A valid policy would not have two symbols (classes, roles, users...)
sharing the same unique identifier. Make policydb_read() rejects such
policy files.
When ..._val_to_name translation tables were allocated with malloc(),
change to calloc() in order to initialize the tables with NULLs.
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
When loading an invalid module which uses a declaration ID 0,
semodule_package crashes in policydb_index_decls():
p->decl_val_to_struct[decl->decl_id - 1] = decl;
gdb shows the following stack trace:
#0 0x00007ffff7aa1bbd in policydb_index_decls (p=p@entry=0x605360)
at policydb.c:1034
#1 0x00007ffff7aaa9fc in policydb_read (p=<optimized out>,
fp=fp@entry=0x605090, verbose=verbose@entry=0) at policydb.c:3958
#2 0x00007ffff7ab4764 in sepol_policydb_read (p=<optimized out>,
pf=pf@entry=0x605090) at policydb_public.c:174
#3 0x0000000000401d33 in main (argc=<optimized out>,
argv=0x7fffffffdc88) at semodule_package.c:220
Change policydb_index_decls() to report an error instead:
libsepol.policydb_index_decls: invalid decl ID 0
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
range transition and name-based type transition rules were originally
simple unordered lists. They were converted to hashtabs in the kernel
by commit 2f3e82d694d3d7a2db019db1bb63385fbc1066f3 ("selinux: convert range
transition list to a hashtab") and by commit
2463c26d50adc282d19317013ba0ff473823ca47 ("SELinux: put name based
create rules in a hashtable"), but left unchanged in libsepol and
checkpolicy. Convert libsepol and checkpolicy to use the same hashtabs
as the kernel for the range transitions and name-based type transitions.
With this change and the preceding one, it is possible to directly compare
a policy file generated by libsepol/checkpolicy and the kernel-generated
/sys/fs/selinux/policy pseudo file after normalizing them both through
checkpolicy. To do so, you can run the following sequence of commands:
checkpolicy -M -b /etc/selinux/targeted/policy/policy.30 -o policy.1
checkpolicy -M -b /sys/fs/selinux/policy -o policy.2
cmp policy.1 policy.2
Normalizing the two files via checkpolicy is still necessary to ensure
consistent ordering of the avtab entries. There may still be potential
for other areas of difference, e.g. xperms entries may lack a well-defined
order.
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Originally object_r's types bitmap was empty since we exempt
object_r from the normal user-role and role-type checks. CIL
however sets object_r's types to all types to avoid special case
logic. However, the kernel does not load object_r types from the
policy file; it predefines object_r and merely validates that the
object_r definition in the policy has the expected value. Thus,
the actual policy file and the /sys/fs/selinux/policy file were
differing in their object_r entry. Fix this by not writing object_r's
types to the policy file, since they are ignored by the kernel
anyway.
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Currently ebitmap_load() accepts loading a bitmap with highbit=192 and
one node {startbit=0, map=0x2}. When iterating over the bitmap,
ebitmap_for_each_bit() is expected to only yield "1" but it gives the
following bits: 1, 65, 129.
This is due to two facts in ebitmap_for_each_bit() implementation:
* ebitmap_next() stays on the first (and only) node of the bitmap
instead of stopping the iteration.
* the end condition of the for loop consists in comparing the bit with
ebitmap_length() (ie. the bitmap highbit), which is above the limit of
the last node here.
These are not bugs when the bitmap highbit is equals to
l->startbit+MAPSIZE, where l is the last node (this is how
ebitmap_set_bit() sets it). So a simple fix consists in making
ebitmap_load() reject bitmaps which are loaded with an invalid highbit
value.
This issue has been found while fuzzing semodule_package with the
American Fuzzy Lop.
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
Commit 02a7d77ef2 ("libsepol: make parsing symbol table headers more
robust") broke refpolicy build, because checkmodule generates avrule
decl blocks with "decl->symtab[i].nprim = 0" for all possible i, even
when decl->symtab[SYM_ROLES] and decl->symtab[SYM_TYPES] are not
empty.
