Merge tag 'keys-next-20160505' of git://git.kernel.org/pub/scm/linux/kernel/git/dhowells/linux-fs into next

Access reporting often happens from atomic contexes. Avoid
lockups when allocating memory for command lines.

Fixes: 8a56038c ("Yama: consolidate error reporting")
Signed-off-by: Sasha Levin <sasha.levin@oracle.com>

Drop accidentally repeated word in comment.
Signed-off-by: Mickaël Salaün <mic@digikod.net>
Cc: Kees Cook <keescook@chromium.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Will Drewry <wad@chromium.org>

Here's a set of patches that changes how certificates/keys are determined
to be trusted.  That's currently a two-step process:

 (1) Up until recently, when an X.509 certificate was parsed - no matter
     the source - it was judged against the keys in .system_keyring,
     assuming those keys to be trusted if they have KEY_FLAG_TRUSTED set
     upon them.

     This has just been changed such that any key in the .ima_mok keyring,
     if configured, may also be used to judge the trustworthiness of a new
     certificate, whether or not the .ima_mok keyring is meant to be
     consulted for whatever process is being undertaken.

     If a certificate is determined to be trustworthy, KEY_FLAG_TRUSTED
     will be set upon a key it is loaded into (if it is loaded into one),
     no matter what the key is going to be loaded for.

 (2) If an X.509 certificate is loaded into a key, then that key - if
     KEY_FLAG_TRUSTED gets set upon it - can be linked into any keyring
     with KEY_FLAG_TRUSTED_ONLY set upon it.  This was meant to be the
     system keyring only, but has been extended to various IMA keyrings.
     A user can at will link any key marked KEY_FLAG_TRUSTED into any
     keyring marked KEY_FLAG_TRUSTED_ONLY if the relevant permissions masks
     permit it.

These patches change that:

 (1) Trust becomes a matter of consulting the ring of trusted keys supplied
     when the trust is evaluated only.

 (2) Every keyring can be supplied with its own manager function to
     restrict what may be added to that keyring.  This is called whenever a
     key is to be linked into the keyring to guard against a key being
     created in one keyring and then linked across.

     This function is supplied with the keyring and the key type and
     payload[*] of the key being linked in for use in its evaluation.  It
     is permitted to use other data also, such as the contents of other
     keyrings such as the system keyrings.

     [*] The type and payload are supplied instead of a key because as an
         optimisation this function may be called whilst creating a key and
         so may reject the proposed key between preparse and allocation.

 (3) A default manager function is provided that permits keys to be
     restricted to only asymmetric keys that are vouched for by the
     contents of the system keyring.

     A second manager function is provided that just rejects with EPERM.

 (4) A key allocation flag, KEY_ALLOC_BYPASS_RESTRICTION, is made available
     so that the kernel can initialise keyrings with keys that form the
     root of the trust relationship.

 (5) KEY_FLAG_TRUSTED and KEY_FLAG_TRUSTED_ONLY are removed, along with
     key_preparsed_payload::trusted.

This change also makes it possible in future for userspace to create a private
set of trusted keys and then to have it sealed by setting a manager function
where the private set is wholly independent of the kernel's trust
relationships.

Further changes in the set involve extracting certain IMA special keyrings
and making them generally global:

 (*) .system_keyring is renamed to .builtin_trusted_keys and remains read
     only.  It carries only keys built in to the kernel.  It may be where
     UEFI keys should be loaded - though that could better be the new
     secondary keyring (see below) or a separate UEFI keyring.

 (*) An optional secondary system keyring (called .secondary_trusted_keys)
     is added to replace the IMA MOK keyring.

     (*) Keys can be added to the secondary keyring by root if the keys can
         be vouched for by either ring of system keys.

 (*) Module signing and kexec only use .builtin_trusted_keys and do not use
     the new secondary keyring.

 (*) Config option SYSTEM_TRUSTED_KEYS now depends on ASYMMETRIC_KEY_TYPE as
     that's the only type currently permitted on the system keyrings.

 (*) A new config option, IMA_KEYRINGS_PERMIT_SIGNED_BY_BUILTIN_OR_SECONDARY,
     is provided to allow keys to be added to IMA keyrings, subject to the
     restriction that such keys are validly signed by a key already in the
     system keyrings.

