[Buildroot] [autobuild.buildroot.net] Daily results for 2021-08-27

[Edgar Bonet]

Hello!

This morning, the autobuilder wrote:
> Detail of defconfig failures for master [...]

I took a look at these failures: some of them are known issues.

>    acmesystems_aria_g25_128mb | https://gitlab.com/buildroot.org/buildroot/-/jobs/1536008767
>    acmesystems_aria_g25_256mb | https://gitlab.com/buildroot.org/buildroot/-/jobs/1536008768
> acmesystems_arietta_g25_128mb | https://gitlab.com/buildroot.org/buildroot/-/jobs/1536008770
> acmesystems_arietta_g25_256mb | https://gitlab.com/buildroot.org/buildroot/-/jobs/1536008771

These have been fixed yesterday in commit
eaece8645b6fb6cbaef6ae8ac6aeacaaa7ced20c, but the tests above ran before
the fix was committed.

>     microchip_sama5d2_icp_mmc | https://gitlab.com/buildroot.org/buildroot/-/jobs/1536008997
> microchip_sama5d2_icp_mmc_dev | https://gitlab.com/buildroot.org/buildroot/-/jobs/1536008999

This is a bug in AT91Bootstrap[1] that was fixed, for the defconfigs
using AT91Bootstrap 3, by upgrading to v3.10.3. C.f. commits
0744eef34d836bc8ef086ed657377bc15ed753a9,
0614f435a09e99c33f418a646eaa7647944c2f4c and
8ac2e5e46fd38b2da8cc292587e91d7b3d4710bf.

The two defconfigs above use AT91Bootstrap v4.0.0-rc2, which has the
same bug (in the 4.x branch, it was fixed between v4.0.0-rc3 and
v4.0.0-rc4). These failures can be fixed by upgrading to the final
release of AT91Bootstrap, namely v4.0.0.

>   microchip_sam9x60ek_mmc_dev | https://gitlab.com/buildroot.org/buildroot/-/jobs/1536008989

This fails with the message:

    mkfs.ext4: Could not allocate block in ext2 filesystem while
    populating file system
    Maybe you need to increase the filesystem size
    (BR2_TARGET_ROOTFS_EXT2_SIZE)

And indeed, adding

    BR2_TARGET_ROOTFS_EXT2_SIZE="100M"

fixes the failure.

I could submit patches, but I cannot test the generated images.

Regards,

Edgar Bonet.

[1] https://github.com/linux4sam/at91bootstrap/issues/127