Convert avc_audit in security/selinux/avc.c to use lsm_audit.h,
for better maintainability.
- changed selinux to use common_audit_data instead of
avc_audit_data
- eliminated code in avc.c and used code from lsm_audit.h instead.
Had to add a LSM_AUDIT_NO_AUDIT to lsm_audit.h so that avc_audit
can call common_lsm_audit and do the pre and post callbacks without
doing the actual dump. This makes it so that the patched version
behaves the same way as the unpatched version.
Also added a denied field to the selinux_audit_data private space,
once again to make it so that the patched version behaves like the
unpatched.
I've tested and confirmed that AVCs look the same before and after
this patch.
Signed-off-by: Thomas Liu <tliu@redhat.com>
Acked-by: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: James Morris <jmorris@namei.org>
Convert avc_audit in security/selinux/avc.c to use lsm_audit.h,
for better maintainability and for less code duplication.
- changed selinux to use common_audit_data instead of
avc_audit_data
- eliminated code in avc.c and used code from lsm_audit.h instead.
I have tested to make sure that the avcs look the same before and
after this patch.
Signed-off-by: Thomas Liu <tliu@redhat.com>
Acked-by: Eric Paris <eparis@redhat.com>
Signed-off-by: James Morris <jmorris@namei.org>
Define three accessors to get/set dst attached to a skb
struct dst_entry *skb_dst(const struct sk_buff *skb)
void skb_dst_set(struct sk_buff *skb, struct dst_entry *dst)
void skb_dst_drop(struct sk_buff *skb)
This one should replace occurrences of :
dst_release(skb->dst)
skb->dst = NULL;
Delete skb->dst field
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Wrap current->cred and a few other accessors to hide their actual
implementation.
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: James Morris <jmorris@namei.org>
Acked-by: Serge Hallyn <serue@us.ibm.com>
Signed-off-by: James Morris <jmorris@namei.org>
Separate the task security context from task_struct. At this point, the
security data is temporarily embedded in the task_struct with two pointers
pointing to it.
Note that the Alpha arch is altered as it refers to (E)UID and (E)GID in
entry.S via asm-offsets.
With comment fixes Signed-off-by: Marc Dionne <marc.c.dionne@gmail.com>
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: James Morris <jmorris@namei.org>
Acked-by: Serge Hallyn <serue@us.ibm.com>
Signed-off-by: James Morris <jmorris@namei.org>
This patch changes xfrm.c to fix whitespace and syntax issues. Things that
are fixed may include (does not not have to include)
whitespace at end of lines
spaces followed by tabs
spaces used instead of tabs
spacing around parenthesis
locateion of { around struct and else clauses
location of * in pointer declarations
removal of initialization of static data to keep it in the right section
useless {} in if statemetns
useless checking for NULL before kfree
fixing of the indentation depth of switch statements
and any number of other things I forgot to mention
Signed-off-by: Eric Paris <eparis@redhat.com>
Signed-off-by: James Morris <jmorris@namei.org>
The xfrm_get_policy() and xfrm_add_pol_expire() put some rather large structs
on the stack to work around the LSM API. This patch attempts to fix that
problem by changing the LSM API to require only the relevant "security"
pointers instead of the entire SPD entry; we do this for all of the
security_xfrm_policy*() functions to keep things consistent.
Signed-off-by: Paul Moore <paul.moore@hp.com>
Acked-by: James Morris <jmorris@namei.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch introduces a mechanism for checking when labeled IPsec or SECMARK
are in use by keeping introducing a configuration reference counter for each
subsystem. In the case of labeled IPsec, whenever a labeled SA or SPD entry
is created the labeled IPsec/XFRM reference count is increased and when the
entry is removed it is decreased. In the case of SECMARK, when a SECMARK
target is created the reference count is increased and later decreased when the
target is removed. These reference counters allow SELinux to quickly determine
if either of these subsystems are enabled.
