When split_args() calls append_arg(), the returned value needs to be
checked in order to detect memory allocation failure. Checks were
missing in two places, which are spotted by clang's static analyzer:
semanage_store.c:1352:7: warning: Value stored to 'rc' is never
read
rc = append_arg(&argv, &num_args, arg);
^ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
semanage_store.c:1368:3: warning: Value stored to 'rc' is never read
rc = append_arg(&argv, &num_args, arg);
^ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
clang's static analyzer reports a potential memory leak because the
buffers allocated in pc and fc are not freed in main(), in sestatus.c.
Free these buffers properly.
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
"sestatus -v" uses /proc/$PID/exe symbolic link in order to find the
context of processes present in /etc/sestatus.conf. For example, this
file includes "/usr/sbin/sshd".
On Arch Linux, /bin, /sbin and /usr/sbin are symbolic links to /usr/bin,
so sshd process is seen as "/usr/bin/sshd" instead of "/usr/sbin/sshd".
This causes "sestatus -v" to show nothing in "Process contexts:" for
sshd, agetty, etc.
Use realpath() to resolve any symlink components in program paths
defined in /etc/sestatus.conf. This makes "sestatus -v" show the
expected result:
Process contexts:
Current context: sysadm_u:sysadm_r:sysadm_t
Init context: system_u:system_r:init_t
/sbin/agetty system_u:system_r:getty_t
/usr/sbin/sshd system_u:system_r:sshd_t
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
This reverts commit 814631d3ae.
As reported by Petr Lautrbach, this commit changed the behavior
of selabel_open() when SELABEL_OPT_VALIDATE is 0, and this would
be an API change.
Reported-by: Petr Lautrbach <plautrba@redhat.com>
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
In getconlist.c's main(), "level" is duplicated from an optional
argument without being ever freed. clang's static analyzer warns about
this memory leak.
Free the allocated memory properly in order to remove a warning reported
by clang's static analyzer.
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
If store_stem() fails to expand the memory allocated on data->stem_arr,
some things go wrong:
* the memory referenced by "buf" is leaked,
* data->alloc_stems has been increased without data->stem_arr having
been expanded. So the next time store_stem() is called, the function
will behave as if the buffer holds enough space, and will write data
after the end of data->stem_arr.
The first issue is being spotted by clang's static analyzer, which warns
about leaking variable "stem" in find_stem_from_spec() (this function
calls store_stem()).
This both issues by freeing buf when realloc(data->stem_arr) fails, and
by not increasing data->alloc_stems when this happens.
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
When load_users() parses an invalid line with an empty level context
(ie. nothing between "level" and "range" keywords), it allocates memory
with malloc(0) and uses it. The behavior of malloc() in this case is
an unspecified behavior: it might return NULL, which would lead to a
segmentation fault.
Fix this issue by reporting the invalid entry instead. While at it,
ensure that the character before "range" is a space, and change the
logic slightly in order to avoid using "--p; ... p++;".
This issue is reported by clang's static analyzer with the following
message:
genusers.c:222:11: warning: Use of zero-allocated memory
*r++ = *s;
^
genusers.c:225:7: warning: Use of zero-allocated memory
*r = 0;
^
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
In cond_expr_to_cil(), when stack_init() fails to allocate a stack, the
function calls stack_pop() with stack = NULL. Then stack_pop()
dereferences the pointer ("if (stack->pos == -1) {"), which is NULL.
Fix this by moving the stack cleaning loop in a "if (stack != NULL)"
block.
This issue is reported by clang's static analyzer with the following
message:
module_to_cil.c:463:6: warning: Access to field 'pos' results in a
dereference of a null pointer (loaded from variable 'stack')
if (stack->pos == -1) {
^~~~~~~~~~
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
This allows sepolgen to generate policy from AVC messages that contain
contexts translated by mcstrans.
Fixes:
\# echo "type=USER_AVC msg=audit(1468415802.940:2199604): pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0-s15:c0.c1023 msg='avc: denied { status } for auid=n/a uid=0 gid=0 cmdline="/usr/lib/systemd/systemd-logind" scontext=system_u:system_r:systemd_logind_t:SystemLow-SystemHigh tcontext=system_u:system_r:init_t:s0-s15:c0.c1023 tclass=system exe="/usr/lib/systemd/systemd" sauid=0 hostname=? addr=? terminal=?'" | audit2allow
libsepol.mls_from_string: invalid MLS context SystemLow-SystemHigh
libsepol.mls_from_string: could not construct mls context structure
libsepol.context_from_record: could not create context structure
libsepol.context_from_string: could not create context structure
libsepol.sepol_context_to_sid: could not convert
system_u:system_r:systemd_logind_t:SystemLow-SystemHigh to sid
Signed-off-by: Vit Mojzis <vmojzis@redhat.com>
Keep track of line numbers for each file context in
selabel_handle. If an error occurs in selabel_fini(), the
line number of an invalid file context is echoed to the user.
