Hi Steve Langasek, thanks for taking a look at the SRU.

> Is that not what this means, or is mqueue access actually denied by
> default and this refers only to how an unqualified 'mqueue' rule is
> interpreted?

Correct, this only refers to how an unqualified 'mqueue' rule is
interpreted.

> In that case how does introducing mqueue support in apparmor benefit
> users of jammy?

Users of jammy will now have the ability to mediate message queues in
their profile if they want, but they will have to opt-in. There is more
than one way to accomplish this, but they can for example add 'abi
<kernel>,' to their profile when using a kernel that provides mqueue
mediation. That means that older policies that were developed when
mqueue mediation was not available will not be broken.

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to apparmor in Ubuntu.
https://bugs.launchpad.net/bugs/1994146

Title:
  [SRU] apparmor - Focal, Jammy

Status in apparmor package in Ubuntu:
  Confirmed
Status in apparmor source package in Focal:
  In Progress
Status in apparmor source package in Jammy:
  Incomplete

Bug description:
  [ Impact ]

  This is a SRU proposal for apparmor in Focal and Jammy.
  For focal, we want to SRU fixes for Bug 1964636 which introduces the
  capability upstream patches. We are also fixing Bug 1728130 and
  Bug 1993353 which are introducing full backport of abi from
  apparmor-3.0 and support for POSIX message queue rules, which are both
  a request from Honeywell.

  Note that specifically for message queue rules, we are overriding the
  abi behavior.
  Message queue mediation is not a part of the 2.13 abi we are
  pinning. Honeywell has a kernel that has message queue mediation,
  but their policy does not contain an abi specified, so when we pin the
  abi for a kernel that does not mediate message queue, it will break
  Honeywell's AppArmor policies. So we are making an exception: when abi
  is not specified in the policy, and the policy contain mqueue rules,
  we are enforcing mqueue rules. When the policy does not contain mqueue
  rules, then they are not being enforced. This is so we do not break
  Honeywell policies and we also are not breaking policies that were
  developed when there was no mqueue or abi support.

  For jammy, we are SRUing fixes for Bug 1993353 which adds message
  queue rules support. 

  
  [ Test Plan ]

  This has been extensively tested by using QA Regression Tests[1] for
  AppArmor. All tests have passed and demonstrated AppArmor to be
  working as expected. We are also adding regression tests for message
  queue rules[2] which guarantees it is working as expected.

  [1] 
https://git.launchpad.net/qa-regression-testing/tree/scripts/test-apparmor.py
  [2] https://gitlab.com/apparmor/apparmor/-/merge_requests/858

  [ Where problems could occur ]

  The message queue rules support could cause issues for AppArmor
  policies that were developed before there was support for mqueues,
  that's why we are also backporting abi support and pinning the abi on
  parser.conf on focal. Jammy already has the abi pinned for a kernel
  that does not have support for mqueue mediation.

  [ Other Info ]

  The patches for both focal and jammy can be found at:
  https://launchpad.net/~georgiag/+archive/ubuntu/mqueue-sru/

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/apparmor/+bug/1994146/+subscriptions


-- 
Mailing list: https://launchpad.net/~touch-packages
Post to     : touch-packages@lists.launchpad.net
Unsubscribe : https://launchpad.net/~touch-packages
More help   : https://help.launchpad.net/ListHelp

Reply via email to