FS based unix sockets have a complicatd interaction with socket
mediation some of the mediation happens in file hooks while other
parts happen in network hooks.
When the kernel doesn't have the unix socket mediation patches the
interactions become largely mediated through the network hooks, as
unix rules get downgraded to socket rules. However some filesystem
operations are needed, and some hooks like bind may be called
differently based on the unix socket type, and not just the address.
Without the kernel patches these variations are not taken into
account.
Changes in the parser networking permission mappings have also
affected the downgrade path, as the parser now supports permissions on
socket rules, downgrades can use permissions and be more faithful to
the original rule but this can also break tests that didn't add all
the permissions needed for the downgrade case.
update unix_socket_pathname.sh to detect whether rule downgrades are
being used, and adjust permissions and expectations based on this.
Fixes: 7ce768244 ("tests: regression: fix regression test for upstream kernels")
Signed-off-by: John Johansen <john.johansen@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1622
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Approved-by: Maxime Bélair <maxime.belair@canonical.com>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit d9ecaf60bc)
The tests on slower systems are occassionally timing out leading to
inconsistent pass/fail runs. The time out failure depending on which
test it occurs in can result in false passes, or failres.
Double the timeout, which hopefully will be enough to avoid the
timeout issue without making the tests wait too long.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit 2c784709b2)
Signed-off-by: John Johansen <john.johansen@canonical.com>
FS based unix sockets have a complicatd interaction with socket
mediation some of the mediation happens in file hooks while other
parts happen in network hooks.
When the kernel doesn't have the unix socket mediation patches the
interactions become largely mediated through the network hooks, as
unix rules get downgraded to socket rules. However some filesystem
operations are needed, and some hooks like bind may be called
differently based on the unix socket type, and not just the address.
Without the kernel patches these variations are not taken into
account.
Changes in the parser networking permission mappings have also
affected the downgrade path, as the parser now supports permissions on
socket rules, downgrades can use permissions and be more faithful to
the original rule but this can also break tests that didn't add all
the permissions needed for the downgrade case.
update unix_socket_pathname.sh to detect whether rule downgrades are
being used, and adjust permissions and expectations based on this.
Fixes: 7ce768244 ("tests: regression: fix regression test for upstream kernels")
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit 08f9466d1c)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The new image-garden snap offers a one-stop-shop for integration
testing, bundling qemu, spread and image-garden build recipes.
Extend the documentation, the run-spread.sh helper script as well as
spread.yaml to support this new method.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit bd500e2391)
Signed-off-by: John Johansen <john.johansen@canonical.com>
... and adjust the tools to raise an exception if such a rule is found.
While this is not nice, it's better than the previous behaviour where
only the last 'fstype' was kept, and the others were lost when writing
the rule.
(cherry picked from commit b5894687ed)
Signed-off-by: John Johansen <john.johansen@canonical.com>
... and adjust the tools to raise an exception if such a rule is found.
While this is not nice, it's better than the previous behaviour where
only the last 'options' was kept, and the others were lost when writing
the rule.
(cherry picked from commit 171e0b1fa9)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The attach_disconnected.sh and deleted.sh tests added expanded their
testing by using unix sockets. This however ever needs support of
unix socket mediation.
Provide a minimal fix by setting bailouts for the the tests if the
requirement is not present. Long term it would be better if the
expected/needed permissions sets could be tweaked to take into
account the permissions required by the use of unix sockets.
The fix f47d5c70a fix af_unix tests for v8 networking, was never
correct, though it worked and was closer before support for fine
grained inet mediation landed. Before finegrained inet mediation
landed unix rules would allow specifying the permission but inet would
not only allowing coarse socket mediation rules. While the backend
supported finegrained permissions in v8 socket mediation the parser
did not.
If af_unix mediation was not supported by the kernel the af_unix
mediation rule would be downgrade to a network rule. All network
socket rules allowed full permission because the parser didn't
support permissions on socket rules. So the "unix create," rule
was being downgraded to a "unix," rule. Thus the "unix create",
rule was enough permissions, in the downgrade even though it
actually wasn't enough permissions.
With support for fine grained inet permissions, support for permissions
on socket rules also landed. When this happend "unix create," was not
enough permissions any more because it was not downgraded to "unix,",
this resulted in failed mediation.
Fixes: cb4a397b1 ("tests: add attach_disconnected tests")
Fixes: f47d5c70a ("fix af_unix tests for v8 networking")
Signed-off-by: John Johansen <john.johansen@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1621
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 0a13a56f5c)
The attach_disconnected.sh and deleted.sh tests added expanded their
testing by using unix sockets. This however ever needs support of
unix socket mediation.
Provide a minimal fix by setting bailouts for the the tests if the
requirement is not present. Long term it would be better if the
expected/needed permissions sets could be tweaked to take into
account the permissions required by the use of unix sockets.
The fix f47d5c70a fix af_unix tests for v8 networking, was never
correct, though it worked and was closer before support for fine
grained inet mediation landed. Before finegrained inet mediation
landed unix rules would allow specifying the permission but inet would
not only allowing coarse socket mediation rules. While the backend
supported finegrained permissions in v8 socket mediation the parser
did not.
If af_unix mediation was not supported by the kernel the af_unix
mediation rule would be downgrade to a network rule. All network
socket rules allowed full permission because the parser didn't
support permissions on socket rules. So the "unix create," rule
was being downgraded to a "unix," rule. Thus the "unix create",
rule was enough permissions, in the downgrade even though it
actually wasn't enough permissions.
With support for fine grained inet permissions, support for permissions
on socket rules also landed. When this happend "unix create," was not
enough permissions any more because it was not downgraded to "unix,",
this resulted in failed mediation.
Fixes: cb4a397b1 ("tests: add attach_disconnected tests")
Fixes: f47d5c70a ("fix af_unix tests for v8 networking")
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit 7ce768244d)
Signed-off-by: John Johansen <john.johansen@canonical.com>
MR: 1561 Added the ability to specify special a keyword to allow
detached mounts. Unfortunately it updated remount to use the device
and devbuffer when remounts current encoding doesn't support it.
This caused the mount.sh regression test to fail in the following
way.
```
$ sudo bash mount.sh
[sudo] password for jj:
using mount rules ...
Error: mount failed. Test 'MOUNT (confined cap bind mount remount rprivate conflict)' was expected to 'pass'. Reason for failure 'FAIL: mount /tmp/sdtest.358520-12403-ASaOnn/mountpoint2 on /tmp/sdtest.358520-12403-ASaOnn/mountpoint failed - Permission denied'
not supported by parser - skipping mount options=(nodirsync),
Error: mount failed. Test 'MOUNT (confined cap mount remount option)' was expected to 'pass'. Reason for failure 'FAIL: mount /dev/loop40 on /tmp/sdtest.358520-12403-ASaOnn/mountpoint failed - Permission denied'
Error: mount failed. Test 'MOUNT (confined cap mount remount)' was expected to 'pass'. Reason for failure 'FAIL: mount /dev/loop40 on /tmp/sdtest.358520-12403-ASaOnn/mountpoint failed - Permission denied'
Error: mount passed. Test 'MOUNT (confined cap mount remount deny option)' was expected to 'fail'
```
Revert the change to remount. This fixes the regression failure.
fa0746f2e parser: add special casing for detached move mounts
Signed-off-by: John Johansen <john.johansen@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1618
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 90938de5c4)
MR: 1561 Added the ability to specify special a keyword to allow
detached mounts. Unfortunately it updated remount to use the device
and devbuffer when remounts current encoding doesn't support it.
This caused the mount.sh regression test to fail in the following
way.
```
$ sudo bash mount.sh
[sudo] password for jj:
using mount rules ...
Error: mount failed. Test 'MOUNT (confined cap bind mount remount rprivate conflict)' was expected to 'pass'. Reason for failure 'FAIL: mount /tmp/sdtest.358520-12403-ASaOnn/mountpoint2 on /tmp/sdtest.358520-12403-ASaOnn/mountpoint failed - Permission denied'
not supported by parser - skipping mount options=(nodirsync),
Error: mount failed. Test 'MOUNT (confined cap mount remount option)' was expected to 'pass'. Reason for failure 'FAIL: mount /dev/loop40 on /tmp/sdtest.358520-12403-ASaOnn/mountpoint failed - Permission denied'
Error: mount failed. Test 'MOUNT (confined cap mount remount)' was expected to 'pass'. Reason for failure 'FAIL: mount /dev/loop40 on /tmp/sdtest.358520-12403-ASaOnn/mountpoint failed - Permission denied'
Error: mount passed. Test 'MOUNT (confined cap mount remount deny option)' was expected to 'fail'
```
Revert the change to remount. This fixes the regression failure.
fa0746f2e parser: add special casing for detached move mounts
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit 89e8fe9c1c)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The picking of 31569f9f7 ("mount: accept fstype and options in any order")
to apparmor-4.1 had a merge conflict, and unfortunately the
mount/ok_opt_86.sd being removed from syntax_failure got dropped.
Remove it.
Since this was already pushed ot the publc tree this is being handled
as a fixup patch instead of revising the original pick as should
normally be done.
Fixes: 31569f9f7 ("mount: accept fstype and options in any order")
Signed-off-by: John Johansen <john.johansen@canonical.com>
Note: If multiple fstype= or options= are given, this is not detected as
an error (to keep the regex simpler). When writing back such a rule,
only one fstype and options will "survive".
Adjust the exclude list in test-parser-simple-tests.py accordingly:
- several valid mount rules no longer fail
- two invalid mount rules which so far accidentally raised an exception
because of the fstype/options order no longer raise this exception
(conflicting mount options, which are the real reason why these rules
are invalid, are not detected in the tools)
Fixes: https://gitlab.com/apparmor/apparmor/-/issues/501
I propose this fix for 4.0, 4.1 and master. (Note: excluding
`mount/ok_opt_86.sd` was added in !1561, so we either need to also
backport this, or have to slightly adjust this MR for backporting.)
Closes#501
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1616
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 37666dd736)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Note: If multiple fstype= or options= are given, this is not detected as
an error (to keep the regex simpler). When writing back such a rule,
only one fstype and options will "survive".
Adjust the exclude list in test-parser-simple-tests.py accordingly:
- several valid mount rules no longer fail
- two invalid mount rules which so far accidentally raised an exception
because of the fstype/options order no longer raise this exception
(conflicting mount options, which are the real reason why these rules
are invalid, are not detected in the tools)
Fixes: https://gitlab.com/apparmor/apparmor/-/issues/501
(cherry picked from commit 7726c86b79)
Signed-off-by: John Johansen <john.johansen@canonical.com>
upsteam move_mount mediation now allows for a detached (disconnected)
mount to be move mounted into a namespace.
Add support for this by detecting 'detached' as a keyword for the
source/device and using it to create a null match. Because existing
mount encoding using a null separator between the mount terms null
match followed by the null seperator will separate detached mounts
within the existing encoding.
```
Eg.
mount detached -> /destination,
mount options=(ro) fstype=ext4 detached -> /destination,
```
This is functionally equivalent to using
```
mount "" -> /destination,
```
However using ```""``` does not provide any context that about what
the rule is allowing or why so the ```detached``` form is preferred.
This is not a perfect solution, but is what can be currently supported
by the kernel without more LSM hooks.
On kernels that don't support detached mount detection, rules using
the detached source conditional will be ignored (never matched).
This encoding also allows the existing
```
mount,
mount options=(move),
mount options=(move) -> /destination,
```
to continue to work with both detached and regular mounts on kernels
that support the move_mount() syscall.
Signed-off-by: John Johansen <john.johansen@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1561
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 2df4bbd39b)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The tools are wrong in parsing the detached mount test.
Until that can be fixed, mark the tools as wrong.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit 2b9f2d2cb7)
Signed-off-by: John Johansen <john.johansen@canonical.com>
upsteam move_mount mediation now allows for a detached (disconnected)
mount to be move mounted into a namespace.
Add support for this by detecting 'detached' as a keyword for the
source/device and using it to create a null match. Because existing
mount encoding using a null separator between the mount terms null
match followed by the null seperator will separate detached mounts
within the existing encoding.
Eg.
mount detached -> /destination,
mount options=(ro) fstype=ext4 detached -> /destination,
This is functionally equivalent to using
mount "" -> /destination,
However using "" does not provide any context that about what the rule is allowing or why so the 'detached' form is preferred.
This is not a perfect solution, but is what can be currently supported
by the kernel without more LSM hooks.
On kernels that don't support detached mount detection, rules using
the detached souce conditional will be ignored (never matched).
This encoding also allows the existing
mount,
mount options=(move),
mount options=(move) -> /destination,
to continue to work with both detached and regular mounts on kernels
that support the move_mount() syscall.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit fa0746f2e2)
Signed-off-by: John Johansen <john.johansen@canonical.com>
When doing testing via LXD VMs and in particular when using "lxc exec" to run
commands in the VM, there is no controlling tty and so the output of last is
missing this column of data. Instead try even harder to parse the timestamp from
the output of "last".
Signed-off-by: Alex Murray <alex.murray@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1582
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit e030ff7ea9)
When doing testing via LXD VMs and in particular when using "lxc exec" to run
commands in the VM, there is no controlling tty and so the output of last is
missing this column of data. Instead try even harder to parse the timestamp from
the output of "last".
Signed-off-by: Alex Murray <alex.murray@canonical.com>
(cherry picked from commit b6654416b0)
Signed-off-by: John Johansen <john.johansen@canonical.com>
When running aa-disable and then aa-enforce passing the binary path as
the argument, aa-enforce fails to enforce the profile with the error:
$ sudo aa-disable /home/foo/test
skipping disabled profile test
Profile for /home/foo/test not found, skipping
According to the man page for aa-enforce, it should work for disabled
profiles.
Note that this does not happen when passing the profile directly to
the tools, so there's a workaround for this issue:
$ sudo /aa-enforce /etc/apparmor.d/test
Setting /etc/apparmor.d/test to enforce mode.
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1579
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 819802cdb3)
When running aa-disable and then aa-enforce passing the binary path as
the argument, aa-enforce fails to enforce the profile with the error:
$ sudo aa-disable /home/foo/test
skipping disabled profile test
Profile for /home/foo/test not found, skipping
According to the man page for aa-enforce, it should work for disabled
profiles.
Note that this does not happen when passing the profile directly to
the tools, so there's a workaround for this issue:
$ sudo /aa-enforce /etc/apparmor.d/test
Setting /etc/apparmor.d/test to enforce mode.
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
(cherry picked from commit b0b45b01c0)
Signed-off-by: John Johansen <john.johansen@canonical.com>
* Make glob_pattern more readable
- replace filename and variable regex parts with RE_PROFILE_PATH_OR_VAR
- split to multiline string
* Move `[\w-]+` into inner match group by removing/moving the ')' after the empty source.
* Prepare source_fileglob_pattern and dest_fileglob_pattern to be customizable by moving adding the closing ')))' into each of them.
* Allow empty source and any word only in mount source
See the individual commits for details.
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1574
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit e065dfb35a)
... by removing/moving the ')' after the empty source.
Also prepare source_fileglob_pattern and dest_fileglob_pattern to be
customizable by moving adding the closing ')))' into each of them.
(cherry picked from commit 537fec8b36)
Signed-off-by: John Johansen <john.johansen@canonical.com>
- replace filename and variable regex parts with RE_PROFILE_PATH_OR_VAR
- split to multiline string
(cherry picked from commit 2843c0f155)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The aa-exec man page makes reference to aa-stack(8) and aa-namespace(8)
manpages that don't exist. For now just remove those references and
add a short blurb on using aa-exec with stacking and namespaces.
Proper full manpages for stacking and namespaces need to be added
but that is beyound the scope of this fix.
Bug: https://gitlab.com/apparmor/apparmor/-/issues/496
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit e64bd1ccdf)
Signed-off-by: John Johansen <john.johansen@canonical.com>
* passwd -e LOGIN was failing
* Allow execution of /usr/sbin/nscd
See: bee77ffc29/lib/nscd.c (L23-L27)
* Allow pam_passwdqc to read /etc/passwdqc.conf and passwdqc filter
files (see https://www.openwall.com/passwdqc/)
* Allow setuid & fsetid capabilities
* Allow locking with /etc/shadow.PID & /etc/shadow.lock
* Allow shadow backup /etc/shadow- and whatever /etc/shadow+ is used for
(cherry picked from commit ab1a455f05)
Signed-off-by: John Johansen <john.johansen@canonical.com>
sbuild is an unconfined profile allowing it to bypass the unprivlieged
user namespace restriction.
unconfined profiles use a pix transition which means that when the
unprivileged_unshare profile is enabled, the binaries in an unconfined
profile calling unshare will cause a transition to the unprivileged_unshare
profile.
This will break sbuild because it needs capabilities within the
user namespace.
However we cannot just add a x transition rule to unconfined profiles, as
the transitions won't be respected. Instead, we have to make the profile
a default allow profile and add a transition that will override
the default pix transition of allow all.
We have to add the attached_disconnected and mediated_deleted flags
because sbuild is manipulating mounts.
Signed-off-by: John Johansen <john.johansen@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1555
Approved-by: Ryan Lee <rlee287@yahoo.com>
Merged-by: Georgia Garcia <georgia.garcia@canonical.com>
(cherry picked from commit 472181ae75)
sbuild is an unconfined profile allowing it to by-pass the unprivlieged
user namespace restritction.
unconfined profiles us a pix transition which means when the
unprivileged_unshare profile is enabled, the binaries in an unconfined
profile calls unshare it will transition to the unprivileged_unshare
profile.
This will break sbuild because it needs capabilities within the
user namespace.
However we can not just add a x transition rule to unconfined profiles,
the transitions won't be respected. Instead we have to make the profile
a default allow profile, and add a transition that will override
the default pix transition of allow all.
We have to add the attached_disconnected and mediated_deleted flags
because sbuild is manipulating mounts.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit 7abfc1baf7)
Signed-off-by: John Johansen <john.johansen@canonical.com>
In the following policy, "ptrace" would be dropped during merging:
```
$FOO=true
/bin/true {
if $FOO {
ptrace,
}
}
```
Current behavior:
```
----- Debugging built structures -----
Name: /bin/true
Local To: <NULL>
Mode:
```
With patch:
```
----- Debugging built structures -----
Name: /bin/true
Local To: <NULL>
Mode:
ptrace,
```
I am quite new to the AA code base, so please let me know if I'm missing something obvious and this is intended behavior :)
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1551
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 46335d58cf)
In the following policy, "ptrace" would be dropped during merging:
$FOO=true
/bin/true {
if $FOO {
ptrace,
}
}
Current behavior:
----- Debugging built structures -----
Name: /bin/true
Local To: <NULL>
Mode:
With patch:
----- Debugging built structures -----
Name: /bin/true
Local To: <NULL>
Mode:
ptrace,
(cherry picked from commit dad66f663b)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The feature matching done in aa_feature_supports calls walk_one to
traverse the features string. This function is supposed to match on
the feature and return, but it matches the feature based on the length
of the feature to check. If the feature to check shorter, then it
would return as if the feature was not present - which was the case
for the following example:
feature_file contains (shortened for example purposes):
```
network_v9 {af_unix {yes
}
}
network_v8 {af_inet {yes
}
}
network {af_unix {yes
}
}
```
if the feature to be checked was simply "network", then walk_one would
return that the feature was not present.
Fix this by restarting the matching if there was not a full match at
the end of the feaure to check.
