From: Claudius Heine <ch@denx.de>
To: Jan Kiszka <jan.kiszka@siemens.com>,
"Scheler, Fabian" <Fabian.Scheler@siemens.com>,
Uladzimir Bely <ubely@ilbers.de>,
Anton Mikanovich <amikan@ilbers.de>,
"isar-users@googlegroups.com" <isar-users@googlegroups.com>
Subject: Re: Accessing sbuild build-directory
Date: Tue, 28 May 2024 10:49:11 +0200 [thread overview]
Message-ID: <b880cb68-7a84-423e-b6c1-9d9a9213c11a@denx.de> (raw)
In-Reply-To: <a67c9d3c-3f1b-4f01-9787-78d6f30eef02@siemens.com>
Hi Jan,
On 2024-05-28 9:58 am, Jan Kiszka wrote:
> On 28.05.24 09:29, Claudius Heine wrote:
>> Hi everyone,
>>
>> On 2024-04-05 3:10 pm, 'Scheler, Fabian' via isar-users wrote:
>>> Am 05.04.2024 um 12:03 schrieb Scheler, Fabian (T CED EDC-DE):
>>>> Am 05.04.2024 um 11:57 schrieb Uladzimir Bely:
>>>>> On Fri, 2024-04-05 at 12:09 +0300, Anton Mikanovich wrote:
>>>>>> 05/04/2024 11:38, 'Scheler, Fabian (T CED SES-DE)' via isar-users
>>>>>> wrote:
>>>>>>> Hi all,
>>>>>>>
>>>>>>> I want to extract the coverage information (i.e. the gcno-files
>>>>>>> created by gcc) that is generated during the package build via
>>>>>>> sbuild
>>>>>>> and I look for some hook that allows to access the content of the
>>>>>>> sbuild build-directory after the build from within a recipe.
>>>>>>>
>>>>>>> Is this possible?
>>>>>>>
>>>>>>> Ciao
>>>>>>> Fabian
>>>>>>
>>>>>> Hello Fabian,
>>>>>>
>>>>>> You can refer to sbuild documentation for hooks explanation.
>>>>>> The last one executed inside session before its cleanup is
>>>>>> '--chroot-cleanup-commands'.
>>>>>> So you can modify dpkg.bbclass adding any custom commands with this
>>>>>> hook.
>>>>>>
>>>>>> There are no already implemented dummy pre/post hooks for now, but we
>>>>>> can try to
>>>>>> implement them in future if needed.
>>>>>>
>>>>>
>>>>> FWIW, perhaps 6d5fbbab could be used to quickly play with custom hooks
>>>>> to sbuild process. A different approach can be chosen for the final
>>>>> implementation (if any).
>>>>>
>>>>
>>>> thanks for your quick feedback. Yes, injecting arguments to sbuild like
>>>> proposed in 6d5fbbab should be an option. I'll give it a try.
>>>
>>> FYI: Injecting --post-build-commands finally did the trick, but it is
>>> not really nice. However, there seems not to be another option to
>>> accomplish that. Unfortunately 6d5fbbab has been reverted and I can
>>> follow Jan's arguments, maybe this is a use-case justifying the
>>> extension.
>>
>> I am currently struggling with a similar issue trying to extract the
>> `.config` used to compile the swupdate recipe from cip-core [1].
>>
>> Because the debian package uses multiple different build profiles as
>> well as a base debian defconfig to generate a `debian_defconfig`, which
>> is then used to generate a `.config` it becomes very difficult to verify
>> if the configuration I patched into the base debian defconfig has
>> actually applied or was overwritten by something else along the way.
>>
>> With openembedded you could just look in the build directory and see if
>> every option that you wanted is set correctly in the `.config` file, but
>> here this is not possible.
>>
>> It might be possible to add the `.config` into a binary package and read
>> it from there, but there should be an option to investigate the build
>> directory in isar itself.
>>
>
> We should add an option to leave the build artifacts behind when leaving
> sbuild, just like we had in the ugly old times with buildchroot. I
> wanted to study kernel .config and similar stuff in the past more than
> once but always found other ad-hoc workarounds to actually solve that
> demand properly.
So what are the options for implementing this?
It would be good that even after a failed build, it is possible to
investigate the build directory.
regards,
Claudius
--
DENX Software Engineering GmbH, Managing Director: Erika Unter
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-54 Fax: (+49)-8142-66989-80 Email: ch@denx.de
next prev parent reply other threads:[~2024-05-28 8:49 UTC|newest]
Thread overview: 12+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-04-05 8:38 Scheler, Fabian (T CED SES-DE)
2024-04-05 9:09 ` Anton Mikanovich
2024-04-05 9:57 ` Uladzimir Bely
2024-04-05 10:03 ` Scheler, Fabian
2024-04-05 13:10 ` Scheler, Fabian
2024-05-28 7:29 ` Claudius Heine
2024-05-28 7:58 ` Jan Kiszka
2024-05-28 8:49 ` Claudius Heine [this message]
2024-05-28 8:52 ` Jan Kiszka
2024-05-29 11:24 ` Anton Mikanovich
2024-05-29 12:33 ` Jan Kiszka
2024-05-29 12:54 ` Baurzhan Ismagulov
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=b880cb68-7a84-423e-b6c1-9d9a9213c11a@denx.de \
--to=ch@denx.de \
--cc=Fabian.Scheler@siemens.com \
--cc=amikan@ilbers.de \
--cc=isar-users@googlegroups.com \
--cc=jan.kiszka@siemens.com \
--cc=ubely@ilbers.de \
/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