Signed-off-by: Yuli Khodorkovskiy <ykhodo@gmail.com>
In permissive mode, calling restorecon with a bad label in file_contexts
does not verify the label's existence in the loaded policy. This
results in any label successfully applying to a file, as long as the
file exists.
This issue has two assumptions:
1) file_contexts must be manually updated with the invalid label.
Running `semanage fcontext` will error when attempting to add
an invalid label to file_contexts.
2) the system must be in permissive. Although applying an invalid label
in enforcing gives an error and fails, successfully labeling a file with a
bad label could cause issues during policy development in permissive.
Instead, as each context is used, verify it is valid before blindly
applying the label. If an error with validation occurs in restorecon,
application of remaining valid labels will be uninterrupted as before.
Signed-off-by: Yuli Khodorkovskiy <ykhodo@gmail.com>
Improve the processing of netifcon, genfscon, ibpkeycon, ibendportcon,
portcon, nodecon, fsuse, filecon, iomemcon, ioportcon, pcidevicecon,
and devicetreecon rules.
If the multiple-decls option is not used then report errors if duplicate
context rules are found. If it is used then remove duplicate context rules
and report errors when two rules are identical except for the context.
This also changes the ordering of portcon and filecon rules. The protocol
of portcon rules will be compared if the port numbers are the same and the
path strings of filecon rules will be compared if the number of meta
characters, the stem length, string length and file types are the same.
Based on an initial patch by Pierre-Hugues Husson (phh@phh.me)
Signed-off-by: James Carter <jwcart2@tycho.nsa.gov>
This commit resolves conflicts in values of expandattribute statements
in policy language and expandtypeattribute in CIL.
For example, these statements resolve to false in policy language:
expandattribute hal_audio true;
expandattribute hal_audio false;
Similarly, in CIL these also resolve to false.
(expandtypeattribute (hal_audio) true)
(expandtypeattribute (hal_audio) false)
A warning will be issued on this conflict.
Motivation
When Android combines multiple .cil files from system.img and vendor.img
it's possible to have conflicting expandattribute statements.
This change deals with this scenario by resolving the value of the
corresponding expandtypeattribute to false. The rationale behind this
override is that true is used for reduce run-time lookups, while
false is used for tests which must pass.
Signed-off-by: Tri Vo <trong@android.com>
Acked-by: Jeff Vander Stoep <jeffv@google.com>
Acked-by: William Roberts <william.c.roberts@intel.com>
Acked-by: James Carter <jwcart2@tycho.nsa.gov>
Unify behaviour for all module actions.
The same behaviour is already present for -i/-u/-r/-e switches.
Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1545218
Signed-off-by: Vit Mojzis <vmojzis@redhat.com>
Unify the way parameters are described in man pages and --help message.
Explain special syntax allowing the user to specify multiple modules when using
-i/u/r/E mods.
Point out that priority has to be specified in order to remove module at
different priority than 400 and that "-d" disables all instances of
given module across priorities.
Resolves: rhbz#1320565, rhbz#1337192
Signed-off-by: Vit Mojzis <vmojzis@redhat.com>
Nicolas Iooss reports:
In sepol_ibendport_key_create(), if sepol_ibendport_alloc_ibdev_name()
fails to allocate tmp_key->ibdev_name, sepol_ibendport_key_free() is
called to free the memory associated with tmp_key, which results in
free() being called on uninitialized tmp_key->ibdev_name.
This issue is reported by clang's static analyzer with the following
message:
ibendport_record.c:115:2: warning: 1st function call argument is an
uninitialized value
free(key->ibdev_name);
^~~~~~~~~~~~~~~~~~~~~
Signed-off-by: James Carter <jwcart2@tycho.nsa.gov>
Fix sizeof calculation in array iteration introduced by commit
6bb8282c4c
"libsemanage: replace access() checks to make setuid programs work"
Signed-off-by: Vit Mojzis <vmojzis@redhat.com>
libselinux and libsemanage Makefiles invoke site.getsitepackages() in
order to get the path to the directory /usr/lib/pythonX.Y/site-packages
that matches the Python interpreter chosen with $(PYTHON). This method
is incompatible with Python virtual environments, as described in
https://github.com/pypa/virtualenv/issues/355#issuecomment-10250452 .
This issue has been opened for more than 5 years.
On the contrary python/semanage/ and python/sepolgen/ Makefiles use
distutils.sysconfig.get_python_lib() in order to get the site-packages
path into a variable named PYTHONLIBDIR. This way of computing
PYTHONLIBDIR is compatible with virtual environments and gives the same
result as PYSITEDIR.
As PYTHONLIBDIR works in more cases than PYSITEDIR, make libselinux and
libsemanage Makefiles use it. And as native code is installed (as part
of the SWIG wrapper), use "plat_specific=1" in order to use /usr/lib64
on systems which distinguish /usr/lib64 from /usr/lib.
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
Acked-by: Petr Lautrbach <plautrba@redhat.com>
Export the sepol_polcap_getnum/name() functions to users of
the shared library. This will enable SETools to stop depending
on the static library.
