Bug: https://bugs.launchpad.net/bugs/1366572
In the move of the apparmor.vim manpage source from the parser
directory to utils/vim/, the creation of the html version of the
manpage was lost. This patch fixes that, as well as fixes the vim
Makefile to use the common/Make.rules _clean target (which clears the
pod2htm*.tmp files created by pod2html as well).
It also fixes a bug in common/Make.rules where the _clean target
would report an error when used in a directory where the Makefile
doesn't set the NAME variable, such as utils/vim/.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
- change --help for files - "Profile(s) to merge" instead of "base profile"
- display the profile to save when asking to save it
- disable searching for existing network rules in abstractions because
it crashes. This doesn't hurt too much, see
https://bugs.launchpad.net/apparmor/+bug/1382241
Acked-by: Steve Beattie <steve@nxnw.org>
- remove some debug output (which Kshitij intentionally kept in the
draft patch)
- add a UI_Info to display which profile will be merged
- disable the mergeprofiles.clear_common() call because it crashes
(https://bugs.launchpad.net/apparmor/+bug/1382236)
- disable (M)ore (CMD_OTHER) because it crashes
- make (F)inish work everywhere
- change the help text so that it doesn't mention 3-way-merge until we
implement it
The python utils didn't know about the hat keyword, for example
hat foo {
This patch changes RE_PROFILE_HAT_DEF to add support for the hat keyword.
Note that this patch only changes profile reading - when saving a
profile, the ^foo syntax will always be used.
While on it, also convert the regex to named matches - the result
numbering changed anyway ;-)
Acked-by: Kshitij Gupta <kgupta8592@gmail.com>
This is the rebased version of the patch by
Kshitij Gupta <kgupta8592@gmail.com>
(mostly) original patch description:
Changes to facilitate 2-way merge (maybe also 3-way) of multiple
profiles as discussed on IRC
This patch
- moves reset method to reset_aa function
- modifies message displayed to user
- allows processing of multiple files in 2-way merge
- disables 3-way merge till new syntax has been decided
The changes reflect the approach of providing arbitrary number of
files using wildcards or explicitly.
The changes map the profiles in the given files to their respective
files in the local directory specified using -d. Then the merges take
place profile-wise.
Acked-by: Steve Beattie <steve@nxnw.org>.
When adding inet rules in aa-logprof, it crashes with
IndexError: list index out of range
The reason is that it doesn't display the options if only the raw rule
is available (aka "no abstraction").
This patch checks if options[] is set and otherwise sets selection to
the raw rule.
Acked-by: Steve Beattie <steve@nxnw.org>
After several hours of debugging on IRC and flooding paste.opensuse.org ;-)
I finally found the reason - reading(!) from log_dict with a wrong
key caused the creation of that strange mode in log_dict.keys().
While finding that bug was very hard, the fix is easy - just replace
"profile" with "aamode". (That probably makes one char per hour of
debugging...)
To improve that ratio, also add a warning to common.py so that this
interesting[tm] behaviour of hasher() is at least documented.
Acked-by: Steve Beattie <steve@nxnw.org>
...
File "/home/cb/apparmor/HEAD-CLEAN/utils/apparmor/severity.py", line 147, in handle_variable_rank
variable = regex_variable.search(resource).groups()[0]
AttributeError: 'NoneType' object has no attribute 'groups'
handle_variable_rank() checked with if '@' in resource:
and if it finds it, expects it can match a variable, which means @{.....}
If a filename contains a @ this fails.
The patch fixes the if condition so that it does a regex match.
It also adds two testcases for filenames containing @ to make sure they
don't cause a crash and result in the exptected severity rank.
Acked-by: Steve Beattie <steve@nxnw.org>
aa-cleanprof (and others?) duplicate quotes in file rules.
If a profile contains
"/bin/foo bar" mrix,
and I run aa-cleanprof on it several times, I end up with
""""""/bin/foo bar"""""" mrix,
This patch calls strip_quotes on the pathname.
(If needed, the quotes are re-added when writing the profile - tested
with aa-cleanprof.)
References: https://bugs.launchpad.net/apparmor/+bug/1328707
Acked-by: Kshitij Gupta <kgupta8592@gmail.com>
This patch changes open_file_read() and open_file_write() to use
errors='surrogateescape' (with fallback to 'replace' for py2).
This avoids a crash when reading a logfile with special characters that
are not utf8-encoded (for example a latin1 "ö"), and also avoids crashes
at several other places we don't know yet ;-)
The patch also changes open_file_read() and open_file_write() to wrapper
functions, and moves the "real" code to the new open_file_anymode()
function.
Also, I removed the try/except - it's superfluous because it throws the
exception without any modifications.
Acked-by: Kshitij Gupta <kgupta8592@gmail.com>
This patch reverts commit 2131 which added support for the newer apache
2.4 ap_hook_check_access_ex() api, based on a report from Christian that
it broke apache's simple authentication.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
With the two recent unix socket test patches to mark dgram tests as
failing but expected to pass, I realized that there's no indication
in the output that there are current expected failures (except for
the single corefile test expected failure)[0]. This patch attempts to
remedy that by emitting the x-type plus the test description for each
test that is marked as such. I've set it to always emit these results.
[0] The test infrastructure does give an alert if there is a test that
has been marked xpass or xfail but has started behaving correctly.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
It's not been tracked down in
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1373172 why
this is happening, but the current unnamed unix socket dgram tests
are failing when only the server is confined, and the peer label is
given as only the confining profile (the stream and seqpacket dgram
tests/permissions don't seem to trigger this revalidation rejection).
Until this bug is diagnosed and addressed, mark these tests as failing
but expected to pass (i.e. 'xpass').
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
Currently the apparmor parser warns about rules that are not enforced or
downgraded. This is a problem for distros that are not carrying the out of
tree kernel patches, as most profile loads result in warnings.
Change the behavior to not output a message unless a warn flag is passed.
This patch adds 2 different warn flags
--warn rule-downgraded # warn if a rule is downgraded
--warn rule-not-enforced # warn if a rule is not enforced at all
If the warnings are desired by default the flags can be set in the
parser.conf file.
v2 of patch
- update man page
- add --warn to usage statement
- make --quiet clear warn flags
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
This patch:
- allows the unsupported rules to be stored when parsing profiles
- writes all the unsupported rules back to profile
Acked-by: Christian Boltz <apparmor@cboltz.de>
match groups.
Also split out RE_OWNER that matches the "owner" keyword.
To make the code easier to understand, I dropped the existing audit
variable and instead directly query the "new" audit variable while
filling path_rule['audit'].
Acked-by: Steve Beattie <steve@nxnw.org>
YaST still uses AppArmor.pm, and now errors out when starting the
profile editor because it doesn't know about signal, unix, ptrace and
dbus rules.
This patch adds basic support for those rules to AppArmor.pm by adding
them to the "ignore those rules" regex.
Note: Rules covered by this regex are lost when writing the profile
therefore the patch adds a comment to at least make this a "known bug".
References:https://bugzilla.novell.com/show_bug.cgi?id=900013
Acked-by: Steve Beattie <steve@nxnw.org>
The following patch:
- creates a class for prompt questions moving away from Perl hash hack
for the purpose.
- moves some functions to the methods for that class
- fix options being incorrectly passed to questionPrompt in aa-mergeprof
Acked-by: Christian Boltz <apparmor@cboltz.de>
socket. Note, DBus mediation is still in effect so rules still need to be added
for accessing the DBus API (LP: #1375067)
Acked-By: Jamie Strandboge <jamie@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
utils/test/runtests-py*.sh always exits with $? = 1 even if there is no
error. This is caused by the last executed command, test -n
This patch changes it to test -z so that we'll get $? = 0 if all tests
succeed.
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Bug: https://bugzilla.novell.com/show_bug.cgi?id=895495
We define the __unused macro as a shortcut for __attribute__((unused))
to quiet compiler warnings for functions where an argument is unused,
for whatever reason. However, on 64 bit architectures, older glibc's
bits/stat.h header defines an array variable with the name __unused
that collides with our macro and causes the parser to fail to build,
because the resulting macro expansion generates invalid C code.
This commit fixes the issue by removing the __unused macro where it's
not needed (mod_apparmor) and renaming it to 'unused' elsewhere. It also
in some instances reorders the arguments so that the unused macro
appears last consistently.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
Bug: https://bugs.launchpad.net/bugs/1339727
LightDM keeps moving the location where it stores xauthority files for
users, when configured to store them in a system directory (e.g. with
[LightDM]
user-authority-in-system-dir=true
set in a lightdm configuration file).
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
BugLink: http://bugs.launchpad.net/bugs/1373085
The parser fails to accept certain characters, even when escaped
or quoted as part of the profile or label name in ipc rules. This
is due to the lexer not accepting those characters as part of the
input pattern.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
(capability is one of the easiest rule types, so it's good as a start.)
The patch also adds basic support for rules containing more than one
capability, like
capability chown dac_override,
Note that this is just a pass-through mode (instead of complaining about
an invalid line). aa-logprof will happily add another "capability chown"
if it hits a log entry for it. (But: we never got a bugreport about not
supporting multi-capability lines, so I guess they are rarely used ;-)
I also added a parse_audit_allow() function to handle the audit and
allow/deny keywords. They are used in most rule types, which means we
can get rid of some duplicated code with this function.
Finally, update utils/test/test-regex_matches.py - RE_PROFILE_CAP now
has 5 instead of 4 match groups because of the added multi-capability
support.
While on it, I also improved the error message in setup_regex_tests()
to also show the rule that causes a problem.
Acked-by: Steve Beattie <steve@nxnw.org>
Bug: https://bugs.launchpad.net/bugs/1375516
The unix_socket test program calls getsockopt() after calling bind().
Because AppArmor continues to use traditional file rules for sockets
bound to a filesystem path, it does not mediate some socket operations
after the socket has been bound to the filesystem path. The getopt
permission is one of those socket operations.
To account for this lack of mediation, the getopt permission should be
removed from the server permissions list.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Tests abstract UNIX domain sockets with various combinations of implied
permissions, explicit permissions, and conditionals. It also tests with
bad permissions and conditionals.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Iterate through the individual client and server AF_UNIX pathname
permissions and remove them, one-by-one, to verify that the test fails.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
The client and server programs require a different set of AF_UNIX
permissions. This patch makes it so that the unix rules are constructed
differently depending on the program under test.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Tests abstract UNIX domain sockets with various combinations of implied
permissions, explicit permissions, and conditionals. It also tests with
bad permissions and conditionals.
The new file unix_socket.inc includes a generic set of tests that can be
reused by another test script in order to test unnamed AF_UNIX socket
mediation. The do_test() function is conditionalized in a way that it
can test confined servers and confined clients depending on the
arguments passed in.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
The unix_socket operations for testing getopt and setopt permissions
were occurring back to back. This patch breaks them up into "pre-bind"
and "post-bind" operations. The setopt operation now occurs pre-bind
while the getopt operation happens post-bind. This allows for the test
policy to test setopt without an addr= conditional and to test getopt
with an addr= conditional.
Additionally, the wrapper functions that call setsockopt()/getsockopt()
are moved into a new file that both unix_socket.c and
unix_socket_client.c can reuse.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
The client will now do a getsockname() on its socket in order to test
the AppArmor 'getattr' unix rule permission.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
The server will now do a shutdown() on its socket in order to test the
AppArmor 'shutdown' unix rule permission.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Use the sendto()/recvfrom() functions when dealing with dgram sockets in
unix_socket_client.
This allows us to test different interfaces besides the typical
write()/read() and will allow for a smaller permissions set for
unix_socket_client.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
The unix_socket_client test program was using an abstract socket, which
was set up using the autobind feature, when testing any socket address
types.
To more accurately test a specific address type, this patch changes the
client code to use whatever address type that the server is using. The
string ".client" will be added to the end of the server's address.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Two weeks ago, I fixed various tests in minitools_test.py which also
included disabling the test if aa-complain deletes the force-complain
symlink because nothing (especially aa-complain) creates those symlinks.
Seth didn't like the removal of that test too much. Therefore this patch
"manually" creates the force-complain symlink and tests that it's removed
by aa-enforce.
Acked-by: Seth Arnold <seth.arnold@canonical.com>
- usr.lib.dovecot.auth needs /{var/,}run/dovecot/auth-token-secret.dat{,.tmp} rw,
- usr.lib.dovecot.imap requests block_suspend, which I propose to deny as usual
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Currently the parser is bailing when it fails to load a profile,
not processing any potential subsequent profiles in the dir or passed
in list. This results in all policy after the first error failing
to load, instead of just the profile(s) with the error.
This is a different behavior than what has been done by initscripts
that have driven it with xargs -n1, passing it a single profile
at a time.
Fix this so that the parser only exits on first error if specifically
told to do so.
Note: this does not fix the various failure points in the parser
that call exit, instead of returning an error.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>, thanks.
With the move to C++-ification of the parser, the parser's makefile was
not updated to take into account .cc files when deriving object files.
This would result in the final linking compilation of the parser binary
including all of the .cc files in its command line, rather than the ,o
files. This patch fixes the issue as well as an additional typo in the
dependency list for af_unix.o that was not triggered because af_unix.o
was not being built independently.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Update unix_socket and unix_socket_client to use setsockopt() in order
to set send and receive timeouts for socket IO operations. This takes
the place of poll(). Poll() was not being used for all potentially
blocking socket operations which could have resulted in test cases
blocking infinitely.
This also has the nice side effect of using getsockopt() and
setsockopt(). These are AppArmor mediation points in kernel ABI v7 so it
is worthwhile to test the calls while under confinement.
This patch updates the existing v7 policy generation to allow the getopt
and setopt accesses.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
The AppArmor kernel ABI v7 requires that a 'unix create,' rule be
granted to confined processes that call socket(AF_UNIX, type, 0). This
is true for pathname, abstract, and unnamed UNIX domain sockets since
the address type of a socket is not yet known when socket(2) is called.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
It is too complicated, due to the number of corner cases, to write a
script that generates the rules for each AF_UNIX address type (pathname,
abstract, and unnamed) and socket type (stream, dgram, and seqpacket).
This patch moves the AF_UNIX pathname tests into their own file with the
intent of having each address type be tested in their own file.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Instead of using the entire sun_path buffer for abstract socket names,
only use the exact length of the string that is specified on the command
line. The nul-terminator is not included for abstract sockets.
The size of sun_path is modified to include the nul-terminator for
pathname address types.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
out common parts (like audit and deny flags or the typical end of the
line (comma and comment).
The patch also introduces the named match groups <audit>, <allow> and
<comment> which we can start to use whenever we want. The group
numbering is kept unchanged, so we can migrate one regex / rule type
after the other to named groups (not only audit, allow and comments, but
changing all regexes will be another patch ;-)
As a side effect, fix RE_PROFILE_CHANGE_PROFILE, RE_PROFILE_ALIAS and
RE_PROFILE_RLIMIT which did not allow ", #" (comma, space, comment) at
the end of the line.
Acked-by: Kshitij Gupta <kgupta8592@gmail.com>
This patch updates the parser code to reject rules that contain local
socket permissions and peer conditional elements. The error message for
that condition is also corrected to resolve a copy and paste mistake
from the D-Bus rule parsing code.
The patch also updates the man page to correctly describe the two sets
of socket permissions and fixes an example rule that resulted in a
parser error after the change described above.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
while it is not possible to specify a rule with local conditionals with
peer conditionals
eg.
unix listen peer=(addr=@foo),
a rule such as
unix peer=(addr=@foo),
is possible, and was setting all permissions for local as well as the peer
condition permissions.
Currently this means the create permission must be specified in a separate
rule from a rule with a peer= condition, if create is to be allowed. This
isn't too much of an issue but it does mean rule such as
unix connect peer=(addr=@foo),
Can not imply the ability to create a socket. Which may indeed be the
behavior if we wish to enforce that the socket was created in another
process and passed in. Is this what we want to do?
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
The new af_unix apparmor kernel patches include the first step towards
implicit labeling. As a result, when a file descriptor is inherited
across one profile boundary to another, both labels' policies are
checked for valid access to the file descriptor. However, due to a quirk
in the linux kernel, when a socket is opened, the file descriptor is
marked as having read and write (aka send and receive) access. When the
crosscheck revalidation occurs, this means that the policy being
inherited from requires read/write access to the socket descriptor, even
if the process never reads or writes to it. This resulted in a few
failures in the socketpair tests.
The following patch adjusts the failing tests to include the neccessary
send and receive permissions, as well as adding additional tests that
are expected to fail when they are not present, to try to ensure that
if our crosscheck behavior changes, we catch it.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
Based on a patch from Felix Geyer who wrote in April:
> On Ubuntu trusty the php package creates config symlinks in
> /etc/php5/cli/conf.d/, /etc/php5/cgi/conf.d/ and
> /etc/php5/fpm/conf.d/ to /etc/php5/mods-available/.
This patch is a simplified version of his patch that allows
/etc/php5/**.ini r and /etc/php5/**/ r
Acked-by: Seth Arnold <seth.arnold@canonical.com> on IRC
(after menacing an Acked-by: <timeout>)
both real [0] and hypothetical (e.g. gui prototypes), as part of
the python utils merge, some namespace packaging bits were added
to apparmor/__init__.py, based on the (not very clear) advice given
in python's pep 0420 [1]. However, a side effect of this is that it
causes system installed versions of python modules to be used over
paths specified via PYTHONPATH [2], which breaks our in-tree tests
when the system versions of the python modules are out of date with
respect to the in-tree version.
It seems based on testing, however, that carrying this code snippet
is no longer necessary to have external modules be found. Thus,
the following patch drops it.
[0] e.g. https://launchpad.net/click-apparmor
[1] http://www.python.org/dev/peps/pep-0420/
[2] a python upstream discussion about this occurred at
https://mail.python.org/pipermail/distutils-sig/2014-March/024049.html
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
This patch adds a 'check_pod_files' make target to the common make
rules, and then fixes the errors it highlighted as well as most of
the warnings. It will cause 'make check' in most of the directories to
fail if there are errors in a pod file (but not if there are warnings).
Common issues were:
- using an '=over/=back' pair for code-like snippets that did not
contain any =items therein; the =over keyword is intended for
indenting lists of =item entries, and generates a warning if
there isn't any.
- not escaping '<' or '>'
- blank lines that contained spaces or tabs
The second -warnings flag passed to podchecker is to add additional
warnings, un-escaped '<' and '>' being of them.
I did not fix all of the warnings in apparmor.d.pod, as I have not come
up with a good warning-free way to express the BNF of the language
similar in format to what is currently generated. The existing
libapparmor warnings (complaints about duplicate =item definition
names) are actually a result of passing the second -warnings flag.
The integration into libapparmor is suboptimal due to automake's
expectation that there will be a test driver program(s) for make check
targets; that's why I added the podchecker call to the manpage
generation point.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
---
changehat/mod_apparmor/Makefile | 3
changehat/mod_apparmor/mod_apparmor.pod | 28 ++-
common/Make.rules | 4
libraries/libapparmor/doc/Makefile.am | 7
parser/Makefile | 2
parser/apparmor.d.pod | 275
+++++++++++++-------------------
utils/Makefile | 3
utils/aa-cleanprof.pod | 2
utils/aa-complain.pod | 2
utils/aa-decode.pod | 2
utils/aa-easyprof.pod | 69 +++-----
utils/aa-enforce.pod | 2
utils/aa-genprof.pod | 2
utils/aa-logprof.pod | 6
utils/aa-sandbox.pod | 64 ++-----
utils/logprof.conf.pod | 2
utils/vim/Makefile | 2
17 files changed, 212 insertions(+), 263 deletions(-)
This patch:
- replaces unnamed arguments with named arguments wherever more than 1
one arguments ware present in a message
- minor fix in aa-unconfined for pname argument in 2 strings
- updated pot files (as a side-effect of testing with make)
Acked-by: Christian Boltz <apparmor@cboltz.de>
In profile.h, flagvals is declared to be class, but then in the
Profile class, the flags field declares it as a struct. This patch
makes the field declaration type consistent.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
In trunk commit 2615, make targets for af_rule.o and af_unix.o were
added. Unfortunately, the af_rule.o target's dependency on rule.h was
missing the .h suffix. This patch fixes the issue and adds some other
headers that the source file are dependent on.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
Put a bare unix rule in the core gendbusprofile() function that all
dbus_*.sh use. We aren't interested in testing AF_UNIX mediation in the
dbus tests, since that's already done elsewhere, so we'll
unconditionally allow full AF_UNIX access to prevent test breakage
caused by any future changes in libdbus.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
A side effect of not including utils/apparmor/*.py in the .pot file was
that some translations were lost. This patch includes backported (or
forward-ported?) translations from r2186. It's not a simple merge, I
reviewed everything I merged and changed it if necessary.
I also removed the outdated
"Language-Team: Novell Language <language@novell.com>\n"
Acked-by: Seth Arnold <seth.arnold@canonical.com>
In the conversion from perl to python, it got overlooked to add the
python-apparmor modules to the set of things to search for translatable
strings in. This patch addresses the issue.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
missing <sys/sysctl.h> header. This header is included by accident, a
vestige of earlier days, and wasn't removed when the sysctls were removed.
(Think Linux 2.0 or Linux 2.2 days.)
See also https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=760378
Thanks to Thorsten Glaser for the discovery and initial fix.
The Debian and Ubuntu Ruby 1.9.1 package is configured like this:
--with-vendordir='/usr/lib/ruby/vendor_ruby' --with-sitedir='/usr/local/lib/site_ruby
These paths are missing in the ruby abstraction.
Patch by Felix Geyer <debfx@ubuntu.com>
Acked-by: Christian Boltz <apparmor@cboltz.de>
> Allow dnsmasq read access to IPv6 config
The commit did not match this part of the commit message
> slightly modified (../conf/**/mtu -> ../conf/*/mtu)
which I'm fixing now.
The IPv6 Neighbor Discovery protocol (RFC 2461) suggests
implementations provide MTU in Router Advertisement (RA)
messages. From section 4.2
MTU SHOULD be sent on links that have a variable MTU
(as specified in the document that describes how to
run IP over the particular link type). MAY be sent
on other links.
dnsmasq supports this option and should have read access
to an interface's MTU.
Patch by James Fehlig <jfehlig@suse.com>
slightly modified (../conf/**/mtu -> ../conf/*/mtu)
Acked-by: Seth Arnold <seth.arnold@canonical.com>
getopt, setopt and shutdown. This was added based on incorrect logging in early
iterations of the abstract kernel patches which have since been fixed. These
options don't make sense with peer=(addr=none), so drop that.
Acked-By: Jamie Strandboge <jamie@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
Unrequested replies are message types that are typically replies, such
as error and method_return message types, but have not been requested by
the recipient.
The AppArmor mediation code in dbus-daemon allows requested reply
messages through if the original message was allowed. However,
unrequested reply messages should be checked against the system policy
to make certain that they should be allowed.
This test verifies that the dbus-daemon is properly querying system
policy when it detects that a message is an unrequested reply.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch adjusts the bison grammer in libapparmor and the parser
to use the %define api.pure directive instead of the deprecated
%pure_parser and %pure-parser keywords. Bison had been warning about
the former:
libraries/libapparmor/src/grammar.y:71.1-12: warning: deprecated directive, use ‘%pure-parser’ [-Wdeprecated]
%pure_parser
^^^^^^^^^^^^
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
- fix typo
- fix whitespace
- add netlink
- update for change from path to addr
- remove TODO items
- add and document examples
- remove undocumented 'unix server addr=@foo,' example
Signed-off-by: Jamie Strandboge <jamie@canonical.com>
Acked-by: Christian Boltz <apparmor@cboltz.de>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch adds support for the mount and pivotroot related keywords,
fstype, flags, and srcname.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
When signals and ptrace mediation were added to apparmor, the aalogparse
routines were not adjusted to compensate. This patch adds support for
the signal and peer keywords.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
The upstream kernel at some point between the 3.13 and 3.16 kernel
adjusted the output of audit messages to include an additional "audit:"
keyword. e.g. a 3.13 message would look like:
kernel: [182243.243324] type=1400 audit(1409684003.960:273342): [SNIP]
whereas in 3.16, it looks like:
kernel: [182243.243324] audit: type=1400 audit(1409684003.960:273342): [SNIP]
^^^^^^
This patch adjust the libapparmor aalogparse grammar and lexer to
compensate for this change.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Convert the signal parse tests to use common AAParseTest super class in
common_test.py.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
Convert the ptrace parse tests to use common AAParseTest super class
in common_test.py.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
Convert the pivotroot parse tests to use common AAParseTest super
class in common_test.py.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
Convert the mount parse tests to use common AAParseTest super class in
common_test.py.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
Convert the DBUS parse tests to use common AAParseTest super class in
common_test.py.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
This patch abstracts out parse tests into a super class to inherit from
and converts the af_unix parse tests to use the super class.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
The patch that adds support for af_unix rules added a _Raw_Rule base
class to inherit from in rules.py. This patch converts the rest of the
raw rules classes to use the same.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
This patch moves the assignment of the regex function into the unittest
setUp() function rather than at script load time. If for some reason
the python utils library does not define the relevant function, without
this patch the script fails entirely; with it, each individual test
class that depends on the missing regex will fail each test case.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
This patch adds limited support for af_unix rules in the python
utilities, of the "don't touch them, but don't throw a python backtrace
when coming across them, either" variety. Testcases are added as well.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
The listen and setopts commands have broken encodings because the
tmp stream they use to handle diverging from the other commands
has does not set its write position to to the end of the copied data.
Instead the write head is set to the beginning so that when the
new data for the command is written it overwrites the begging of
the command instead of appending to it.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Fix to allow specifying the unix perm with peer perms. This is allowed
now and even supported, since for unix sockets the peer accept is
mediated in the unix_stream_connect hook (something that is not
possible in the lsm accept hook).
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This changes/fixes the encoding for unix socket rules. The changes
look larger than they are because it refactors the code, instead
of duplicating.
The major changes are:
- it changes where the accept perm is stored
- it moves anyone_match_pattern to default_match_pattern
- it fixes the layout of the local addr only being written when local
perms are present
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Let unix keyword accept bare send, receive keywords and add more
simple unix acceptance test cases.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
The old dfa table format has 2 64 bit permission field used to store
all of allow, quiet, audit, owner/!owner and transition mask. This
leaves 7 bits for entry + a few other special bits.
Since policydb entries when using old style dfa permission format
don't use support the !owner permission entries we can map, the
high net work permission bits to these entries.
This allows us to enforce base network permissions on system with
only support for the old dfa table format.
Bits 0-7 inclusive stay put
Bits 8-9 inclusive move (14 - 8) = 6 to 14-15 GETATTR | SETATTR
Bits 20-22 inclusive move -4 to 16-18 ACCEPT | BIND | LISTEN (notice 22 not 23)
Bit 23 is skipped, hence the need to shift 5 for 24-25 instead of 4
Bits 24-25 inclusive move -5 to 19-20
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Refactor add_new_state into two versions, one that splits anodes from
nnodes, and one for use when anodes and nnodes are presplit
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
The shared node type will be used in the future to add new capabilities
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
We need to rework permission type mapping to nodesets, which means we
need to move the nodeset computations earlier in the dfa creation
processes, instead of a post step of follow(), so move the nodeset
into expr-tree
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch tells the parser to do af_unix processing while running the
parser sanity tests, letting the af_unix tests generate the correct
results.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
This patch fixes a segfault that was occurring in testing over the
weekend. The problem existed in the original patch that adds af_unix
rules (lp:apparmor commit 2615).
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
This patch converts the path= modifier to the af_unix rules to use
addr= instead.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch implements parsing of fine grained mediation for unix domain
sockets, that have abstract and anonymous paths. Sockets with file
system paths are handled by regular file access rules.
The unix network rules follow the general fine grained network
rule pattern of
[<qualifiers>] af_name [<access expr>] [<rule conds>] [<local expr>] [<peer expr>]
specifically for af_unix this is
[<qualifiers>] 'unix' [<access expr>] [<rule conds>] [<local expr>] [<peer expr>]
<qualifiers> = [ 'audit' ] [ 'allow' | 'deny' ]
<access expr> = ( <access> | <access list> )
<access> = ( 'server' | 'create' | 'bind' | 'listen' | 'accept' |
'connect' | 'shutdown' | 'getattr' | 'setattr' |
'getopt' | 'setopt' |
'send' | 'receive' | 'r' | 'w' | 'rw' )
(some access modes are incompatible with some rules or require additional
parameters)
<access list> = '(' <access> ( [','] <WS> <access> )* ')'
<WS> = white space
<rule conds> = ( <type cond> | <protocol cond> )*
each cond can appear at most once
<type cond> = 'type' '=' ( <AARE> | '(' ( '"' <AARE> '"' | <AARE> )+ ')' )
<protocol cond> = 'protocol' '=' ( <AARE> | '(' ( '"' <AARE> '"' | <AARE> )+ ')' )
<local expr> = ( <path cond> | <attr cond> | <opt cond> )*
each cond can appear at most once
<peer expr> = 'peer' '=' ( <path cond> | <label cond> )+
each cond can appear at most once
<path cond> = 'path' '=' ( <AARE> | '(' '"' <AARE> '"' | <AARE> ')' )
<label cond> = 'label' '=' ( <AARE> | '(' '"' <AARE> '"' | <AARE> ')')
<attr cond> = 'attr' '=' ( <AARE> | '(' '"' <AARE> '"' | <AARE> ')' )
<opt cond> = 'opt' '=' ( <AARE> | '(' '"' <AARE> '"' | <AARE> ')' )
<AARE> = ?*[]{}^ ( see man page )
unix domain socket rules are accumulated so that the granted unix
socket permissions are the union of all the listed unix rule permissions.
unix domain socket rules are broad and general and become more restrictive
as further information is specified. Policy may be specified down to
the path and label level. The content of the communication is not
examined.
Some permissions are not compatible with all unix rules.
unix socket rule permissions are implied when a rule does not explicitly
state an access list. By default if a rule does not have an access list
all permissions that are compatible with the specified set of local
and peer conditionals are implied.
The 'server', 'r', 'w' and 'rw' permissions are aliases for other permissions.
server = (create, bind, listen, accept)
r = (receive, getattr, getopt)
w = (create, connect, send, setattr, setopt)
In addition it supports the v7 kernel abi semantics around generic
network rules. The v7 abi removes the masking unix and netlink
address families from the generic masking and uses fine grained
mediation for an address type if supplied.
This means that the rules
network unix,
network netlink,
are now enforced instead of ignored. The parser previously could accept
these but the kernel would ignore anything written to them. If a network
rule is supplied it takes precedence over the finer grained mediation
rule. If permission is not granted via a broad network access rule
fine grained mediation is applied.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
mistakenly did not incorporate feedback from Seth Arnold. Specifically, don't
specify label=unconfined on the abstract sockets.
Acked-By: Jamie Strandboge <jamie@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Update mdnsd for fine-grained netlink mediation. A mdnsd binary was not
available to test but code inspection showed it set up the socket the same as
avahi, which uses SOCK_DGRAM type instead of SOCK_RAW with netlink.
Acked-By: Jamie Strandboge <jamie@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
- the base abstraction for common abstract and anonymous rules (comments
included per rule)
- dbus-session-strict to add a rule for connecting to the dbus session
abstract
socket. I used 'peer=(label=unconfined)' here, but I could probably lose the
explicit label if people preferred that
- X to add a rule for connecting to the X abstract socket. Same as for
dbus-session-strict
- nameservice to add a rule for connecting to a netlink raw. This change could
possibly be excluded, but applications using networking (at least on Ubuntu)
all seem to need it. Excluding it would mean systems using nscd would need to
add this and ones not using it would have a noisy denial
Acked-By: Jamie Strandboge <jamie@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
before the af type protocol mappings patch was applied, a single rule could
result in multiple rule entries being created. The af type protocol mappings
patch broke this by apply only the first of the mappings that could be
found.
Restore the previous behavior by search through the entire table until
all matches have been made.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
add a add_to_options() helper function to aa.py which
- adds newpath to options if it's not already there
- returns the updated options and the index of newpath
This removes duplicated code for CMD_GLOB and CMD_GLOBEXT in
ask_the_question()
It also adds duplicate prevention to CMD_NEW.
Acked-by: Kshitij Gupta <kgupta8592@gmail.com>
When reaching EOF while still in a profile (syntax-wise), there are two
possible reasons:
- missing "}"
- missing "," in the last rule (which means that, thanks to multiline
rule handling, the "}" is considered to be part of the last rule)
This patch improves the error message in aa.py to cover a missing ","
Acked-by: Kshitij Gupta <kgupta8592@gmail.com>.
but should look for '//null-' instead.
Also remove some code duplication by merging with the next condition,
which executes the same self.add_to_tree code.
Acked-by: Kshitij Gupta <kgupta8592@gmail.com>.
File "/home/cb/apparmor/HEAD-CLEAN/utils/apparmor/aa.py", line 126, in check_for_LD_XXX
for line in f_in:
[...]
UnicodeDecodeError: 'ascii' codec can't decode byte 0xf8 in position 40: ordinal not in range(128)
While on it, also replace usage of the "found" variable by "return"
statements, which should bring a small performance improvement - if we
have a match, it's superfluous to continue searching.
The patch also adds me to the copyright header ;-)
Acked-by: Kshitij Gupta <kgupta8592@gmail.com>.
they only fail because of one (expected) reason and we notice if they
don't fail anymore. Complex profiles have the risk to fail for multiple
reasons, which also means nobody will notice if they fail for one reason
less.
The simplification is done by
- removing #include lines
- in some cases, replace the #include line with "/foo/bar r," to avoid
empty hats
Acked-by: Steve Beattie <steve@nxnw.org>
Kernel ABI v6 only required 'w' permissions for the parent process that
creates the socket, accepts a connection, writes to the socket, and
reads from the socket.
Kernel ABI v7 will require 'rw' permissions for the parent process. This
change detects the current kernel ABI version and adjusts the parent
process's confinement appropriately. It also performs a negative test to
make sure that 'w' is not sufficient.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This change only sets up unix_socket.sh to test abstract sockets.
Unconfined processes are tested while using an abstract socket but
the test function returns before testing with confinement.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Rename the test in preparation for expanding its capabilities to cover
all UNIX domain socket address format types.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
openSUSE now compiles samba --with-cachedir=/var/lib/samba (instead of
the default /var/cache/samba). This patch updates the smbd profile to
match this change.
Acked by: Seth Arnold <seth.arnold@canonical.com>
user/password files (everybody will use a different filename for the
user/password list - and when you allow reading the password list,
allowing to read the config doesn't add any harm ;-)
References: https://bugzilla.novell.com/show_bug.cgi?id=874094
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Subject: perl-apparmor: Properly handle bare 'file' keyword
References: bnc#889652
The bare file keyword is a shortcut for /{**,}. There are also implied
permissions that go with it.
This patch accepts the file keyword as well as allowing for missing mode
specifiers.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Subject: perl-apparmor: Handle bare 'capability' keyword
References: bnc#889651
Specifying 'capability' implies all capabilities, but the perl code didn't
recognize it.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Subject: perl-apparmor: Fix bare 'network' keyword handling
References: bnc#889650
The 'network' bare keyword was being printed as "audit network all" due to
two different bugs:
1) {audit}{all} was always being set to 1, regardless of whether the audit
keyword was used
2) {rule} eq 'all' is the wrong test - it should be {rule}{all}
With these fixed, 'network' is properly handled.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Acked-by: Steve Beattie <steve@nxnw.org>
- allow users to merge two profiles (2-way merge) using aa-mergeprof by
making the third profile optional
- re-enable code that cleaned up base and other profile and using it in
deleted count (was disabled due to pyflakes thinking it was unused)
Patch by Kshitij Gupta <kgupta8592@gmail.com>
Acked-by: Christian Boltz <apparmor@cboltz.de>
This patch adjusts the verbosity of several of the utils tests,
to make them all consistently verbose.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
tree python modules.
Also remove "sys.path.append('../')" (and now-unused "import sys") from
all tests that contain it for consistency and to make testing with the
installed modules possible (even if we don't have a USE_SYSTEM option
yet).
Acked-by: Steve Beattie <steve@nxnw.org>
saving the remaining profile after using "save selected profile".
When using "save selected profile", write_profile() (called by
write_profile_ui_feedback()) does "changed.pop(profile_name)".
However, saving the remaining profiles in save_profiles() uses a cached
version of changed.keys() and therefore tries to save a profile that
was already saved and is no longer part of "changed".
