Here are some more thoughts from me. This took me a long time to write,
read, and rewrite. Consider this me brainstorming on pros and cons. I
like apparmor and am happy to see a push to confine more applications,
and thanks for offering a strategy for doing that.

> The number of policies in this package is very large. When no policy cache 
> exists (as on first 
> installation), building it can be very long. Even when a cache, loading all 
> policies is not 
> instantaneous.

The upgrade took double the time, and there were no changes. I just
repeated "dpkg -i" with the same package. I also noticed the package
doesn't use dh_apparmor, so no dh_apparmor snippets are in the rendered
postinst, maybe that is where some debhelper smarts are missing. I
didn't investigate further.

>  - It allows the AppArmor team to review carefully profiles, maintain them, 
> ensure their coherency and 
> how they interact with each other.
> - It allows to decouple profiles from the application maintainers, that don't 
> necessarily have the 
> necessary AppArmor knowledge.

But I think we would want coupling. Without it, the profile can evolve
in one direction, and the application in another, and the confinement
will break. You could have an old version of the profile installed and a
new version of the application package installed. Users could have
pinned an application package to a specific version. Users could want a
profile fix from bin:apparmor.d-N+1, but keep another profile shipped in
bin:apparmor.d-N for another application because in N+1 it broke their
use case.

>  - That also allows updating profiles without needing to update the
application package.

Conversely, you would be updating a package that ships 1500 profiles.
You would be fixing a bug for one profile, and could be introducing a
bug in another (bugs happen).

I think at the core I have two objections to this whole approach:

1) all profiles loaded even when not needed, leading to the problems in comment 
#6. You explained several optimizations, but to me the best optimization is to 
not load what is not needed :)
2) decoupling with the application: high risk of the profile being meant for 
one version of the app, but a later one has different requirements that do not 
match the profile anymore. This discrepancy looks easier and quicker to catch 
if the profile is together with the application. Tee risk of updates to this 
single package approach also seems much higher.

Now, you make a good point about package maintainers not necessarily
having the apparmor knowledge, or even a desire to confine their
application. Us suddenly injecting an apparmor profile into their
package is rude and disruptive. And we would also have potentially up to
1500 new delta pieces added to debian packages.

How can we crack this nut?

Have you guys thought of ways to still ship all profiles in a separate
binary package, but not load them unless they are needed? Unless the
application they are meant to confine is installed? Can we play some
tricks with triggers?

I guess similar problems and discussions were had in the past about the
kernel modules package (we have two binary packages for kernel modules
IIRC), and linux-firmware (which also installs a whole bunch of binary
blobs regardless if you have that hardware or not: you *could* have it
in the future). But none of these are loaded by default: they are just
files available on disk, in case they are needed.

Some other thoughts:
a) a promotion plan: what happens once a profile matures, and can be shipped 
with the application? What are the conditions? What packaging changes will be 
needed then? We will have to add careful breaks/replaces, following 
https://wiki.debian.org/PackageTransition to avoid conflicts like in comment #5
b) or is the plan to always ship the profile in the distro via bin:apparmor.d, 
to be available in case the application package is installed, and never ship it 
in the application package itself? Counting on the fact that the current 
installation times can be made faster and have it consume less memory?
c) testing plan: how can the profiles from src:apparmor.d be tested? We would 
have to have an autopkgtest in src:apparmor.d for package bin:FOO that would 
install *both* bin:apparmor.d and bin:FOO, and from your comments looks like 
that fails due to OOM, going back to the optimization problem.
d) what about more restricted systems like raspberry PIs, are they of scope for 
this package, at this stage?
e) What will SRUs look like for src:apparmor.d? How many profiles would you be 
updating in one go? How many applications would have to be tested separately?
f) What happens if I have a host spawning dozens of LXD containers, and all 
those containers install bin:apparmor.d? "Don't do it"? :)

I also understand this is following an upstream project, which has all
these profiles in a git repository/tarball, and having one source debian
package mimicking that makes sense. But even with optimizations, unless
they are really fantastic, I don't see right now what this will look
like in the long term.

Now, I'm not the final word on this. This just appeared on my radar for
sponsorship reasons, and I have a passion for application confinement,
having written some apparmor profiles in the recent past. I truly
welcome others to join the discussion, and have no objections to be
proven wrong.

-- 
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/2121409

Title:
  [FFE] add a new apparmor.d package containing several apparmor
  profiles

Status in apparmor package in Ubuntu:
  Triaged

Bug description:
  ## FFE ##

  This is a Feature Freeze Exception request for questing for the
  apparmor package and for a new source package called apparmor.d:

  I'd like to add a new source package called apparmor.d which contains
  over 1500 profiles from the upstream project apparmor.d [1]

  These profiles will be added in "complain" mode, which means that for
  a given action, if the profile rules do not grant permission the
  action will be allowed, but the violation will be logged with a tag of
  the access being ALLOWED. This is done because we want to test these
  profiles and enable others to test and add new rules to eventually
  improve the profiles.

  By adding these profiles in a new package which is not installed by
  default, regular users will not be affected. But users that would like
  to test and contribute to the profiles can install it.