Note that we may want to move polcaps.h up one level since
the convention is that headers directly under include/sepol are
shared library APIs while headers under include/sepol/policydb
are limited to static users. However, this will unnecessarily
break the build for existing static users so it is deferred.
Suggested-by: Chris PeBenito <pebenito@ieee.org>
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
access() uses real UID instead of effective UID which causes false
negative checks in setuid programs.
Replace access() calls (mostly tests for file existence) by stat().
Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1186431
Signed-off-by: Vit Mojzis <vmojzis@redhat.com>
F_OK access checks only work properly as long as all directories along
the path are accessible to real user running the program.
Replace F_OK access checks by testing return value of open, write, etc.
Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1186431
Signed-off-by: Vit Mojzis <vmojzis@redhat.com>
clang's static analyzer reports an out-of-bound array access in
semanage_user_roles() when num_roles is zero, with the following
statement:
strcpy(roles,roles_arr[0]);
When num_roles is zero, roles_arr[0] is not uninitialized and roles is
the result of malloc(0) so this strcpy is dangerous. Make
semanage_user_roles() return an empty string instead.
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
cil_tree_print_expr() calls cil_expr_to_string() in order to compute a
string expression into expr_str. If this function fails, expr_str is
left unitialized but its value is dereferenced with:
cil_log(CIL_INFO, "%s)", expr_str);
Prevent such an issue by checking cil_expr_to_string()'s return value
before using expr_str.
This issue has been found with clang's static analyzer.
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
"Edit" and "add" dialogues weren't closed after successful transaction
("add" and "edit" methods return "None" if successful).
Signed-off-by: Vit Mojzis <vmojzis@redhat.com>
Since Darwin systems do not have GNU sed installed, the Darwin sed is
missing the "regexp-extended" flag needed to modify the secilc markdown
files before processing with pandoc.
A quick fix for Mac users is to `brew install gnu-sed` and to use gsed.
Signed-off-by: Yuli Khodorkovskiy <ykhodo@gmail.com>
in class policy, ports is usually initialized with the result of
get_all_ports(), which is a dict. If this function failed, the error was
not fatal but ports was left initialized as a list, which would trigger
other errors when accessing it.
Initialize ports as an empty dict in order to fix this issue.
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
When using sepolgen GUI on a system with a non-MLS policy,
sepolicy.info(sepolicy.PORT) does not define "range" attributes (since
commit 908340e8e7 ("sepolicy: support non-MLS policy in manpage")).
Replace them with None in get_all_ports() result.
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
When a translated string takes parameters with %, the % and the
parameters need to be after the call to the underline function in order
to work properly.
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
It was removed from Python 3
Fixes:
File "polgengui.py", line 390, in forward
self.generate_policy()
File "polgengui.py", line 491, in generate_policy
my_policy.set_use_syslog(self.syslog_checkbutton.get_active() == 1)
File "/home/plautrba/devel/github/bachradsusi/SELinuxProject-selinux/python/sepolicy/sepolicy/generate.py", line 468, in set_use_syslog
if not isinstance(val, types.BooleanType):
AttributeError: module 'types' has no attribute 'BooleanType'
Signed-off-by: Petr Lautrbach <plautrba@redhat.com>
map() returns an iterator in python3, list in python2
Fixes:
File "/usr/lib/python3.6/site-packages/sepolicy/generate.py", line 114, in get_all_users
users.remove("system_u")
AttributeError: 'map' object has no attribute 'remove'
Signed-off-by: Petr Lautrbach <plautrba@redhat.com>
- $ gtk-builder-convert polgen.glade polgen.ui
- use get_object instead of get_widget
- use connect_signals instead of signal_connect
Signed-off-by: Petr Lautrbach <plautrba@redhat.com>
b43991f9 added direct import of sepolicy but it forgot to import
sepolicy.generate and didn't change use of generate to sepolicy.generate
Fixes:
Traceback (most recent call last):
File "/usr/bin/selinux-polgengui", line 778, in <module>
app = childWindow()
File "/usr/bin/selinux-polgengui", line 205, in __init__
self.all_types = sepolicy.generate.get_all_types()
AttributeError: 'module' object has no attribute 'generate'
Traceback (most recent call last):
File "/usr/share/system-config-selinux/polgengui.py", line 365, in forward
if self.on_in_net_page_next():
File "/usr/share/system-config-selinux/polgengui.py", line 701, in on_in_net_page_next
generate.verify_ports(self.in_tcp_entry.get_text())
NameError: global name 'generate' is not defined
Signed-off-by: Petr Lautrbach <plautrba@redhat.com>
Recent commits removed $DESTDIR from $PYSITEDIR in libselinux and
libsemanage:
-PYSITEDIR ?= $(DESTDIR)$(shell $(PYTHON) -c 'import site;
print(site.getsitepackages()[0])')
+PYSITEDIR ?= $(shell $(PYTHON) -c 'import site;
print(site.getsitepackages()[0])')
As "site.getsitepackages()" does not work within virtualenvs,
.travis.yml defines PYSITEDIR's value in it and this definition needs to
be updated too.
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