NetLabel already has a similar mechanism which provides the netlbl_enabled()
function.
This patch also renames the selinux_relabel_packet_permission() function to
selinux_secmark_relabel_packet_permission() as the original name and
description were misleading in that they referenced a single packet label which
is not the case.
Signed-off-by: Paul Moore <paul.moore@hp.com>
Signed-off-by: James Morris <jmorris@namei.org>
On PowerPC allmodconfig build we get this:
security/selinux/xfrm.c:214: warning: comparison is always false due to limited range of data type
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: James Morris <jmorris@namei.org>
Get rid of sparse related warnings from places that use integer as NULL
pointer.
[akpm@linux-foundation.org: coding-style fixes]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Cc: Andi Kleen <ak@suse.de>
Cc: Jeff Garzik <jeff@garzik.org>
Cc: Matt Mackall <mpm@selenic.com>
Cc: Ian Kent <raven@themaw.net>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Davide Libenzi <davidel@xmailserver.org>
Cc: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Convert LSM into a static interface, as the ability to unload a security
module is not required by in-tree users and potentially complicates the
overall security architecture.
Needlessly exported LSM symbols have been unexported, to help reduce API
abuse.
Parameters for the capability and root_plug modules are now specified
at boot.
The SECURITY_FRAMEWORK_VERSION macro has also been removed.
In a nutshell, there is no safe way to unload an LSM. The modular interface
is thus unecessary and broken infrastructure. It is used only by out-of-tree
modules, which are often binary-only, illegal, abusive of the API and
dangerous, e.g. silently re-vectoring SELinux.
[akpm@linux-foundation.org: cleanups]
[akpm@linux-foundation.org: USB Kconfig fix]
[randy.dunlap@oracle.com: fix LSM kernel-doc]
Signed-off-by: James Morris <jmorris@namei.org>
Acked-by: Chris Wright <chrisw@sous-sol.org>
Cc: Stephen Smalley <sds@tycho.nsa.gov>
Cc: "Serge E. Hallyn" <serue@us.ibm.com>
Acked-by: Arjan van de Ven <arjan@infradead.org>
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
xfrm_audit_log() expects the context string to be null-terminated
which currently doesn't happen with user-supplied contexts.
Signed-off-by: Venkat Yekkirala <vyekkirala@TrustedCS.com>
Acked-by: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: James Morris <jmorris@namei.org>
Now that labeled IPsec makes use of the peer_sid field in the
sk_security_struct we can remove a lot of the special cases between labeled
IPsec and NetLabel. In addition, create a new function,
security_skb_extlbl_sid(), which we can use in several places to get the
security context of the packet's external label which allows us to further
simplify the code in a few places.
Signed-off-by: Paul Moore <paul.moore@hp.com>
Signed-off-by: James Morris <jmorris@namei.org>
Fix the selection of an SA for an outgoing packet to be at the same
context as the originating socket/flow. This eliminates the SELinux
policy's ability to use/sendto SAs with contexts other than the socket's.
With this patch applied, the SELinux policy will require one or more of the
following for a socket to be able to communicate with/without SAs:
1. To enable a socket to communicate without using labeled-IPSec SAs:
allow socket_t unlabeled_t:association { sendto recvfrom }
2. To enable a socket to communicate with labeled-IPSec SAs:
allow socket_t self:association { sendto };
allow socket_t peer_sa_t:association { recvfrom };
Signed-off-by: Venkat Yekkirala <vyekkirala@TrustedCS.com>
Signed-off-by: James Morris <jmorris@namei.org>
Fix SO_PEERSEC for tcp sockets to return the security context of
the peer (as represented by the SA from the peer) as opposed to the
SA used by the local/source socket.
Signed-off-by: Venkat Yekkirala <vyekkirala@TrustedCS.com>
Signed-off-by: James Morris <jmorris@namei.org>
Since the upstreaming of the mlsxfrm modification a few months back,
testing has resulted in the identification of the following issues/bugs that
are resolved in this patch set.
