Foundations Team Updates - Thursday 13 October 2022

Last week’s update can be found here: Foundations Team Updates - Thursday 06 October 2022

  • Bug triage
  • APT solver bug analysis (nvidia) - LP: #1992513
  • Small grub patch
  • Dismayed at the Ubuntu Pro news in the apt hook, and talking to people about improving that




  • email & backlog
  • Linux skills interview
  • M&G interview
  • Take-home test review
  • DX interview

Testing images for RISC-V board

  • Initrd for LicheeRV and Nezha is too big
    @alexghiti indentified missing stripping of modules as a root cause
    I created a merge request for livecd-rootfs to use MODULES=list for initramfs-tools

PolarFire Icicle Kit

  • Microchip updated their HSS firmware. Due to a changed memory map U-Boot and LInux have to be updated.
  • Packaged U-Boot with two patches for Microchip LP #1992776
  • Currently testing kernel change on a upstream 5.19.14 kernel
  • subiquity
    • write isotracker test case covering install automation from netboot
    • File LP: #1992533 regarding an error around ubuntu-drivers apt handling and unmounting
    • PR: 1449 (LP: #1992531) - above ubuntu-drivers unmount does trigger a crash later, make that non-fatal
    • PR: 1450 (LP: #1992541) - fix issue with partition layouts being bad during guided LVM for certain disk sizes
    • PR: 1451 - merge these fixes for kinetic
    • Roadmap planning
  • probert
    • PR: 123 (LP: #1992115) - while investigating a crash when netinstalling Kinetic, I found that curtin was trying to match partitions to a disk that did not exist. This was due to probert not outputting the disk info because it was really a loopback device. Update probert to also not output partitions on a loopback device.
  • candidate interview


  • Reviewed MPs from ddstreet for LP 1991829
  • systemd upload to provide a stop gap fix to LP 1992178
  • Investigated LP 1990742 “screen not consistently locking when lid is closed”



  • Fixed LP 1992206 " bash: non existent locale crashes bash"
    • Came across when debugging TEST-69-SHUTDOWN failures in systemd autopkgtest. This crash turned out to be the cause.
1 Like
  • Uploaded glibc 2.36-0ubuntu4 to Kinetic on Friday
    • Working on getting it through -proposed
  • Fixed ruby-ragged against libgit 1.3.2
  • Looked into tzdata 2022e with Benjamin
  • Multiple interviews & Take-Home Tests


  • opened PR 1446 to replace deprecated calls to asyncio.get_event_loop. Integrations tests fail on focal and this requires more investigation.
  • PR reviews

+1 maintenance

  • LP: #1992331 fixed FTBFS for libcamera on ppc64el. Thanks @schopin for sponsoring the patch in Ubuntu! Sponsored in Debian by @bdrung, thanks! Also forwarded the patch upstream.
  • LP: #1992431 - filed removal bug for aseba binary packages on s390x and ppc64el. The last uploads reduced the list of architectures supported.
  • LP: #1992672 - uploaded debdiff to fix src:aft FTBFS. The failing package is in -proposed and a version that builds exists in the -release pocket. No need to sponsor for kinetic AFAICT. Forwarded the patch to Debian.
  • LP: #1992769 - filed bug and investigated slow install of haskell-hoogle on armhf making autopkgtest time out on kinetic. The main binary package takes forever to install. Adding to long_tests would not help AFAIK.
  • Added libfirefox-marionette-perl to never_run list for jammy. The package is incompatible with firefox shipped as a snap and autopkgtest times out.
  • LP: #1992816 investigating FTBFS on armhf of src:ball. On armhf only, the package build-depends on both libgles-dev and libglew-dev. Those two libraries have conflicting type declarations. Seems to be a known issue without a clear path forward.
  • Investigating asmjit build failure on s390x. Seems to be an endianness issue but upstream only supports aarch64, x86 and x86_64 backends. Considering changing Architecture: any to a more reduced list.