Andrey Ryabinin [Sat, 8 Nov 2014 14:48:05 +0000 (17:48 +0300)]
security: smack: fix out-of-bounds access in smk_parse_smack()
Setting smack label on file (e.g. 'attr -S -s SMACK64 -V "test" test')
triggered following spew on the kernel with KASan applied:
==================================================================
BUG: AddressSanitizer: out of bounds access in strncpy+0x28/0x60 at addr ffff8800059ad064
=============================================================================
BUG kmalloc-8 (Not tainted): kasan error
-----------------------------------------------------------------------------
Disabling lock debugging due to kernel taint
INFO: Slab 0xffffea0000166b40 objects=128 used=7 fp=0xffff8800059ad080 flags=0x4000000000000080
INFO: Object 0xffff8800059ad060 @offset=96 fp=0xffff8800059ad080
Rohit [Wed, 15 Oct 2014 12:10:41 +0000 (17:40 +0530)]
Security: smack: replace kzalloc with kmem_cache for inode_smack
The patch use kmem_cache to allocate/free inode_smack since they are
alloced in high volumes making it a perfect case for kmem_cache.
As per analysis, 24 bytes of memory is wasted per allocation due
to internal fragmentation. With kmem_cache, this can be avoided.
Accounting of memory allocation is below :
total slack net count-alloc/free caller
Before (with kzalloc) 1919872 719952 1919872 29998/0 new_inode_smack+0x14
After (with kmem_cache) 1201680 0 1201680 30042/0 new_inode_smack+0x18
>From above data, we found that 719952 bytes(~700 KB) of memory is
saved on allocation of 29998 smack inodes.
Casey Schaufler [Thu, 9 Oct 2014 23:18:55 +0000 (16:18 -0700)]
Smack: Lock mode for the floor and hat labels
The lock access mode allows setting a read lock on a file
for with the process has only read access. The floor label is
defined to make it easy to have the basic system installed such
that everyone can read it. Once there's a desire to read lock
(rationally or otherwise) a floor file a rule needs to get set.
This happens all the time, so make the floor label a little bit
more special and allow everyone lock access, too. By implication,
give processes with the hat label (hat can read everything)
lock access as well. This reduces clutter in the Smack rule set.
Dmitry Kasatkin [Mon, 6 Oct 2014 16:25:45 +0000 (17:25 +0100)]
KEYS: output last portion of fingerprint in /proc/keys
Previous version of KEYS used to output last 4 bytes of fingerprint.
Now it outputs 8 last bytes of raw subject, which does not make any
visual meaning at all. This patch restores old behavior.
Signed-off-by: Dmitry Kasatkin <d.kasatkin@samsung.com> Signed-off-by: David Howells <dhowells@redhat.com>
Dmitry Kasatkin [Mon, 6 Oct 2014 15:52:12 +0000 (16:52 +0100)]
KEYS: use swapped SKID for performing partial matching
Earlier KEYS code used pure subject key identifiers (fingerprint)
for searching keys. Latest merged code removed that and broke
compatibility with integrity subsytem signatures and original
format of module signatures.
David Howells [Fri, 3 Oct 2014 15:17:02 +0000 (16:17 +0100)]
X.509: If available, use the raw subjKeyId to form the key description
Module signing matches keys by comparing against the key description exactly.
However, the way the key description gets constructed got changed to be
composed of the subject name plus the certificate serial number instead of the
subject name and the subjectKeyId. I changed this to avoid problems with
certificates that don't *have* a subjectKeyId.
Instead, if available, use the raw subjectKeyId to form the key description
and only use the serial number if the subjectKeyId doesn't exist.
Reported-by: Dmitry Kasatkin <d.kasatkin@samsung.com> Signed-off-by: David Howells <dhowells@redhat.com>
Restructure to keyword=value pairs without spaces. Drop superfluous words in
text. Make invalid_context a keyword. Change result= keyword to seresult=.
Signed-off-by: Richard Guy Briggs <rgb@redhat.com>
[Minor rewrite to the patch subject line] Signed-off-by: Paul Moore <pmoore@redhat.com>
David Howells [Sun, 21 Sep 2014 23:02:01 +0000 (00:02 +0100)]
KEYS: Check hex2bin()'s return when generating an asymmetric key ID
As it stands, the code to generate an asymmetric key ID prechecks the hex
string it is given whilst determining the length, before it allocates the
buffer for hex2bin() to translate into - which mean that checking the result of
hex2bin() is redundant.
Unfortunately, hex2bin() is marked as __must_check, which means that the
following warning may be generated if the return value isn't checked:
crypto/asymmetric_keys/asymmetric_type.c: In function
asymmetric_key_hex_to_key_id:
crypto/asymmetric_keys/asymmetric_type.c:110: warning: ignoring return
value of hex2bin, declared with attribute warn_unused_result
The warning can't be avoided by casting the result to void.
Instead, use strlen() to check the length of the string and ignore the fact
that the string might not be entirely valid hex until after the allocation has
been done - in which case we can use the result of hex2bin() for this.
