[PATCH v3 0/2] Enable systemd units based on systemd presets

[PATCH v3 0/2] Enable systemd units based on systemd presets

[alexander.heinisch via isar-users]

From: Alexander Heinisch <alexander.heinisch@siemens.com>

By default population of the presets is automatically done by systemd
on first-boot.

There were several issues with that:

1. The rootfs we get as a build artifact does not reflect the actual
system running in the field.

2. For setups without writeable /etc this fails. With that addition
it happens already at build time.

Note: Additional services are enabled only. Services already enabled
during the package installation won't be changed.

Opt-out: `ROOTFS_FEATURES:remove = "populate-systemd-preset"`

Changes since v2:
  - Enable postprocessing to populate systemd presets via ROOTFS_FEATURES

Alexander Heinisch (2):
  Minor cleanup: Refactor ROOTFS_FEATURES assignment for image to make
    upcoming changes more easily trackable.
  classes/image-postproc: Enable systemd units based on systemd presets

 RECIPE-API-CHANGELOG.md     | 18 ++++++++++++++++++
 meta/classes/image.bbclass  | 10 +++++++++-
 meta/classes/rootfs.bbclass |  6 ++++++
 3 files changed, 33 insertions(+), 1 deletion(-)

--
2.39.5

-- 
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/20250917103506.295010-1-alexander.heinisch%40siemens.com.

[PATCH v3 1/2] Minor cleanup: Refactor ROOTFS_FEATURES assignment for image to make upcoming changes more easily trackable.

[alexander.heinisch via isar-users]

From: Alexander Heinisch <alexander.heinisch@siemens.com>

Signed-off-by: Alexander Heinisch <alexander.heinisch@siemens.com>
---
 meta/classes/image.bbclass | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/meta/classes/image.bbclass b/meta/classes/image.bbclass
index bd1b8552..d833564f 100644
--- a/meta/classes/image.bbclass
+++ b/meta/classes/image.bbclass
@@ -66,7 +66,14 @@ inherit multiarch
 inherit essential
 
 ROOTFSDIR = "${IMAGE_ROOTFS}"
-ROOTFS_FEATURES += "clean-package-cache clean-pycache generate-manifest export-dpkg-status clean-log-files clean-debconf-cache"
+ROOTFS_FEATURES += "\
+    clean-package-cache \
+    clean-pycache \
+    generate-manifest \
+    export-dpkg-status \
+    clean-log-files \
+    clean-debconf-cache \
+    "
 # when using a custom initrd, do not generate one as part of the image rootfs
 ROOTFS_FEATURES += "${@ '' if d.getVar('INITRD_IMAGE') == '' else 'no-generate-initrd'}"
 ROOTFS_PACKAGES += "${IMAGE_PREINSTALL} ${@isar_multiarch_packages('IMAGE_INSTALL', d)}"
-- 
2.39.5

-- 
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/20250917103506.295010-2-alexander.heinisch%40siemens.com.

[PATCH v3 2/2] classes/image-postproc: Enable systemd units based on systemd presets

[alexander.heinisch via isar-users]

From: Alexander Heinisch <alexander.heinisch@siemens.com>

By default population of the presets is automatically done by systemd
on first-boot.

There were several issues with that:

1. The rootfs we get as a build artifact does not reflect the actual
system running in the field.

2. For setups without writeable /etc this fails. With that addition
it happens already at build time.

Note: Additional services are enabled only. Services already enabled
during the package installation won't be changed.

Opt-out: `ROOTFS_FEATURES:remove = "populate-systemd-preset"`

Signed-off-by: Alexander Heinisch <alexander.heinisch@siemens.com>
---
 RECIPE-API-CHANGELOG.md     | 18 ++++++++++++++++++
 meta/classes/image.bbclass  |  1 +
 meta/classes/rootfs.bbclass |  6 ++++++
 3 files changed, 25 insertions(+)