Also remove two commented lines we'll never need again:
- #changed.pop is done in write_profile()
- q['options'] is set some lines above
References: https://bugs.launchpad.net/apparmor/+bug/1341178
Acked-by: Steve Beattie <steve@nxnw.org>
This patch adds the new capability CAP_AUDIT_READ, added in the 3.16
kernel, to the utils severity.db. I set the severity level at 7,
since it seemed to offer less exposure than CAP_AUDIT_CONTROL and
CAP_AUDIT_WRITE, which are both considered severity 8.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
testcases) to
- sleep 10 seconds after each failed test to make failures more annoying
^W^W^W^Wgive people a chance to read failure details
- print a list of failed tests at the end
Also avoid duplicate code by letting runtests-py2.sh call runtests-py3.sh.
Acked-by: Steve Beattie <steve@nxnw.org>
LOG_MODE_RE is also defined (and used) in aamode.py.
This patch removes the superfluous definition from logparser.py.
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch fixes some memory leaks in the libapparmor log parsing
functions, specifically around handling records obtained from syslog
and records containing network addresses.
Bug: https://bugs.launchpad.net/bugs/1340927
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch is cosmetic; it cleans up a lot of whitespace issues:
removing trailing spaces, converting tabs into spaces, and removing
unneeded spaces around function arguments.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
This patch is a cosmetic set of changes to remove references to immunix
from the source code (except in the case of handling deprecated
keywords), as well as correcting my email address.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
This patch adds an additional hat to try in the mod_apparmor processing
sequence, constructed from the host's ServerName + '-' + URI
(e.g. 'www.example.com-/some/uri'). This hat is attempted before the raw
URI hat is attempted, leaving the ordering as follows:
(1) to a hatname in a location/directory directive
(2) to the server name or a defined per-server default
(3) to the server name + "-" + uri
(4) to the uri
(5) to DEFAULT_URI
(6) back to the parent profile
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
In trunk revno 2335, a bug was fixed in mod_apparmor that corrected
the storage location for AADefaultHatName. The incorrect storage
caused the hat specified by the AADefaultHatName keyword to be the
default value for AAHatName, and meant that if both an AAHatName and
an AADefaultHatName entry were given in a vhost, mod_apparmor would
not fall back to trying AADefaultHatName if the hat specified in
AAHatName did not exist in the apache apparmor profile.
However, because the value specified in AADefaultHatName was the
default, if no AAHatName was specified, it would be attempted first,
before a hat based on the passed URI, rather than after as the
documentation stated and the code intended. By fixing the storage bug,
the attempted hat ordering now matched the documentation. But a number
of users came to rely on AADefaultHatName being attempted before
the URI. For trunk, this issue is less severe because mod_apparmor
passes a vector of hats to aa_change_hatv(), and thus missing URI
hats are not logged by the kernel apparmor bits. It still represents
a behavioral change to users, though.
This patch re-adjusts the ordering so that the URI-based hat is
attempted after the hat specified by AADefaultHatName is attempted,
thus maintaining the actual behavior before the bug addressed in
revno 2335 was fixed.
Patch history:
v1: initial revision
v2: no code changes; adjust comments and improve the man page
documentation
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
Some updates for the dovecot profiles, based on a patch from
Christian Wittmer <chris@computersalat.de> (he sent it as SR for the
openSUSE package, which uses a slightly older version of the dovecot
profiles)
Fix problems with dovecot and managesieve:
* usr.lib.dovecot.managesieve-login: network inet6 stream
* usr.lib.dovecot.managesieve:
+#include <tunables/dovecot>
/usr/lib/dovecot/managesieve {
+ capability setgid, # covered by abstractions/dovecot-common, therefore not part of this patch
+ capability setuid,
+ network inet stream,
+ network inet6 stream,
+ @{DOVECOT_MAILSTORE}/ rw,
+ @{DOVECOT_MAILSTORE}/** rwkl,
* add #include <abstractions/wutmp> to usr.lib.dovecot.auth
apparmor="DENIED" operation="open" parent=18310 \
profile="/usr/lib/dovecot/auth" name="/var/run/utmp" pid=20939 \
comm="auth" requested_mask="r" denied_mask="r" fsuid=0 ouid=0
Acked-by: Steve Beattie <steve@nxnw.org>
Bug: https://launchpad.net/bugs/1322778
When configured with "clustering = yes", Samba needs to be able to
connect to the local ctdbd daemon socket, and directly manipulate .tdb
database files managed by ctdb.
Signed-off-by: David Disseldorp <ddiss@suse.de>
This commit adds a dovecot-common abstraction, as well as adjusting
the profiles for dovecot's helper binaries to make use of it. The
important addition is the ability for the dovecot master process to
send signals to the helpers.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Move postfix-common from program-chunks/ to abstractions/; remove
program-chunks directory since postfix-common was the last resident of
that directory (and had been since 2007), and adjust the includes of all
the profiles that include postfix-common.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
- Allow reciprocal ptrace readby to everyone (requires peer unconfined or to
ptrace read to us)
- same for ptrace tracedby
- allow us to ptrace read ourselves
- receive all signals from unconfined
- allow us to signal ourselves
- allow sending and receiving "exists" (for pid existence)
Acked-By: Jamie Strandboge <jamie@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
This patch improves the error messages in aa.py store_list_var() to make
debugging of profile syntax problems easier. It also adds an additional
parameter for the profile filename (used in the error message)
Acked-by: Steve Beattie <steve@nxnw.org>
Earlier fixes to the parser's handling of escape sequences involving '\'
caused a behavioral change that profiles no longer needed to contain
'\\' before an octal escape sequence. However, the regression tests were
never modified to take this change into account, and thus the i18n.sh
octal tests would fail. This patch fixes that.
Also, with the changes, the parser no longer accepts _\_ as a valid
sequence, so we skip this character.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com> (on IRC)
The change to processing escape sequences in trunk commit r2537 requires
a corresponding change to the unit tests in parser_misc.c.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
escape sequences that result in special character that will be interpreted
by later processing need to be passed through as well.
Eg. previously \\ was fixed to be passed through, but other chars
get interpretted as well.
*?[]{}
and ^, in character classes
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Also for characters that are not recognized as a valid escape seq
make sure that the character is emitted.
previously
\$ resulted in \
where it should have been \$ if $ wasn't recognized
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
r2456 unified escape sequence processing but it results in the \\
sequence being processed multiple times (lexer, regex conversion,
backend pcre parsing).
What used to happen was the lexer would only convert octal sequences
and a few special escapes, \\ would be passed through the lexer and
the regex conversion, thus only being handled in the pcre backend.
r2456 changed that so that \\ is handled by the lexer, converting it
to \, which is handled as an escape sequence in both the regex
conversion and the pcre backend.
This means
\\001 instead of being treated as the literal \001 is treated
as an octal escape sequence which is rejected by the regex conversion
(it only allows for certain special chars).
etc.
Fix this by ensuring the lexer does not processes \\ and passes it
through so it is only handled in the backend as was done in the past.
Also fix front end escape sequence processing of octals etc from resulting
in a later escape sequence. That is \134, \d92, .. would get converted
to \ in the lexer and then treated as an escape sequence in the regex
conversion or pcre processing.
We fix this by converting them to the equivalent \\ sequence in the
lexer and letting the backend processes it.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
This patch fixes a crash in aa-complain when a profile name is quoted.
It also makes sure aa-complain actually adds the complain flag in such
cases. (aa-enforce etc. will also benefit from this fix.)
Note: superfluous quotes will be removed when saving the profile (for
example with aa-cleanprof), but they are kept if needed, like in
profile "/bin/foo bar"
(tested with aa-complain and aa-cleanprof - and also with "rcapparmor
reload", where the initscript bailed out because my profile filename
contained a space...)
The patch also adds some TODO notes.
References: https://bugs.launchpad.net/apparmor/+bug/1296218
Acked-by: Steve Beattie <steve@nxnw.org>.
For some strange reason our caching use ctime instead of mtime.
However this can lead to odd cases of the cache missing even though
neither the profile data nor cache data have changed.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Update the apache2 profile so that the parent apache process can kill
worker processes inside of hats. Update the example comments and the
DEFAULT_URI and HANDLING_UNTRUSTED_INPUT hats to include the
apache2-common abstraction to allow them to receive signals from the
parent process.
Author: Kees Cook <kees@ubuntu.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Bug: https://bugs.launchpad.net/apparmor/+bug/1322764
Update the apache2-common abstraction so that the parent apache process
can kill worker processes inside of hats, as well as handle the updated
mod_apparmor behavior that invokes aa_change_hatv() and then checks
which hat it ended up in via aa_getconn() (which reads from
{PROC}/@{pid}/attr/current).
Author: Kees Cook <kees@ubuntu.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Bug: https://bugs.launchpad.net/apparmor/+bug/1322764
The child process changes into a hat while the parent process stays in
the main profile.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Add two tests that verify AppArmor denials when one end of the pipe has
bad access permissions to the pipe.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
The named_pipe parent process kills the child process at exit. A
"signal," rule must be added to all confinement profiles when the test
is running under a kernel that performs signal mediation.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Allow for the parent and child processes to change into separate hats to
verify named pipe communications between hats with varying permissions.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Add debugging info to test binaries and disable optimizations.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
When creating a child profile while using genprof, I get a backtrace:
Traceback (most recent call last):
File "aa-genprof", line 160, in <module>
lp_ret = apparmor.do_logprof_pass(logmark, passno)
File "/home/cb/apparmor/HEAD-CLEAN/utils/apparmor/aa.py", line 2291, in do_logprof_pass
save_profiles()
File "/home/cb/apparmor/HEAD-CLEAN/utils/apparmor/aa.py", line 2309, in save_profiles
for prof_name in changed.keys():
RuntimeError: dictionary changed size during iteration
(See https://bugs.launchpad.net/apparmor/+bug/1014304 for more details.)
After digging into the code, it seems for some reason the child profile
is added to "changed" - I doubt this is correct (guess why it's removed
later... ;-)
After digging a bit more, I found out that create_new_profile() is
(ab)used to create a new stub profile to be used as child profile.
create_new_profile then adds the new child (which looks like a normal
profile to it) to "changed".
This patch most probably makes the cleanup round in save_profile()
superfluous by adding a is_stub parameter to create_new_profile(). If
this parameter is set, the new (child) profile is not added to "created"
and "changed".
I intentionally added the two print() lines in safe_profile because
a) I think they will never be displayed
b) I want to know if a) is wrong ;-)
c) it's always nice to have a "nice" error message before displaying
a backtrace ;-)
Acked-by: Steve Beattie <steve@nxnw.org>
(unlimited) because the "if not value:" check matches 0.
This patch replaces the check with "... is None".
It also prints a warning if the old value is None (could in theory
happen if reading the old value failed).
Acked-by: Steve Beattie <steve@nxnw.org>. Thanks.
preprocessor and is not as thorough as -QTK (--skip-kernel-load,
--skip-read-cache, --skip-cache). Like with '-p', '-QTK' can be run without
privilege but it will catch things like conflictings 'x' modifiers.
Acked-By: Jamie Strandboge <jamie@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Commit r2456 fixes a bug in the parsers compilation that can result
policy failures. Unfortunately this Bug slipped into the wild and
shipped in at least one distro.
Bump the parser abi so that parsers that have the fix will invalid
existing cache files, and recompile policy to ensure the fix is applied.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Bug: https://bugs.launchpad.net/bugs/1235478
This is a test to check the label on file descriptors returned from
socketpair().
In its simple form, it simply calls socketpair() and checks the
labels on both fds.
In its complex form, it has the ability to do the simple test, then set
up an exec transition using aa_change_onexec(), and re-exec itself to
check the labeling after the file descriptors have been passed across an
exec transition.
The complex form is meant to test revalidation at exec. AppArmor
currently keeps the original labeling in place across the exec
transition.
Note that this test does not currently test read/write access to the
file descriptors. It only checks the label, as returned by
aa_getpeercon(2).
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
Allow for the regression tests to specify arbitrary profile names
without hitting fatal errors or getting warnings from mkprofile.pl.
This allows for a test to have a line like this:
genprofile change_profile->':arbitrary_name -- \
image=arbitrary_name addimage:$test
In the example above, $test can call aa_change_onexec("arbitrary_name")
and then re-exec itself to test behavior across exec transitions.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
Ubuntu 14.04's chromium-browser has changed paths in a way that prevents
evince from opening clicked links in chromium-browser windows.
This patch adds a new path for the chrome-sandbox executable to the
sanitized_helper profile, so chromium will get its own tailored profile if
necessary.
The reporter who said this patch helped included some further DENIED lines
for signals that indicates this is probably not sufficient but did make
the links work as expected.
https://bugs.launchpad.net/ubuntu/+source/apparmor/+bug/1282314
Signed-off-by: Seth Arnold <seth.arnold@canonical.com>
Acked-By: Jamie Strandboge <jamie@canonical.com>
- convert "tail" result from byte to string to avoid TypeError crash
- use apparmor.filename instead of hardcoded /var/log/audit/audit.log
Acked-by: Kshitij Gupta <kgupta8592@gmail.com>
v3: fix freeing of filename when undefined
v2: address tyhicks feedback
refactor to have a common write routine
fix issue with set profile load being done even if !kernel_load
Profile loads from cache files that contain multiple profiles can
result in multiple reloads of the same profile or error messages about
failure to load profiles if the --add option is used. eg.
apparmor="STATUS" operation="profile_load"
name="/usr/lib/apache2/mpm-prefork/apache2" pid=8631
comm="apparmor_parser"
<sth0R> [82932.058388] type=1400 audit(1395415826.937:616):
apparmor="STATUS" operation="profile_load" name="DEFAULT_URI" pid=8631
comm="apparmor_parser"
<sth0R> [82932.058391] type=1400 audit(1395415826.937:617):
apparmor="STATUS" operation="profile_load"
name="HANDLING_UNTRUSTED_INPUT" pid=8631 comm="apparmor_parser"
<sth0R> [82932.058394] type=1400 audit(1395415826.937:618):
apparmor="STATUS" operation="profile_load" name="phpsysinfo" pid=8631
comm="apparmor_parser"
<sth0R> [82932.059058] type=1400 audit(1395415826.937:619):
apparmor="STATUS" operation="profile_replace" info="profile can not be
replaced" error=-17
name="/usr/lib/apache2/mpm-prefork/apache2//DEFAULT_URI" pid=8631
comm="apparmor_parser"
<sth0R> [82932.059574] type=1400 audit(1395415826.937:620):
apparmor="STATUS" operation="profile_replace" info="profile can not be
replaced" error=-17
name="/usr/lib/apache2/mpm-prefork/apache2//HANDLING_UNTRUSTED_INPUT"
pid=8631 comm="apparmor_parser"
The reason this happens is that the cache file is a container that
can contain multiple profiles in sequential order
profile1
profile2
profile3
The parser loads the entire cache file to memory and the writes the
whole file to the kernel interface. It then skips foward in the file
to the next profile and reloads the file from that profile into
the kernel.
eg. First load
profile1
profile2
profile3
advance to profile2, do second load
profile2
profile3
advance to profile3, do third load
profile3
With older kernels the interface would stop after the first profile and
return that it had processed the whole file, thus while wasting compute
resources copying extra data no errors occurred. However newer kernels
now support atomic loading of multipe profiles, so that all the profiles
passed in to the interface get processed.
This means on newer kernels the current parser load behavior results
in multiple loads/replacements when a cache file contains more than
one profile (note: loads from a compile do not have this problem).
To fix this, detect if the kernel supports atomic set loads, and load
the cache file once. If it doesn't only load one profile section
from a cache file at a time.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Mention, in the apparmor.d man page, that pivot_root arguments must end
with a '/' character since they are directories.
The parser currently allows pivot_root arguments that do not end in '/',
but those rules will always fail to match.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
It may not be obvious that the peer label can be "unconfined". Provide
an example rule, in the apparmor.d man page, demonstrating the
peer=(label=unconfined) conditional.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
(in a more serious version: add some tests for dbus, *mount, signal,
ptrace and pivot_root and make sure a space after the keyword is enforced.
The tools shouldn't accept a "dbusdriver" or "pivot_rootbeer" rule. ;-)
Acked-by: Tyler Hicks <tyhicks@canonical.com>
(dbus, *mount, signal, ptrace, pivot_root) except if the line only
contains the bare keyword.
Note that in most cases (except *mount) I used an alternation - this has
the advantage that it doesn't change the match group numbering, with the
small disadvantage of having to mention the keyword twice in the regex.
I chose this way to avoid that I have to change lots of other places and
possibly introduce bugs by overlooking something.
For the *mount rules, I read the code - it shouldn't need any changes
because it uses only matches[0..2]
Acked-by: Tyler Hicks <tyhicks@canonical.com>
This patch extends the coverage of the parser's simple dbus language
tests.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch adds basic signal tests to the parser's simple language
test suite.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
The mount.sh script mixes calls to the regression test 'mount' binary
and /sbin/mount. This can result in stale mtab entries being left around
after a test run because /sbin/mount adds an mtab entry but the test
'mount' binary, which is also used for unmounting, does not remove mtab
entries.
To solve this problem, the -n option is passed to /sbin/mount so that it
doesn't add an mtab entry when mounting.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
The end of the mount.sh regression test script contained cleanup
commands to unmount and detach the loop device used for testing.
However, the second losetup command fails and, with the recent
regression test suite fix to not ignore failed shell commands, an error
is triggered at the end of the test run.
Additionally, these cleanup commands are not ran when the test fails
during the test run and an immediate exit is requested upon failure
(with the -r flag).
This patch fixes and moves the cleanup logic into a function that is
assigned to do_onexit so that the cleanup is always performed at exit
and the test can run successfully.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
When there was a test error, such as a shell command failure, the
function used for the ERR trap, error_handler(), was causing the error
to be silently ignored by the test runner.
It was calling exit_handler() directly, before calling fatalerror().
This caused $_fatal to be left unset when exit_handler() was called.
exit_handler() sources epilogue.inc and the last bit of epilogue.inc
exits with $num_testfailures if $_fatal was unset. The fatalerror() call
site in error_hanlder() was never reached. So, as long as there were no
test failures, then an error in a test script would cause the test to
exit early with 0.
It is safe to simply call fatalerror() from error_handler() because
fatalerror() sets $_fatal to true and exits. This causes exit_handler()
to be called and since $_fatal is set to true, prologue.inc exits with
127.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
With the recent addition of features like ptrace and signals that
give warnings and then ignore the subset of rules when the features
directory indicates that the kernel does not support mediating such
features, at least one of the language tests fails in a chroot
environment where the apparmor securityfs tree is not mounted
inside it.
To compensate, a features file containing the current supported features
is included, and the simple.pl test driver is modified to pass it as an
argument to the parser, so that it will act as if the environment
supports all our current features.
A simple python script is included that was used to generate the
features file based on the current feature set.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
table
This patch adds the creation of an implicit set variable
@{profile_name} for use within policy. It expands to:
- a given profile name if specified; e.g. for
'profile flappy_bird /some/pattern/match* { [...] }'
@{profile_name} would expand to 'flappy_bird'
- if no given name, the match pattern; e.g. for
'/usr/bin/doge_bird { [...] }'
@{profile_name} would expand to '/usr/bin/doge_bird'
- hats and child profiles will include the fully qualified name; e.g.
the 'doge' hat in the /usr/bin/flappy_bird profile would cause
@{profile_name} to expand to '/usr/bin/flappy_bird//doge' within the
'doge' hat, and '/usr/bin/flappy_bird' outside of it in the profile.
There are some parsing tests added, but more tests are needed to verify
that expansion occurs properly (I've verified manually using parser
dumps of the added tests, but automated checks are needed).
The @{profile_name} variable is expected to be most useful in the
context of signal and ptrace rules (e.g. for specifying that an app
can send itself signals).
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
seenlist bug
This patch:
- refactors the parser_symtab.c unit tests a bit in preparation for
the patch to add an implicit autofilled @{profile_name} variable
- expands coverage of the unit tests such that all code paths that
don't result in an exit() or are due to memory allocation errors are
exercised (this doesn't mean the tests are complete; the
__expand_variable() could use more tests for correctness).
- it fixes a bug where variables were not being removed from the
seenlist when a problem was detected in __expand_variable().
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-By: Seth Arnold <seth.arnold@canonical.com>
Remove duplicated test code by adding a simple way for regex test
classes to declare a regex to use and a list of tuples consisting of
(line, expected_result). The setup_regex_tests() method generates test
methods for each tuple in a classes list. The test methods are based on
the regex_test() method, which performs the regex search and compares
the results to the expected_result.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Christian Boltz <apparmor@cboltz.de>
The regexes for signal, ptrace, and pivot_root rules each had an extra
grouping around the terms 'signal', 'ptrace', and 'pivot_root'. Those
extra groupings can be safely removed.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Christian Boltz <apparmor@cboltz.de>
This patch backs out most of the changes from r2448 in favor of a better
approach.
The optional "file" keyword is handled under the pre-existing
RE_PROFILE_PATH_ENTRY regex and a new regex, RE_PROFILE_BARE_FILE_ENTRY,
is created for handling bare file rules.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Christian Boltz <apparmor@cboltz.de>
The first conditional around string being set is not needed. If string
is not set, the while loop will be skipped and mode will be returned.
The variable tmp was being overloaded by being the regex search result
and then being reassigned to be the first match group in the regex
search result. This patch keeps tmp as the regex search result and then
uses mode_char to represent the first match group of the search.
Group the search and replace actions together at the beginning of the
loop and group the mode character processing at the end of the loop.
Finally, remove the unnecessary check of tmp (now mode_char) before
calling MODE_HASH.get(tmp, False). If tmp is None or '', get() will
do the right thing and return False.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Christian Boltz <apparmor@cboltz.de>
Bug: https://bugs.launchpad.net/bugs/1307665
When str_to_mode() was given a string containing unknown mode
characters, it entered an infinite loop. The case of the MODE_MAP_RE
regex string not matching the mode string was being ignored.
This patch makes it so that the loop breaks when MODE_MAP_RE no longer
matches the mode string. This occurs when all of the valid mode
characters have been processed and only invalid mode characters remain.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Christian Boltz <apparmor@cboltz.de>
Bug: https://bugs.launchpad.net/bugs/1298678
This patch does bare bones parsing of pivot_root rules and stores the raw
strings for writing them out later. It is meant to be a simple change to
prevent aa.py from emitting a traceback when encountering pivot_root rules.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Acked-By: Christian Boltz <apparmor@cboltz.de>
Bug: https://bugs.launchpad.net/bugs/1300317
This patch does bare bones parsing of ptrace rules and stores the raw
strings for writing them out later. It is meant to be a simple change to
prevent aa.py from emitting a traceback when encountering ptrace rules.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Acked-By: Christian Boltz <apparmor@cboltz.de>
Bug: https://bugs.launchpad.net/bugs/1300316
This patch does bare bones parsing of signal rules and stores the raw
strings for writing them out later. It is meant to be a simple change to
prevent aa.py from emitting a traceback when encountering signal rules.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Acked-By: Christian Boltz <apparmor@cboltz.de>
This patch adds a bunch of language parsing tests for ptrace rules.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
This patch adds the newer rules class entries (e.g. ptrace, signals)
when dumping profiles (invoking the parser with the -dd argument).
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
The previous test patches where done with the hardcoded bypass for
unconfined.
This semantic was changed so that a confined app can now block unconfined
processes from tracing or sending signals to it.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
ptrace rules currently take the form of
ptrace [<ptrace_perms>] [<peer_profile_name>],
ptrace_perm := read|trace|readby|tracedby
ptrace_perms := ptrace_perm | '(' ptrace_perm+ ')'
After having used the cross check (permission needed in both profiles)
I am not sure it is correct for ptrace.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
The match
{VARIABLE_NAME}/{WS}*={WS}*\(
is too broad causing mount and dbus rules to fail for sets of values eg.
mount options=(ro bind)
Instead of doing a broad match, for now lets lock it down to just
peer=(...) being the only cond that can cause entry into CONDLISTID
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
The match
{VARIABLE_NAME}/{WS}*={WS}*\(
is too broad causing mount and dbus rules to fail for sets of values eg.
mount options=(ro bind)
Instead of doing a broad match, for now lets lock it down to just
peer=(...) being the only cond that can cause entry into CONDLISTID
Signed-off-by: John Johansen <john.johansen@canonical.com>
Add signal rules and make sure the parser encodes support for them
if the supported feature set reports supporting them.
The current format of the signal rule is
[audit] [deny] signal [<signal_perms>] [<signal_set>] <target_profile>,
signal_perm := 'send'|'receive'|'r'|'w'|'rw'
signal_perms := <signal_perm> | '(' <signal_perm> ([,]<signal_perm>)* ')'
signal := ("hup"|"int"|"quit"|"ill"|"trap"|"abrt"|"bus"|"fpe"|"kill"|
"usr1"|"segv"|"usr2"|"pipe"|"alrm"|"term"|"tkflt"|"chld"|
"cont"|"stop"|"stp"|"ttin"|"ttou"|"urg"|"xcpu"|"xfsz"|"vtalrm"|
"prof"|"winch"|"io"|"pwr"|"sys"|"emt"|"exists")
signal_set := set=<signal> | '(' <signal> ([,]<signal>)* ')'
it does not currently follow the peer=() format, and there is some question
as to whether it should or not. Input welcome.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
For some rules the output of apparmor_parser -p has a double comma
Eg.
ptrace (tracedby),
dbus (send,receive),
is output as
ptrace (tracedby),,
dbus (send,receive),,
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
apparmor_parser -p is broken. Outputting garbage charcters after every
include statement.
eg.
##included <tunables/multiarch>
^@^@V><A8>^?^@^@<C8>^NV><A8>^?^@^@<A0>^Pu^@# -----------------------------------
-------------------------------
#
This is happening because includes are handled specially and should not
go through the usual preprocessing output dump.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
The addition of the dbus tests requires dbus dev libraries be installed
to run the test suite. This is not always desirable or even possible.
So make building and running the dbus tests conditional on the
pkg-config info from those libs. If they are not present output a
message about skipping the tests.
This patch contains the review fix from sbeattie
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch makes use of the htoleXX() functions (see endian(3))
defined as part of endian.h (already included in parser_interface.c),
instead of defining a function differently based on the detection of
endian related macros.
This fixes a build failure experienced on powerpc with John's patch
set applied. This patch has been updated with John's feedback to use
letoh16() in the le16_to_cpu() macro.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
alone when the labeling is not path dependent.
Some rules will not generate label entries, some will generate only
label entries and some will generate both label and path entries.
This is left to the particular rule encoding.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Using the parser timestamp was a work around to force recompilation of
policy that was built with a buggy parser. There are better ways to
handle this so remove checking of the parser timestamp.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This will allow for the parser to invalidate its caches separate of whether
the kernel policy version has changed. This can be desirable if a parser
bug is discovered, a new version the parser is shipped and we need to
force cache files to be regenerated.
Policy current stores a 32 bit version number in the header binary policy.
For newer policy (> v5 kernel abi) split this number into 3 separate
fields policy_version, parser_abi, kernel_abi.
If binary policy with a split version number is loaded to an older
kernel it will be correctly rejected as unsupported as those kernels
will see it as a none v5 version. For kernels that only support v5
policy on the kernel abi version is written.
The rules for policy versioning should be
policy_version:
Set by text policy language version. Parsers that don't understand
a specified version may fail, or drop rules they are unaware of.
parser_abi_version:
gets bumped when a userspace bug is discovered that requires policy be
recompiled. The policy version could be reset for each new kernel version
but since the parser needs to support multiple kernel versions tracking
this is extra work and should be avoided.
kernel_abi_version:
gets bumped when semantic changes need to be applied. Eg unix domain
sockets being mediated at connect.
the kernel abi version does not encapsulate all supported features.
As kernels could have different sets of patches supplied. Basic feature
support is determined by the policy_mediates() encoding in the policydb.
As such comparing cache features to kernel features is still needed
to determine if cached policy is best matched to the kernel.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Tag start of entries in the policydb as being mediated. This makes
the start state for any class being mediated be none 0. The kernel
can detect this to determine whether the parser expected mediation
for the class.
This is just a way of encoding what features expect mediation within
the policydb it self so that a separate table isn't needed.
This is also used to indicate the new unix semantics for mediation of
unix domain sockets on connect should be applied.
Note: this does cause a fail open on situation on Ubuntu Saucy, which
did not properly indicate support. That is if a kernel using this patch
is installed on an Ubuntu Saucy system, unix domain socket mediation
on connect won't happen, instead the older behavior will be applied.
This won't cause policy failures as it is less strict than what
Ubuntu Saucy applies.
This is necessary so that AppArmor can properly function on older
userspaces without a compile time configuration on the kernel to determine
behavior. A kernel expecting this behavior will function correctly
with all old userspaces expect it will not enforce connect time mediation
on Ubuntu Saucy. However Ubuntu does not support Trusty (or newer)
kernels as backports to Saucy, so this does not break them.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This cleans things up a bit and fixes a bug where not all rules are
getting properly counted so that the addition of policy_mediation
rules fails to generate the policy dfa in some cases.
Because the policy dfa is being generated correctly now we need to
fix some tests to use the new -M flag to specify the expected features
set of the test.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This is not the cleanup this code needs, but a quick hack to add the
-M flag so we can specify a feature file (or directory) to use for
the compile.
It mostly just moves around existing code and adds the -M option,
though it does introduce a few changes.
While I didn't do it in this patch I propose we drop support for
the match file without create support. This is several years old
now and would clean things up a lot.
Note: that the manually input -m or -M drop support for it already
I just can't see a good way to support a single input stream indicating
the result/existance of two separate files.
This needs more work but is needed to support tests and the policy_mediates
frame work depends on the policydb getting generated with the special
stub rules to indicate whether policy was compiled expecting a certain
feature. But this can break the current tests, at least once a bug
in the policy rule counting is fixed in a follow on patch.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Another issue with commit 2456 is that with older versions of glibc and
g++, a definition for SIZE_MAX was not being found; e.g. on Ubuntu 12.04
LTS and 12.10, the parser fails to compile with the following error:
g++ -g -O2 -pipe -Wall -Wsign-compare -Wmissing-field-initializers -Wformat-security -Wunused-parameter -std=gnu++0x -D_GNU_SOURCE -DPACKAGE=\"apparmor-parser\" -DLOCALEDIR=\"/usr/share/locale\" -DSUBDOMAIN_CONFDIR=\"/etc/apparmor\" -I../libraries/libapparmor//include -c -o lib.o lib.c
lib.c: In function 'int str_escseq(const char**, const char*)':
lib.c:292:32: error: 'SIZE_MAX' was not declared in this scope
The following patch addresses the issue by explicitly including the C stdint
header which contains the definition for SIZE_MAX.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
Trunk commit 2456 broke the builds on i386 with the following compiler
error:
g++ -g -O2 -pipe -Wall -Wsign-compare -Wmissing-field-initializers -Wformat-security -Wunused-parameter -std=gnu++0x -D_GNU_SOURCE -DPACKAGE=\"apparmor-parser\" -DLOCALEDIR=\"/usr/share/locale\" -DSUBDOMAIN_CONFDIR=\"/etc/apparmor\" -I../libraries/libapparmor//include -c -o lib.o lib.c
lib.c: In function 'int strn_escseq(const char**, const char*, size_t)':
lib.c:236:47: error: no matching function for call to 'min(long unsigned int, size_t&)'
tmp = strntol(*pos, &end, 8, 255, min(3ul, n));
^
This is due to size_t differing in size on i386 and amd64. The
following patch addresses the issue by casting the constant values
to size_t (and removing the ul suffix since the constant values are
getting cast anyway), satisfying C++'s types (and the patch removes
the unnecessary min macro).
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Policy enforcement needs to be able to support older userspaces and
compilers that don't know about new features. The absence of a feature
in the policydb indicates that feature mediation is not present for
it.
We add stub rules, that provide a none 0 start state for features that
are supported at compile time. This can be used by the kernel to
indicate that it should enforce a given feature. This does not indicate
the feature is allowed, in an abscence of other rules for the feature
the feature will be denied.
Note: this will break the minimize tests when run with kernels that
support mount or dbus rules. A patch to specify these features to
the parser is needed to fix this.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Fix the octal escape sequence that was broken, so that short escapes \0,
\00 \xa, didn't work and actually resulted in some encoding bugs.
Also we were missing support for the decimal # conversion \d123
Incorporate and update Steve Beattie's unit tests of escape sequences
patch
v2
- unify escape sequence processing, creating lib fns.
- address Steve Beattie's feedback
- incorporate Steve Beattie's feedback
v3
- address Seth's feedback
- add missing strn_escseq tests
- expand strn_escseq to take a 3rd parameter to allow specifying chars to
convert straight across. . eg "+" will cause it to convert \+ as +
- fix libapparmor/parse.y failed escape pass through to match processunqoted
Unit tests by Steve Beattie
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch separates pivot_root rules from mount rules, since the syntax
of the two types of rules is very different. It also documents the
missing "oldroot=" prefix required for the conditional corresponding to
the put_old parameter. Finally, it briefly describes pivot_root rules
and provides some examples.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This test attempts to clone itself in a new mount namespace, pivot root
into a new filesystem (ext2 disk image mounted over loopback), and then
verify that a profile transition, if one was specified in the pivot_root
rule, has properly occurred.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Yes its seems pointless because these will eventually get replaced by
stl. But until then
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Valgrind is offering complaints like the following when dealing with
profiles with mount rules:
==27919== Conditional jump or move depends on uninitialised value(s)
==27919== at 0x805CDC1: mnt_rule::mnt_rule(cond_entry*, char*, cond_entry*
==27919== by 0x805674E: do_mnt_rule(cond_entry*, char*, cond_entry*, char*
==27919== by 0x8057937: yyparse() (parser_yacc.y:1133)
==27919== by 0x8053916: process_profile(int, char const*) (parser_main.c:1
==27919== by 0x804B20E: main (parser_main.c:1340)
Doing this consistently with the other initializers for the mount
class instead:
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
This will simplify add new features as most of the code can reside in
its own class. There are still things to improve but its a start.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Bug: https://bugs.launchpad.net/bugs/1295346
Add the ability to read and write path rules containing the file prefix.
This also includes bare "file," rules.
The ALL global is updated to include a preceding NUL char to eliminate
possibilities of a real file path colliding with the ALL global.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
The mount.sh regression test script was not testing with actual AppArmor
mount rules. This patch improves mkprofile.pl by adding the ability to
generate mount rules and adds tests to mount.sh that verify mount
mediation is working properly.
Signed-off-by: John Johansen <john.johansen@canonical.com>
[tyhicks: Fixed a couple typos and added fstype tests]
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
This updates the regression tests for v6 policy. It refactors the
required_features test into a have_features fn, and a new
requires_features fn (renamed to catch all instances make sure they
where right)
The have_features fn is then applied to several test to make them
conditionally apply based off of availability of the feature
and policy version.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
With the conversion of the python utils, aa-easyprof got added to the
list of tools to be installed (in /usr/sbin/), but is already installed
(in /usr/bin) by the python-tools-setup.py distutils script, leaving two
copies of the tool in place. This patch filters out aa-easyprof from the
list of tools for the makefile to install itself, leaving it to
(continue to) be installed by the distutils script.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
The valgrind test script would happily chug along even if if valgrind
was not installed, not doing anything of use. This patch fixes that, and
offers up the ability to specify an alternate location for valgrind if
it does not exist in the usual /usr/bin location.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
\s is a new feature of GNU grep 2.6 (released on 2010-03-23) and
it does not work in older versions. By using [[:space:]] instead,
AppArmor can compile on systems with older versions of grep.
Signed-off-by: Alban Crequy <alban.crequy@collabora.co.uk>
Acked-by: Steve Beattie <steve@nxnw.org>
This branch removes the deprecated/management subdirectory as the code
there has languished there for four years without interest. It then
drops excluding the deprecated/ tree from the toplevel tarball creation,
while adding a mechanism for adding back in exclusions to tar.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
After removing the tools that had lingered in the deprecated directory
for too long, don't exclude the deprecated/ subdirectory from tarball
creation, as SUSE needs access to the deprecated perl modules for YaST.
Add a make variable for adding back in exclusions if needed.
profile editor. They've been under the deprecated tree since Feb 2010,
and were placed there because they were already problematic to support.
No one has taken the mantle to resurrect support after 4 years, so
remove them from the tree entirely. (They will live on in the history,
if anyone does decide to resurrect them.)