     If this option is enabled, but secondary keyrings aren't, additions to
     the IMA keyrings will be restricted to signatures verifiable by keys in
     the builtin system keyring only.
Signed-off-by: David Howells <dhowells@redhat.com>

Commit 3034a146 "ima: pass 'opened' flag to identify newly created files"
stopped identifying empty files as new files.  However new empty files
can be created using the mknodat syscall.  On systems with IMA-appraisal
enabled, these empty files are not labeled with security.ima extended
attributes properly, preventing them from subsequently being opened in
order to write the file data contents.  This patch defines a new hook
named ima_post_path_mknod() to mark these empty files, created using
mknodat, as new in order to allow the file data contents to be written.

In addition, files with security.ima xattrs containing a file signature
are considered "immutable" and can not be modified.  The file contents
need to be written, before signing the file.  This patch relaxes this
requirement for new files, allowing the file signature to be written
before the file contents.

Changelog:
- defer identifying files with signatures stored as security.ima
  (based on Dmitry Rozhkov's comments)
- removing tests (eg. dentry, dentry->d_inode, inode->i_size == 0)
  (based on Al's review)
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
Cc: Al Viro <<viro@zeniv.linux.org.uk>
Tested-by: Dmitry Rozhkov <dmitry.rozhkov@linux.intel.com>

Changing file metadata (eg. uid, guid) could result in having to
re-appraise a file's integrity, but does not change the "new file"
status nor the security.ima xattr.  The IMA_PERMIT_DIRECTIO and
IMA_DIGSIG_REQUIRED flags are policy rule specific.  This patch
only resets these flags, not the IMA_NEW_FILE or IMA_DIGSIG flags.

With this patch, changing the file timestamp will not remove the
file signature on new files.
Reported-by: Dmitry Rozhkov <dmitry.rozhkov@linux.intel.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
Tested-by: Dmitry Rozhkov <dmitry.rozhkov@linux.intel.com>

This patch is based on top of the "vfs: support for a common kernel file
loader" patch set.  In general when the kernel is reading a file into
memory it does not want anything else writing to it.

The kernel currently only forbids write access to a file being executed.
This patch extends this locking to files being read by the kernel.

Changelog:
- moved function to kernel_read_file() - Mimi
- updated patch description - Mimi
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
Reviewed-by: Luis R. Rodriguez <mcgrof@kernel.org>
Acked-by: Kees Cook <keescook@chromium.org>

When I was fixing up const recommendations from checkpatch.pl, I went
overboard. This fixes the warning (during a W=1 build):

include/linux/fs.h:2627:74: warning: type qualifiers ignored on function return type [-Wignored-qualifiers]
static inline const char * const kernel_read_file_id_str(enum kernel_read_file_id id)
Reported-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: James Morris <james.l.morris@oracle.com>

The execstack check was only being applied on the main
process stack.  Thread stacks allocated via mmap were
only subject to the execmem permission check.  Augment
the check to apply to the current thread stack as well.
Note that this does NOT prevent making a different thread's
stack executable.
Suggested-by: Nick Kralevich <nnk@google.com>
Acked-by: Nick Kralevich <nnk@google.com>
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: Paul Moore <paul@paul-moore.com>

Distinguish capability checks against a target associated
with the init user namespace versus capability checks against
a target associated with a non-init user namespace by defining
and using separate security classes for the latter.

This is needed to support e.g. Chrome usage of user namespaces
for the Chrome sandbox without needing to allow Chrome to also
exercise capabilities on targets in the init user namespace.
Suggested-by: Dan Walsh <dwalsh@redhat.com>
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: Paul Moore <paul@paul-moore.com>

This LSM enforces that kernel-loaded files (modules, firmware, etc)
must all come from the same filesystem, with the expectation that
such a filesystem is backed by a read-only device such as dm-verity
or CDROM. This allows systems that have a verified and/or unchangeable
filesystem to enforce module and firmware loading restrictions without
needing to sign the files individually.
Signed-off-by: Kees Cook <keescook@chromium.org>
Acked-by: Serge Hallyn <serge.hallyn@canonical.com>
Signed-off-by: James Morris <james.l.morris@oracle.com>

A string representation of the kernel_read_file_id enumeration is
needed for displaying messages (eg. pr_info, auditing) that can be
used by multiple LSMs and the integrity subsystem.  To simplify
keeping the list of strings up to date with the enumeration, this
patch defines two new preprocessing macros named __fid_enumify and
__fid_stringify to create the enumeration and an array of strings.
kernel_read_file_id_str() returns a string based on the enumeration.
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
[kees: removed removal of my old version, constified pointer values]
Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: James Morris <james.l.morris@oracle.com>