More precisely, decl->symtab[i].nprim seems to be only updated in
libsepol/src/link.c (in *_copy_callback() functions).
Revert the buggy part of commit 02a7d77ef2 to fix this regression.
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
There is no reason to modify the number of roles defined in a policy
when no role is being inserted.
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
When running hll/pp on some invalid policy module, it can output:
libsepol.sepol_module_package_read: unknown magic number at section
1, offset: 251, number: 0x
The last number looks funny and was caused by using "%ux". "u" is not a
prefix like "l", "h", "z"... and "%x" already expects an unsigned
integer (cf. http://man7.org/linux/man-pages/man3/printf.3.html).
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
When fuzzing hll/pp, the fuzzer created a policy module with a block
which has no declaration. With block->branch_list = NULL,
typealias_list_create() triggered a NULL pointer dereference when
computing max_decl_id.
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
When hll/pp reads an invalid policy module where some scopes use
required symbols which are not defined, the program crashes with a
segmentation fault in required_scopes_to_cil():
Program received signal SIGSEGV, Segmentation fault.
required_scopes_to_cil (decl_stack=0x6040b0, block=0x607780,
pdb=0x6042e0, indent=0) at module_to_cil.c:3479
3479 for (j = 0; j < scope_datum->decl_ids_len; j++)
{
=> 0x00007ffff7a7b1a8 <block_to_cil+5224>: 44 8b 58 10 mov
0x10(%rax),%r11d
(gdb) bt
#0 required_scopes_to_cil (decl_stack=0x6040b0, block=0x607780,
pdb=0x6042e0, indent=0) at module_to_cil.c:3479
#1 block_to_cil (pdb=pdb@entry=0x6042e0,
block=block@entry=0x607780, stack=stack@entry=0x6040b0,
indent=indent@entry=0) at module_to_cil.c:3622
#2 0x00007ffff7a85a18 in global_block_to_cil (stack=0x6040b0,
block=0x607780, pdb=0x6042e0) at module_to_cil.c:3738
#3 blocks_to_cil (pdb=0x6042e0) at module_to_cil.c:3764
#4 sepol_module_policydb_to_cil (fp=fp@entry=0x7ffff79d05e0
<_IO_2_1_stdout_>, pdb=0x6042e0, linked=linked@entry=0) at
module_to_cil.c:4051
#5 0x00007ffff7a86b55 in sepol_module_package_to_cil
(fp=fp@entry=0x7ffff79d05e0 <_IO_2_1_stdout_>, mod_pkg=0x604280) at
module_to_cil.c:4080
#6 0x0000000000401acc in main (argc=<optimized out>,
argv=<optimized out>) at pp.c:150
(gdb) p scope_datum
$1 = (struct scope_datum *) 0x0
Detect such errors and exit with an error return value.
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
When hll/pp operates on an invalid policy module which defines blocks
with non-empty decl->symtab[SYM_COMMONS], additive_scopes_to_cil_map()
calls func_to_cil[SYM_COMMONS], which is NULL.
In additive_scopes_to_cil(), filter out NULL elements of func_to_cil
before calling additive_scopes_to_cil_map().
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
1. Use the new helper to convert from AVRULE to AVTAB values.
2. Only check once for invalid AVRULE specified parameter.
3. Drop assert and just return error on invalid specification.
Signed-off-by: William Roberts <william.c.roberts@intel.com>
General clean up for expand_avrule_helper:
1. Minimize the conversions of AVRULE specification to AVTAB specification,
they are almost the same, the one exception is AVRULE_DONTAUDIT.
2. Clean up the if/else logic, collapse with a switch.
3. Move xperms allocation and manipulation to its own helper.
4. Only write avkey for values that change.
5. Return error rather than assert on invalid specification.
Signed-off-by: William Roberts <william.c.roberts@intel.com>
Rather than having multiple copies of the AVTAB and AVRULE
defines, consolidate them.
This makes it clear that AVRULE to AVTAB conversion no longer
need to occur.