This patch adds some simple tests of the capability regex in
apparmor/aa.py.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
Bug: https://bugs.launchpad.net/bugs/1294819
This patch as minimal support for bare capability rules ("capability,").
It prevents aa.py from emitting a traceback when encountering such a
rule.
It only adds the ability to parse and write the bare rule. It doesn't
attempt to be clever when deleting duplicate rules, such as realizing
that "capability audit_control," can be deleted if "capability," is also
present.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
Bug: https://bugs.launchpad.net/bugs/1294825
This patch is inspired by sbeattie's patch to add limited dbus rule
support. It adds does very dumb parsing of mount rules. Basically, it
stores mount, remount, and umount rules as raw strings wrapped in a
class.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
On ppc64el platforms, the minimum swapfile size is 640KiB. Our swap
test aborts there because it creates a swapfile of size 512KiB. This
patch adjusts the size to 768KiB, to satisfy ppc64el and to try
to keep the size down for embedded and otherwise limited platforms
(e.g. phones).
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
Some kernels have CONFIG_SYSCALL_SYSCTL disabled, which is something to
be encouraged. This patch separates out the two different kind of sysctl
tests (syscall based and /proc/sys based) into separate shell functions,
and then checks to see that the test environment supports each before
invoking each shell function, issuing a warning (but not failing the
tests) if not available.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
Bug: https://bugs.launchpad.net/bugs/1294848
This patch fixes the testsuite for aa-easyprof when the easyprof
utilities and configuration files are not installed in the system.
What was happening was that verify_manifest was calling parse_args()
without the synthetic arguments created by the test case and passing
the result to AppArmorEasyProfile(). Because parse_args() didn't
have the synthetic arguments, it would parse the actual command line
arguments passed to the testscript, which of course didn't specify the
alternate configuration file location. This would work when easyprof
had been installed in the system, because the fallback configuration
file in /etc/apparmor/easyprof.conf would exist and specify template
and policy group locations. Without that, though, the tests would abort
due to not knowing the location of the templates and policy groups.
This patch fixes the issue by passing the synthetic argument list
to verify_manifest, which uses that when calling parse_args(). A
debugging statement that states which conffile is being used when
AppArmorEasyProfile is being instantiated.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
last_audit_entry_time() was waiting forever because
subprocess.check_output() started tail without any parameters.
Fixed by removing shell=True (default is shell=False).
Also fix the regex ("^.*", the dot was missing)
Acked-by: Steve Beattie <steve@nxnw.org>
The regression swap test attempts to activate a swap file in a
directory under where tmpdir is set in uservars.inc; if this is a
tmpfs filesystem, this will fail (it's kind of silly to create a
swap file on a tmpfs, a memory-backed filesystem). This patch adds a
check to the swap test script and skips the tests if it detects it's
on tmpfs and marks the test as a failure if the check fails.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
- added beginning of utils translations for Polish and Swedish
- Some rejiggering of existing utils translations; I don't think any
existing translations got lost, but there are new missing entries
- A whole bunch of comment updating for the parser translations
Acked-by: Steve Beattie <steve@nxnw.org>
Current libapparmor python bindings are very "unpythonic". Also lack
ability to access "why" information in case of failure.
In python when something fail the normal behaviour is exception
to occur. In case of apparmor functions die silently and require
user to verify returned value.
And here comes second problem. In C api when return value is -1
(and the same value is returned in python API) we can access errno
to get information why this occured. Unfortunately in python there
is no way to access the same information. Pythonic way of accessing
errno is via exception (which is never raised in python bindings
currently).
So the patch adds exceptions on failures. First %exception creates
a wrapper that swig adds to each function listed below. Empty %exception
causes that the rest of code (beside listed functions) won't be wrapped.
How this works? Example on apparmor disabled system:
Before:
>>> LibAppArmor.aa_change_hat(hat, random.randint(1, sys.maxint))
-1
After:
>>> LibAppArmor.aa_change_hat(hat, random.randint(1, sys.maxint))
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
OSError: [Errno 22] Invalid argument
so pythonic way of accessing "why":
>>> try:
... LibAppArmor.aa_change_hat(hat, random.randint(1, sys.maxint))
... except OSError, e:
... print e.errno
...
22
Signed-off-by: Arkadiusz Miśkiewicz <arekm@maven.pl>
Acked-by: Steve Beattie <steve@nxnw.org>
D-Bus rules in particular seem to get written as multi-line rules. This
patch adds very simple hackish support for multiple lines. Essentially,
what it does is if the parsing of a line doesn't match anything and
falls all the way through, it saves the line and prepends it to the next
line that occurs in the profile, but *only* if the line does not have a
trailing comma to indicate the end of a rule. If the trailing comma
exists, then it assumes that it's a rule that it doesn't understand and
aborts.
With this patch, the simpler tools (aa-enforce, aa-complain, etc.) can
parse policies containing multi-line rules to an extent and continue to
function correctly. Again, aa-logprof and aa-genprof may have issues on
the writing back of profiles, so some assistance testing here would be
appreciated.
Some testcases are added to exercise the regex that looks for a rule
with a trailing comma but can still handle rules that have (,) or {,}
in them.
Patch history:
v1 - initial version
v2 - simplify and rearrange rule-ending comma search regex, since
we only care about the trailing comma
- add a new regex to search for trailing comments to filter out
- simplify reset of lastline variable
- restructure tests into a new script, and add more tests
v3 - add additional testcases, most of which are problematic and thus
commented out :(
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Acked-by: Christian Boltz <apparmor@cboltz.de>
This patch adds very limited support for very dumb parsing of dbus
rules. Basically, it stores dbus rules as raw strings wrapped in
a class.
There's class structure to indicate how I'd like to see fuller future
support for dbus rules to be implemented and act as a guidepost for
how to handle most rules, moving away from the giant structure of
nested dictionaries. A stub test script is included as well, with a
modification to the make check target to set the PYTHONPATH to point
in the right place.
With this patch, aa-audit, aa-autodep, aa-complain, aa-disable,
and aa-enforce all function for me. aa-logprof and aa-genprof have
functionality issues for me at the moment (one of them dumps a
backtrace even without this patch), and I'm not sure the writing out
of dbus rules is completely implemented for modified profiles.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch splits out the genprof tool functionality into a separate
command function, merging with the use_autodep function that already
existed.
Patch history:
v1 - initial revision
v2 - mark strings for translation and modify message when a profile
name is passed to aa-autodep, rather than a program name/path.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch moves the audit functionality to an audit specific command
function.
As an aside, the -r option is left in place here, because aa-audit
is a bit orthogonal to aa-enforce, aa-complain, and aa-disable.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch creates a separate tool.cmd_complain function, as well as
removes the -r remove option, to match aa-enforce and aa-disable.
It also cleans up some bits in aa-enforce now that aa-complain and
aa-enforce have been separated.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch splits out the aa-enforce functionality into a separate
method in the aa_tools class. It also removes one last reference to
the no-longer-existent -r option in the aa-enforce manpage.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch modifies the aa-disable tool implementation to allow it to
take a profile name (rather than a program name) as the argument(s)
for what to disable, as this was supported behavior in the perl
tools. (The rest of the commands that make use of the aa_tools.act()
method have not been exercised with this patch in place, as further
patches will separate those out.)
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch removes a debugging print statement accidentally left in,
as well as a duplicated initialization to a variable, and moves the
variable init closer to the declaration that the variable is a global.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
It's not useful to report the location of the temporary directory for
each test if you're going to immediately delete it.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
aa-genprof was incorrectly trying to refer to UI_xxx functions in
apparmor.aa rather than the correct apparmor.ui. This patch fixes the
issue.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
The relative directory tests in test-aa-easyprof.py were failing when
TMPDIR pointed to a directory that was a symlink, because the generated
tmpdir path was not the same as the realpath that easyprof resolved to.
This patch sets the tmpdir to the realpath of the result of the
tempfile.mkdtemp() to avoid the issue.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
The magic token used in the test suite is incorrectly stored
as an int rather than unsigned long leading to failure like this:
running changehat_misc
/tmp/testlibCTcwOe/source/trusty/apparmor-2.8.95~2411/tests/regression/apparmor/prologue.inc: line 176: 20184
Killed $testexec "$@" > $outfile 2>&1
Error: changehat_twice failed. Test 'CHANGEHAT (subprofile->subprofile)' was expected to 'pass'. Reason for
failure 'killed by signal 9'
Signed-off-by: Seth Arnold <seth.arnold@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
This patch removes the '--remove' option on aa-enforce as well as from
the man page. It also removes the test entry that contains it, but I
don't think this is entirely correct because I think the second half
of the test is dependent on the (now deleted) first half of the test.
(It also removes a missed reference to --revert in the aa-disable man
page.)
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch removes the non-funcional -r option for aa-disable, as
well as the test and manpage documentation for it.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
The test-aa-easyprof.py script has '/bin/ls' hardcoded as a path;
however, on systems that have undergone UsrMove, this is a symlink to
/usr/bin/ls. This patch fixes the issue by getting the realpath of
/bin/ls and storing it as an instance field.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
apparmor/tools.py:act() method into a separate cmd_disable()
method. The intent is to unwind the logic in act() into smaller, more
digestible chunks, while sharing commonality via helper functions
(e.g. the added get_next_to_profile() function).
A secondary driver of this change is that the tools fail when used
against the trunk profiles, due to act() forcing all the profiles to
be read and the tools not understanding the recently added dbus rules
(they were intentionally ignored as part of scoping the rewrite).
Unfortunately, this is not a solution for aa-enforce, aa-complain, etc.
as they are expected to know enough about profiles to understand and
update profile flags.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
In the course of developing apparmor dbus mediation, the aa_query_label
symbol was added to libapparmor on trunk, and given the symbol version
(via libapparmor.map) of APPARMOR_3.0. As apparmor upstream, we have
not made a release where this would have been exported.
Unfortunately, in Ubuntu, a version was released in 13.10 that included
the aa_query_label() symbol with a version of APPARMOR_1.1. This
can cause a breakage on that platform with the incorporation of the
impending apparmor 2.9 release.
This patch provides both versions (APPARMOR_1.1 and APPARMOR_2.9)
of the aa_query_label() symbol. It requires the function name in
kernel_interface.c to be renamed (similar to how the deprecated
change_hat() symbol is named in the source as __change_hat()),
otherwise linking fails with duplicated symbols. The default symbol
used will still be the APPARMOR_2.9 version, but binaries linked with
the APPARMOR_1.1 version would still continue to work unchanged.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
It prints a data structure in an easily readable output and is quite
useful for debugging. However, I don't recommend to call it in
production code ;-)
Acked-by: Kshitij Gupta <kgupta8592 at gmail.com>
Update path to tools in minitools_test.py
Fix assert values in config_test.py
Signed-off: Kshitij Gupta <kgupta8592@gmail.com>
Acked-by: Christian Boltz <apparmor@cboltz.de>
packaging systems that do a make clean while trying to represent
changes.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
AttributeError: 'module' object has no attribute 'UI_Info'
AttributeError: 'module' object has no attribute 'open_file_read'
AttributeError: 'module' object has no attribute 'check_for_apparmor'
Signed-off-by: Seth Arnold <seth.arnold@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
profile files with only tunables/global, but no profile. This patch
makes sure that the profile itsself is also written to the profile file.
Without the added line, filelist[prof_filename]['profiles'].keys()) in
serialize_profile was empty, which means the loop that writes the
profile was never executed.
Acked-by: Steve Beattie <steve@nxnw.org>
This commit adjusts the trunk libtool versions in accordance
with the outlined rules. It also adds clarifying text to the
documentation comment to explain how to use the rules and what the
resulting outcome is. Finally, it removes a bogus argument to linker
to forcibly set the SONAME in the library, as libtool will do this
automatically (and override the passed argument).
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
This patch fixes up the parser command invocation via
apparmor/common.py:cmd(), as it handles stdout/stderr redirection,
and the redirection that was being attempted were being handed as
arguments to the parser.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
Let "(F)inish" ask the user if he wants to save the changed profiles
before exiting, instead of aborting without saving (we already have
Abo(r)t for that ;-)
Acked-by: Kshitij Gupta <kgupta8592@gmail.com>
- add some debug logging in valid_path()
- fix a py2 incompability in DebugLogger.__init__ (OSError vs. IOError)
Acked-by: Kshitij Gupta <kgupta8592@gmail.com>
- use the (new) "parts" variable for the line.split result to make the
code less confusing
- change the line.startswith to check for "REPOSITORY:" (note the
added ":") like it was in the code before my previous patch.
- make the check for "NEVERSUBMIT" more exact
- print a warning on invalid REPOSITORY: lines and make sure to keep
them as unmodified line (it might just be a "normal" comment someone
added manually)
Acked-by: Kshitij Gupta <kgupta8592@gmail.com> (on IRC)
This patch
- preserves the complete initial comment
- makes sure whitespace inside the comment is kept (except leading
whitespace - line.trim() is still applied).
- no longer removes the "# vim:syntax" line
Note: I didn't test if handling the "REPOSITORY" line still works (in
theory it should), but without a working repo, I don't care too much ;-)
Acked-by: Steve Beattie <steve@nxnw.org>
The parser currently indicates that it exited successfully if invalid
arguments are passed to it, which makes it difficult to detect when
other tools are calling it incorrectly. This patch causes it to return
'1' indicating a failure.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
This is an updated version of the previous dnsmasq profile patch, again
from develop7 [at] develop7.info
Acked-by: John Johansen <john.johansen@canonical.com>
the suggestion to use @{XDG_DOWNLOAD_DIR} in abstractions/user-download as
well as the existing entries.
Acked-By: Jamie Strandboge <jamie@canonical.com>
Acked-By: Christian Boltz <apparmor@cboltz.de>
The xdg-user-dirs specification[1] allows for translatable and movable common
directories. While this may be beneficial for users who for example want to have
~/Pictures translated into their own language, this flexibility provides
challenges for AppArmor. Untranslated xdg user directories are typically (see
~/.config/user-dirs.dirs):
XDG_DESKTOP_DIR="$HOME/Desktop"
XDG_DOWNLOAD_DIR="$HOME/Downloads"
XDG_TEMPLATES_DIR="$HOME/Templates"
XDG_PUBLICSHARE_DIR="$HOME/Public"
XDG_DOCUMENTS_DIR="$HOME/Documents"
XDG_MUSIC_DIR="$HOME/Music"
XDG_PICTURES_DIR="$HOME/Pictures"
XDG_VIDEOS_DIR="$HOME/Videos"
On an Ubuntu system with the fr_CA locale installed, these become:
XDG_DESKTOP_DIR="$HOME/Desktop"
XDG_DOWNLOAD_DIR="$HOME/Téléchargements"
XDG_TEMPLATES_DIR="$HOME/Templates"
XDG_PUBLICSHARE_DIR="$HOME/Public"
XDG_DOCUMENTS_DIR="$HOME/Documents"
XDG_MUSIC_DIR="$HOME/Musique"
XDG_PICTURES_DIR="$HOME/Images"
XDG_VIDEOS_DIR="$HOME/Vidéos"
While the kernel and AppArmor parser handle these translations fine, the
profiles do not.
As an upstream, we can vastly improve the situation by simply creating the
xdg-user-dirs tunable using the default 'C' xdg-user-dirs values:
$ cat /etc/apparmor.d/tunables/xdg-user-dirs
@{XDG_DESKTOP_DIR}=Desktop
@{XDG_DOWNLOAD_DIR}=Downloads
@{XDG_TEMPLATES_DIR}=Templates
@{XDG_PUBLICSHARE_DIR}=Public
@{XDG_DOCUMENTS_DIR}=Documents
@{XDG_MUSIC_DIR}=Music
@{XDG_PICTURES_DIR}=Pictures
@{XDG_VIDEOS_DIR}=Videos
# Also, include files in tunables/xdg-user-dirs.d for site-specific adjustments
# to the various XDG directories
#include <tunables/xdg-user-dirs.d>
and then create the /etc/apparmor.d/tunables/xdg-user-dirs.d directory. With
that alone, we can start using rules like this in policy:
owner @{HOME}/@{XDG_MUSIC_DIR}/** r,
and users/admins can adjust /etc/apparmor.d/tunables/xdg-user-dirs or drop files
into /etc/apparmor.d/tunables/xdg-user-dirs.d, providing a welcome convenience.
This of course doesn't solve everything. Because users can modify their
~/.config/user-dirs.dirs file at will and have it point anywhere, so we can't
examine those files and do anything automatic there (when we have user policy we
can revisit this). This patch handles translations well though since use of
translations for these directories happens outside of the user's control. Users
who modify ~/.config/user-dirs.dirs can update policy like they need to now (ie,
this patch doesn't change anything for them).
[0] https://lists.ubuntu.com/archives/apparmor/2013-August/004183.html
[1] http://freedesktop.org/wiki/Software/xdg-user-dirs/
This patch adds basic support for XDG user dirs:
1. Update profiles/apparmor.d/tunables/global to include xdg-user-dirs.
2. Create the xdg-user-dirs tunable using the default 'C' xdg-user-dirs values
and includes tunables/xdg-user-dirs.d
3. Add profiles/apparmor.d/tunables/xdg-user-dirs.d/site.local with commented
out examples on how to use the directory.
Acked-By: Jamie Strandboge <jamie@canonical.com>
Acked-By: Christian Boltz <apparmor@cboltz.de>
Description: Allow applications run under sanitized_helper to connect to DBus
This was originally 0076_sanitized_helper_dbus_access.patch in the Ubuntu
apparmor packaging.
jdstrand: +1 (this is in the Ubuntu namespace, so feel free to commit)
apparmor packaging.
These were originally 0030-easyprof-sdk.patch and
0037-easyprof-sdk-pt2.patch. Jamie posted an updated
0030-easyprof-sdk_v2.patch and I squashed both patches into one commit.
Acked-By: Jamie Strandboge <jamie@canonical.com>
Grant access to specific files in the /var/run/user/UID/pulse/ directory to
remove access to potentially dangerous and non-essential files such as the
debug (cli) socket provided by the module-cli-protocol-unix module.
Author: Tyler Hicks <tyhicks@canonical.com>
Bug-Ubuntu: https://launchpad.net/bugs/1211380
Acked-by: Steve Beattie <steve@nxnw.org>
Description: update mod_apparmor man page for Apache 2.4 and add new
apparmor.d/usr.sbin.apache2 profile (based on the prefork profile)
Acked-by: Steve Beattie <steve@nxnw.org>
Differs from original 0036-libapache2-mod-apparmor-profile-2.4.patch
ubuntu patch -- I've deleted the "delete the apache 2.2 profile" part of
the patch. So apache 2.2's profile is also still supported.
Author: Micah Gersten <micah@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Modified by Seth Arnold; nvidia nvpau_wrapper.cfg permission was hoisted
up into an nvidia abstraction.
Author: Stéphane Graber <stgraber@ubuntu.com>
Acked-by: Steve Beattie <steve@nxnw.org>
This was originally patch 0018-lp1056391.patch in the Ubuntu apparmor
packaging; Steve noticed the now-redundant line for /var/lib/sss/mc/passwd
so I removed that at the same time.
After testing the dovecot profiles on a new server, I noticed
/usr/lib/dovecot/dict and /usrlib/dovecot/lmtp need more nameservice-
related permissions.
Therefore include abstractions/nameservice instead of adding more and
more files.
Acked-by: John Johansen (on IRC)
logprof/genprof and related utilities in python. Because the branch that
was worked on was not based on the apparmor tree, not all of the history
can be maintained for files that are not newly created or entirely
rewritten in the branch.
(This merge also includes a subsequent commit to the branch
I was merging from which includes my missed bzr add of
utils/apparmor/translations.py)
perl utilities to the deprecated to directory; a couple of perl
utilities remain, but they are still useful and do not depend on the
Immunix module (just the LibAppArmor perl module).
pep8 --ignore=E501,E302
on individual files. This uncovered a bug where the type of an object
was being compared to a type of a list. However, a python string is a
list of characters, and so would cause the test to be true.
the utility python modules to be used inside another tool with another
textdomain binding and still have the translations for that tool and the
stuff internal to the apparmor module convert properly.
When passing an include directory on the command line to
apparmor_parser, valgrind emits a warning:
Invalid read of size 4
at 0x404DA6: add_search_dir(char const*) (parser_include.c:152)
by 0x40BB37: process_arg(int, char*) (parser_main.c:457)
by 0x403D43: main (parser_main.c:590)
Address 0x572207c is 28 bytes inside a block of size 29 alloc'd
at 0x4C2A420: malloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
by 0x53E31C9: strdup (strdup.c:42)
by 0x404D94: add_search_dir(char const*) (parser_include.c:145)
by 0x40BB37: process_arg(int, char*) (parser_main.c:457)
by 0x403D43: main (parser_main.c:590)
This patch quiets the warning by removing strlen() calls on the t char
array. Instead, it only calls strlen() on the dir char array. t is a
dupe of dir and strlen(dir) does not trigger the valgrind warning.
Additionally, this patch adds a bit of defensive programming to the
while loop to ensure that index into the t array is never negative.
Finally, the valgrind suppression is removed from valgrind_simple.py.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
With commit 2364 addressing one of valgrind's false positives, we can
remove the related valgrind suppression entry from the test script.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
strlen() assumes that it can read an entire word but when a char array
does not end on a word boundary, it reads past the end of the array.
This results in the following valgrind warning:
Invalid read of size 4
at 0x40A162: yylex() (parser_lex.l:277)
by 0x40FA14: yyparse() (parser_yacc.c:1487)
by 0x40C5B9: process_profile(int, char const*) (parser_main.c:1003)
by 0x404074: main (parser_main.c:1340)
Address 0x578d870 is 16 bytes inside a block of size 18 alloc'd
at 0x4C2A420: malloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
by 0x53E31C9: strdup (strdup.c:42)
by 0x40A145: yylex() (parser_lex.l:276)
by 0x40FA14: yyparse() (parser_yacc.c:1487)
by 0x40C5B9: process_profile(int, char const*) (parser_main.c:1003)
by 0x404074: main (parser_main.c:1340)
This patch quiets the warning by not using strlen(). This can be done
because yyleng already contains the length of string.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
When the --cache-loc option was added in trunk commit 1916, it was
intended that -L would be the short form of the option (based on
documentation and usage changes). However, the commit mistakenly
did not include the short option in the list include in the call
to getopt_long(3). This patch adds it along with the indicator
that it requires an argument (the different cache location) to the
getopt_long() call.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
includes come out like
#include
##included <tunables/home>
which is wrong because #include by itself is broken, and since -p is
supposed to be removing includes, it should not be directly echoed
any keyword in the keyword table is double echoed
ownerowner /{run,dev}/shm/pulse-shm* rwk
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
abstractions/mysql.
This binary/profile seems to be the only one that needs to do this, so
add it to this profile (instead of abstractions/mysql) to avoid superfluous
permissions for other programs with abstractions/mysql
Acked-by: John Johansen <john.johansen@canonical.com>
after testing the dovecot profiles on a new server, I noticed
/usr/sbin/dovecot needs some more permissions:
- mysql access
- execution permissions for /usr/lib/dovecot/dict and lmtp
- write access to some postfix sockets, used to
- provide SMTP Auth via dovecot
- deliver mails to dovecot via LMTP
- and read access to /proc/filesystems
Acked-by: John Johansen <john.johansen@canonical.com>
logprof.conf contains a list of binaries in the [qualifiers] section
that should for example never have their own profile.
Since some distributions moved lots of files from /bin/ to /usr/bin/
("UsrMove"), this list is outdated.
The patch adds copies of all /bin/ (and /sbin/) lines with /usr
prepended.
Acked-by: John Johansen <john.johansen@canonical.com>
The usr.sbin.dovecot profile needs several updates for dovecot 2.x,
including
- capability dac_override and kill
- Px for various binaries in /usr/lib/dovecot/
The patch also adds a nice copyright header (I hope I got the bzr log
right ;-)
Acked-by: John Johansen <john.johansen@canonical.com>
dovecot 2.x comes with several new binaries in /usr/lib/dovecot.
This patch adds profiles for
/usr/lib/dovecot/anvil
/usr/lib/dovecot/auth
/usr/lib/dovecot/config
/usr/lib/dovecot/dict
/usr/lib/dovecot/dovecot-lda
/usr/lib/dovecot/lmtp
/usr/lib/dovecot/log
/usr/lib/dovecot/managesieve
/usr/lib/dovecot/ssl-params
References: https://bugzilla.novell.com/show_bug.cgi?id=851984
Acked-by: John Johansen <john.johansen@canonical.com>
Introduces tunables/dovecot (with @{DOVECOT_MAILSTORE}) and replace
the mail storage location in various dovecot-related profiles with
this variable.
Also add nice copyright headers (I hope I got the bzr log right ;-)
Acked-by: John Johansen <john.johansen@canonical.com>
From: Felix Geyer <debfx@ubuntu.com>
AppArmor requires read and write permission to connect to
unix domain sockets but the nameservice abstraction only
grants write access to the avahi socket.
As a result mdns name resolution fails.
Acked-by: John Johansen <john.johansen@canonical.com>
This patch adds support for the rttime rlimit (aka RLIMIT_RTTIME),
available since the 2.6.25 kernel, according to the getrlimit(2)
man page; see that man page for more details on this rlimit.
An acceptance test is also added, as well as an update to the
apparmor.vim input template.
While reviewing to see what made sense in apparmor.vim for the rttime
rlimit, I discovered that RLIMIT_RTTIME's units are microseconds, not
seconds like RLIMIT_CPU (according to the setrlimit(2) manpage). This
necessitated not sharing the case switch with RLIMIT_CPU. I didn't add
a keyword for microseconds, but I did for milliseconds. I also don't
accept any unit larger than minutes, as it didn't seem appropriate
(and even minutes felt... gratuitous). I would appreciate feedback
on what keywords would be useful here.
Patch History:
v1: initial submission
v2: - add apparmor.vim support for rttime keyword
- adjust RLIMIT_TIME value assignment due to its units being
microseconds, not seconds, and add milliseconds keyword.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
This patch replaces explicitly named output targets with the make
variable $@ as well as an instance where dbus_common.h was being added
to the compile command line due to the use of $^ rather than $<.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
Close file handle left opened if parser.cfg is found and read from.
Found by cppcheck.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
Seth Arnold noticed an ugly string.clear(); convert_entry(string,
NULL) pattern occurred frequently following the conversion to using
std::string. This patch replaces that by using a static pointer to
a constant string matching pattern, and also converts other uses of
that pattern. It also adds a function wrapper that will clear the
passed buffer before calling convert_entry().
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
As noted by Seth Arnold, in expand_by_alternations() if our set
variable has at least one value, then we're going to rewrite the entry,
so rather than sprinkle the free()s near where the reallocation occurs,
use one free() once we're guaranteed to need to do so.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
As suggested by Seth Arnold, we can use string::find_last_not_of()
instead of using C++'s hideous reverse iterators.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
As noted by Seth Arnold, e_buffer_overflow is no longer set in
convert_aaregex_to_pcre(), so remove it and the sole check for it.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
This patch converts a stack allocated buffer into an std::ostringstream
object. The stringstream interface for specifying the equivalent of
a printf %02x conversion is a bit of an awkward construction, however.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
Based on feedback from Seth Arnold, the convert_aaregex_to_pcre()'s
first argument is const char *, and thus the unit test macros don't need
to pass a copy of the input string to it, as it's guaranteed to be
unmodified by the function.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
This patch eliminates the bison warning about "%name-prefix =" being
deprecated.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
This patch includes the errno in the log messages generated by two
different failed aa_change_hat() calls and the failure to open
/dev/urandom to get the random token, to further ease failure
diagnosis.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
This patch removes unnecessary back out aa_change_hat() calls that occur
if the prior call to aa_change_hat() call failed. It used to be case
that an aa_change_hat() call that failed would result in the task being
placed in a profile with no permissions except the ability to
aa_change_hat() back out, but this behavior has been removed from
apparmor for many, many years now.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
This patch adds code that checks the resulting hat that apache gets
placed into, and verifies that if the apache configuration specified
that an AAHatName or AADefaultHatName should have been the resulting
hat. If it wasn't, emit a warning message to the apache log, as this
likely indicates a mismatch between the apache configuration and its
apparmor policy (i.e. why define AAHatName if you aren't going to
create the corresponding hat in the apparmor policy?)
Note for AADefaultHatName, a message is not logged if a defined
AAHatName would also apply or if there is a hat defined for the uri,
as each of those come first in the order of attempted hats.
Also note that the way the hat name is manually calculated will break
for nested profiles and stacking. It should be fine for all current
deployments as we don't allow nesting beyond the first subprofile level
in policy yet. And stacking will likely only be used between namespaces
where aa_getcon() will not report parent namespace info. However, when
libapparmor adds functionality to query the hatname, the code that
computes it here should be replaced by a call to that library function.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
This patch converts the request entry point from using multiple (if
necessary) aa_change_hat() calls into a single aa_change_hatv() call,
simplifying the code a bit, requiring fewer round trips between
mod_apparmor and the kernel for each request, as well as providing more
information when the apache profile is in complain mode.
Patch history:
v1: initial version
v2: - the server config (scfg) code accidentally re-added the
directory config (dcfg) hat to the vector of hats, fix that
- actually add the DEFAULT_URI hat to the vector of hats, instead
of only logging that that is happening.
- pass errno to ap_log_rerror() if aa_change_hatv() call fails.
- don't call aa_change_hat again if aa_change_hatv() call fails,
as this is no longer necessary.
v3: - Based on feedback from jjohansen, convert exit point
aa_change_hat() call to aa_change_hatv(), in order to work
around aa_change_hat() bug addressed in trunk rev 2329,
which causes the exiting aa_change_hat() call to fail and
results in the apache process being killed by the kernel.
When it's no longer likely that mod_apparmor could run into
a system libapparmor that still contains this bug, this can
be converted back.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
Bug: https://bugs.launchpad.net/ubuntu/+source/apparmor/+bug/1207424
This patch makes the default value for AADefaultHatName be the
server/vhost name, which can be specified in apache via the ServerName
configuration declaration. It can be overridden by setting
AADefaultHatName directly. Thus, with this patch applied, the order of
attempted hats will be:
1. try to aa_change_hat(2) into a matching AAHatName hat if it exists
and applies, otherwise
2. try to aa_change_hat(2) into the URI itself, otherwise
3. try to aa_change_hat(2) into the value of ServerName, unless
AADefaultHatName has been explicitly set for this server/vhost, in
which case that value will be used, otherwise
4. try to aa_change_hat(2) into the DEFAULT_URI hat, if it exists,
otherwise
5. fall back to the global Apache policy
This should eliminate the need for most admins to define both
ServerName and AADefaultHatName, unless there's a specific need for
the values to deviate.
Man page documentation is updated as well, though probably more
wordsmithing is needed there for clarity.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
When defining an AADefaultHatName entry, it was being stored in the
passed mconfig location, which is not the module specific server
config, but instead the top level (i.e. no path defined) default
directory/location config. This would be superceded by a more specific
directory config if it applied to the request. Thus, if an AAHatName was
defined that applied, but the named hat was not defined in the apparmor
policy, mod_apparmor would not attempt to fall back to the defined
AADefaultHatName, but instead jump directly to trying the DEFAULT_URI
hat.
This patch fixes it by storing the defined AADefaultHatName correctly in
the module specific storage in the related server data structure. It
also adds a bit of developer debugging statements.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
Bug: https://launchpad.net/bugs/1207424
This patch adds the name of the hat to the log message about the
initial aa_change_hat call, just to be explicit about what's happening
when debugging and changes the formatting slightly of the exiting
change_hat log message.
Patch history:
v1: initial version
v2: tweak output of exit trace message
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
mod_apparmor never got converted to use the renamed aa_change_hat()
call (there's a compatibility macro in sys/apparmor.h); this patch does
that as well as converting the type of the magic_token to long from int.
(This patch is somewhat mooted by a later patch in the series to
convert to using aa_change_hatv(), but would be a safer candidate
for e.g. the 2.8 branch.)
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
This patch converts the debug_dump_uri() function to use the trace
loglevels and enable it all the time, rather than just when DEBUG is
defined at compile time.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
Apache 2.4 added addition logging levels. This patch converts some of
the log messages that are more intended for mod_apparmor development
and debugging than for sysadmins configuring mod_apparmor to use trace1
(APLOG_TRACE1) level instead. Since apache 2.2. does not contain this
level (or define), we define it back to APLOG_DEBUG.
Patch history:
v1: initial version
v2: mark a couple of additional log messages as trace1 level
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
The apache2 mod_apparmor module was failing to log debugging messages
when the apache loglevel was set to debug or lower (i.e. traceN). This
patch fixes it by using ap_log_rerror() (for request specific messages,
with the request passed for context) and ap_log_error() (more general
messages outside of a request context).
Also, the APLOG_USE_MODULE macro is called, to mark the log messages as
belonging to the apparmor module, so that the apache 2.4 feature of
enabling debug logging for just the apparmor module will work, with an
apache configuration entry like:
LogLevel apparmor:debug
See
http://ci.apache.org/projects/httpd/trunk/doxygen/group__APACHE__CORE__LOG.html
for specific about the ap_log_*error() and APLOG_USE_MODULE functions
and macros, and
http://httpd.apache.org/docs/2.4/mod/core.html.en#loglevel
for the bits about module specific logging.
Patch history:
v1: initial version
v2: - revert to using ap_log_error with (the 2.4 specific)
ap_server_conf outside of a request specific context, as the
pool specific ap_log_perror messages weren't being reported.
- add compatibility workaround for apache 2.2
v3: keep commented out merge function's log call consistent with the
others
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
This patch fixes the format string for the magic token in aa_change_hat
to match the type of the magic token (long). Without this, on 64
bit platforms, only the bottom 32 bits of the token would be used.
aa_change_hatv() has the correct format string, so an aa_change_hatv()
call followed by an exiting aa_change_hat() call would result in the
latter having a different token, which would cause the process to be
killed by apparmor.
(Hat tip to John Johansen for spotting the actual bug.)
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
(collected in the openSUSE package over the last months)
- add abstractions/samba to usr.sbin.winbindd profile
(and cleanup things that are included in the abstraction - the cleanup
part is not in the openSUSE package)
- add capabilities ipc_lock and setuid to usr.sbin.winbindd profile
(bnc#851131)
- updates for samba 4.x and kerberos (bnc#846586#c12 and #c15,
bnc#845867, bnc#846054)
- drop always-outdated "Last Modified" comment
References: see the bnc# above (they are bug numbers at
bugzilla.novell.com)
Acked-by: John Johansen <john.johansen@canonical.com>
This patch eliminates the complaints from running:
pep8 --ignore=E501 aa-easyprof vim/
(E501 is 'line too long', which I'm not too chuffed about.)
Mostly, it's a lot of whitespace touchups, with a few conversions from
'==' to 'is'.
Commit includes applied feedback from cboltz.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
Current builds include many warnings when building translations message
files like so:
msgfmt -c -o ja.mo ja.po
ja.po:5: warning: header field 'Language' missing in header
According to what I read in the entry for Language in
http://www.gnu.org/software/gettext/manual/gettext.html#Header-Entry
the language entry should be (in our case) the same as the file name
minus the .po suffix. This patch adds the language field for those
po files that were missing it.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
This patch adds several assorted language tests, to exercise various
parts of the parser that were not being covered by the language tests
previously. Areas lacking were found using the coverage compilation
option; coverage from the language tests is still incomplete.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
This patch updates the Report-Msgid-Bugs-To: to point to the apparmor
list instead of the old Novell forge address. It also makes the
Project-Id-Version: field consistent.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
The rlimits syntax checking support in apparmor.vim was broken in
various unhelpful ways:
- lacked support for the 'infinity' keyword (aka RLIM_INFINITY)
- lacked support for the 'ofile' rlimit, an alias for the nofile
rlimit
- lacked support for the 'cpu' rlimit (aka RLIMIT_CPU)
- incorrect syntax for nofile|nproc|rtprio rlimits (didn't include
required '<=' between the limit name and value)
- incorrect syntax for specifying optional SI units for size based
rlimits (e.g. 'MB' is required, but syntax only allowed incorrect
'M'; that said, one could argue the parser is overly strict here,
and the pattern should be '[KMG]B?')
(See the setrelimit(2) man page for more details on the specifics of the
rlimit definitions.)
