Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
/*-
|
|
|
|
* Copyright (c) 1999, 2000, 2001, 2002 Robert N. M. Watson
|
2003-08-21 17:28:45 +00:00
|
|
|
* Copyright (c) 2001, 2002, 2003 Networks Associates Technology, Inc.
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
* All rights reserved.
|
|
|
|
*
|
|
|
|
* This software was developed by Robert Watson for the TrustedBSD Project.
|
|
|
|
*
|
2002-11-04 01:53:12 +00:00
|
|
|
* This software was developed for the FreeBSD Project in part by Network
|
|
|
|
* Associates Laboratories, the Security Research Division of Network
|
|
|
|
* Associates, Inc. under DARPA/SPAWAR contract N66001-01-C-8035 ("CBOSS"),
|
|
|
|
* as part of the DARPA CHATS research program.
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
*
|
|
|
|
* Redistribution and use in source and binary forms, with or without
|
|
|
|
* modification, are permitted provided that the following conditions
|
|
|
|
* are met:
|
|
|
|
* 1. Redistributions of source code must retain the above copyright
|
|
|
|
* notice, this list of conditions and the following disclaimer.
|
|
|
|
* 2. Redistributions in binary form must reproduce the above copyright
|
|
|
|
* notice, this list of conditions and the following disclaimer in the
|
|
|
|
* documentation and/or other materials provided with the distribution.
|
|
|
|
*
|
|
|
|
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
|
|
|
|
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
|
|
|
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
|
|
|
|
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
|
|
|
|
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
|
|
|
|
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
|
|
|
|
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
|
|
|
|
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
|
|
|
|
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
|
|
|
|
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
|
|
|
|
* SUCH DAMAGE.
|
|
|
|
*
|
|
|
|
* $FreeBSD$
|
|
|
|
*/
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Developed by the TrustedBSD Project.
|
|
|
|
* Generic mandatory access module that does nothing.
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include <sys/types.h>
|
|
|
|
#include <sys/param.h>
|
|
|
|
#include <sys/acl.h>
|
|
|
|
#include <sys/conf.h>
|
Slightly change the semantics of vnode labels for MAC: rather than
"refreshing" the label on the vnode before use, just get the label
right from inception. For single-label file systems, set the label
in the generic VFS getnewvnode() code; for multi-label file systems,
leave the labeling up to the file system. With UFS1/2, this means
reading the extended attribute during vfs_vget() as the inode is
pulled off disk, rather than hitting the extended attributes
frequently during operations later, improving performance. This
also corrects sematics for shared vnode locks, which were not
previously present in the system. This chances the cache
coherrency properties WRT out-of-band access to label data, but in
an acceptable form. With UFS1, there is a small race condition
during automatic extended attribute start -- this is not present
with UFS2, and occurs because EAs aren't available at vnode
inception. We'll introduce a work around for this shortly.