Signed-off-by: David Howells <dhowells@redhat.com>
This patch fixes the detection of the 'open_writers' violation for mmaped
files.
before) an 'open_writers' violation is detected if the policy contains
a rule with the criteria: func=FILE_CHECK mask=MAY_READ
after) an 'open_writers' violation is detected if the current event
matches one of the policy rules.
With the old behaviour, the 'open_writers' violation is not detected
in the following case:
policy:
measure func=FILE_MMAP mask=MAY_EXEC
steps:
1) open a shared library for writing
2) execute a binary that links that shared library
3) during the binary execution, modify the shared library and save
the change
result:
the 'open_writers' violation measurement is not present in the IMA list.
Only binaries executed are protected from writes. For libraries mapped
in memory there is the flag MAP_DENYWRITE for this purpose, but according
to the output of 'man mmap', the mmap flag is ignored.
Since ima_rdwr_violation_check() is now called by process_measurement()
the information about if the inode must be measured is already provided
by ima_get_action(). Thus the unnecessary function ima_must_measure()
has been removed.
Changes in v3 (Dmitry Kasatkin):
- Violation for MMAP_CHECK function are verified since this patch
- Changed patch description a bit
ima: fix race condition on ima_rdwr_violation_check and process_measurement
This patch fixes a race condition between two functions that try to access
the same inode. Since the i_mutex lock is held and released separately
in the two functions, there may be the possibility that a violation is
not correctly detected.
Suppose there are two processes, A (reader) and B (writer), if the
following sequence happens:
the ToMToU violation (a reader may be accessing a content different from
that measured, due to a concurrent modification by a writer) will not be
detected. To avoid this issue, the violation check and the measurement
must be done atomically.
This patch fixes the problem by moving the violation check inside
process_measurement() when the i_mutex lock is held. Differently from
the old code, the violation check is executed also for the MMAP_CHECK
hook (other than for FILE_CHECK). This allows to detect ToMToU violations
that are possible because shared libraries can be opened for writing
while they are in use (according to the output of 'man mmap', the mmap()
flag MAP_DENYWRITE is ignored).
Changes in v5 (Roberto Sassu):
* get iint if action is not zero
* exit process_measurement() after the violation check if action is zero
* reverse order process_measurement() exit cleanup (Mimi)
Changes in v4 (Dmitry Kasatkin):
* iint allocation is done before calling ima_rdrw_violation_check()
(Suggested-by Mimi)
* do not check for violations if the policy does not contain 'measure'
rules (done by Roberto Sassu)
Changes in v3 (Dmitry Kasatkin):
* no violation checking for MMAP_CHECK function in this patch
* remove use of filename from violation
* removes checking if ima is enabled from ima_rdrw_violation_check
* slight style change
This patch introduces the new variable 'ima_policy_flag', whose bits
are set depending on the action of the current policy rules. Only the
flags IMA_MEASURE, IMA_APPRAISE and IMA_AUDIT are set.
The new variable will be used to improve performance by skipping the
unnecessary execution of IMA code if the policy does not contain rules
with the above actions.
Changes in v6 (Roberto Sassu)
* do not check 'ima_initialized' before calling ima_update_policy_flag()
in ima_update_policy() (suggested by Dmitry)
* calling ima_update_policy_flag() moved to init_ima to co-locate with
ima_initialized (Dmitry)
* add/revise comments (Mimi)
Changes in v5 (Roberto Sassu)
* reset IMA_APPRAISE flag in 'ima_policy_flag' if 'ima_appraise' is set
to zero (reported by Dmitry)
* update 'ima_policy_flag' only if IMA initialization is successful
(suggested by Mimi and Dmitry)
* check 'ima_policy_flag' instead of 'ima_initialized'
(suggested by Mimi and Dmitry)
ima: return an error code from ima_add_boot_aggregate()
This patch modifies ima_add_boot_aggregate() to return an error code.
This way we can determine if all the initialization procedures have
been executed successfully.
Dmitry Kasatkin [Thu, 8 May 2014 10:11:29 +0000 (13:11 +0300)]
ima: provide 'ima_appraise=log' kernel option
The kernel boot parameter "ima_appraise" currently defines 'off',
'enforce' and 'fix' modes. When designing a policy and labeling
the system, access to files are either blocked in the default
'enforce' mode or automatically fixed in the 'fix' mode. It is
beneficial to be able to run the system in a logging only mode,
without fixing it, in order to properly analyze the system. This
patch adds a 'log' mode to run the system in a permissive mode and
log the appraisal results.
Dmitry Kasatkin [Fri, 27 Jun 2014 10:01:32 +0000 (13:01 +0300)]
ima: move keyring initialization to ima_init()
ima_init() is used as a single place for all initializations.
Experimental keyring patches used the 'late_initcall' which was
co-located with the late_initcall(init_ima). When the late_initcall
for the keyring initialization was abandoned, initialization moved
to init_ima, though it would be more logical to move it to ima_init,
where the rest of the initialization is done. This patch moves the
keyring initialization to ima_init() as a preparatory step for
loading the keys which will be added to ima_init() in following
patches.
