linux-toradex.git/security/apparmor/include/match.h, branch v6.19

apparmor: fix loop detection used in conflicting attachment resolution

2025-05-26T03:14:53+00:00

Conflicting attachment resolution is based on the number of states
traversed to reach an accepting state in the attachment DFA, accounting
for DFA loops traversed during the matching process. However, the loop
counting logic had multiple bugs:

 - The inc_wb_pos macro increments both position and length, but length
   is supposed to saturate upon hitting buffer capacity, instead of
   wrapping around.
 - If no revisited state is found when traversing the history, is_loop
   would still return true, as if there was a loop found the length of
   the history buffer, instead of returning false and signalling that
   no loop was found. As a result, the adjustment step of
   aa_dfa_leftmatch would sometimes produce negative counts with loop-
   free DFAs that traversed enough states.
 - The iteration in the is_loop for loop is supposed to stop before
   i = wb->len, so the conditional should be < instead of <=.

This patch fixes the above bugs as well as the following nits:
 - The count and size fields in struct match_workbuf were not used,
   so they can be removed.
 - The history buffer in match_workbuf semantically stores aa_state_t
   and not unsigned ints, even if aa_state_t is currently unsigned int.
 - The local variables in is_loop are counters, and thus should be
   unsigned ints instead of aa_state_t's.

Fixes: 21f606610502 ("apparmor: improve overlapping domain attachment resolution")

Signed-off-by: Ryan Lee 
Co-developed-by: John Johansen 
Signed-off-by: John Johansen

apparmor: ensure WB_HISTORY_SIZE value is a power of 2

2025-05-18T01:20:10+00:00

WB_HISTORY_SIZE was defined to be a value not a power of 2, despite a
comment in the declaration of struct match_workbuf stating it is and a
modular arithmetic usage in the inc_wb_pos macro assuming that it is. Bump
WB_HISTORY_SIZE's value up to 32 and add a BUILD_BUG_ON_NOT_POWER_OF_2
line to ensure that any future changes to the value of WB_HISTORY_SIZE
respect this requirement.

Fixes: 136db994852a ("apparmor: increase left match history buffer size")

Signed-off-by: Ryan Lee 
Signed-off-by: John Johansen

apparmor: fix some kernel-doc issues in header files

2025-05-17T08:52:25+00:00

Fix kernel-doc warnings in apparmor header files as reported by
scripts/kernel-doc:

cred.h:128: warning: expecting prototype for end_label_crit_section(). Prototype was for end_current_label_crit_section() instead
file.h:108: warning: expecting prototype for aa_map_file_perms(). Prototype was for aa_map_file_to_perms() instead

lib.h:159: warning: Function parameter or struct member 'hname' not described in 'basename'
lib.h:159: warning: Excess function parameter 'name' description in 'basename'

match.h:21: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
 * The format used for transition tables is based on the GNU flex table

perms.h:109: warning: Function parameter or struct member 'accum' not described in 'aa_perms_accum_raw'
perms.h:109: warning: Function parameter or struct member 'addend' not described in 'aa_perms_accum_raw'
perms.h:136: warning: Function parameter or struct member 'accum' not described in 'aa_perms_accum'
perms.h:136: warning: Function parameter or struct member 'addend' not described in 'aa_perms_accum'

Signed-off-by: Randy Dunlap 
Reviewed-by: Ryan Lee 
Cc: John Johansen 
Cc: John Johansen 
Cc: apparmor@lists.ubuntu.com
Cc: linux-security-module@vger.kernel.org
Cc: Paul Moore 
Cc: James Morris 
Cc: "Serge E. Hallyn" 
Signed-off-by: John Johansen

apparmor: add support for 2^24 states to the dfa state machine.

2024-11-27T03:21:05+00:00

Currently the dfa state machine is limited by its default, next, and
check tables using u16. Allow loading of u32 tables, and if u16 tables
are loaded map them to u32.

The number of states allowed does not increase to 2^32 because the
base table uses the top 8 bits of its u32 for flags. Moving the flags
into a separate table allowing a full 2^32 bit range wil be done in
a separate patch.

Link: https://gitlab.com/apparmor/apparmor/-/issues/419
Signed-off-by: John Johansen

apparmor: refcount the pdb

2023-10-18T22:30:47+00:00

With the move to permission tables the dfa is no longer a stand
alone entity when used, needing a minimum of a permission table.
However it still could be shared among different pdbs each using
a different permission table.

Instead of duping the permission table when sharing a pdb, add a
refcount to the pdb so it can be easily shared.

Reviewed-by: Georgia Garcia 
Signed-off-by: John Johansen

apparmor: preparse for state being more than just an integer

2022-10-03T21:49:03+00:00

Convert from an unsigned int to a state_t for state position. This is
a step in prepping for the state position carrying some additional
flags, and a limited form of backtracking to support variables.

Signed-off-by: John Johansen

apparmor: add outofband transition and use it in xattr match

2020-01-21T14:00:20+00:00

There are cases where the a special out of band transition that can
not be triggered by input is useful in separating match conditions
in the dfa encoding.

The null_transition is currently used as an out of band transition
for match conditions that can not contain a \0 in their input
but apparmor needs an out of band transition for cases where
the match condition is allowed to contain any input character.

Achieve this by allowing for an explicit transition out of input
range that can only be triggered by code.

Signed-off-by: John Johansen

apparmor: add a valid state flags check

2020-01-18T23:37:24+00:00

Add a check to ensure only known state flags are set on each
state in the dfa.

Signed-off-by: John Johansen

Merge tag 'apparmor-pr-2019-12-03' of git://git.kernel.org/pub/scm/linux/kernel/git/jj/linux-apparmor

2019-12-03T20:51:35+00:00

Pull apparmor updates from John Johansen:
 "Features:

   - increase left match history buffer size to provide improved
     conflict resolution in overlapping execution rules.

   - switch buffer allocation to use a memory pool and GFP_KERNEL where
     possible.

   - add compression of policy blobs to reduce memory usage.

  Cleanups:

   - fix spelling mistake "immutible" -> "immutable"

  Bug fixes:

   - fix unsigned len comparison in update_for_len macro

   - fix sparse warning for type-casting of current->real_cred"

* tag 'apparmor-pr-2019-12-03' of git://git.kernel.org/pub/scm/linux/kernel/git/jj/linux-apparmor:
  apparmor: make it so work buffers can be allocated from atomic context
  apparmor: reduce rcu_read_lock scope for aa_file_perm mediation
  apparmor: fix wrong buffer allocation in aa_new_mount
  apparmor: fix unsigned len comparison with less than zero
  apparmor: increase left match history buffer size
  apparmor: Switch to GFP_KERNEL where possible
  apparmor: Use a memory pool instead per-CPU caches
  apparmor: Force type-casting of current->real_cred
  apparmor: fix spelling mistake "immutible" -> "immutable"
  apparmor: fix blob compression when ns is forced on a policy load
  apparmor: fix missing ZLIB defines
  apparmor: fix blob compression build failure on ppc
  apparmor: Initial implementation of raw policy blob compression

apparmor: increase left match history buffer size

2019-06-20T17:33:31+00:00

There have been cases reported where a history buffer size of 8 was
not enough to resolve conflict overlaps. Increase the buffer to and
get rid of the size element which is currently just storing the
constant WB_HISTORY_SIZE.

Signed-off-by: John Johansen