diff --git a/RECIPE-API-CHANGELOG.md b/RECIPE-API-CHANGELOG.md
index 92e7811c..f2e21088 100644
--- a/RECIPE-API-CHANGELOG.md
+++ b/RECIPE-API-CHANGELOG.md
@@ -741,3 +741,21 @@ By setting `MS_TPM_20_REF_DIR` in an optee-ftpm recipe, it is now possible to
 use the new optee_ftpm code base from the OP-TEE project. That variable has to
 point to a subdir in `WORKDIR` which contains the unpacked ms-tpm-20-ref source
 code.
+
+### Populate systemd units based on presets during image postprocessing
+
+By default population of the presets is automatically done by systemd
+on first-boot.
+
+There were several issues with that:
+
+1. The rootfs we get as a build artifact does not reflect the actual
+system running in the field.
+
+2. For setups without writeable /etc this fails. With that addition
+it happens already at build time.
+
+**Note**: Additional services are enabled only. Services already enabled
+during the package installation won't be changed.
+
+Opt-out: `ROOTFS_FEATURES:remove = "populate-systemd-preset"`
diff --git a/meta/classes/image.bbclass b/meta/classes/image.bbclass
index d833564f..44b59420 100644
--- a/meta/classes/image.bbclass
+++ b/meta/classes/image.bbclass
@@ -73,6 +73,7 @@ ROOTFS_FEATURES += "\
     export-dpkg-status \
     clean-log-files \
     clean-debconf-cache \
+    populate-systemd-preset \
     "
 # when using a custom initrd, do not generate one as part of the image rootfs
 ROOTFS_FEATURES += "${@ '' if d.getVar('INITRD_IMAGE') == '' else 'no-generate-initrd'}"
diff --git a/meta/classes/rootfs.bbclass b/meta/classes/rootfs.bbclass
index ebe3bf4a..fdfad2fa 100644
--- a/meta/classes/rootfs.bbclass
+++ b/meta/classes/rootfs.bbclass
@@ -15,6 +15,7 @@ ROOTFS_BASE_DISTRO ?= "${BASE_DISTRO}"
 # 'export-dpkg-status' - exports /var/lib/dpkg/status file to ${ROOTFS_DPKGSTATUS_DEPLOY_DIR}
 # 'clean-log-files' - delete log files that are not owned by packages
 # 'no-generate-initrd' - do not generate debian default initrd
+# 'populate-systemd-preset' - enable systemd units according to systemd presets
 ROOTFS_FEATURES ?= ""
 
 ROOTFS_APT_ARGS="install --yes -o Debug::pkgProblemResolver=yes"
@@ -429,6 +430,11 @@ rootfs_cleanup_base_apt() {
 EOSUDO
 }
 
+ROOTFS_POSTPROCESS_COMMAND += "${@bb.utils.contains('ROOTFS_FEATURES', 'populate-systemd-preset', 'image_postprocess_populate_systemd_preset', '', d)}"
+image_postprocess_populate_systemd_preset() {
+    sudo chroot '${ROOTFSDIR}' systemctl preset-all --preset-mode="enable-only"
+}
+
 do_rootfs_postprocess[vardeps] = "${ROOTFS_POSTPROCESS_COMMAND}"
 do_rootfs_postprocess[network] = "${TASK_USE_SUDO}"
 python do_rootfs_postprocess() {
-- 
2.39.5

-- 
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/20250917103506.295010-3-alexander.heinisch%40siemens.com.

Re: [PATCH v3 0/2] Enable systemd units based on systemd presets

['Heinisch, Alexander' via isar-users]

Any updates on this patchset? Seems it got lost.

BR Alexander

On Wed, 2025-09-17 at 12:35 +0200, alexander.heinisch@siemens.com
wrote:
> From: Alexander Heinisch <alexander.heinisch@siemens.com>
> 
> By default population of the presets is automatically done by systemd
> on first-boot.
> 
> There were several issues with that:
> 
> 1. The rootfs we get as a build artifact does not reflect the actual
> system running in the field.
> 
> 2. For setups without writeable /etc this fails. With that addition
> it happens already at build time.
> 
> Note: Additional services are enabled only. Services already enabled
> during the package installation won't be changed.
> 
> Opt-out: `ROOTFS_FEATURES:remove = "populate-systemd-preset"`
> 
> Changes since v2:
>   - Enable postprocessing to populate systemd presets via
> ROOTFS_FEATURES
> 
> Alexander Heinisch (2):
>   Minor cleanup: Refactor ROOTFS_FEATURES assignment for image to
> make
>     upcoming changes more easily trackable.
>   classes/image-postproc: Enable systemd units based on systemd
> presets
> 
>  RECIPE-API-CHANGELOG.md     | 18 ++++++++++++++++++
>  meta/classes/image.bbclass  | 10 +++++++++-
>  meta/classes/rootfs.bbclass |  6 ++++++
>  3 files changed, 33 insertions(+), 1 deletion(-)
> 
> --
> 2.39.5
> 

-- 
Alexander Heinisch
Siemens AG
www.siemens.com

-- 
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/d79451250668d72535e7869e175878aff890eb14.camel%40siemens.com.

