1. 11 Jul, 2020 1 commit
  2. 24 Jun, 2020 2 commits
    • sajolida's avatar
      Clarify beta state · 709d6b3e
      sajolida authored
      It will make it easier to migrate to full persistence of the Welcome
      Screen. See #15635#note_151069
      709d6b3e
    • sajolida's avatar
      Simplify · 41e62c70
      sajolida authored
      For the other features, we are not specifying in their name whether they
      relate to settings, data, files, etc. We rely on the description of the
      feature to clarify this if needed.
      41e62c70
  3. 19 Jun, 2020 1 commit
  4. 18 Jun, 2020 2 commits
  5. 23 May, 2020 5 commits
  6. 18 May, 2020 3 commits
  7. 26 Apr, 2020 2 commits
    • Cyril 'kibi' Brulebois's avatar
      IUK creation: use a lock file around mksquashfs calls · eb56115f
      Cyril 'kibi' Brulebois authored
      Let callers like wrap_tails_create_iuks spawn as many tails-create-iuk
      jobs as they wish to parallelize the I/Os (rsync), while still ensuring
      CPU-intensive tasks (mksquashfs, twice) are run sequentially.
      
      The caller gets to clean up the lock, but doesn't need to create it in
      the first place (even if doing so would be preferable, to ensure
      ownership and ability to remove it afterwards).
      eb56115f
    • Cyril 'kibi' Brulebois's avatar
      IUK creation: instrument lengthy operations · df8e1f35
      Cyril 'kibi' Brulebois authored
      Loop-mounting costs seem to be very close to 0, and attribute-related
      fixups are negligible compared to the rsync/mksquashfs operations, so
      concentrate on those only.
      df8e1f35
  8. 05 Apr, 2020 2 commits
  9. 28 Mar, 2020 1 commit
  10. 21 Mar, 2020 3 commits
  11. 20 Mar, 2020 3 commits
  12. 22 Feb, 2020 1 commit
  13. 21 Feb, 2020 2 commits
  14. 19 Jan, 2020 2 commits
  15. 13 Jan, 2020 1 commit
    • anonym's avatar
      IUK: various fixes. · c37af6ec
      anonym authored
      When doing the "Prepare upgrade-description files" section for Tails
      4.2.2 I encountered multiple issues, which this commit fixes. Many
      thanks to kibi for finding me the `canonpath` method!
      c37af6ec
  16. 11 Jan, 2020 3 commits
    • intrigeri's avatar
      Upgrader: don't assume that all offered upgrades are relevant (refs: #17425) · f2a786f1
      intrigeri authored
      Before this commit, the Upgrader blindly assumed that _any_ upgrade path that's
      proposed for the initially installed version actually contains an upgrade to
      a version that's strictly newer than the currently running one.
      
      This assumption used to hold in the v1 scheme, but in the v2 scheme:
      
       - When running version == initially installed version, i.e. when one runs
         a Tails that was never upgraded, this assumption is true.
      
       - Else, this assumption is wrong.
      
      As a consequence, any Tails >= 4.2 that hasn't this fix, and that ever got an
      incremental upgrade applied, will keep proposing an upgrade even if it's
      up-to-date already.
      
      To fix this, let's filter the proposed upgrade paths to ignore those that would
      "upgrade" to a lower-or-equal version than the currently running one.
      f2a786f1
    • intrigeri's avatar
      Upgrader test suite: add steps to reproduce bug (refs: #17425) · f122cc07
      intrigeri authored
      These steps are currently failing, as expected.
      f122cc07
    • intrigeri's avatar
      IUK generation: don't make all files in the SquashFS diff owned by root (refs: #17422) · 90d6dc1e
      intrigeri authored
      This fixes a regression introduced while implementing the IUK v2
      format (iuk.git:b578872ba2167da7dbc106a9d20c88e523ab73b3).
      
      We need to pass -all-root for the 1st-level SquashFS (the IUK itself)
      but we need to *not* pass it for the SquashFS diff that's included
      in the IUK, otherwise we're going to break systems when upgrading them.
      
      Thanks kibi for independently doing the same root cause analysis,
      and for testing that dropping -all-root indeed fixes the problem!
      90d6dc1e
  17. 31 Dec, 2019 3 commits
  18. 27 Dec, 2019 2 commits
  19. 26 Dec, 2019 1 commit
    • intrigeri's avatar
      IUK creation: delete temporary directory on success (refs: #15287) · 1501f638
      intrigeri authored
      tails-create-iuk previously left temporary files behind. In particular, even on
      success, the "squashfs_src" directory would remain on the filesystem.
      
      It's no big deal in general, but our build_IUKs Jenkins job runs
      tails-create-iuk as root with sudo, so the leftover temporary files are owned by
      root, and then the "workspace-cleanup" step (itself run as the "jenkins" user)
      can't delete them, which causes 2 problems:
      
       - These temporary files would endlessly accumulate on isobuilders,
         without any mechanism to clean them up automatically.
      
       - Confusing output in the Jenkins console:
      
            cannot remove path when cwd is /var/lib/jenkins/workspace/build_IUKs/j4gMvgS3wH for /var/lib/jenkins/workspace/build_IUKs/j4gMvgS3wH:  at /usr/share/perl/5.24/File/Temp.pm line 1583.
            Archiving artifacts
            [WS-CLEANUP] Deleting project workspace...
            Cannot delete workspace: null
            Option not to fail the build is turned on, so let's continue
      1501f638