Use a common error reporting function for Yama violation reports, and give
more detail into the process command lines.
Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: James Morris <james.l.morris@oracle.com>

Allocate a NULL-terminated file path with special characters escaped,
safe for logging.
Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: James Morris <james.l.morris@oracle.com>

Provide an escaped (but readable: no inter-argument NULLs) commandline
safe for logging.
Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: James Morris <james.l.morris@oracle.com>

Handle allocating and escaping a string safe for logging.
Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: James Morris <james.l.morris@oracle.com>

There is no point in trying to revalidate an inode's security label if
the security server is not yet initialized.
Signed-off-by: Paul Moore <paul@paul-moore.com>

Since looking up an inode's label can result in revalidation, delay
the lookup as long as possible to limit the performance impact.
Signed-off-by: Paul Moore <paul@paul-moore.com>

There is no point in attempting to revalidate an inode's security
label when we are in the process of setting it.
Reported-by: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: Paul Moore <paul@paul-moore.com>

security_get_bool_value(int bool) argument "bool" conflicts with
in-kernel macros such as BUILD_BUG().  This patch changes this to
index which isn't a type.

Cc: Paul Moore <paul@paul-moore.com>
Cc: Stephen Smalley <sds@tycho.nsa.gov>
Cc: Eric Paris <eparis@parisplace.org>
Cc: James Morris <james.l.morris@oracle.com>
Cc: "Serge E. Hallyn" <serge@hallyn.com>
Cc: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Cc: Andrew Perepechko <anserper@ya.ru>
Cc: Jeff Vander Stoep <jeffv@google.com>
Cc: selinux@tycho.nsa.gov
Cc: Eric Paris <eparis@redhat.com>
Cc: Paul Moore <pmoore@redhat.com>
Cc: David Howells <dhowells@redhat.com>
Signed-off-by: Prarit Bhargava <prarit@redhat.com>
Acked-by: David Howells <dhowells@redhat.com>
[PM: wrapped description for checkpatch.pl, use "selinux:..." as subj]
Signed-off-by: Paul Moore <paul@paul-moore.com>

These commits do the following:

 (1) Retain a signature in an asymmetric-type key and associate with it the
     identifiers that will match a key that can be used to verify it.

 (2) Differentiate an X.509 cert that cannot be used versus one that cannot
     be verified due to unavailable crypto.  This is noted in the
     structures involved.

 (3) Determination of the self-signedness of an X.509 cert is improved to
     include checks on the subject/issuer names and the key
     algorithm/signature algorithm types.

 (4) Self-signed X.509 certificates are consistency checked early on if the
     appropriate crypto is available.
Signed-off-by: David Howells <dhowells@redhat.com>

Miscellaneous keyrings changes.
Signed-off-by: David Howells <dhowells@redhat.com>

This adds userspace access to Diffie-Hellman computations through a
new keyctl() syscall command to calculate shared secrets or public
keys using input parameters stored in the keyring.

Input key ids are provided in a struct due to the current 5-arg limit
for the keyctl syscall. Only user keys are supported in order to avoid
exposing the content of logon or encrypted keys.

The output is written to the provided buffer, based on the assumption
that the values are only needed in userspace.

Future support for other types of key derivation would involve a new
command, like KEYCTL_ECDH_COMPUTE.

Once Diffie-Hellman support is included in the crypto API, this code
can be converted to use the crypto API to take advantage of possible
hardware acceleration and reduce redundant code.
Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: David Howells <dhowells@redhat.com>

Solved TODO task: big keys saved to shmem file are now stored encrypted.
The encryption key is randomly generated and saved to payload[big_key_data].
Signed-off-by: Kirill Marinushkin <k.marinushkin@gmail.com>
Signed-off-by: David Howells <dhowells@redhat.com>

The payload preparsing routine for user keys makes a copy of the payload
provided by the caller and stashes it in the key_preparsed_payload struct for
->instantiate() or ->update() to use.  However, ->update() takes another copy
of this to attach to the keyring.  ->update() should be using this directly
and clearing the pointer in the preparse data.
Signed-off-by: David Howells <dhowells@redhat.com>

Commit d43de6c7 ("akcipher: Move the RSA DER encoding check to
the crypto layer") removed the Kconfig option PUBLIC_KEY_ALGO_RSA,
but forgot to remove a 'select' to this option in the definition of
INTEGRITY_ASYMMETRIC_KEYS.

