public inbox for isar-users@googlegroups.com
 help / color / mirror / Atom feed
From: Jan Kiszka <jan.kiszka@siemens.com>
To: Uladzimir Bely <ubely@ilbers.de>,
	isar-users <isar-users@googlegroups.com>
Subject: Re: [PATCH v6] dpkg: Restore support for replacing pre-installed packages in sbuild-chroot
Date: Fri, 19 Jan 2024 10:42:49 +0100	[thread overview]
Message-ID: <9f23f168-f5ee-44f5-aaa5-a242632dcbc3@siemens.com> (raw)
In-Reply-To: <932bf7a167b931150719479296a509996c8b8536.camel@ilbers.de>

On 19.01.24 10:15, Uladzimir Bely wrote:
> On Fri, 2024-01-19 at 08:49 +0100, Jan Kiszka wrote:
>> On 19.01.24 08:44, Jan Kiszka wrote:
>>> On 19.01.24 08:40, Jan Kiszka wrote:
>>>> On 19.01.24 08:37, Uladzimir Bely wrote:
>>>>> On Fri, 2024-01-19 at 10:20 +0300, Uladzimir Bely wrote:
>>>>>> On Thu, 2024-01-18 at 14:50 +0100, Jan Kiszka wrote:
>>>>>>> From: Jan Kiszka <jan.kiszka@siemens.com>
>>>>>>>
>>>>>>> During the migration to sbuild, support for using self-
>>>>>>> built
>>>>>>> packages
>>>>>>> in
>>>>>>> the build environment was lost if those were already part
>>>>>>> of the
>>>>>>> sbuild-chroot. This restores it by adding --apt-distupgrade
>>>>>>> to the
>>>>>>> sbuild call. But that is not enough because sbuild will
>>>>>>> only
>>>>>>> upgrade
>>>>>>> packages from already configured sources, not those
>>>>>>> specified via
>>>>>>> --extra-repository. We therefore have to switch back to
>>>>>>> configuring
>>>>>>> isar-apt during sbuild-chroot creation.
>>>>>>>
>>>>>>> As rootfs_configure_isar_apt configures the isar repo under
>>>>>>> /isar-apt, we bind-mount the one in /home/builder/${PN} to
>>>>>>> that
>>>>>>> folder.
>>>>>>> Another difference is that we now need to run apt-get
>>>>>>> update
>>>>>>> explicitly,
>>>>>>> but only for isar-apt.
>>>>>>>
>>>>>>> For apt fetching, we neither need nor want isar-apt to be
>>>>>>> available.
>>>>>>> Rebuilding self-generated apt packages is generally
>>>>>>> pointless as
>>>>>>> the
>>>>>>> needs are better addressed in the generating recipe itself.
>>>>>>> Exposing
>>>>>>> isar-apt to the fetch may furthermore lead to fetching a
>>>>>>> previously
>>>>>>> built source package of the same recipe, rather than
>>>>>>> pulling the
>>>>>>> external version.
>>>>>>>
>>>>>>> And because the sbuild-chroots are now left behind with
>>>>>>> isar-apt
>>>>>>> configured, the configuration in imager_run can be removed
>>>>>>> (credits
>>>>>>> to
>>>>>>> Srinuvasan Arjunan).
>>>>>>>
>>>>>>> Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
>>>>>>> ---
>>>>>>>
>>>>>>> Changes in v6:
>>>>>>>  - basically, going back to v4 but disabling isar-apt in
>>>>>>> do_apt_fetch
>>>>>>>
>>>>>>> I was able to resolve the scenario that Uladzimir was
>>>>>>> sharing this
>>>>>>> way. 
>>>>>>> And I also realized that apt fetching as well as unpacking
>>>>>>> should
>>>>>>> have 
>>>>>>> no deal with isar-apt. So, v5 was taking a wrong turn.
>>>>>>>
>>>>>>>  meta/classes/dpkg-base.bbclass                | 29
>>>>>>> +++++++++++----