Re: [PATCH v3 0/2] Enable systemd units based on systemd presets

['cedric.hombourger@siemens.com' via isar-users]

On Tue, 2025-10-21 at 11:21 +0000, 'Heinisch, Alexander' via isar-users
wrote:
> Any updates on this patchset? Seems it got lost.
> 
> BR Alexander
> 
> On Wed, 2025-09-17 at 12:35 +0200, alexander.heinisch@siemens.com
> wrote:
> > From: Alexander Heinisch <alexander.heinisch@siemens.com>
> > 
> > By default population of the presets is automatically done by
> > systemd
> > on first-boot.
> > 
> > There were several issues with that:
> > 
> > 1. The rootfs we get as a build artifact does not reflect the
> > actual
> > system running in the field.
> > 
> > 2. For setups without writeable /etc this fails. With that addition
> > it happens already at build time.
> > 
> > Note: Additional services are enabled only. Services already
> > enabled
> > during the package installation won't be changed.
> > 
> > Opt-out: `ROOTFS_FEATURES:remove = "populate-systemd-preset"`

Since we have ROOTFS_FEATURES in LIST_VARIABLES (see bitbake.conf), a
more elegant way to opt is IMO:

ROOTFS_FEATURES:append = " ~populate-systemd-preset" # (or
alternatively " -populate-systemd-preset")

> > 
> > Changes since v2:
> >   - Enable postprocessing to populate systemd presets via
> > ROOTFS_FEATURES
> > 
> > Alexander Heinisch (2):
> >   Minor cleanup: Refactor ROOTFS_FEATURES assignment for image to
> > make
> >     upcoming changes more easily trackable.
> >   classes/image-postproc: Enable systemd units based on systemd
> > presets
> > 
> >  RECIPE-API-CHANGELOG.md     | 18 ++++++++++++++++++
> >  meta/classes/image.bbclass  | 10 +++++++++-
> >  meta/classes/rootfs.bbclass |  6 ++++++
> >  3 files changed, 33 insertions(+), 1 deletion(-)
> > 
> > --
> > 2.39.5
> > 
> 
> -- 
> Alexander Heinisch
> Siemens AG
> www.siemens.com
> 

-- 
Cedric Hombourger
Siemens AG
www.siemens.com

-- 
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/c2106cd12b6d5781b86276ce5ba7b4679ef3e1a8.camel%40siemens.com.

Re: [PATCH v3 0/2] Enable systemd units based on systemd presets

[Zhihang Wei]

Hi,

when testing this patch on CI, the following test failed:
- citest.py:DevTest.test_dev

Error log from avocado:
2025-10-22 17:42:41,056 cibuilder        L0322 ERROR| ERROR: 
mc:qemuarm-bookworm:isar-image-base-sdk-1.0-r0 do_rootfs_postprocess: 
ExecutionError('/work/build/tmp/work/debian-bookworm-armhf/isar-image-base-sdk-qemuarm/1.0-r0/temp/run.image_postprocess_populate_systemd_preset.287029', 
127, None, None)
2025-10-22 17:42:41,059 cibuilder        L0322 ERROR| ERROR: Logfile of 
failure stored in: 
/work/build/tmp/work/debian-bookworm-armhf/isar-image-base-sdk-qemuarm/1.0-r0/temp/log.do_rootfs_postprocess.287029
......
2025-10-22 17:42:41,983 cibuilder        L0322 ERROR| ERROR: Task 
(mc:qemuarm-bookworm:virtual:sdk:/work/meta-isar/recipes-core/images/isar-image-base.bb:do_rootfs_postprocess) 
failed with exit code '1'
......

Error log from log.do_rootfs_postprocess.287029:
......
DEBUG: Executing shell function rootfs_cleanup_isar_apt
DEBUG: Shell function rootfs_cleanup_isar_apt finished
DEBUG: Executing shell function image_postprocess_populate_systemd_preset
chroot: failed to run command ‘systemctl’: No such file or directory
WARNING: exit code 127 from a shell command.
DEBUG: Executing shell function rootfs_do_umounts
DEBUG: Shell function rootfs_do_umounts finished
DEBUG: Python function do_rootfs_postprocess finished

You can redo the test on your machine using avocado:
1. Have a clean clone of isar, checkout to branch next and apply your 
patches:
$ git clone -b next https://github.com/ilbers/isar.git
$ cd isar
$ git am /path-to/0001-my-contribution-to-isar.patch
2.Run kas shell, setup CI prerequisites (avocado, qemu) and cleanup:
$ ./kas/kas-container shell kas/isar.yaml --command \
     "rm -rf /work/build/conf && /work/scripts/ci_setup.sh"
3.Run the failed test:
$ cd /work/testsuite
$ avocado run citest.py:DevTest.test_dev --max-parallel-tasks=1

On my machine (CPU: AMD 7840U, 32G RAM) it took around 16 minutes for the
error to show up.