Fixes: https://bugs.launchpad.net/apparmor/+bug/2105986
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1608
Approved-by: Ryan Lee <rlee287@yahoo.com>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 69355d41f7)
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
--configdir is meant for testing and should override all other configs,
instead of being combined with them. Config combination causes aa-notify
test failures if e.g. the user-local config sets filtering options.
Also supply a notify.conf file for exclusive use during testing.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1610
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit abe588cea4)
--configdir is meant for testing and should override all other configs,
instead of being combined with them. Config combination causes aa-notify
test failures if e.g. the user-local config sets filtering options.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit 096c8df52b)
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
Both the unconfined profile and unprivileged_userns are part of the
default notify.conf, so the default fallback when no configurations are
present should also match this default.
MR: !1609
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit 4623da695e)
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
The original location of the error count checks in the middle of the priority function helper was completely nonsensical. We can instead do this check just once after running all the tests.
In addition, some tests in the priority helper don't use the priority variables at all, and are moved out of the helper to avoid repeating the exact same sequence of tests 16 times.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1604
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit a54e4e2cf9)
The original location in the middle of the priority function helper was
completely nonsensical. We can instead do this check just once after
running all the tests.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit 1f32c1175e)
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
We are seeing some test failures caused by the fact that a fixed kernel,
while available, is not installed the CI image. Since cloud-init does
not itself offer a way to express precise dependency on a package
version we may use a crude replacement of upgrading all the packages at
image construction time.
The next time this happens all we need is to touch the .image-garden.mk
file, so that it is more recent than the image kept in CI cache for the
re-generation to occur.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit f43221d945)
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
The command shell builtin is not recognized by older versions of make, so
switch back to using the which binary instead.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit 52fc40a9cb)
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
linux/mount.h is only used in the move_mount test, which exercises the
move_mount syscall that was introduced sometime in 2018 or later. Older
systems without the header also lack the syscall, so we can just skip the
test in those cases.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit 765d5b87ef)
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
The header was being generated incorrectly in 2 cases:
When the profile/hat contained the parent profile in its name, as in
profile firefox//dash {
hat ^firefox//dash {
and in the unit tests, the child profile or hat was being named as the
parent profile. This was not caught by the general case because the
code has not yet been fully adapted to handle multiple nested child
profiles.
Fixes: https://gitlab.com/apparmor/apparmor/-/issues/493
(backported from MR https://gitlab.com/apparmor/apparmor/-/merge_requests/1592)
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
Add quotes if a mount source or mountpoint includes whitespace.
Also explicitely handle empty mount source (known from
1f33fc9b29)
As usual, some tests can't hurt ;-)
I propose this fix for 4.0..master
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1573
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 2fb0fa9964)
d0626085 add support for writing quoted mount source and mountpoints
Co-authored-by: John Johansen <john@jjmx.net>
First expand nested `(...)` in glob_pattern. This duplicates a few bytes, but makes the regex easier to read.
With that done, allow `-` in glob_pattern.
One of the possible matches in glob_pattern was `\w+` which matched for example `none`.
However, it doesn't match `revokefs-fuse` because of the `-`. Therefore change `\w+` to [\w-]+.
While on it, add two more tests - one for `none` with some options, and one with `revokefs-fuse`.
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1565
Approved-by: Maxime Bélair <maxime.belair@canonical.com>
Approved-by: Ryan Lee <rlee287@yahoo.com>
Merged-by: Christian Boltz <apparmor@cboltz.de>
(cherry picked from commit 2afdf1b214)
c1712c88 Expand nested (...) in glob_pattern
ce87b44b glob_pattern: allow `-`
Co-authored-by: Christian Boltz <apparmor@cboltz.de>
The utils cannot parse some of the newer profile constructs yet, so
generalize a pre-existing mechanism for skipping profiles to use that mechanism in the other tests that need it
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit df0358062d)
Signed-off-by: John Johansen <john.johansen@canonical.com>
... instead of erroring out at the first conflict.
Also display all options involved in a hotkey conflict to make fixing it
easier.
(cherry picked from commit 3e7e9bf01f)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Similarly to apparmor/apparmor!403, we don't really need to pass these flags
here, but if we don't, blhc raises a false positive, and I don't want to get
used to ignoring blhc failures on Debian's GitLab CI.
(cherry picked from commit 187c244056)
Signed-off-by: John Johansen <john.johansen@canonical.com>
When a test fails because of an unexpected success (XFAIL), do not display the empty error log as that may confuse the reader just as it had confused the author.
In addition, when something legitimately fails then display tail of trace log as that may show some useful information.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1548
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 8711c7754b)
We should be using apparmor controlled domains for these files.
Rename the template file from
com.ubuntu.pkexec.aa-notify.policy
to
net.apparmor.pkexec.aa-notify.policy
And update the template file and the install file so that the files
that are generated use net.apparmor instead of com.ubuntu
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit a410f347a3)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The install of the polkit action files for aa-notify leaks build root
information.
From OBS
apparmor-utils.noarch: E: file-contains-buildroot (Badness: 10000) /usr/share/polkit-1/actions/com.ubuntu.pkexec.aa-notify.policy
this is present on Ubuntu as well
<annotate key="org.freedesktop.policykit.exec.path">/build/apparmor-ZUzkoL/apparmor-4.1.0~beta4/debian/tmp/usr/lib/python3/dist-packages/apparmor/update_profile.py</annotate>
this occurs because the {LIB_PATH} template variable is being replaced
with the self.install_lib. Make sure we strip the build prefix if
we are generating the files in a build environment instead of doing
a direct install.
Closes: https://gitlab.com/apparmor/apparmor/-/issues/486
Co-Author: Ryan Lee <ryan.lee@canonical.com>
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit b4e6f0449b)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The install of the polkit action files for aa-notify leaks build root
information.
From OBS
apparmor-utils.noarch: E: file-contains-buildroot (Badness: 10000) /usr/share/polkit-1/actions/com.ubuntu.pkexec.aa-notify.policy
this is present on Ubuntu as well
<annotate key="org.freedesktop.policykit.exec.path">/build/apparmor-ZUzkoL/apparmor-4.1.0~beta4/debian/tmp/usr/lib/python3/dist-packages/apparmor/update_profile.py</annotate>
this occurs because the {LIB_PATH} template variable is being replaced
with the self.install_lib. Make sure we strip the build prefix if
we are generating the files in a build environment instead of doing
a direct install.
Closes: https://gitlab.com/apparmor/apparmor/-/issues/486
Signed-off-by: John Johansen <john.johansen@canonical.com>
Closes#486
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1540
Approved-by: Ryan Lee <rlee287@yahoo.com>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 697e53d752)
This PR only adds the digit `@{d}` and integer `@{int}` variables.
It provides two improvements from the use of the `[0-9]*` glob:
- security: the glob means "a digit followed by anything but `/`", whereas `@{int}` means "up to 10 digits"
Next to the
- stability: using glob in path with `x` can expose to path conflict, removing the glob fixed a lot of issues.
These variables are used by a lot of abstractions that could be upstream here from apparmor.d (PR will follow). It is an import from 33681e14f2/apparmor.d/tunables/multiarch.d/system where other similar variables are in use: `@{hex}`, `@{rand}`, `@{word}`, `@{u8}`, `@{u16}`, `@{u64}`, `@{int2}...@{int64}` ...
They also all could be upstreamed here.
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1544
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 783f012074)
On certain lxc containers, when aa-genprof tries to set
printk_ratelimit, it fails with the OSError exception, with the
message "OSError: [Errno 30] Read-only file system" instead of
PermissionError.
Since PermissionError is a subclass of OSError, replace it by broader
OSError exception to include both cases in which running aa-genprof
fails.
Reported-by: Paulo Flabiano Smorigo <paulo.smorigo@canonical.com>
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1539
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 226ab5f050)
On certain lxc containers, when aa-genprof tries to set
printk_ratelimit, it fails with the OSError exception, with the
message "OSError: [Errno 30] Read-only file system" instead of
PermissionError.
Since PermissionError is a subclass of OSError, replace it by broader
OSError exception to include both cases in which running aa-genprof
fails.
Reported-by: Paulo Flabiano Smorigo <paulo.smorigo@canonical.com>
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
(cherry picked from commit e1ae6fa81c)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Instead of setting those variables unconditionally, set them if they
aren't externally set by environment variables. This will allow for usages
like DESTDIR=/some/other/dir make install in the utils directory.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit 2747013d9b)
Signed-off-by: John Johansen <john.johansen@canonical.com>
ttkthemees may not be installed on some systems, and if not present
will cause aa-notify to fail. Instead of making ttkthemes a required
dependency, make its use conditional on it being present.
Backport by: Christian Boltz <apparmor@cboltz.de>
Signed-off-by: John Johansen <john.johansen@canonical.com>
Changes include:
- using `long` instead of `intmax_t` for `pid_t` typemap (32-bit build failure); see commit message for more details
- specifying messages for `static_assert` declarations (required up until C23, was accepted as a compiler extension on the systems I had tested this on previously)
- removing label-followed-by-declaration instance (also a C23 feature supported as extension)
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1536
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 55889ef783)
The message being optional is apparently a C23 thing that was available as an extension on the systems I tested on previously
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit 87b60e4e94)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The previous code using intmax_t failed to build on armhf because
intmax_t was long long int instead of long int on that platform.
As to shrinking down to a long: not only does SWIG lack a
SWIG_AsVal_intmax_t, but aalogparse also assumes PIDs fit in a long
by storing them as unsigned longs in aa_log_record. Thus, we can
assume that sizeof(pid_t) <= sizeof(long) right now and deal with
the big headache that a change to pid_t would cause if it becomes
larger than a long in the future.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit c5016e1227)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The documentation was missing information about path sanitization, and
why you shouldn't do a leading @{VAR} on path rules. While the example
doing this was fixed, actual information about why you shouldn't do
this was missing.
Document how apparmor will collapse consecutive / characters into a
single character for paths, except when this occurs at the start of
the path.
Signed-off-by: John Johansen <john.johansen@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1532
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Approved-by: Christian Boltz <apparmor@cboltz.de>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 0c4e452b46)
The documentation was missing information about path sanitization, and
why you shouldn't do a leading @{VAR} on path rules. While the example
doing this was fixed, actual information about why you shouldn't do
this was missing.
Document how apparmor will collapse consecutive / characters into a
single character for paths, except when this occurs at the start of
the path.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit cce5bd6e95)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The unshare-userns-restrict profile contained a cx transition to
transition to a profile that allows most things while denying
capabilities:
audit allow cx /** -> unpriv,
However, this transition does not stack the unshare//unpriv profile
against any other profile the target binary might have had. As a result,
the lack of stacking resulted in a non-namespace-related sandboxing
bypass in which attachments of other profiles that should have confined
the target binary do not get applied. Instead, we adopt a stack similar
to the one in bwrap-userns-restrict, with the exception that unshare
does not use no-new-privs and therefore only needs a two-layer stack
instead of a three-layer stack.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1533
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 8e586e5492)
The unshare-userns-restrict profile contained a cx transition to
transition to a profile that allows most things while denying
capabilities:
audit allow cx /** -> unpriv,
However, this transition does not stack the unshare//unpriv profile
against any other profile the target binary might have had. As a result,
the lack of stacking resulted in a non-namespace-related sandboxing
bypass in which attachments of other profiles that should have confined
the target binary do not get applied. Instead, we adopt a stack similar
to the one in bwrap-userns-restrict, with the exception that unshare
does not use no-new-privs and therefore only needs a two-layer stack
instead of a three-layer stack.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit ab3ca1a93f)
Signed-off-by: John Johansen <john.johansen@canonical.com>
There are minor tweak around the lib, constify some vars etc. That
don't justify a large bump. Bumpt there have been changes to swig
such that we want to force it to be linked against the new lib.
Signed-off-by: John Johansen <john.johansen@canonical.com>
The current behavior of priority rules can be non-intuitive with
higher priority rules completely overriding lower priority rules even in
permissions not held in common. This behavior does have use cases but
its can be very confusing, and does not normal policy behavior
Eg.
priority=0 allow r /**,
priority=1 deny w /**,
will result in no allowed permissions even though the deny rule is
only removing the w permission, beause the higher priority rule
completely over ride lower priority permissions sets (including
none shared permissions).
Instead move to tracking the priority at a per permission level. This
allows the w permission to still override at priority 1, while the
read permission is allowed at priority 0.
The final constructed state will still drop priority for the final
permission set on the state.
Note: this patch updates the equality tests for the cases where
the complete override behavior was being tested for.
The complete override behavior will be reintroduced in a future
patch with a keyword extension, enabling that behavior to be used
for ordered blocks etc.
Signed-off-by: John Johansen <john.johansen@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1522
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 0ab4fc0580)
The current behavior of priority rules can be non-intuitive with
higher priority rules completely overriding lower priority rules even in
permissions not held in common. This behavior does have use cases but
its can be very confusing, and does not normal policy behavior
Eg.
priority=0 allow r /**,
priority=1 deny w /**,
will result in no allowed permissions even though the deny rule is
only removing the w permission, beause the higher priority rule
completely over ride lower priority permissions sets (including
none shared permissions).
Instead move to tracking the priority at a per permission level. This
allows the w permission to still override at priority 1, while the
read permission is allowed at priority 0.
The final constructed state will still drop priority for the final
permission set on the state.
Note: this patch updates the equality tests for the cases where
the complete override behavior was being tested for.
The complete override behavior will be reintroduced in a future
patch with a keyword extension, enabling that behavior to be used
for ordered blocks etc.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit 1ebd991155)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The original patch adding priority to the set of prefixes failed to
update the prefix dump to include the priority priority field.
Fixes: e3fca60d1 ("parser: add the ability to specify a priority prefix to rules")
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit e56dbc2084)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The priority field is only used during state construction, and can
even prevent later optimizations like minimization. The parser already
explcitily clears the states priority field as part of the last thing
done during construction so it doesn't prevent minimization
optimizations.
This means the state priority not only wastes storage because it is
unused post construction but if used it could introduce regressions,
or other issues.
The change to the minimization tests just removes looking for the
priority field that is no longer reported.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit cc31a0da22)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Like was done for the other MatchFlags switch to using a node type
instead of dynamic_cast as this will result in a performance
improvement.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit 9221d291ec)
Signed-off-by: John Johansen <john.johansen@canonical.com>
This requires a runner with the tags: linux, x86_64, kvm. One needs to
be provisioned for the AppArmor project for the pipeline to function.
It is possible to run the same tests on SAAS runners offered by GitLab
but due to issue gitlab-org/gitlab-runner#6208 there is no way to expose
/dev/kvm on the host to the guest. Without this feature emulation works
but is rather slow as to be impractical.
Note that there's some overlap between the build-all job and spread that
might be avoided in the future. At present this is made more difficult
by the fact that the path where build-all job builds libapparmor is
stored internally by autotools. This prevents us from using GitLab
artifacts from moving the built files across to the spread testing jobs
without extra work.
In addition to adding the spread job, remove test-build-regression job.
This job is now redundant since the same operation is done when spread
builds and runs regression tests.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1512
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 12787648a7)
We were not building or caching the .seed.iso target, causing make to re-create
the image, as seen in the make --debug --dry-run output:
```
Updating goal targets....
File ubuntu-cloud-24.04.user-data does not exist.
Must remake target ubuntu-cloud-24.04.user-data.
echo "${USER_DATA}" | tee ubuntu-cloud-24.04.user-data
Successfully remade target file ubuntu-cloud-24.04.user-data.
File ubuntu-cloud-24.04.meta-data does not exist.
Must remake target ubuntu-cloud-24.04.meta-data.
echo "${META_DATA}" | tee ubuntu-cloud-24.04.meta-data
Successfully remade target file ubuntu-cloud-24.04.meta-data.
Prerequisite ubuntu-cloud-24.04.user-data is newer than target ubuntu-cloud-24.04.seed.iso.
Prerequisite ubuntu-cloud-24.04.meta-data is newer than target ubuntu-cloud-24.04.seed.iso.
Must remake target ubuntu-cloud-24.04.seed.iso.
/usr/bin/genisoimage \
-input-charset utf-8 \
-output ubuntu-cloud-24.04.seed.iso \
-volid CIDATA \
-joliet \
-rock \
-graft-points \
user-data=ubuntu-cloud-24.04.user-data \
meta-data=ubuntu-cloud-24.04.meta-data
Successfully remade target file ubuntu-cloud-24.04.seed.iso.
Prerequisite ubuntu-cloud-24.04.seed.iso is newer than target ubuntu-cloud-24.04.x86_64.qcow2.
```
Build and cache the cloud-init seed iso to prevent that.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit 4cfeb4a9ad)
Signed-off-by: John Johansen <john.johansen@canonical.com>
We are seeing images cached and then re-constructed as if something had
changed in the meanitime. Debug image construction with make --dry-run --debug.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit b3ce87af23)
Signed-off-by: John Johansen <john.johansen@canonical.com>
This way there's somewhat less repetition and the flow of job definitions is,
at least to me, easier to read.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit bcf8c968db)
Signed-off-by: John Johansen <john.johansen@canonical.com>
A new explicit, non-parallel job is injected when the .image-garden.mk or
.spread.yaml file changes. This job warms up the cache for the subsequent
parallel testing jobs.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit 14ceb92ca0)
Signed-off-by: John Johansen <john.johansen@canonical.com>
As a security measure, GitLab splits cache into two broad pools: protected and
non-protected. Any job running in a protected branch has access to the
protected cache pool. All other jobs run in the non-protected cache pool.
This effectively forces us to push to cache in non-protected branches, like all
the merge requests, in order to actually use the cache.
Ideally we'd disable this protection and only push from the default branch and
pull otherwise, as changes to dependency set is rather rare.
[1] https://docs.gitlab.com/ee/ci/caching/#use-the-same-cache-for-all-branches
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit a0adb01631)
Signed-off-by: John Johansen <john.johansen@canonical.com>
This requires a runner with the tags: linux, x86_64, kvm. One needs to
be provisioned for the AppArmor project for the pipeline to function.
It is possible to run the same tests on SAAS runners offered by GitLab
but due to issue gitlab-org/gitlab-runner#6208 there is no way to expose
/dev/kvm on the host to the guest. Without this feature emulation works
but is rather slow as to be impractical.
Note that there's some overlap between the build-all job and spread that
might be avoided in the future. At present this is made more difficult
by the fact that the path where build-all job builds libapparmor is
stored internally by autotools. This prevents us from using GitLab
artifacts from moving the built files across to the spread testing jobs
without extra work.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit 7f68ed174c)
Signed-off-by: John Johansen <john.johansen@canonical.com>
There's contention between running spread across many nodes, in chunks,
in a CI/CD pipeline, and running spread on one machine, across many
instances at the same time. The case with CI/CD needs one worker, as
parallelism is provided by GitLab. The case with local spread needs many
workers as parallelism is provided locally by spread allocating new
instances.
At present we need to focus on the CI/CD case. I have a plan on how to
avoid the problem entirely down the line, by running multiple copies of
spread locally, as if everything was done in a CI/CD pipeline.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit dfa331dfff)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Unfortunately similar to bwrap unshare will need the mediate_deleted
flag in some cases.
see
commit 6488e1fb7 "profiles: add mediate_deleted to bwrap"
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit c157eb0cb6)
Signed-off-by: John Johansen <john.johansen@canonical.com>
This allows the use of sparse allocation on filesystems that support it,
allowing a fallback when the underlying filesystem doesn't.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit 98c60e477d)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Calling losetup -f first and passing its result to create the loop device
creates a race condition in which the loop device might be claimed first
in between the two losetup calls. Instead, create the device atomically
and then obtain the loop device /dev/ handle afterwards.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit 95f3bdf66b)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Introduce the EXPECT_DENIALS environment variable for profile tests.