>>>>>>> --
>>>>>>> --
>>>>>>>  meta/classes/dpkg.bbclass                     |  5 ++--
>>>>>>>  meta/classes/image-tools-extension.bbclass    | 13 -------
>>>>>>> --
>>>>>>>  .../sbuild-chroot/sbuild-chroot.inc           |  2 --
>>>>>>>  4 files changed, 20 insertions(+), 29 deletions(-)
>>>>>>>
>>>>>>> diff --git a/meta/classes/dpkg-base.bbclass
>>>>>>> b/meta/classes/dpkg-
>>>>>>> base.bbclass
>>>>>>> index 7b054d3f..80686677 100644
>>>>>>> --- a/meta/classes/dpkg-base.bbclass
>>>>>>> +++ b/meta/classes/dpkg-base.bbclass
>>>>>>> @@ -19,8 +19,6 @@ DEPENDS:append:riscv64 = "${@'
>>>>>>> crossbuild-
>>>>>>> essential-riscv64' if d.getVar('ISAR_C
>>>>>>>  DEB_BUILD_PROFILES ?= ""
>>>>>>>  DEB_BUILD_OPTIONS ?= ""
>>>>>>>  
>>>>>>> -ISAR_APT_REPO ?= "deb [trusted=yes]
>>>>>>> file:///home/builder/${PN}/isar-
>>>>>>> apt/${DISTRO}-${DISTRO_ARCH}/apt/${DISTRO} ${DEBDISTRONAME}
>>>>>>> main"
>>>>>>> -
>>>>>>>  python do_adjust_git() {
>>>>>>>      import subprocess
>>>>>>>  
>>>>>>> @@ -115,6 +113,8 @@ do_apt_fetch() {
>>>>>>>      trap 'exit 1' INT HUP QUIT TERM ALRM USR1
>>>>>>>      trap 'schroot_cleanup' EXIT
>>>>>>>  
>>>>>>> +    schroot -d / -u root -c ${SBUILD_CHROOT} -- \
>>>>>>> +        rm /etc/apt/sources.list.d/isar-apt.list
>>>>>>> /etc/apt/preferences.d/isar-apt
>>>>>>
>>>>>> The issue is still reproducible with the steps I posted
>>>>>> below, as
>>>>>> well
>>>>>> as CI still fails with v6.
>>>>>> I think, the steps above are not sufficient to completely
>>>>>> stop
>>>>>> dealing
>>>>>> with isar-apt in do_apt_fetch() and we need some kind of 'apt
>>>>>> update'
>>>>>> here to consider removed list and preferences. The question
>>>>>> is how to
>>>>>> say apt to forget about isar-apt and do not run apt update
>>>>>> for other
>>>>>> source lists... maybe, simply remove isar-apt files from
>>>>>> /var/lib/apt/lists/.
>>>>>
>>>>> An addition:
>>>>>
>>>>> It seems that files are not really removed from the session for
>>>>> some
>>>>> reason. I added some kind of "ls" near remove code:
>>>>>
>>>>> ```
>>>>> +    schroot -d / -u root -c ${SBUILD_CHROOT} -- \
>>>>> +        sh -c 'ls -la /etc/apt/sources.list.d/'
>>>>>      schroot -d / -u root -c ${SBUILD_CHROOT} -- \
>>>>>          rm /etc/apt/sources.list.d/isar-apt.list
>>>>> /etc/apt/preferences.d/isar-apt
>>>>> +    schroot -d / -u root -c ${SBUILD_CHROOT} -- \
>>>>> +        sh -c 'ls -la /etc/apt/sources.list.d/'
>>>>> ```
>>>>> , but still see the following in the log:
>>>>> ```
>>>>>> DEBUG: Executing shell function do_apt_fetch
>>>>>> total 16
>>>>>> drwxr-xr-x 2 root root 4096 Jan 19 07:09 .
>>>>>> drwxr-xr-x 8 root root 4096 Jan 19 07:09 ..
>>>>>> -rw-r--r-- 1 root root  569 Jan 19 07:09 bootstrap.list
>>>>>> -rw-r--r-- 1 root root   94 Jan 19 07:09 isar-apt.list
>>>>>> total 16
>>>>>> drwxr-xr-x 2 root root 4096 Jan 19 07:09 .
>>>>>> drwxr-xr-x 8 root root 4096 Jan 19 07:09 ..
>>>>>> -rw-r--r-- 1 root root  569 Jan 19 07:09 bootstrap.list
>>>>>> -rw-r--r-- 1 root root   94 Jan 19 07:09 isar-apt.list
