1. 18 May, 2015 1 commit
  2. 10 Apr, 2014 1 commit
  3. 03 Feb, 2014 1 commit
  4. 12 Nov, 2013 6 commits
  5. 07 Nov, 2013 1 commit
    • Ray Strode's avatar
      kerberos: don't crash if keyring credentials disappear · cb1d1b1e
      Ray Strode authored
      It's possible to make gnome-online-accounts crash by:
      
      1) creating a kerberos account in control-center
      2) deleting the stored credentials from gnome keyring using seahorse
      3) running kdestroy in the terminal to make the sign in button show up
      in the control-center panel
      4) clicking sign in
      
      This is because the provider makes the assumption that there will always
      be stored credentials when signing in interactively, which is incorrect
      in the above scenario (because of step 2).
      
      This commit hardens the get_ticket_sync function against credentials
      disappearing.
      
      https://bugzilla.gnome.org/show_bug.cgi?id=711572
      cb1d1b1e
  6. 06 Nov, 2013 3 commits
  7. 01 Nov, 2013 3 commits
  8. 30 Oct, 2013 4 commits
  9. 08 Oct, 2013 1 commit
  10. 07 Oct, 2013 3 commits
  11. 01 Oct, 2013 3 commits
  12. 30 Sep, 2013 2 commits
    • Ray Strode's avatar
      alarm: Check proper cancellable from timer source ready callback · f3052079
      Ray Strode authored
      on_timer_source_ready is called in two cases:
      
      1) When the timer fires
      2) When the input stream cancellable is cancelled
      
      We attempt to check for this latter case up front at the start of the
      function and react appropriately. We do this by checking if
      self->priv->cancellable is cancelled.
      
      Unforunately, checking self->priv->cancellable isn't always right,
      because self->priv->cancellable isn't always the cancellable associated
      with the input stream.  They can get out of sync, when for instance, the
      alarm expiration time is changed. In this case the old input stream
      cancellable will get cancelled, and self->priv->cancellable will get set
      to the new cancellable.
      
      This commit makes sure the on_timer_source_ready callback always checks
      the cancellable associated with the input stream.  It accomplishes this
      by allocating a GTask and using that as the callback data instead of
      the alarm object directly.
      
      https://bugzilla.gnome.org/show_bug.cgi?id=708975
      f3052079
    • Ray Strode's avatar
      alarm: Print warning if timerfd I/O fails · 8f8f8ccb
      Ray Strode authored
      Reading from a timerfd shouldn't fail, but if it
      does we should print what's going on.
      
      https://bugzilla.gnome.org/show_bug.cgi?id=708975
      8f8f8ccb
  13. 29 Sep, 2013 1 commit
  14. 20 Sep, 2013 2 commits
  15. 16 Sep, 2013 1 commit
  16. 15 Sep, 2013 1 commit
  17. 30 Aug, 2013 1 commit
  18. 29 Aug, 2013 1 commit
    • Debarshi Ray's avatar
      kerberos: Don't crash when an identity could not be looked up · 92930750
      Debarshi Ray authored
      Set an error in look_up_identity instead of only setting the op_res to
      NULL, so that ensure_credentials_sync catches it.
      
      Hold a reference to the GoaKerberosIdentity object when queueing a
      notify. It is safer than trying to disconnect the source in dispose. We
      were not checking for all the idle sources, and even if we did, it can
      be racy if the object was getting disposed from a different thread.
      
      Fixes: https://bugzilla.gnome.org/699491
      92930750
  19. 28 Aug, 2013 1 commit
    • Alban Browaeys's avatar
      goaidentity: Fix deadlock in goaalarm on_cancelled · 88eecb0f
      Alban Browaeys authored
      Calling g_cancellable_disconnect in a cancelled handler causes a
      deadlock. This happens here as in clear_scheduled_timer_wakeups we
      destroy the cancellable source that emitted the cancelled signal.
      (goaalarm: on_cancelled-> clear_scheduled_wakeups
      -> clear_scheduled_timer_wakeups -> g_source_destroy
      -> g_source_destroy_internal -> g_child_source_remove_internal
      -> g_source_unref_internal -> cancellable_source_finalize
      -> g_cancellable_disconnect).
      
      Thus based on Matthew Barnes , evolution commit
      c155b170bcf644ab2cdf7b11afdda0d9c100772e :
      EPhotoCache: Fix deadlock when cancelling subtasks.
      
      That is, we move the cancelled task into the main thread in an idle
      source, to prevent the disconnect from happening inside the cancelled
      handler.
      
      Fixes: https://bugzilla.gnome.org/705395
      88eecb0f
  20. 15 Aug, 2013 1 commit
  21. 18 Jun, 2013 2 commits