  We want to add these profiles, even in complain mode, as a new package
  (and not part of the apparmor package) because labeling certain
  binaries could cause issues with existing policy, specially those that
  use "peer". Additionally, the large amount of profiles do take a while
  to compile by the parser in the first boot. After that, a cached
  version of the profiles can be loaded directly into the kernel by the
  parser which takes considerably less time. Note again that apparmor.d
  will not be installed by default, so this will only affect users that
  choose to install it.

  The benefits of this change is the ability to increase the amount of
  testing for these profiles, which will then enable us to eventually
  ship them in enforce mode. More profiles means more confined
  applications, which could lead to higher security. This is the first
  step towards that.

  This FFE also includes the apparmor package because we want to change
  the suggestion from the apparmor-profiles-extra package, which is no
  longer maintained and will be deprecated in the future, to the new
  apparmor.d.

  This is the PPA containing a built version of apparmor and apparmor.d:

  https://launchpad.net/~georgiag/+archive/ubuntu/apparmor.dinapparmor5/

  These are the installation logs:
  georgia@sec2-questing-amd64:~/qrt-test-apparmor$ sudo apt install apparmor.d
  The following packages were automatically installed and are no longer 
required:
    apg                      libllvm19               
linux-headers-6.15.0-3-generic  xbitmaps
    cpp-14                   libopengl0              
linux-modules-6.15.0-3-generic  xinit
    cpp-14-x86-64-linux-gnu  libsframe1              linux-tools-6.15.0-3       
     xorg
    gcc-14-base              libxcb-damage0          
linux-tools-6.15.0-3-generic
    libclang1-19             libxkbcommon-x11-0      x11-apps
    libglu1-mesa             linux-headers-6.15.0-3  x11-session-utils
  Use 'sudo apt autoremove' to remove them.

  Upgrading:
    apparmor

  Installing:
    apparmor.d

  Summary:
    Upgrading: 1, Installing: 1, Removing: 0, Not Upgrading: 86
    Download size: 1,116 kB
    Space needed: 3,418 kB / 6,269 MB available

  Continue? [Y/n]
  WARNING: The following packages cannot be authenticated!
    apparmor  apparmor.d

  Install these packages without verification? [y/N] y
  Get:1 http://192.168.122.1/debs/testing questing/ apparmor 
5.0.0~alpha1-0ubuntu5 [853 kB]
  Get:2 http://192.168.122.1/debs/testing questing/ apparmor.d 0.015-1ubuntu1 
[264 kB]
  Fetched 1,116 kB in 0s (20.6 MB/s)
  Preconfiguring packages ...
  (Reading database ... 240702 files and directories currently installed.)
  Preparing to unpack .../apparmor_5.0.0~alpha1-0ubuntu5_amd64.deb ...
  Unpacking apparmor (5.0.0~alpha1-0ubuntu5) over (5.0.0~alpha1-0ubuntu4) ...
  Selecting previously unselected package apparmor.d.
  Preparing to unpack .../apparmor.d_0.015-1ubuntu1_amd64.deb ...
  Unpacking apparmor.d (0.015-1ubuntu1) ...
  Setting up apparmor (5.0.0~alpha1-0ubuntu5) ...
  Installing new version of config file /etc/apparmor.d/hostname ...
  Reloading AppArmor profiles
  Skipping profile in /etc/apparmor.d/disable: brave
  Skipping profile in /etc/apparmor.d/disable: chrome
  Skipping profile in /etc/apparmor.d/disable: chromium
  Skipping profile in /etc/apparmor.d/disable: dig
  Skipping profile in /etc/apparmor.d/disable: element-desktop
  Skipping profile in /etc/apparmor.d/disable: epiphany
  Skipping profile in /etc/apparmor.d/disable: firefox
  Skipping profile in /etc/apparmor.d/disable: flatpak
  Skipping profile in /etc/apparmor.d/disable: foliate
  Skipping profile in /etc/apparmor.d/disable: free
  Skipping profile in /etc/apparmor.d/disable: fusermount3
  Skipping profile in /etc/apparmor.d/disable: hostname
  Skipping profile in /etc/apparmor.d/disable: locale
  Skipping profile in /etc/apparmor.d/disable: loupe
  Skipping profile in /etc/apparmor.d/disable: lsblk
  Skipping profile in /etc/apparmor.d/disable: lsusb
  Skipping profile in /etc/apparmor.d/disable: msedge
  Skipping profile in /etc/apparmor.d/disable: nslookup
  Skipping profile in /etc/apparmor.d/disable: openvpn
  Skipping profile in /etc/apparmor.d/disable: opera
  Skipping profile in /etc/apparmor.d/disable: os-prober
  Skipping profile in /etc/apparmor.d/disable: plasmashell
  Skipping profile in /etc/apparmor.d/disable: signal-desktop
  Skipping profile in /etc/apparmor.d/disable: slirp4netns
  Skipping profile in /etc/apparmor.d/disable: steam
  Skipping profile in /etc/apparmor.d/disable: systemd-coredump
  Skipping profile in /etc/apparmor.d/disable: systemd-detect-virt
  Skipping profile in /etc/apparmor.d/disable: thunderbird
  Skipping profile in /etc/apparmor.d/disable: transmission
  Skipping profile in /etc/apparmor.d/disable: unix-chkpwd
  Warning: found usr.sbin.sssd in /etc/apparmor.d/force-complain, forcing 
complain mode
  Warning from /etc/apparmor.d (/etc/apparmor.d/usr.sbin.sssd line 69): Caching 
disabled for: 'usr.sb
  in.sssd' due to force complain
  Skipping profile in /etc/apparmor.d/disable: virtiofsd
  Skipping profile in /etc/apparmor.d/disable: wg
  Skipping profile in /etc/apparmor.d/disable: wg-quick
  Skipping profile in /etc/apparmor.d/disable: who
  Setting up apparmor.d (0.015-1ubuntu1) ...
  Processing triggers for systemd (257.7-1ubuntu3) ...
  Processing triggers for man-db (2.13.1-1) ...
  Processing triggers for procps (2:4.0.4-8ubuntu2) ...

