From: Jan Kiszka <jan.kiszka@siemens.com>
To: Uladzimir Bely <ubely@ilbers.de>, isar-users@googlegroups.com
Subject: Re: [PATCH v2 04/24] linux-module: Do not use shell environment
Date: Wed, 24 Nov 2021 07:13:17 +0100 [thread overview]
Message-ID: <25fe32be-fbbd-5f21-177a-bcb31aa20fd8@siemens.com> (raw)
In-Reply-To: <2392353.XAFRqVoOGU@home>
On 23.11.21 13:24, Uladzimir Bely wrote:
> In the email from Friday, 19 Nov 2021. 15:44:12 +03 user Jan Kiszka wrote:
>> On 19.11.21 13:13, Uladzimir Bely wrote:
>>
>>> +KDIR := $(shell dpkg -L $(KERNEL_DEP) | grep "/lib/modules/.*/build")
>>> +endif
>>> +
>>> +# With sbuild `dh clean` is called before chroot's apt database updated.
>>> So, +# KDIR is empty at that moment and the override causes error. Will
>>> skip it. +ifneq ($(KDIR),)
>>>
>>> override_dh_auto_clean:
>>> $(MAKE) -C $(KDIR) M=$(PWD) clean
>>>
>>> +endif
>>
>> This looks wrong, or at least fragile. Why can we live without a proper
>> clean on the kernel tree?
>>
>> Jan
>
> Actually, this workaround with `ifneq ($(KDIR),)` I added to the patch after
> the commit https://github.com/ilbers/isar/commit/248436880eb2 was applied to
> `next`.
>
> This change doesn't cancel a proper clean on the kernel. Here are some example
> logs:
>
> `next`: https://gist.github.com/WiseLord/0fb98bc036532ec03e4f763b2ba06a41
> `sbuild`: https://gist.github.com/WiseLord/3c35d477efc025bcc9cd3c5cbd3dbc5d
>
> On current `next` code dh_clean (and override-related `make ... clean` on
> kernel code) is executed by dpkg-buildpackage. At this point KDIR is set to a
> proper value, so it works as expected.
>
> With sbuild dh_clean happens to be executed twice:
> - 1: immediately after sbuild is run, executed by sbuild itself
> - 2: after few stages (like downloading deps and so on), executed by dpkg-
> buildpackage.
>
> The issue is that at 1st execution KDIR is not set, so `make ... clean`
> (coming with override) produces wrong command.
>
> That's why this workaround (skip `make ... clean` in case of empty KDIR) was
> added. Proper cleaning is still performed, but on the second call of dh_clean.
>
> Probably, there is a way to say sbuild not to run this 'generic' dh_clean and
> use just one from dpkg-buildpackage later...
>
Thanks for the explanation. How about clarifying that no clean is lost
in the commit message?
Jan
--
Siemens AG, T RDA IOT
Corporate Competence Center Embedded Linux
next prev parent reply other threads:[~2021-11-24 6:13 UTC|newest]
Thread overview: 41+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-11-19 12:13 [PATCH v2 00/24] Sbuild/Schroot migration Uladzimir Bely
2021-11-19 12:13 ` [PATCH v2 01/24] dpkg: Install raw package files to source root Uladzimir Bely
2021-11-19 12:13 ` [PATCH v2 02/24] dpkg-gbp: Use separate command to export tarball Uladzimir Bely
2021-11-19 12:13 ` [PATCH v2 03/24] isar-bootstrap: Export bootstrap to schroot config Uladzimir Bely
2021-11-19 12:13 ` [PATCH v2 04/24] linux-module: Do not use shell environment Uladzimir Bely
2021-11-19 12:44 ` Jan Kiszka
2021-11-19 12:45 ` Jan Kiszka
2021-11-23 12:24 ` Uladzimir Bely
2021-11-24 6:13 ` Jan Kiszka [this message]
2021-11-25 5:47 ` Uladzimir Bely
2021-11-19 12:13 ` [PATCH v2 05/24] u-boot: " Uladzimir Bely
2021-11-19 12:13 ` [PATCH v2 06/24] trusted-firmware: " Uladzimir Bely
2021-11-19 12:13 ` [PATCH v2 07/24] optee-os: " Uladzimir Bely
2021-11-19 12:13 ` [PATCH v2 08/24] kselftest: " Uladzimir Bely
2021-11-19 12:13 ` [PATCH v2 09/24] dpkg: Build packages with sbuild Uladzimir Bely
2021-11-19 12:13 ` [PATCH v2 10/24] sbuild: Introduce environment variables export API Uladzimir Bely
2021-11-21 9:07 ` Jan Kiszka
2021-11-19 12:13 ` [PATCH v2 11/24] dpkg-gbp: Migrate to schroot Uladzimir Bely
2021-11-19 12:13 ` [PATCH v2 12/24] linux-mainline: Move cfg fragment test to debian/rules Uladzimir Bely
2021-11-19 12:13 ` [PATCH v2 13/24] linux-custom: Prepare kernel config inside sbuild Uladzimir Bely
2021-11-19 12:13 ` [PATCH v2 14/24] sbuild: Add recipes for host and target rootfs to run sbuild Uladzimir Bely
2021-11-19 12:13 ` [PATCH v2 15/24] sbuild: Mount base-apt in schroot Uladzimir Bely
2021-11-19 12:13 ` [PATCH v2 16/24] sbuild: Add sbuildshell task Uladzimir Bely
2021-11-19 12:13 ` [PATCH v2 17/24] dpkg-gbp: Preinstall gbp utils in schroot Uladzimir Bely
2021-11-19 12:13 ` [PATCH v2 18/24] dpkg: Remove builddeps install task Uladzimir Bely
2021-11-19 12:13 ` [PATCH v2 19/24] dpkg-base: Switch devshell to use schroot Uladzimir Bely
2021-11-19 12:13 ` [PATCH v2 20/24] dpkg-base: Switch apt_fetch and apt_unpack " Uladzimir Bely
2021-11-19 12:13 ` [PATCH v2 21/24] dpkg-base: Cleanup from buildchroot parts Uladzimir Bely
2021-11-19 12:13 ` [PATCH v2 22/24] dpkg-gbp: Use host tools for dsc preparation Uladzimir Bely
2021-11-19 12:13 ` [PATCH v2 23/24] doc: Add sbuild-related documentation Uladzimir Bely
2021-11-19 12:13 ` [PATCH v2 24/24] sbuild: Replace isar-apt mounting with copying Uladzimir Bely
2021-11-19 21:48 ` [PATCH v2 00/24] Sbuild/Schroot migration Henning Schild
2021-11-21 9:07 ` Jan Kiszka
2021-11-23 13:05 ` Uladzimir Bely
2021-11-26 6:43 ` Jan Kiszka
2021-11-26 8:03 ` Uladzimir Bely
2021-11-26 8:50 ` Jan Kiszka
2021-12-01 12:11 ` Jan Kiszka
2021-11-26 12:09 ` Michael Adler
2021-11-26 12:57 ` Uladzimir Bely
2021-11-26 14:58 ` Jan Kiszka
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=25fe32be-fbbd-5f21-177a-bcb31aa20fd8@siemens.com \
--to=jan.kiszka@siemens.com \
--cc=isar-users@googlegroups.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