Each line of EXPECT_DENIALS is a regex that must match an AppArmor
denial for the corresponding test, and conversely.
This ensures that problematic behaviors are correctly blocked and logged.
Signed-off-by: Maxime Bélair <maxime.belair@canonical.com>
(cherry picked from commit fc3f27e255)
Signed-off-by: John Johansen <john.johansen@canonical.com>
A number of tests are failing and since spread does not contain a native
XFAIL facility, we have to maintain a silent-failure feature code
ourselves. A few of those have been fixed since the first iteration of
this patch. The remaining known failures are being fixed.
Later on I would like to separate XFAIL from SKIP so that if a test is
known to exercise kernel feature unavailable on the given system, the
test is just not executed.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1483
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit d482aab419)
This is so that we get a baseline that passes to enable testing in CI/CD
but also to spark a discussion around what to do with a profile that
indirectly relies on a kernel feature that is not available on a given
system.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit 32bf95bb1e)
Signed-off-by: John Johansen <john.johansen@canonical.com>
A number of tests are failing and since spread does not contain a native
XFAIL facility, we have to maintain a silent-failure feature code
ourselves. A few of those have been fixed since the first iteration of
this patch. The remaining known failures are being fixed.
Later on I would like to separate XFAIL from SKIP so that if a test is
known to exercise kernel feature unavailable on the given system, the
test is just not executed.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit b0422d5572)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The prefix can be done in higher-level languages via slicing and having an explicit length exposes an out-of-bounds memory read footgun to those higher level languages
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit a2df3143d1)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Surprisingly, SWIG did not pick up the "typedef int pid_t" from the C headers.
As such, we need to provide our own wrapper. We don't just replicate the typdef
because we still support systems that have 16-bit PIDs.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit d199c2ae33)
Signed-off-by: John Johansen <john.johansen@canonical.com>
In Python, return status is signalled by exceptions (or lack thereof)
instead of int. Keep the typemap portable for any other languages we may
add in the future.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit a15768b0bf)
Signed-off-by: John Johansen <john.johansen@canonical.com>
This includes a custom typemap to handle (char **label, char **mode)
pairs and a cstring_output_allocate declaration for char **mnt.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit 50d26beb00)
Signed-off-by: John Johansen <john.johansen@canonical.com>
This is one of those functions that never worked anyways, because it
modified the passed-in label in place. Moreover, it is a low-level
interface that requires its callers to manually construct a binary query.
As such, it would be better not to expose it and to add wrappers like
aa_query_file_path for the other query classes if that functionality is
needed later.
The removal of this function from the bindings was dropped from !1337 because it exposed functionality that was not present in wrappers around aa_query_label. However, upon further discussion, we decided that it'd be better to remove it now and add other wrappers to libapparmor itself if the functionality provided by the existing wrappers became insufficient.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1352
Approved-by: John Johansen <john@jjmx.net>
Merged-by: Ryan Lee <rlee287@yahoo.com>
(cherry picked from commit 5b141dd580)
This is one of those functions that never worked anyways, because it
modified the passed-in label in place. Moreover, it is a low-level
interface that requires its callers to manually construct a binary query.
As such, it would be better not to expose it and to add wrappers like
aa_query_file_path for the other query classes if that functionality is
needed later.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit d3603a1f20)
Signed-off-by: John Johansen <john.johansen@canonical.com>
It doesn't make sense to expose the *_raw functions or the varg version
of aa_change_hatv to higher-level languages. While technically a breaking
change, the generated bindings for these functions never actually worked
anyways:
- aa_change_hat_vargs uses C varargs, which SWIG passes in NULL for by
default. It does not attempt to process the passed-in arguments at all
(and in fact caused an unused-argument compiler warning when compiling
the generated bindings).
- aa_getprocattr_raw and aa_getpeercon_raw both place output into a ``char
**mode`` pointer. SWIG by default generates these as opaque pointer
object arguments, rendering them unusable for getting output. Future
patches would be needed to fix ``char**`` arguments for the other functions
that use them. Moreover, these functions expect their caller to handle
memory allocation, which is also not possible from a higher-level
language point of view.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1337
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: Ryan Lee <rlee287@yahoo.com>
(cherry picked from commit d35a6939be)
It doesn't make sense to expose the *_raw functions or the varg version
of aa_change_hatv to higher-level languages. While technically a breaking
change, the generated bindings for these functions never actually worked
anyways:
- aa_change_hat_vargs uses C varargs, which SWIG passes in NULL for by
default. It does not attempt to process the passed-in arguments at all
(and in fact caused an unused-argument compiler warning when compiling
the generated bindings).
- aa_getprocattr_raw and aa_getpeercon_raw both place output into a char
**mode pointer. SWIG by default generates these as opaque pointer
object arguments, rendering them unusable for getting output. Future
patches would be needed to fix char** arguments for the other functions
that use them. Moreover, these functions expect their caller to handle
memory allocation, which is also not possible from a higher-level
language point of view.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit 2bd1884654)
Signed-off-by: John Johansen <john.johansen@canonical.com>
This patchset adds annotations so that SWIG can automatically manage the memory lifetimes of aa_log_record objects, and ensures proper cleanup is done in the %exception handler.
This is the first of a sequence of MRs to overhaul the SWIG bindings and fix pieces that never actually worked in the first place. As fixing those other pieces will require breaking changes, I am separating out the non-breaking changes into separate MRs.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1334
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: Ryan Lee <rlee287@yahoo.com>
(cherry picked from commit bcab725670)
Swig generates a "thisown" attribute, which is an escape hatch in case
higher-level code does something weird and needs to tell SWIG whether to
free the C object when Python garbage collects it. Adding this attribute
is not a breaking change w.r.t access to the other attributes of the parsed
record.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit e5fd0fc636)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Unfortunately SWIG_exception does not support throwing OSError, so this
still requires Python-specific code.
Unlike just returning NULL, this will clean up intermediate allocations.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit 436ebda9b5)
Signed-off-by: John Johansen <john.johansen@canonical.com>
This change matches the names in the .c source and the man page for aa_query_label,
and also simplifies the typemap annotations needed to make the SWIG versions usable.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit 0c4cda2f1c)
Signed-off-by: John Johansen <john.johansen@canonical.com>
By making the test a file to be included as a helper, we can reuse most of the code for a fuse_overlayfs test without copy-pasting
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit 9413658277)
Signed-off-by: John Johansen <john.johansen@canonical.com>
We now have GitLab CI/CD pipeline co-existing with spread, coupled with
image-garden and the cloud-init profile defined for each distribution.
To avoid duplicating list of required dependencies, re-use cloud-init
profile as the reference list of dependencies (superset between build
and test) to install.
In addition to the dependency list, the build_all job now re-uses spread
prepare section in similar fashion. If it builds in spread, it should
build in CI as well.
A small quality-of-life improvement is the shape of a collapsible
section around dependency installation should make reading job logs
easier.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1494
Approved-by: Christian Boltz <apparmor@cboltz.de>
Merged-by: Zygmunt Krynicki <me@zygoon.pl>
(cherry picked from commit 4c8c4a1d77)
We now have GitLab CI/CD pipeline co-existing with spread, coupled with
image-garden and the cloud-init profile defined for each distribution.
To avoid duplicating list of required dependencies, re-use cloud-init
profile as the reference list of dependencies (superset between build
and test) to install.
In addition to the dependency list, the build_all job now re-uses spread
prepare section in similar fashion. If it builds in spread, it should
build in CI as well.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit f01a40a77c)
Signed-off-by: John Johansen <john.johansen@canonical.com>
This should be a more readable example to follow in other tests. The
toybox test was special given the fact that it is a shell itself, and is
fairly programmable.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit ffd38b7ac4)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Using gdb in batch mode, put a breakpoint on _start and spawn the
process. Then using the built-in python interpreter print the
confinement label on the process and terminate everything.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit 23df780544)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Hopefully more and more profiles will come with smoke tests. Since the
pattern of those tests is likely to be very similar (compile profile,
run some programs, remove profile) it will be good to check if the
profile had caused any denials to be logged. Having this at the suite
level should make writing actual tests easier.
The prepare-each and restore-each logic compile the profile, check for
errors and finally remove the profile. The debug-each logic shows the
program name (with full path).
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit 2ab2c8f8a1)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Tests that interact with the kernel are skipped (tests/regression and
tests/snapd) but everything else is green. Most of the tests are
actually passing. The only exception is the aa-notify test that was
broken by Python 3.13 stdlib change. The fix for that has been posted
separately.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1496
Approved-by: Christian Boltz <apparmor@cboltz.de>
Merged-by: Zygmunt Krynicki <me@zygoon.pl>
(cherry picked from commit 1462e1c4b0)
Tests that interact with the kernel are skipped (tests/regression and
tests/snapd) but everything else is green. Most of the tests are
actually passing. The only exception is the aa-notify test that was
broken by Python 3.13 stdlib change. The fix for that has been posted
separately.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit 7ce6819c53)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The openSUSE project has decided to switch to security=selinux by
default. For the purpose of continuing to test AppArmor on the
distribution, alter the cloud-init profile to switch to booting with
security=apparmor.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit 2b44cc09a6)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The artifact restoration step does not preserve mtime, resulting in source files newer than built files, resulting in a needless rebuild of everything before actually running the tests.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit 030f991320)
Signed-off-by: John Johansen <john.johansen@canonical.com>
This allows building the tst_* binaries in parallel independently of running the parser test suite
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit 2e841655cf)
Signed-off-by: John Johansen <john.johansen@canonical.com>
This pipeline only makes sense to run in the upstream project where
the coverity variables are defined, so they currently fail in forks.
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
(cherry picked from commit c382efe119)
Signed-off-by: John Johansen <john.johansen@canonical.com>
This makes the snapd/mount-control test pass on all the currently tested systems. Note that there's a somewhat complex problem with the new mount APIs (https://lwn.net/Articles/753473/) from 2018 that are now being used on, for example, Debian 13.
I will need to make similar changes to the profiles generated by snapd, so any insight on what to do there is strongly appreciated.
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1479
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit f171f5ebc8)
In addition allow linking to libeconf, generalize locale paths to cover
values other than C.UTF-8 and allow reading system-wide locale.alias and
gconv modules.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit cff25b8d17)
Signed-off-by: John Johansen <john.johansen@canonical.com>
This is not the best of fixes but it seems that on Debian 13, with new
libmount calling fsopen/fsconfig/move_mount, the current apparmor mount
rule is insufficient to allow the call to go through.
The key problems are:
- the fstype is not visible to LSM
- the source directory is an empty string
- the mount is moved to final position
I don't know the extent of "new" mount API coverage by LSM hooks but
I think we should either synthesize new permissions from old rules,
.e.g match each of the system calls against what the mount class
expression, or somehow allow the exceptions better.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit 5556de53c0)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The package is required by the file_unbindable_mount regression test.
To properly re-generate affected images please update image-garden
to version containing 9714dc45d0ef06862ffe7037193dc43386db48ea
(Tie .user-data and .meta-data to MAKEFILE_LIST).
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit 215fab71a5)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The BASH_XTRACEFD variable can be used to redirect "set -x" traces
to a dedicated file. We can use it to split the execution trace
(what has actually happened) from the failure messages.
On a failing test this does provide improved clarity when debugging
interactively with "spread -debug". On non-interactive runs the now
shorter error list is also implicitly printed.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1481
Approved-by: Christian Boltz <apparmor@cboltz.de>
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: Christian Boltz <apparmor@cboltz.de>
(cherry picked from commit b4cb33b488)
The BASH_XTRACEFD variable can be used to redirect "set -x" traces
to a dedicated file. We can use it to split the execution trace
(what has actually happened) from the failure messages.
On a failing test this does provide improved clarity when debugging
interactively with "spread -debug". On non-interactive runs the now
shorter error list is also implicitly printed.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit 2c2e0478f8)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Construction of the chfa can reorder states from what the numbering
given during the hfa constuctions because of reordering for better
compression, dead state removal to ensure better packing etc.
This however means the dfa dump is difficult (it is possible using
multiple dumpes) to match up to the chfa that the kernel is
using. Make this easier by making the dfa dump be able to take
the remapping as input, and provide an option to dump the
chfa equivalent hfa.
Renumbered states will show up as {new <== {orig}} in the dump
Eg.
```
--D dfa-states
{1} <== priority (allow/deny/prompt/audit/quiet)
{5} 0 (0x 4/0//0/0/0)
{1} perms: none
0x2 -> {5} 0 (0x 4/0//0/0/0)
0x4 -> {5} 0 (0x 4/0//0/0/0)
\a 0x7 -> {5} 0 (0x 4/0//0/0/0)
\t 0x9 -> {5} 0 (0x 4/0//0/0/0)
\n 0xa -> {5} 0 (0x 4/0//0/0/0)
\ 0x20 -> {5} 0 (0x 4/0//0/0/0)
4 0x34 -> {3}
{3} perms: none
0x0 -> {6}
{6} perms: none
1 0x31 -> {5} 0 (0x 4/0//0/0/0)
```
```
-D dfa-compressed-states
{1} <== priority (allow/deny/prompt/audit/quiet)
{2 == {5}} 0 (0x 4/0//0/0/0)
{1} perms: none
0x2 -> {2 == {5}} 0 (0x 4/0//0/0/0)
0x4 -> {2 == {5}} 0 (0x 4/0//0/0/0)
\a 0x7 -> {2 == {5}} 0 (0x 4/0//0/0/0)
\t 0x9 -> {2 == {5}} 0 (0x 4/0//0/0/0)
\n 0xa -> {2 == {5}} 0 (0x 4/0//0/0/0)
\ 0x20 -> {2 == {5}} 0 (0x 4/0//0/0/0)
4 0x34 -> {3}
{3} perms: none
0x0 -> {4 == {6}}
{4 == {6}} perms: none
1 0x31 -> {2 == {5}} 0 (0x 4/0//0/0/0)
```
Signed-off-by: John Johansen <john.johansen@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1474
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 72f9952a5f)
Construction of the chfa can reorder states from what the numbering
given during the hfa constuctions because of reordering for better
compression, dead state removal to ensure better packing etc.
This however means the dfa dump is difficult (it is possible using
multiple dumpes) to match up to the chfa that the kernel is
using. Make this easier by making the dfa dump be able to take the
emapping as input, and provide an option to dump the chfa equivalent
hfa.
Renumbered states will show up as {new <== {orig}} in the dump
Eg.
--D dfa-states
{1} <== priority (allow/deny/prompt/audit/quiet)
{5} 0 (0x 4/0//0/0/0)
{1} perms: none
0x2 -> {5} 0 (0x 4/0//0/0/0)
0x4 -> {5} 0 (0x 4/0//0/0/0)
\a 0x7 -> {5} 0 (0x 4/0//0/0/0)
\t 0x9 -> {5} 0 (0x 4/0//0/0/0)
\n 0xa -> {5} 0 (0x 4/0//0/0/0)
\ 0x20 -> {5} 0 (0x 4/0//0/0/0)
4 0x34 -> {3}
{3} perms: none
0x0 -> {6}
{6} perms: none
1 0x31 -> {5} 0 (0x 4/0//0/0/0)
-D dfa-compressed-states
{1} <== priority (allow/deny/prompt/audit/quiet)
{2 == {5}} 0 (0x 4/0//0/0/0)
{1} perms: none
0x2 -> {2 == {5}} 0 (0x 4/0//0/0/0)
0x4 -> {2 == {5}} 0 (0x 4/0//0/0/0)
\a 0x7 -> {2 == {5}} 0 (0x 4/0//0/0/0)
\t 0x9 -> {2 == {5}} 0 (0x 4/0//0/0/0)
\n 0xa -> {2 == {5}} 0 (0x 4/0//0/0/0)
\ 0x20 -> {2 == {5}} 0 (0x 4/0//0/0/0)
4 0x34 -> {3}
{3} perms: none
0x0 -> {4 == {6}}
{4 == {6}} perms: none
1 0x31 -> {2 == {5}} 0 (0x 4/0//0/0/0)
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit 50452e1147)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Our ubuntu packaging builds Python-enabled libapparmor's in the directories `libapparmor/libapparmor.python[version_identifier]`. In order for the util's `make check` to pick up on the correct libapparmor during the Ubuntu build process, we need the ability to override its search path. This patch introduces a `LIBAPPARMOR_BASEDIR` variable to allow for that.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1497
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: Georgia Garcia <georgia.garcia@canonical.com>
(cherry picked from commit 17a09d2987)
90143494 Allow overrides and preservation of some environment variables in utils make check
Co-authored-by: Georgia Garcia <georgia.garcia@canonical.com>
The first patch fixes a `test-aa-notify.py` `TypeError` when `APPARMOR_NOTIFY` and `__AA_CONFDIR` are both specified, which is something that was broken all this time.
The second patch ensures that `aa-notify` in the test suite is run using the same Python interpreter that the test suite itself is run with, which is necessary for testing the utils under different Pythons.
The third patch does analogous modifications to the minitools tests that launch `aa-audit`, `aa-complain`, etc.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1498
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: Georgia Garcia <georgia.garcia@canonical.com>
(cherry picked from commit 625a919bb8)
3365e492 utils: test: test-aa-notify: Ensure aanotify_bin is always a list
77cabf7d utils: test: use sys.executable when launching aa-notify in tests
e32c2673 utils: test: use sys.executable when launching minitools in tests
Co-authored-by: Georgia Garcia <georgia.garcia@canonical.com>
This extra functionality is to be used in a different regression test that reuses the binary
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit 8212fa8be4)
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
lastlog2 is the 2038-safe replacement for wtmp, and in the meantime
became part of util-linux.
Adjust get_last_login_timestamp() to use the lastlog2 database
(/var/lib/lastlog/lastlog2.db) if it exists, and adjust
get_last_login_timestamp_lastlog2() to actually do that.
(If lastlog2.db doesn't exist, aa-notify will read wtmp as usual.)
Unfortunately lastlog2 doesn't have a way to get machine-readable output
(for example json), therefore - after trying and failing to parse the
lastlog2 output - directly read from lastlog2.db. Let's hope the format
never changes ;-)
Fixes: https://bugzilla.opensuse.org/show_bug.cgi?id=1228378
Fixes: https://bugzilla.opensuse.org/show_bug.cgi?id=1216660
Fixes: https://gitlab.com/apparmor/apparmor/-/issues/372
I propose this patch for 4.0 and master.
Closes#372
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1282
Approved-by: Christian Boltz <apparmor@cboltz.de>
Merged-by: Christian Boltz <apparmor@cboltz.de>
(cherry picked from commit 692e6850ba)
7d537efc Rename get_last_login_timestamp to get_last_login_timestamp_wtmp
371a9ff9 Add support for lastlog2 to get last login
45e4c27c Add support for lastlog2 to get last login
Co-authored-by: Christian Boltz <apparmor@cboltz.de>
Tested by using Valgrind's Helgrind and DRD against the reentrancy test that I wrote: they both report no errors with the changes while reporting many errors with the old versions.
