Re: [OE-core] [PATCH v3 00/11] systemd based initrd and modular kernel support

[Mikko Rapeli <mikko.rapeli@linaro.org>]

Hi,

On Mon, Apr 07, 2025 at 09:53:45AM +0200, Mathieu Dubois-Briand wrote:
> On Fri Apr 4, 2025 at 6:29 PM CEST, Mikko Rapeli via lists.openembedded.org wrote:
> > systemd based initrd supports more security, encryption etc features
> > than custom shell script ones but oe-core lacks support for it.
> > Convert core-image-initramfs-boot to create systemd based initrd
> > if "systemd-initramfs" is set in DISTRO_FEATURES. Includes a test
> > for this initrd to mount rootfs via Unified Kernel Image uki.py.
> >
> > Sadly the boot time is really slow, e.g. over 30 seconds with
> > systemd based initrd. One of the major reasons is the large
> > amount of kernel modules built into the linux-yocto kernel and
> > processing of their udev events in userspace. Thus add
> > kernel-initrd-modules meta package to easily install
> > subset of all kernel modules to initrd. The subset
> > supports "mount rootfs from any block device" but
> > doesn't include graphics, UBS etc support. Without this
> > generated meta package, initrd recipe needs to manually define
> > which exact kernel modules to include which breaks if
> > those modules are built into the kernel. Thus the initrd
> > becomes machine and kernel config specific and hard to maintain.
> > With this meta package a generic initrd can be created.
> >
> > Separate changes to linux-yocto compile a lot more drivers
> > as modules but they depend on these initrd changes.
> >
> 
> Hi Mikko,
> 
> Thanks for your patch.
> 
> I believe one of these wic related patches is responsible of some
> oe-selftest failures we can see on the autobuilder:
> 
> 2025-04-06 15:51:49,317 - oe-selftest - INFO - wic.Wic.test_permissions (subunit.RemotedTestCase)
> 2025-04-06 15:51:49,318 - oe-selftest - INFO -  ... FAIL
> 
> ...
> 
> ERROR: When reparsing /tmp/selftest-fetchsrayuxoq/test.bb:do_checkuri, the basehash value changed from 84b9e23a83a764e7a5d8a96ee523199e2b1c604f4280a45030e248de0ff52962 to 4123da244f465feffecb2b4fd2b132b64851981ec2fb70bf9445df8b0ea6a69e. The metadata is not deterministic and this needs to be fixed.
> ERROR: The following commands may help:
> ERROR: $ bitbake test -cdo_checkuri -Snone
> ERROR: Then:
> ERROR: $ bitbake test -cdo_checkuri -Sprintdiff
> 
> ERROR: When reparsing /tmp/selftest-fetchsrayuxoq/test.bb:do_fetch, the basehash value changed from 6d55d6743729b615749dc016857d7e5f9c884a8b92f0d57f68e743f4910333d3 to fa7d6ea0563b60362418d2a6c5e41a6684183d01ec578b65abd0c64a9cfba7cf. The metadata is not deterministic and this needs to be fixed.
> ERROR: The following commands may help:
> ERROR: $ bitbake test -cdo_fetch -Snone
> ERROR: Then:
> ERROR: $ bitbake test -cdo_fetch -Sprintdiff
> 
> ERROR: When reparsing /tmp/selftest-fetchsrayuxoq/test.bb:do_unpack, the basehash value changed from ae9d94aafc721d891f8754df82ed90da8191357178f3db64221a49d5c1de3bcd to 1ac8f87a71db151b8eb2ae8d13f9a6ea1a0c884ef98555d61d2c2fe07a476a8a. The metadata is not deterministic and this needs to be fixed.
> ERROR: The following commands may help:
> ERROR: $ bitbake test -cdo_unpack -Snone
> ERROR: Then:
> ERROR: $ bitbake test -cdo_unpack -Sprintdiff
> 
> ...
> 
> https://autobuilder.yoctoproject.org/valkyrie/#/builders/23/builds/1391
> 
> Can you have a look at this please?

I can have a look but how can these patches change the basehash'es when testing?

I've seen some of the wic selftets failing on x86_64 build machine with zfs
but changes to fix this was rejected multiple times.

Then I've seen some of these failing with odd pseudo failures on aarch64
build machine.

With my patches the end result was the same.

I ran wic and uki selftests on x86_64 and aarch64 build machines for targets qemuarm64,
qemuarm and genericx86-64.

The wic patches have actually nothing to do with this series except I ran into bugs
in wic which resulted in unbootable images, e.g. the UEFI bootloader binaries
missing from ESP partition. These were impacting some builds but not all
which was really troubling.

Cheers,

-Mikko