This patch fixes the above issues.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
The parser was lacking language tests for rlimits. This test adds
several, one for each rlimit type.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
As noted by Seth Arnold, there's now only one failure case in the
function and thus does not warrant a goto target (especially since
there's no cleanup to occur).
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
Change uservars.inc subdomain variable to use the in-tree parser by
default.
Also, clean up some commented out subdomain values that don't look to be
in use any longer and add one commented out value pointing to the system
parser.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Subtle change to remove the "..." between the test description and
result and also to single-space the output. This brings the output in
line with what minimize.sh outputs, which is the test that runs just
before equality.sh.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
The regression test README examples use sh when showing how to run
individual tests but bash is needed, instead.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
This test ensures that the proper DFA minimization occurs when a
permissive D-Bus abstraction #include's the corresponding strict
abstraction.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-By: Christian Boltz <apparmor@cboltz.de>
Create a new strict accessibility bus abstraction.
The strict abstraction only allows for calling the Hello, AddMatch,
RemoveMatch, GetNameOwner, NameHasOwner, and StartServiceByName methods
that are exported by the D-Bus daemon.
The permissive abstraction reuses the strict abstraction and then allows
all communications on the accessibility bus.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
Move some of the file rules from the existing permissive session bus
abstraction into a new strict session bus abstraction. Leave the
dbus-launch rule in the permissive profile since not all applications
will need it.
The strict abstraction only allows for calling the Hello, AddMatch,
RemoveMatch, GetNameOwner, NameHasOwner, and StartServiceByName methods
that are exported by the D-Bus daemon.
The permissive abstraction reuses the strict abstraction and then allows
all communications on the session bus.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
Acked-By: Jamie Strandboge <jamie@canonical.com>
Move the file rule from the existing permissive system bus abstraction
into a new strict system bus abstraction.
The strict abstraction only allows for calling the Hello, AddMatch,
RemoveMatch, GetNameOwner, NameHasOwner, and StartServiceByName methods
that are exported by the D-Bus daemon.
The permissive abstraction reuses the strict abstraction and then allows
all communications on the system bus.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
std::max in C++ requires that both arguments be the same type. The
previous fix added std::max comparisons between unsigned long numeric
constants and size_t, this fix casts the numeric constants to size_t.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
Make the accept information dump output be in hexidecimal like the
other dumps so its easier to reference between them.
Signed-off-by: John Johansen <john.johansen@canonical.com>
This was part of the original minimization patch, but got dropped when
applying to bzr. Again bzr status didn't show any files out of place
nor did the patching fail :(
Signed-off-by: John Johansen <john.johansen@canonical.com>
This diff is part of the diffencode patch but was dropped when it was
applied to bzr. I have no idea why and status showed a clean tree.
Signed-off-by: John Johansen <john.johansen@canonical.com>
So DFA minimization has a bug and feature that keeps it from minimizing
some dfas completely. This feature/bug did not result in incorrect dfas,
it just fails to result in full minimization.
The same mappings comparison is wrong. Or more correctly it is right when
transitions are not remapped to minimization partitions, but it may be
wrong when states are remapped. This means it will cause excess
partitioning (not removing all the states it should).
The trans hashing does a "guess" at partition splitting as a performance
enhancement. Basically it leverages the information that states that have
different transitions or transitions on different characters are not the
same. However this isn't always the case, because minimization can cause
some of those transitions to be altered. In previous testing this was
always a win, with only a few extra states being added some times. However
this changes with when the same mappings are fixed, as the hashing that was
done was based on the same flawed mapping as the broken same mappings.
If the same mappings are fixed and the hashing is not removed then there
is little to no change. However with both changes applied some dfas see
significant improvements. These improvements often result in performance
improvements despite minimization doing more work, because it means less
work to be done in the chfa comb compression
eg. test case that raised the issue (thanks tyler)
/t { mount fstype=ext2, mount, }
used to be minimized to
{1} <== (allow/deny/audit/quiet)
{6} (0x 2/0/0/0)
{1} -> {2}: 0x7
{2} -> {3}: 0x0
{2} -> {2}: []
{3} -> {4}: 0x0
{3} -> {3}: []
{4} -> {6}: 0x0
{4} -> {7}: 0x65 e
{4} -> {5}: []
{5} -> {6}: 0x0
{5} -> {5}: []
{6} (0x 2/0/0/0) -> {6}: [^\0x0]
{7} -> {6}: 0x0
{7} -> {8}: 0x78 x
{7} -> {5}: []
{8} -> {6}: 0x0
{8} -> {5}: 0x74 t
{8} -> {5}: []
with the patch it is now properly minimized to
{1} <== (allow/deny/audit/quiet)
{6} (0x 2/0/0/0)
{1} -> {2}: 0x7
{2} -> {3}: 0x0
{2} -> {2}: []
{3} -> {4}: 0x0
{3} -> {3}: []
{4} -> {6}: 0x0
{4} -> {4}: []
{6} (0x 2/0/0/0) -> {6}: [^\0x0]
The evince profile set sees some significant improvements picking a couple
example from its "minimized" dfas (it has 12) we see a reduction from 9720
states to 6232 states, and 6537 states to 3653 states. All told seeing the
performance/profile size going from
2.8 parser: 4.607s 1007267 bytes
dev head: 3.48s 1007267 bytes
min fix: 2.68s 549603 bytes
of course evince is an extreme example so a few more
firefox
2.066s 404549 bytes
to
1.336s 250585 bytes
cupsd
0.365s 90834 bytes
to
0.293s 58855 bytes
dnsmasq
0.118s 35689 bytes
to
0.112s 27992 bytes
smbd
0.187s 40897 bytes
to
0.162s 33665 bytes
weather applet profile from ubuntu touch
0.618s 105673 bytes
to
0.432s 89300 bytes
I have not seen a case where the parser regresses on performance but it is
possible. This patch will not cause a regression on generated policy size,
at worst it will result in policy that is the same size
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Differential state compression encodes a state's transitions as the
difference between the state and its default state (the state it is
relative too).
This reduces the number of transitions that need to be stored in the
transition table, hence reducing the size of the dfa. There is a
trade off in that a single input character may have to traverse more
than one state. This is somewhat offset by reduced table sizes providing
better locality and caching properties.
With carefully encoding we can still make constant match time guarentees.
This patch guarentees that a state that is differentially encoded will do at
most 3m state traversal to match an input of length m (as opposed to a
non-differentially compressed dfa doing exactly m state traversals).
In practice the actually number of extra traversals is less than this becaus
we selectively choose which states are differentially encoded.
In addition to reducing the size of the dfa by reducing the number of
transitions that have to be stored. Differential encoding reduces the
number of transitions that need to be considered by comb compression,
which can result in tighter packing, due to a reduction in sparseness, and
also reduces the time spent in comb compression which currently uses an
O(n^2) algorithm.
Differential encoding will always result in a DFA that is smaller or equal
in size to the encoded DFA, and will usually improve compilation times,
with the performance improvements increasing as the DFA gets larger.
Eg. Given a example DFA that created 8991 states after minimization.
* If only comb compression (current default) is used
52057 transitions are packed into a table of 69591 entries. Achieving an
efficiency of about 75% (an average of about 7.74 table entries per state).
With a resulting compressed dfa16 size of 404238 bytes and a run time for
the dfa compilation of
real 0m9.037s
user 0m8.893s
sys 0m0.036s
* If differential encoding + comb compression is used, 8292 of the 8991
states are differentially encoded, with 31557 trans removed. Resulting in
20500 transitions are packed into a table of 20675 entries. Acheiving an
efficiency of about 99.2% (an average of about 2.3 table entries per state
With a resulting compressed dfa16 size of 207874 bytes (about 48.6%
reduction) and a run time for the dfa compilation of
real 0m5.416s (about 40% faster)
user 0m5.280s
sys 0m0.040s
Repeating with a larger DFA that has 17033 states after minimization.
* If only comb compression (current default) is used
102992 transitions are packed into a table of 137987 entries. Achieving
an efficiency of about 75% (an average of about 8.10 entries per state).
With a resultant compressed dfa16 size of 790410 bytes and a run time for d
compilation of
real 0m28.153s
user 0m27.634s
sys 0m0.120s
* with differential encoding
39374 transition are packed into a table of 39594 entries. Achieving an
efficiency of about 99.4% (an average of about 2.32 entries per state).
With a resultant compressed dfa16 size of 396838 bytes (about 50% reduction
and a run time for dfa compilation of
real 0m11.804s (about 58% faster)
user 0m11.657s
sys 0m0.084s
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
By raising an error for being unable to find libapparmor any time
a make command is run, we break things like make clean and other
targets that don't strictly depend on libapparmor existing (note that
Tyler's implementation for the parser did not do this). This patch
fixes this for the regression tests, mod_apparmor and pam_apparmor
by making a separate libapparmor_check target that looks to see if
an error message should be generated.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch adds support for the USE_SYSTEM make flag and adjusts
search paths for mod_apparmor and pam_apparmor, as well as fixing up
a couple of the (probably ought to be deprecated) tomcat locations
where apparmor.h is included.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch converts the regression tests to build against the in-tree
libapparmor, giving an error if libapparmor has not already been built.
It also maintains support for building against the system libapparmor
via passing
USE_SYSTEM=1
on the make command line. An error is also given if the system
libapparmor cannot be found, indicating that development packages need
to be installed. The check to look for libapparmor is also updated
to make use of libapparmor pkg-config data if available.
Patch history:
v1: initial submission
v2: convert from including the apparmor.h on the command line to
specifying an in-tree libapparmor header include path, now
that their location has been adjusted to make it safe to do
so. Remove work around related to defining _GNU_SOURCE.
v3: fix LDLIBS to use output of pkg-config et al tests instead of
hardcoding -lapparmor in the USE_SYSTEM case.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
By default, statically link against the in-tree libapparmor. If the
in-tree libapparmor is not yet built, print a helpful error message. To
build against the system libapparmor, the USE_SYSTEM make
variable can be set on the command line like so:
$ make USE_SYSTEM=1
This patch also fixes issues around the inclusion of the apparmor.h
header. Previously, the in-tree apparmor.h was always being included
even if the parser was being linked against the system libapparmor.
It modifies the apparmor.h include path based on the previous patch
separating them out in the libapparmor source. This was needed because
header file name collisions were already occurring.
For source files needing to include apparmor.h, the make targets were
also updated to depend on the local apparmor.h when building against
the in-tree libapparmor. When building against the system libapparmor,
the variable used in the dependency list is empty. Likewise, a
libapparmor.a dependency is added to the apparmor_parser target when
building against the in-tree apparmor.
Patch history:
v1: from Tyler Hicks <tyhicks@canonical.com>
- initial version
v2: revert to altering the include search path rather than including
the apparmor.h header directly via cpp arguments, alter the
include statements to <sys/apparmor.h> which will work against
either in-tree or (default) system paths.
v3: convert controlling variable to USE_SYSTEM from SYSTEM_LIBAPPARMOR
to unify between the parser and the regression tests.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Signed-off-by: Steve Beattie <steve@nxnw.org>
This patch moves the apparmor.h and aalogparse.h headers
from the libapparmor/src/ directory to a new directory
libapparmor/include/. The apparmor.h header is stored in a sys/
directory within libapparmor/include/ to match its usual install
location in /usr/include/sys/, simplifying the #include statements of
source that wishes to include either the in-tree or system installed
version of the header (i.e. #include <sys/apparmor.h> can be used
everywhere).
The patch size is inflated by the movements of the header files, which
are unchanged except for their locations. Otherwise, the rest of the
changes are to modify the include search path or to stop looking in
$CWD for one of the headers.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
libtoolize is the standard mechanism for incorporating libtool support
into a library; however, libapparmor's autogen.sh script specifically
looks for the existence of the libtool binary rather than libtoolize.
The libtoolize tool automatically generates a libtool script and does
not require the existence of the libtool binary, so we no longer need
to check for it.
The autogen.sh script aborting because it can't find the libtool
binary causes a spurious build failure in Ubuntu 14.04, due to the
libtool binary being separated out into its own package, and which
is not a strict dependency for libtool.
(I also added setting the package variable so that the error message
emitted will indicate that the failure is in building libapparmor.)
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
The AppArmor kernel now checks for both read and write permissions when
a process calls connect() on a UNIX domain socket.
The patch updates four abstractions that were found to be needing
changes after the change in AF_UNIX kernel mediation.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
The accessibility bus uses an abstract socket, so there hasn't been a
need for an accessibility bus abstraction in the past. Now that D-Bus
mediation is supported, an abstraction becomes a useful place to put
accessibility bus D-Bus rules.
This patch follows the lead of the dbus and dbus-session abstraction by
granting full access to the accessibility bus.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Before D-Bus mediation support was added to AppArmor, the dbus and
dbus-session abstractions granted full access to the system and session
buses, respectively.
In order to continue granting full access to those buses, bus-specific
D-Bus mediation rules need to be added to the abstractions.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
With the previous patch to switch to using alternations for variable
expansion, the clone_and_chain set of functions are no longer needed
and no longer need to be passed around. This patch removes them.
(I kept this patch separate to keep the previous patch smaller and more
easily reviewed.)
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch converts the parser's variable expansion from adding new
entries for each additional variable value to incorporating an
alternation that includes all the values for the variable; e.g. given:
@{BINS}=/bin /usr/bin /sbin /usr/sbin
@{BINS}/binary ix,
rather than expanding to exntries for
/bin/binary
/usr/bin/binary
/sbin/binary
/usr/sbin/binary
one entry would remain that looks like:
{/bin,/usr/bin,/sbin,/usr/sbin}/binary
One complication with this patch is that we try to prevent mistakes for
our users with variable expansion around '/'s; it's common for people to
write profiles that contain things like:
@{BAR}=/bingo/*/ /bango/
/foo/@{BAR}/baz
We already have a post-processing step that walks entries looking
for multiple sequences of '/'s and filters them into single
'/' which worked when creating new entries for each variable
expansion. Converting to alternation expansion breaks this filtering,
so code is added that removes leading and trailing slashes in variable
values in the expansion if the character immediately preceding or
following the variable is also a slash.
The intent behind this is to reduce the amount of memory allocations
and structure walking that needed to occur in when converting from the
entry strings to the back end nodes. Examples with real world profiles
showed performance improvements ranging from 2.5% to 10%. However,
because the back end operations are sensitive to the front end inputs,
it is possible for worse results to occur; for example, it takes the
simple_tests/vars/vars_stress_0[123].sd tests significantly longer to
complete after this patch is applied (vars_stress_03.sd in particular
takes ~23 times longer). An initial analysis of profiling output in
this negative case looks like it causes the tree simplification in
the back end to do more work for unknown reasons.
On the other hand, the test simple_tests/vars/vars_dbus_9.sd
(introduced in "[patch 09/12] parser: more dbus variable testcases")
takes ~1 sec to complete on my laptop before this patch, and roughly
0.01s with this patch applied.
(One option would be to keep the "expand entries" approach as an
alternative, but I couldn't come up with a good heuristic for when
to use it instead.)
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch addresses the FIXMEs from the last patch by converting
process_mnt_entry's typebuf from a char[] to std::string. As a side
effect, the code in build_list_val_expr() is greatly simplified.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch removes the string length limit in convert_aaregex_to_pcre()
usage. One of the benefits to moving to C++ is the ability to use
std::strings, which dynamically resize themselves. While it's a large
patch, a non-trivial amount is due to needing to get a char * string
back out via the c_str() method.
The unit tests are modified to include checks to ensure that
convert_aaregex_to_pcre only appends to the passed pcre string,
it never resets it.
As the test case with overlong alternations added in the previous
patch now passes, the TODO status is removed from it.
(Note: there's a couple of FIXME comments related to converting typebuf
to std::string that are added by this patch that are addressed in the
next patch. I kept that conversion separate to try to reduce the size
of this patch a little.)
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch adds a test case with an extremely large set of alternations.
It is marked TODO, because it fails with the current parser due to
strings used in convert_aaregex_to_pcre() being limited to (roughly)
PATH_MAX.
While contrived, it is possible to have alternations that are longer
than PATH_MAX that always match paths that are shorter than PATH_MAX.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
At least that's how this patch started ;-)
The updated (and much bigger) patch
- removes the note about can ?not mknod
- also removes mount and umount from the can ?not list which are covered
by mount rules now
- updates the example audit.log lines to the current log format
- updates the description of the log format
Acked-By: Seth Arnold <seth.arnold@canonical.com> (on IRC)
Seth also promised a follow-up patch with the remaining changes.
Patch history:
v1: initial version
v2: based on feedback from cboltz and sarnold:
- fix bad grammar when mentioning *.gcno and *.gcda files
- mention that distros generally don't need other options besides
verbose builds
- fix 'the valgrind' grammar messup.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch adds more testcases around variables used in dbus rules.
In particular, it
- attempts to verify that variable expansion and alternation
expansion results in identical DFA blobs,
- tests that variables can be expanded within alternations,
- tests that alternations can occur in variable definitions, and
- that having alternations inside variable declarations that are
used inside alternations results in parsing success
Note that vars/vars_dbus_9.sd veers into stress test land, as the
combinatoric expansion results in over 1000 dbus rule entries being
generated, which means that DFA reduction on all the fields takes
noticeable amounts of time (around 1s on my i5 ivy-core laptop).
Patch history:
v1: initial version
v2: based on feedback:
- add more alternation tests for cases where only part of the
alternation is defined within a variable
- mark test with nested alternations as being successful now that
the patch that implements it was accepted
v3: based on feedback from cboltz:
- tst/simple_tests/vars/vars_dbus_9.sd: reference all variables
declared, including a variable that references another variable
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
When compiling the parser, g++ currently emits warnings like so:
profile.h: In constructor ‘Profile::Profile()’:
profile.h:177:11: warning: missing initializer for member ‘aa_rlimits::limits’ [-Wmissing-field-initializers]
rlimits = { 0 };
^
This patch fixes the issue.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
The parser was not checking for an error when reading from
/proc/sys/kernel/osrelease. Additionally, valgrind was complaining
because of the uninitialized space in the buffer in between where
the read(2) had deposited its data and where the parser was writing
a trailing NUL to close the string. This patch fixes the above by
writing the NUL byte at the position at the end of the read characters
and checks for a negative result from the read() call.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
The parser was converting alternation characters ('{', '}', and ',')
to their pcre versions ('(', ')', and '|', respectively) that occurred
inside of character class patterns (i.e. inside '[ ]'). This patch
fixes the issue and adds a few unit tests around character classes.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Currently, D-Bus rules are the only type of policy that we expect to be
queried from userspace. Therefore, we do not need to export other
mediation types at this time.
This patch removes all AA_CLASS_* macros, except AA_CLASS_DBUS, from
libapparmor's apparmor.h header. These macros are already defined in the
parser's policydb.h header.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Now that the parser links against libapparmor, it makes sense to move
all public permission types and flags to libapparmor's apparmor.h. This
prevents duplication across header files for the parser and libapparmor.
Additionally, this patch breaks the connection between
AA_DBUS_{SEND,RECEIVE,BIND} and AA_MAY_{WRITE,READ,BIND} by using raw
values when defining the AA_DBUS_{SEND,RECEIVE,BIND} macros. This makes
sense because the two sets of permission flags are from two distinctly
different mediation types (AA_CLASS_DBUS and AA_CLASS_FILE). While it is
nice that they share some of the same values, the macros don't need to
be linked together. In other words, when you're creating a D-Bus rule,
it would be incorrect to use permission flags from the AA_CLASS_FILE
type.
The change mentioned above allows the AA_MAY_{WRITE,READ,BIND} macros
to be removed from public-facing apparmor.h header.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Simple regression test that calls AddMatch using a match string that
sets up eavesdropping on all method call messages.
The shell script file runs the test unconfined and under a variety of
confinement profiles to make sure that eavesdropping confinement is
working as intended.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Rules using implied permissions may pick up the eavesdropping
permission, depending on the conditionals present in the rule.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Allows for the policy writer to grant permission to eavesdrop on the
specified bus. Some example rules for granting the eavesdrop permission
are:
# Grant send, receive, bind, and eavesdrop
dbus,
# Grant send, receive, bind, and eavesdrop on the session bus
dbus bus=session,
# Grant send and eavesdrop on the system bus
dbus (send eavesdrop) bus=system,
# Grant eavesdrop on any bus
dbus eavesdrop,
Eavesdropping rules can contain the bus conditional. Any other
conditionals are not compatible with eavesdropping rules and the parser
will return an error.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch adds a warning when quote characters '\' are added
unnecessarily, generates an error when a single quote is the last
character in a pattern, and uncomments and corrects the relevant unit
test cases.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch adds a parser make variable and a make target for building
the compiler with coverage compilation flags. With this, coverage
information can be generated by running tests/test suites against the
built parser and run through tools like gcovr.
Patch History:
v1: initial version
v2: refreshed/no change
v3: address feedback from sarnold:
- mark coverage target as phony
- correct missing '.' typo in clean target
- make coverage extensions consistent in clean targets
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
equivalents. (v2)
This patch verifies basic alternation usage.
Patch history:
v1: initial revision
v2: mark nested alternation tests as passing, as it was deemed a bug
that the parser didn't support them.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-By: Christian Boltz <apparmor@cboltz.de>
This patch adds a test that verifies the parser considers an emty
character class regex as a parse arror.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-By: Christian Boltz <apparmor@cboltz.de>
This patch annotates that a couple of values emitted on failure are
of type size_t, eliminating a couple of compiler warnings.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
This patch adds unit tests and macros for the convert_aaregex_to_pcre()
function.
Patch history:
v1: initial version
v2: - give more verbose output on failures
- free memory used in tests
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
useful for people without the tools installed, and is against Debian
packaging policy (symlink pointing outside the source tree).
Signed-off-by: Kees Cook <kees@ubuntu.com>
Acked-by: Steve Beattie <steve@nxnw.org>
This is patch tries to reduce the number of dynamic_cast<>s needed
during normalization by pushing the operations of normalize_tree()
into the expr-tree classes themselves rather than perform it as
an external function. This eliminates the need for dynamic_cast<>
checks on the current object under inspection and reduces the number
of checks needing to be performed on child Nodes as well.
In non-strict benchmarking, doing the dynamic_cast<> reduction
for just the tree normalization operation resulted in a ~10-15%
improvement in overall time on a couple of different hosts (amd64,
armel), as measured against apparmor_parser -Q. Valgrind's callgrind
tool indicated a reduction in the number of calls to dynamic_cast<>
on the tst/simple_tests/vars/dbus_vars_9.sd test profile from ~19
million calls to ~12 million.
In comparisons with dumped expr trees over both the entire
tst/simple_tests/ tree and from 1000 randomly generated profiles via
stress.rb, the generated trees were identical.
Patch history:
v1: initial version of patch
v2: update patch to take into account the infinite loop fix in
trunk rev 1975 and refresh against current code.
v3: no change
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
Enabling the python caching test by default broke the build tests when
running in environments that do not contain the apparmor securityfs
mounted (think build chroots). This is because an initial check from the
shell script version of the tests was not reproduced within the python
version. This patch adds a check in the base class setUp function that
marks each testcase as skipped if apparmor's securityfs cannot be found.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch:
- incorporates the new python caching test into the make check/make
caching target, and removes the older shell based test script
- adjusts the python scripts to give verbose output when the VERBOSE
flag is set
- reorders the tests so that the tests that take a shorter amount of
time to run come first, leaving the language sanity test with its
69000+ testcases last
Patch history:
v1: initial revision
v2: add gen_xtrans/gen_dbus dependency to valgrind test
v3: drop gen_xtrans/gen_dbus as that was committed as a separate fix
Acked-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
With the C++-ization of the parser, some functions were renamed or
eliminated; this patch fixes the relevant valgrind false positive
suppression
pattern to match.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch converts the problematic-with-g++ 4.6 state_names array
into a C++ unordered_map type. Using this depends on using the c++0x
(aka c++11) standard, and as we have gnuisms elsewhere (using the
typeof builtin), the patch also adds/converts to using -std=gnu++c0x
in the build rules (which conveniently eliminates some other warnings
we had due to other c++11-isms).
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-By: Seth Arnold <seth.arnold@canonical.com>
A bug existed in the parser that it would not detect the error case
where an unquoted ']' is given without a matching '[' (the quoted
cases are accepted properly). This patch fixes the issue.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Currently alternations are limited to a single level, make it so we can
nest alternations.
Note: this is a temporary solution to the problem. Long term this routine
to convert to pcre will go away when native parsing of aare is added to
the backend.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
The previous patch for removing libimmunix support from the regression
tests wasn't complete. Also, the 2.2 and 2.4 kernel support code is
closely related and can be removed considering how old those kernels
are.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
This typo allowed apparmor.h to be pulled in multiple times.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
The libapparmor_wrap.c target generates libapparmor_wrap.c and
LibAppArmor.pm. The Perl module must exist before `perl Makefile.PL`
under the Makefile.perl target, otherwise the generated Makefile.perl
ends up with an empty $(TO_INST_PM) variable and the pm_to_blib target's
dependencies are incomplete. That results in the Perl module not getting
copied to the blib directory and a build that is missing LibAppArmor.pm.
A build missing LibAppArmor.pm only occurred while building with
multiple threads.
Thanks to Seth Arnold for the suggestion on how to best fix the
dependencies.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Only attempt to link against libapparmor since libimmunix has been
deprecated for 5+ years.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
The multiarch filesystem layout for Ubuntu uses directories such as
/usr/lib/{i386-linux-gnu,x86_64-linux-gnu,arm-linux-gnueabihf} so
peeking into /usr/{lib,lib64} is no longer sufficient.
This patch uses ldconfig to print out the libraries that it knows about
and grep that output for libapparmor.so or libimmunix.so.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
The parser sanity test make target does not directly depend on the make
targets that generate the tests consumed by the sanity test, leading to
runs that did not verify all the test cases when make check is invoked
with parallelism (e.g. make check -j4). This patch against trunk fixes
the issue.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
dnsmasq needs read access to more files in /var/lib/libvirt/dnsmasq/
(at least *.conf and *.addnhosts)
Since this directory contains only files that are intended for dnsmasq
(also confirmed by Jim Fehlig, the SUSE libvirt maintainer), the best
way is to just allow "/var/lib/libvirt/dnsmasq/* r,"
References: https://bugzilla.novell.com/show_bug.cgi?id=848215
+1'd for trunk and 2.8 by Jamie Strandboge <jamie@canonical.com>
The log parsing in the Immunix::AppArmor perl module has fallen behind
when it comes to audit events from some of the newer rule types
supported by apparmor_parser.
When an unsupported event is found, it causes aa-logprof to error out.
This patch creates a list of valid, but unsupported, event operations
that should be ignored by the perl module when parsing logs.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
When a parser that is aware of dbus rules is running under a kernel
that is unaware of dbus rules, the parser should ignore the dbus rules
instead of attempting to load them into the kernel. Otherwise, the
kernel will reject the entire profile, leaving the application
unconfined.
Similar to what is done for mount rules, the features listed in
apparmorfs should be checked to see if dbus is supported under the
current kernel.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
The purpose is to provide test coverage for accessing UNIX domain socket
files. AppArmor write permissions are needed to create the socket files
and both read and write permissions are needed to connect to the socket.
This patch adds a test to the UNIX file descriptor passing tests and
creates an entirely new set of tests for sending and receiving messages
using path-based SOCK_STREAM, SOCK_DGRAM, and SOCK_SEQPACKET UNIX domain
sockets.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Allow directories to be passed directly to the parser and handled instead
of needing an initscript to find the files in the directory.
eg. load all profiles in profiles dir
apparmor_parser -r /etc/apparmor.d/
eg. load all binary files in the cache dir
apparmor_parser -Br /etc/apparmor.d/cache/
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
This patch adds a python metaclass to wrap the test methods in the
subclasses of the template class AATestTemplate with the keep_on_fail
function, which sets the do_cleanup attribute to False when a testcase
failure occurs (i.e. an Exception is raised), and removes the manually
applied decorators to the caching tests that made use of this.
The downside to this approach is that the way metaclasses are declared
changed between python 2 and python 3 in an incompatible way. Since
python 3 is The Future™, I chose that approach and made the caching
and valgrind tests which use testlib be python3 (until this change,
they would have worked under either python 2 or python 3).
(An output message when a failure occurs is tweaked, to make the
output a little cleaner when verbose test output is requested and
failures occur.)
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
This patch modifies testlib.write_file() to take a directory and a file
name instead of a path and return the joined result for callers to use
if necessary.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
This patch adds the command run to the reported message when a valgrind
failure is detected. This makes reproducing the failure outside of the
test suite easier, for easier diagnosis of what problem is occurring.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
When converting the valgrind tests from optparse to argparse, I managed
to not verify that the resulting code actually worked :( . This patch
fixes it by adding a positional argument to handle the optional passed
directory location.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
[previous commit forgot to bzr add caching.py; this commit fixes that]
This patch rewrites the caching test in python, using python's unittest
framework. It has been used with python 2.7 and python 3.3; python2.6
may have issues. It covers the tests in the existing caching.sh
test script (with the exception of the test that checks for when the
parser in $PATH is newer), as well as adding additional tests that
more extensively cover using a cache in an alternate location from
basedir. It also adds simple tests for the --create-cache-dir option
(along with that option's interaction with the alt-cache option).
(Some further work to be done is listed under TODO.)
Patch history:
v1: - initial version
v2: - create template base class
- add keep_on_fail() decorator to keep temporary test files
around after a test fails
- don't dump raw cache file to failure output in
test_cache_writing_updates_cache_file()
- push run_cmd into template class
- create run_cmd_check wrapper to run_cmd that adds an assertion
check based on whether return code matches the expected rc
(the valgrind tests only want to verify that the rc is not a
specific set of values, hence the separate wrapper function)
- similarly, add a check to run_cmd_check for verifying the output
contains a specific string, also simplifying many of the caching
tests.
- create testlib.write_file() to simplify writing file
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
This patch rewrites the caching test in python, using python's unittest
framework. It has been used with python 2.7 and python 3.3; python2.6
may have issues. It covers the tests in the existing caching.sh
test script (with the exception of the test that checks for when the
parser in $PATH is newer), as well as adding additional tests that
more extensively cover using a cache in an alternate location from
basedir. It also adds simple tests for the --create-cache-dir option
(along with that option's interaction with the alt-cache option).
(Some further work to be done is listed under TODO.)
Patch history:
v1: - initial version
v2: - create template base class
- add keep_on_fail() decorator to keep temporary test files
around after a test fails
- don't dump raw cache file to failure output in
test_cache_writing_updates_cache_file()
- push run_cmd into template class
- create run_cmd_check wrapper to run_cmd that adds an assertion
check based on whether return code matches the expected rc
(the valgrind tests only want to verify that the rc is not a
specific set of values, hence the separate wrapper function)
- similarly, add a check to run_cmd_check for verifying the output
contains a specific string, also simplifying many of the caching
tests.
- create testlib.write_file() to simplify writing file
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
This patch adds a test wrapper that runs valgrind on the parser over the
simple_tests tree (or other directory tree if passed on the command
line). An alternate parser location can also be passed on the command
line.
Like the libapparmor python bindings test, this test uses a bit of magic
to generate tests that doesn't work with auto-detecting test utilities
like nose.
Running valgrind on the parser over all 69000+ testcases takes several
hours, so while this patch includes a make target 'make valgrind', it
does not add it to the set of tests run when 'make check' is called.
Perhaps a 'make extra-tests' target is in order.
Patch history:
v1: - initial version.
v2: - add some valgrind suppressions for overaggressive 4 byte reads
past the end of allocated storage (not completed).
v3: - add ability to dump valgrind suppressions to stdout, to use
diagnosis runs of valgrind for determining whether a given
failure is a false positive or not.
- correctly return 0 on a successful run and an error code if one
or more test cases fail.
- point LD_LIBRARY_PATH at the in-tree libapparmor build.
- split out some utility functions into testlib.py, for possible
use by other to be written test scripts
v4: - convert optparse to argparse
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Tyler Hicks <tyhicks@canonical.com> (for v2 version)
This patch converts to statically linking libapparmor with
whichever static libapparmor it can find on its library search path
(and verified to choose the in-tree version over the system one if both
are available)
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
With trunk commit 2205 "use libapparmor's find mountpoint fn",
the parser now builds against and uses libapparmor at runtime. However,
it currently builds against the system installed libapparmor library and
header files, which fails if either aren't installed, and is thus
painful for bootstrapping in a new environment.
Instead, the parser, like pam_apparmor and mod_apparmor, should build
against the in-tree libapparmor header and library. This patch does
that and adjusts the tests to point LD_LIBRARY_PATH at the location
of the built library as well.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
$HOME/.config/fontconfig/conf.d/* and
$HOME/.config/fontconfig/fonts.conf
/etc/fonts/conf.d/50-user.conf:
<!--
Load per-user customization files where stored on XDG Base Directory
specification compliant places. it should be usually:
$HOME/.config/fontconfig/conf.d
$HOME/.config/fontconfig/fonts.conf
-->
<include ignore_missing="yes" prefix="xdg">fontconfig/conf.d</include>
<include ignore_missing="yes" prefix="xdg">fontconfig/fonts.conf</include>
abstractions/fonts should allow read access to those files:
From: Felix Geyer debfx@ubuntu.com
Acked-by: John Johansen <john.johansen@canonical.com>
cache files will be written out even if the '--skip-bad-cache' option
is given and the cached features file differs from the features of
the currently running kernel. The patch below fixes the regression.
From: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Fix Makefile to rebuild dbus object when yacc definitions change.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
With the conversion to c++, the use of void* pointers for the parser
interface buffers generates several warnings. This patch converts the
types from void* to u8* for the buffer pointers, to clean up those
warnings.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
This patch is a very minor optimization to the search to determine
whether a given rule is an exact match or not. If a wildcard rule
(i.e. an inexact match) is discovered, exact_match is set to 0,
so we don't need to continue the tree traversal.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
This patch fixes a few more parser memory leaks as identified by the
simple valgrind test script. These mostly occur during cleanup of
structs and classes and as such, don't represent very serious leaks
for common usages of the parser.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: John Johansen <john.johansen@canonical.com>
Our simple language tests did not include any file deny rule tests. This
patch adds a few simple ones.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
The README in the parser directory was woefully out of date; this patch
updates the information to contain the current mail list, wiki, and bug
tracking locations.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Christian Boltz <apparmor@cboltz.de>
From: Kshitij Gupta <kgupta8592@gmail.com>
This patch removes:-
2 rules covered by abstractions in smbd profile:
- /var/log/samba/cores/smbd/ rw and /var/log/samba/cores/smbd/** rw
are in abstractions/samba covered by /var/log/samba/cores/** rw
1 superfluous rule:
- /var/lib/samba/printers/** rw is covered by /var/lib/samba/** rwk
Acked-by: Steve Beattie <steve@nxnw.org>
An abstraction to grant the ability to query dconf settings. It does
not grant the ability to update or add settings, due to our current
inability to restrict where within the dconf hierarchy updates
can occur.
From: intrigeri <intrigeri@boum.org>
Acked-by: Steve Beattie <steve@nxnw.org>
From: Kshitij Gupta <kgupta8592@gmail.com>
This patch removes rules covered by abstractions in nscd profile:
- the network rules are in abstractions/nameservice
- @{PROC}/filesystems is in abstractions/base
- /{,var/}run/avahi-daemon/socket is in abstractions/nameservice
- /tmp/.winbindd/pipe and /var/lib/samba/winbindd_privileged/pipe are
in abstractions/winbind via abstractions/nameservice
Acked-by: Steve Beattie <steve@nxnw.org>
covered by abstractions:
- the network rules are in abstractions/nameservice
- /etc/gai.conf is also in abstractions/nameservice
- @{PROC}/sys/kernel/ngroups_max is in abstractions/base
Acked-by: Steve Beattie <steve@nxnw.org>
This patch addresses a bunch of the compiler string conversion warnings
that were introduced with the C++-ification patch.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
Fix this and unify creation and use of cacheloc so that we can hopefully
avoid these bugs.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
--cache-loc= is specified. This results in using the .features file from
/etc/apparmor.d/cache or always recompiling policy.
The former case is particularly bad as the .features file in
/etc/apparmor.d/cache/ may not correspond to the file in the specified
cache location.
bug: launchpad.net/bugs/1229393
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
This is a regression test to load a profile, query it from userspace
using aa_query_label(), and then verify the results.