Commits "Inline _parse_yacc in libaalogparse" and "Make parse_record take a const char pointer since it never modified str anyways" have a tiny potential to be backwards-incompatible changes: I have justified why they shouldn't be in the commit messages, but it's worth looking over in case I was mistaken and we need to back those out.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1322
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 37cac653d1)
The entry AA_RECORD_SYNTAX_V1 is only there for API compatibility reasons.
If we wanted to remove it, we could just renumber the other two entries
to preserve ABI compatibility. However, it seems easier to just delete the
entry if we ever break backcompat with a libapparmor2.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit 79670745d6)
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
This shouldn't be a breaking change because it's fine to pass a
non-const pointer to a function taking a const pointer, but not the other way round
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit 78f138c37f)
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
This function was only ever called once inside libaalogparse.c, and it looks
simple enough to not need to be split out into its own helper function.
As this function was never exposed publicly in installed header files, removing it
is not a breaking API change.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit 6a55fb5613)
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
The generated grammar.h already sets the correct YYDEBUG value regardless
of whether parse.trace is defined
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit 7ff045583d)
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
I had this message in my log
```
Dez 30 08:14:46 kernel: audit: type=1400 audit(1735542886.787:307): apparmor="DENIED" operation="open" class="file" profile="/usr/sbin/cupsd" name="/etc/paperspecs" pid=317509 comm="cupsd" requested_mask="r" denied_mask="r" fsuid=0 ouid=0
```
If the second commit is bad, I can drop it.
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1472
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit e5a960a685)
While cups itself writes to /etc the others require only read-only access
and might therefore live in /usr/etc.
(cherry picked from commit c3af6228fd)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Gtk applications like Firefox request write access to the file
`/run/user/1000/dconf/user`. The code in `dconf_shm_open` opens the file
with `O_RDWR | O_CREAT`.
4057f8c84f/shm/dconf-shm.c (L68)
(cherry picked from commit 318fb30446)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Fix priority for file rules, and the ability to dump the dfa at different stages, and update and fix the equality tests.
This in particular adds the ability to better debug the equality tests. Instead of just piping the parser output into the hash it creates a tmp dir and drops the binary files there so they can be manually examined. It adds new options particularly the -r option making so the tests will exit on first failure to make it easier to isolate and examine a failure.
Eg.
```
./equality.sh -r -d -v
Equality Tests:
................................................................................................................................................................................................................................
Binary inequality 'priority=-1'x'priority=-1' change_hat rules automatically inserted
FAIL: Hash values match
parser: ./../apparmor_parser -QKSq --features-file=./features_files/features.all
known-good (ee4f926922ecd341f1389a79dd155879) == profile-under-test (ee4f926922ecd341f1389a79dd155879) for the following profiles:
known-good /t { priority=-1 owner /proc/[0-9]*/attr/{apparmor/,}current a, ^test { priority=-1 owner /proc/[0-9]*/attr/{apparmor/,}current a, /f r, }}
profile-under-test /t { priority=-1 owner /proc/[0-9]*/attr/{apparmor/,}current w, ^test { priority=-1 owner /proc/[0-9]*/attr/{apparmor/,}current w, /f r, }}
files retained in "/tmp/eq.3240859-deHu10/"
```
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1455
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 40e9b2a961)
There is a general industry wide effort to move off of md5 and even
sha1 (see recent kernel changes). While in this particular use case it
doesn't make a difference (besides slightly lowering the chance of a
collision) switch to sha256sum to make sure our code doesn't depend on
tools that are deprecated and there is an effort to remove.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit 027b508da8)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Similar to the deny x permission tests, the tests that test carving
out r permissions need to be updated to be conditional on what
priority is being used on the rule.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit bf7b80c478)
Signed-off-by: John Johansen <john.johansen@canonical.com>
With priority rules, deny does not carve out permissions from the
higher priority rule. Technically it doesn't from lower priority either
as it completely overrides them, but that case already results in
an inequality so does not cause the tests to fail.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit 25f16b239d)
Signed-off-by: John Johansen <john.johansen@canonical.com>
cx rules using a specified profile transition, may be emulated by
using px and a hierarchical profile name. That is
cx -> b
may be transformed into
px -> profile//b
which will generate an xtable entry of
profile//b
which means the previous patch using
pivot_root -> b,
to reliably add b to the xtable will not cover this case.
transition to using two pivot_root rules to provide the xtable entries
pivot_root /a -> b,
pivot_root /c -> /t//b,
the paths /a and /c are irrelavent as long as they don't have an
overlap with the generic globbing expression in the test, Two table
entries will be generated. We guarantee no overlap by converting the
/** to /f**
Also the xtable reserving rules are moved to the end of the profile so
the table order can be reliably created. A follow on MR around xtable
improvements should add reliability to xtable order.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit 369029dc07)
Signed-off-by: John Johansen <john.johansen@canonical.com>
exec rules that specify an specific target profile generate an entry
in the xtable. The test entries containing " -> b" are an example of
this.
Currently the parser allocates the xtable entry before priorities are
applied in the backend, or minimization is done. Further more the
parser does not ref count the xtable entry to know what it is no
longer referenced.
The equality tests generate rules that are designed to completely
override and remove a lower priority rule, and remove it. Eg.
/t { priority=1 /* ux, /f px -> b, }
and then compares the generated profile to the functionaly equivalent
profile eg.
/t { priority=1 /* ux, }
To verify the overridden rule has been completely removed.
Unfortunately the compilation is not removing the unused xtable entry
for the specified transition, causing the equality comparison to fail.
Ideally the parser should be fixed so unused xtable entries are removed,
but that should be done in a different MR, and have its own test.
To fix the current tests, and another rule that adds an xtable entry
to the same target that can not be overriden by the x rule using
pivot_root. The parser will dedup the xtable entry resulting in the
known and test profile both having the same xtable. So the test will
pass and meet the original goal of verifying the x rule being overriden
and eliminated.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit 84650beb2f)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Failed equality tests can be hard to debug. The profiles aren't always
enough to figure out what is going on. Add several options that will
help in debugging, and developing new tests.
Add switches and arg parsing.
Add the ability to run tests individually
Add a -r flag to allow retaining the test and output
similar to the regression tests, so the exact output from the
tests can be examined.
Add a -d flag to dump dfa build information.
Allow overriding the parser, features, and description for a given
test run.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit cca842b897)
Signed-off-by: John Johansen <john.johansen@canonical.com>
printf of failure/error info should be going to stderr. Unfortunately
the test has a mix of 2>&1 and 1>&2. Having a mix is just wrong, we
could standardize on either but since the info is error info 1>&2
seems to be the better choice.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit 31e60baab2)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The test was passing because the file priority was always zero bug
resulting in the priority rule always being correctly combined
with the specific match x rule, instead of overriding it.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit 57c57f198c)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The test was passing because the file priority always being zero bug,
the supplied rule always had the same priority as the implied
rule. Resulting in binary_equality always passing even though the
specified priority should have resulted in a failure.
Fix this by checking if the priorities are equal to the implied
rule other wise it should result in an inequality.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit 4b410b67f1)
Signed-off-by: John Johansen <john.johansen@canonical.com>
When there is a failure output the exact call info used to invoke the
parser. To facilitate manually recreating the test.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit d275dfdd42)
Signed-off-by: John Johansen <john.johansen@canonical.com>
With the file priority fix the xequality (expected equal but known
failure) tests are now passing. So convert them to regular equality
tests.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit fcee32a37e)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The dfa goes through several stages during the build. Allow dumping it
at the various stages instead of only at the end.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit 5d2a38e816)
Signed-off-by: John Johansen <john.johansen@canonical.com>
File rules could drop priority info when rule matched a rule
that was the same except for having different priority. For now
fix this by treating them as a different rule.
The priority was also be dropped when add_prefix was used to
add the priority during the parse resulting in file rules always
getting a default priority of 0.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit 9d5b86bc9d)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Starting with Python 3.8, you can use the PYTHONPYCACHEPREFIX environment
variable to define a cache directory for Python [1]. I think most people would set
this dir to @{HOME}/.cache/python/ , so the python abstraction should allow
writing to this location.
[1]: https://docs.python.org/3/using/cmdline.html#envvar-PYTHONPYCACHEPREFIX
(cherry picked from commit 03b5a29b05)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Depending on the system, copying echo to the loop device fails because the echo binary is too large.
Especially on systems that have echo be just a symlink to coreutils (e.g. busybox) (as opposed to echo being its own binary) 16k is just not enough.
2M seems fine on my system, but this might need yet a higher value depending on what coreutils other people actually run.
The crash in question:
```
cp: error writing '/tmp/sdtest.3937422-31490-Bxvi6g/mount_target/echo': No space left on device
Fatal Error (file_unbindable_mount): Unexpected shell error. Run with -x to debug
rm: cannot remove '/tmp/sdtest.3937422-31490-Bxvi6g/mount_target': Device or resource busy
```
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1469
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 8e431ebcd9)
Depending on the system, copying echo to the loop device fails because the echo binary is too large.
Especially on systems that have echo be just a symlink to coreutils (e.g. busybox) 16k is just not enough.
2M seems fine on my system, but this might need yet a higher value depending on what coreutils other people actually run.
The actual loop device needs to be larger to properly fit the allocated file size. Testing shows 4M is sufficient, but this is basically arbitrary.
(cherry picked from commit 1cc2a3bd86)
Signed-off-by: John Johansen <john.johansen@canonical.com>
- previously, aa-status --json --show profiles would return non-standard json
- adding the --pretty flag would crash completely
- closes#470
Things done:
- removed trailing ", " in json generation
- generate json seperator (", ") for each new json field
(profiles/processes) after the header if json is enabled
Tested on NixOS and apparmor 4.0.3 base, but should work on any version the patch applies on.
Closes#470
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1451
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit c489631770)
- previously, aa-status --json --show profiles would return non-standard json
- adding the --pretty flag would crash completely
- closes#470
Things done:
- removed trailing ", " in json generation
- generate json seperator (", ") for each new json field
(profiles/processes) after the header if json is enabled
Tested on NixOS and apparmor 4.0.3 base, but should work on any version the patch applies on.
(cherry picked from commit 4f006a660c)
Signed-off-by: John Johansen <john.johansen@canonical.com>
This MR is meant to resolve warnings such as "Warning: execname '/home/username/Documents/apparmor/tests/regression/apparmor/file_unbindable_mount': no such file or directory" when running tests like the one in the current version of !1448.
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1450
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 59957aa1d8)
When settest was called with two parameters, one for the test name and
the other for the test wrapper/binary, the profile created with
genprofile would show the test name, causing an error if the file
didn't exist.
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
(cherry picked from commit b4adff2ce0)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Some of the tests using the --stdin option of mkprofile.pl are adding
more than one profile at a time. Whenever a profile is created in the
test, its name is added to the file profile.names so the test
infrastructure can tell if the profile is loaded or removed when
appropriately. The issue is that the name of the second profile
created by --stdin is not added, so these checks are not applied.
This patch adds the option of appending a second profile (not rules).
The option --append was used instead of a short -A because the short
options are arguments of mkprofile.pl, which --append is not.
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
(cherry picked from commit 0307619ed9)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Due to how the tests were implemented in the past, permissions could
be passed along with the image name, and the permission part would be
discarded. The issue is that permissions are usually separated by ':',
but namespaces also contain ':', which would cause a conflict.
Since permissions are no longer passed as part of the image name,
remove that description so profile names in namespaces can be
supported.
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
(cherry picked from commit 9cc40e2dca)
Signed-off-by: John Johansen <john.johansen@canonical.com>
While the mount syscall documentation disallows this, the kernel silently
ignores make-* flags when doing a remount, and real applications were
passing this conflicting set of flags. Because changing the kernel to
reject this combination would break userspace, we should allow them
instead.
For an example: see https://bugs.launchpad.net/apparmor/+bug/2091424.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit 52babe8054)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The previous code would concatenate all of them together without spacing.
While dump_flags and the corresponding operator<< function aren't currently used,
this will help for when dump_flags is used to debug parser problems.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
(cherry picked from commit 96718ea4d1)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Swap on ZFS is *weird*. Getting it working needs some special casing, see e.g. https://askubuntu.com/questions/1198903/can-not-use-swap-file-on-zfs-files-with-holes
Currently, the swap regression test fails on my system (with /tmp in zfs):
```bash
tests/regression/apparmor ❯ ./swap.sh
Error: swap failed. Test 'SWAPON (unconfined)' was expected to 'pass'. Reason for failure 'FAIL: swapon /tmp/sdtest.872368-19048-kN4FN2/swapfile failed - Invalid argument'
Error: swap failed. Test 'SWAPOFF (unconfined)' was expected to 'pass'. Reason for failure 'FAIL: swapoff /tmp/sdtest.872368-19048-kN4FN2/swapfile failed - Invalid argument'
swapon: /tmp/sdtest.872368-19048-kN4FN2/swapfile: skipping - it appears to have holes.
Fatal Error (swap): Unexpected shell error. Run with -x to debug
```
However, just doing a file mount does make the test work on zfs, similar to how it is done with tmpfs. This means we don't need any special-casing for zfs beyond what is already there for working around (similar) tmpfs limitations.
Also, while researching this, it is possible a similar patch is needed for btrfs, but i currently don't have an easy way to test that.
This is non-breaking for anyone *not* using zfs, and it is currently broken with zfs anyways.
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1462
Approved-by: Ryan Lee <rlee287@yahoo.com>
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: Georgia Garcia <georgia.garcia@canonical.com>
(cherry picked from commit e8f1ac4791)
previously, this check would fail if the setuptools version would contain non-integers.
On my system, that is the case: `setuptools.__version__` is `'75.1.0.post0'`
I believe it is entirely fair to just check the relevant bits and refuse to continue if those can not be checked properly.
Having some extra slug on the version should not immediately cause issues (e.g. the `post0` here, or slugs like `beta`, `alpha` and the likes).
Probably only very few systems are running setuptools with weird version info, but supporting this is a simple one-line change i figured i might as well MR.
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1460
Approved-by: Christian Boltz <apparmor@cboltz.de>
Merged-by: Christian Boltz <apparmor@cboltz.de>
(cherry picked from commit b3de4ef022)
previously, this check would fail if the setuptools version would contain non-integers.
On my system, that is the case: `setuptools.__version__` is `'75.1.0.post0'`
I believe it is entirely fair to just check the relevant bits and refuse to continue if those can not be checked properly.
But haviong something extra on the version should not immediately cause issues (e.g. the `post0` here, or slugs like `beta`, `alpha` and the likes).
Probably only very few systems are running setuptools with weird version info, but supporting this doesn't cost much, i believe.
(cherry picked from commit 3302ae98e4)
Signed-off-by: John Johansen <john.johansen@canonical.com>
MS_SYNC is a flag for msync(2) while MS_SYNCHRONOUS is a flag for mount(2).
The header used to define MS_SYNC but IMO this is confusing since that's an
unrelated flag.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit d164e877f5)
Signed-off-by: John Johansen <john.johansen@canonical.com>
By placing a bzImage into the top level of the AppArmor git repository one can
instruct spread and image-garden to use that image instead of booting
traditionally with an EFI / full disk image pair.
In addition, make error handling in qemu more robust, so failures are both
surfaced and do not cause endless attempts to allocate.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1452
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 239ae21b69)
By placing a bzImage into the top level of the AppArmor git repository one can
instruct spread and image-garden to use that image instead of booting
traditionally with an EFI / full disk image pair.
In addition, make error handling in qemu more robust, so failures are both
surfaced and do not cause endless attempts to allocate.
Please update image-garden to at least 5a00ead9964df6463e19432ae50e7760fc6da755
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit 7031b5aeee)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The test adds a very small and simple smoke test that shows that a mount rule
with both fstype and options allows mounts to be performed on a real running
kernel.
The test is structured in a way that should make it easy to extend with new
variants (flags, fstype) in the future.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit 1f60021979)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Spread is a full-system, or integration test suite runner initially developed
to test snapd. Over time it has spread to other projects where it provides a
structured way to organize, run and debug complex full-system interactions.
Spread is documented on https://github.com/canonical/spread and is used in
production since late 2016.
Spread has a notion of backends which are responsible for allocating and
discarding test machines. For the purpose of running AppArmor regression tests,
I've combined spread with my own tool, image garden. The tool provides
off-the-shelf images, constructed on-the-fly from freely available images, and
makes them easily available to spread.
The reason for doing it this way is so that using non-free cloud systems is not
required and anyone can repeat the test process locally, on their own computer.
Vanilla spread is somewhat limited to x86-64 systems but the way I've used it
here makes it equally possible to test x86_64 *and* aarch64 systems. I've done
most of the development on an ARM single-board-computer running on my desk.
Spread requires a top-level spread.yaml file and a collection of task.yaml
files that describe individual tasks (for us, those are just tests). Tasks have
no implied dependency except that to reach a given task, spread will run all
the _prepare_ statements leading to that task, starting from the project, test
suite and then task. With proper care one can then run a specific individual
test with a one-line command, for example:
```
spread -v garden:ubuntu-cloud-24.04:tests/regression/apparmor:at_secure
```
This will prepare a fresh ubuntu-cloud-24.04 system (matching the CPU
architecture of the host), copy the project tree into the test machine, install
all the build dependencies, build all the parts of apparmor and then run one
specific variant of the regression test, namely the at_secure program.
Importantly the same test can also run on, say debian-cloud-13 (Debian Trixie),
but also, if you have a Google cloud account, on Google Compute Engine or in
one of the other backends either built into spread or available as a fork of
spread or as a helper for ad-hoc backend. Spread can also create more than one
worker per system and distribute the tests to all of the available instances.
In no way are we locking ourselves out of the ability to run our test suite on
our target of choice.
Spread has other useful switches, such as:
- `-reuse` for keeping machines around until discarded with -discard
- `-resend` for re-sending updated copy of the project (useful for -reuse)
- `-debug` for starting an interactive shell on any failure
- `-shell` for starting an interactive shell instead of the `execute` phase
This first patch contains just the spread elements, assuming that both spread
and image-garden are externally installed. A GitLab continuous integration
installing everything required and running a subset of tests will follow
shortly.
I've expanded the initial selection of systems to allow running all the tests
on several versions of Ubuntu, Debian and openSUSE, mainly as a sanity check
but also to showcase how practical spread is at covering real-world systems.
A number of tests are currently failing:
- garden:debian-cloud-12:tests/regression/apparmor:attach_disconnected
- garden:debian-cloud-12:tests/regression/apparmor:deleted
- garden:debian-cloud-12:tests/regression/apparmor:unix_fd_server
- garden:debian-cloud-12:tests/regression/apparmor:unix_socket_pathname
- garden:debian-cloud-13:tests/regression/apparmor:attach_disconnected
- garden:debian-cloud-13:tests/regression/apparmor:deleted
- garden:debian-cloud-13:tests/regression/apparmor:unix_fd_server
- garden:debian-cloud-13:tests/regression/apparmor:unix_socket_pathname
- garden:opensuse-cloud-15.6:tests/regression/apparmor:attach_disconnected
- garden:opensuse-cloud-15.6:tests/regression/apparmor:deleted
- garden:opensuse-cloud-15.6:tests/regression/apparmor:e2e
- garden:opensuse-cloud-15.6:tests/regression/apparmor:unix_fd_server
- garden:opensuse-cloud-15.6:tests/regression/apparmor:unix_socket_pathname
- garden:opensuse-cloud-15.6:tests/regression/apparmor:xattrs_profile
In addition, only on openSUSE, I've skipped the entire test suite of the utils
directory, as it requires python3 ttk themes, which I cannot find in packaged
form.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1432
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: Georgia Garcia <georgia.garcia@canonical.com>
(cherry picked from commit d9304c7653)
- Tests defined in utils/test are now described by a task.yaml in the same
directory and can run concurrently across many machines.
