Foundations Team Updates - Thursday 31 March 2022

Foundations
1

Hello everyone, the Foundations team will post their updates from the past week here.

Last week’s update is available here: Foundations Team Updates - Thursday 24 March 2022

1 Like
2

autopkgtest

  • bumped the number of amd64 workers in lcy01 for the autopkgtest env
  • discovered never_run in autopkgtest-package-configs added gfan, booth, debci
    and sabnzbdplus to it
  • modified the autopkgtest running page so package names link to their results
  • modified the results page to use denylisted instead of blacklisted

ISO Tracker

  • updated manual test cases for RISCV64 installations for @xypron
  • removed unleashed images from the Beta milestone

Ubuntu

  • rebuilt Ubuntu desktop ISOs
  • install testing of the Ubuntu desktop image
  • resolved an issue where daily builds were creating .img files instead of .iso
  • created a JammyUpgrades wiki page
  • pinged flavors about adding upgrade instructions to the JammyUpgrades page
  • special SRU review of cloud-init for the Server team
  • special SRU review of systemd
  • helped @enron report LP: #1966317
  • found a possible version of the same crash in the Error Tracker
  • reported a bug regarding KDE frontend for ubiquity not loading the release notes in a browser
  • reported a bug regarding KDE frontend for ubiquity showing incorrect disk utilization
  • sponsored a casper change fixing LP: #1892369
  • investigation into, testing and fixing of LP: #1966204
  • tested, uploaded the apport fix for LP: #1965830
  • found, tested, uploaded a fix for LP: #1966811 which was supposed to be fixed by the previous upload of update-notifier
  • modified bugbot to no longer move no package bugs filed by apport to xorg
  • submitted MP for ubuntu-archive-scripts to clean up vintage packages
  • submitted MP which modifies sru-review to do what was intended (who wrote this in the first place‽)
  • did a nochange rebuild of mypaint fixing LP: #1967163
  • retesting python3-defaults, livecd-rootfs autopkgtest “regressions”
  • reviews of FFe and UIFe bug reports
  • ISO testing
  • SRU reviews
1 Like
3
  • +1 Maintenance
    • SoftHSM v2 (the last remaining reason for openssl 1.1 to be in the archive)
    • golang-github-hhatto-gorst
    • pytest-mock related autopkgtest regressions
    • gst-omx: new version in Debian should be synced upon kk archive opening
  • Support more Raspberry Pi Hats
    • Packaged the rest of the unicorn hat family in a PPA
  • Support Piboot on UC22
    • Did lots of testing and merged the pi-gadget changes
    • Cherry picked changes to 22-armhf branch
    • Discovered that snapd is still broken with armhf and reported
    • Tested snapd changes related to armhf dtb location
    • Update pi-gadget to not include any u-boot files
    • Submitted IS ticket about failing pi-gadget builds with 403 errors
  • ubuntu-image
    • Investigated bug reported by Sergio. Turned out to be a snapd bug instead
    • Meetings and discussion on classic image redesign
    • Uploaded a piboot capable deb to a PPA for building UC22 images
  • General distro work
    • Sponsored python-uvicorn and pytest-mock for @enr0n
    • Troubleshot missing symlinks in apport upload I had done
  • Beta iso testing
  • Proposed migration:
    • kopanocore (just needed tests restarted)
    • pam vs stunnel and exim4 (just needed tests restarted)
1 Like
4
  • subiquity
    • PR: 1219 - Address feedback.
    • PR: 1241 - Pick up probert change below.
    • PR: 1245 - Fix a bug noted in review and improve integration tests around v2 API and adding partitions.
    • PR: 1246 - Fix regression in speed of API test
    • PR: 1248 - Drop the now obsolete free_for_partitions, as the same info can be obtained by the Gap objects.
    • PR: 1249 - Draft for calculation of the maximum size of a partition, to counter for the different locations where a gap can be.
    • Review PR: 1232
    • Move Feature Freeze things along for build for beta - LP: #1966186 help ping and merge, LP: #1965760 merge
  • probert
    • PR: 113 - In integration test of a different change, found that probert is not actually correctly running os-prober from the snap, fix that.
  • curtin
    • MP: 417829 - Open draft implementation of resize. Change definition of v2 storage config to allow a resize keyword. Can handle both resize up and down for ext{2,3,4}.
  • proposed-migration - request retest of nagios-plugins-contrib, but it seems to be afflicted by existing issues with bos01
1 Like
5
  • proposed-migration
    • stunnel4 vs openssl - flaky test on s390x. Retried, and tests passed.
    • swupdate vs openssl: Another flaky test. Retried, and waiting for the result.
    • systemd vs libselinux: Investigating the failing logind test. This may be worth re-trying, but there is not a noticeable history of this test failing.
    • systemd vs linux-meta-raspi: This has been failing (and retried) many times. I noticed that other autopkgtest regressions triggered by 5.15.0.1004.4 are failing the same way, so it seems this needs to be investigated in linux-meta-raspi further.
  • @jawn-smith sponsored python-uvicorn for me, and it has since migrated.
  • @jawn-smith sponsored pytest-mock for me. Working on getting this migrated.
  • Reported LP 1966317 to assist in investigation of impish->jammy upgrade crashes.
  • Started investigating LP 1966381 for systemd.
  • Looked into 1966203 since it was reported against systemd. It seems at the moment there is no need to change systemd-udevd behavior, and snapd will fix their udev rules.
  • systemd SRUs for focal and impish
    • investigating autopkgtest regressions. So far, nothing looks like a real regression, but is instead related to recent network/infrastructure issues. Continuing to investigate.
    • Did SRU verification for LP 1958284 and LP 1962038.
1 Like
6
  • short week after some annual leave
  • subiquity reviews
  • 20.10 planning and spec writing
1 Like
7
2 Likes
8

short week:

  • RISC-V kernel workaround for live-installer seems to fix observed live-installer stalls and work on proper fix
  • gpgme1.0 FTBFS fix ongoing
1 Like
9
  • Was off for half of the week
  • Subiquity:
    • Worked on new design for third-party drivers. Mostly finished. Answers mechanism to improve.
    • Investigating LP 1966672. First exposition to console_conf. Bug not yet reproduced at the moment.
    • Got third-party drivers in Jammy. Tests were successful with Nvidia GPU hardware & FFe approved.
    • Got ubuntu-pro removed from 22.04.0.
  • Investigated ltrace FTBFS on ppc64el. Goto causing jump over variable declaration.
  • pexpect from proposed migration - times out on arm64 while running performance test (generating large chunks) of random data.
1 Like
10

I’m sure I’m missing something but the cloud kept me busy so I didn’t manage to note down during the week :frowning:

  • Finished preparing and uploading the ppc64el rebuilds during +1
  • Bought the autopkgtest cloud back to a somewhat working state.
    • Filed tickets
    • restricted PPAs and upstream tests to bos02 as they failed on bos01
    • Doubled arm64 worker count on bos02
    • Various improvements to throughput
  • Half days on Monday (:partly_sunny: cycling fun) and Thursday (:cloud_with_rain: )
  • (private)
1 Like
11
  • Holiday last week
  • Tested Unicorn HAT HD packages, and rejigged a few bits
  • Beta image testing
  • Investigating boot times LTS->LTS
  • Worked on flash-kernel performance (LP: #1965129)
  • Worked on rshell packaging for pico support (not finished yet)
  • Reviewed mediatek firmware merge
  • Reviewed status of debhelper backport (LP: #1965758)

Off-topic:

  • snapped structa so go kick the tires if you have any massive JSON/YAML you want decoded
1 Like
12
1 Like