Signed-off-by: William Roberts <william.c.roberts@intel.com>
When hll/pp loads a policy file which has been modified so that the
nprim field of one of its non-empty symbol table was changed to zero, it
crashes with a segmentation fault. A quick analysis leads to
"p->sym_val_to_name[i] = (char **)alloc(p->symtab[i].nprim, sizeof(char
*));" in policydb_index_others(), which is not executed when
p->symtab[i].nprim is zero even though there are items in
p->symtab[i].table.
Detect such an oddity in the policy file early to exit with a clean
error message.
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
While fuzzing hll/pp, the fuzzer (AFL) crafted a policy which triggered
the following message without making the policy loading fail (the
program crashed with a segmentation fault later):
security: ebitmap: map size 192 does not match my size 64 (high bit
was 0)
This is because ebitmap_read() returned -EINVAL and this value was
handled as a successful return value by scope_index_read() because it
was not -1.
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
When fuzzing hll/pp inputs, a policy module where the value of
scope->decl_ids_len has been modified to zero makes the program abort
(when it has been compiled without -DNDEBUG).
Change the behavior to report an error message instead. This eases
fuzzing functions like policydb_read().
Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
The combining logic for dontaudit rules was wrong, causing
a dontaudit A B:C *; rule to be clobbered by a dontaudit A B:C p;
rule.
This is a reimplementation of:
commit 6201bb5e25 ("libsepol:
fix checkpolicy dontaudit compiler bug")
that avoids the cumbersome pointer assignments on alloced.
Reported-by: Nick Kralevich <nnk@google.com>
Signed-off-by: William Roberts <william.c.roberts@intel.com>
The flex skeleton often triggers compiler warnings; make these
non-fatal for building. We already do likewise for checkpolicy.
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
The combining logic for dontaudit rules was wrong, causing
a dontaudit A B:C *; rule to be clobbered by a dontaudit A B:C p;
rule.
Reported-by: Nick Kralevich <nnk@google.com>
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
The sepol_{bool|iface|user}_key_create() functions were not
copying the name. This produces a use-after-free in the
swig-generated code for python3 bindings. Copy the name
in these functions, and free it upon sepol_{bool|iface|user}_key_free().
Reported-by: Nicolas Iooss <nicolas.iooss@m4x.org>
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Nicholas Iooss discovered that using an unknown permission with a
map class will cause a segfault.
CIL will only give a warning when it fails to resolve an unknown
permission to support the use of policy module packages that use
permissions that don't exit on the current system. When resolving
the unknown map class permission an empty list is used to represent
the unknown permission. When it is evaluated later the list is
assumed to be a permission and a segfault occurs.
There is no reason to allow unknown class map permissions because
the class maps and permissions are defined by the policy.
Exit with an error when failing to resolve a class map permission.
Reported-by: Nicolas Iooss <nicolas.iooss@m4x.org>
Signed-off-by: James Carter <jwcart2@tycho.nsa.gov>
The blank default symver fails to compile with ld.gold. This updates the
symver from blank to LIBSEPOL_1.0. The dynamic linker will first look
for the symbol with the explicit version specified. If there is none, it
will pick the first listed symbol so there is no breakage.
This also matches how symvers are defined in libsemanage.
Signed-off-by: Jason Zaman <jason@perfinion.com>
cil_strpool currently provides an interface to a statically stored
global data structure. This interface does not accomodate multiple
consumers, however, as two calls to cil_strpool_init() will lead to a
memory leak and a call to cil_strpool_destroy() by one consumer will
remove data from use by others, and subsequently lead to a segfault on
the next cil_strpool_destroy() invocation.
Add a reference counter so that the strpool is only initialized once and
protect the exported interface with a mutex.
Tested by calling cil_db_init() on two cil_dbs and then calling
cil_db_destroy() on each.
Signed-off-by: Daniel Cashman <dcashman@android.com>
Nicolas Iooss found while fuzzing secilc with AFL that using an attribute
as a child in a typebounds statement will cause a segfault.
This happens because the child datum is assumed to be part of a cil_type
struct when it is really part of a cil_typeattribute struct. The check to
verify that it is a type and not an attribute comes after it is used.
This bug effects user and role bounds as well because they do not check
whether a datum refers to an attribute or not.