1. Fix the security context used in the IKE negotiation to be the context
of the socket as opposed to the context of the SPD rule.
2. Fix SO_PEERSEC for tcp sockets to return the security context of
the peer as opposed to the source.
3. Fix the selection of an SA for an outgoing packet to be at the same
context as the originating socket/flow.
The following would be the result of applying this patchset:
- SO_PEERSEC will now correctly return the peer's context.
- IKE deamons will receive the context of the source socket/flow
as opposed to the SPD rule's context so that the negotiated SA
will be at the same context as the source socket/flow.
- The SELinux policy will require one or more of the
following for a socket to be able to communicate with/without SAs:
1. To enable a socket to communicate without using labeled-IPSec SAs:
allow socket_t unlabeled_t:association { sendto recvfrom }
2. To enable a socket to communicate with labeled-IPSec SAs:
allow socket_t self:association { sendto };
allow socket_t peer_sa_t:association { recvfrom };
This Patch: Pass correct security context to IKE for use in negotiation
Fix the security context passed to IKE for use in negotiation to be the
context of the socket as opposed to the context of the SPD rule so that
the SA carries the label of the originating socket/flow.
Signed-off-by: Venkat Yekkirala <vyekkirala@TrustedCS.com>
Signed-off-by: James Morris <jmorris@namei.org>
Currently when an IPSec policy rule doesn't specify a security
context, it is assumed to be "unlabeled" by SELinux, and so
the IPSec policy rule fails to match to a flow that it would
otherwise match to, unless one has explicitly added an SELinux
policy rule allowing the flow to "polmatch" to the "unlabeled"
IPSec policy rules. In the absence of such an explicitly added
SELinux policy rule, the IPSec policy rule fails to match and
so the packet(s) flow in clear text without the otherwise applicable
xfrm(s) applied.
The above SELinux behavior violates the SELinux security notion of
"deny by default" which should actually translate to "encrypt by
default" in the above case.
This was first reported by Evgeniy Polyakov and the way James Morris
was seeing the problem was when connecting via IPsec to a
confined service on an SELinux box (vsftpd), which did not have the
appropriate SELinux policy permissions to send packets via IPsec.
With this patch applied, SELinux "polmatching" of flows Vs. IPSec
policy rules will only come into play when there's a explicit context
specified for the IPSec policy rule (which also means there's corresponding
SELinux policy allowing appropriate domains/flows to polmatch to this context).
Secondly, when a security module is loaded (in this case, SELinux), the
security_xfrm_policy_lookup() hook can return errors other than access denied,
such as -EINVAL. We were not handling that correctly, and in fact
inverting the return logic and propagating a false "ok" back up to
xfrm_lookup(), which then allowed packets to pass as if they were not
associated with an xfrm policy.
The solution for this is to first ensure that errno values are
correctly propagated all the way back up through the various call chains
from security_xfrm_policy_lookup(), and handled correctly.
Then, flow_cache_lookup() is modified, so that if the policy resolver
fails (typically a permission denied via the security module), the flow
cache entry is killed rather than having a null policy assigned (which
indicates that the packet can pass freely). This also forces any future
lookups for the same flow to consult the security module (e.g. SELinux)
for current security policy (rather than, say, caching the error on the
flow cache entry).
This patch: Fix the selinux side of things.
This makes sure SELinux polmatching of flow contexts to IPSec policy
rules comes into play only when an explicit context is associated
with the IPSec policy rule.
Also, this no longer defaults the context of a socket policy to
the context of the socket since the "no explicit context" case
is now handled properly.
Signed-off-by: Venkat Yekkirala <vyekkirala@TrustedCS.com>
Signed-off-by: James Morris <jmorris@namei.org>
This automatically labels the TCP, Unix stream, and dccp child sockets
as well as openreqs to be at the same MLS level as the peer. This will
result in the selection of appropriately labeled IPSec Security
Associations.
