From: Florian Bezdeka <florian.bezdeka@siemens.com>
To: "Moessbauer, Felix (T CED INW-CN)" <felix.moessbauer@siemens.com>,
"isar-users@googlegroups.com" <isar-users@googlegroups.com>
Cc: "amikan@ilbers.de" <amikan@ilbers.de>,
"Kiszka, Jan (T CED)" <jan.kiszka@siemens.com>,
"Schild, Henning (T CED SES-DE)" <henning.schild@siemens.com>
Subject: Re: [PATCH] image: remove / entry from fstab template
Date: Fri, 6 Jan 2023 09:54:48 +0100 [thread overview]
Message-ID: <5e6f0d51-1079-9c0e-0c33-7c266d6ecde0@siemens.com> (raw)
In-Reply-To: <5983d0b4bf27504cca5f0288e7ad33e56c97aec8.camel@siemens.com>
On 06.01.23 05:58, Moessbauer, Felix (T CED INW-CN) wrote:
> On Thu, 2023-01-05 at 14:08 +0100, Florian Bezdeka wrote:
>> On Tue, 2022-10-04 at 13:46 +0200, Henning Schild wrote:
>>> An fstab entry for / should usually not be required. It just makes
>>> a
>>> rootfs less portable (USB vs. NFS, vs. SSD). The kernel is the one
>>> that should know which fs to mount and how. That is done via a
>>> bootloader.
>>>
>>> Should additional mount options be needed for a concrete image, an
>>> imager like wic might add an entry. i.e. for adding fspassno or
>>> x-systemd.growfs
>>>
>>> So to make the rootfs more generic and to allow imagers to add
>>> their
>>> individual / lines, instead of having to modify an existing one,
>>> remove
>>> the line from the template.
>>>
>>> Signed-off-by: Henning Schild <henning.schild@siemens.com>
>>
>> Hi all,
>>
>> FYI: While trying to update Isar in one layer using LAVA tests we
>> recognized that there is now a breaking change:
>>
>> Previously the rootfs was mounted rw on ext4 images because we had
>> the
>> / line in /etc/fstab, triggering systemd-remount logic.
>>
>> As the entry for the root (/) has been removed in /etc/fstab
>> remounting
>> does no longer happen. This breaks several LAVA tests here as well as
>> start-qemu scripts. Adding "rw" to the kernel cmdline would be one
>> option but we have a lot of LAVA job descriptions and even more
>> start-
>> qemu scripts in downstream layers.
>>
>> Felix already send out patches to OE (open embedded) to address that
>> in
>> wic. Necessary Isar patches (wic update, ...) will follow soon.
>> Thanks
>> a lot Felix for taking over.
>
> Hi Devs,
>
> let me add some additional information:
>
> - in OE, the fstab in the rootfs contains the / entry
> - we added a patch to OE to let WIC handle the root partition, but this
> had been reverted because it led to duplicated entries (which break
> systemd-remount)
>
> By that, we are no longer in sync with OE. Also considering the revert
> of the WIC patch, this shows that OE want to keep the / entry in the
> rootfs fstab for the meantime.
>
> Instead of discussing fundamental aspects (like if the line should be
> in the rootfs fstab or not) I vote for just integrating the
> deduplication logic in WIC and re-add the / entry in the rootfs fstab.
> This brings us in sync with OE again and from then on we can discuss
> how the general design should look like.
+1. Let's fix the current bug (duplicate entry for / in /etc/fstab)
first. Anything else can be addressed afterwards - if necessary at all.
>
> Felix
>
>>
>> Best regards,
>> Florian
>>
>
prev parent reply other threads:[~2023-01-06 8:55 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-10-04 11:46 Henning Schild
2022-10-04 11:53 ` Henning Schild
2022-10-04 11:54 ` Jan Kiszka
2022-10-04 12:21 ` Henning Schild
2022-10-11 18:03 ` Anton Mikanovich
2022-10-13 15:00 ` Henning Schild
2022-10-14 12:08 ` Anton Mikanovich
2022-10-14 16:27 ` Klincov, Wadim
2022-10-14 17:18 ` Anton Mikanovich
2022-10-14 12:10 ` Anton Mikanovich
2023-01-05 13:08 ` Florian Bezdeka
2023-01-06 4:58 ` Moessbauer, Felix
2023-01-06 8:54 ` Florian Bezdeka [this message]
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=5e6f0d51-1079-9c0e-0c33-7c266d6ecde0@siemens.com \
--to=florian.bezdeka@siemens.com \
--cc=amikan@ilbers.de \
--cc=felix.moessbauer@siemens.com \
--cc=henning.schild@siemens.com \
--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