Changes for next to improve the matching of asymmetric keys and to improve the
handling of PKCS#7 certificates:
(1) Provide a method to preparse the data supplied for matching a key. This
permits they key type to extract out the bits it needs for matching once
only.
Further, the type of search (direct lookup or iterative) can be set and
the function used to actually check the match can be set by preparse
rather than being hard coded for the type.
(2) Improves asymmetric keys identification.
Keys derived from X.509 certs now get labelled with IDs derived from their
issuer and certificate number (required to match PKCS#7) and from their
SKID and subject (required to match X.509).
IDs are now binary and match criterion preparsing is provided so that
criteria can be turned into binary blobs to make matching faster.
(3) Improves PKCS#7 message handling to permit PKCS#7 messages without X.509
cert lists to be matched to trusted keys, thereby allowing minimally sized
PKCS#7 certs to be used.
(4) Improves PKCS#7 message handling to better handle certificate chains that
are broken due to unsupported crypto that can otherwise by used to
intersect a trust keyring.
These must go on top of the PKCS#7 parser cleanup fixes.
Signed-off-by: David Howells <dhowells@redhat.com>
David Howells [Tue, 16 Sep 2014 16:36:17 +0000 (17:36 +0100)]
PKCS#7: Handle PKCS#7 messages that contain no X.509 certs
The X.509 certificate list in a PKCS#7 message is optional. To save space, we
can omit the inclusion of any X.509 certificates if we are sure that we can
look the relevant public key up by the serial number and issuer given in a
signed info block.
This also supports use of a signed info block for which we can't find a
matching X.509 cert in the certificate list, though it be populated.
Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Vivek Goyal <vgoyal@redhat.com>
David Howells [Tue, 16 Sep 2014 16:36:15 +0000 (17:36 +0100)]
PKCS#7: Better handling of unsupported crypto
Provide better handling of unsupported crypto when verifying a PKCS#7 message.
If we can't bridge the gap between a pair of X.509 certs or between a signed
info block and an X.509 cert because it involves some crypto we don't support,
that's not necessarily the end of the world as there may be other ways points
at which we can intersect with a ring of trusted keys.
Instead, only produce ENOPKG immediately if all the signed info blocks in a
PKCS#7 message require unsupported crypto to bridge to the first X.509 cert.
Otherwise, we defer the generation of ENOPKG until we get ENOKEY during trust
validation.
Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Vivek Goyal <vgoyal@redhat.com>
David Howells [Tue, 16 Sep 2014 16:36:13 +0000 (17:36 +0100)]
KEYS: Overhaul key identification when searching for asymmetric keys
Make use of the new match string preparsing to overhaul key identification
when searching for asymmetric keys. The following changes are made:
(1) Use the previously created asymmetric_key_id struct to hold the following
key IDs derived from the X.509 certificate or PKCS#7 message:
id: serial number + issuer
skid: subjKeyId + subject
authority: authKeyId + issuer
(2) Replace the hex fingerprint attached to key->type_data[1] with an
asymmetric_key_ids struct containing the id and the skid (if present).
(3) Make the asymmetric_type match data preparse select one of two searches:
(a) An iterative search for the key ID given if prefixed with "id:". The
prefix is expected to be followed by a hex string giving the ID to
search for. The criterion key ID is checked against all key IDs
recorded on the key.
(b) A direct search if the key ID is not prefixed with "id:". This will
look for an exact match on the key description.
(4) Make x509_request_asymmetric_key() take a key ID. This is then converted
into "id:<hex>" and passed into keyring_search() where match preparsing
will turn it back into a binary ID.
(5) X.509 certificate verification then takes the authority key ID and looks
up a key that matches it to find the public key for the certificate
signature.
(6) PKCS#7 certificate verification then takes the id key ID and looks up a
key that matches it to find the public key for the signed information
block signature.
Additional changes:
(1) Multiple subjKeyId and authKeyId values on an X.509 certificate cause the
cert to be rejected with -EBADMSG.
(2) The 'fingerprint' ID is gone. This was primarily intended to convey PGP
public key fingerprints. If PGP is supported in future, this should
generate a key ID that carries the fingerprint.
(3) Th ca_keyid= kernel command line option is now converted to a key ID and
used to match the authority key ID. Possibly this should only match the
actual authKeyId part and not the issuer as well.
Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Vivek Goyal <vgoyal@redhat.com>
David Howells [Tue, 16 Sep 2014 16:36:11 +0000 (17:36 +0100)]
KEYS: Implement binary asymmetric key ID handling
Implement the first step in using binary key IDs for asymmetric keys rather
than hex string keys.
The previously added match data preparsing will be able to convert hex
criterion strings into binary which can then be compared more rapidly.
Further, we actually want more then one ID string per public key. The problem
is that X.509 certs refer to other X.509 certs by matching Issuer + AuthKeyId
to Subject + SubjKeyId, but PKCS#7 messages match against X.509 Issuer +
SerialNumber.
This patch just provides facilities for a later patch to make use of.
Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Vivek Goyal <vgoyal@redhat.com>
David Howells [Tue, 16 Sep 2014 16:36:06 +0000 (17:36 +0100)]
KEYS: Remove key_type::match in favour of overriding default by match_preparse
A previous patch added a ->match_preparse() method to the key type. This is
allowed to override the function called by the iteration algorithm.
Therefore, we can just set a default that simply checks for an exact match of
the key description with the original criterion data and allow match_preparse
to override it as needed.
The key_type::match op is then redundant and can be removed, as can the
user_match() function.
Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Vivek Goyal <vgoyal@redhat.com>
David Howells [Tue, 16 Sep 2014 16:36:04 +0000 (17:36 +0100)]
KEYS: Remove key_type::def_lookup_type
Remove key_type::def_lookup_type as it's no longer used. The information now
defaults to KEYRING_SEARCH_LOOKUP_DIRECT but may be overridden by
type->match_preparse().
Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Vivek Goyal <vgoyal@redhat.com>
David Howells [Tue, 16 Sep 2014 16:29:03 +0000 (17:29 +0100)]
PKCS#7: Fix the parser cleanup to drain parsed out X.509 certs
Fix the parser cleanup code to drain parsed out X.509 certs in the case that
the decode fails and we jump to error_decode.
The function is rearranged so that the same cleanup code is used in the success
case as the error case - just that the message descriptor under construction is
only released if it is still pointed to by the context struct at that point.
Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Vivek Goyal <vgoyal@redhat.com>
David Howells [Tue, 16 Sep 2014 16:29:03 +0000 (17:29 +0100)]
PKCS#7: Provide a single place to do signed info block freeing
The code to free a signed info block is repeated several times, so move the
code to do it into a function of its own. This gives us a place to add clean
ups for stuff that gets added to pkcs7_signed_info.
Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Vivek Goyal <vgoyal@redhat.com>
Paul Moore [Wed, 10 Sep 2014 21:09:57 +0000 (17:09 -0400)]
selinux: make the netif cache namespace aware
While SELinux largely ignores namespaces, for good reason, there are
some places where it needs to at least be aware of namespaces in order
to function correctly. Network namespaces are one example. Basic
awareness of network namespaces are necessary in order to match a
network interface's index number to an actual network device.
This patch corrects a problem with network interfaces added to a
non-init namespace, and can be reproduced with the following commands:
[NOTE: the NetLabel configuration is here only to active the dynamic
networking controls ]
# ip link add type veth
# ip netns add myns
# ip link set veth1 netns myns
# ip a add dev veth0 10.250.13.100/24
# ip netns exec myns ip a add dev veth1 10.250.13.101/24
# ip l set veth0 up
# ip netns exec myns ip l set veth1 up
integrity: make integrity files as 'integrity' module
The kernel print macros use the KBUILD_MODNAME, which is initialized
to the module name. The current integrity/Makefile makes every file
as its own module, so pr_xxx messages are prefixed with the file name
instead of the module. Similar to the evm/Makefile and ima/Makefile,
this patch fixes the integrity/Makefile to use the single name
'integrity'.
integrity: base integrity subsystem kconfig options on integrity
The integrity subsystem has lots of options and takes more than
half of the security menu. This patch consolidates the options
under "integrity", which are hidden if not enabled. This change
does not affect existing configurations. Re-configuration is not
needed.
Changes v4:
- no need to change "integrity subsystem" to menuconfig as
options are hidden, when not enabled. (Mimi)
- add INTEGRITY Kconfig help description
Changes v3:
- dependency to INTEGRITY removed when behind 'if INTEGRITY'
Changes v2:
- previous patch moved integrity out of the 'security' menu.
This version keeps integrity as a security option (Mimi).
Dmitry Kasatkin [Tue, 19 Aug 2014 13:48:39 +0000 (16:48 +0300)]
ima: remove usage of filename parameter
In all cases except ima_bprm_check() the filename was not defined
and ima_d_path() was used to find the full path. Unfortunately,
the bprm filename is a relative pathname (eg. ./<dir>/filename).
ima_bprm_check() selects between bprm->interp and bprm->filename.
The following dump demonstrates the differences between using
filename and interp.
Dmitry Kasatkin [Wed, 20 Aug 2014 09:37:57 +0000 (12:37 +0300)]
ima: remove unnecessary appraisal test
ima_get_action() sets the "action" flags based on policy.
Before collecting, measuring, appraising, or auditing the
file, the "action" flag is updated based on the cached
iint->flags.
This patch removes the subsequent unnecessary appraisal
test in ima_appraise_measurement().
Dmitry Kasatkin [Fri, 22 Aug 2014 06:43:55 +0000 (09:43 +0300)]
ima: simplify conditional statement to improve performance
Precede bit testing before string comparison makes code
faster. Also refactor statement as a single line pointer
assignment. Logic is following: we set 'xattr_ptr' to read
xattr value when we will do appraisal or in any case when
measurement template is other than 'ima'.
Dmitry Kasatkin [Tue, 19 Aug 2014 11:56:18 +0000 (14:56 +0300)]
integrity: remove declaration of non-existing functions
Commit f381c27 "integrity: move ima inode integrity data management"
(re)moved few functions but left their declarations in header files.