- Tests for utils/ are now executed on openSUSE Tumbleweed since ttk themes is
no longer a hard dependency in master.
- Tests no longer run on openSUSE Leap 15.6 due to the age of default
Python (3.6) and gcc/g++. The tight integration with SWIG which does
not seem to support other Python versions very well. Perl hard-codes
old GCC for extension modules. The upcoming openSUSE Leap 16 should be
a viable target. In the meantime we can still test everything through
rolling-release Tumbleweed.
- Formatting of YAML files is now more uniform, at four spaces per tab.
- The run-spread.sh script is now in the root of the tree. The script allows
running all spread tests sequentially on one system, while collecting logs
and artifacts for convenient analysis after the fact.
- All systems are adjusted to run _four_ workers in parallel with _two_ virtual
cores each and equipped with 1.5GB of virtual memory. This aims to best
utilize the capacity of a typical CI worker with two to four cores and about
8GB of available memory.
- Failing tests are marked as such, so that as a whole the entire spread suite
can pass and be useful at catching regressions.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit 1df91e2c8c)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Compared to v1 the following improvements have been made:
- The cost of installing packages have been shifted from each startup to image
preparation phase, thanks to the integration of custom cloud-init profiles
into image-garden. This has dramatic impact on iteration time while also
entirely removing requirement to be online to run once a prepared image is
available.
- Support for running on Google Compute Engine has been removed since it would
not be able to use cloud-init the same way would currently only complicate
setup.
- The number of workers have been tuned for local iteration, aiming for
comfortable work with 16GB of memory on the host. Once CI/CD pipeline
support is introduced I will add a dedicated entry so that resources are
utilized well both locally and when running in CI.
- The set of regression tests listed in tests/regression/apparmor/task.yaml is
now cross-checked so introduction of a new test to the makefile there is
automatically flagged and causes spread to fail with a clear message.
- The task tests/unit/utils has been improved to generate profiles. Thanks to
Christian Boltz for explaining this relationship between tests.
- A number of comments have been improved and cleaned up for readability,
accuracy and sometimes better grammar.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit c95ac4d350)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Spread is a full-system, or integration test suite runner initially developed
to test snapd. Over time it has spread to other projects where it provides a
structured way to organize, run and debug complex full-system interactions.
Spread is documented on https://github.com/canonical/spread and is used in
production since late 2016.
Spread has a notion of backends which are responsible for allocating and
discarding test machines. For the purpose of running AppArmor regression tests,
I've combined spread with my own tool, image garden. The tool provides
off-the-shelf images, constructed on-the-fly from freely available images, and
makes them easily available to spread.
The reason for doing it this way is so that using non-free cloud systems is not
required and anyone can repeat the test process locally, on their own computer.
Vanilla spread is somewhat limited to x86-64 systems but the way I've used it
here makes it equally possible to test x86_64 *and* aarch64 systems. I've done
most of the development on an ARM single-board-computer running on my desk.
Spread requires a top-level spread.yaml file and a collection of task.yaml
files that describe individual tasks (for us, those are just tests). Tasks have
no implied dependency except that to reach a given task, spread will run all
the _prepare_ statements leading to that task, starting from the project, test
suite and then task. With proper care one can then run a specific individual
test with a one-line command, for example:
```
spread -v garden:ubuntu-cloud-24.04:tests/regression/apparmor:at_secure
```
This will prepare a fresh ubuntu-cloud-24.04 system (matching the CPU
architecture of the host), copy the project tree into the test machine, install
all the build dependencies, build all the parts of apparmor and then run one
specific variant of the regression test, namely the at_secure program.
Importantly the same test can also run on, say debian-cloud-13 (Debian Trixie),
but also, if you have a Google cloud account, on Google Compute Engine or in
one of the other backends either built into spread or available as a fork of
spread or as a helper for ad-hoc backend. Spread can also create more than one
worker per system and distribute the tests to all of the available instances.
In no way are we locking ourselves out of the ability to run our test suite on
our target of choice.
Spread has other useful switches, such as:
- `-reuse` for keeping machines around until discarded with -discard
- `-resend` for re-sending updated copy of the project (useful for -reuse)
- `-debug` for starting an interactive shell on any failure
- `-shell` for starting an interactive shell instead of the `execute` phase
This first patch contains just the spread elements, assuming that both spread
and image-garden are externally installed. A GitLab continuous integration
installing everything required and running a subset of tests will follow
shortly.
I've expanded the initial selection of systems to allow running all the tests
on several versions of Ubuntu, Debian and openSUSE, mainly as a sanity check
but also to showcase how practical spread is at covering real-world systems.
A number of systems and tests are currently failing:
- garden:debian-cloud-12:tests/regression/apparmor:attach_disconnected
- garden:debian-cloud-12:tests/regression/apparmor:deleted
- garden:debian-cloud-12:tests/regression/apparmor:unix_fd_server
- garden:debian-cloud-12:tests/regression/apparmor:unix_socket_pathname
- garden:debian-cloud-13:tests/regression/apparmor:attach_disconnected
- garden:debian-cloud-13:tests/regression/apparmor:deleted
- garden:debian-cloud-13:tests/regression/apparmor:unix_fd_server
- garden:debian-cloud-13:tests/regression/apparmor:unix_socket_pathname
- garden:opensuse-cloud-15.6:tests/regression/apparmor:attach_disconnected
- garden:opensuse-cloud-15.6:tests/regression/apparmor:deleted
- garden:opensuse-cloud-15.6:tests/regression/apparmor:e2e
- garden:opensuse-cloud-15.6:tests/regression/apparmor:unix_fd_server
- garden:opensuse-cloud-15.6:tests/regression/apparmor:unix_socket_pathname
- garden:opensuse-cloud-15.6:tests/regression/apparmor:xattrs_profile
- garden:opensuse-cloud-tumbleweed:tests/regression/apparmor:attach_disconnected
- garden:opensuse-cloud-tumbleweed:tests/regression/apparmor:deleted
- garden:opensuse-cloud-tumbleweed:tests/regression/apparmor:unix_fd_server
- garden:opensuse-cloud-tumbleweed:tests/regression/apparmor:unix_socket_pathname
- garden:ubuntu-cloud-22.04:tests/regression/apparmor:attach_disconnected
In addition, only on openSUSE, I've skipped the entire test suite of the utils
directory, as it requires python3 ttk themes, which I cannot find in packaged
form.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit cc04181578)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The new check-one-test-% pattern rule allows running individual test scripts.
This allows them to be tested in parallel across many Make worker threads or
across many distinct machines with spread.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit 9588b06e0f)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Tests #466 but is marked as expected fail due to that bug not being resolved.
Depends on !1441 which adds the xfail infrastructure to the parser equality testing framework, and should be rebased on top of master once that MR is merged.
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1443
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit e1d8bf1888)
currently the equality tests require the tests to PASS as known equality
or inequality. Add the ability to add tests that are a known problem
and are expected to fail the equality, or inequality test.
This is done by using
verify_binary_xequality
verify_binary_xinequality
This allows new tests to be added to document a known issue, without
having to develop the fix for the issue. The use of this facility
is expected to be temporary, so any test marked as xequality or
xinequality will be noisy but not fail the other tests until they
are fixed, at which point they will cause the tests to fail to
force them to be updated to the correct equality or inequality
test.
Signed-off-by: John Johansen <john.johansen@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1441
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 53e322b755)
currently the equality tests require the tests to PASS as known equality
or inequality. Add the ability to add tests that are a known problem
and are expected to fail the equality, or inequality test.
This is done by using
verify_binary_xequality
verify_binary_xinequality
This allows new tests to be added to document a known issue, without
having to develop the fix for the issue. The use of this facility
is expected to be temporary, so any test marked as xequality or
xinequality will be noisy but not fail the other tests until they
are fixed, at which point they will cause the tests to fail to
force them to be updated to the correct equality or inequality
test.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit b81ea65c1c)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The regression test suite uses root with capabilities restricted in
several tests. This can cause the test suite to fail in weird and
confusing ways.
Add a test to check for DAC permissiosns from / to the testsuite
and abort running the tests with an error message if DAC permissions
are going to cause the test suite to fail.
Currently the test is pretty basic, but is better than nothing.
Signed-off-by: John Johansen <john.johansen@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1411
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit c5b17d85ea)
The regression test suite uses root with capabilities restricted in
several tests. This can cause the test suite to fail in weird and
confusing ways.
Add a test to check for DAC permissiosns from / to the testsuite
and abort running the tests with an error message if DAC permissions
are going to cause the test suite to fail.
Currently the test is pretty basic, but is better than nothing.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit 82e4b4ba00)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Using this version of make:
```
GNU Make 4.2.1
Built for x86_64-suse-linux-gnu
```
I'm not entirely sure why but the alternative syntax I've used works correctly.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit 4caf0aff81)
Signed-off-by: John Johansen <john.johansen@canonical.com>
This fixes the test to pass on openSUSE Tumbleweed, where the small size
prevented alloction of an inode for the `lost+found` directory:
```
garden:opensuse-cloud-tumbleweed .../tests/regression/apparmor# mkfs.ext2 -F -m 0 -N 10 /tmp/sdtest.32929-21402-6x826m/image.ext3
mke2fs 1.47.0 (5-Feb-2023)
Discarding device blocks: done
Creating filesystem with 512 1k blocks and 8 inodes
Allocating group tables: done
Writing inode tables: done
ext2fs_mkdir: Could not allocate inode in ext2 filesystem while creating /lost+found
```
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit 32ee85cef8)
Signed-off-by: John Johansen <john.johansen@canonical.com>
This fixes an error with Python 3.11:
```
test/test-parser-simple-tests.py:420:21: E502 the backslash is redundant between brackets
```
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit 92fcdcab9e)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Which is technically not POSIX and command -v works everywhere. This fixes
building and running the test suite on openSUSE Tumbleweed.
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit 4b0adc63f5)
Signed-off-by: John Johansen <john.johansen@canonical.com>
On a test system without bison installed, make setup fails with:
/bin/sh: 1: bison: not found
/bin/sh: 1: test: -ge: unexpected operator
Signed-off-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
(cherry picked from commit f58fe9cd52)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The file was quoted with the following space, making the test broken.
Signed-off-by: Zygmunt Krynicki <me@zygoon.pl>
(cherry picked from commit 8c16fb2700)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Match Flags convert output to hex but don't restore after outputting
the flag resulting in following numbers being hex encoded. This
results in dumps that can be confusing eg.
rule: \d2 -> \x2 priority=1001 (0x4/0)< 0x4>
rule: \d7 -> \a priority=3e9 (0x4/0)< 0x4>
rule: \d10 -> \n priority=3e9 (0x4/0)< 0x4>
rule: \d9 -> \t priority=3e9 (0x4/0)< 0x4>
rule: \d14 -> \xe priority=1001 (0x4/0)< 0x4>
where priority=3e9 is the hex encoded priority 1001.
Signed-off-by: John Johansen <john.johansen@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1419
Approved-by: Maxime Bélair <maxime.belair@canonical.com>
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 2e77129e15)
Match Flags convert output to hex but don't restore after outputting
the flag resulting in following numbers being hex encoded. This
results in dumps that can be confusing eg.
rule: \d2 -> \x2 priority=1001 (0x4/0)< 0x4>
rule: \d7 -> \a priority=3e9 (0x4/0)< 0x4>
rule: \d10 -> \n priority=3e9 (0x4/0)< 0x4>
rule: \d9 -> \t priority=3e9 (0x4/0)< 0x4>
rule: \d14 -> \xe priority=1001 (0x4/0)< 0x4>
where priority=3e9 is the hex encoded priority 1001.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit a31343c5f7)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Most `tests/regression/apparmor/*.sh` scripts contain
. $bin/prologue.inc
This will explode if one of the parent directories contains a space.
Minimized reproducer:
```
# cat test.sh
pwd=`dirname $0`
pwd=`cd $pwd ; /bin/pwd`
bin=$pwd
echo "pwd: $bin"
. $bin/prologue.inc
# ./test.sh
pwd: /tmp/foo bar
./test.sh: line 9: /tmp/foo: No such file or directory
```
Notice that test.sh tries to source `/tmp/foo` instead of `/tmp/foo bar/prologue.inc`.
The fix is to quote the prologue.inc path:
. "$bin/prologue.inc"
While on it, also fix other uses of $bin - directly and indirectly - by quoting them.
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1418
Approved-by: Ryan Lee <rlee287@yahoo.com>
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit a422d2ea17)
... to avoid issues with spaces in a parent directory's name.
"Indirect uses" means usage of $bin via another variable, for example
`foo=$bin/whatever`
(cherry picked from commit 55db4af979)
Signed-off-by: John Johansen <john.johansen@canonical.com>
... to avoid issues with spaces in a parent directory's name.
(cherry picked from commit 22cf88b7c7)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Most `tests/regression/apparmor/*.sh` scripts contain
. $bin/prologue.inc
This will explode if one of the parent directories contains a space.
Minimized reproducer:
```
pwd=`dirname $0`
pwd=`cd $pwd ; /bin/pwd`
bin=$pwd
echo "pwd: $bin"
. $bin/prologue.inc
pwd: /tmp/foo bar
./test.sh: line 9: /tmp/foo: No such file or directory
```
Notice that test.sh tries to source `/tmp/foo` instead of `/tmp/foo bar/prologue.inc`.
The fix - as done in this commit - is to quote the prologue.inc path:
. "$bin/prologue.inc"
(cherry picked from commit e1972eb22f)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Instead of encoding permissions in the accept and accept2 tables
extended perms uses a permissions table and accept becomes an index
into the table.
Add the ability to dump the permissions table so that it can be
compared and debugged.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit 45964d34e7)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The chfa dump is missing information about the accept2 entry. The
accept2 information is necessary to help with debugging state machine
builds as accept2 is used to store quiet and audit information in the
old format or conditional information in the extended perms format.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit 00dedf10ad)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The Makefile is missing some of its .h depenedncies causing compiles
to either fail or worse result in bad builds when rebuilding in an
already built tree.
Move the header dependencies into a variable and use it for each
target. While some targets don't need every include as a dependency
and this will result in unnecessary rebuilds in some cases, it makes
the Makefile cleaner, easier to maintain and makes sure a dependency
isn't accidentally missed.
Signed-off-by: John Johansen <john.johansen@canonical.com>
(cherry picked from commit 7cc7f47424)
Signed-off-by: John Johansen <john.johansen@canonical.com>
The parser recently changed how/where deny information is applied.
commit 1fa45b7c1 ("parser: dfa minimization prepare for extended
permissions") removed the implicit filtering of explicit denies during
the minimization pass. The implicit clear allowed the explicit
information to be carried into the minimization pass and merged with
implicit denies. The end result being a minimized dfa with the explicit
deny information available to be applied post minimization, and
then dropped later at permission encoding in the accept entries.
Extended permission however enable carrying explicit deny information
into the kernel to fix certain bugs like complain mode not being
able to distinguish between implicit and explicit deny rules (ie.
deny rules get ignored in complain mode). However keeping explicit
deny information when unnecessary result in a larger state machine
than necessary and slower compiles.
commit 179c1c1ba ("parser: fix minimization check for filtering_deny")
Moved the explicit apply_and_clear_deny() pass to before minimization
to restore mnimization's ability to create a minimized dfa with
explicit and implicit deny information merged but this also cleared
the explicit deny information that used to be carried through
minimization. This meant that when the deny information was applied
post minimization it resulted in the audit and quiet information
being cleared.
This resulted in the query_label tests failing as they are checking
for the expected audit infomation in the permissions.
Fixes: 179c1c1ba ("parser: fix minimization check for filtering_deny")
Bug: https://gitlab.com/apparmor/apparmor/-/issues/461
Signed-off-by: John Johansen <john.johansen@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1408
Approved-by: Ryan Lee <rlee287@yahoo.com>
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit eb365b374d)
Signed-off-by: John Johansen <john.johansen@canonical.com>
If a user specifies a non-existing file to merge into the profiles
(`aa-mergeprof /file/not/found`), this results in a backtrace showing an
AppArmorBug because that file unsurprisingly doesn't end up in the
active_profiles filelist.
Handle this more gracefully by adding a read_error_fatal parameter to
read_profile() that, if set, forwards the exception. With that,
aa-mergeprof doesn't try to list the profiles in this non-existing file.
Note that all other callers of read_profile() continue to ignore read
errors, because aborting just because a single file in /etc/apparmor.d/
(for example a broken symlink) isn't readable would be a bad idea.
This bug was introduced in 4e09f315c3, therefore I propose this patch for 3.0..master
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1403
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 5ebbe788ea)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Bash will try to read the passwd database to find the shell of a user if
$SHELL is not set. This causes zgrep to trigger
```
apparmor="DENIED" operation="open" class="file" profile="zgrep" name="/etc/nsswitch.conf" comm="zgrep" requested_mask="r" denied_mask="r" fsuid=0 ouid=0
apparmor="DENIED" operation="open" class="file" profile="zgrep" name="/etc/passwd" comm="zgrep" requested_mask="r" denied_mask="r" fsuid=0 ouid=0
```
if called in a sanitized environment. As the functionality of zgrep is
not impacted by a limited Bash environment, add deny rules to avoid the
potentially misleading AVC messages.
Signed-off-by: Georg Pfuetzenreuter <mail@georg-pfuetzenreuter.net>
(cherry picked from commit 48483f2ff8)
Signed-off-by: John Johansen <john.johansen@canonical.com>
Instead of always storing the name of the main profile, store the child
profile/hat name if we are in a child profile or hat.
As a result, we always get the correct "profile xy" header even for
child profiles when dumping the ProfileStorage object.
Also extend the tests to check that the name gets stored correctly.
(cherry picked from commit cb943e4efc)
Signed-off-by: John Johansen <john.johansen@canonical.com>
This is neeed for "inherit owner = yes" in smb.conf.
From man smb.conf:
inherit owner (S)
The ownership of new files and directories is normally governed by
effective uid of the connected user. This option allows the Samba
administrator to specify that the ownership for new files and
directories should be controlled by the ownership of the parent
directory.
Fixes: https://bugzilla.suse.com/show_bug.cgi?id=1234327
I propose this fix for 3.x, 4.x and master.
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1456
Approved-by: Ryan Lee <rlee287@yahoo.com>
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit a315d89a2b)
d3050285 smbd: allow capability chown
Co-authored-by: John Johansen <john@jjmx.net>
I am upstreaming this patch that is part of the nix package of apparmor for close to a year now.
This fixes the issue at https://github.com/NixOS/nixpkgs/issues/273164 for more distros than just NixOS.
The original merge Request on the nix side patching this was https://github.com/NixOS/nixpkgs/pull/285915.
However, people had issues with gitlab, so this never hit apparmor upstream until now. This does however also mean this patch has seen production and seems to work quite well.
## Original reasoning/message of the patch author:
This check is intended for ensuring that the profiles file can actually
be opened. The *actual* check is performed by the shell, not the read
utility, which won't even be executed if the input redirection (and
hence the test) fails.