Best regards,
Zhihang

On 10/21/25 13:21, 'Heinisch, Alexander' via isar-users wrote:
> Any updates on this patchset? Seems it got lost.
>
> BR Alexander
>
> On Wed, 2025-09-17 at 12:35 +0200, alexander.heinisch@siemens.com
> wrote:
>> From: Alexander Heinisch <alexander.heinisch@siemens.com>
>>
>> By default population of the presets is automatically done by systemd
>> on first-boot.
>>
>> There were several issues with that:
>>
>> 1. The rootfs we get as a build artifact does not reflect the actual
>> system running in the field.
>>
>> 2. For setups without writeable /etc this fails. With that addition
>> it happens already at build time.
>>
>> Note: Additional services are enabled only. Services already enabled
>> during the package installation won't be changed.
>>
>> Opt-out: `ROOTFS_FEATURES:remove = "populate-systemd-preset"`
>>
>> Changes since v2:
>>    - Enable postprocessing to populate systemd presets via
>> ROOTFS_FEATURES
>>
>> Alexander Heinisch (2):
>>    Minor cleanup: Refactor ROOTFS_FEATURES assignment for image to
>> make
>>      upcoming changes more easily trackable.
>>    classes/image-postproc: Enable systemd units based on systemd
>> presets
>>
>>   RECIPE-API-CHANGELOG.md     | 18 ++++++++++++++++++
>>   meta/classes/image.bbclass  | 10 +++++++++-
>>   meta/classes/rootfs.bbclass |  6 ++++++
>>   3 files changed, 33 insertions(+), 1 deletion(-)
>>
>> --
>> 2.39.5
>>

-- 
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/1dfe2925-928e-4cde-a4ae-42c61a121603%40ilbers.de.

Re: [PATCH v3 0/2] Enable systemd units based on systemd presets

['Heinisch, Alexander' via isar-users]

On Tue, 2025-10-21 at 16:07 +0000, Hombourger, Cedric (FT FDS CES LX)
wrote:
> On Tue, 2025-10-21 at 11:21 +0000, 'Heinisch, Alexander' via isar-
> users
> wrote:
> > Any updates on this patchset? Seems it got lost.
> > 
> > BR Alexander
> > 
> > On Wed, 2025-09-17 at 12:35 +0200, alexander.heinisch@siemens.com
> > wrote:
> > > From: Alexander Heinisch <alexander.heinisch@siemens.com>
> > > 
> > > By default population of the presets is automatically done by
> > > systemd
> > > on first-boot.
> > > 
> > > There were several issues with that:
> > > 
> > > 1. The rootfs we get as a build artifact does not reflect the
> > > actual
> > > system running in the field.
> > > 
> > > 2. For setups without writeable /etc this fails. With that
> > > addition
> > > it happens already at build time.
> > > 
> > > Note: Additional services are enabled only. Services already
> > > enabled
> > > during the package installation won't be changed.
> > > 
> > > Opt-out: `ROOTFS_FEATURES:remove = "populate-systemd-preset"`
> 
> Since we have ROOTFS_FEATURES in LIST_VARIABLES (see bitbake.conf), a
> more elegant way to opt is IMO:
> 
> ROOTFS_FEATURES:append = " ~populate-systemd-preset" # (or
> alternatively " -populate-systemd-preset")

Thx, for pointing that out, but I decided, to keep opting out in v4 as
is. Imo ":remove" is easier to read than "-" inbetween and also more
intuitive than "~".

BR Alexander

> 
> > > 
> > > Changes since v2:
> > >   - Enable postprocessing to populate systemd presets via
> > > ROOTFS_FEATURES
> > > 
> > > Alexander Heinisch (2):
> > >   Minor cleanup: Refactor ROOTFS_FEATURES assignment for image to
> > > make
> > >     upcoming changes more easily trackable.
> > >   classes/image-postproc: Enable systemd units based on systemd
> > > presets
> > > 
> > >  RECIPE-API-CHANGELOG.md     | 18 ++++++++++++++++++
> > >  meta/classes/image.bbclass  | 10 +++++++++-
> > >  meta/classes/rootfs.bbclass |  6 ++++++
> > >  3 files changed, 33 insertions(+), 1 deletion(-)
> > > 
> > > --
> > > 2.39.5
> > > 
> > 
> > -- 
> > Alexander Heinisch
> > Siemens AG
> > www.siemens.com
> > 
> 

-- 
Alexander Heinisch
Siemens AG
www.siemens.com

-- 
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/9e2d98f243ee30b988fd9368e6387cbe2b710219.camel%40siemens.com.