  georgia@sec2-questing-amd64:~/qrt-test-apparmor$ systemctl status apparmor
  \u25cf apparmor.service - Load AppArmor profiles
       Loaded: loaded (/usr/lib/systemd/system/apparmor.service; enabled; 
preset: enabled)
       Active: active (exited) since Fri 2025-08-29 12:09:41 -03; 21min ago
   Invocation: 7acd3f71e5084f50a7893334f2c2addf
         Docs: man:apparmor(7)
               https://gitlab.com/apparmor/apparmor/wikis/home/
      Process: 13802 ExecReload=/lib/apparmor/apparmor.systemd reload 
(code=exited, status=0/SUCCESS)
     Main PID: 535 (code=exited, status=0/SUCCESS)
     Mem peak: 156.1M (swap: 268K)
          CPU: 5min 18.046s

  Aug 29 12:29:57 sec2-questing-amd64 apparmor.systemd[15293]: Skipping profile 
in /etc/apparmor.d/d>
  Aug 29 12:30:02 sec2-questing-amd64 apparmor.systemd[15328]: Skipping profile 
in /etc/apparmor.d/d>
  Aug 29 12:30:05 sec2-questing-amd64 apparmor.systemd[15373]: Skipping profile 
in /etc/apparmor.d/d>
  Aug 29 12:30:08 sec2-questing-amd64 apparmor.systemd[15437]: Warning: found 
usr.sbin.sssd in /etc/>
  Aug 29 12:30:08 sec2-questing-amd64 apparmor.systemd[15437]: Warning from 
/etc/apparmor.d (/etc/ap>
  Aug 29 12:30:13 sec2-questing-amd64 apparmor.systemd[15456]: Skipping profile 
in /etc/apparmor.d/d>
  Aug 29 12:30:19 sec2-questing-amd64 apparmor.systemd[15483]: Skipping profile 
in /etc/apparmor.d/d>
  Aug 29 12:30:19 sec2-questing-amd64 apparmor.systemd[15484]: Skipping profile 
in /etc/apparmor.d/d>
  Aug 29 12:30:19 sec2-questing-amd64 apparmor.systemd[15492]: Skipping profile 
in /etc/apparmor.d/d>
  Aug 29 12:30:31 sec2-questing-amd64 systemd[1]: Reloaded apparmor.service - 
Load AppArmor profiles.

  For testing, I ran the QA Regression Tests [2]:

  Steps:
  $ git clone https://git.launchpad.net/qa-regression-testing
  $ ./scripts/make-test-tarball ./scripts/test-apparmor.py
  Copying: test-apparmor.py
  Copying: testlib.py
  Copying: install-packages
  Copying: packages-helper
  Copying: apparmor/

  Test files: /tmp/qrt-test-apparmor.tar.gz

  To run, first install the apparmor.d package introduced in this FFE, then 
copy the tarball somewhere, then do:
  $ tar -zxf qrt-test-apparmor.tar.gz
  $ cd ./qrt-test-apparmor
  $ sudo ./install-packages test-apparmor.py
  $ ./test-apparmor.py -v

  This script runs various tests against the installed apparmor
  package

  The result was:

  FAILED: disconnected_mount_complain socketpair
  make: *** [Makefile:487: alltests] Error 1

  
  ----------------------------------------------------------------------
  Ran 62 tests in 3949.185s

  FAILED (failures=1, skipped=4)

  Note that these failures are not related to the apparmor.d package and
  are also reproducible with apparmor version 5.0.0~alpha1-0ubuntu4 from
  the archive.

  [1] https://github.com/roddhjav/apparmor.d
  [2] 
https://git.launchpad.net/qa-regression-testing/tree/scripts/test-apparmor.py

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/apparmor/+bug/2121409/+subscriptions


-- 
Mailing list: https://launchpad.net/~touch-packages
Post to     : [email protected]
Unsubscribe : https://launchpad.net/~touch-packages
More help   : https://help.launchpad.net/ListHelp

Reply via email to