If the test succeeds, though, using `read` here might actually
jeopardize the test result if there are no profiles loaded and the file
is empty.
This commit fixes that case by simply using `true` instead of `read`.
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1438
Approved-by: Christian Boltz <apparmor@cboltz.de>
Merged-by: Christian Boltz <apparmor@cboltz.de>
(cherry picked from commit 93c7035148)
b4aa00de aa-remove-unknown: fix readability check
Co-authored-by: Christian Boltz <apparmor@cboltz.de>
Seen on various VMs, my guess is that bash wants to translate a uid to a
username.
Log events (slightly shortened)
apparmor="DENIED" operation="open" class="file" profile="zgrep" name="/etc/nsswitch.conf" comm="zgrep" requested_mask="r" denied_mask="r" fsuid=0 ouid=0
apparmor="DENIED" operation="open" class="file" profile="zgrep" name="/etc/passwd" comm="zgrep" requested_mask="r" denied_mask="r" fsuid=0 ouid=0
I propose this patch for 3.0..master
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1357
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit ab16377838)
Signed-off-by: John Johansen <john.johansen@canonical.com>
When the find fails but the insertion also fails, we leak the new node
that we generated. Delete the new node in this case to avoid leaking
memory.
The question remains, however, as to whether we should implement `operator==` in addition to `operator<` so that they are consistent with each other and `find` works correctly.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1399
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit 99261bad11)
Signed-off-by: John Johansen <john.johansen@canonical.com>
As I have read multiple MR mentioning the `nameservice-strict`. Therefore, I thought it would make sense to directly import it here.
To give some context, this abstraction is probably the most commonly included abstraction (after `base`). In `apparmor.d`, it is used by over 700 profiles (only counting direct import). Therefore, adding new rules can have an important impact over a lot of profiles.
Note: the abstraction is a direct import from https://gitlab.com/roddhjav/apparmor.d. The license is the same, I obviously kept Morfikov copyright line. However, I am not sure either or not the SPDX identifier can be used here.
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1368
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Approved-by: Christian Boltz <apparmor@cboltz.de>
Approved-by: Ryan Lee <rlee287@yahoo.com>
Merged-by: Georgia Garcia <georgia.garcia@canonical.com>
(cherry picked from commit 68376e7fee)
Signed-off-by: John Johansen <john.johansen@canonical.com>
commit 1fa45b7c1 ("parser: dfa minimization prepare for extended
permissions") removed implicit filtering of explicit denies in the
minimization pass (the information was ignored in building the set of
final accept states).
The filtering of explicit denies reduces the size of the produced
dfa. Since we need to be smarter about when explicit denies are
kept (eg. during complain mode), and most dfas are limited to 65k
states we currently need to filter explicit deny perms by default.
To compensate commit 2737cb2c2 ("parser: minimization - remove
unnecessary second minimization pass") moved the
apply_and_clear_deny() to before minimization. However its check to
apply removal denials before minimization is broken. Remove minimization
triggering apply_and_clear_deny() and just set the FILTER_DENY flag
by default, until we have better selection of rules/conditions where
explicit deny information should be carried through to the backend.
Fixes: 2737cb2c2 ("parser: minimization - remove unnecessary second minimization pass")
Signed-off-by: John Johansen <john.johansen@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1397
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit e9d6e0ba14)
Signed-off-by: John Johansen <john.johansen@canonical.com>
io_uring and userns mediation are encoding permissions on the class
byte. This is a mistake that should never have been allowed.
With the addition of rule priorities the class byte mediates rule,
that ensure the kernel can determine a class is being mediated is
given the highest priority possible, to ensure class mediation can not
be removed by a deny rule. See
61b7568e1 ("parser: bug fix mediates_X stub rules.")
for details.
Unfortunately this breaks rule classes that encode permissions on the
class byte, because those rules will always have a lower priority and
the class mediates rule will always be selected over them resulting in
only the class mediates permission being on the rule class state.
Fix this by adding the mediaties class rules for these rule classes
with the lowest priority possible. This means that any rule mediating
the class will wipe out the mediates class rule. So add a new mediates
class rule at the same priority, as the rule being added.
This is a naive implementation and does result in more mediates rules
being added than necessary. The rule class could keep track of the
highest priority rule that had been added, and use that to reduce the
number of mediates rules it adds for the class.
Technically we could also get away with not adding the rules for allow
rules, as the kernel doesn't actually check the encoded permission but
whether the class state is not the trap state. But it is required with
deny rules to ensure the deny rule doesn't result in permissions being
removed from the class, resulting in the kernel thinking it is
unmediated. We also want to ensure that mediation is encoded for other
rule types like prompt, and in the future the kernel could check the
permission so we do want to guarantee that the class state has the
MAY_READ permission on it.
Note: there is another set of classes (file, mqueue, dbus, ...) which
encodes a default rule permission as
class .* <perm>
this encoding is unfortunate in that it will also add the permission
to the class byte, but also sets up following states with the permission.
thankfully, this accespt anything, including nothing generally isn't
valid in the nothing case (eg. a file without any absolute name). For
this set of classes, the high priority mediates rule just ensures
that the null match case does not have permission.
Fixes: 61b7568e1 parser: bug fix mediates_X stub rules.
Signed-off-by: John Johansen <john.johansen@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1307
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: John Johansen <john@jjmx.net>
(cherry picked from commit b6e9df3495)
Signed-off-by: John Johansen <john.johansen@canonical.com>
af_protos.h is a generated table of the protocols created by looking
for definitions of IPPROTO_* in netinet/in.h. Depending on the
architecture, the order of the table may change when using -dM in the
compiler during the extraction of the defines.
This causes an issue because there is more than one IPPROTO defined
by the value 0: IPPROTO_IP and IPPROTO_HOPOPTS which is a header
extension used by IPv6. So if IPPROTO_HOPOPTS was first in the table,
then protocol=0 in the audit logs would be translated to hopopts.
This caused a failure in arm 32bit:
Output doesn't match expected data:
--- ./test_multi/testcase_unix_01.out 2024-08-15 01:47:53.000000000 +0000
+++ ./test_multi/out/testcase_unix_01.out 2024-08-15 23:42:10.187416392 +0000
@@ -12,7 +12,7 @@
Peer Addr: @test_abstract_socket
Network family: unix
Socket type: stream
-Protocol: ip
+Protocol: hopopts
Class: net
Epoch: 1711454639
Audit subid: 322
By the time protocol is resolved in grammar.y, we don't have have
access to the net family to check if it's inet6. Instead of making
protocol dependent on the net family, make the order of the
af_protos.h table consistent between architectures using -dD.
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1309
Approved-by: John Johansen <john@jjmx.net>
Merged-by: Georgia Garcia <georgia.garcia@canonical.com>
(cherry picked from commit 0ec0e2b035)
Signed-off-by: John Johansen <john.johansen@canonical.com>
This enables adding a priority to a rules in policy.
Rules have a default priority of 0. The priority prefix can be added
before the other currently support rule prefixes, ie.
[priority prefix][audit qualifier][rule mode][owner]
If present a numerical priority can be assigned to the rule, where the
greater the number the higher the priority. Eg.
priority=1 audit file r /etc/passwd,
priority=-1 deny file w /etc/**,
Rule priority allows the rule with the highest priority to completely
override lower priority rules where they overlap. Within a given
priority level rules will accumulate in standard apparmor fashion.
Eg. given
priority=1 w /*c,
priority=0 r /a*,
priority=-1 k /*b*,
/abc, /bc, /ac .. will have permissions of w
/ab, /abb, /aaa, .. will have permissions of r
/b, /bcb, /bab, .. will have permissions of k
User specified rule priorities are currently capped at the arbitrary
values of 1000, and -1000.
Notes:
* not all rule types support the priority prefix. Rukes like
- network
- capability
- rlimits
need to be reworked to properly preserve the policy rule structure.
* this patch does not support priority on rule blocks
* this patch does not support using a variable in the priority value.
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1261
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
the ix portion of file, causes x conflicts in regular priority. The
long term goal is to fix this by using dominance for x rules. But in
the mean time we can fix by giving the ix portion of the rule a
reduced priority.
Signed-off-by: John Johansen <john.johansen@canonical.com>
This enables adding a priority to a rules in policy, finishing out the
priority work done to plumb priority support through the internals in
the previous patch.
Rules have a default priority of 0. The priority prefix can be added
before the other currently support rule prefixes, ie.
[priority prefix][audit qualifier][rule mode][owner]
If present a numerical priority can be assigned to the rule, where the
greater the number the higher the priority. Eg.
priority=1 audit file r /etc/passwd,
priority=-1 deny file w /etc/**,
Rule priority allows the rule with the highest priority to completely
override lower priority rules where they overlap. Within a given
priority level rules will accumulate in standard apparmor fashion.
Eg. given
priority=1 w /*c,
priority=0 r /a*,
priority=-1 k /*b*,
/abc, /bc, /ac .. will have permissions of w
/ab, /abb, /aaa, .. will have permissions of r
/b, /bcb, /bab, .. will have permissions of k
User specified rule priorities are currently capped at the arbitrary
values of 1000, and -1000.
Notes:
* not all rule types support the priority prefix. Rukes like
- network
- capability
- rlimits need to be reworked
need to be reworked to properly preserve the policy rule structure.
* this patch does not support priority on rule blocks
* this patch does not support using a variable in the priority value.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Currently mediates_X stub rules are added to the dfa to ensure a valid
transition state will exist if X should be mediated. The kernel uses
this to test whether the dfa supports certain mediation classes.
Unfortunately the mediates stub rules can be removed by other rules,
combined with minimization. In the allow case this is not a problem,
as if the stub rule is removed it will be due to state merging and the
test will still be valid. Unfortunately the deny case can wipe out the
stub rule in a couple of cases, meaning the when the kernel tests that
its in a valid state for mediation it will fail and treat the dfa as
not mediating the rule type, which results in allowing instead of
denying.
Fix this by making sure mediated stub rules can't be overridden by a
deny rule by giving them maximum priority.
Note: there is another issue with stub rule elimination in the allow
case. It will can cause equality tests to fail when combined
with priority rules, because the stub rules where added at
priority 0 and an actual rule of higher priority could
completely override it removing the permission on the stub rule.
This issue will be caught by the equality.sh tests in the
following patch that exposes priority to rules in policy.
Signed-off-by: John Johansen <john.johansen@canonical.com>
The prefix comparison doesn't need to do as many operations as it is
doing, and the operator< can be based on the cmp() fn further reducing
the chance that the code will get out of sync if prefixes are changed.
Signed-off-by: John Johansen <john.johansen@canonical.com>
the parser front end boolean is used for both boolean and integer
values. This is confusing when integer values different than 1 or 0
are being assigned to and from boolean.
Split its uses into the correct semantic boolean and integer cases.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Currently use of extended perms are dependent on prompt rules being present
in policy. Switch to using extended perms if they are supported.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Moving apply_and_clear_deny() before the first minimization pass, which
was necessary to propperly support building accept information for
older none extended permission dfas, allows us to also get rid of doing a
second minimization pass if we want to force clearing explicit deny
info from extended permission tables.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Instead of compressing the permission set into 128 bit and using that
as the index in the permission map, just use the permissions directly
as the index into the permission map.
Note: this will break equality and minimization tests. Because deny
is not being cleared it will result in more partitions in the initial
setup. This will be addressed and the tests will be fixed in a follow
on patch.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Hash minimization was removed in
f0b154528 Fix dfa minimization
however some remnants of minimization remained. A comment and the use
of the hash but only as a 0 value. Drop this dead code and comment.
Signed-off-by: John Johansen <john.johansen@canonical.com>
The hfa stores next/check transitions in 16 bit fields to reduce memory
usage. However this means the state machine can on contain 2^16
states.
Allow the next/check tables to be 32 bit. This theoretically could allow
for 2^32 states however the base table uses the top 8 bits as flags
giving us only 2^24 bits to index into the next/check tables. With
most states having at least 1 transition this effectively caps the
number of states at 2^24.
To obtain 2^32 possible states a flags table needs to be added. Add
a skeleton around supporting a flags table, so we can note the remaining
work that needs to be done. This patch will only allow for 2^24 states.
Bug: https://gitlab.com/apparmor/apparmor/-/issues/419
Signed-off-by: John Johansen <john.johansen@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1303
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: John Johansen <john@jjmx.net>
The hfa stores next/check transitions in 16 bit fields to reduce memory
usage. However this means the state machine can on contain 2^16
states.
Allow the next/check tables to be 32 bit. This theoretically could allow
for 2^32 states however the base table uses the top 8 bits as flags
giving us only 2^24 bits to index into the next/check tables. With
most states having at least 1 transition this effectively caps the
number of states at 2^24.
To obtain 2^32 possible states a flags table needs to be added. Add
a skeleton around supporting a flags table, so we can note the remaining
work that needs to be done. This patch will only allow for 2^24 states.
Bug: https://gitlab.com/apparmor/apparmor/-/issues/419
Signed-off-by: John Johansen <john.johansen@canonical.com>
Older kernels do not support an xtable grouped with the policy dfa.
The presence of a policy.dfa does not indicate whether we should create
an xtable with the policy dfa.
Instead the check should be if the kernel supports the extended
permstable32 format.
Signed-off-by: John Johansen <john.johansen@canonical.com>
__uint128 is not supported by gcc on 32 bit architectures so rework
the 128 bit map key to be a pair of 64bit numbers.
Signed-off-by: John Johansen <john.johansen@canonical.com>
switch permission bits to use perm32_t type. This is just annotating
the code as it is no different than uint32_t at this time.
We do not convert the accept values as they may be mapped permission
bits or they may be and index value.
Signed-off-by: John Johansen <john.johansen@canonical.com>
The use of xbits can not pass verification so we need to leave them
off this makes the profile a leaf profile.
Signed-off-by: John Johansen <john.johansen@canonical.com>
v1 of permstable32 has some broken verification checks. By using two
copies of a merged dfa and an xtable the same size of the permstable
we can work around the issue.
Signed-off-by: John Johansen <john.johansen@canonical.com>
There are two distinct declarations of perms_t.
rule.h: typedef uint32_t perms_t
hfa.h: class perms_t
these definitions clash when the front end and backend share more info.
To avoid this rename rule.h to perm32_t, and move the definition into
perms.h and use it in struct aa_perms.
Signed-off-by: John Johansen <john.johansen@canonical.com>
If extended permissions are supported use them. We need to build a
permission table and set the accept state of the chfa up as an index
into the table.
For now map the front end permission layout into the old format and
then convert that to the perms table just as the kernel does.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Add the ability to parse the prompt qualifier but do not provide
functionality because the backend does not currently support prompt
permissions.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Remove conditional logic from the parser and move it to its own class,
that way any improvements or conditional features will make cleaner
changes.
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
I couldn't figure out why the show info window was using a different
font color than the theme default but this forces its use.
Also, add padding when "Show Current Profile" button is not shown.
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
In this change, I'm also removing the messagebox window and reusing
the more info GUI already implemented
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
aa-notify: Enhanced Graphical User Interfaces
- Added support for --prompt-filter=userns: a popup GUI now appears when an unprivileged, unconfined process attempts to create a user namespace, enabling automatic generation of specific unconfined profiles.
- Added GUIs for easy rule addition.
- Upgraded notifications to two-button format, enabling extended information display and direct rule addition.
- Initial support for customized notification messages based on rule type.
Signed-off-by: Maxime Bélair <maxime.belair@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1281
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: John Johansen <john@jjmx.net>
Code assumes full username would be printed, but this actually requires an extra command line option:
Please double check that this is the only place where `last` is called as a binary before merging this MR.
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1293
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
POSIX states that d_name has up to NAME_MAX (255) characters, and glibc
stores d_name as an array of size NAME_MAX+1 (256). Thus, supplying
PATH_MAX (4096) as the max length could trigger a buffer overrun. This
could be an even bigger issue on other libcs, as POSIX states that d_name
can be unsized.
Fortunately, this does not seem to cause actual issues, as the length is
only used to compare d_name to a short fixed string. However, it'd be better
to pass the actual correct max length to strnlen.
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1290
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: Georgia Garcia <georgia.garcia@canonical.com>
POSIX states that d_name has up to NAME_MAX (255) characters, and glibc
stores d_name as an array of size NAME_MAX+1 (256). Thus, supplying
PATH_MAX (4096) as the max length could trigger a buffer overrun. This
could be an even bigger issue on other libcs, as POSIX states that d_name
can be unsized.
Fortunately, this does not seem to cause actual issues, as the length is
only used to compare d_name to a short fixed string. However, it'd be better
to pass the actual correct max length to strnlen.
Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
The linkage of aa-load with the dynamic libapparmor fails with:
aa_load.c:273: undefined reference to `aa_split_overlay_str'
That is because when aa_split_overlay_str was added to libapparmor,
the function was not added to the library map.
Fixes: 50054ff0 ("add aa_split_overlay_str")
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1288
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
The linkage of aa-load with the dynamic libapparmor fails with:
aa_load.c:273: undefined reference to `aa_split_overlay_str'
That is because when aa_split_overlay_str was added to libapparmor,
the function was not added to the library map.
Fixes: 50054ff0 ("add aa_split_overlay_str")
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
Commonly used by applications to determine if Linux is running in
FIPS mode. As we already allow access to FIPS specific library files
as part of base, allow this there as well.
Signed-off-by: Georg Pfuetzenreuter <mail@georg-pfuetzenreuter.net>
If the test ran under a fs mounted with nosuid option, then these bits
would be ignored and the test would fail. In that case, detect it and
run the test in a tmpfs mountpoint without nosuid.
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
When /tmp is of type tmpfs, the test didn't run because you can't
mount a swapfile on it. This patch mounts an ext2 mountpoint on
$tmpdir so that the swapfile can be mounted on top of it instead of
tmpfs.
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
When /tmp is mounted, remounting / as private for tests that don't
work when shared still fail because /tmp remains as shared. The option
-T in findmnt helps determine the mountpoint in a certain directory,
so use that with $tmpdir to determine the root.
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
The tests that use pivot_root or move mountpoints with mount have to
make sure that / is private for the tests to work. Refactor that logic
into a file to be sourced by the test scripts
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
The abi is not being respected by mqueue rules in many cases. If policy
does ot specify an mqueue rule the abi is correctly applied but if
an mqueue rule is specified explicitly or implicitly (eg. allow all).
without setting the mqueue type OR setting the mqueue type to sysv.
The abi will be ignored and mqueue will be enforced for policy regadless.
Known good mqueue rule that respects abi
mqueue type=posix,
# and all variations that keep type=posix
Known bad mqueue rules that do not respect abi
mqueue,
# and all variants that do not specify the type= option
mqueue type=sysv,
# and all variants that specify the type=sysv option
Issue: https://gitlab.com/apparmor/apparmor/-/issues/412
Fixes: d98c5c4cf ("parser: add parser support for message queue mediation")
Signed-off-by: John Johansen <john.johansen@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1277
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
The kernel does not expect a name and it is not used even within the
parser so drop it. Correct the padding calculation.
sizeof(th_version)
includes the trailing \0 in the count so we should not be adding it
explicitly. Doing so made it seem like we were writing an extra byte
and messing things up, because the string write below did not include
the \0 which we had to add explicitly.
Switch to writing the th_version using size_of() bytes as is used in
the pad calculation, to avoid confusion around the header padding.