Let's remove the select, as it's ineffective now.
Signed-off-by: Andreas Ziegler <andreas.ziegler@fau.de>
Signed-off-by: David Howells <dhowells@redhat.com>

Add a config option (IMA_KEYRINGS_PERMIT_SIGNED_BY_BUILTIN_OR_SECONDARY)
that, when enabled, allows keys to be added to the IMA keyrings by
userspace - with the restriction that each must be signed by a key in the
system trusted keyrings.

EPERM will be returned if this option is disabled, ENOKEY will be returned if
no authoritative key can be found and EKEYREJECTED will be returned if the
signature doesn't match.  Other errors such as ENOPKG may also be returned.

If this new option is enabled, the builtin system keyring is searched, as is
the secondary system keyring if that is also enabled.  Intermediate keys
between the builtin system keyring and the key being added can be added to
the secondary keyring (which replaces .ima_mok) to form a trust chain -
provided they are also validly signed by a key in one of the trusted keyrings.

The .ima_mok keyring is then removed and the IMA blacklist keyring gets its
own config option (IMA_BLACKLIST_KEYRING).
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>

Add a secondary system keyring that can be added to by root whilst the
system is running - provided the key being added is vouched for by a key
built into the kernel or already added to the secondary keyring.

Rename .system_keyring to .builtin_trusted_keys to distinguish it more
obviously from the new keyring (called .secondary_trusted_keys).

The new keyring needs to be enabled with CONFIG_SECONDARY_TRUSTED_KEYRING.

If the secondary keyring is enabled, a link is created from that to
.builtin_trusted_keys so that the the latter will automatically be searched
too if the secondary keyring is searched.
Signed-off-by: David Howells <dhowells@redhat.com>

Remove KEY_FLAG_TRUSTED and KEY_ALLOC_TRUSTED as they're no longer
meaningful.  Also we can drop the trusted flag from the preparse structure.

Given this, we no longer need to pass the key flags through to
restrict_link().

Further, we can now get rid of keyring_restrict_trusted_only() also.
Signed-off-by: David Howells <dhowells@redhat.com>

Move the point at which a key is determined to be trustworthy to
__key_link() so that we use the contents of the keyring being linked in to
to determine whether the key being linked in is trusted or not.

What is 'trusted' then becomes a matter of what's in the keyring.

Currently, the test is done when the key is parsed, but given that at that
point we can only sensibly refer to the contents of the system trusted
keyring, we can only use that as the basis for working out the
trustworthiness of a new key.

With this change, a trusted keyring is a set of keys that once the
trusted-only flag is set cannot be added to except by verification through
one of the contained keys.

Further, adding a key into a trusted keyring, whilst it might grant
trustworthiness in the context of that keyring, does not automatically
grant trustworthiness in the context of a second keyring to which it could
be secondarily linked.

To accomplish this, the authentication data associated with the key source
must now be retained.  For an X.509 cert, this means the contents of the
AuthorityKeyIdentifier and the signature data.


If system keyrings are disabled then restrict_link_by_builtin_trusted()
resolves to restrict_link_reject().  The integrity digital signature code
still works correctly with this as it was previously using
KEY_FLAG_TRUSTED_ONLY, which doesn't permit anything to be added if there
is no system keyring against which trust can be determined.
Signed-off-by: David Howells <dhowells@redhat.com>

Make the system trusted keyring depend on the asymmetric key type as
there's not a lot of point having it if you can't then load asymmetric keys
onto it.

This requires the ASYMMETRIC_KEY_TYPE to be made a bool, not a tristate, as
the Kconfig language doesn't then correctly force ASYMMETRIC_KEY_TYPE to
'y' rather than 'm' if SYSTEM_TRUSTED_KEYRING is 'y'.