This patch removes them and also removes duplicated declaration of
integrity_iint_find().
Commit c7de7ad "ima: remove unused cleanup functions". This patch
removes these definitions as well.
Dmitry Kasatkin [Fri, 27 Jun 2014 15:15:44 +0000 (18:15 +0300)]
ima: pass 'opened' flag to identify newly created files
Empty files and missing xattrs do not guarantee that a file was
just created. This patch passes FILE_CREATED flag to IMA to
reliably identify new files.
evm: properly handle INTEGRITY_NOXATTRS EVM status
Unless an LSM labels a file during d_instantiate(), newly created
files are not labeled with an initial security.evm xattr, until
the file closes. EVM, before allowing a protected, security xattr
to be written, verifies the existing 'security.evm' value is good.
For newly created files without a security.evm label, this
verification prevents writing any protected, security xattrs,
until the file closes.
Following is the example when this happens:
fd = open("foo", O_CREAT | O_WRONLY, 0644);
setxattr("foo", "security.SMACK64", value, sizeof(value), 0);
close(fd);
While INTEGRITY_NOXATTRS status is handled in other places, such
as evm_inode_setattr(), it does not handle it in all cases in
evm_protect_xattr(). By limiting the use of INTEGRITY_NOXATTRS to
newly created files, we can now allow setting "protected" xattrs.
Changelog:
- limit the use of INTEGRITY_NOXATTRS to IMA identified new files
Dmitry Kasatkin [Fri, 27 Jun 2014 15:04:27 +0000 (18:04 +0300)]
ima: provide flag to identify new empty files
On ima_file_free(), newly created empty files are not labeled with
an initial security.ima value, because the iversion did not change.
Commit dff6efc "fs: fix iversion handling" introduced a change in
iversion behavior. To verify this change use the shell command:
This patch defines the IMA_NEW_FILE flag. The flag is initially
set, when IMA detects that a new file is created, and subsequently
checked on the ima_file_free() hook to set the initial security.ima
value.
Paul Moore [Wed, 3 Sep 2014 14:51:59 +0000 (10:51 -0400)]
selinux: fix a problem with IPv6 traffic denials in selinux_ip_postroute()
A previous commit c0828e50485932b7e019df377a6b0a8d1ebd3080 ("selinux:
process labeled IPsec TCP SYN-ACK packets properly in
selinux_ip_postroute()") mistakenly left out a 'break' from a switch
statement which caused problems with IPv6 traffic.
Thanks to Florian Westphal for reporting and debugging the issue.
Reported-by: Florian Westphal <fwestpha@redhat.com> Signed-off-by: Paul Moore <pmoore@redhat.com>
Dmitry Kasatkin [Mon, 23 Jun 2014 17:32:56 +0000 (20:32 +0300)]
ima: fix fallback to use new_sync_read()
3.16 commit aad4f8bb42af06371aa0e85bf0cd9d52c0494985
'switch simple generic_file_aio_read() users to ->read_iter()'
replaced ->aio_read with ->read_iter in most of the file systems
and introduced new_sync_read() as a replacement for do_sync_read().
Most of file systems set '->read' and ima_kernel_read is not affected.
When ->read is not set, this patch adopts fallback call changes from the
vfs_read.
Dmitry Kasatkin [Fri, 15 Aug 2014 10:28:52 +0000 (13:28 +0300)]
ima: prevent buffer overflow in ima_alloc_tfm()
This patch fixes the case where the file's signature/hash xattr contains
an invalid hash algorithm. Although we can not verify the xattr, we still
need to measure the file. Use the default IMA hash algorithm.
The patch 3bcced39ea7d: "ima: use ahash API for file hash
calculation" from Feb 26, 2014, leads to the following static checker
warning:
security/integrity/ima/ima_crypto.c:204 ima_alloc_atfm()
        error: buffer overflow 'hash_algo_name' 17 <= 17
Unlike shash tfm memory, which is allocated on initialization, the
ahash tfm memory allocation is deferred until needed.
This patch fixes the case where ima_ahash_tfm has not yet been
allocated and the file's signature/hash xattr contains an invalid hash
algorithm. Although we can not verify the xattr, we still need to
measure the file. Use the default IMA hash algorithm.
Changelog:
- set valid algo before testing tfm - based on Dmitry's comment
Mark Rustad [Thu, 28 Aug 2014 11:43:09 +0000 (04:43 -0700)]
security: Silence shadow warning
Renaming an unused formal parameter in the static inline function
security_inode_init_security eliminates many W=2 warnings.
Signed-off-by: Mark Rustad <mark.d.rustad@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: James Morris <james.l.morris@oracle.com>
Lukasz Pawelczyk [Fri, 29 Aug 2014 15:02:55 +0000 (17:02 +0200)]
Make Smack operate on smack_known struct where it still used char*
Smack used to use a mix of smack_known struct and char* throughout its
APIs and implementation. This patch unifies the behaviour and makes it
store and operate exclusively on smack_known struct pointers when managing
labels.