Signed-off-by: John Johansen <john.johansen@canonical.com>
The abi is not being respected by mqueue rules in many cases. If policy
does ot specify an mqueue rule the abi is correctly applied but if
an mqueue rule is specified explicitly or implicitly (eg. allow all).
without setting the mqueue type OR setting the mqueue type to sysv.
The abi will be ignored and mqueue will be enforced for policy regadless.
Known good mqueue rule that respects abi
mqueue type=posix,
# and all variations that keep type=posix
Known bad mqueue rules that do not respect abi
mqueue,
# and all variants that do not specify the type= option
mqueue type=sysv,
# and all variants that specify the type=sysv option
Issue: https://gitlab.com/apparmor/apparmor/-/issues/412
Fixes: d98c5c4cf ("parser: add parser support for message queue mediation")
Signed-off-by: John Johansen <john.johansen@canonical.com>
`execpath` allows to reliably store the path of the binary that triggered a log.
This is useful because comm was not sufficient to reliably identify a binary
Signed-off-by: Maxime Bélair <maxime.belair@canonical.com>
This reverts commit 78ae956087.
Commit 78ae956087 causes policy to not
to conform to protocol as determined by the kernel. Technically the
reverted patch is correct and the kernel is wrong but we can not
change 15 years of history.
The reason it breaks the policy in the kernel is because the kernel
does not use the name field, and does not expect it. It just expects
the size with a single trailing 0. This doesn't break because this
section is all padded to 64 bytes so writing the extra 0 doesn't
hurt as it is effectively just manually adding to the padding.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Some applications using the bwrap profile don't function properly due to "Failed name lookup - deleted entry". The following denials trying to start flatpak KeePassXC is an example showing that it happens for both bwrap and unpriv_bwrap profiles:
Jul 12 09:44:37 ubuntu2404 kernel: audit: type=1400 audit(1720741477.106:310): apparmor="DENIED" operation="link" class="file" info="Failed name lookup - deleted entry" error=-2 profile="bwrap" name="/home/\*\*\*\*/.var/app/org.keepassxc.KeePassXC/config/keepassxc/#317211" pid=4021 comm="keepassxc" requested_mask="l" denied_mask="l" fsuid=1000 ouid=1000
Jul 12 09:44:37 ubuntu2404 kernel: audit: type=1400 audit(1720741477.341:317): apparmor="DENIED" operation="link" class="file" profile="unpriv_bwrap" name="/home/**/.var/app/org.keepassxc.KeePassXC/config/keepassxc/keepassxc.ini" pid=4021 comm="keepassxc" requested_mask="l" denied_mask="l" fsuid=1000 ouid=1000 target="/home/**/.var/app/org.keepassxc.KeePassXC/config/keepassxc/#317214"
Fixes: https://launchpad.net/bugs/2072811
I propose this fix for master and apparmor-4.0
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1272
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
Without AA_MAY_MOUNT, mount was not allowed by the allow all
rule. AA_DUMMY_REMOUNT does become AA_MAY_MOUNT, but it fixes the
flags to remount only, so other options are not included. Also, add
allow all rule testcases to the mount regression tests.
Fixes: https://gitlab.com/apparmor/apparmor/-/issues/410
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
By specifying 0 in the unix type, all rules were allowing only the
"none" type, when it wanted to allow all types, so replace it by
0xffffffff. Also, add this testcase to the unix regression tests.
Fixes: https://gitlab.com/apparmor/apparmor/-/issues/410
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
Some applications using the bwrap profile don't function properly due
to "Failed name lookup - deleted entry". The following denials trying
to start flatpak KeePassXC is an example showing that it happens for
both bwrap and unpriv_bwrap profiles:
Jul 12 09:44:37 ubuntu2404 kernel: audit: type=1400 audit(1720741477.106:310): apparmor="DENIED" operation="link" class="file" info="Failed name lookup - deleted entry" error=-2 profile="bwrap" name="/home/****/.var/app/org.keepassxc.KeePassXC/config/keepassxc/#317211" pid=4021 comm="keepassxc" requested_mask="l" denied_mask="l" fsuid=1000 ouid=1000
Jul 12 09:44:37 ubuntu2404 kernel: audit: type=1400 audit(1720741477.341:317): apparmor="DENIED" operation="link" class="file" profile="unpriv_bwrap" name="/home/****/.var/app/org.keepassxc.KeePassXC/config/keepassxc/keepassxc.ini" pid=4021 comm="keepassxc" requested_mask="l" denied_mask="l" fsuid=1000 ouid=1000 target="/home/****/.var/app/org.keepassxc.KeePassXC/config/keepassxc/#317214"
Fixes: https://launchpad.net/bugs/2072811
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
Saw these couple of accesses fail recently on my Ubuntu 22.04 system:
`Jun 3 15:29:24 darkstar kernel: [5401883.070129] audit: type=1107 audit(1717442964.884:9223): pid=729 uid=102 auid=4294967295 ses=4294967295 subj=unconfined msg='apparmor="DENIED" operation="dbus_method_call" bus="system" path="/org/freedesktop/UPower" interface="org.freedesktop.DBus.Properties" member="GetAll" mask="send" name=":1.28" pid=2164500 label="firefox" peer_pid=2502 peer_label="unconfined"`
`Jun 3 15:29:24 darkstar kernel: [5401883.070588] audit: type=1107 audit(1717442964.884:9224): pid=729 uid=102 auid=4294967295 ses=4294967295 subj=unconfined msg='apparmor="DENIED" operation="dbus_method_call" bus="system" path="/org/freedesktop/UPower" interface="org.freedesktop.UPower" member="EnumerateDevices" mask="send" name=":1.28" pid=2164500 label="firefox" peer_pid=2502 peer_label="unconfined"`
Also, I noticed that the `firefox` profile in the Ubuntu 24.04 package has a rule for `/etc/writable/timezone` that is not present in Git. Figured that should be in here.
Fixes: https://gitlab.com/apparmor/apparmor/-/issues/409Closes#409
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1253
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
This is needed to avoid a "Conflicting profiles" error if there are two
profiles for an application, with one of them disabled.
This is not a theoretical usecase - for example, apparmor.d ships some
profiles that replace our "userns+unconfined" profiles. These profiles
use a different filename, and apparmor.d also creates a disable symlink
for the "userns+unconfined" profile it replaces.
I propose this patch for 4.0 and master.
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1264
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
... and not only for events in missing hats.
This fixes a crash if the log contains exec events for a hat where not
even the parent profile exists.
I propose this patch for master, 4.0 and 3.1.
(In 3.0, `aa` is still a `hasher` which avoids the crash, therefore it doesn't really need this patch.)
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1265
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: Christian Boltz <apparmor@cboltz.de>
Contrary to what the name would imply aa-unconfined displays info for
both confined and unconfined processes. Add a --short option that only
output processes that are not confined. Eg.
$ ./utils/aa-unconfined
17192 /snap/chromium/2890/usr/lib/chromium-browser/chrome (/snap/chromium/2890/usr/lib/chromium-browser/chrome --password-store=basic --disable-features=TFLiteLanguageDetectionEnabled) confined by 'snap.chromium.chromium (enforce)'
17395 /snap/chromium/2890/usr/lib/chromium-browser/chrome (/snap/chromium/2890/usr/lib/chromium-browser/chrome --type=utility --utility-sub-type=network.mojom.NetworkService --lang=en-US --service-sandbox-type=none --crashpad-handler-pid=17337 --enable-crash-reporter=,snap --change-stack-guard-on-fork=enable --shared-files=v8_context_snapshot_data:100 --field-trial-handle=3,i,16674663885832976354,18417931519279121981,262144 --disable-features=TFLiteLanguageDetectionEnabled --variations-seed-version) confined by 'snap.chromium.chromium (enforce)'
17981 /snap/firefox/4451/usr/lib/firefox/firefox confined by 'snap.firefox.firefox (enforce)'
1353664 /tmp/.mount_OrcaSl7G1va5/bin/orca-slicer not confined
is trimmed to
$ ./utils/aa-unconfined --short
1353664 /tmp/.mount_OrcaSl7G1va5/bin/orca-slicer not confined
Signed-off-by: John Johansen <john.johansen@canonical.com>
The prompt/user upcall mode shows up as a mode of (user). And for
stacked policy with different modes (mixed) is used. Add these to the
list of modes to screen.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Add the ability to list applications that are unconfined and have
any open network socket open, both listening and non-listening.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Add the abiity to list applications that are unconfined and have
open connection ports that are not listening.
Signed-off-by: John Johansen <john.johansen@canonical.com>
The documentation of --paranoid is wrong. It lists all processes and
does not exclude based on whether it has a network port open.
Signed-off-by: John Johansen <john.johansen@canonical.com>
This is needed to avoid a "Conflicting profiles" error if there are two
profiles for an application, with one of them disabled.
This is not a theoretical usecase - for example, apparmor.d ships some
profiles that replace our "userns+unconfined" profiles. These profiles
use a different filename, and apparmor.d also creates a disable symlink
for the "userns+unconfined" profile it replaces.
* UnixRule: Fix handling of peers with a ? and peers that are/need to be quoted
`?` is a valid AARE char, add it to the regexes that match the AARE.
Also add some tests to ensure this is really fixed, and make the error
output of the tests more useful/verbose.
* Fix handling of quoted peers in UnixRule (and others)
In UnixRule (and probably also in other rules that use
print_dict_values()` and `initialize_cond_dict()`), the handling of
peers with a value that is quoted and/or needs to be quoted was broken
because
- quotes didn't get stripped in `initialize_cond_dict()`
- `print_dict_values()` didn't use `quote_if_needed()`
Note: print_dict_values also handles integers (like network ports).
Convert them to a string so that `if ' ' in data` in `quote_if_needed()`
doesn't explode.
Also enable the test that uncovered this bug.
Fixes: https://gitlab.com/apparmor/apparmor/-/issues/404Closes#404
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1262
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
In UnixRule (and probably also in other rules that use
print_dict_values()` and `initialize_cond_dict()`), the handling of
peers with a value that is quoted and/or needs to be quoted was broken
because
- quotes didn't get stripped in `initialize_cond_dict()`
- `print_dict_values()` didn't use `quote_if_needed()`
Note: print_dict_values also handles integers (like network ports).
Convert them to a string so that `if ' ' in data` in `quote_if_needed()`
doesn't explode.
Also enable the test that uncovered this bug.
`?` is a valid AARE char, add it to the regexes that match the AARE.
Also add some tests to ensure this is really fixed, and make the error
output of the tests more useful/verbose.
Note: One of the added tests (with a space in the peer name) uncovered a
bug in quote handling. This will be fixed in the next commit.
Fixes: https://gitlab.com/apparmor/apparmor/-/issues/404
Update the state machine readme to better reflect how the chfa is
encoded and works. It still needs a lot more but fixes several errors
in the doc and adds some info about state differential encoding, oobs,
and comb compression.
In addition fix an off by own error during chfa encoding. This has
likely never triggered as it gets hidden by being in a section that
is being in a section that is padded to an 8 byte boundary.
Signed-off-by: John Johansen <john.johansen@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1244
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
Expression simplification can get into an infinite loop due to eps
pairs hiding behind and alternation that can't be caught by
normalize_eps() (which exists in the first place to stop a similar
loop).
The loop in question happens in AltNode::normalize when a subtree has
the following structure.
1. elseif (child[dir]->is_type(ALT_NODE)) rotate_node too
alt
/\
/ \
/ \
eps alt
/\
/ \
/ \
alt eps
/\
/ \
/ \
eps eps
2. if (normalize_eps(dir)) results in
alt
/\
/ \
/ \
alt eps
/\
/ \
/ \
alt eps
/\
/ \
/ \
eps eps
3. elseif (child[dir]->is_type(ALT_NODE)) rotate_node too
alt
/\
/ \
/ \
alt alt
/\ /\
/ \ / \
/ \ / \
eps eps eps eps
4. elseif (child[dir]->is_type(ALT_NODE)) rotate_node too
alt
/\
/ \
/ \
eps alt
/\
/ \
/ \
eps alt
/\
/ \
/ \
eps eps
5. if (normalize_eps(dir)) results in
alt
/\
/ \
/ \
alt eps
/\
/ \
/ \
eps alt
/\
/ \
/ \
eps eps
6. elseif (child[dir]->is_type(ALT_NODE)) rotate_node too
alt
/\
/ \
/ \
eps alt
/\
/ \
/ \
alt eps
/\
/ \
/ \
eps eps
back to beginning of cycle
Fix this by detecting the creation of an eps_pair in rotate_node(),
that pair can be immediately eliminated by simplifying the tree in that
step.
In the above cycle the pair creation is caught at step 3 resulting
in
3. elseif (child[dir]->is_type(ALT_NODE)) rotate_node too
alt
/\
/ \
/ \
alt eps
/\
/ \
/ \
eps eps
4. elseif (child[dir]->is_type(ALT_NODE)) rotate_node too
alt
/\
/ \
/ \
eps alt
/\
/ \
/ \
eps eps
which gets reduced to
alt
/\
/ \
/ \
eps eps
breaking the normalization loop. The degenerate alt node will be caught
in turn when its parent is dealt with.
This needs to be backported to all releases
Closes: https://gitlab.com/apparmor/apparmor/-/issues/398
Fixes: 846cee506 ("Split out parsing and expression trees from regexp.y")
Reported-by: Christian Boltz <apparmor@cboltz.de>
Signed-off-by: John Johansen <john.johansen@canonical.com>
Closes#398
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1252
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: John Johansen <john@jjmx.net>
Expression simplification can get into an infinite loop due to eps
pairs hiding behind and alternation that can't be caught by
normalize_eps() (which exists in the first place to stop a similar
loop).
The loop in question happens in AltNode::normalize when a subtree has
the following structure.
1. elseif (child[dir]->is_type(ALT_NODE)) rotate_node too
alt
/\
/ \
/ \
eps alt
/\
/ \
/ \
alt eps
/\
/ \
/ \
eps eps
2. if (normalize_eps(dir)) results in
alt
/\
/ \
/ \
alt eps
/\
/ \
/ \
alt eps
/\
/ \
/ \
eps eps
3. elseif (child[dir]->is_type(ALT_NODE)) rotate_node too
alt
/\
/ \
/ \
alt alt
/\ /\
/ \ / \
/ \ / \
eps eps eps eps
4. elseif (child[dir]->is_type(ALT_NODE)) rotate_node too
alt
/\
/ \
/ \
eps alt
/\
/ \
/ \
eps alt
/\
/ \
/ \
eps eps
5. if (normalize_eps(dir)) results in
alt
/\
/ \
/ \
alt eps
/\
/ \
/ \
eps alt
/\
/ \
/ \
eps eps
6. elseif (child[dir]->is_type(ALT_NODE)) rotate_node too
alt
/\
/ \
/ \
eps alt
/\
/ \
/ \
alt eps
/\
/ \
/ \
eps eps
back to beginning of cycle
Fix this by detecting the creation of an eps_pair in rotate_node(),
that pair can be immediately eliminated by simplifying the tree in that
step.
In the above cycle the pair creation is caught at step 3 resulting
in
3. elseif (child[dir]->is_type(ALT_NODE)) rotate_node too
alt
/\
/ \
/ \
alt eps
/\
/ \
/ \
eps eps
4. elseif (child[dir]->is_type(ALT_NODE)) rotate_node too
alt
/\
/ \
/ \
eps alt
/\
/ \
/ \
eps eps
whch gets reduces to
alt
/\
/ \
/ \
eps eps
breaking the normalization loop. The degenerate alt node will be caught
in turn when its parent is dealt with.
This needs to be backported to all releases
Closes: https://gitlab.com/apparmor/apparmor/-/issues/398
Fixes: 846cee506 ("Split out parsing and expression trees from regexp.y")
Reported-by: Christian Boltz <apparmor@cboltz.de>
Signed-off-by: John Johansen <john.johansen@canonical.com>
Adding mediation classes in unconfined profiles caused nested profiles
to be mediated, inside a container for example.
As a first step, skip the addition of mediation classes into the dfa.
The creation of unprivileged user namespaces is an exception, where we
always want to mediate it.
Fixes: https://bugs.launchpad.net/apparmor/+bug/2067900
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
technically a # leading a value in an assignment expression is allowed,
however people are also using it to a comment at the end of a line.
ie.
```
@{var1}=value1 # comment about this value or for a given system
```
this unsurprisingly leads to odd/unexpected behavior when the variable
is used.
```
allow rw /@{var1},
```
expands into
```
allow rw /{value1,#,comment,about,this,value,or,for,a,given,system},
```
change a leading # of a value in an assignment expression to a comment.
If the # is really supposed to lead the value, require it to be escaped
or in quotes.
ie.
```
@{var1}=value1 \#not_a_comment
```
Note: this could potentially break som policy if the # was used as the
leading character for a value in an assignment expression, but
is worth it to avoid the confusion.
Signed-off-by: John Johansen <john.johansen@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1255
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: John Johansen <john@jjmx.net>
The parser writes
sizeof(th)) + th_version + (char)0 + name + (char)0;
but the padding currently is computed as
sizeof(th)) + th_version + name + (char)0;
missing the internal (char)0, add 1 to the pad and fill to ensure
this is correct.
Reported-by: Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
Signed-off-by: John Johansen <john.johansen@canonical.com>
Update the state machine readme to better reflect how the chfa is
encoded and works. It still needs a lot more but fixes several errors
in the doc and adds some info about state differential encoding, oobs,
and comb compression.
Signed-off-by: John Johansen <john.johansen@canonical.com>
technically a # leading a value in an assignment expression is allowed,
however people are also using it to a comment at the end of a line.
ie.
@{var1}=value1 # comment about this value or for a given system
this unsurprisingly leads to odd/unexpected behavior when the variable
is used.
allow rw /@{var1},
expands into
allow rw /{value1,#,comment,about,this,value,or,for,a,given,system},
change a leading # as value in an assignment expression to a comment.
If the # is really supposed to lead the value, require it to be escaped
or in quotes.
ie.
@{var1}=value1 \#not_a_comment
Note: this could potentially break som policy if the # was used as the
leading character for a value in an assignment expression, but
is worth it to avoid the confusion.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Similarly to https://gitlab.com/apparmor/apparmor/-/merge_requests/689, use the
global CFLAGS when building Python library, so we honor extra flags set by
distributions, such
as -fstack-protector-strong -fstack-clash-protection -Werror=format-security -fcf-protection.
Spotted by blhc on Debian.
Gbp-Pq: Name Honor-global-CFLAGS-when-building-Python-library.patch
This file contains the same kind of information as @{PROC}/@{pid}/net/route
and both files are world readable:
```
$ ls -l /proc/self/net/*route
-r--r--r-- 1 root root 0 Jun 3 15:33 /proc/self/net/ipv6_route
-r--r--r-- 1 root root 0 Jun 3 15:33 /proc/self/net/route
```
Signed-off-by: Simon Deziel <simon.deziel@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1246
Approved-by: Christian Boltz <apparmor@cboltz.de>
Merged-by: Christian Boltz <apparmor@cboltz.de>
This file contains the same kind of information as @{PROC}/@{pid}/net/route
and both files are world readable:
```
$ ls -l /proc/self/net/*route
-r--r--r-- 1 root root 0 Jun 3 15:33 /proc/self/net/ipv6_route
-r--r--r-- 1 root root 0 Jun 3 15:33 /proc/self/net/route
```
Signed-off-by: Simon Deziel <simon.deziel@canonical.com>
aa-remove-unknown doesn't deal properly with profiles that contain
spaces in their names.