Making SYSTEM_TRUSTED_KEYRING *select* ASYMMETRIC_KEY_TYPE instead doesn't
work as the Kconfig interpreter then wrongly complains about dependency
loops.
Signed-off-by: David Howells <dhowells@redhat.com>

Move the X.509 trust validation code out to its own file so that it can be
generalised.
Signed-off-by: David Howells <dhowells@redhat.com>

We should call verify_signature() rather than directly calling
public_key_verify_signature() if we have a struct key to use as we
shouldn't be poking around in the private data of the key struct as that's
subtype dependent.
Signed-off-by: David Howells <dhowells@redhat.com>

Generalise x509_request_asymmetric_key().  It doesn't really have any
dependencies on X.509 features as it uses generalised IDs and the
public_key structs that contain data extracted from X.509.
Signed-off-by: David Howells <dhowells@redhat.com>

Move x509_request_asymmetric_key() to asymmetric_type.c so that it can be
generalised.
Signed-off-by: David Howells <dhowells@redhat.com>

Add a facility whereby proposed new links to be added to a keyring can be
vetted, permitting them to be rejected if necessary.  This can be used to
block public keys from which the signature cannot be verified or for which
the signature verification fails.  It could also be used to provide
blacklisting.

This affects operations like add_key(), KEYCTL_LINK and KEYCTL_INSTANTIATE.

To this end:

 (1) A function pointer is added to the key struct that, if set, points to
     the vetting function.  This is called as:

	int (*restrict_link)(struct key *keyring,
			     const struct key_type *key_type,
			     unsigned long key_flags,
			     const union key_payload *key_payload),

     where 'keyring' will be the keyring being added to, key_type and
     key_payload will describe the key being added and key_flags[*] can be
     AND'ed with KEY_FLAG_TRUSTED.

     [*] This parameter will be removed in a later patch when
     	 KEY_FLAG_TRUSTED is removed.

     The function should return 0 to allow the link to take place or an
     error (typically -ENOKEY, -ENOPKG or -EKEYREJECTED) to reject the
     link.

     The pointer should not be set directly, but rather should be set
     through keyring_alloc().

     Note that if called during add_key(), preparse is called before this
     method, but a key isn't actually allocated until after this function
     is called.

 (2) KEY_ALLOC_BYPASS_RESTRICTION is added.  This can be passed to
     key_create_or_update() or key_instantiate_and_link() to bypass the
     restriction check.

 (3) KEY_FLAG_TRUSTED_ONLY is removed.  The entire contents of a keyring
     with this restriction emplaced can be considered 'trustworthy' by
     virtue of being in the keyring when that keyring is consulted.

 (4) key_alloc() and keyring_alloc() take an extra argument that will be
     used to set restrict_link in the new key.  This ensures that the
     pointer is set before the key is published, thus preventing a window
     of unrestrictedness.  Normally this argument will be NULL.

 (5) As a temporary affair, keyring_restrict_trusted_only() is added.  It
     should be passed to keyring_alloc() as the extra argument instead of
     setting KEY_FLAG_TRUSTED_ONLY on a keyring.  This will be replaced in
     a later patch with functions that look in the appropriate places for
     authoritative keys.
Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Mimi Zohar <zohar@linux.vnet.ibm.com>

The skb_owned_by hook was added with the commit ca10b9e9
("selinux: add a skb_owned_by() hook") and later removed
when said commit was reverted.

Later on, when switching to list of hooks, a field named
'skb_owned_by' was included into the security_hook_head struct,
but without any users nor caller.

This commit removes the said left-over field.

Fixes: b1d9e6b0 ("LSM: Switch to lists of hooks")
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Acked-by: Casey Schaufler <casey@schaufler-ca.com>
Acked-by: Paul Moore <pmoore@paul-moore.com>
Signed-off-by: James Morris <james.l.morris@oracle.com>

Make the determination of the trustworthiness of a key dependent on whether
a key that can verify it is present in the supplied ring of trusted keys
rather than whether or not the verifying key has KEY_FLAG_TRUSTED set.

verify_pkcs7_signature() will return -ENOKEY if the PKCS#7 message trust
chain cannot be verified.
Signed-off-by: David Howells <dhowells@redhat.com>

Generalise system_verify_data() to provide access to internal content
through a callback.  This allows all the PKCS#7 stuff to be hidden inside
this function and removed from the PE file parser and the PKCS#7 test key.

If external content is not required, NULL should be passed as data to the
function.  If the callback is not required, that can be set to NULL.

The function is now called verify_pkcs7_signature() to contrast with
verify_pefile_signature() and the definitions of both have been moved into
linux/verification.h along with the key_being_used_for enum.
Signed-off-by: David Howells <dhowells@redhat.com>