Approved by: re
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-26 14:38:24 +00:00
|
|
|
#include <sys/extattr.h>
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
#include <sys/kernel.h>
|
|
|
|
#include <sys/mac.h>
|
2003-04-15 21:20:34 +00:00
|
|
|
#include <sys/malloc.h>
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
#include <sys/mount.h>
|
|
|
|
#include <sys/proc.h>
|
|
|
|
#include <sys/systm.h>
|
|
|
|
#include <sys/sysproto.h>
|
|
|
|
#include <sys/sysent.h>
|
|
|
|
#include <sys/vnode.h>
|
|
|
|
#include <sys/file.h>
|
|
|
|
#include <sys/socket.h>
|
|
|
|
#include <sys/socketvar.h>
|
|
|
|
#include <sys/sysctl.h>
|
|
|
|
|
|
|
|
#include <fs/devfs/devfs.h>
|
|
|
|
|
|
|
|
#include <net/bpfdesc.h>
|
|
|
|
#include <net/if.h>
|
|
|
|
#include <net/if_types.h>
|
|
|
|
#include <net/if_var.h>
|
|
|
|
|
|
|
|
#include <vm/vm.h>
|
|
|
|
|
|
|
|
#include <sys/mac_policy.h>
|
|
|
|
|
|
|
|
SYSCTL_DECL(_security_mac);
|
|
|
|
|
|
|
|
SYSCTL_NODE(_security_mac, OID_AUTO, test, CTLFLAG_RW, 0,
|
|
|
|
"TrustedBSD mac_test policy controls");
|
|
|
|
|
2002-12-10 16:20:34 +00:00
|
|
|
static int mac_test_enabled = 1;
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, enabled, CTLFLAG_RW,
|
|
|
|
&mac_test_enabled, 0, "Enforce test policy");
|
|
|
|
|
|
|
|
#define BPFMAGIC 0xfe1ad1b6
|
|
|
|
#define DEVFSMAGIC 0x9ee79c32
|
|
|
|
#define IFNETMAGIC 0xc218b120
|
|
|
|
#define IPQMAGIC 0x206188ef
|
|
|
|
#define MBUFMAGIC 0xbbefa5bb
|
|
|
|
#define MOUNTMAGIC 0xc7c46e47
|
|
|
|
#define SOCKETMAGIC 0x9199c6cd
|
|
|
|
#define PIPEMAGIC 0xdc6c9919
|
2003-03-25 01:20:56 +00:00
|
|
|
#define PROCMAGIC 0x3b4be98f
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
#define CREDMAGIC 0x9a5a4987
|
|
|
|
#define VNODEMAGIC 0x1a67a45c
|
|
|
|
#define EXMAGIC 0x849ba1fd
|
|
|
|
|
|
|
|
#define SLOT(x) LABEL_TO_SLOT((x), test_slot).l_long
|
2003-08-21 17:28:45 +00:00
|
|
|
|
|
|
|
#define ASSERT_BPF_LABEL(x) KASSERT(SLOT(x) == BPFMAGIC || \
|
|
|
|
SLOT(x) == 0, ("%s: Bad BPF label", __func__ ))
|
|
|
|
#define ASSERT_DEVFS_LABEL(x) KASSERT(SLOT(x) == DEVFSMAGIC || \
|
|
|
|
SLOT(x) == 0, ("%s: Bad DEVFS label", __func__ ))
|
|
|
|
#define ASSERT_IFNET_LABEL(x) KASSERT(SLOT(x) == IFNETMAGIC || \
|
|
|
|
SLOT(x) == 0, ("%s: Bad IFNET label", __func__ ))
|
|
|
|
#define ASSERT_IPQ_LABEL(x) KASSERT(SLOT(x) == IPQMAGIC || \
|
|
|
|
SLOT(x) == 0, ("%s: Bad IPQ label", __func__ ))
|
|
|
|
#define ASSERT_MBUF_LABEL(x) KASSERT(SLOT(x) == MBUFMAGIC || \
|
|
|
|
SLOT(x) == 0, ("%s: Bad MBUF label", __func__ ))
|
|
|
|
#define ASSERT_MOUNT_LABEL(x) KASSERT(SLOT(x) == MOUNTMAGIC || \
|
|
|
|
SLOT(x) == 0, ("%s: Bad MOUNT label", __func__ ))
|
|
|
|
#define ASSERT_SOCKET_LABEL(x) KASSERT(SLOT(x) == SOCKETMAGIC || \
|
|
|
|
SLOT(x) == 0, ("%s: Bad SOCKET label", __func__ ))
|
|
|
|
#define ASSERT_PIPE_LABEL(x) KASSERT(SLOT(x) == PIPEMAGIC || \
|
|
|
|
SLOT(x) == 0, ("%s: Bad PIPE label", __func__ ))
|
|
|
|
#define ASSERT_PROC_LABEL(x) KASSERT(SLOT(x) == PROCMAGIC || \
|
|
|
|
SLOT(x) == 0, ("%s: Bad PROC label", __func__ ))
|
|
|
|
#define ASSERT_CRED_LABEL(x) KASSERT(SLOT(x) == CREDMAGIC || \
|
|
|
|
SLOT(x) == 0, ("%s: Bad CRED label", __func__ ))
|
|
|
|
#define ASSERT_VNODE_LABEL(x) KASSERT(SLOT(x) == VNODEMAGIC || \
|
|
|
|
SLOT(x) == 0, ("%s: Bad VNODE label", __func__ ))
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
static int test_slot;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, slot, CTLFLAG_RD,
|
|
|
|
&test_slot, 0, "Slot allocated by framework");
|
|
|
|
|
|
|
|
static int init_count_bpfdesc;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, init_count_bpfdesc, CTLFLAG_RD,
|
|
|
|
&init_count_bpfdesc, 0, "bpfdesc init calls");