This also uses the sock's sid (as opposed to the isec sid) in SELinux
enforcement of secmark in rcv_skb and postroute_last hooks.
Signed-off-by: Venkat Yekkirala <vyekkirala@TrustedCS.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This defaults the label of socket-specific IPSec policies to be the
same as the socket they are set on.
Signed-off-by: Venkat Yekkirala <vyekkirala@TrustedCS.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This labels the flows that could utilize IPSec xfrms at the points the
flows are defined so that IPSec policy and SAs at the right label can
be used.
The following protos are currently not handled, but they should
continue to be able to use single-labeled IPSec like they currently
do.
ipmr
ip_gre
ipip
igmp
sit
sctp
ip6_tunnel (IPv6 over IPv6 tunnel device)
decnet
Signed-off-by: Venkat Yekkirala <vyekkirala@TrustedCS.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This implements a seemless mechanism for xfrm policy selection and
state matching based on the flow sid. This also includes the necessary
SELinux enforcement pieces.
Signed-off-by: Venkat Yekkirala <vyekkirala@TrustedCS.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add new per-packet access controls to SELinux, replacing the old
packet controls.
Packets are labeled with the iptables SECMARK and CONNSECMARK targets,
then security policy for the packets is enforced with these controls.
To allow for a smooth transition to the new controls, the old code is
still present, but not active by default. To restore previous
behavior, the old controls may be activated at runtime by writing a
'1' to /selinux/compat_net, and also via the kernel boot parameter
selinux_compat_net. Switching between the network control models
requires the security load_policy permission. The old controls will
probably eventually be removed and any continued use is discouraged.
With this patch, the new secmark controls for SElinux are disabled by
default, so existing behavior is entirely preserved, and the user is
not affected at all.
It also provides a config option to enable the secmark controls by
default (which can always be overridden at boot and runtime). It is
also noted in the kconfig help that the user will need updated
userspace if enabling secmark controls for SELinux and that they'll
probably need the SECMARK and CONNMARK targets, and conntrack protocol
helpers, although such decisions are beyond the scope of kernel
configuration.
Signed-off-by: James Morris <jmorris@namei.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch contains a fix for the previous patch that adds security
contexts to IPsec policies and security associations. In the previous
patch, no authorization (besides the check for write permissions to
SAD and SPD) is required to delete IPsec policies and security
assocations with security contexts. Thus a user authorized to change
SAD and SPD can bypass the IPsec policy authorization by simply
deleteing policies with security contexts. To fix this security hole,
an additional authorization check is added for removing security
policies and security associations with security contexts.
Note that if no security context is supplied on add or present on
policy to be deleted, the SELinux module allows the change
unconditionally. The hook is called on deletion when no context is
present, which we may want to change. At present, I left it up to the
module.
LSM changes:
The patch adds two new LSM hooks: xfrm_policy_delete and
xfrm_state_delete. The new hooks are necessary to authorize deletion
of IPsec policies that have security contexts. The existing hooks
xfrm_policy_free and xfrm_state_free lack the context to do the
authorization, so I decided to split authorization of deletion and
memory management of security data, as is typical in the LSM
interface.
Use:
The new delete hooks are checked when xfrm_policy or xfrm_state are
deleted by either the xfrm_user interface (xfrm_get_policy,
xfrm_del_sa) or the pfkey interface (pfkey_spddelete, pfkey_delete).
SELinux changes:
The new policy_delete and state_delete functions are added.
Signed-off-by: Catherine Zhang <cxzhang@watson.ibm.com>
Signed-off-by: Trent Jaeger <tjaeger@cse.psu.edu>
Acked-by: James Morris <jmorris@namei.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
security/selinux/xfrm.c: In function 'selinux_socket_getpeer_dgram':
security/selinux/xfrm.c:284: error: 'struct sec_path' has no member named 'x'
security/selinux/xfrm.c: In function 'selinux_xfrm_sock_rcv_skb':
security/selinux/xfrm.c:317: error: 'struct sec_path' has no member named 'x'
Signed-off-by: Dave Jones <davej@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch implements an application of the LSM-IPSec networking
controls whereby an application can determine the label of the
security association its TCP or UDP sockets are currently connected to
via getsockopt and the auxiliary data mechanism of recvmsg.