The query interface is tested by the dbus mediation regression tests,
but this test helps in finding bugs specific to AppArmor, which may
possibly be caused by the parser, kernel, and/or libapparmor.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
AppArmor dbus rules are split into two classes. The first is
(send receive) rules and the second in bind rules. When the parser was
creating its internal representation of dbus rules, it wasn't separating
the overlapping bitmasks for (send receive) perms and bind perms.
(send receive) perms are 0x06 and bind perms are 0x40. Here's the old
parser output for an audit dbus rule that has accept states for
(send receive) and for bind:
$ dbus="/t { audit dbus, }"
$ echo $dbus | apparmor_parser -qQD dfa-states 2>&1 | sed '/^$/,$d'
{1} <== (allow/deny/audit/quiet)
{3} (0x 40/0/40/0)
{7} (0x 46/0/46/0)
The {3} state is the accept state for the bind perms. The {7} state is
the accept state for the (send receive) perms. Note that the bind perm
mask bled over into the (send receive) accept state's mask.
With this patch, the masks for the two accept states do not overlap:
$ echo $dbus | apparmor_parser -qQD dfa-states 2>&1 | sed '/^$/,$d'
{1} <== (allow/deny/audit/quiet)
{3} (0x 40/0/40/0)
{7} (0x 6/0/6/0)
Additionally, this patch makes the rule creation for (send receive)
perms more strict to keep any future perm bits from unintentionally
slipping into the (send receive) accept states.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
Rules that have the audit and deny modifiers are to be explicitly denied
and audited. However, accept states were incorrectly being generated
with the deny and quiet masks set. This resulted in actions being denied
but not audited.
Here's the old parser output for audit deny dbus and mount rules:
$ dbus="/t { audit deny dbus, }"
$ mount="/t { audit deny mount, }"
$ echo $dbus | apparmor_parser -qQD dfa-states 2>&1 | sed '/^$/,$d'
{1} <== (allow/deny/audit/quiet)
{3} (0x 0/40/0/40)
{7} (0x 0/46/0/46)
$ $ echo $mount | apparmor_parser -qQD dfa-states 2>&1 | sed '/^$/,$d'
{1} <== (allow/deny/audit/quiet)
{5} (0x 0/2/0/2)
With this patch, no accept states are generated which means that actions
will be denied and audited:
$ echo $dbus | apparmor_parser -qQD dfa-states 2>&1 | sed '/^$/,$d'
{1} <== (allow/deny/audit/quiet)
$ echo $mount | apparmor_parser -qQD dfa-states 2>&1 | sed '/^$/,$d'
{1} <== (allow/deny/audit/quiet)
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
When using the deny rule modifier, accept states were not being
generated for dbus and mount rules. This means that the actions were
being denied, but it was not possible to quiet the auditing of the
actions.
The problem is that the deny and audit members of the dbus_entry and
mnt_entry structs were being used incorrectly. The deny member is a
boolean, not a bitmask. When the deny modifier is exclusively used in a
rule, the deny boolean should be true and the audit mask should be equal
to the perm mask.
Here's the old parser output for denied dbus and mount rules:
$ dbus="/t { deny dbus, }"
$ mount="/t { deny mount, }"
$ echo $dbus | apparmor_parser -qQD dfa-states 2>&1 | sed '/^$/,$d'
{1} <== (allow/deny/audit/quiet)
$ echo $mount | apparmor_parser -qQD dfa-states 2>&1 | sed '/^$/,$d'
{1} <== (allow/deny/audit/quiet)
With this patch, the accept states are generated correctly with deny and
quiet masks:
$ echo $dbus | apparmor_parser -qQD dfa-states 2>&1 | sed '/^$/,$d'
{1} <== (allow/deny/audit/quiet)
{3} (0x 0/40/0/40)
{7} (0x 0/46/0/46)
$ echo $mount | apparmor_parser -qQD dfa-states 2>&1 | sed '/^$/,$d'
{1} <== (allow/deny/audit/quiet)
{5} (0x 0/2/0/2)
https://launchpad.net/bugs/1226356
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
Convert the codomain to a class, and the policy lists that store
codomains to stl containers instead of glibc twalk.
Signed-off-by: John Johansen <john.johansen@canonical.com>
[tyhicks: Merge with dbus changes and process_file_entries() cleanup]
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
change_hat 1.4 was an experiement is more directly controlling change_hat
by adding hat rulles to the profile. It has not been used since the
original experiment (4 years). So remove it
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
remove old dead code that used to fail compilation if regular expressions
where detected in the rules and the apparmor kernel module did not support
regular expression matching.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
This conversion is nothing more than what is required to get it to
compile. Further improvements will come as the code is refactored.
Unfortunately due to C++ not supporting designated initializers, the auto
generation of af names needed to be reworked, and "netlink" and "unix"
domain socket keywords leaked in. Since these where going to be added in
separate patches I have not bothered to do the extra work to replace them
with a temporary place holder.
Signed-off-by: John Johansen <john.johansen@canonical.com>
[tyhicks: merged with dbus changes and memory leak fixes]
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
From: John Johansen <john.johansen@canonical.com>
let allow be used as a prefix in place of deny. Allow is the default
and is implicit so it is not needed but some user keep tripping over
it, and it makes the language more symmetric
eg.
/foo rw,
allow /foo rw,
deny /foo rw,
Patch history:
v1: - initial revision
v2: - rename yacc target rule from opt_deny to opt_perm_mode to
reflect
that it can be either an allow or deny modifier
- break apart tests into more digestible chunks and to clarify
their purpose
- fix some tests to exercise 'audit allow'
- add negative tests for 'allow' and 'deny' in the same rule
- add support for 'allow' keyword to apparmor.vim
- fix a bug in apparmor.vim to let it recognize multiple
capability entries in a single line.
v3: - add support for optional keywords on capability rules in
regression tests, as well as the bare capability keyword (via
'cap:ALL')
- add allow, deny, and conflicting capability behavioral
regression tests
- fix vim syntax modeline to refer to apparmor in parser tests
- adjust FILE regex in vim syntax file creator script
Signed-off-by: John Johansen <john.johansen@canonical.com>
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
example with LANG=pt_BR) because a regex relies on netstat output.
Enforce LANG=C to make sure aa-unconfined always sees the expected output.
Acked-by: Steve Beattie <steve@nxnw.org>
(The broken URLs were introduced in r1582.)
for utils/*.pod:
Acked-by: Steve Beattie <steve@nxnw.org>
for the other directories:
Patch by Steve Beattie
Acked-by: Christian Boltz <apparmor@cboltz.de>
Patch-Author: Stefan Seyfried <seife+obs@b1-systems.com>
After this change in ntp:
* Mo Aug 19 2013 crrodriguez@opensuse.org
- Build with -DOPENSSL_LOAD_CONF , ntp must respect and use
the system's openssl configuration.
we need to read openssl.cnf or starting of ntpd will fail silently(!)
Patch v2 by Christian Boltz: use abstractions/openssl instead of
allowing /etc/ssl/openssl.cnf directly
Acked-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
The python bindings were using the wrong data type cast (long long
instead of just long) on the value '-1' that is used to indicate no
value for the 'fsuid' and 'ouid' fields in the returned data structure.
Thus a bunch of the tests were failing in 32bit environments.
This patch corrects the issue.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
- Make indenting consistent
- Move common match + fn patterns into a single shared entry with mulitstate
headers
- add names table to convert lexer state #s to state names used in the code
- Create/use macros for common patterns of DUMP, DEBUG, return ...
this fixes a few places where preprocess or DEBUG output was not
available
- update RE patterns for bugs eg. {WS} inside a character class [] does
not match whitespace
all told despite adding code to provide better support to debug and
preprocessing, the code is about 150 lines shorter, and has few corner
cases cleaned up.
Signed-off-by: John Johansen <john.johansen@canonical.com>
[tyhicks: Added small fix for D-Bus parsing bug]
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Removes an unnecessary variable, simplifies and unifies some of the loop
logic, and removes commented out code.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
When merging file entries in process_file_entries(), an error condition
can leave the entries list in a bad state which can cause invalid reads
and/or double frees when freeing the codomain and entries list memory.
The problem comes from the need to sort the entries linked list. An
array of pointers is created to represent the linked list, then the
array is sorted, then the linked list and the array coexist while the
entries are merged, then the linked list is reconstructed and the array
is freed. While the entries are being merged, an error condition can
occur and the function can return while the linked list is partially
modified.
The solution is to complete the sorting, reconstruct the linked list,
and free the array immediately. Once the linked list is in a good state,
the entries can be merged. Care is taken to adjust the linked list
pointers as entries are merged. An error condition can occur but the
linked list is always in a good state and proper cleanup can be
performed without any memory access issues.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
If a parser test case causes the parser to produce a core dump, the
simple.pl test runner incorrectly treats the test as a success.
This patch treats tests that cause core dumps as failures, even when the
tests are marked as #=TODO. The only way to ignore tests that fail in
this manner is with #=DISABLED.
Note that this patch changes the meaning of the $result variable.
Previously, it held a true or false status returned when closing the
apparmor_parser pipe. Now, it holds the exit status of apparmor_parser.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
This patch fixes a few memory leaks found by valgrind. Most of these
occur in error cases and as such, are not a big deal. The dbus TOK_MODE
and flags TOK_CONDID leaks in parser_yacc.y are legitimate leaks,
if of very small amounts of memory.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
This patch adds tests for the swig generated python library bindings
that reuse the C language tests.
Fitting it into autotools was a bit of a trick, and is likely pretty
brittle, as before the test script runs, it needs to know the location
of the built libapparmor.so library, the built _LibAppArmor.so library
and the python wrapper bits (thankfully, the latter two are the same
directory). It's also unclear how to get autotools to emit the output of
the test_python.py script when building, rather than just summarizing it
as one test run.
Also note that test_python.py is doing a bit of magic to automatically
generate test case methods based on the contents of the test_multi/
directory. This has the disadvantage of breaking tools like nosetests
and other external tools that try to automatically detect testcases.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
This patch gives a more pythonish whitespace cleanup to the swig python
setup.py.in configuration file. It also updates the wiki url.
(That said, pep8 will still probably barf all over it.)
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
This patch converts the C test program to only emit the basename(3) of
the test input file under consideration, rather than the entire path as
passed on the command line, and fixes up all the expected outputs to
match.
The reason to do this is to make it easier for other tools located
in other directories (e.g. under libapparmor/swig) to use these same
test cases with reduced special casing.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
This patch wraps the event record output cases in a macro, for
consistent generation.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
This patch converts most of the fields to using the existing macros for
output, to make consistent and simplify the code a bit.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
The test program didn't make use of the existing print_long() macro for
printing long values, which meant that they were always emitted for
every testcase. This patch makes them consistent with all the other
emitted fields and fixes up the expected output where they shouldn't be
emitted.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
The oldstyle name configure.in has been deprecated by autotools. This
patch renames configure.in to the new standard name configure.ac.
The AC_INIT() entry was adjusted as well to point to configure.ac
instead of configure.in.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
This patch makes the binding check messages a little more understandable
and coherent. The original messages looked like (for a perl binding only
build):
checking Checking for Python... no
checking Checking for perl... yes
checking for perl... /usr/bin/perl
Note the duplicated 'checking Checking' and '[cC]hecking for perl'
statements. With the patch applied, the output looks like thus:
checking whether python bindings are enabled... no
checking whether perl bindings are enabled... yes
checking for perl... /usr/bin/perl
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
This patch causes libapparmor's configure script to exit with an error
if a language binding is asked for and the relevant interpreter is not
found. The previous behavior was to *silently* disable the binding.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
This patch causes libapparmor's configure script to exit with an error
if any of the language bindings are requested but swig has not been
found earlier in the configure script. Without this script, configure
would bury the inability to find swig in its output, without informing
the user that building any of the language bindings would fail.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
The function new_dbus_entry() free()s the conds argument but not the
peer_conds argument.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
This patch moves the DUP_STRING macro to parser.h and modifies
it to accept a goto error target, that will be jumped to if the
call to strdup(3) fails. It also uses it in additional locations
where copying structures occurs, as well as detecting additional
cases where a structure duplication might have failed but not been
propagated outward.
Signed-off-by: Steve Beattie <steve@nxnw.org>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
The aa_getcon man page only implies that the *mode strings returned by
aa_getprocattr(), aa_gettaskcon(), aa_getcon(), and aa_getpeercon()
should not be freed. A developer using the man page to build against
libapparmor may miss that subtlety and end up hitting double free issues.
This patch makes the man page more clear, makes the function comments
more clear, and changes the aa_getprocattr() *buf param to *con. The use
of *buf should reserved for the aa_get*_raw() functions that do not
allocate a buffer for the confinement context and all documents now
clearly mention that *con must be freed.
Additionally, this patch removes the line wrapping of the
aa_getprocattr_raw() prototype in the aa_getcon man page source. The
line wrapping caused incorrect formatting of the function prototype when
viewing the man page.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
r2125 caused a regression in aa_getpeercon_raw() when a NULL pointer was
passed into the mode parameter. Instead of unconditionally
NUL-terminating the con string before the mode portion of the security
context, it made it to where the NUL byte was only put into place when
mode was non-NULL.
This resulted in the con string incorrectly containing the label and the
mode.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Bug: https://bugs.launchpad.net/bugs/1218099
This patch adds support for expanding variables with dbus rules.
Specifically, they can expanded within the bus, name, path, member,
interface, and peer label fields.
Parser test cases and regression test cases are added as well.
Patch history:
v1: initial version of patch
v2: add equality.sh tests to verify that the results of using
variable expansion is the same as what should be equivalent rules
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
In aa_query_label(), errors encountered during a write() to the AppArmor
filesystem's .access file results in an unintentional file descriptor
leak outside of aa_query_label(). Callers don't expect aa_query_label()
to return with a newly opened file descriptor so they can't be expected
to close the fd.
This flaw was introduced in r2147, which has not yet been included in an
official release.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <sbeattie@ubuntu.com>
The profile parsing in the Immunix::AppArmor perl module has fallen
behind when it comes to some of the newer rule types and syntax
supported by apparmor_parser.
When an unsupported rule is found, it causes aa-logprof and aa-genprof
to error out. This patch creates a list of valid, but unsupported rule
types that should be ignored by the perl module when parsing policy.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
When dnsmasq is started with the --enable-dbus option, it uses the
system bus.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Steve Beattie <sbeattie@ubuntu.com>
Integrate dbus tests into the regression testing framework.
This started out as dbus-send.c, from the dbus source, and then grew
from there.
dbus_message is an example "client" program that only sends out
messages. dbus_service binds to a well-known name and then listens and
responds to incoming messages. They share some code in dbus_common.c.
The test scripts, dbus_message.sh and dbus_service.sh, share some
functionality in dbus.inc.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
It may be useful to applications that do AppArmor queries to know if the
subject label in the query is unknown to the kernel. For example, the
corresponding profile may have been removed/renamed.
This patch eliminates all potential return locations of aa_query_label()
that may have errno set to ENOENT, except for the write() to
apparmorfs/.access that sets ENOENT when the subject label isn't found
by the kernel.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Test a set of send, bind, and receive denials routed through syslog,
as well as a set routed through auditd.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-By: Seth Arnold <seth.arnold@canonical.com>
Add a new set of tests that tests delegation of file descriptors when
inherited across combinations of confined and unconfined processes.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch broadens the testing of file descriptor passing over Unix
domain sockets, but the real focus is on passing a file descriptor from
an unconfined server to a confined client. The confined client should
have full access to the file descriptor, despite not having a
corresponding file rule in its profile, due to delegation.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
required for a test to run. This will help keep the regression suite
from reporting a lot of failures when it hits a kernel that doesn't
support the feature being tested.
its current iteration is pretty brain dead, only testing for the
existance of feature files/dirs (and not contents) but I think it is
probably sufficient for now.
To use it, just call the required_features fn right after sourcing
prologue.inc in the bash test script that is called by the make file.
If any of the requested features don't exist the bash script will exit
reporting the first feature that was missing
Eg.
. $bin/prologue.inc
required_features dbus
required_features dbus dbus/mask network domain/change_hatv
Acked-by: Tyler Hicks <tyhicks@canonical.com>
Allows for the test script to specify a final check to be performed
after checking the output of the test binary. This may be useful, for
example, if the test script wants to compare logging output of the test
binary to known-good logging output.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
Add an interface for trusted applications to use when they need to query
AppArmor kernel policy to determine if an action should be allowed.
This is a simplified interface that tries to make it as easy as possible
for applications to use. They provide a permissions mask and query
string and they get a pair of booleans back that let them know if the
action should be allowed and/or audited.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This test is to verify that a list of profiles compile down into the
same binary representation. This is useful, for example, when testing a
rule syntax that includes permission aliases, as well as implied and
explicit accesses.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This is a test in the style of gen-xtrans.pl that attempts to run
through the most commonly constructed DBus rules. It also attempts to
run through some common mistakes to ensure that the parser fails
appropriately.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch implements the parsing of DBus rules.
It attempts to catch all corner cases, such as specifying a bind
permission with an interface conditional or specifying a subject name
conditional and a peer name conditional in the same rule.
It introduces the concept of conditional lists to the lexer and parser
in order to handle 'peer=(label=/usr/bin/foo name=com.foo.bar)', since
the existing list support in the lexer only supports a list of values.
The DBus rules are encoded as follows:
bus,name<bind_perm>,peer_label,path,interface,member<rw_perms>
Bind rules stop matching at name<bind_perm>. Note that name is used for
the subject name in bind rules and the peer name in rw rules. The
function new_dbus_entry() is what does the proper sanitization to make
sure that if a name conditional is specified, that it is the subject
name in the case of a bind rule or that it is the peer name in the case
of a rw rule.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
This patch fixes problems in the handling of both the final cache
name location and the temporary cache file when an alternate location
is specified.
The first issue is that if the alternate cache directory location was
specified, the alternate directory name would be used as the final location for
the cache file, rather than the alternate directory + the basename of
the profile.
The second issue is that it would generate the temporary file that it
stores the cache file in [basedir]/cache even if an alternate cache
location was specified on the command line. This causes a problem
if [basedir]/cache is on a separate device than the alternate cache
location, because the rename() of the tempfile into the final location
would fail (which the parser would not check the return code of).
This patch fixes the above by incorporating the basename into the cache
file name if the alternate cache location has been specified, bases the
temporary cache file name on the destination cache name (such that they
end up in the same directory), and finally detects if the rename fails
and unlinks the temporary file if that happens (rather than leave it
around). It also has been updated to add a couple of testcases to verify
that writing and reading from an alternate cache location work.
Patch history:
v1: first draft of patch
v2: add testcases, convert PERROR() to pwarn() if rename() fails for
placing cachefile into place.
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-by: John Johansen <john.johansen@canonical.com>
Debian sid's fonts-mathjax ships fonts in
/usr/share/javascript/mathjax/fonts, that are now used by default by
fontconfig-enabled software.
Acked-by: Seth Arnold <seth.arnold@canonical.com>
At least on Debian, with recent versions of fontconfig-config
(>= 2.10), files in /etc/fonts/conf.d/ are symlinks pointing to
/usr/share/fontconfig/.
This was reported by Jakub Wilk <jwilk@debian.org> on Debian bug #714843.
Acked-by: Seth Arnold <seth.arnold@canonical.com>
abstractions/ubuntu-browsers.d/ubuntu-integration.
Patch by Felix Geyer <debfx@ubuntu.com>
Acked-by: Seth Arnold <seth.arnold at canonical.com>
Acked-by: Christian Boltz <apparmor@cboltz.de>
Using --subdomainfs without an argument triggers a segfault. This was due
to the long option missing the "has_arg" flag.
Signed-off-by: Kees Cook <kees@ubuntu.com>
Acked-by: John Johansen <john.johansen@canonical.com>
Add aa_getpeercon_raw() to the man page and adjust aa_getpeercon()
prototype to include the new mode parameter.
Also, explain the significance of ERANGE for aa_getpeercon_raw() and fix
a misspelling in the meaning of ERANGE.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
The parameter names are slightly different in the two functions. Rename
buffer to buf and rename size to len to make the two function prototypes
look similar.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
functions
The functions that return the confinement information of a peer socket
connection should parse and return the mode like the task-based
functions.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
The getpeercon functions need to parse the mode from the confinement
string. This patch creates a function that aa_getpeercon_raw() and
aa_getprocattr_raw() can both use.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Protect against bugs in AppArmor's getsockopt() LSM hook from sending
aa_getpeercon() into an infinite loop.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
file is larger than the feature buffer used for cache version comparison.
Ideally this would be dynamically allocated but for 2.8 just bumping the
buffer size is the quick fix.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <sbeattie@ubuntu.com>
1) make sure that the xpra socket exists before trying to attach to it
2) make sure that the client has attached before we start the application
The fix for '1' solves a problem when the system is under load and the
one for '2' fixes a problem with firefox starting too soon and not
having system themes applied.
Exercising the 1 week rule. Seth Arnold commented on the added sleeps and I
adjusted one based on his comments and replied to the list that the other is
needed and that this improves the sandbox/xpra code but that there are
limitations with driving xpra.
Acked-By: Jamie Strandboge <jamie@canonical.com>
Author: Dmitrijs Ledkovs <dmitrij.ledkov@ubuntu.com>
Modifiy the libapparmor macro for python to use python-config if it
exists to determine what CPPFLAGS and LDFLAGS to use when building
the python swig libraries. Without this addition, python detection
fails on ubuntu 13.04. I've confirmed that with this patch applied,
the python libraries still build successfully on older releases as well
(as far back as ubuntu 11.10).
Acked-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-by: Kees Cook <kees@ubuntu.com>
messages neglected to include the empty .err files in the testsute
directory, resulting in ERROR output. These files were included in the
patch submitted to the mail list. This commit adds them.
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
= How it works =
There are basically two modes:
1. using an existing profile with --profile
2. dynamically generating a profile
For '1', aa-sandbox is just a wrapper around aa-exec.
For '2', aa-sandbox leverages easyprof and allows you to specify policy
in a limited way on the command line. It then loads the policy into the
kernel as a profile (ie, 'profile <foo> { ... }') so it doesn't get in
the way of existing profiles. It currently calls apparmor_parser via
sudo or pkexec. Once the profile is loaded, aa-exec the application
under the profile.
When -X is specified, the application is launched inside its own X
server using either xpra (the default, which uses Xvfb), xephyr and
xpra3d (xpra, but using Xorg with the xdummy[1] driver for now[2].
xpra3d doesn't currently perform well, but works ok with newer Gnome
applications that now require GLX). When using '-X', it:
- adds an explicit deny rule for ~/.Xauthority
- generates a dynamic Xauthority file for the session in
~/.Xauthority-sandbox<DISPLAYNUMBER>
- adds an allow rule for ~/.Xauthority-sandbox<DISPLAYNUMBER>
- adds checks for xhost being properly setup
- honors the --with-xauthority option which can be used with --profile
With the above, the :0.0 display should no longer be accessible. Eg:
$ ./aa-sandbox -t ~/sandbox-xterm -X /usr/bin/xterm
$ XAUTHORITY=~/.Xauthority DISPLAY=:0.0 xinput
No protocol specified
Unable to connect to X server
This requires a specifically configured xauth/xhost setup, which is less common
on modern distributions. The man page details how to get this setup.
= Trying it out =
Apply the patch, then:
$ cd ./utils
# cli
$ ./aa-sandbox --templates-dir=`pwd`/easyprof/templates --read-path=/proc/ /usr/bin/uptime
# 2d only
$ ./aa-sandbox --templates-dir=`pwd`/easyprof/templates -X /usr/bin/xeyes
$ ./aa-sandbox --templates-dir=`pwd`/easyprof/templates -X /usr/bin/gedit
# 2d alternate (xephyr)
$ ./aa-sandbox --templates-dir=`pwd`/easyprof/templates -X --with-xserver=xephyr /usr/bin/xeyes
$ ./aa-sandbox --templates-dir=`pwd`/easyprof/templates -X --with-xserver=xephyr /usr/bin/gedit
# 3d
$ ./aa-sandbox --templates-dir=`pwd`/easyprof/templates -X --with-xserver=xpra3d /usr/bin/xeyes
$ ./aa-sandbox --templates-dir=`pwd`/easyprof/templates -X --with-xserver=xpra3d /usr/bin/glxgears
# With an existing profile:
$ ./aa-sandbox --profile=/usr/bin/evolution -X --with-xserver=xpra3d /usr/bin/evolution
= The Patch =
The patch itself is pretty self contained:
utils/aa-easyprof:
- adjusted to import optparse
utils/easyprof/templates/sandbox*
- add two new templates to easyprof
utils/apparmor/easyprof.py:
- use 'profile <foo>' if '<foo>' is not an absolute path
- adjust parser handling so we can reuse it
utils/aa-sandbox:
- small script to drive utils/apparmor/sandbox.py
utils/apparmor/common.py:
- the start of our python library. aa-easyprof would eventually use
this (along with the various rewrites), but for now, only the
sandboxing uses it.
utils/apparmor/sandbox.py:
- the sandboxing code itself. Of particular note is the use of classing
to support different X servers
utils/aa-sandbox.pod:
- the corresponding man page
= Improvements =
* don't use sudo
* make pulseaudio in xpra opt-in (currently it is off)
* take advantage of upstream's 3D patches when they stabilize
* investigate how applications can work with the Unity global menu
* surely lots more
[1]http://xpra.org/Xdummy.html
[2]http://xpra.org/trac/ticket/147
abstractions/mysql contains
/var/lib/mysql/mysql.sock rw,
/usr/share/mysql/charsets/ r,
/usr/share/mysql/charsets/*.xml r,
but the files moved (at least on openSUSE) to
/usr/share/mysql-community-server/charsets/*.xml
/var/run/mysql/mysql.sock
This causes denials for all applications using MySQL on 12.2 and
Factory.
MariaDB has the *.xml files in
/usr/share/mariadb/charsets/*.xml
and also seems to use /var/run/mysql/ for the socket.
Since MariaDB is basically a drop-in replacement for MySQL, it makes
sense to allow access to it via abstractions/mysql.
References: https://bugzilla.novell.com/show_bug.cgi?id=798183
Acked-by: Seth Arnold <seth.arnold@canonical.com>
The libimmunix library is a historical artifact and has generated a
deprecation warning when used to syslog for over 4 years. This patch
removes it entirely from the libapparmor tree.
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-by: John Johansen <john.johansen@canonical.com>
I was testing out a profile for pulseaudio and hit an issue where my
pulseaudio process was getting the firefox profile applied to it. This
is because in abstractions/ubuntu-browsers.d/multimedia the rule for
pulseaudio is /usr/bin/pulseaudio ixr; attached is a patch to change it
to Pixr, so as to use a global pulseaudio policy if it exists.
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-by: John Johansen <john.johansen@canonical.com>
time formats.
currently the only supported format is
<Month> ## hh:mm:ss
extend this to
<Month> ## hh:mm:ss(.ms)?((+|-)timezone)?
yyyy-mm-dd hh:mm:ss(.ms)?((+|-)timezone)?
yyyy-mm-ddThh:mm:ss(.ms)?((+|-)timezone)?
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <sbeattie@ubuntu.com>
From: Simon Deziel <simon.deziel@gmail.com>
A fair number of the rules that apply to files in @{HOME} predate the
existence of the 'owner' qualifier. This patch adds the 'owner'
qualifier in several places.
Acked-by: Steve Beattie <sbeattie@ubuntu.com>
target in the profiles Makefile, for future archaeological spelunking.
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-by: Christian Boltz <apparmor@cboltz.de>
This patch adds the kernelvars tunable to the global set that is usually
included by default in apparmor policies. It then converts the rules
that are intended to match /proc/pid to use this tunable.
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-By: Seth Arnold <seth.arnold@canonical.com>
This patch finishes the conversion from /proc to the @{PROC}
tunable within profiles and abstractions. It also adjusts some of
the /proc/*/something usages to @{PROC}/[0-9]*/something to restrict
things to just the /proc/pid directories. (A followup patch will
convert these to use @{pid} from the kernelvars tunable.)
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-By: Jamie Strandboge <jamie@canonical.com>
In testing the skype profile, I found access to my @{HOME}/.XCompose
was being rejected. This patch updates the X abstraction to take a
user's defined XCompose key shortcuts into account.
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-By: Jamie Strandboge <jamie@canonical.com>
Author: Jamie Strandboge <jamie@canonical.com>
Bug-Ubuntu: https://launchpad.net/bugs/933440 Forwarded: yes
This is a very slightly updated version of the skype profile
update that Jamie Strandboge submitted, but did not get a review.
The only addition over the previously submitted version is rw access
to @{HOME}/.config/Skype/Skype.conf.
(This commit incorporates the additional @{HOME}/.kde4 change proposed
by Christian Boltz <apparmor@cboltz.de>)
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-By: Jamie Strandboge <jamie@canonical.com>
The apparmor_api abstractions make the mistake of including tunables
directly, which is a no-no since the variable definitions in tunables
need to occur in the preamble of a profile, not embedded within it.
This patch removes those includes, and replaces them documentation of
tunables are necessary, as some of the expected ones are not part of
tunables/global.
It also adjust the kernelvars tunable's definition of the @{pid}
regex, as the current parser does not support nesting of {} groupings,
which breaks any profile that attempts to use the tunable.
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-By: Seth Arnold <seth.arnold@canonical.com>
This patch modifies the nvidia abstraction to add the livdpau wrapper
config file for nvidia workarounds. It also converts the /proc/
rules to use the @{PROC} tunable. And finally, it converts the
ubuntu-browsers.d/multimedia abstraction to use the nvidia abstraction.
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-By: Jamie Strandboge <jamie@canonical.com>
This patch separates out make check in the profiles/ directory into
two sub targets, for checking profiles against the built parser
and aa-logprof respectively. The logprof check currently makes some
assumptions about the environment that make it difficult to run in
a minimal chroot environment.
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-By: Jamie Strandboge <jamie@canonical.com>
When I corrected the profiles/Makefile to automatically find files to
install, I converted one variable name but missed a later location where
that variable was used, which broke the 'make check' target, because
directories would be handed to the apparmor parser. This patch corrects
that and also makes the VERBOSE flag report each profile name as it's
being handed to the parser.
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-By: Jamie Strandboge <jamie@canonical.com>
Acked-By: Seth Arnold <seth.arnold@canonical.com>
fix a nasty little bug that can surface in apparmor 2.8 when
Hats/children profiles are used.
the matchflags in the dfa backend are not getting properly reset, which
results in a previously processed profiles match flags being used. This is
not a problem for most permissions but can result in x conflict errors.
Note: this should not result in profiles with the wrong x transitions loaded
as it causes compilation to file with an x conflict.
This is a minimal patch targeted at the 2.8 release. As such I have just
updated the delete_ruleset routine to clear the flags as it is already
being properly called for every rule set.
Apparmor 2.9/3.0 will have a different approach where it is not possible
to reuse the flags.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Steve Beattie <sbeattie@ubuntu.com>
chance to run before verifying it's current and future confinement
state. In testing the combined sleeps added roughly a second to
onexec.sh's total time on relatively reasonable hardware.
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-By: John Johansen <john.johansen@canonical.com> (via IRC)
whether or not the filesystem has a fine enough timestamp resolution.
Occasionally even on filesystems like ext3, the two files' creation
dates would differ when created less than a second apart, which would
typically cause the 'Cache is used when cache is newer' test to fail
because the cached file would have the same timestamp as the profile.
The fix creates 10 files 0.1 seconds apart and ensures that all ten
have distinct timestamps.
(The occasional failure was caught in testing runs like
https://bugs.launchpad.net/qa-regression-testing/+bug/1087061/ )
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-by: John Johansen <john.johansen@canonical.com>
chance to run before verifying it's current and future confinement
state. In testing the combined sleeps added roughly a second to
onexec.sh's total time on relatively reasonable hardware.
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-By: John Johansen <john.johansen@canonical.com> (via IRC)
to the apparmor_api subtree not getting added in the Makefile. Rather
Rather than require every sub-directory that gets added to be
enumerated, it uses find to determine what directories and files to
install, to avoid future breakage. It is admittedly slower than the
original code because install(1) is being invoked for every file in
the apparmor.d tree, rather than acting on wildcard globs. That said,
I think it's an acceptable tradeoff.
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-by: John Johansen <john.johansen@canonical.com>
mediation, file picker, etc.), making it easier for other source bases
to detect the presence of libapparmor would be beneficial. This patch
adds pkg-config support to the build infrastructure for libapparmor.
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-by: Kees Cook <kees@ubuntu.com>
I added this profile to the openSUSE apparmor-profiles package in Feb 2012.
Until now I didn't receive any bugreports so I'd say it's complete ;-)
References: https://bugzilla.novell.com/show_bug.cgi?id=748499
Acked-By: jdstrand (on IRC)
This patch adds a test script/driver for the aa-decode utility. The only
change from the previous versions is to support overriding the location
of the aa-decode to test via the APPARMOR_DECODE environment variable
and documenting the utils/ tests in the top level README.
The aa-decode test can be run directly from the commandline in the utils
directory like so:
test/test-aa-decode.py -v
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-By: Christian Boltz <apparmor@cboltz.de>
This patch fixes two issue with the simple test driver. The first is
that child exec that actually ran the parser was located inside the
eval statement. This meant that if the exec failed for some reason
(like the parser didn't exist), the child wouldn't actually die,
but would pop out of the eval and continue running through the loop
of test profiles (while the parent process does the same). This meant
that if the script ran on the full testsuite with a misconfiguration,
it would explode creating O(n^2) processes, where n is the number of
testcase files -- with over 25k testcases, that's a lot. The fis is to
lift the child exec outside the eval{}, then an exec() failure causes
the child process to die correctly.
The second fix is that several of the testcases were added with the
DESCRIPTION field added in lower case (i.e. #=Description blah blah).
This fix makes the regex that pulls out the description not be
case-sensitive.
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-By: John Johansen <john.johansen@canonical.com>
This patch replaces the hardcoded path to the in-tree apparmor parser
in several of the script based test scripts with the APPARMOR_PARSER
environment variable, keeping the hardcoded location as the default.
It also adds support for overriding the location of the parser via the
same environment variable. The make infrastructure is updated to use
this, though uses a different variable (PARSER) to drive it.
Thus 'make check PARSER=/some/path/to/an/alternate/apparmor_parser'
will run all the parser tests on that binary. This is useful for
running the testsuite in an automated post-install environment.
(It should be noted that doing so will still build and run the unit
test binaries based on the source tree.)
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-By: John Johansen <john.johansen@canonical.com>
It turns out that PAGE_SIZE isn't defined on all architectures.
This fixes a regression test failure happening on Ubuntu quantal
on the arm ti-omap4 architecture.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Handling stdin was totally broken (= no output) with the current log
format because aa-decode expected name= to be the last entry in the
log line.
This patch for stdin handling
- fixes the pattern to match the current log format (name= is NOT the
last part in the log entry)
- uses bash replacement to avoid some sed calls (which also means the
script now needs an explicit "#!/bin/bash")
- prints decoded filenames in double instead of single quotes to be
consistent with filenames that were not encoded
- also prints lines that do not contain an encoded filename (instead of
grepping them away)
- replace tr calls by perl's uc() (also for non-stdin mode)
- also handle encoded profile names (introduced by Steve)
- don't fail if a file or profile name contains a '
In other words: you can pipe your audit.log through aa-decode, and the
only difference to the raw audit.log is that filenames are decoded.
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
- It failed to remove coredump files named "core"
- It failed to properly detect "core.<pid>" files
- And it would fail if the coredump_pattern had been modified to
a different location.
This lead one of the tests to report it was passing when it
wasn't because it was detecting the previous tests core file.
- Fix the test to set the coredump_pattern, to dump into the
tmpdir used for the test.
- Make it so it will only detect the core file for the pid of
the last test run.
- And extend the test to have a couple of extra test cases.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
/etc/apparmor/profiles/extras/, and update the path at various places.
Also update the mailinglist address in extra-profiles README and
recommend cp instead of mv.
