jan Kiszka' via isar-users
>>>> wrote:
>>>>> Hi,
>>>>>
>>>>> for some days, I'm playing with a trixie-based kas container
>>>>> (currently
>>>>> available via kas-isar:next from upstream), and I it seems I ran
>>>>> into
>>>>> an
>>>>> issue that might be isar-related: Building e.g. qemu-
>>>>> arm64:bookworm
>>>>> with
>>>>> that container fails:
>>>>>
>>>>>> Setup apt archive
>>>>>> -----------------
>>>>>>
>>>>>> Merged Build-Depends: debhelper-compat (= 13), build-
>>>>>> essential:amd64,
>>>>> crossbuild-essential-arm64:amd64, apt-utils:amd64, libc-
>>>>> dev:arm64,
>>>>> libstdc++-dev:arm64
>>>>>> Filtered Build-Depends: debhelper-compat (= 13),
>>>>> build-essential:amd64, crossbuild-essential-arm64:amd64,
>>>>> apt-utils:amd64, libc-dev:arm64, libstdc++-dev:arm64
>>>>>> dpkg-deb: building package 'sbuild-build-depends-main-dummy' in
>>>>> '/build/sshd-regen-keys-QAjoa6/resolver-
>>>>> ruuPrM/apt_archive/sbuild-
>>>>> build-depends-main-dummy.deb'.
>>>>>> Ign:1 copy:/build/sshd-regen-keys-QAjoa6/resolver-
>>>>>> ruuPrM/apt_archive
>>>>> ./ InRelease
>>>>>> Get:2 copy:/build/sshd-regen-keys-QAjoa6/resolver-
>>>>>> ruuPrM/apt_archive
>>>>> ./ Release [609 B]
>>>>>> Ign:3 copy:/build/sshd-regen-keys-QAjoa6/resolver-
>>>>>> ruuPrM/apt_archive
>>>>> ./ Release.gpg
>>>>>> Get:4 copy:/build/sshd-regen-keys-QAjoa6/resolver-
>>>>>> ruuPrM/apt_archive
>>>>> ./ Sources [690 B]
>>>>>> Get:5 copy:/build/sshd-regen-keys-QAjoa6/resolver-
>>>>>> ruuPrM/apt_archive
>>>>> ./ Packages [733 B]
>>>>>> Fetched 2032 B in 0s (0 B/s)
>>>>>> Reading package lists...
>>>>>> Reading package lists...
>>>>>>
>>>>>> Install main build dependencies (apt-based resolver)
>>>>>> ----------------------------------------------------
>>>>>>
>>>>>> Installing build dependencies
>>>>>> Reading package lists...
>>>>>> Building dependency tree...
>>>>>> Reading state information...
>>>>>> E: Unable to locate package sbuild-build-depends-main-dummy
>>>>>> apt-get failed.
>>>>>> E: Package installation failed
>>>>>> Not removing build depends: cloned chroot in use
>>>>>> Reading package lists...
>>>>>> Building dependency tree...
>>>>>> Reading state information...
>>>>>> E: Unable to locate package sbuild-build-depends-main-dummy
>>>>>> apt-get failed.
>>>>>>
>>>>>
>>>>> That was part of the sshd-regen-keys package build, but all
>>>>> package
>>>>> seem
>>>>> to fail this way.
>>>>>
>>>>> I suspect we are improperly using the host dpkg tooling here,
>>>>> rather
>>>>> than that from the target distro (see other recent discussions we
>>>>> had
>>>>> here)...
>>>>>
>>>>> Jan
>>>>>
>>>>> -- 
>>>>> Siemens AG, Foundational Technologies
>>>>> Linux Expert Center
>>>>
>>>> Hello, Jan.
>>>>
>>>> I took a quick look at this. It seems cross-building is broken for
>>>> trixie host only for bookworm target. Buster, bullseye and trixie
>>>> targets looks OK.
>>>>
>>>
>>> Hmm, maybe we are lucky with the other ones, but the same issue
>>> applies
>>> in principle.
>>>
>>>> The reason is still not clear for me - some failure happens with
>>>> apt/dependencies, even before dpkg-buildpackage is called.
>>>>
>>>> Will try to debug.
>>>>
>>>
>>> Thanks in advance!
>>>
>>> Jan
>>
>> Hello again.
>>
>> Got some time to debug it a bit more, but with no success.
> 
> I also stumbled over that issue after upgrading to KAS 5.0 (which now is
> based on trixie).
> 
> I suspect the root cause to be this sbuild commit/PR [1], which should
> solve this bug [2] regarding "libglib2.0-dev" builds/deps.
> 
> What confuses me a bit is that IMHO generally no cross-build originating
> from trixie should work anymore.
> 
> 
> [1]
> https://salsa.debian.org/debian/sbuild/-/merge_requests/69/diffs?commit_id=3fe5278e47de798ef9cafabb8674cd8e1f41ab0b
> 
> [2] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=%231071246
> 
>>
>> So, the picture is still the same:
>>
>> - Cross building trixie/bookworm (host/target) fails.
>> - It fails on any target architecture
>> - Non-ross building trixie/bookworm (host/target) is OK
>> - Cross building trixie/buster, trixie/bullseye, trixie/trixie is OK.
> 
> Can I reproduce that locally somehow or is this a result of the CI?
> 
> 
> Regards,
> Benedikt
>>
>> For me it looks like sbuild issue. But the reason remains unclear. ALl
>> build depependencies sbuild lists are already in sbuild chroot, but if
>> fails to generate dummy package...
>>
>> Probably, it could be useful to try trixie/bookworm cross-building
>> outside of isar to get clear understanding if sbuild itself is in blame
>> or the way we use it is incorrect.
>>
>>
> 

Hi again,

after some testing it turns out that only cross-building from trixie for 
bookworm fails. So I can confirm what Uladzimir reported.

Adding this patch [1] to sbuild (0.89.3+deb13u1) fixes this error. I 
created a bug report for that [2].

Not sure yet if this patch will make it into the current Debian trixie 
release but it seems that we depend on it. At least I have currently no 
idea howto workaround this issue without using the patch.


[1] 
https://salsa.debian.org/debian/sbuild/-/commit/0da1ccda97e6a4134c2aadf85913680fd8274e2b

[2] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1119344

Regards,
Benedikt

-- 
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/745cf04f-a281-4aea-a253-8147278992c7%40siemens.com.

  reply	other threads:[~2025-11-05 10:27 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-09-18  5:49 'Jan Kiszka' via isar-users
2025-09-19  8:23 ` Uladzimir Bely
2025-09-19  9:32   ` 'Jan Kiszka' via isar-users
2025-10-03 13:50     ` Uladzimir Bely
2025-10-27  9:43       ` 'Niedermayr, BENEDIKT' via isar-users
2025-11-05 10:27         ` 'Niedermayr, BENEDIKT' via isar-users [this message]
2025-11-05 13:04           ` 'Quirin Gylstorff' via isar-users
2025-11-05 14:45             ` 'Jan Kiszka' 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=745cf04f-a281-4aea-a253-8147278992c7@siemens.com \
    --to=isar-users@googlegroups.com \
    --cc=benedikt.niedermayr@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