Lukasz Pawelczyk [Fri, 29 Aug 2014 15:02:54 +0000 (17:02 +0200)]
Fix a bidirectional UDS connect check typo
The 54e70ec5eb090193b03e69d551fa6771a5a217c4 commit introduced a
bidirectional check that should have checked for mutual WRITE access
between two labels. Due to a typo the second check was incorrect.
Casey Schaufler [Wed, 27 Aug 2014 21:51:27 +0000 (14:51 -0700)]
Smack: Bring-up access mode
People keep asking me for permissive mode, and I keep saying "no".
Permissive mode is wrong for more reasons than I can enumerate,
but the compelling one is that it's once on, never off.
Nonetheless, there is an argument to be made for running a
process with lots of permissions, logging which are required,
and then locking the process down. There wasn't a way to do
that with Smack, but this provides it.
The notion is that you start out by giving the process an
appropriate Smack label, such as "ATBirds". You create rules
with a wide range of access and the "b" mode. On Tizen it
might be:
ATBirds System rwxalb
ATBirds User rwxalb
ATBirds _ rwxalb
User ATBirds wb
System ATBirds wb
Accesses that fail will generate audit records. Accesses
that succeed because of rules marked with a "b" generate
log messages identifying the rule, the program and as much
object information as is convenient.
When the system is properly configured and the programs
brought in line with the labeling scheme the "b" mode can
be removed from the rules. When the system is ready for
production the facility can be configured out.
This provides the developer the convenience of permissive
mode without creating a system that looks like it is
enforcing a policy while it is not.
Stephen Smalley [Mon, 4 Aug 2014 17:36:49 +0000 (13:36 -0400)]
selinux: Permit bounded transitions under NO_NEW_PRIVS or NOSUID.
If the callee SID is bounded by the caller SID, then allowing
the transition to occur poses no risk of privilege escalation and we can
therefore safely allow the transition to occur. Add this exemption
for both the case where a transition was explicitly requested by the
application and the case where an automatic transition is defined in
policy.
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov> Reviewed-by: Andy Lutomirski <luto@amacapital.net> Signed-off-by: Paul Moore <pmoore@redhat.com>
Smack: remove unneeded NULL-termination from securtity label
Values of extended attributes are stored as binary blobs. NULL-termination
of them isn't required. It just wastes disk space and confuses command-line
tools like getfattr because they have to print that zero byte at the end.
This patch removes terminating zero byte from initial security label in
smack_inode_init_security and cuts it out in function smack_inode_getsecurity
which is used by syscall getxattr. This change seems completely safe, because
function smk_parse_smack ignores everything after first zero byte.
Signed-off-by: Konstantin Khlebnikov <k.khlebnikov@samsung.com>
Smack: handle zero-length security labels without panic
Zero-length security labels are invalid but kernel should handle them.
This patch fixes kernel panic after setting zero-length security labels:
# attr -S -s "SMACK64" -V "" file
And after writing zero-length string into smackfs files syslog and onlycp:
# python -c 'import os; os.write(1, "")' > /smack/syslog
The problem is caused by brain-damaged logic in function smk_parse_smack()
which takes pointer to buffer and its length but if length below or equal zero
it thinks that the buffer is zero-terminated. Unfortunately callers of this
function are widely used and proper fix requires serious refactoring.
Signed-off-by: Konstantin Khlebnikov <k.khlebnikov@samsung.com>
Security operation ->inode_listsecurity is used for generating list of
available extended attributes for syscall listxattr. Currently it's used
only in nfs4 or if filesystem doesn't provide i_op->listxattr.
The list is the set of NULL-terminated names, one after the other.
This method must include zero byte at the and into result.
Also this function must return length even if string does not fit into
output buffer or it is NULL, see similar method in selinux and man listxattr.
Signed-off-by: Konstantin Khlebnikov <k.khlebnikov@samsung.com>
Linus Torvalds [Sun, 3 Aug 2014 16:58:20 +0000 (09:58 -0700)]
Merge branch 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull timer fixes from Thomas Gleixner:
"Two fixes in the timer area:
- a long-standing lock inversion due to a printk
- suspend-related hrtimer corruption in sched_clock"
* 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
timer: Fix lock inversion between hrtimer_bases.lock and scheduler locks
sched_clock: Avoid corrupting hrtimer tree during suspend
Linus Torvalds [Sat, 2 Aug 2014 17:57:39 +0000 (10:57 -0700)]
Merge branch 'fixes' of git://ftp.arm.linux.org.uk/~rmk/linux-arm
Pull ARM fixes from Russell King:
"A few fixes for ARM. Some of these are correctness issues:
- TLBs must be flushed after the old mappings are removed by the DMA
mapping code, but before the new mappings are established.
- An off-by-one entry error in the Keystone LPAE setup code.
Fixes include:
- ensuring that the identity mapping for LPAE does not remove the
kernel image from the identity map.
- preventing userspace from trapping into kgdb.
- fixing a preemption issue in the Intel iwmmxt code.
- fixing a build error with nommu.