|
|
|
|
static int init_count_cred;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, init_count_cred, CTLFLAG_RD,
|
|
|
|
&init_count_cred, 0, "cred init calls");
|
|
|
|
static int init_count_devfsdirent;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, init_count_devfsdirent, CTLFLAG_RD,
|
|
|
|
&init_count_devfsdirent, 0, "devfsdirent init calls");
|
|
|
|
static int init_count_ifnet;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, init_count_ifnet, CTLFLAG_RD,
|
|
|
|
&init_count_ifnet, 0, "ifnet init calls");
|
|
|
|
static int init_count_ipq;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, init_count_ipq, CTLFLAG_RD,
|
|
|
|
&init_count_ipq, 0, "ipq init calls");
|
|
|
|
static int init_count_mbuf;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, init_count_mbuf, CTLFLAG_RD,
|
|
|
|
&init_count_mbuf, 0, "mbuf init calls");
|
|
|
|
static int init_count_mount;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, init_count_mount, CTLFLAG_RD,
|
|
|
|
&init_count_mount, 0, "mount init calls");
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
static int init_count_mount_fslabel;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, init_count_mount_fslabel, CTLFLAG_RD,
|
|
|
|
&init_count_mount_fslabel, 0, "mount_fslabel init calls");
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
static int init_count_socket;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, init_count_socket, CTLFLAG_RD,
|
|
|
|
&init_count_socket, 0, "socket init calls");
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
static int init_count_socket_peerlabel;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, init_count_socket_peerlabel,
|
|
|
|
CTLFLAG_RD, &init_count_socket_peerlabel, 0,
|
|
|
|
"socket_peerlabel init calls");
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
static int init_count_pipe;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, init_count_pipe, CTLFLAG_RD,
|
|
|
|
&init_count_pipe, 0, "pipe init calls");
|
2003-03-25 01:20:56 +00:00
|
|
|
static int init_count_proc;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, init_count_proc, CTLFLAG_RD,
|
|
|
|
&init_count_proc, 0, "proc init calls");
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
static int init_count_vnode;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, init_count_vnode, CTLFLAG_RD,
|
|
|
|
&init_count_vnode, 0, "vnode init calls");
|
|
|
|
|
|
|
|
static int destroy_count_bpfdesc;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, destroy_count_bpfdesc, CTLFLAG_RD,
|
|
|
|
&destroy_count_bpfdesc, 0, "bpfdesc destroy calls");
|
|
|
|
static int destroy_count_cred;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, destroy_count_cred, CTLFLAG_RD,
|
|
|
|
&destroy_count_cred, 0, "cred destroy calls");
|
|
|
|
static int destroy_count_devfsdirent;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, destroy_count_devfsdirent, CTLFLAG_RD,
|
|
|
|
&destroy_count_devfsdirent, 0, "devfsdirent destroy calls");
|
|
|
|
static int destroy_count_ifnet;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, destroy_count_ifnet, CTLFLAG_RD,
|
|
|
|
&destroy_count_ifnet, 0, "ifnet destroy calls");
|
|
|
|
static int destroy_count_ipq;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, destroy_count_ipq, CTLFLAG_RD,
|
|
|
|
&destroy_count_ipq, 0, "ipq destroy calls");
|
|
|
|
static int destroy_count_mbuf;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, destroy_count_mbuf, CTLFLAG_RD,
|
|
|
|
&destroy_count_mbuf, 0, "mbuf destroy calls");
|
|
|
|
static int destroy_count_mount;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, destroy_count_mount, CTLFLAG_RD,
|
|
|
|
