canonical-ubuntu-qa team mailing list archive

[Launchpad Bug Tracker <2003554@xxxxxxxxxxxxxxxxxx>]

** Merge proposal linked:
   https://code.launchpad.net/~agherzan/+git/adt-matrix-hints/+merge/437946

-- 
You received this bug notification because you are a member of
Canonical's Ubuntu QA, which is subscribed to autopkgtest in Ubuntu.
https://bugs.launchpad.net/bugs/2003554

Title:
  Second test run failing to install dependencies

Status in autopkgtest package in Ubuntu:
  Confirmed

Bug description:
  The kernel dep8 tests define two tests to be run

  Tests: rebuild
  Depends: @builddeps@, fakeroot
  Restrictions: allow-stderr, skippable

  Tests: ubuntu-regression-suite
  Depends: build-essential, gcc-multilib [amd64 armhf i386], gdb, git, python2 | python, bzr
  Restrictions: allow-stderr, isolation-machine, breaks-testbed, skippable

  Since about last year December, we see the ubuntu-regression-suite
  failing because the generated dependency package (autopkgtest-satdep)
  gets removed as a result of the resolver (apt install -f) running. In
  the example case this is because build-essential is considered
  uninstallable (however that package got installed for the rebuild run.

  Following steps to resolve this uncover a second bug. It takes the
  dependency string and just deletes the comma separators and uses that
  string as input for "apt get install". However the included
  alternative is treated as a shell pipe, resulting in "python not
  found" in the log.

  https://autopkgtest.ubuntu.com/results/autopkgtest-
  jammy/jammy/amd64/l/linux/20230116_105942_58155@/log.gz

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/autopkgtest/+bug/2003554/+subscriptions