Note: if you want to have a symlink
/etc/apparmor/profiles/extras -> /usr/share/apparmor/extra-profiles/
for backward compability, you'll have to create it yourself (for example
in the .spec file)
This also fixes https://bugzilla.novell.com/show_bug.cgi?id=713647
Acked-by: John Johansen <john.johansen@canonical.com>
@{HOME}/.Xauthority
utils/apparmor/sandbox.py: verify the above rule is any any dynamic templates
that use -X
utils/aa-sandbox.pod: update man page to warn about /.Xauthority access
- use signal.<signal> instead of hardcoding a number
- add --with-xauthority option
- remove '-r' and '--with-geometry' and use --with-xephyr-geometry instead
- allow passing arguments to the application when using aa-exec
- kill with SIGTERM, then try again with SIGKILL
- always use os.execv() in forks. Using cmd() when not specifying '-d' created
different behaviors between debug and non-debug mode
- better cleanup Xpra when aa-exec command fails
- use the full dummy.xorg.conf, which gives us the correct modelines for large
displays. This fixes the issue "Server's virtual screen is too small .... You
may see strange behavior." which should up when the window's size was bigger
than the 'current server resolution'
utils/apparmor/common.py: adjust for python3 (ie, make bi-lingual)
utils/apparmor/sandbox.py:
- set reasonable default template
- gen_policy_name() uses full pathname
- adjust for python3
The apparmor_parser has 3 different directory walking routines. Abstract
them out and use a single common routine.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
/usr/share/poppler/cMap/**. These files are included in the poppler-data
package on Ubuntu, and their 'r' denials create quite a bit of noise.
Apparently they are needed to display PDF documents containing CJK
characters with libpoppler. I added it to the gnome abstraction because
several applications not linked against poppler are consulting this
data.
Acked-By: Jamie Strandboge <jamie@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
/usr/share/poppler/cMap/**. These files are included in the poppler-data
package on Ubuntu, and their 'r' denials create quite a bit of noise.
Apparently they are needed to display PDF documents containing CJK
characters with libpoppler. I added it to the gnome abstraction because
several applications not linked against poppler are consulting this
data.
Acked-By: Jamie Strandboge <jamie@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Fix the apparmor_parsers -N command (which dumps the list of profile
names found in a policy file) to be available without privilege and
also make it be recognized as a command instead of an option so that
it can conflict with -a -r -R -S and -o.
Currently it can be specified with these commands but will cause the
parser to short circuit just dumping the names and not doing the actual
profile compile or load.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Rework and update the apparmor_parser man page. It reworks some of the
text but mostly just reorganizes the commands and options into logical
grouping to make it easier to sort out how the various commands and
options work.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Add the ability to clear out the binary profile cache. This removes the
need to have a separate script to handle the logic of checking and
removing the cache if it is out of date.
The parser already does all the checking to determine cache validity
so it makes sense to allow the parser to clear out inconsistent cache
when it has been instructed to update the cache.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
serious flaw. The test for the network flag was being applied against both
the kernel flags and the cache flags. This means that if either the kernel
or the cache did not have the flag set then network mediation would be
turned off.
Thus if a kernel was booted without the flag, and a cache was generated
based on that kernel and then the system was rebooted into a kernel with
the network flag present, the parser on generating the new policy would
detect the old cache did not support network and turn it off for the
new policy as well.
This can be fixed by either removing the old cache first or regenerating
the cache twice. As the first generation will write that networking is
supported in the cache (even though the policy will have it disabled), and
the second generation will generate the correct policy.
The following patch moves the test so that it is only applied to the kernel
flags set.
Signed-off-by: John Johansen <john.johansen@canonical.com>
compatibility interface. Previously it was assuming that if the compatibility
interface was present that network rules where also present, this is not
necessarily true and causes apparmor to break when only the compatibility
patch is applied.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
weren't being processed. It ultimately boiled down to a kernel issue
but I found it useful to see what the parser thought it was working
with. Since the parser already has a debugging mode that will show things
like capabilities, it was an obvious extension to add network rules.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Acked-by: John Johansen <john.johansen@canonical.com>
This patch moves the generation of file rules from apparmor.vim.in to
create-apparmor.vim.py. It also adds support for
- filenames in quotes
- reverse syntax (permissions first)
The patch also removes an outdated $Id header in apparmor.vim.in and
updates the copyright year.
Acked-By: John Johansen <john.johansen@canonical.com>
- make table of contents, footnotes etc. clickable hyperlinks
- use timestamp of techdoc.tex (instead of build time) as creationdate
in the PDF metadata
- don't include build date on first page of the PDF
- make clean:
- delete techdoc.out (created by pdftex)
- fix deletion of techdoc.txt (was techdo_r_.txt)
The initial target was to get reproduceable PDF builds (therefore the
timestamp-related changes), the other things came up during discussing
this patch with David Haller.
The only remaining difference in the PDF from build to build is the /ID
line. This line can't be controlled in pdflatex and is now filtered
out by build-compare in the openSUSE build service (bnc#760867).
Credits go to David Haller for writing large parts of this patch
(but he didn't notice the techdo_r_.txt ;-)
Signed-Off-By: Christian Boltz <apparmor@cboltz.de>
* the application, library, documentation and installation script
* the initial templates and policy groups. This will undoubtedly need
refinement as we get feedback from users. Initial policy is based on Ubuntu's
Application Review Board (ARB) requirements[2].
* tests for the library
* Makefile integration
Templates are stored in /usr/share/apparmor/easyprof/templates and policy
groups in /usr/share/apparmor/easyprof/policygroups. This can be adjusted via
/etc/apparmor/easyprof.conf.
The aa-easyprof.pod has complete documentation on usage with some
additional information in utils/easyprof/README (mostly duplicated
here).
Testing can be performed in a number of ways:
$ cd utils ; make check # runs unit tests and pyflakes
Unit tests manually:
$ ./test/test-aa-easyprof.py
In source manual testing:
$ ./aa-easyprof --templates-dir=./easyprof/templates \
--policy-groups-dir=./easyprof/policygroups \
... \
/opt/foo/bin/foo
Post-install manual testing:
$ make DESTDIR=/tmp/test PERLDIR=/tmp/test/usr/share/perl5/Immunix install
$ cd /tmp/test
$ PYTHONPATH=/tmp/test/usr/local/.../dist-packages ./usr/bin/aa-easyprof \
--templates-dir=/tmp/test/usr/share/apparmor/easyprof/templates \
--policy-groups-dir=/tmp/test/usr/share/apparmor/easyprof/policygroups \
/opt/bin/foo
(you may also adjust /tmp/test/etc/apparmor/easyprof.conf to avoid
specifying --templates-dir and --policy-groups-dir).
Committing this now based on conversation with John and Steve.
Acked-By: Jamie Strandboge <jamie@canonical.com>
the chromium and chrome sandboxes are setuid root, they only link in limited
libraries so glibc's secure execution should be enough to not require the
santized_helper (ie, LD_PRELOAD will only use standard system paths (man
ld.so)). Also allow some paths in /opt for Chrome.
Ubuntu-Bug: https://launchpad.net/bugs/964510
Acked-By: Jamie Strandboge <jamie@canonical.com>
sanitized_helper. For now this only allows software-center scripts in
/usr/share, but we may need to increase what is allowed in /usr/share if more
things are denied when they shouldn't be.
Ubuntu-Bug: https://launchpad.net/bugs/972367
Acked-By: Jamie Strandboge <jamie@canonical.com>
$CPPFLAGS. Additionally, do not repeat compiler flags for automake
targets that already include them, and pass more flags to the Perl build.
Signed-off-by: Kees Cook <kees@ubuntu.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
When writing out a profile, aa-logprof incorrectly converts PUx execute
permission modes to the syntactically invalid UPx mode, because the
function that converts the internal representation of permissions to
a string emits the U(nconfined) mode bit before the P bit.
This patch corrects this by reordering the way the exec permissions
are emitted, so that P and C modes come before U and i. Based on
http://wiki.apparmor.net/index.php/AppArmor_Core_Policy_Reference#Execute_rules
this should emit the modes correctly in all combined exec modes.
Other approaches to fixing this would require adjusting the data
structure that contains the permission modes, resulting in a more
invasive patch.
Bug: https://launchpad.net/bugs/982619
access to /proc/*/attr/{current,exec}, the onexec testcase that
attempted to do things without explicit access granted to
/proc/*/attr/exec in the testsuite passes instead of fails. This commit
takes that into account.
http://bugs.launchpad.net/bugs/979135
Currently a change_profile rule does not grant access to the
/proc/<pid>/attr/{current,exec} interfaces that are needed to perform
a change_profile or change_onexec, requiring that an explicit rule allowing
access to the interface be granted.
Make it so change_profile implies the necessary
/proc/@{PID}/attr/{current,exec} w,
rule just like the presence of hats does for change_hat
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
This fix is needed for the userspace portion of both
BugLink: http://bugs.launchpad.net/bugs/963756
BugLink: http://bugs.launchpad.net/bugs/978038
change_onexec fails for profiles that don't have an attachment specification
eg. unconfined
This is because change_onexec goes through 2 permission checks. The first
at the api call point, which is a straight match of the profile name
eg.
/bin/foo
unconfined
and a second test at exec time, tying the profile to change to to the
exec. This allows restricting the transition to specific execs. This
is mapped as a two entry check
/executable/name\x00profile_name
where the executable name must be marked with the change_onexec permission
and the subsequent profile name as well.
The previous "fix" only covered adding onexec to executable names and
also works for the initial change_onexec request when the profile is
an executable.
However it does not fix the case for when the profile being transitioned
to is not an executable.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
This extends the auto-profile generation so that it can take profiles formated
in standard profile language augemented by a few special variables for
the automatically generated rules. This will all extended the regression
tests in ways that are not currently supported, because mkprofile format
does not match of the profile language.
the special apparmorish variables are
@{gen_elf name} - generate rules for elf binaries
@{gen_bin name} - generate rules for a binary
@{gen_def} - generate default rules
@{gen name} - do @{gen_def} @{gen_bin name}
To generate a profile you do
genprofile --stdin <<EOF
/profile/name {
@{gen /profile/name}
}
EOF
eg. to generate the equivalent of
genprofile
you would do
genprofile --stdin <<EOF
$test {
@{gen $test}
}
EOF
and the equiv of
genprofile $file:rw
would be
genprofile --stdin <<EOF
$test {
@{gen $test}
$file rw,
}
while it takes a little more to generate a base profile than the old syntax, it
use the actual profile language (augmented with the special variables), it is a
lot more flexible, and a lot easier to expand when new rule types are added.
eg. of something not possible with the current auto generation
Generate a profile with a child profile and hat and a trailing profile
genprofile --stdin <<EOF
$test {
@{gen $test}
profile $bin/open {
@{gen $bin/open}
}
^hatfoo {
$file rw,
}
}
profile $bin/exec {
@{gen $bin/exec}
}
EOF
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
manpages (and adjust it so that it's one rule instead of eight). It
also fixes the above problem and a similar problem in the aa-exec
manpage.
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Acked-By: Jamie Strandboge <jamie@canonical.com>
Bugs: https://bugs.launchpad.net/ubuntu/+source/apparmor/+bug/800826https://bugzilla.novell.com/show_bug.cgi?id=755923
This patch modifies the libapparmor log parsing code to add support
for the additional ip address and port keywords that can occur in
network rejection rules. The laddr and faddr keywords stand for local
address and foreign address respectively.
The regex used to match an ip address is not very strict, to hopefully
catch the formats that the kernel emits for ipv6 addresses; however,
because this is in a context triggered by the addr keywords, it should
not over-eagerly consume non-ip addresses. Said addresses are returned
as strings in the struct to be processed by the calling application.
Bug: https://launchpad.net/bugs/800826
This patch calls autodep on the 'exec'ed binary when the user selects
to place that execution in a child profile. Previously, logprof would
create an entirely empty child profile in complain mode (this fix
still leaves the child profile in complain mode).
This patch fixes a couple of issue with autodep:
1) The initial profile construction had not been adjusted to include
the 'allow' or 'deny' hash prefixing the path elements. This
fixes it by eliminating the path portion entirely and pushing
the path based accesses to the later analysis section of code.
2) the mode of the original binary was accidentally getting reset
to 0, when it was intended to initialize the audit field to 0.
Bug #963756
The kernel has an extended test for change_profile when used with
onexec, that allows it to only work against set executables.
The parser is not correctly mapping change_profile for this test
update the mapping so change_onexec will work when confined.
Note: the parser does not currently support the extended syntax
that the kernel test allows for, this just enables it to work
for the generic case.
Signed-off-by: John Johansen <john.johansen@canonical.com>
The capabilities tests where failing in the changehat_wrapper test. This was because
they could not the changehat_wrapper sub executable, which trying to exec a binary
in the tmpdir.
Specifically if the test was for syscall_ptrace. It would generate a profile with
a hat for ^syscall_ptrace and attempt to execute ./syscall_ptrace. However this
was failing in some situations, including when trying to debug from the tmpdir,
as the syscall_XXX binary is no longer local.
Instead use the fully qualified path for the hat name, and the exec path.
Signed-off-by: John Johansen <john.johansen@canonical.com>
The retaining of the tmpdir is used during debugging of test failures, but currently
when a test fails, the next test is run overwritting the previous tmpdir value. This
is a problem even when manually running individual test shell scripts if the failure
is not the last test in the script.
Instead cause testing to about when retaintmpdir is true, which will cover the debugging
needs for the majority of failure cases.
Signed-off-by: John Johansen <john.johansen@canonical.com>
This patch adds a make install target for the generated apparmor.vim
file, installing by default into /usr/share/apparmor based on IRC
discussions; alternate suggestions welcome. (Installing directly
into the vim syntax tree is difficult as the system path by default
contains the vim version number.)
This patch replaces the apparmor.vim generating script with a python
version that eliminates the need for using the replace tool from the
mysql-server package. It makes use of the automatically generated
lists of capabilities and network protocols provided by the build
infrastructure. I did not capture all the notes and TODOs that
Christian had in the shell script; I can do so if desired.
It also hooks the generation of the apparmor.vim file into the utils/
build and clean stages.
This patch adds several missing capabilities to the utils/
severity.db file as detected by the newly added make check target,
along with corresponding severity levels that I believe :re appropriate
(discussion welcome):
CAP_MAC_ADMIN 10
CAP_MAC_OVERRIDE 10
CAP_SETFCAP 9
CAP_SYSLOG 8
CAP_WAKE_ALARM 8
The latter two are undocumented in the capabilities(7) man page
provided in Ubuntu 12.04; the syslog one is the separation out of
accessing the dmesg buffer from CAP_SYSADMIN, and the CAP_WAKE_ALARM
allows setting alarms that would wake a system from a suspended state,
if my reading is correct.
This also fixes a trailing whitespace on CAP_CHOWN, moves
CAP_DAC_READ_SEARCH to the end of the section of capabilities it's
in due to its lower priority level (7).
capabilities
This patch adds a new make target, check_severity_db, to the
utils/Makefile. It greps the severity.db for the presence of each
capability, as computed by the newly abstracted out variable in
common/Make.rules, and issues a build time error if it finds any
missing.
It also silences the check targets, so that only the output from them
will be emitted.
This patch abstracts out the generation of the lists of capabilities
and network protocol names to the common Make.rules file that is
included in most locations in the build tree, to allow it to be
re-used in the utils/ tree and possibly elsewhere.
It provides the lists in both make variables and as make targets.
It also sorts the resulting lists, which causes it to output differently
than the before case. I did confirm that the results for the generated
files used in the parser build were the same after taking the sorting
into account.
aa_getprocattr is returning the size of the buffer not the size of the
data read that it is supposed to return. Also update the man page to
reflect the return value as documented in the functions, and update
the test cases to check the return value.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Steve Beattie <sbeattie@ubuntu.com>
While aa_getprocattr does return the documented error code on failure
the **buf and **mode parameters can point into the buffer that was
allocated and then discarded on failure.
Set them to null on failure so that even if the error code is ignored
they do not point to heap data.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Remount should not be screening off the set of flags it is. They are
the set of flags that the kernel is masking out for make_type and
should not be used on remount. Instead just screen off the other cmds
that can have their own rules generated.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
The deny information is not used as valid accept state information,
so remove it from the is_null test. This does not change the dfa
generated but does result in the dumped information changing,
as states that don't have any accept information are no longer
reported as accepting. This is what changes the number of states
reported in the minimize tests.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
The same mappings routine had two bugs in it, that in practice haven't
manifested because of partition ordering during minimization. The
result is that some states may fail comparison and split, resulting
in them not being eliminated when they could be.
The first is that direct comparison to the nonmatching state should
not be done as it is a candiate for elimination, instead its partion
should be compared against. This simplifies the first test
The other error is the comparison
if (rep->otherwise != nonmatching)
again this is wrong because nomatching should not be directly
compared against. And again can result in the current rep->otherwise
not being eliminated/replaced by the partion. Again resulting in
extra trap states.
These tests where original done the way they were because
->otherwise could be null, which was used to represent nonmatching.
The code was cleaned up a while ago to remove this, ->otherwise is
always a valid pointer now.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Also make sure the perms method properly switches to hex and back to dec
as some of the previous perm dump code did not.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
The aa-exec command can be used to launch an application under a specified
confinement, which may be different for what regular profile attachment
would apply.
Signed-off-by: John Johansen <john.johansen@canonical.com>
The changes are around how user data is handled.
1. permissions are mapped before data is matched
2. If data is to be mapped a AA_CONT_MATCH flag is set in the permissions
which allows data matching to continue.
3. If data auditing is to occur the AA_AUDIT_MNT_DATA flag is set
This allows better control over matching and auditing of data which can
be binary and should not be matched or audited
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
file, should grant access to all files paths on the system but it does
not currently allow access to /
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Rename the pivotroot rule to pivot_root to match the command and the fn
and fix it to support named transition correctly leveraging the parsing
action used for exec transitions.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Currently the backend doesn't like it (blows up) when the a vector entry is
empty. For the case where no flags match build_mnt_flags generates an
alternation of an impossible entry and nothing
(impossible|)
This provides the effect of a null entry without having an empty vector
entry. Unfortunately the impossible entry is not correct.
Note: how this is done needs to be changed and fixed in the next release
this is just a minimal patch to get it working for 2.8
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
When generating the flag set the parser was not generating the complete
set when flags where not consecutive. This is because the len value
was not being reset for each flag considered, so once it was set for
a flag, then the next flag would have to be set to reset it else the
output string was still incremented by the old len value.
Eg.
echo "/t { mount options=rbind, }" | apparmor_parser -QT -D rule-exprs
results in
rule: \x07[^\000]*\x00[^\000]*\x00[^\000]*\x00\x0d ->
however \x0d only covers the bind and not the recursive flag
This is fixed by adding a continue to the flags generation loop for the
else case.
resulting the dump from above generating
rule: \x07[^\000]*\x00[^\000]*\x00[^\000]*\x00\x0d\x0f ->
\x0d\x0f covers both of the required flags
Also fix the flags output to allow for the allow any flags case. This
was being screened out. By masking the flags even when no flags where
specified.
this results in a difference of
echo "/t { mount, }" | apparmor_parser -QT -D rule-exprs
rule: \x07[^\000]*\x00[^\000]*\x00[^\000]*\x00(\x01|)(\x02|)(\x03|)(\x04|)(\x05|)\x00[^\000]*
becoming
\x07[^\000]*\x00[^\000]*\x00[^\000]*\x00[^\000]*\x00[^\000]*
which is simplified and covers all permissions vs. the first rule output
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
This commit causes policy problems because we do not have chroot rules
and policy extension to support it.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
the cache test is failing because it assumes that kernel features are
stored in a file instead of a directory
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
On newer kernels the features directory causes the creation of a
cache/.feature file that contains newline characters. This causes the
feature comparison to fail, because get_flags_string() uses fgets
which stop reading in the feature file after the first newline.
This caches the features comparision to compare a single line of the
file against the full kernel feature directory resulting in caching
failure.
Worse this also means the cache won't get updated as the parser doesn't
change what set gets caches after the .feature file gets created.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
file,
was not given the correct permissions. It was only being given the owner
set of permissions. This would result in rejects when trying look at
files owned by other users
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
There are some rare occassions, when lots of alternations are used that
tree simplification can result in an expression of
(E | (E | E)) or (E . (E . E)) where E is the epsnode
both of these expressions will lead to an inifinite loop in normalize_tree
as the epsnode test
if ((&epsnode == t->child[dir]) &&
(&epsnode != t->child[!dir]) &&
dynamic_cast<TwoChildNode *>(t)) {
and the tree node rotation test
} else if ((dynamic_cast<AltNode *>(t) &&
dynamic_cast<AltNode *>(t->child[dir])) ||
(dynamic_cast<CatNode *>(t) &&
dynamic_cast<CatNode *>(t->child[dir]))) {
end up undoing each others work, ie.
eps flip rotate
(E | (E | E)) --------> ((E | E) | E) -------> (E | (E | E))
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Minimization was failing because it was too agressive. It was minimizing
as if there was only 1 accept condition. This allowed it to remove more
states but at the cost of loosing unique permission sets, they where
being combined into single commulative perms. This means that audit,
deny, xtrans, ... info on one path would be applied to all other paths
that it was combined with during minimization.
This means that we need to retain the unique accept states, not allowing
them to be combined into a single state. To do this we put each unique
permission set into its own partition at the start of minimization.
The states within a partition have the same permissions and can be combined
within the other states in the partition as the loss of unique path
information is will not result in a conflict.
This is similar to what perm hashing used to do but deny information is
still being correctly applied and carried.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
The in x intersection consistency test for minimization was failing because
it was screening off the AA_MAY_EXEC permission before passing the exec
information to the consistency test fn. This resulted in the consistency
test fn not testing the consistency because it treated the permission set
as not having x permissions.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Make them report a hex value strings instead of the default C++
\vvvvv
Make them consistent,
- Dump to report the default transition and what isn't transitioned
on it.
Signed-off-by: John Johansen <john.johansen@canonical.com>
The permission reporting was not reporting the full set of permission
flags and was inconsistent between the dump routines.
Report permissions as the quad (allow/deny/audit/quiet) in hex.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Fix the transitions states output so that they output the state label
instead of the state address. That is
{1} -> 0x10831a0: /
now becomes
{1} -> {2}: /
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
The pcre parser in the dfa backend is not correctly converting escaped
hex string like
\0x0d
This is the minimal patch to fix, and we should investigate just using
the C/C++ conversion routines here.
I also I nominated for the 2.7 series.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@gmail.com>
Add the ability to control mounting and unmounting
The basic form of the rules are.
[audit] [deny] mount [conds]* [device] [ -> [conds] path],
[audit] [deny] remount [conds]* [path],
[audit] [deny] umount [conds]* [path],
[audit] [deny] pivotroot [oldroot=<value>] <path> -> <profile>
remount is just a short cut for mount options=remount
where [conds] can be
fstype=<expr>
options=<expr>
conds follow the extended conditional syntax of allowing either:
* a single value after the equals, which has the same character range as
regular IDS (ie most anything but it can't be terminated with a , (comma)
and if spaces or other characters are needed it can be quoted
eg.
options=foo
options = foo
options="foo bar"
* a list of values after the equals, the list of values is enclosed within
parenthesis () and its has a slightly reduced character set but again
elements can be quoted.
the separation between elements is whitespace and commas.
eg.
options=(foo bar)
options=(foo, bar)
options=(foo , bar)
options=(foo,bar)
The rules are flexible and follow a similar pattern as network, capability,
etc.
mount, # allow all mounts, but not umount or pivotroot
mount fstype=procfs, # allow mounting procfs anywhere
mount options=(bind, ro) /foo -> /bar, # readonly bind mount
mount /dev/sda -> /mnt,
mount /dev/sd** -> /mnt/**,
mount fstype=overlayfs options=(rw,upperdir=/tmp/upper/,lowerdir=/) overlay -> /mnt/
umount,
umount /m*,
Currently variables and regexs are are supported on the device and mount
point. ie.
mount <devince> -> <mount point>,
Regexes are supported in fstype and options. The options have a further
caveat that regexs only work if the option is fs specific option.
eg. options=(upperdir=/tmp/*,lowerdir=/)
regex's will not currently work against the standard options like ro, rw
nosuid
Conditionals (fstype) can only be applied to the device (source) at this
time and will be disregarded in situations where the mount is manipulating
an existing mount (bind, remount).
Options can be specified multiple times
mount option=rw option=(nosuid,upperdir=/foo),
and will be combined together into a single set of values
The ordering of the standard mount options (rw,ro, ...) does not matter
but the ordering of fs specific options does.
Specifying that the value of a particular option does not matter can be
acheived by providing both the positive and negative forms of and option
option=(rw,ro) options=(suid,nosuid)
For the fs specific options specifying that a particular value does not
matter is achieve using a regex with alternations.
Improvements to the syntax and order restrictions are planned for the
future.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Newer versions of AppArmor use a features directory instead of a file
update the parser to use this to determine features and match string
This is just a first pass at this to get things up quickly. A much
more comprehensive rework that can parse and use the full information
set is needed.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Due to changes in path looks and the work going forward default profiles
to resolve relative to the chroot instead of the namespace.
This will only affect profiles that are used on tasks within a chroot.
For now it will be possible to get the old default namespace relative
behavior by passing the namespace_relative flag to the profile
eg.
profile /example (namespace_relative) { .. }
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
policydb is the new matching format, that combines the matching portions
of different rules into a single dfa/hfa. This patch only lays some ground
work it does not add encoding of any rules into the policydb
Signed-off-by: John Johansen <john.johansen@canonical.com>
value_list can be reused by conditionals and list values, so pull it out
and abstract it some more.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
Add the optional 'file' keyword to the language/grammer. The main reason
for doing this is to support false token injection. Which is needed
to move towards the parser being broken out into an api that can be
used to parse individual rule types, separate from parsing the whole file.
Since we are adding the token to the grammar expose it to userspace with
the 'file' keyword. While not needed it helps bring consistency, as all
the other rule types start with a keyword (capability, network, rlimit, ...).
Also allow the bare keyword to be used to represent allowing all file
operations, just as with network and capability. Domain transitions are
defaulted to ix. Thus
file,
is equivalent to
/** rwlkmix,
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
The ability to set capabilities from a profile has been removed from the
kernel for several releases. Remove it from the parser as well.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
Allow the capability rule to be bare to represent all capabilities similar
to how network, and other rule types work.
capability,
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
Currently the parser can not directly influence the lexer output. This
limits the grammar and also how the parser can be invoked. Allow the
parser to pass the next TOKEN that the lexer will return.
This is has two uses: It allows us to trick the bison parser into having
multiple start symbols, allowing us to say invoke the parser on an
individual network or file rule. It also allows the semantic analysis of
the parser to change the language recognized. This can be leveraged to
overcome some of the limitation of bison's LALR parse generator.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
mode strings overlap with other potential commands, or strings, and as
currently written can be match as a leading substring of an ID. Eliminate
the leading substring case by requiring that for a mode string to be
recognized it must be terminated by whitespace, eol, eof, or
comma (end of rule).
The other cases where modes string overlap are ambiguous and the ID should
be quoted to remove the ambiguity.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
extended conditionals use a syntax of
cond=value
cond=(value1 value2)
cond=(value1,value2)
where the comma is optional and supported as such because the
flags syntax used them
The mount patch extends, and improves on this patch, the changes probably
should have been separated out but ...
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
Signed-off-by: John Johansen <john.johansen@canonical.com>
jj@ortho:~/apparmor/aa-test/parser$ guilt header
Convert FLAGS_MODE start condition to a generic list of values start cond
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
ID and POST_VAR_ID define a set of characters that is reused, pull this
out to avoid making mistakes when updating the character set.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
The removal of deny information is a one way operation, that can result
in a smaller dfa, but also results in a dfa that should not be used in
future operations because the deny rules from the precomputed dfa would
not get applied.
For now default filtering out of deny information to off, as it takes
extra time and seldom results in further state reduction.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
Previously permission information was thrown away early and permissions
where packed to their CHFA form at the start of DFA construction. Because
of this permissions hashing to setup the initial DFA partitions was
required as x transition conflicts, etc. could not be resolved.
Move the mapping of permissions to CHFA construction, and track the full
permission set through DFA construction. This allows removal of the
perm_hashing hack, which prevented a full minimization from happening
in some DFAs. It also could result in x conflicts not being correctly
detected, and deny rules not being fully applied in some situations.
Eg.
pre full minimization
Created dfa: states 33451
Minimized dfa: final partitions 17033
with full minimization
Created dfa: states 33451
Minimized dfa: final partitions 9550
Dfa minimization no states removed: partitions 9550
The tracking of deny rules through to the completed DFA construction creates
a new class of states. That is states that are marked as being accepting
(carry permission information) but infact are non-accepting as they
only carry deny information. We add a second minimization pass where such
states have their permission information cleared and are thus moved into the
non-accepting partion.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
Delay the packing of audit and quiet permissions until chfa construction,
and track deny and quiet perms during DFA construction, so that we will
be able to do full minimization. Also delay the packing of audit and
Signed-off-by: John Johansen <john.johansen@canonical.com>
From: Kees Cook <kees@ubuntu.com>
This is a trivial manpage fix that makes pod2man stop yelling at me.
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
From: Steve Beattie <sbeattie@ubuntu.com>
This patch adds --stderr to pod2man to make it report errors, as well as
fixes a few other minor text issues I noticed.
Acked-by: Kees Cook <kees@ubuntu.com>
capabilities). The rule will be marked in the "dangerous capability"
color.
Additionally, the patch removes the (already commented out) code for
"set capability".
Acked-by: Kees Cook <kees@ubuntu.com>
Description: glibc's __get_nprocs() now checks /sys/devices/system/cpu/online
in addition to /proc/stat for the number of processors. This is used in the
_SC_NPROCESSORS_ONLN implementation, a part of sysconf. This was introduced in
upstream glibc commit:
84e2a551a7
Bug-Ubuntu: https://launchpad.net/bugs/929531
Acked-By: Jamie Strandboge <jamie@canonical.com>
Acked-By: Christian Boltz <apparmor@cboltz.de>
(At least) openSUSE uses ~/.kde4 to store KDE4 settings.
This patch changes ~/.kde/ to ~/.kde{,4} in all abstractions.
The patch is mostly from Velery Valery, I only fixed a merge conflict
and added the kmail{,2} part in private-files-strict.
References: https://bugzilla.novell.com/show_bug.cgi?id=741592
Acked-By: Steve Beattie <sbeattie@ubuntu.com> for both trunk and 2.7.
A bug in Ubuntu reported that the aspell abstraction does
not allow write access to the user customizable dictionaries, the
personal dictionary (~/.aspell.$LANG.pws) and the personal replacement
dictionary (~/.aspell.$LANG.prepl). It also adjusts the abstraction
to add the owner modifier to the personal dictionaries.
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-by: John Johansen <john.johansen@canonical.com>
Bug: https://launchpad.net/bugs/917859
it is needed by pretty much all of the browser abstractions. aa-update-browser
unconditionally adds the plugins-common abstraction, so this should be
sufficient.
properly quote the _known variable (set when the tests are marked as
expected failures) when the expectation was that the testcase would
produce a corefile. This would result in a failed testcase reporting
XFAIL incorrectly.
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-by: John Johansen <john.johansen@canonical.com>
suffix of an image= flag, as it's no longer needed. It also eliminates
code that emitted the permissions based on the result of the parse.
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-by: John Johansen@canonical.com
add 'rix' permissions on executable images (but still auto-generate
ldd dependencies), for use when specifying alternate permissions
on executables.
Where appropriate, it also converts a few testcases to make use of
the option.
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-by: John Johansen <john.johansen@canonical.com>
generated profile and have the ldd auto-generation of rules occur on
it. It also kills all testcase usage of $dynlibs, which had stopped
being generated by prologue.inc in a prior patch.
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-by: John Johansen <john.johansen@canonical.com>
mkprofile.pl helper and take the convoluted code out of the bash
prologue.inc. It also detects if the binary is a script and performs
ldd analysis on the interpreter.
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-by: John Johansen <john.johansen@canonical.com>
their purpose a little more accurately; renames the dump_flags to
emit_flags for the same reason, and also adds a modicum a function
prototype information to the function declarations.
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
Acked-by: John Johansen <john.johansen@canonical.com>
Profile loads when specifying namespaces currently conflict with caching.
If the profile (ignoring the specified namespace) is in the cache, then
the cached profile will be loaded, replacing the profile in the current
namespace instead of loading the profile to the new namespace.
Fix this by disabling caching when a namespace is specified, forcing the
profile to be compiled.
NOTE: this will not affect profiles loaded from within a namespace using
either the same or a separate directory as the base to load a namespac
from. This only affects loading profiles directly into a child
namespace.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
Currently the cache location is fixed and links are needed to move it.
Add an option that can be set in the apparmor_parser.conf file so distros
can locate the cache where ever makes sense for them.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
The behavior for revalidation/revocation of open files has changed
with the current kernel code, resulting in these tests being reported
as failing even though they are showing expected behavior.
Under the current kernel module this form of revalidation/revocation
can not be tested reliably, so just changing the expected result is
not enough, completely disable the tests for now.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Lenient profile that is intended to be used when 'Ux' is desired but
does not provide enough environment sanitizing. This effectively is an
open profile that blacklists certain known dangerous files and also
does not allow any capabilities. For example, it will not allow 'm' on files
owned be the user invoking the program. While this provides some additional
protection, please use with care as applications running under this profile
are effectively running without any AppArmor protection. Use this profile
only if the process absolutely must be run (effectively) unconfined.
Limitations:
1. This does not work for root owned processes, because of the way we use
owner matching in the sanitized helper. We could do a better job with
this to support root, but it would make the policy harder to understand
and going unconfined as root is not desirable anyway.
2. For this sanitized_helper to work, the program running in the sanitized
environment must open symlinks directly in order for AppArmor to mediate
it. This is confirmed to work with:
- compiled code which can load shared libraries
- python imports
It is known not to work with:
- perl includes
3. Going forward it might be useful to try sanitizing ruby and java
Use at your own risk. This profile was developed as an interim workaround for
LP: #851986 until AppArmor implements proper environment filtering.
Acked-by: Jamie Strandboge <jamie@canonical.com>
Adjust ubuntu abstractions to use sanitized_helper instead of (P)Ux.
Acked-by: Jamie Strandboge <jamie@canonical.com>
Update launchpad-integration to use a sanitized helper in a similar manner
as that in ubuntu-helpers.
Acked-by: Jamie Strandboge <jamie@canonical.com>
creating owner writes on things like ~/.cache and ~/.config
Acked-by: Jamie Strandboge <jamie@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
add machine script = /usr/sbin/smbldap-useradd -t 5 -w "%u"
smbd obviously needs x permissions for smbldap-useradd.
The commit also adds a new profile for usr.sbin.smbldap-useradd (based on
the audit.log from Alexis Pellicier).
Additionally, I moved the "/etc/samba/* rwk" rule next to the other
/etc-related rules in the smbd profile.
References: https://bugzilla.novell.com/show_bug.cgi?id=738041
Lenient profile that is intended to be used when 'Ux' is desired but
does not provide enough environment sanitizing. This effectively is an
open profile that blacklists certain known dangerous files and also
does not allow any capabilities. For example, it will not allow 'm' on files
owned be the user invoking the program. While this provides some additional
protection, please use with care as applications running under this profile
are effectively running without any AppArmor protection. Use this profile
only if the process absolutely must be run (effectively) unconfined.
Limitations:
1. This does not work for root owned processes, because of the way we use
owner matching in the sanitized helper. We could do a better job with
this to support root, but it would make the policy harder to understand
and going unconfined as root is not desirable any way.
2. For this sanitized_helper to work, the program running in the sanitized
environment must open symlinks directly in order for AppArmor to mediate
it. This is confirmed to work with:
- compiled code which can load shared libraries
- python imports
It is known not to work with:
- perl includes
3. Going forward it might be useful to try sanitizing ruby and java
Use at your own risk. This profile was developed as an interim workaround for
LP: #851986 until AppArmor implements proper environment filtering.
From the README in the toplevel source:
"[P11-KIT] Provides a way to load and enumerate PKCS#11 modules. Provides a
standard configuration setup for installing PKCS#11 modules in such a way that
they're discoverable."
File locatations are described in [1]. There is a global configuration file in
/etc/pkcs11/pkcs11.conf. Per module configuration happens in
/etc/pkcs11/<module name>. There is also user configuration in ~/.pkcs11, but
IMO this should not be allowed in the abstraction. Example configuration can be
seen in the upstream documentation[2].
This will likely need to be refined as more applications use p11-kit.
[1]http://p11-glue.freedesktop.org/doc/p11-kit/config-locations.html
[2]http://p11-glue.freedesktop.org/doc/p11-kit/config-example.html
Acked-by: Jamie Strandboge <jamie@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Also add p11-kit to authentication abstraction
Acked-by: Jamie Strandboge <jamie@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
From the README in the toplevel source:
"[P11-KIT] Provides a way to load and enumerate PKCS#11 modules. Provides a
standard configuration setup for installing PKCS#11 modules in such a way that
they're discoverable."