Other changes include:
- Adding a note about which areas of memory are expected to be
accessible while the identity mapping tables are in place"
* 'fixes' of git://ftp.arm.linux.org.uk/~rmk/linux-arm:
ARM: 8124/1: don't enter kgdb when userspace executes a kgdb break instruction
ARM: idmap: add identity mapping usage note
ARM: 8115/1: LPAE: reduce damage caused by idmap to virtual memory layout
ARM: fix alignment of keystone page table fixup
ARM: 8112/1: only select ARM_PATCH_PHYS_VIRT if MMU is enabled
ARM: 8100/1: Fix preemption disable in iwmmxt_task_enable()
ARM: DMA: ensure that old section mappings are flushed from the TLB
Omar Sandoval [Fri, 1 Aug 2014 17:14:06 +0000 (18:14 +0100)]
ARM: 8124/1: don't enter kgdb when userspace executes a kgdb break instruction
The kgdb breakpoint hooks (kgdb_brk_fn and kgdb_compiled_brk_fn)
should only be entered when a kgdb break instruction is executed
from the kernel. Otherwise, if kgdb is enabled, a userspace program
can cause the kernel to drop into the debugger by executing either
KGDB_BREAKINST or KGDB_COMPILED_BREAK.
Acked-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Omar Sandoval <osandov@osandov.com> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Russell King [Tue, 29 Jul 2014 11:18:34 +0000 (12:18 +0100)]
ARM: idmap: add identity mapping usage note
Add a note about the usage of the identity mapping; we do not support
accesses outside of the identity map region and kernel image while a
CPU is using the identity map. This is because the identity mapping
may overwrite vmalloc space, IO mappings, the vectors pages, etc.
Acked-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Linus Torvalds [Sat, 2 Aug 2014 00:37:01 +0000 (17:37 -0700)]
Merge branch 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull x86 fix from Peter Anvin:
"A single fix to not invoke the espfix code on Xen PV, as it turns out
to oops the guest when invoked after all. This patch leaves some
amount of dead code, in particular unnecessary initialization of the
espfix stacks when they won't be used, but in the interest of keeping
the patch minimal that cleanup can wait for the next cycle"
* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86_64/entry/xen: Do not invoke espfix64 on Xen
Linus Torvalds [Sat, 2 Aug 2014 00:16:05 +0000 (17:16 -0700)]
Merge tag 'staging-3.16-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging
Pull staging driver bugfixes from Greg KH:
"Here are some tiny staging driver bugfixes that I've had in my tree
for the past week that resolve some reported issues. Nothing major at
all, but it would be good to get them merged for 3.16-rc8 or -final"
* tag 'staging-3.16-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging:
staging: vt6655: Fix disassociated messages every 10 seconds
staging: vt6655: Fix Warning on boot handle_irq_event_percpu.
staging: rtl8723au: rtw_resume(): release semaphore before exit on error
iio:bma180: Missing check for frequency fractional part
iio:bma180: Fix scale factors to report correct acceleration units
iio: buffer: Fix demux table creation
Linus Torvalds [Fri, 1 Aug 2014 19:49:02 +0000 (12:49 -0700)]
Merge tag 'fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc
Pull ARM straggler SoC fix from Olof Johansson:
"A DT bugfix for Nomadik that had an ambigouos double-inversion of a
gpio line, and one MAINTAINER URL update that might as well go in now.
We could hold off until the merge window, but then we'll just have to
mark the DT fix for stable and it just seems like in total causing
more work"
* tag 'fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc:
MAINTAINERS: Update Tegra Git URL
ARM: nomadik: fix up double inversion in DT
Anssi Hannula [Fri, 1 Aug 2014 15:55:47 +0000 (11:55 -0400)]
dm cache: fix race affecting dirty block count
nr_dirty is updated without locking, causing it to drift so that it is
non-zero (either a small positive integer, or a very large one when an
underflow occurs) even when there are no actual dirty blocks. This was
due to a race between the workqueue and map function accessing nr_dirty
in parallel without proper protection.
People were seeing under runs due to a race on increment/decrement of
nr_dirty, see: https://lkml.org/lkml/2014/6/3/648
Fix this by using an atomic_t for nr_dirty.
Reported-by: roma1390@gmail.com Signed-off-by: Anssi Hannula <anssi.hannula@iki.fi> Signed-off-by: Joe Thornber <ejt@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
1d3d4437eae1 ("vmscan: per-node deferred work") added a flags field to
struct shrinker assuming that all shrinkers were zero filled. The dm
bufio shrinker is not zero filled, which leaves arbitrary kmalloc() data
in flags. So far the only defined flags bit is SHRINKER_NUMA_AWARE.
But there are proposed patches which add other bits to shrinker.flags
(e.g. memcg awareness).
Rather than simply initializing the shrinker, this patch uses kzalloc()
when allocating the dm_bufio_client to ensure that the embedded shrinker
and any other similar structures are zeroed.
This fixes theoretical over aggressive shrinking of dm bufio objects.
If the uninitialized dm_bufio_client.shrinker.flags contains
SHRINKER_NUMA_AWARE then shrink_slab() would call the dm shrinker for
each numa node rather than just once. This has been broken since 3.12.