Patch purpose:
This patch enables a security-aware application to retrieve the
security context of an IPSec security association a particular TCP or
UDP socket is using. The application can then use this security
context to determine the security context for processing on behalf of
the peer at the other end of this connection. In the case of UDP, the
security context is for each individual packet. An example
application is the inetd daemon, which could be modified to start
daemons running at security contexts dependent on the remote client.
Patch design approach:
- Design for TCP
The patch enables the SELinux LSM to set the peer security context for
a socket based on the security context of the IPSec security
association. The application may retrieve this context using
getsockopt. When called, the kernel determines if the socket is a
connected (TCP_ESTABLISHED) TCP socket and, if so, uses the dst_entry
cache on the socket to retrieve the security associations. If a
security association has a security context, the context string is
returned, as for UNIX domain sockets.
- Design for UDP
Unlike TCP, UDP is connectionless. This requires a somewhat different
API to retrieve the peer security context. With TCP, the peer
security context stays the same throughout the connection, thus it can
be retrieved at any time between when the connection is established
and when it is torn down. With UDP, each read/write can have
different peer and thus the security context might change every time.
As a result the security context retrieval must be done TOGETHER with
the packet retrieval.
The solution is to build upon the existing Unix domain socket API for
retrieving user credentials. Linux offers the API for obtaining user
credentials via ancillary messages (i.e., out of band/control messages
that are bundled together with a normal message).
Patch implementation details:
- Implementation for TCP
The security context can be retrieved by applications using getsockopt
with the existing SO_PEERSEC flag. As an example (ignoring error
checking):
getsockopt(sockfd, SOL_SOCKET, SO_PEERSEC, optbuf, &optlen);
printf("Socket peer context is: %s\n", optbuf);
The SELinux function, selinux_socket_getpeersec, is extended to check
for labeled security associations for connected (TCP_ESTABLISHED ==
sk->sk_state) TCP sockets only. If so, the socket has a dst_cache of
struct dst_entry values that may refer to security associations. If
these have security associations with security contexts, the security
context is returned.
getsockopt returns a buffer that contains a security context string or
the buffer is unmodified.
- Implementation for UDP
To retrieve the security context, the application first indicates to
the kernel such desire by setting the IP_PASSSEC option via
getsockopt. Then the application retrieves the security context using
the auxiliary data mechanism.
An example server application for UDP should look like this:
toggle = 1;
toggle_len = sizeof(toggle);
setsockopt(sockfd, SOL_IP, IP_PASSSEC, &toggle, &toggle_len);
recvmsg(sockfd, &msg_hdr, 0);
if (msg_hdr.msg_controllen > sizeof(struct cmsghdr)) {
cmsg_hdr = CMSG_FIRSTHDR(&msg_hdr);
if (cmsg_hdr->cmsg_len <= CMSG_LEN(sizeof(scontext)) &&
cmsg_hdr->cmsg_level == SOL_IP &&
cmsg_hdr->cmsg_type == SCM_SECURITY) {
memcpy(&scontext, CMSG_DATA(cmsg_hdr), sizeof(scontext));
}
}
ip_setsockopt is enhanced with a new socket option IP_PASSSEC to allow
a server socket to receive security context of the peer. A new
ancillary message type SCM_SECURITY.
When the packet is received we get the security context from the
sec_path pointer which is contained in the sk_buff, and copy it to the
ancillary message space. An additional LSM hook,
selinux_socket_getpeersec_udp, is defined to retrieve the security
context from the SELinux space. The existing function,
selinux_socket_getpeersec does not suit our purpose, because the
security context is copied directly to user space, rather than to
kernel space.
