Clean up the existing export namespace code along the same lines of
commit 33def8498fdd ("treewide: Convert macro and uses of __section(foo)
to __section("foo")") and for the same reason, it is not desired for the
namespace argument to be a macro expansion itself.

Scripted using

  git grep -l -e MODULE_IMPORT_NS -e EXPORT_SYMBOL_NS | while read file;
  do
    awk -i inplace '
      /^#define EXPORT_SYMBOL_NS/ {
        gsub(/__stringify\(ns\)/, "ns");
        print;
        next;
      }
      /^#define MODULE_IMPORT_NS/ {
        gsub(/__stringify\(ns\)/, "ns");
        print;
        next;
      }
      /MODULE_IMPORT_NS/ {
        $0 = gensub(/MODULE_IMPORT_NS\(([^)]*)\)/, "MODULE_IMPORT_NS(\"\\1\")", "g");
      }
      /EXPORT_SYMBOL_NS/ {
        if ($0 ~ /(EXPORT_SYMBOL_NS[^(]*)\(([^,]+),/) {
  	if ($0 !~ /(EXPORT_SYMBOL_NS[^(]*)\(([^,]+), ([^)]+)\)/ &&
  	    $0 !~ /(EXPORT_SYMBOL_NS[^(]*)\(\)/ &&
  	    $0 !~ /^my/) {
  	  getline line;
  	  gsub(/[[:space:]]*\\$/, "");
  	  gsub(/[[:space:]]/, "", line);
  	  $0 = $0 " " line;
  	}

  	$0 = gensub(/(EXPORT_SYMBOL_NS[^(]*)\(([^,]+), ([^)]+)\)/,
  		    "\\1(\\2, \"\\3\")", "g");
        }
      }
      { print }' $file;
  done

Requested-by: Masahiro Yamada <masahiroy@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://mail.google.com/mail/u/2/#inbox/FMfcgzQXKWgMmjdFwwdsfgxzKpVHWPlc
Acked-by: Greg KH <gregkh@linuxfoundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

the kernel test robot reports a C23 extension
warning: label followed by a declaration is a C23 extension
[-Wc23-extensions]
     696 |                 struct aa_profile *new_profile = NULL;

Instead of adding a null statement creating a C99 style inline var
declaration lift the label declaration out of the block so that it no
longer immediatedly follows the label.

Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202411101808.AI8YG6cs-lkp@intel.com/
Fixes: ee650b3820f3 ("apparmor: properly handle cx/px lookup failure for complain")
Signed-off-by: John Johansen <john.johansen@canonical.com>

The wording of 'scrubbing environment' implied that all environment
variables would be removed, when instead secure-execution mode only
removes a small number of environment variables. This patch updates the
wording to describe what actually occurs instead: setting AT_SECURE for
ld.so's secure-execution mode.

Link: https://gitlab.com/apparmor/apparmor/-/merge_requests/1315 is a
merge request that does similar updating for apparmor userspace.

Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
Signed-off-by: John Johansen <john.johansen@canonical.com>

The macros for label combination XXX_comb are no longer used and there
are no plans to use them so remove the dead code.

Signed-off-by: John Johansen <john.johansen@canonical.com>

In the macro definition of next_comb(), a parameter L1 is accepted,
but it is not used. Hence, it should be removed.

Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>
Signed-off-by: John Johansen <john.johansen@canonical.com>

The previous audit_cap cache deduping was based on the profile that was
being audited. This could cause confusion due to the deduplication then
occurring across multiple processes, which could happen if multiple
instances of binaries matched the same profile attachment (and thus ran
under the same profile) or a profile was attached to a container and its
processes.

Instead, perform audit_cap deduping over ad->subj_cred, which ensures the
deduping only occurs across a single process, instead of across all
processes that match the current one's profile.

Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
Signed-off-by: John Johansen <john.johansen@canonical.com>

When auditing capabilities, AppArmor uses a per-CPU, per-profile cache
such that the same capability for the same profile doesn't get repeatedly
audited, with the original goal of reducing audit logspam. However, this
cache does not have an expiration time, resulting in confusion when a
profile is shared across binaries (for example) and an expected DENIED
audit entry doesn't appear, despite the cache entry having been populated
much longer ago. This confusion was exacerbated by the per-CPU nature of
the cache resulting in the expected entries sporadically appearing when
the later denial+audit occurred on a different CPU.

To resolve this, record the last time a capability was audited for a
profile and add a timestamp expiration check before doing the audit.

v1 -> v2:
 - Hardcode a longer timeout and drop the patches making it a sysctl,
   after discussion with John Johansen.
 - Cache the expiration time instead of the last-audited time. This value
   can never be zero, which lets us drop the kernel_cap_t caps field from
   the cache struct.

Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
Signed-off-by: John Johansen <john.johansen@canonical.com>

The profile_capabile function takes a struct apparmor_audit_data *ad,
which is documented as possibly being NULL. However, the single place that
calls this function never passes it a NULL ad. If we were ever to call
profile_capable with a NULL ad elsewhere, we would need to rework the
function, as its very first use of ad is to dereference ad->class without
checking if ad is NULL.

Thus, document profile_capable's ad parameter as not accepting NULL.

Signed-off-by: Ryan Lee <ryan.lee@canonical.com>
Signed-off-by: John Johansen <john.johansen@canonical.com>

Multiple profiles shared 'ent->caps', so some logs missed.

Fixes: 0ed3b28ab8bf ("AppArmor: mediation of non file objects")
Signed-off-by: chao liu <liuzgyid@outlook.com>
Signed-off-by: John Johansen <john.johansen@canonical.com>

Add a comment to unpack_perm to document the first entry in the packed
perms struct is reserved, and make a non-functional change of unpacking
to a temporary stack variable named "reserved" to help suppor the
documentation of which value is reserved.

Suggested-by: Serge E. Hallyn <serge@hallyn.com>
Signed-off-by: John Johansen <john.johansen@canonical.com>