File locatations are described in [1]. There is a global configuration file in
/etc/pkcs11/pkcs11.conf. Per module configuration happens in
/etc/pkcs11/<module name>. There is also user configuration in ~/.pkcs11, but
IMO this should not be allowed in the abstraction. Example configuration can be
seen in the upstream documentation[2].
This will likely need to be refined as more applications use p11-kit.
[1]http://p11-glue.freedesktop.org/doc/p11-kit/config-locations.html
[2]http://p11-glue.freedesktop.org/doc/p11-kit/config-example.html
Acked-by: Jamie Strandboge <jamie@canonical.com>
Currently hfa::match calls hfa::match_len to do matching. However this
requires walking the input string twice. Instead provide a match routine
for input that is supposed to terminate at a given input character.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Add the ability to match strings directly from the hfa instead of needing
to build a cfha.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
asprintf is marked with warn_unused_result and its return value should
not be ignored, even casting to (void) will not remove this warning.
The current code ignored the result and used the value of newfmt to
make a decision. This is however not correct in that according to the
asprintf man page newfmt is undefined if asprintf returns an error.
Fix the warning and error by using the return value of asprintf
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
abstractions/apache2-common. Additionally, add read permissions
for /**/.htaccess and /dev/urandom to apache2-common.
The patch is based on a profile abstraction from darix. I made some
things more strict (compared to darix' profile), and OTOH added some
things that are needed on my servers.
*** BACKWARDS-INCOMPATIBLE CHANGES ***
^HANDLING_UNTRUSTED_INPUT
- don't allow /.htaccess (.htaccess files in subdirectories are still allowed)
- don't allow *.htaccess files (the old /**.htaccess rule was too generous)
Reworking this code is a step to getting rid of the SUB_NAME2 start
condition.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
Change how we handle the parsing of the hat and profile keywords this allows
us to get rid of the SUB_NAME2 start condition because the the whitespace
that is allowed by these rules are now consumed by matching the keyword
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
There is a lot of duplication of code calling processqunquoted and
processquoted. Move all this code to use the new processid fn.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
This is the first step in reducing the number of shared rules between the
different start conditions.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
The affected comment rule is already in the INITIAL start condition
so BEGIN(INITIAL) is extraneous and will cause problems when switching
to a stack of start conditions.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
The rlimit start condition was separating different rules of the base
set making the lexer grammer harder to read than necessary.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
The module interface calls for names with namespaces to be in the format of
:namespace:profile or :namespace://profile
but the parser was generating
namespace:profile
causing profile lookup to fail, or removal of the wrong profile as it was
done against the current namespace, instead of the specified namespace
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
The module interface calls for names with namespaces to be in the format of
:namespace:profile or :namespace://profile
but the parser was generating
namespace:profile
causing profile lookup to fail, or removal of the wrong profile as it was
done against the current namespace, instead of the specified namespace
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
instead of a NodeSet.
We need to store sets of Nodes, to compute the dfa but the C++ set is
not the most efficient way to do this as, it has a has a lot of overhead
just to store a single pointer.
Instead we can use an array of tightly packed pointers + a some header
information. We can do this because once the Set is finalized it will
not change, we just need to be able to reference and compare to it.
We don't use C++ Vectors as they have more overhead than a plain array
and we don't need their additional functionality.
We only replace the use of hashedNodeSets for non-accepting states as
these sets are only used in the dfa construction, and dominate the memory
usage. The accepting states still may need to be modified during
minimization and there are only a small number of entries (20-30), so
it does not make sense to convert them.
Also introduce a NodeVec cache that serves the same purpose as the NodeSet
cache that was introduced earlier.
This is not abstracted this out as nicely as might be desired but avoiding
the use of a custom iterator and directly iterating on the Node array
allows for a small performance gain, on larger sets.
This patch reduces the amount of heap memory used by dfa creation by about
4x - overhead. So for small dfas the savings is only 2-3x but on larger
dfas the savings become more and more pronounced.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
non-accepting, and have the proto-state use them.
To reduce memory overhead each set gains its own "cache" that make sure
there is only a single instance of each NodeSet generated. And since
we have a cache abstraction, move relavent stats into it.
Also refactor code slightly to make caches and work_queue etc, DFA member
variables instead of passing them as parameters.
The split + caching results in a small reduction in memory use as the
cost of ProtoState + Caching is less than the redundancy that is eliminated.
However this results in a small decrease in performance.
Sorry I know this really should have been split into multiple patches
but the patch evolved and I got lazy and decided to just not bother
splitting it.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
It is the functional equivalent of ProtoState. We do this to provide a
new level of abstraction that ProtoState can leverage, when the node types
are split.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
Create a new ProtoState class that will encapsulate the split, but for
this patch it will just contain what was done previously with NodeSet
Signed-off-by: John Johansen <john.johansen@canonical.com>
is done to be clear what TransitionTable is, as we will then add matching
capabilities. Renaming the files is just to make them consistent with
the class in the file.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
entries where the comm entry has been hex-encoded. This occurs when the
binary being confined contains a space or other problematic character in
its filename. A test case is included.
When executing apparmor_status from rc functions and utils are not installed, this message is received:
AppArmor is enabled,
Install the apparmor-utils package to receive more detailed
status information here (or examine directly).
Signed-off-by: John Johansen <john.johansen@canonical.com>
parser/Makefile has a number of issues.
* Some warnings are produced: "make[1]: warning: jobserver unavailable: using -j1. Add `+' to parent make rule."
* CXX is not always respected
* LDFLAGS are not always respected
modified to apply and retain $(Q) by John Johansen
Signed-off-by: John Johansen <john.johansen@canoncial.com>
When running installing apparmor-utils from Makefile, some warnings are
generated: make[1]: warning: jobserver unavailable: using -j1. Add
`+' to parent make rule.
Use $(MAKE) instead of make
Signed-off-by: John Johansen <john.johansen@canonical.com>
Original openSUSE changelog entry:
Thu Jan 6 16:23:19 UTC 2011 - rhafer@suse.de
- Splitted ldap related things from nameservice into separate
profile and added some missing paths (bnc#662761)
If $DISPLAY is not set and --display is not used, aa-notify prints a
warning that notifications won't be shown (exact warning text depends if
using sudo or not).
Acked-by: John Johansen <john.johansen@canonical.com>
requested and fails to be created. Also don't make the
warning output conditional on the showcache flag as we
should be showing warning/errors by default.
Signed-off-by: John Johansen <john.johansen@canonical.com>
openSUSE, which results in non-working desktop notifications in aa-notify
because $notify_exe is unable to connect to DBUS to display the message.
This patch sets the correct value for $HOME.
The code for setting $DISPLAY is still under discussion, therefore only
a TODO note is included in this commit for $DISPLAY.
Acked-By: John Johansen <john.johansen@canonical.com>
rc.apparmor.suse), I noticed that "rcapparmor restart" is
totally silent.
The attached patch prints a message in __apparmor_restart().
It also replaces the hardcoded "return 0" with $?. I'm quite sure this
won't catch all errors, but it's still better than the hardcoded success
message.
Acked-by: John Johansen <john.johansen@canonical.com>
- fix permissions for additional-log-sockets.conf (the comma in {var/,}
was at the wrong place, which broke the /var/run/ case)
- add read permissions for /sys/devices/system/cpu/online
(that was even new for Peter, but I trust him not to post faked
audit.log lines ;-)
Acked-by: John Johansen <john.johansen@canonical.com>
Fix the build so
make DEBUG=1
results in a compile with DEBUG turned on.
Also fix build errors in the compile with DEBUG is defined
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees.cook@canonical.com>
This documentation should have been checked in as part of the patches
that added aa_is_confined and aa_get_con.
Signed-off-by: John Johansen <john.johansen@canonical.com>
The define for pid_t is missing in apparmor.h so that if it is included
in programs that don't also include sys/types.h the compile will break.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees.cook@canonical.com>
Subject: apparmor-profiles: Add samba config files
References: bnc#679182 bnc#666450
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
- updated to match trunk
- added changed path to nmbd profile (/var/cache/samba has moved to
/var/lib/samba on (at least) openSUSE 11.4), bnc#679182#c8
For backward compability, it also allows /var/spool/samba.
- Note: The smbd profile already contains both locations.
by Christian Boltz <apparmor@cboltz.de>
updated according to the comments from Steve Beattie
by Christian Boltz <apparmor@cboltz.de>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
- allow /var/spool/mail, not only the /var/mail symlink
- allow @{HOME}/Mail/
- allow capability fsetid, read access to /etc/lsb-release and
SuSE-release and k for /var/{lib,run}/dovecot in usr.bin.dovecot
References:
- dovecot: Added support for /var/spool/mail (bnc#691072)
- Updated dovecot profile (bnc#681267).
Patch taken from openSUSE:11.4:Update:Test, file apparmor-profiles-dovecot
updated to match trunk by Christian Boltz <apparmor@cboltz.de>
Change compared to the patch posted to the ML:
- link rule instead of adding l permissions for /var/lib/dovecot and
/var/run/dovecot (as proposed by John Johansen)
Acked-By: John Johansen <john.johansen@canonical.com> on IRC
by converting the comm(1) usage on temporary files to an embedded
awk script. On both Ubuntu and OpenSUSE, a version of awk (mawk in
Ubuntu, gawk in OpenSUSE) is either a direct or indirect dependency
on the minimal or base package set, and the original reporter also
mentioned that an awk-based solution would be palatable in a way that
converting to bash, or using perl or python here would not be.
In the embedded awk script, I've tried to avoid gawk or mawk specific
behaviors or extensions; e.g. this is the reason for the call to sort
on the output of the awk script, rather than using gawk's asort(). But
please let me know if you see anything that shouldn't be portable
across awk implementations.
An additional issue that is fixed in both scripts is handling child
profiles (e.g. hats) during reload. If child profiles are filtered
out (via grep -v '//') of the list to consider, then on reloading
a profile where a child profile has been removed or renamed, that
child profile will continue to stick around. However, if the profile
containing child profiles is removed entirely, if the initscript
attempts to unload the child profiles after the parent is removed,
this will fail because they were unloaded when the parent was unloaded.
Thus I removed any filtering of child profiles out, but do a post-awk
reverse sort which guarantees that any child profiles will be removed
before their parent is. I also added the LC_COLLATE=C (based on the
Ubuntu version) to the sort call to ensure a consistent sort order.
To restate, the problem with the existing code is that it creates
temporary files in $TMPDIR (by default /tmp) and if that partition
is full, problems with the reload action ensue. Alternate solutions
include switching the initscript to use bash and its <$() extension
or setting TMPDIR to /dev/shm/. The former is unpalatable to some
(particularly for an initscript), and for the latter, /dev/shm is
only guaranteed to exist on GNU libc based systems (glibc apparently
expects /dev/shm to exist for its POSIX shared memory implementation;
see shm_overview(7)). So to me, awk (sans GNU extensions) looks to
be the least bad option here.
Bug: https://launchpad.net/bugs/775785
to switch to a non-root user. unscd is installed as /usr/sbin/nscd
at least at openSUSE.
Original changelog entry from unscd package:
Mon Sep 7 17:30:36 CEST 2009 - pbaudis[at]suse.cz
- Provide the /etc/apparmor.d/usr.sbin.nscd file and make it allow
for change to the nobody user [bnc#535467]
Currently the nscd package from glibc and the unscd package both contain
a usr.sbin.nscd profile which needs to maintained/updated manually.
With this patch, the profile could be moved back to the
apparmor-profiles package.
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Subject: apparmor-utils: Add support for creds and path operations
References: bnc#564316
2.6.29 introduced the path security_operations and credentials
This patch adds support for those operations to the log parser.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Resolved merge conflict and removal operation already supported by
the log parser.
Acked-by: John Johansen <john.johansen@canonical.com>
changing. Since we arguably shouldn't be hardcoding this kind of
thing, this changes a path around to use the C preprocessor to do the
work of finding the kernel definitions.
Signed-off-by: Kees Cook <kees.cook@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
instead.
Needed at least on upgraded Ubuntu machines that went from messages to
syslog recently. If this causes problems, we can easily revert it.
Signed-off-by: Kees Cook <kees.cook@canonical.com>
Acked-by: John Johansen <john.johansen@canonical.com>
syslog, otherwise audit events will get dropped.
This runs the risk of having the kernel log wrap around, but that
is a less common case that what this solves. This is a work-around
that will go away when complain info takes a different path in the
future.
I intentionally don't allow pUx and Pux since the behaviour of those is
very unexpected (the first letter decides if the environment is cleaned
up or not - at least that's the result of the discussion in April) and
the average user won't know this.
Acked-by: John Johansen <john.johansen@canonical.com>
profiles.
It also adds a comment to the klogd profile that capability sys_admin is
only needed for backward compatibility with older kernels.
Acked-by: John Johansen <john.johansen@canonical.com>
the modifiers as lowercase (meaning to pass on sensitive environment
variables to the exec'ed process) even if the user told them not to
when prompted. This patch fixes the issue.
aa-notify would abort if it could not stat the logfile, as can happen
when using auditd and the directory perms for the logfile do not allow access
(x). Add raise_privileges() and drop_privileges() helper functions and adjust
get_logfile_size() and get_logfile_inode() to raise then drop privileges if the
logfile parent directory is not executable. Also adjust reopen_logfile() to use
these helpers.
When error checking in these helpers, use '$> == ...' instead of '$> = ... or
die...' since perl always dies when raising privs in this manner even though
the euid did change (and $!, $@, $^E, and $? are all the same). Not sure why
this is happening but the '==' check should be sufficient.
aa-notify would abort if it could not stat the logfile, as can happen
when using auditd and the directory perms for the logfile do not allow access
(x). Add raise_privileges() and drop_privileges() helper functions and adjust
get_logfile_size() and get_logfile_inode() to raise then drop privileges if the
logfile parent directory is not executable. Also adjust reopen_logfile() to use
these helpers.
When error checking, use '$> == ...' instead of '$> = ... or die...' since perl
always dies when raising privs in this manner even though the euid did change
(and $!, $@, $^E, and $? are all the same). Not sure why this is happening but
the '==' check should be sufficient.
- drop supplemental group privileges too. While POSIX::setgid() works nice in
that it will set both the real uid and euid, it doesn't do anything with the
supplemental groups (sigh). Instead, assign to $( and $) in a manner that
clears the supplemental groups.
in merging in the opensuse fixes to the dhclient profile. It does so
by merging them, using the profile for dhclient-script if it exists
and inheriting dhclient's profile if it does not.
Subject: apparmor-utils: Fix handling of files in /
References: bnc#397883
The separate handling of files and directories with realpath is broken.
For files e.g. /foo, $dir ends up being empty since the / is eaten by
the regex. realpath resolves an empty argument as the current directory,
resulting in an incorrect path.
There's no explanation of why the separate handling was used in the
first place.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Subject: Fix for sshd profile
References: bnc#457072
Without this patch, sshd won't work in enforce mode.
libselinux accesses /proc/filesystems to determine if it's enabled
bash won't execute
audit_control is probably from libselinux too
Updated by Christian Boltz <apparmor@cboltz.de>:
- add /proc/*/oom_adj and oom_score_adj rw
- add /var/log/btmp r
- add /var/log/lastlog k
- removed capability sys_ptrace - doesn't seem to be needed
- changed all login shells to rUx, not only bash
- removed /proc/filesystems (already part of abstractions/base)
Acked-By: John Johansen <john.johansen@canonical.com>
with the following note:
ACK because I don't see a choice right now but for the 3.0 release
(next year) I'll ask you to retest and add newer audit controls.
Subject: apparmor-utils: Add check_for_apparmor helper.
This should be an alias but those get complicated quickly in perl.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Subject: apparmor-utils: setprofileflags() drops leading whitespace
References: bnc#480795
setprofileflags() drops leading whitespace for subprofiles. writeheader()
properly indents subprofiles 2 spaces per nesting level but when
genprof sets the profile to enforce mode at completion, the whitespace
is removed.
This patch adds the whitespace globbing to the regexp and uses it to
prefix the sub-profile with the correct spacing.
Reported at: https://bugzilla.novell.com/show_bug.cgi?id=480795
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Allow for rlimit cpu to specified which is now supported by the kernel.
Previously the rlimit units where limited to K, M, G and would fail when
KB, MB, GB where used. Allow for both, also allow for units on lengths
of time, by specifying "seconds", "minutes", "hours".. or any unique subset
eg. "s", "sec", "m", "min", "h", "hour" ..
NOTE:
This patch does not extend rlimits to be able to handle setting of tasks
that are confined by other profiles.
Signed-off-by: John Johansen <john.johansen@canonical.com>
to /etc/apparmor/parser.conf (NOTE option to allow changing this is not
provided currently).
Signed-off-by: John Johansen <john.johansen@canonical.com>
create the generated_* directories themselves if they don't exist before
running the script to generate them.
Also modify the default invocation of prove to add -f, which reports the
details of failing test cases.
Changed /var/run/cups/** rw, to
/{,var/}run/cups/ rw,
/{,var/}run/cups/** rw,
as requested by Steve Beattie
With this change:
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Subject: dnsmasq: Profile fixes
References: bnc#666090 bnc#678749
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Updated to match master by
Christian Boltz <apparmor@cboltz.de>
Updated for systemd (/{,var/},run/ instead of /var/run/) by
Christian Boltz <apparmor@cboltz.de> as requested by Steve Beattie
With this change:
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
(final confirmation on IRC in #apparmor)
updated to match master by
Christian Boltz <apparmor@cboltz.de>
updated to work with systemd (/{,var/}run/ instead of /var/run)
Christian Boltz <apparmor@cboltz.de> as requested by Steve Beattie
With this change:
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Subject: dhcpd: Fix apparmor profile
References: bnc#692428
This patch adds the network rules needed, corrects the path to dhcpd.leases,
and adds the path for TSIG DNS keys.
Reported-by: Andrew Beames <suseforum@roocomputing.co.uk>
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
updated to match trunk by
Christian Boltz <apparmor@cboltz.de>
updated to use
/var/lib/dhcp/{db/,}dhcpd.leases* rwl,
(instead of just /var/lib/dhcp/db/dhcpd.leases* rwl) to keep the profile
Ubuntu-compatible as requested by Steve Beattie.
With this change:
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Subject: profiles: Add openssl abstraction
References: bnc#623886
Profiles that use openssl have been adding the openssl files piecemeal.
This patch creates a new openssl abstraction that can be inherited by
all profiles that use it.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Patch for
- profiles/apparmor.d/abstractions/ssl_certs
- profiles/apparmor/profiles/extras/usr.sbin.httpd2-prefork (second chunk)
updated by Christian Boltz <apparmor@cboltz.de>
(didn't apply to trunk)
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Copyright header in profiles/apparmor.d/abstractions/openssl added by
Christian Boltz <apparmor@cboltz.de>
Subject: profile: ntpd -N needs sys_nice
References: bnc#657054
ntpd -N allows the administrator to increase or decrease priority of the
ntp server. Since the profile doesn't allow it, the operation is denied.
This patch adds support for that operation.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
along with a link to the wiki page. This helps users locate profiles
that possibly already exist for the application they are attempting to
confine, and suggests they contribute the profile when they're done.
apparmor_2.6.1~rc1 to apparmor_2.6.1-rc1
apparmor_2.6.0~rc1 to apparmor_2.6.0-rc1
to be consistent with the new format and to enable mirroring to a git
repository
'~' replaced by '-'. This is needed for mirroring to git as git can't
handle '~'s embedded in tag or branch names.
Tested by setting up a separate tag_version target like so:
tag_version:
echo ${TAG_VERSION}
This patch fixes the init scripts helper functions file to
filter out the hat/child process separator as currently used
by the parser, '//' rather than what used to be used, the '^'
symbol. This fixes bugs where profiles that covered regexs (e.g.
'/usr/lib/firefox-4.0.1/firefox{,*[^s][^h]}') and thus were being
improperly filtered away and unloaded when reloading apparmor policy.
binary portions (Hurd). This patch splits up the build targets so this is
possible:
"main" becomes "arch"
"indep" is created and depends on "docs"
po building is moved from "main" to "indep"
"all" has "tests" removed (standard build practices are to "make" then
"make check" so I think "tests"/"check" should stay separate from "all").
redundant chunk is removed (this exists twice in the Makefile):
-.SILENT: check
-check: tests
"install" is split into "install-indep" and "install-arch"
"install-arch" requires "arch" and only installs the binaries
"install-indep" requires "indep" and only install non-binaries
Additionally, update the README to mention the "check" target both for the
parser and the utils.
Signed-off-by: Kees Cook <kees.cook@canonical.com>
Acked-by: Steve Beattie <steve@nxnw.org>
Some non-Linux systems do not define PATH_MAX (Hurd). Since I have no
interest in supporting a fully dynamic PATH_MAX in AppArmor, work around
this by just defining a static value that matches Linux's limits.h value.
Signed-off-by: Kees Cook <kees.cook@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
translations to be built via the LANGS make argument whitelist. For
example:
cd parser; make all install "LANGS=en_US fr"
will build and install the en_US and fr .mo files for the parser.
Allow dumping out which states where dropped during partition minimization
and which state became the partitions representative state.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
The dfa graph dump was broken by previous dfa cleanups so that the graph
transition target is the output of a pointer instead of the dfa state
number.
Signed-off-by: John Johansen <john.johansen@canonical.com>
This is a rather large rearrangement of how a subset of the parser global
variables are defined. Right now, there are unit tests built without
linking against parser_main.c. As a result, none of the globals defined in
parser_main.c could be used in the code that is built for unit tests
(misc, regex, symtab, variable). To get a clean build, either stubs needed
to be added to "#ifdef UNIT_TEST" blocks in each .c file, or we had to
depend on link-time optimizations that would throw out the unused routines.
First, this is a problem because all the compile-time warnings had to be
explicitly silenced, so reviewing the build logs becomes difficult on
failures, and we can potentially (in really unlucky situations) test
something that isn't actually part of the "real" parser.
Second, not all compilers will allow this kind of linking (e.g. mips gcc),
and the missing symbols at link time will fail the entire build even though
they're technically not needed.
To solve all of this, I've moved all of the global variables used in lex,
yacc, and main to parser_common.c, and adjusted the .h files. On top of
this, I made sure to fully link the tst builds so all symbols are resolved
(including aare lib) and removedonly tst build-log silencing (for now,
deferring to another future patchset to consolidate the build silencing).
Signed-off-by: Kees Cook <kees.cook@canonical.com>
Johansen, use 'ix' instead of 'Pix' for dbus-launch since if someone happens to
define a profile for dbus-launch and it is loosely confined, then users of this
abstraction could end up launching a program via dbus-launch in a less confined
manner than intended. This sort of thing should not be possible via an
abstraction (and people are always free to profile using Pix if they prefer).
since things go extremely badly when capabilities disappear. If someone
wants to work on it, I have some initial patch attempts, but it was getting
too time-consuming, so I back-burnered the parser. A very small change was
needed to get the libraries to build, and this is it.
Description: Workaround non-Linux environments to build everything but the
parser.
Author: Kees Cook <kees@debian.org>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
- move cap_sys_module and cap_sys_rawio to "dangerous" capabilities
- sorted sdKapKeyDanger
Proposed by Seth Arnold,
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
about an actual bug in the parser; namely that when handling strings
encapsulated in quotes, that our handling of octals is busted. It
fixes this by fixing the case entries so that 3 digit octals will
get parsed correctly, rather than dropped.
It also adds a bunch of unit tests for the processquoted() function.
32bit arch, due to size_t objects being passed to fprintf with format
strings expecting longs. It does this by adjusting the fprintf rules
to expect size_t objects.
the default compilation rules when compiling C++ files, so that things
like CFLAGS et al will be honored. Without this, doing 'make DEBUG=y'
in the parser/ tree will not have its added -pg flag honored, breaking
profiling of the parser.
Basically the files will generate apparmor.vim as included in openSUSE
11.4 (and posted here before at the end of january). The only difference
is that the patch that Steve posted some days ago is already included
(patch summary: sdGlob: first character of variable name has to be
:alpha:, followed by any number of :alnum: or _)
stress tests on the parser, by dropping the maximum number of rules
each profile can have, as well as reducing the number of profiles to
generate by default to 50. It also cleans up the emitted profiles
a little, creates the profile names with the suffix .sd [1], fixes
stress.sh to actually honor the -p (alternate parser) argument, fixes
the profile flags generation to not generate duplicates flags, and
fixes the file rules to always start with a constant randomly-generated
prefix element (rather than a regex or variable) to greatly reduce
the possibility of X dominance collisions in the parser
'[[:alpha:]][[:alnum:]_]*' (i.e. a single alpha followed by any number
of alphanumerics or underscores). Unfortunately, the code that expends
variables inside a profile does not match this, it incorrectly matched
'([[:alpha:]]|_)+' (one or more alphas or underscores). This patch
corrects the behavior there as well as synchronizing the expected
variable names in the apparmor.d manpage and apparmor.vim syntax file.
It also adds unit tests and testcases to verify the behavior.
Signed-off-by: Steve Beattie <sbeattie@ubuntu.com>
process does not generate local files for things in extras, and even if
it did, this one is named in a non-standard fashion (usr.bin.firefox vs.
usr.lib.firefox.firefox).
the extras directory as intended and fail the make if a parse failure
occurs. Also, set the default parser and logprof to be the intree ones;
the system ones can still be used by setting environment variables.
Finally, have the 'all' target generate the local files. Also, set the
parser base directory to the apparmor.d directory (rather than as an
added include, to avoid outside contamination from system profiles and
includes).
With these changes, make && make check should verify the profile set is
compilable and mostly consistent. (Alas, the current profiles are not
quite consistent).
built libapparmor, as well as working around libtool so that the
libapparmor library build directory does not get added as an rpath to
the module.
Bug: https://launchpad.net/bugs/737074
Split hfa into hfa and compressed_hfa files. The hfa portion focuses on
creating an manipulating hfas, while compressed_hfa is used for creating
compressed hfas that can be used/reused at run time with much less memory
usage than the full blown hfa.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Split out the aare_rule bits that encapsulate the convertion of apparmor
rules into the final compressed dfa.
This patch will not compile because of the it needs hfa to export an interface
but hfa is going to be split so just delay until hfa and transtable are
split and they can each export their own interface.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Start of splitting regexp.y into logical components instead of the mess
it is today. Split out the expr-tree and parsing components from regexp.y
int expr-tree.x and parse.y and since regexp.y no longer does parsing
rename it to hfa.cc
Some code cleanups snuck their way into this patch and since I am to
lazy to redo it, I have left them in.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
If the apparmor_parser is updated (outside of current packaging), when
doing profile loads it will use the existing cache of compiled profiles,
instead of forcing a recompile on profiles.
This can cause apparmor to load bad policy if the parser contains a bug
fix for the previous version of the parser.
This can be worked around in packaging by invalidating the cache and
forcing a profile reload when the parser is upgraded.
Signed-off-by: John Johansen <john.johansen@canonical.com>
attachment specification doesn't contain globbing.
eg.
# profile name and attachment the same - attaches as expected
profile /usr/lib/chromium-browser/chromium-browser
# profile without attachment specification - does not attach as expected
profile chromium-browser
# profile with name and attachment specification where the attachment specification uses globbing - attaches as expected
profile chromium-browser /usr/lib/chromium-browser/chromium-broswer*
# profile with name and attachment specification without globbing - FAILS to attach when it should
profile chromium-browser /usr/lib/chromium-browser/chromium-browser
This occurs because the xmatch_len is not set correctly for the profiles that specify
a name and an attachment specification, where the attachment specification does not
contain globbing characters.
In this situation the correct length for the xmatch_len is the length of the name, as
the shortest possible unambiguous match is the name length.
This patch does not fix a related bug where an attachment specification of ** will not
match (/**) will.
Older versions of the apparmor kernel patches didn't handle receiving
network tables of a larger size than expected.
Allow the parser to detect the kernel version and override the AF_MAX
value for those kernels.
This also replaces the hack using a hardcoded limit of 36 for kernels
missing the features flag.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees.cook@canonical.com>
reload. For now just special-case libvirt's profiles. If more applications
use dynamic profiles, this should be generalized in some way to flag profiles
as dynamic. (LP: #702774)
reload. For now just special-case libvirt's profiles. If more applications
use dynamic profiles, this should be generalized in some way to flag profiles
as dynamic. (LP: #702774)
reload. For now just special-case libvirt's profiles. If more applications
use dynamic profiles, this should be generalized in some way to flag profiles
as dynamic.
Rename change_hat.c to kernel_interface.c to better reflect that it
is providing multiple kernel_interfaces.
Signed-off-by: John Johansen <john.johansen@canonical.com>
The apparmor_parser translation files where using the old subdomain_parser
domain, but the parser was binding to apparmor-parser. Create a new
apparmor-parser.pot file and remove the subdomain_parser.pot file.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Output a better failure message when a conflict of x permissions cause
policy compilation to fail. We don't have enough information available
to output which rules during the dfa compilation so just improve the
message to let people know that it means there are conflicting x modifiers
in the rules.
Signed-off-by: John Johansen <john.johansen@canonical.com>
portion of apparmor has gone upstream. These patches had already been
dropped from the 2.5.x tree.
The compatibility kernel patches are still provided.
Subject: apparmor-utils: Inherit flags in sub-profiles when generating profiles
References: bnc#496204
When creating profiles with cx subprofiles, genprof will set the
sub-profile in enforce mode. When genprof cycles multiple times, it
prohibits the sub-profile from working correctly.
e.g.
# Last Modified: Mon Jan 24 13:52:26 2011
#include <tunables/global>
/home/jeffm/mycat flags=(complain) {
#include <abstractions/base>
#include <abstractions/bash>
#include <abstractions/consoles>
/bin/bash ix,
/bin/cat cx,
/home/jeffm/mycat r,
profile /bin/cat {
#include <abstractions/base>
/bin/cat r,
/home/jeffm/mycat r,
}
}
This patch allows sub-profiles to inherit the flags from the parent
profile, which allows it to be created in complain mode (if appropriate).
The temporary complain flags are cleaned up at genprof completion as
expected.
This issue was reported at: https://bugzilla.novell.com/show_bug.cgi?id=496204
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Bug: https://launchpad.net/bugs/707092
Subject: Subdomain.pm: Fix for null path
References: bnc#407959
When handling the following log entry, logprof will spew perl errors and
ultimately generate an invalid config: "r,"
Since there is nothing to do with a null path, just skip to the next entry.
type=APPARMOR_DENIED msg=audit(1214497030.421:39): operation="inode_permission" info="Failed name resolution - object not a valid entry" requested_mask="r" denied_mask="r" pid=31367 profile="/usr/sbin/httpd2-worker
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Subject: apparmor: Subdomain.pm: Fix handling of audits of unconfined processes
The version of AppArmor that was accepted into the mainline kernel
issues audit events for things like change_hat while unconfined.
Previous versions just returned -EPERM without the audit.
This results in logprof and friends spewing uninitialized value errors
when it hits events like:
type=AVC msg=audit(1291742101.899:220): apparmor="DENIED" operation="change_hat" info="unconfined" error=-1 pid=28005 comm="cron
... which happen any time an unconfined process does something with pam
when pam_apparmor is installed.
This patch skips those events.
[Note that the second half of the OpenSUSE patch had already been applied.]
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Subject: apparmor: Fix incorrect /proc/*/sys usage in usr.sbin.ntpd
References: bnc#634801
/proc/sys/kernel exists, but /proc/*/sys/kernel doesn't. This patch
fixes the profile.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Description: the Ubuntu buildds do not have the AppArmor securityfs mounted, so
the cache tests fail. This patch skips these tests if the introspection
directory is not mounted, but runs them if it is. This should allow testing of
local builds while still allowing builds on the official buildds.
Acked-By: Steve Beattie <sbeattie@ubuntu.com> - both Ubuntu and
OpenSUSE were carrying patches that disabled the caching test,
though OpenSUSE's disabled it completely rather than checking. The
parser builds need to complete even when the kernel it's building on
doesn't support AppArmor or all the extensions that the parser needs
at runtime.
Subject: apparmor-docs: Fix grammar error in techdoc.pdf
References: bnc#588235
This patch fixes a grammar error in techdoc.pdf.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Subject: apparmor-utils: Translation unification
References: bnc#586072
This patch removes small inconsistencies between identical strings to
allow for easier translation.
Reported-by: Isis Binder <isis.binder@gmail.com>
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Subject: [PATCH] apparmor-utils: cleanup after abort in genprof
References: bnc#307067
The initial generation of the base profile is required to be written out
to put the process in complain mode for observation. If the user
decides to abort the profiling session, that base profile is left
behind.
This patch removes all profiles created during the run up to an abort.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
aa-enforce and aa-complain exist to put a profile into enforce or
complain mode respectively. The /etc/apparmor.d/disable directory
already exists to drop files into it to disable profile load via
apparmor_parser (and therefore via the apparmor initscript). What
doesn't exist is aa-disable to add a file to the disable/ directory and
unload the profile. This patch does that. This version of aa-disable is
based on aa-complain (in fact doing a diff between aa-complain and
aa-disable might make review easier) and works as well as aa-enforce and
aa-complain. In other words, aa-disable has the same limitations of not
handling the specified binary properly if the specified attachment does
not match the path naming scheme (eg, the profile doesn't use the
conventional path.to.binary naming scheme, globbing is used for
attachment within the profile, etc). Also adjust documentation to
reference aa-disable.
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
This patch fixes a logprof bug where when profiles with variable
declarations at the top level (not hidden in an include) were written
back to a file, a trailing comma was being added to the declaration
statement, which is invalid apparmor policy syntax. This patch corrects
this and no longer adds the trailing comma.
Subject: apparmor: Fix network event parsing
References: bnc#665483
The upstream version of AppArmor had network mediation but it was
removed. There's a compability patch floating around that both openSUSE
and Ubuntu have applied to their kernels. Unfortunately, one part was
overlooked. The socket operation event names where changed from the
socket_ prefixed names they had when AppArmor was out-of-tree and
utils/SubDomain.pm was never updated to understand them.
This patch adds an operation-type table so that the code can just
do a optype($operation) call to discover what type of operation a
particular name refers to. It then uses this in place of the socket_
checks to decide whether an event is a network operation.
This allows genprof and logprof to work with networking rules again.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Bug: https://launchpad.net/bugs/706733
The testcases that were in place for the old style log messages have
had their expected output modified such that they are expected to
return invalid results, rather than deleting the testcases outright.
"SubDomain" in some way. This leaves only "subdomain.conf" and the
function names internally.
Additionally, I added a "make check" rule to the utils/Makefile to do a
simple "perl -c" sanity check just for good measure.
allows write to the ~/Desktop directory, which could conceivably allow writing
of .desktop files which could be clicked on and executed by the user. This is
based on the firefox base profile as included in Ubuntu. Notable features:
- allows for using the browser to navigate through directories
- allows reads from @{HOME}/Public/**
- allows writes to @{HOME}/Downloads/**
The intent of this profile is to restrict code execution, writes to $HOME
and information leaks while allowing basic web browsing and reading of
system documentation. It does not allow for plugins, extensions or other
helpers (but these can be added via the local/ mechanism).
- allow net_admin capability for DHCP server
- allow net_raw and network inet raw for ICMP pings when used as a DHCP
server
- allow read and write access to libvirt pid files for dnsmasq
See the FAQ in the dnsmasq source for details. This fixes
https://launchpad.net/bugs/697239
toolchain has gotten stricter about linking order; in short, linked
libraries need to come after the objects referring to them. Adding to
LDLIBS is the correct solution for this.
See https://wiki.ubuntu.com/NattyNarwhal/ToolchainTransition for more
details.
use by more and more applications, including empathy and evolution. It
is listed on freedesktop.org. See:
http://www.abisource.com/projects/enchant/
This abstraction gives access to enchant itself, files in the user's home
directory for enchant and various dictionaries for:
- aspell
- ispell
- hunspell
- myspell
- hspell
- zemberek
- voikko
start to use it. Additionally, the 'rw' on the @{HOME}/.config/ibus/bus/
probably only needs 'create' and 'chmod', so that could be tightened up once
those are exposed in the tools. LP: #649497.