Testing:
We have tested the patch by setting up TCP and UDP connections between
applications on two machines using the IPSec policies that result in
labeled security associations being built. For TCP, we can then
extract the peer security context using getsockopt on either end. For
UDP, the receiving end can retrieve the security context using the
auxiliary data mechanism of recvmsg.
Signed-off-by: Catherine Zhang <cxzhang@watson.ibm.com>
Acked-by: James Morris <jmorris@namei.org>
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
security/selinux/xfrm.c:155:10: warning: Using plain integer as NULL pointer
Signed-off-by: Luiz Capitulino <lcapitulino@mandriva.com.br>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch contains two corrections to the LSM-IPsec Nethooks patches
previously applied.
(1) free a security context on a failed insert via xfrm_user
interface in xfrm_add_policy. Memory leak.
(2) change the authorization of the allocation of a security context
in a xfrm_policy or xfrm_state from both relabelfrom and relabelto
to setcontext.
Signed-off-by: Trent Jaeger <tjaeger@cse.psu.edu>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch series implements per packet access control via the
extension of the Linux Security Modules (LSM) interface by hooks in
the XFRM and pfkey subsystems that leverage IPSec security
associations to label packets. Extensions to the SELinux LSM are
included that leverage the patch for this purpose.
This patch implements the changes necessary to the SELinux LSM to
create, deallocate, and use security contexts for policies
(xfrm_policy) and security associations (xfrm_state) that enable
control of a socket's ability to send and receive packets.
Patch purpose:
The patch is designed to enable the SELinux LSM to implement access
control on individual packets based on the strongly authenticated
IPSec security association. Such access controls augment the existing
ones in SELinux based on network interface and IP address. The former
are very coarse-grained, and the latter can be spoofed. By using
IPSec, the SELinux can control access to remote hosts based on
cryptographic keys generated using the IPSec mechanism. This enables
access control on a per-machine basis or per-application if the remote
machine is running the same mechanism and trusted to enforce the
access control policy.
Patch design approach:
The patch's main function is to authorize a socket's access to a IPSec
policy based on their security contexts. Since the communication is
implemented by a security association, the patch ensures that the
security association's negotiated and used have the same security
context. The patch enables allocation and deallocation of such
security contexts for policies and security associations. It also
enables copying of the security context when policies are cloned.
Lastly, the patch ensures that packets that are sent without using a
IPSec security assocation with a security context are allowed to be
sent in that manner.
A presentation available at
www.selinux-symposium.org/2005/presentations/session2/2-3-jaeger.pdf
from the SELinux symposium describes the overall approach.
Patch implementation details:
The function which authorizes a socket to perform a requested
operation (send/receive) on a IPSec policy (xfrm_policy) is
selinux_xfrm_policy_lookup. The Netfilter and rcv_skb hooks ensure
that if a IPSec SA with a securit y association has not been used,
then the socket is allowed to send or receive the packet,
respectively.
The patch implements SELinux function for allocating security contexts
when policies (xfrm_policy) are created via the pfkey or xfrm_user
interfaces via selinux_xfrm_policy_alloc. When a security association
is built, SELinux allocates the security context designated by the
XFRM subsystem which is based on that of the authorized policy via
selinux_xfrm_state_alloc.
When a xfrm_policy is cloned, the security context of that policy, if
any, is copied to the clone via selinux_xfrm_policy_clone.
When a xfrm_policy or xfrm_state is freed, its security context, if
any is also freed at selinux_xfrm_policy_free or
selinux_xfrm_state_free.
Testing:
The SELinux authorization function is tested using ipsec-tools. We
created policies and security associations with particular security
contexts and added SELinux access control policy entries to verify the
authorization decision. We also made sure that packets for which no
security context was supplied (which either did or did not use
security associations) were authorized using an unlabelled context.
Signed-off-by: Trent Jaeger <tjaeger@cse.psu.edu>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