Add checks to verify that neither the child nor the parent datum refer
to an attribute before using them in user, role, and type bounds.
Signed-off-by: James Carter <jwcart2@tycho.nsa.gov>
Nicolas Iooss found while fuzzing secilc with AFL that the statement
"(sensitivityaliasactual SENS SENS)" will cause a segfault.
The segfault occurs because when the aliasactual is resolved the first
identifier is assumed to refer to an alias structure, but it is not.
Add a check to verify that the datum retrieved is actually an alias
and exit with an error if it is not.
Signed-off-by: James Carter <jwcart2@tycho.nsa.gov>
Nicolas Iooss found while fuzzing secilc with AFL that the statement
"(class C (()))" will cause a segfault.
CIL expects a list of permissions in the class declaration and "(())"
is a valid list. Each item of the list is expected to be an identifier
and as the list is processed each item is checked to see if it is a
list. An error is given if it is a list, otherwise the item is assumed
to be an identifier. Unfortunately, the check only works if the list
is not empty. In this case, the item passes the check and is assumed
to be an identifier and a NULL is passed as the string for name
verification. If name verification assumes that a non-NULL value will
be passed in, a segfault will occur.
Add a check for an empty list when processing a permission list and
improve the error handling for permissions when building the AST.
Signed-off-by: James Carter <jwcart2@tycho.nsa.gov>
Nicolas Iooss found while fuzzing secilc with AFL that the statement
"(class C (()))" will cause a segfault.
When CIL checks the syntax of the class statement it sees "(())" as a
valid permission list, but since "()" is not an identifier a NULL is
passed as the string for name verification. A segfault occurs because
name verification assumes that the string being checked is non-NULL.
Check if identifier is NULL when verifying name.
Signed-off-by: James Carter <jwcart2@tycho.nsa.gov>
Nicolas Iooss found while fuzzing secilc with AFL that the statement
"(classpermissionset CPERM (CLASS (and unknow PERM)))" will cause a
segfault.
In order to support a policy module package using a permission that
does not exist on the system it is loaded on, CIL will only give a
warning when it fails to resolve an unknown permission. CIL itself will
just ignore the unknown permission. This means that an expression like
"(and UNKNOWN p1)" will look like "(and p1)" to CIL, but, since syntax
checking has already been done, CIL won't know that the expression is not
well-formed. When the expression is evaluated a segfault will occur
because all expressions are assumed to be well-formed at evaluation time.
Use an empty list to represent an unknown permission so that expressions
will continue to be well-formed and expression evaluation will work but
the unknown permission will still be ignored.
Signed-off-by: James Carter <jwcart2@tycho.nsa.gov>
Nicolas Iooss found while fuzzing secilc with AFL that the following
policy will cause a segfault.
(category c0)
(category c1)
(categoryorder (c0 c1))
(sensitivity s0)
(sensitivitycategory s0 (not (all)))
The expression "(not (all))" is evaluated as containing no categories.
There is a check for the resulting empty list and the category datum
expression is set to NULL. The segfault occurs because the datum
expression is assumed to be non-NULL after evaluation.
Assign the list to the datum expression even if it is empty.
Signed-off-by: James Carter <jwcart2@tycho.nsa.gov>
Nicolas Iooss found while fuzzing secilc with AFL that the following
policy will cause a segfault.
(category c0)
(category c1)
(categoryorder (c0 c1))
(sensitivity s0)
(sensitivitycategory s0 (range c1 c0))
The category range "(range c1 c0)" is invalid because c1 comes after c0
in order.
The invalid range is evaluated as containing no categories. There is a
check for the resulting empty list and the category datum expression is
set to NULL. The segfault occurs because the datum expression is assumed
to be non-NULL after evaluation.
Add a check for an invalid range when evaluating category ranges.
Signed-off-by: James Carter <jwcart2@tycho.nsa.gov>
Correct the build issues on mac, mostly flags for tools.
libsepol and cil now build completley on Mac with a
simple make command.
Signed-off-by: William Roberts <william.c.roberts@intel.com>