$config->{repository}{enabled} = "no" (LP: #692406). We need to do this
since opensuse's site is down and there is no current alternative. Can
reenable once we have an alternative.
$config->{repository}{enabled} = "no" (LP: #692406). We need to do this
since opensuse's site is down and there is no current alternative. Can
reenable once we have an alternative.
When doing permission merging in the dfa minimization phase the information
about whether a rule is dominant or not has been lost so the merge of
xtransitions can not be handled correctly.
When two conflicting x transitions are merged the results are unpredicitable
and not currently detected. So default dfa minimization to set up its
initial partitions with permission hashing, this ensures that dfa states
that have different xtransitions in the minimization stage will never
be merged thus will not result in a conflict.
x permission checking is still enforced at the dfa creation phase where
the originial information is available to check whether the conflicting
permissions came from exact match or re rules so that conflict resolution
can be properly applied.
The end result is that dfa minimization does not result in a truely minimal
dfa (the minimization phase is also slightly faster).
Signed-off-by: John Johansen <john.johansen@canonical.com>
Currently apparmor provides the unsafe keyword to indicate an xtransition
is not scrubbing its environment variables. This can be used to be
explicit about which transition are unsafe instead of relying on people
remembering which of px Px is safe or unsafe.
Add the orthogonal keyword safe to allow specifying a transition is
safe.
Signed-off-by: John Johansen <john.johansen@canonical.com>
x Permissions when specified as a the start of the rule had a differnt
meaning than when they appeared at the tail of a rule.
Specifically px,cx,ux were not treated as unsafe when they appeared at
the start of the rule.
px /foo,
instead of at the tail of the rule
/foo px,
the keyword unsafe had to be used to force the rule to cause the x transitio
to be its unsafe variant.
Fix leading permissions so that they are consistent with file rules that
use trailing permissions.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Test the leading permission form of an xrule against its trailing permission
form, to verify that they are generating the same xtransition and thus
don't conflict (assumes xtransition conflict checking is working).
eg.
px /foo,
/foo px,
should generate the same rule and thus not result in any conflicts
Signed-off-by: John Johansen <john.johansen@canonical.com>
All the combiniation of xtransition conflics where not well represented in
the regression test suite. Instead of relying on multiple static test
files, automatically generate all possible conflicts.
Signed-off-by: John Johansen <john.johansen@canonical.com>
The is_merged_x_consistend macro was incorrect in that is tested for
USER_EXEC_TYPE to determine if there was an x transition. This fails
for unconfined execs so an unconfined exec would not correctly conflict
with another exec type.
The dfa match flag table for xtransitions was not large enough and not
indexed properly for pux, and cux transitions. The index calculation did
not take into account the pux flag so that pux and px aliased to the same
location and cux and cx aliased to the same location.
This would result in the first rule being processed defining what the
transition type was for all following rules of the type following. So
if a px transition was processed first all pux, transitions in the profile
would be treated pux.
Signed-off-by: John Johansen <john.johansen@canonical.com>
The dfa engine uses the defines from immunix.h for permission conflict
checking, so make the build depend on it.
Signed-off-by: John Johansen <john.johansen@canonical.com>
During some of the dfa cleanups, the checks for conflicting xtransition
was removed. This adds the conflict checking back in and makes it part
of dfa creation.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Several of the x-trans tests where failing because of the include file was
bad. This kept the test from testing what it was supposed as the test
was expected to fail. Thus hidding a bug :(
Signed-off-by: John Johansen <john.johansen@canonical.com>
Add the ability to specify the name and attachment of the profile
separately. It does not allow for the attachment specification to
begin with a variable however since variables in profile names is not
currently support this shouldn't be and issue.
Signed-off-by: John Johansen <john.johansen@canonical.com>
clean up profile parsing by merging profile and :namespace:profile parsing
into a single rule.
This also fixes a bug where the profile keyword was not allowed to proceed
profiles with a namespace declaration.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Short summary: Unloading of profiles with a space in the name fails,
therefore "rcapparmor stop" (or restart) causes a funny message - and
the profile is still loaded.
Thanks to Christian Boltz <apparmor@cboltz.de>
parser that interact with the regex DFA generation library, and thus
need to be recompiled when the header file changes.
(This patch isn't particularly of interest to distros, as they
typically won't be doing incremental compilation.)
The other changes have made it so that using a macro really isn't justified
so rework the code to get rid of the hiddeous update_for_nodes macro.
Signed-off-by: John Johansen <john.johansen@canonical.com>
With the addition of the nodes field to the state we can make the work
queue, be based off of the state instead of the node, and avoid doing
the node to map lookup to get back to the state.
This means that the NodeMap is now only used for duplicate elimination.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Factoring the updating of the state transitions doesn't save on any code
but it provides a nice logical seperation and makes the dfa work_queue
loop and the updating of the state transitions easier to understand as
units.
Signed-off-by: John Johansen <john.johansen@canonical.com>
The match_count variable is a sum of the number of duplicates node sets
that have been encountered and discarded. Rename it to better reflect what
it is doing.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Embedding the nodes are part of the state gives fast back reference from
the state to the nodes that created it. This is useful for the state to
nodes mapping dump as it lets us output the states in order. It will also
let us avoid certain nodemap lookup in the future.
Overlay the nodes field (used only in dfa construction) with the partition
field which is only used during dfa minimization to avoid making the state
any larger.
Signed-off-by: John Johansen <john.johansen@canonical.com>
commits were made (as well as a few other minor warnings elsewhere).
The Makefile change is to avoid passing -Wstrict-prototypes and
-Wnested-externs to the C++ compiler, which the compiler yells about and
then ignores.
Since we compile with -Wmissing-field-initializers I dropped the
unreferenced zero-width fields in the header structs, and then explicitly
initialized the remaining fields.
I tagged several unused function parameters to silence those warnings.
And finally, I dropped the unused filter_escapes() too.
Embedding the the partition mapping into the State structure significantly
speeds up dfa minimization, by converting rbtree finds to straight direct
references when checking for same mappings.
The overall time improvement is small but it can half the time spent in
minimization.
The nodemap.size() increases by one with each node added, every time we
add a state we label it so this provides the proper labeling without needing
a separate variable.
add short options to turn on all stats, and all progress indicators,
also allow adding "no-" prefix to dump options to allow subtracting
individual options when short options are used.
eg.
-D stats -D no-expr-simplify
Move the -O and -D options into tables, that keep the option and its
description. This will help keep the options consistent and the description
up to date, as all information is now in one place.
Previously the options, and descriptions kept getting out of sync as all
relavent parts were spread out.
help reduce peak memory usage in some cases.
Also disbale remove_unreachable, as the current dfa code isn't generating
unreachable states, and minimization removes any states that are connected
but redundant.
hold permission information. We currently keep them in a table with a
refcount so that they don't go away, until we delete the table.
We can simulate this by getting rid of the refcount, and making dup and release
virtual, and overriding it for the special accept nodes.
improves minimization performance, it can slow down total creation time and
result in larger compressed dfas.
This is because it results in the dfa not being completely minimized which
with the current O(n2) dfa table compression algorithm can result in slower
compressed dfa generation.
first hash does hashing on state just state transitions, which always results
in a performance improvement.
The second does hashing based off of accept permissions, which can create
more initial states but can result in not being able to achieve a true
minimum dfa. This can also lead to slowing down total dfa creation because
while minimization, compression can take longer if the dfa isn't completely
minimized.
permission hashing is currently required, as minimization does not accumulate
redundant Node permissions.
memory than just using the NodeSet size to short circuit comparison but it
improves on the case where compared sets have the same size. It is possible
that this will slow down small dfa generation slightly but the trade off for
large dfa's (which are the slow ones to generate) is worth it.
This results in another performance bump over using the NodeSize is NodeSet
comparison, and the amount of improvement increases with larger dfas
of pointers when it isn't necessary. This results in a nice little
performance increase in dfa creation.
This is more of a proof of concept patch, and is replaced by the next
patch which does better short circuiting via hashing
the large side, and I experimented with different ways to split this up but in
the end, anything I could do would result in a series of dependent patches
that would require all of them to be applied to get meaningful functional
changes.
The patch structural reworks the dfa so that
- there is a new State class, it takes the place of sets of nodes in the
dfa, and allows storing state information within the state
- removes the dfa transition table, which mapped sets of nodes to a
transition table, by moving the transition into the new state class
- computes dfa state permissions once (stored in the state)
- expression tree nodes are independent from a created dfa. This allows
computed expression trees, and sets of Nodes (used as protostates when
computing the dfa). To be managed independent of the dfa life time.
This will allow reducing the amount of memory used, in the future,
and will also allow separating the expression tree logic out into
its own file.
The patch has some effect on reducing peak memory usage, and computation
time. The actual amount of reduction is dependent on the number of states
in the dfa with larger saving being achieved on larger dfas. Eg. for
the test evince profile I was using it makes the parser about 7% faster with a
peak memory usage about 12% less.
This patch changes the initial partition hashing of minimization resulting
in slightly smaller dfas.
of a mix of symlinks to non-prefixed comands, and "apparmor_" prefixed
commands.
This also refactors the manpage generation slightly since we no longer
need special cases for the manpages, and drops aa-eventd from the default
list of tools to install (it also lacks a manpage).
exported in the environment. Without it, merely setting the CFLAGS
environment variable would not affect the compilation of the parser,
though it was still possible to override it by passing the variable
as an argument (e.g. make all CFLAGS="-Oinsane -Wextra-special").
It also makes the default CFLAGS for the parser consistent with
the default for the C++ dfa library, and passes the flags on to
the library.
An audit of the other bits of C showed that they either supported
CFLAGS during configure or were otherwise honoring CFLAGS when set
as environment variable.
the common/Make.rules file into common/Version so that libapparmor's
configure.in can make use of it, meaning there's one less thing to
adjust when updating the version. It also bumps the trunk version
from 2.5 to 2.5.90 in (perhaps excessively long) preparation for the
2.6.0 release, and to indicate that it's newer than the 2.5.x branch.
tarballs and converts some of the common/Make.rules targets to get
version information from bzr. As part of this, the tarball generation
creates a .stamp_rev file in the common directory which contains both
the name of the bzr repo exported from as well as the revision.
This just adds prototypes to all functions to make further cleanup
slightly easier by getting perl to complain if not enough args are
passed to a function. Perl doesn't appear to complain about this in
every case even with prototypes, which is kind of annoying.
One of the uses of eval { } wasn't checking $@ for errors, so if
something bad happened, it'd be silently ignored. This just adds in
an extra check to die if we hit a failure.
loopback mounted filesystem to operate on, to guarantee that the mount
option user_xattr is enabled (it's disabled by default on Ubuntu).
With this change, a number of the user xattr testcases that were
expected to pass but weren't started working; however, some of the
ones that were failing as expected are now passing. I've touched up
the expectations as well.
script and add an additional sleep before the parser invocation that
generates the cache file for the first time, to avoid failures in the
"Profiles are cached when requested:" test on ext3 and other filesystems
without fine-grained enough timestamps.
parsing, and precompilation of policy. This allows finding the most
recent text time stamp during parsing and this is then compared to
the cache file time stamp.
While this is slightly slower than the cache file check that only
validated against the profile file it fixes the bug where abstraction
updates do not cause the cache file to become invalid.
capability is reported by LSM_AUDIT and is just the capability number.
capname is reported by the apparmor module and is the name the kernel
knows the capability as.
For now just use capname and silently drop capability when it is found.
started. Since apparmor_notify is not installed by default and not started
by default, the act of installing and starting it implies the desire to
get messages.
https://bugs.launchpad.net/apparmor/+bug/623467
This patch adds some additional testcases to the log parsing
testsuite, to cover rejections for operations that aren't covered by
other testcase (truncate, rename_src, rename_dest, mkdir) as well
as fixing SubDomain.pm to take those operations into account when
parsing log files.
The operations link, unlink, and possibly setattr still need to be
covered by SubDomain.pm
kernel when the hat that was passed does not exist in the profile (but
other hats exist). It also removes the very old EPERM case, which hasn't
been accurate for a while. (LP: #619521)
this results in
Unable to open output file - Success
to be output to standard error.
This occurs because despite specifying kernel_load = 0, the kernel load
parts are still being done, and failing.
- create profiles/apparmor.d/local/README to explain it all
- adjust shipped profiles in profiles/apparmor.d to include the local changes
- adjust profiles/Makefile for local files
- add profiles/local/README
- adjust profiles/apparmor.d/{bin,sbin,usr}* to include a file from local/
- adjust profiles/apparmor.d/{bin,sbin,usr}* for for copyright, some whitespace
and svn conventions
the testsuite. It looks like coredump mediation may have been removed,
since it is rather a corner-case, so I have currently marked it as XFAIL.
In hooking it back up, the "prologue.inc" was reviewed, dead code dropped,
and the "image=" argument changed to correctly handle the imageperms
syntax used elsewhere. It was working in other tests out of coincidence.
tree. It is limited in that it doesn't currently handle the permissions of a rule.
conversion output presents an aare -> prce conversion followed by 1 or more expression
tree rules, governed by what the rule does.
eg.
aare: /** -> /[^/\x00][^\x00]*
rule: /[^/\x00][^\x00]* -> /[^\0000/]([^\0000])*
eg.
echo "/foo { /** rwlkmix, } " | ./apparmor_parser -QT -D rule-exprs -D expr-tree
aare: /foo -> /foo
aare: /** -> /[^/\x00][^\x00]*
rule: /[^/\x00][^\x00]* -> /[^\0000/]([^\0000])*
rule: /[^/\x00][^\x00]*\x00/[^/].* -> /[^\0000/]([^\0000])*\0000/[^/](.)*
DFA: Expression Tree
(/[^\0000/]([^\0000])*(((((((((((((<513>|<2>)|<4>)|<8>)|<16>)|<32>)|<64>)|<8404992>)|<32768>)|<65536>)|<131072>)|<262144>)|<524288>)|<1048576>)|/[^\0000/]([^\0000])*\0000/[^/](.)*((<16>|<32>)|<262144>))
This simple example shows many things
1. The profile name under goes pcre conversion. But since no regular expressions where found
it doesn't generate any expr rules
2. /** is converted into the pcre expression /[^\0000/]([^\0000])*
3. The pcre expression /[^\0000/]([^\0000])* is converted into two rules that are then
converted into expression trees.
The reason for this can not be seen by the output as this is actually triggered by
permissions separation for the rule. In this case the link permission is separated
into what is shown as the second rule: statement.
4. DFA: Expression Tree dump shows how these rules are combined together
You will notice that the rule conversion statement is fairly redundant currently as it just
show pcre to expression tree pcre. This will change when direct aare parsing occurs,
but currently serves to verify the pcre conversion step.
It is not the prettiest patch, as its touching some ugly code that is schedule to be cleaned
up/replaced. eg. convert_aaregex_to_pcre is going to replaced with native parse conversion
from an aare straight to the expression tree, and dfaflag passing will become part of the
rule set.
It changes the table resizing so that there is always sufficient
high entries in the table, preventing bounds violations from
occurring.
Previously the resize allocation was always based on the character
set range for a state, which could be more or less than actually
required, and packing would waste some space when over allocation
was done.
As a result this patch in general results in slightly smaller
transition tables even though it enforcing the minimum required
padding to avoid bounds violations.
are computed and stored in a map, that is not cleaned up. This means that the labeling
is retained across different dfas.
Move the labeling into expr node as this takes less memory than using a map and will
also separates node labeling so its per dfa instead of global. In addition this means
the labeling is cleanedup/freed when the expr tree is freed without any extra work.
each expression tree node and then used as input to create the dfa states.
Currently they are not being freed until the nodes are destroyed, but the information
is no longer needed once the dfa has been created. Cleaning them up early reduces
peak memory usage.
* a non-include related syntax error (errors/modefail.sd)
* multiple successful includes followed by a failed include
(errors/multi_include.sd)
It also fixes two issues with the parser's line counting:
* the count began at 0 (demonstrated by the first testcase's error
being reporting on one line less than it should be), and
* an extra line increment when includes were detected (demonstrated
by the second testcase's error being reported at a line beyond the
correct linenumber.
The existing testcases did not catch these because they were all
based on the first include in the file failing and so the start of
the count from 0 counteracted the extra counted line.
paths. Secondly, the /lib64 -> /lib symlink would mean the
/lib64/ld-linux symlink would incorrectly be generated as
/lib64/ld-N.NN.so which still has a symlink in its path, and thus
apparmor wouldn't permit the access. Fixing by having readlink
canonicalize the entire path.
ack thppt.
when auditd logs get rotated)
- use getgrnam() with setgid when dropping to nobody_group
- add '-u USER' option to drop to this user when running priviliged but
not under sudo. Useful for starting when logged in as root.
- add a read access check before get_logfile_inode() so we don't have to
wait for the timeout in get_logfile_inode()
- set euid only when dropping privileges, instead of using POSIX::setuid()
which sets uid, euid and saved id when starting privileged
- create send_message() function which fork/execs so that we can set the
real uid before calling notify-send (notify-send looks at the real uid
when trying to connect to dbus)
- adjust reopen_logfile() to raise privileges (via euid) before accessing
logfile when $< != $>. Drop them again after open().
- also check for inode change
- update size to use stat
- treat logfile_size like logfile_inode
- update logfile_size and logfile_inode in reopen_logfile()
- add -f option to optionally specify the logfile
- when polling, check to see if the logfile size decreased, and if so, reopen
it. Currently this only works if you can read the file after dropping
privileges
tests. I don't like the solution because it exposes a data structure
definition outside of the only file that should know it's layout.
Also, fixed the Makefile to fail the build when one of the unit test
programs fails. :-(
killall-ing a few things in order to make it stop. And alas, it does seem
to eventually cause kernel hangs with 2.6.32-16. (Committing now before ext4
eats my changes and brain.)
Instead of updating the profile name, allow a profile to have multiple
alternate names. Aliases are now added as alternate names and matched
through the xmatch dfa.
Alias was broken because it when an alias was made the old path was completely
removed and there was no way to specify it. Update it so aliases just add
an new duplicate rule instead.
These replacement routines allow an application to avoid the probing
behavior of earlier version of change_hat. Allowing them to be faster
and have better learning characteristics.
xfail instead of known_{pass,fail}, also have it only reports unexpected
results, error for when result != what it should, and Alert for when it
result is what is should be but is a known problem and hence expected
to report something else.
Also update the regression tests for known problems under AppArmor 2.5,
this does not fix all known problems, (ie hats being removed differently
and hence resulting in unable to load profile errors, and the mknod
problem on alternate runs of the test suite, nor xattrs tests not ensuring
that the fs supports xattrs).
imediately after the current partition being considered, instead of
at the back of the parition list. This does two things, it makes it
more likely the data is in cache, and it also in general results in
more partitions being created in a single pass.
not an algorithmic improvement. It does the same basic algorithm of
test until it can insert the data, but instead of only tracking the
first free entry (and recomputing it each pass). It tracks all
free entries reducing the number of comparisons done and the table
grows in size.
This may actually result in a small loss on small tables, but is a win
for larger tables.
Update the hash calculation to guarentee that states with a different
number of transition entries will be placed in seperate partitions.
This will allow for a better character transition based state comparison.
Add basic Hopcroft based dfa minimization. It currently does a simple
straight state comparison that can be quadratic in time to split partitions.
This is offset however by using hashing to setup the initial partitions so
that the number of states within a partition are relative few.
The hashing of states for initial partition setup is linear in time. This
means the closer the initial partition set is to the final set, the closer
the algorithm is to completing in a linear time. The hashing works as
follows: For each state we know the number of transitions that are not
the default transition. For each of of these we hash the set of letters
it can transition on using a simple djb2 hash algorithm. This creates
a unique hash based on the number of transitions and the input it can
transition on. If a state does not have the same hash we know it can not
the same as another because it either has a different number of transitions
or or transitions on a different set.
To further distiguish states, the number of transitions of each transitions
target state are added into the hash. This serves to further distiguish
states as a transition to a state with a different number of transitions
can not possibly be reduced to an equivalent state.
A further distinction of states is made for accepting states in that
we know each state with a unique set of accept permissions must be in
its own partition to ensure the unique accept permissions are in the
final dfa.
The unreachable state removal is a basic walk of the dfa from the start
state marking all states that are reached. It then sweeps any state not
reached away. This does not do dead state removal where a non accepting
state gets into a loop that will never result in an accepting state.
For now just look at 'name=...' which is usually the last in the log entry,
so validate input against this and output based on it.
TODO: better handle other cases too
This will allow turning on and off various debug dumps as needed.
Multiple dump options can be specified as needed by using multiple
options.
eg. apparmor_parser -D variables
apparmor_parser -D dfa-tree -D dfa-simple-tree
The help option has also been updated to take an optional argument
to display help about give parameters, currently only dump is supported.
eg. apparmor_parser -h # standard help
apparmor_parser -h=dump # dump info about --dump options
Also Enable the dfa expression tree dumps
thing again. Fix to use the kernel's definition of AF_MAX in
linux/socket.h if it's larger than glibc's AF_MAX definition in
sys/socket.h and add a wrapper function so that we don't have include
af_names.h everywhere.
Also, fix memory leaks around the handling of network entries of
policies.
Change_profile was broken so that it couldn't parse expressions that
weren't path based or started with a variable. Furthermore if the name
held any expressions it was not hanlded correctly, as it was being passed
directly to dfa conversion without going through glob -> pcre conversion.
* move network families to filter out into a separate variable to
so that the list doesn't get lost in a complex sed invocation
* pull out the actual macro definitions from linux/socket.h and use
them if glibc's sys/socket.h (really bit/socket.h) hasn't caught up
with the family definitions.
- Update matching regex for reordered kernel audit messages (when they
come through syslog). Ideally, rather than use a regex, the utils would
just use the log parsing library to determine whether it's a log even
of interest.
- fix debugging code write a logfile in /var/log/apparmor and not a
predictable location in /tmp; File::Temp would be the right solution
except that the log file is created in a BEGIN clause, and
File::Temp.new() ends up returning an unopened filehandle in that
situation, so logging fails. Someone with more perl-fu may know how to
fix that.
- fix "Namespcae" tyop
- get rid of sub_name and default_deny from the main profile struct as
they haven't been used for a long time; also eliminates their output
from the debugging output.
- emit dumped parsing structure with only one -d, users were confuzzled
and it was not documented that you needed to use -dd to get it to
output anything if DEBUG wasn't set when compiling.
doesn't work. Making the parser match the documentation, though either
form should still work.
(Based on a secondary element of https://bugs.launchpad.net/bugs/341205)
Change the globbing conversion to include [^\x00]. This reduces cases of
artifical overlap between globbing rules, and link rules. Link rules
are encoded to use a \0 char to seperate the 2 match parts of the rule.
Before this fix a glob * or ** could match against the \0 seperator
resulting the generation of dfa states for that overlap. This of course
can never happen as \0 is not a valid path name character.
In one example stress policy when adding the rule
owner /** rwl,
this change made the difference between having a dfa with 55152 states
and one with 30019
- disable charter, charset merging. This can actually hamper optimization
in some cases and needs special cases added to the factoring code.
The charset code is merged off into its own routines that can be
reenabled at a later time.
- fix a couple bugs in tree simplifications that would cause earlier
exit before the tree had even reached a local minima
I particular the t != c portion of the simplify_tree, would cause
the loop to exit early if it didn't change but other modifications
had been made.
- remove the extra epsnode that was getting added to the created tree
- optimize the forward factor alt loop so that it will find all left
factor matches down the alt subtree without having to loop and recompare
against nodes that were already checked
These changes result in small improvements in most cases, but in some
policies the changes result in very large wins. The early bailout of
optimizations was causing 2.5* as many dfa states in one particular
stress test policy.
- fix-complain.diff
Fixes deny rules in complain mode so that they don't reject events
- mount-capability.diff
Allow confined applications to mount and unmount as long as they
have capability sys_admin
- fix-config.diff
Add the missing SECURITY_NETWORK dependency
- fix-security-param.diff
Make apparmor respect the security= parameter
- securit_default.diff
Add a new kernel config option to allow setting the default LSM,
When multiple LSMs are compiled into the kernel this is often
more desirable than taking the first LSM to register
- fork-tracking.diff
Newer kernels have changed the allocation of child pid until after
the security_clone hook. This breaks AppArmor's fork tracking
for processes that enter the null-complain-profile.
To fix this the parent pid is output with every message. A corresponding
update in the tools also must be done.
- fix-d_namespace_path.diff
It is possible that the root.mnt->mnt_ns has been unmounted, resulting
in an oops. In this case just test for it, and if it happens the
ns_root.mnt passed to __d_path will be NULL resulting in a disconnected
path.
- AppArmor-misc-cleanups.diff
Some miscelleanous cleanups from Miklos Szeredi, covering some
kernel coding style and defaults cleanups
- AppArmor-checkpatch.diff
patch from Miklos Szeredi, to cleanup sparse warnings, and other misc
coding style errors.
Needs love in the form of enhancements to support regexs, all the added
features in upcoming 2.3 release, etc.
Could also stand a bit of refactoring to make the ruby program not suck
up so much ram by writing out profiles as things go along rather than
generating a bunch of large objects and keeping them around.
Fork tracking is broken by the kernel message for clone no longer supplying
the child pid correctly. Instead the parent pid will be output with each
message and the tools will fake a fork when they detect a previously
unknow parent child relationship.
/usr/bin/rsyncd) bnc#408869
The unconfined tool shows:
[...]
29799 /usr/bin/rsync not confined
29799 /usr/bin/rsync not confined
This is because unconfined is grabbing the post symlink resolved exe filename
which for /usr/sbin/rsyncd is /usr/bin/rsync.
To fix this provide both the cmdline and exec name in parenthesis when the
exe name and the cmdline name differ.
For the above example you would see
29799 /usr/bin/rsync (/usr/sbin/rsyncd) not confined
been made but only from the top level. This allows us to get the
optimizations that were missed, while not causing the massive recursive call
explosion we had before.
Description: fix compile on build
Patch from Gentoo community:
- fix up a couple of missing semicolons in syntax (bison compensates
by emitting it's own)
- Fix yet another variable tyop in rc.apparmor.functions
- dump stderr of ls in rc.apparmor.functions to /dev/null
- add an install-unknown make target
alias rules and variable declarations within the preamble of a profile.
This commit adds another testcase for alias rules; one in which there is
an overlapping pair of aliases. The parser parses it, but based on -dd
output, I don't believe it's treating it properly.
Apply tree factoring and simplification techniques to reduce the number of
states used in computing the dfa. This can have an exponential impact
on both space and time for dfa generation.
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
The format of audit messages that are redirected to syslog because
auditd isn't running changed between Hardy and Intrepid and now have
the type=NNNN field before the audit tag like--
Nov 1 22:24:43 box kernel: [ 158.113592] type=1503
audit(1225603483.635:5): operation="inode_permission" requested_mask="r::"
denied_mask="r::" fsuid=7 name="/proc/7034/net/" pid=7034
profile="/usr/sbin/cupsd"
I believe this patch will address the moved type=NNNN field as well as
capturing non-matching logfile input instead of printing it to stdout.
Patch modified by Steve Beattie <sbeattie@ubuntu.com> to take into
account a couple of different situations.
https://bugs.launchpad.net/bugs/271252https://bugzilla.novell.com/show_bug.cgi?id=441381
- flags being dropped from hats
- rules can be poorly split on writing the profile
- identical rules with different permissions are not properly combined, so
that only permissions of the last rule are kept
Added missing capabilities names.
Simple rlimits support. It doesn't care about range of individual limit,
you can add ie -100G stack size or 100M nice. But maybe sufficient?
of rlimits supported by the kernel.
- remove hat rules
- add hat flag for each profile
- fix apparmorfs profile listing code. Used to only return the first
80 or so profiles, and then refuse to output more
removing profiles, as the list is unstable after additions or removals.
- Add the ability to loaded precompiled policy by specifying the -B
option, which can be combined with --add or --replace
- rc.apparmor.functions were not correctly removing profiles on replace and
reload, also convert to using the module interface directly bypassing the
parser.
- fix cx -> named transitions
- fix apparmor_parser -N so that it emits hats as profiles under new kernel
modules. This is the correct behavior as hats are promoted to profiles.
- rpc was failing when passing arrays because the perl is_utf8 string flag
was set even though its only sending numbers but newer HTTP::Message
checks for this is_utf8 and if it finds it aborts.
- fix local profiles
local profiles were failing because
1.) the parameters to serialize_profile were bad
2.) the file location was not getting updated so they would get written
back to the inactive profiles directory
# Copyright (c) 2005 Novell, Inc. All Rights Reserved.
# Copyright (c) 2011 Canonical, Ltd.
#
# This program is free software; you can redistribute it and/or
# modify it under the terms of version 2 of the GNU General Public
@@ -25,7 +23,7 @@ use strict;
use FindBin;
use Getopt::Long;
use Immunix::SubDomain;
use Immunix::AppArmor;
use Data::Dumper;
@@ -53,7 +51,7 @@ GetOptions(
$profiledir = get_full_path($profiledir);
unless (-d $profiledir) {
UI_Important("Can't find subdomain profiles in $profiledir.");
UI_Important("Can't find AppArmor profiles in $profiledir.");
exit 1;
}
@@ -119,7 +117,7 @@ for my $profiling (@profiling) {
UI_Info(sprintf(gettext('Can\'t find %s in the system path list. If the name of the application is correct, please run \'which %s\' as a user with the correct PATH environment set up in order to find the fully-qualified path.'), $profiling, $profiling));
exit 1;
} else {
UI_Info(sprintf(gettext('%s does not exist, please double-check the path.') . $profiling));
UI_Info(sprintf(gettext('%s does not exist, please double-check the path.'), $profiling));
# Copyright (c) 2005 Novell, Inc. All Rights Reserved.
# Copyright (c) 2011 Canonical, Ltd.
#
# This program is free software; you can redistribute it and/or
# modify it under the terms of version 2 of the GNU General Public
@@ -25,7 +23,7 @@ use strict;
use FindBin;
use Getopt::Long;
use Immunix::SubDomain;
use Immunix::AppArmor;
use Data::Dumper;
@@ -60,7 +58,7 @@ my $sd_mountpoint = check_for_subdomain();
$profiledir = get_full_path($profiledir);
unless (-d $profiledir) {
UI_Important(sprintf(gettext('Can\'t find subdomain profiles in %s.'), $profiledir));
UI_Important(sprintf(gettext('Can\'t find AppArmor profiles in %s.'), $profiledir));
exit 1;
}
@@ -109,7 +107,7 @@ for my $profiling (@profiling) {
UI_Info(sprintf(gettext('Can\'t find %s in the system path list. If the name of the application is correct, please run \'which %s\' as a user with the correct PATH environment set up in order to find the fully-qualified path.'), $profiling, $profiling));
exit 1;
} else {
UI_Info(sprintf(gettext('%s does not exist, please double-check the path.') . $profiling));
UI_Info(sprintf(gettext('%s does not exist, please double-check the path.'), $profiling));
UI_Info(sprintf(gettext('Can\'t find %s in the system path list. If the name of the application is correct, please run \'which %s\' as a user with the correct PATH environment set up in order to find the fully-qualified path.'), $profiling, $profiling));
exit 1;
} else {
UI_Info(sprintf(gettext('%s does not exist, please double-check the path.'), $profiling));
exit 1;
}
}
}
exit 0;
sub usage {
UI_Info(sprintf(gettext("usage: \%s [ -d /path/to/profiles ] [ program to have profile disabled ]"), $0));
@@ -119,7 +127,7 @@ for my $profiling (@profiling) {
UI_Info(sprintf(gettext('Can\'t find %s in the system path list. If the name of the application is correct, please run \'which %s\' as a user with the correct PATH environment set up in order to find the fully-qualified path.'), $profiling, $profiling));
exit 1;
} else {
UI_Info(sprintf(gettext('%s does not exist, please double-check the path.') . $profiling));
UI_Info(sprintf(gettext('%s does not exist, please double-check the path.'), $profiling));
# Copyright (c) 2005 Novell, Inc. All Rights Reserved.
#
@@ -24,13 +21,33 @@
use strict;
use Getopt::Long;
use Immunix::SubDomain;
use Immunix::AppArmor;
use Data::Dumper;
use Locale::gettext;
use POSIX;
sub sysctl_read($) {
my $path = shift;
my $value = undef;
if (open(SYSCTL, "<$path")) {
$value = int(<SYSCTL>);
}
close(SYSCTL);
return $value;
}
sub sysctl_write($$) {
my $path = shift;
my $value = shift;
return if (!defined($value));
if (open(SYSCTL, ">$path")) {
print SYSCTL $value;
close(SYSCTl);
}
}
# force $PATH to be sane
$ENV{PATH} = "/bin:/sbin:/usr/bin:/usr/sbin";
@@ -52,14 +69,14 @@ GetOptions(
my $sd_mountpoint = check_for_subdomain();
unless ($sd_mountpoint) {
fatal_error(gettext("SubDomain does not appear to be started. Please enable SubDomain and try again."));
fatal_error(gettext("AppArmor does not appear to be started. Please enable AppArmor and try again."));
}
# let's convert it to full path...
$profiledir = get_full_path($profiledir);
unless (-d $profiledir) {
fatal_error "Can't find subdomain profiles in $profiledir.";
fatal_error "Can't find AppArmor profiles in $profiledir.";
}
# what are we profiling?
@@ -112,6 +129,17 @@ if ($helpers{$fqdbin} eq "enforce") {
reload($fqdbin);
}
# When reading from syslog, it is possible to hit the default kernel
# printk ratelimit. This will result in audit entries getting skipped,
# making profile generation inaccurate. When using genprof, disable
# the printk ratelimit, and restore it on exit.
my $ratelimit_sysctl = "/proc/sys/kernel/printk_ratelimit";
my $ratelimit_saved = sysctl_read($ratelimit_sysctl);
END { sysctl_write($ratelimit_sysctl, $ratelimit_saved); }
sysctl_write($ratelimit_sysctl, 0);
UI_Info(gettext("\nBefore you begin, you may wish to check if a\nprofile already exists for the application you\nwish to confine. See the following wiki page for\nmore information:\nhttp://wiki.apparmor.net/index.php/Profiles"));
UI_Important(gettext("Please start the application to be profiled in \nanother window and exercise its functionality now.\n\nOnce completed, select the \"Scan\" button below in \norder to scan the system logs for AppArmor events. \n\nFor each AppArmor event, you will be given the \nopportunity to choose whether the access should be \nallowed or denied."));
my $syslog = 1;
@@ -166,7 +194,8 @@ for my $p (sort keys %helpers) {
}
}
UI_Info(gettext("Reloaded SubDomain profiles in enforce mode."));
UI_Info(gettext("Reloaded AppArmor profiles in enforce mode."));
UI_Info(gettext("\nPlease consider contributing your new profile! See\nthe following wiki page for more information:\nhttp://wiki.apparmor.net/index.php/Profiles\n"));
UI_Info(sprintf(gettext('Finished generating profile for %s.'), $fqdbin));
Subject: Call lsm hook before unhashing dentry in vfs_rmdir()
If we unhash the dentry before calling the security_inode_rmdir hook,
we cannot compute the file's pathname in the hook anymore. AppArmor
needs to know the filename in order to decide whether a file may be
deleted, though.
Signed-off-by: John Johansen <jjohansen@suse.de>
Signed-off-by: Andreas Gruenbacher <agruen@suse.de>
---
fs/namei.c | 13 +++++++------
1 file changed, 7 insertions(+), 6 deletions(-)
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -2097,6 +2097,10 @@ int vfs_rmdir(struct inode *dir, struct
if (!dir->i_op || !dir->i_op->rmdir)
return -EPERM;
+ error = security_inode_rmdir(dir, dentry, mnt);
+ if (error)
+ return error;
+
DQUOT_INIT(dir);
mutex_lock(&dentry->d_inode->i_mutex);
@@ -2104,12 +2108,9 @@ int vfs_rmdir(struct inode *dir, struct
if (d_mountpoint(dentry))
error = -EBUSY;
else {
- error = security_inode_rmdir(dir, dentry, mnt);
- if (!error) {
- error = dir->i_op->rmdir(dir, dentry);
- if (!error)
- dentry->d_inode->i_flags |= S_DEAD;
- }
+ error = dir->i_op->rmdir(dir, dentry);
+ if (!error)
+ dentry->d_inode->i_flags |= S_DEAD;
}
mutex_unlock(&dentry->d_inode->i_mutex);
if (!error) {
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.