Using profile "MongoDB Compass" as an example, awk's sub returns the
number of matches - either 1 or 0 and replaces the actual string ($0)
with the substitution. By accessing the return of sub with $, awk
would be accessing $1 which would return only "MongoDB".
Fix this by using $0 instead of $str.
Fixes: https://gitlab.com/apparmor/apparmor/-/issues/395
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
Closes#395
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1243
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
In the course of preparing !1207, I found that the validation rules in `profiles/Makefile` did not take kindly to the new `profiles/apparmor/profiles/extras/abstractions/` directory. I tried a couple rounds of quick fixes, but it became clear that the rules as currently written were just not amenable to the new addition, and needed more attention than I could give it by-the-by.
So I separated out that commit, and revised the makefile more thoroughly. The updated rules now rely more on `find(1)` than `$(wildcard)`, and have a number of [what I believe to be] small quality-of-life improvements. Taken together, `make check` passes cleanly with the new files from my other MR present.
One thing I noticed was that the profiles under `apparmor.d/` were not previously being checked for the `include if exists <local/*>` bit---only the ones under `extras/`. I've thus included a fix to the `sbuild-shell` profile, which fortunately was the only one that failed the check.
Note that at present, you'll get a couple of harmless `find: ‘./apparmor/profiles/extras/abstractions’: No such file or directory` errors when running the checks, since that directory won't appear until the other MR is merged. I figure, better to bear that for now, and not have to touch the makefile again later.
NOTE: The CI pipeline here will need to be updated to invoke the `check-local` target instead of `check-extras`. This target was renamed as it is no longer limited in scope to the profiles under `extras/`.
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1214
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
aa-remove-unknown doesn't deal properly with profiles that contain
spaces in their names.
Using profile "MongoDB Compass" as an example, awk's sub returns the
number of matches - either 1 or 0 and replaces the actual string ($0)
with the substitution. By accessing the return of sub with $, awk
would be accessing $1 which would return only "MongoDB".
Fix this by using $0 instead of $str.
Fixes: https://gitlab.com/apparmor/apparmor/-/issues/395
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
Without this patch, aa-remove-unknown uses 'profile_name (unconfined)'
when trying to unload unconfined profiles, which fails for obvious
reasons with (picking a random example)
Removing 'busybox (unconfined)'
/sbin/aa-remove-unknown: line 112: echo: write error: No such file or directory
I propose this patch for 4.0 and master.
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1240
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: Georgia Garcia <georgia.garcia@canonical.com>
Without this patch, aa-remove-unknown uses 'profile_name (unconfined)'
when trying to unload unconfined profiles, which fails for obvious
reasons with (picking a random example)
Removing 'busybox (unconfined)'
/sbin/aa-remove-unknown: line 112: echo: write error: No such file or directory
... and tests for it.
This replaces the old code that just stores the full rule as text.
We also get rid of the old ['allow'] and ['deny'] items in
ProfileStorage, the handling of old write functions, and the last usage
of _Raw_Rule (and therefore _Raw_Rule itsself).
Also delete the old test-pivot_root_parse.py which relied on the ancient
code, and even used a wrong syntax in its test rules.
Oh, and aa-logprof can now ask about pivot_root events.
See the individual commits for details.
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1232
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: Christian Boltz <apparmor@cboltz.de>
Mount Rules with options in { remount, [make-] { [r]unbindable, [r]shared, [r]private, and [r]slave }} do not support specifying a source. This commit aligns utils implementation to apparmor_parser's, which prohibits having a both source and a destination simultaneously, instad of just prohibiting source.
Therefore, both `mount options=(unbindable) /a,` and `mount options=(unbindable) -> /a,` are now supported (and equivalent for apparmor_parser). However, `mount options=(unbindable) /a -> /b,` is invalid.
For the same reason, specifying a fstype in these cases is also prohibited.
Similarly, we prohibit to specify a fstype for bind mount rules.
Fixes: https://bugs.launchpad.net/ubuntu/+source/apparmor/+bug/2065685
Signed-off-by: Maxime Bélair <maxime.belair@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1236
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: Georgia Garcia <georgia.garcia@canonical.com>
Mount Rules with options in { remount, [make-] { [r]unbindable, [r]shared, [r]private, and [r]slave }} do not support specifying a source. This commit aligns utils implementation to apparmor_parser's, which prohibits having a both source and a destination simultaneously, instad of just prohibiting source.
Therefore, both `mount options=(unbindable) /a,` and `mount options=(unbindable) -> /a,` are now supported (and equivalent for apparmor_parser). However, `mount options=(unbindable) /a -> /b,` is invalid.
For the same reason, specifying a fstype in these cases is also prohibited.
Similarly, we prohibit to specify a fstype for bind mount rules.
Fixes: https://bugs.launchpad.net/ubuntu/+source/apparmor/+bug/2065685
Signed-off-by: Maxime Bélair <maxime.belair@canonical.com>
... for handling pivot_root rules.
This replaces the old code that just stores the full rule as text.
We also get rid of the old ['allow'] and ['deny'] items in
ProfileStorage, the handling of old write functions, and the last usage
of _Raw_Rule (and therefore _Raw_Rule itsself).
Also delete the old test-pivot_root_parse.py which relied on the ancient
code, and even used a wrong syntax in its test rules.
The following exceptions were added to flake8 since they have several
expected uses in the tools and their tests:
E501: Line lengths are recommended to be no greater than 79 characters.
E241: Multiple spaces after ','
W503: Line break occurred before a binary operator
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
ProfileStorage knows a whole profile, therefore it should also include the profile header in `__repr__()`.
Also add a test for this.
While on it, add a test for an invalid type change for a type that doesn't have special handling in `__setitem__()` to increase test coverage.
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1233
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: Christian Boltz <apparmor@cboltz.de>
Since musl 1.2.5, basename(3) prototype is only provided in libgen.h
(as mandated by POSIX) and not in strings.h. Also there is a major
difference between the gnu basename and the one defined in libgen.h,
the latter modify the argument string making them incompatible.
Fix this by defining a portable version of basename using strchr.
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1234
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: Georgia Garcia <georgia.garcia@canonical.com>
Since musl 1.2.5, basename(3) prototype is only provided in libgen.h
(as mandated by POSIX) and not in strings.h. Also there is a major
difference between the gnu basename and the one defined in libgen.h,
the latter modify the argument string making them incompatible.
Fix this by defining a portable version of basename using strchr.
audit.log entries for mount events don't always include `class=mount`,
but can still be the base for mount rules.
Change logparser.py to also consider `operation=mount` as a mount event.
Actually we already had such a log and profile in our collection
(testcase_mount_01), but since it existed years before MountRule was
implemented, it was excluded in test-libapparmor-test_multi.py.
Therefore we didn't notice that it failed to produce a profile rule when
MountRule was introduced.
Remove testcase_mount_01 from the list of known failures so that it gets
tested - and fix the syntax error in the hand-written
testcase_mount_01.profile.
Also add testcase_mount_02 which is a mount event without fstype,
srcname and class.
I propose this fix for 4.0 and master.
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1229
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: Georgia Garcia <georgia.garcia@canonical.com>
audit.log entries for mount events don't always include `class=mount`,
but can still be the base for mount rules.
Change logparser.py to also consider `operation=mount` as a mount event.
Actually we already had such a log and profile in our collection
(testcase_mount_01), but since it existed years before MountRule was
implemented, it was excluded in test-libapparmor-test_multi.py.
Therefore we didn't notice that it failed to produce a profile rule when
MountRule was introduced.
Remove testcase_mount_01 from the list of known failures so that it gets
tested - and fix the syntax error in the hand-written
testcase_mount_01.profile.
Also add testcase_mount_02 which is a mount event without fstype,
srcname and class.
Did some testing on a fresh post-release image of noble, and uncovered some new denials:
#### Xorg
`2024-05-06T19:55:36.782484-04:00 image-ubuntu64 kernel: audit: type=1400 audit(1715039736.765:174): apparmor="DENIED" operation="link" class="file" profile="Xorg" name="/tmp/.X0-lock" pid=1366 comm="Xorg" requested_mask="r" denied_mask="r" fsuid=0 ouid=0 target="/tmp/.tX0-lock"`
#### chromium_browser
`2024-05-06T21:17:09.674963-04:00 image-ubuntu64 kernel: audit: type=1400 audit(1715040834.256:168): apparmor="DENIED" operation="userns_create" class="namespace" profile="chromium_browser" pid=2133 comm="chromium" requested="userns_create" denied="userns_create"`
#### firefox
`2024-05-06T21:33:09.387356-04:00 image-ubuntu64 kernel: audit: type=1400 audit(1715045589.369:505): apparmor="DENIED" operation="userns_create" class="namespace" profile="firefox" pid=3610 comm="firefox" requested="userns_create" denied="userns_create"`
`2024-05-06T21:36:48.911280-04:00 image-ubuntu64 kernel: audit: type=1400 audit(1715045808.884:682): apparmor="DENIED" operation="open" class="file" profile="firefox" name="/sys/fs/cgroup/user.slice/user-1000.slice/session-c2.scope/cpu.max" pid=4348 comm="firefox" requested_mask="r" denied_mask="r" fsuid=1000 ouid=0`
`2024-05-06T21:42:34.152955-04:00 image-ubuntu64 dbus-daemon[1628]: apparmor="DENIED" operation="dbus_bind" bus="session" name="org.mpris.MediaPlayer2.firefox.instance_1_82" mask="bind" pid=4348 label="firefox"`
#### Xorg (second commit)
I neglected to set `abi/4.0` when this went in originally. (I was using the profile on jammy, hence the `3.0`.)
Also, testing on an older laptop that *doesn't* support KMS revealed that Xorg needs some pretty serious permissions then. I've added them in commented-out form, with an explanatory comment. (The `#nokms#` bit is meant to simplify uncommenting those two lines mechanically, e.g. `sed -i 's/#nokms#//'`)
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1227
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Approved-by: Christian Boltz <apparmor@cboltz.de>
Merged-by: Christian Boltz <apparmor@cboltz.de>
Balena Etcher runs in a degraded sandbox mode when unprivileged userns
is not available. Add an unconfined profile so it's properly
sandboxed.
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
Shellcheck is complaining that some of the functions are never called,
but they are called from rc.apparmor.functions, causing a false
positive.
This issue only appears in shellcheck version 0.9.0, which is the one
used in ubuntu 24.04, that's why it only failed in the pipeline now.
Fixes: https://gitlab.com/apparmor/apparmor/-/issues/388
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
Since we are using ubuntu:latest, and noble was released, some tests
are failing.
shellcheck needs python3 to run, which was possibly installed by
default in previous ubuntu images and is no longer the case.
Ignore dist-packages python files during our coverage tests.
Fixes: https://gitlab.com/apparmor/apparmor/-/issues/388
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
Rename the "check-extras" target to "check-local" as it is no longer
limited to the extra profiles, and also fix a local include in the
sbuild-shell profile so that it passes the newly-applied CI check.
Add a flag that allows setting the error code AppArmor will send when
an operation is denied. This should not be used normally.
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
This is a profile to contain the Xorg X11 server, which still runs as root in many scenarios (not least under [LightDM](https://github.com/canonical/lightdm/issues/18)).
I've tested this under every X display manager available in Debian/Ubuntu, as well as plain `startx(1)`. Both rootful and rootless modes are covered. The hardware I've tried this on predominantly uses Intel integrated graphics, with one Nouveau system represented. If someone has an Nvidia GPU running the proprietary driver, that would be a good data point to double-check, owing to the different driver architecture.
As you can see, I avoided going too far into the weeds enumerating everything the X server needs to run. The general pattern I found was that it needs read access to a lot of things, but write access to relatively few.
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1075
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
This updates the man page for the recent inet mediation patch.
This is an extension of MR 1202, it adds a patch that changes the anonymous ip address anon to be ip address none which is a better fit.
This patch adds documentation of the recent network changes which extended all network rules to support access permissions, and added address and port matching for inet and inet6 families.
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1213
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
inet mediation allows specifying rules for sockets that don't have
a known address, whether because it is unbound or because the
kernel doesn't make the address available.
The current code uses the word anon for anonymous, but that has
proven to be unclear. Switch from using anon to none, to emphasize
that this is a case where there just isn't an address to use as
part of mediation.
Signed-off-by: John Johansen <john.johansen@canonical.com>
When a family is specified in the network rules, we have to make sure
the conditionals match the family. A netlink rule should not be able
to specify ip and port for local and remote (peer) sockets, for example.
When type or protocol is specified in network rules along with inet
conditionals, we should only generate rules for the families that
support those conditionals.
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
The bwrap and unshare profiles are special profiles in the same
vein as the unconfined profiles but they actual enforce restrictions
on the applications that are launched.
As such they have come to late in the 4.0 dev cycle to consider enabling
by default. Disable them but ship them so users or distros can easily
enable them.
Signed-off-by: John Johansen <john.johansen@canonical.com>
This adds a bwrap profile to allow it to function on a system with
user namespace restrictions enabled.
The child task of bwrap will enter into a profile without capabilities
thus preventing bwrap from being able to be used to arbitrarily
by-pass user namespace restrictions.
This profile does prevent applications launch with privilege (eg.
sudo bwrap ...) from functioning so it may break some use cases.
Note: The unpriv_bwrap profile is deliberately stacked against the
bwrap profile due to bwraps uses of no-new-privileges.
Fixes: https://bugs.launchpad.net/ubuntu/+source/pageedit/+bug/2046844
Signed-off-by: John Johansen <john.johansen@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1205
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
This adds a bwrap profile to allow it to function on a system with
user namespace restrictions enabled.
The child task of bwrap will enter into a profile without capabilities
thus preventing bwrap from being able to be used to arbitrarily
by-pass user namespace restrictions.
This profile does prevent applications launch with privilege (eg.
sudo bwrap ...) from functioning so it may break some use cases.
Note: The unpriv_bwrap profile is deliberately stacked against the
bwrap profile due to bwraps uses of no-new-privileges.
Fixes: https://bugs.launchpad.net/ubuntu/+source/pageedit/+bug/2046844
Signed-off-by: John Johansen <john.johansen@canonical.com>
The version of tarball version of firefox downloaded from mozilla.org
installs to /opt/firefox/firefox. Support this location so that the
firefox from the tarball works.
Note this does not support running firefox from the users home directory
in this case the user must update the profile accordingly.
Signed-off-by: John Johansen <john.johansen@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1203
Approved-by: John Johansen <john@jjmx.net>
Merged-by: John Johansen <john@jjmx.net>
This adds an unshare profile to allow it to function on a system
with user namespace restrictions enabled.
The child task of unshare will enter into a profile without capabilities
thus preventing unshare from being able to arbitrarily being used to
by-pass the user namespace restriction.
This profile does prevent applications launch with privilege (eg.
sudo unshare ...) from functioning so it may break some use cases.
Signed-off-by: John Johansen <john.johansen@canonical.com>
The version of tarball version of firefox downloaded from mozilla.org
installs to /opt/firefox/firefox. Support this location so that the
firefox from the tarball works.
Note this does not support running firefox from the users home directory
in this case the user must update the profile accordingly.
Signed-off-by: John Johansen <john.johansen@canonical.com>
In some cases, ldd might obtain information by executing the given
binary (see ldd(1)) - which is not something we should do on potentially
unknown binaries, especially because aa-genprof and aa-autodep (and
therefore also ldd) are often started as root.
Additionally, the ldd result typically listed libraries already covered
by abstractions/base, which makes the ldd call superfluous.
While on it,
- remove all references to ldd
- remove code only used for calling ldd and handling its results
- remove tests checking ldd results, and the fake_ldd script
- adjust a test where fake_ldd had added some libraries
- remove ldd path from logprof.conf [settings]
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1201
Approved-by: Georgia Garcia <georgia.garcia@canonical.com>
Merged-by: Christian Boltz <apparmor@cboltz.de>
- Before this commit, fstype had to match a known fs. However, having and maintaining the exhaustive list of fstypes proved challenging (see !1195 and !1176). Therefore, we add support for any filesystem name.
- Completing AARE support for fstype (brace expressions like ext{3,4} are now supported).
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1198
Approved-by: Christian Boltz <apparmor@cboltz.de>
Merged-by: Christian Boltz <apparmor@cboltz.de>
- Before this commit, fstype had to match a known fs. However, having and maintaining the exhaustive list of fstypes proved challenging (see !1195 and !1176). Therefore, we add support for any filesystem name.
- Completing AARE support for fstype (brace expressions like ext{3,4} are now supported).
In some cases, ldd might obtain information by executing the given
binary (see ldd(1)) - which is not something we should do on potentially
unknown binaries, especially because aa-genprof and aa-autodep (and
therefore also ldd) are often started as root.
Additionally, the ldd result typically listed libraries already covered
by abstractions/base, which makes the ldd call superfluous.
While on it,
- remove all references to ldd
- remove code only used for calling ldd and handling its results
- remove tests checking ldd results, and the fake_ldd script
- adjust a test where fake_ldd had added some libraries
- remove ldd path from logprof.conf [settings]
Give access to @{HOMEDIRS}, just like in usr.sbin.smbd, so that
usershares in /home/ can be accessed.
apparmor="DENIED" operation="open" class="file" profile="samba-rpcd-classic" name="/home/user/path/to/usershare/" pid=4781 comm="rpcd_classic" requested_mask="r" denied_mask="r" fsuid=0 ouid=1000
- Adding support for --output-dir in aa-logprof and aa-genprof, allowing to work on profiles without applying the modified version
- Adding support for --allow-all in aa-logprof that creates non-interactively 'allow' rules for all logs
- Adding support for --no-abstraction in aa-logprof and aa-genprof
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1177
Approved-by: Christian Boltz <apparmor@cboltz.de>
Merged-by: Christian Boltz <apparmor@cboltz.de>
When the setup of the notify options failed, they were exiting the
program without cleaning up the mqueue. Fix this by returning instead
of exiting, since the main function does the cleanup in case of any
failures. If the test succeeds, then it exits successfully.
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
mq_notify only notifies if the queue is empty, so if the sender wins
the race and sends a message before mq_notify is set up, mq_notify
will timeout.
Adding synchronization using pipes the same way it was used in the
setns tests should fix it. The pipe now needs rw permissions, so add
that to the tests.
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
The child which sends the message was winning the race and causing a
timeout when the receiver was waiting for the message.
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
net_inet makes more sense since other finegrained network types can be
added in the future.
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
The abstraction lxc/start-container shipped by the liblxc-common
package uses the following mount rule which was not allowed by our
regexes:
mount options=(rw, make-slave) -> **,
mount options=(rw, make-rslave) -> **,
Since in AppArmor regex ** includes '/' but * by itself doesn't, I'm
adding explicit support for **.
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
MR: https://gitlab.com/apparmor/apparmor/-/merge_requests/1195
Approved-by: Christian Boltz <apparmor@cboltz.de>
Merged-by: Georgia Garcia <georgia.garcia@canonical.com>
The abstraction lxc/start-container shipped by the liblxc-common
package uses the following mount rule which was not allowed by our
regexes:
mount options=(rw, make-slave) -> **,
mount options=(rw, make-rslave) -> **,
Since in AppArmor regex ** includes '/' but * by itself doesn't, I'm
adding explicit support for **.
Signed-off-by: Georgia Garcia <georgia.garcia@canonical.com>
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