>>>>>> Reading package lists...
>>>>> ```
>>>>> Fith the apt_fetch failure followed, because of existing of
>>>>> previously
>>>>> built hello source package in the isar-apt repo.
>>>>>
>>>>
>>>> OK... last-minute change of mine, splitting up the purging and
>>>> the
>>>> actual fetching into two schroot runs to avoid having to run the
>>>> latter
>>>> with "-u root". I didn't find a way to de-configure isar-apt
>>>> unprivileged. Back to the drawing board. :(
>>>>
>>>
>>> Comparing the split run of schroot with image-tool-extension, I
>>> think
>>> I'm just missing a proper session ID here.
>>>
>>
>> Before sending v7: can you try this on top?
>>
>> diff --git a/meta/classes/dpkg-base.bbclass b/meta/classes/dpkg-
>> base.bbclass
>> index 80686677..1a3333ac 100644
>> --- a/meta/classes/dpkg-base.bbclass
>> +++ b/meta/classes/dpkg-base.bbclass
>> @@ -113,10 +113,13 @@ do_apt_fetch() {
>>      trap 'exit 1' INT HUP QUIT TERM ALRM USR1
>>      trap 'schroot_cleanup' EXIT
>>  
>> -    schroot -d / -u root -c ${SBUILD_CHROOT} -- \
>> +    session_id=$(schroot -q -b -c ${SBUILD_CHROOT})
>> +    echo "Started session: ${session_id}"
>> +
>> +    schroot -r -c ${session_id} -d / -u root -- \
>>          rm /etc/apt/sources.list.d/isar-apt.list
>> /etc/apt/preferences.d/isar-apt
>>      for uri in "${SRC_APT}"; do
>> -        schroot -d / -c ${SBUILD_CHROOT} -- \
>> +        schroot -r -c ${session_id} -d / -- \
>>              sh -c 'mkdir -p /downloads/deb-src/"$1"/"$2" && cd
>> /downloads/deb-src/"$1"/"$2" && apt-get -y --download-only --only-
>> source source "$2"' my_script "${BASE_DISTRO}-
>> ${BASE_DISTRO_CODENAME}" "${uri}"
>>      done
>>      schroot_delete_configs
>>
>> Jan
>>
> 
> Yes, with persistent session it works, at least with reproduce steps I
> used. I just added session cleanup (e.g, "schroot -e -c ${session_id}")
> at the end.
> 
> Before sending v7 I could also check if it is not reproduced anymore in
> CI.
> 
> Also, due to using schroot session we probably also need to care about
> proper session removing in case commands under schroot fail, similar to
> how it's done in image-tools-extention by imager_cleanup().

Yeah, still need to check that - schroot is also still new to me,
learned a lot about it these days.

Jan

-- 
Siemens AG, Technology
Linux Expert Center


  reply	other threads:[~2024-01-19  9:42 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-01-18 13:50 Jan Kiszka
2024-01-19  7:20 ` Uladzimir Bely
2024-01-19  7:37   ` Uladzimir Bely
2024-01-19  7:40     ` Jan Kiszka
2024-01-19  7:44       ` Jan Kiszka
2024-01-19  7:49         ` Jan Kiszka
2024-01-19  9:15           ` Uladzimir Bely
2024-01-19  9:42             ` Jan Kiszka [this message]
2024-01-19  9:49               ` Jan Kiszka
2024-01-22  7:59                 ` Uladzimir Bely
2024-01-27  7:35                   ` Uladzimir Bely

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=9f23f168-f5ee-44f5-aaa5-a242632dcbc3@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