Paul Moore [Fri, 1 Aug 2014 15:17:37 +0000 (11:17 -0400)]
netlabel: shorter names for the NetLabel catmap funcs/structs
Historically the NetLabel LSM secattr catmap functions and data
structures have had very long names which makes a mess of the NetLabel
code and anyone who uses NetLabel. This patch renames the catmap
functions and structures from "*_secattr_catmap_*" to just "*_catmap_*"
which improves things greatly.
There are no substantial code or logic changes in this patch.
Signed-off-by: Paul Moore <pmoore@redhat.com> Tested-by: Casey Schaufler <casey@schaufler-ca.com>
Paul Moore [Fri, 1 Aug 2014 15:17:29 +0000 (11:17 -0400)]
netlabel: fix the catmap walking functions
The two NetLabel LSM secattr catmap walk functions didn't handle
certain edge conditions correctly, causing incorrect security labels
to be generated in some cases. This patch corrects these problems and
converts the functions to use the new _netlbl_secattr_catmap_getnode()
function in order to reduce the amount of repeated code.
Cc: stable@vger.kernel.org Signed-off-by: Paul Moore <pmoore@redhat.com> Tested-by: Casey Schaufler <casey@schaufler-ca.com>
Paul Moore [Fri, 1 Aug 2014 15:17:17 +0000 (11:17 -0400)]
netlabel: fix the horribly broken catmap functions
The NetLabel secattr catmap functions, and the SELinux import/export
glue routines, were broken in many horrible ways and the SELinux glue
code fiddled with the NetLabel catmap structures in ways that we
probably shouldn't allow. At some point this "worked", but that was
likely due to a bit of dumb luck and sub-par testing (both inflicted
by yours truly). This patch corrects these problems by basically
gutting the code in favor of something less obtuse and restoring the
NetLabel abstractions in the SELinux catmap glue code.
Everything is working now, and if it decides to break itself in the
future this code will be much easier to debug than the code it
replaces.
One noteworthy side effect of the changes is that it is no longer
necessary to allocate a NetLabel catmap before calling one of the
NetLabel APIs to set a bit in the catmap. NetLabel will automatically
allocate the catmap nodes when needed, resulting in less allocations
when the lowest bit is greater than 255 and less code in the LSMs.
Cc: stable@vger.kernel.org Reported-by: Christian Evans <frodox@zoho.com> Signed-off-by: Paul Moore <pmoore@redhat.com> Tested-by: Casey Schaufler <casey@schaufler-ca.com>
Paul Moore [Fri, 1 Aug 2014 15:17:03 +0000 (11:17 -0400)]
netlabel: fix a problem when setting bits below the previously lowest bit
The NetLabel category (catmap) functions have a problem in that they
assume categories will be set in an increasing manner, e.g. the next
category set will always be larger than the last. Unfortunately, this
is not a valid assumption and could result in problems when attempting
to set categories less than the startbit in the lowest catmap node.
In some cases kernel panics and other nasties can result.
This patch corrects the problem by checking for this and allocating a
new catmap node instance and placing it at the front of the list.
Cc: stable@vger.kernel.org Reported-by: Christian Evans <frodox@zoho.com> Signed-off-by: Paul Moore <pmoore@redhat.com> Tested-by: Casey Schaufler <casey@schaufler-ca.com>
Jan Kara [Fri, 1 Aug 2014 10:20:02 +0000 (12:20 +0200)]
timer: Fix lock inversion between hrtimer_bases.lock and scheduler locks
clockevents_increase_min_delta() calls printk() from under
hrtimer_bases.lock. That causes lock inversion on scheduler locks because
printk() can call into the scheduler. Lockdep puts it as:
======================================================
[ INFO: possible circular locking dependency detected ] 3.15.0-rc8-06195-g939f04b #2 Not tainted
-------------------------------------------------------
trinity-main/74 is trying to acquire lock:
(&port_lock_key){-.....}, at: [<811c60be>] serial8250_console_write+0x8c/0x10c
but task is already holding lock:
(hrtimer_bases.lock){-.-...}, at: [<8103caeb>] hrtimer_try_to_cancel+0x13/0x66
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
Fix the problem by using printk_deferred() which does not call into the
scheduler.
Reported-by: Fengguang Wu <fengguang.wu@intel.com> Signed-off-by: Jan Kara <jack@suse.cz> Cc: stable@vger.kernel.org Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Eric Biggers [Wed, 25 Jun 2014 04:45:08 +0000 (23:45 -0500)]
vfs: fix check for fallocate on active swapfile
Fix the broken check for calling sys_fallocate() on an active swapfile,
introduced by commit 0790b31b69374ddadefe ("fs: disallow all fallocate
operation on active swapfile").
Signed-off-by: Eric Biggers <ebiggers3@gmail.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
The direct-io.c rewrite to use the iov_iter infrastructure stopped updating
the size field in struct dio_submit, and thus rendered the check for
allowing asynchronous completions to always return false. Fix this by
comparing it to the count of bytes in the iov_iter instead.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reported-by: Tim Chen <tim.c.chen@linux.intel.com> Tested-by: Tim Chen <tim.c.chen@linux.intel.com>