&destroy_count_mount, 0, "mount destroy calls");
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
static int destroy_count_mount_fslabel;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, destroy_count_mount_fslabel,
|
|
|
|
CTLFLAG_RD, &destroy_count_mount_fslabel, 0,
|
|
|
|
"mount_fslabel destroy calls");
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
static int destroy_count_socket;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, destroy_count_socket, CTLFLAG_RD,
|
|
|
|
&destroy_count_socket, 0, "socket destroy calls");
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
static int destroy_count_socket_peerlabel;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, destroy_count_socket_peerlabel,
|
|
|
|
CTLFLAG_RD, &destroy_count_socket_peerlabel, 0,
|
|
|
|
"socket_peerlabel destroy calls");
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
static int destroy_count_pipe;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, destroy_count_pipe, CTLFLAG_RD,
|
|
|
|
&destroy_count_pipe, 0, "pipe destroy calls");
|
2003-03-25 01:20:56 +00:00
|
|
|
static int destroy_count_proc;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, destroy_count_proc, CTLFLAG_RD,
|
|
|
|
&destroy_count_proc, 0, "proc destroy calls");
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
static int destroy_count_vnode;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, destroy_count_vnode, CTLFLAG_RD,
|
|
|
|
&destroy_count_vnode, 0, "vnode destroy calls");
|
|
|
|
|
|
|
|
static int externalize_count;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, externalize_count, CTLFLAG_RD,
|
|
|
|
&externalize_count, 0, "Subject/object externalize calls");
|
|
|
|
static int internalize_count;
|
|
|
|
SYSCTL_INT(_security_mac_test, OID_AUTO, internalize_count, CTLFLAG_RD,
|
|
|
|
&internalize_count, 0, "Subject/object internalize calls");
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Policy module operations.
|
|
|
|
*/
|
|
|
|
static void
|
|
|
|
mac_test_destroy(struct mac_policy_conf *conf)
|
|
|
|
{
|
|
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_init(struct mac_policy_conf *conf)
|
|
|
|
{
|
|
|
|
|
|
|
|
}
|
|
|
|
|
2002-08-20 02:53:35 +00:00
|
|
|
static int
|
|
|
|
mac_test_syscall(struct thread *td, int call, void *arg)
|
|
|
|
{
|
|
|
|
|
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
/*
|
|
|
|
* Label operations.
|
|
|
|
*/
|
|
|
|
static void
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
mac_test_init_bpfdesc_label(struct label *label)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
|
|
|
SLOT(label) = BPFMAGIC;
|
|
|
|
atomic_add_int(&init_count_bpfdesc, 1);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
mac_test_init_cred_label(struct label *label)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
|
|
|
SLOT(label) = CREDMAGIC;
|
|
|
|
atomic_add_int(&init_count_cred, 1);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
mac_test_init_devfsdirent_label(struct label *label)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
|
|
|
SLOT(label) = DEVFSMAGIC;
|
|
|
|
atomic_add_int(&init_count_devfsdirent, 1);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
mac_test_init_ifnet_label(struct label *label)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
|
|
|
SLOT(label) = IFNETMAGIC;
|
|
|
|
atomic_add_int(&init_count_ifnet, 1);
|
|
|
|
}
|
|
|
|
|
2003-03-26 15:12:03 +00:00
|
|
|
static int
|
|
|
|
mac_test_init_ipq_label(struct label *label, int flag)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
2003-04-15 21:20:34 +00:00
|
|
|
if (flag & M_WAITOK)
|
|
|
|
WITNESS_WARN(WARN_GIANTOK | WARN_SLEEPOK, NULL,
|
|
|
|
"mac_test_init_ipq_label() at %s:%d", __FILE__,
|
|
|
|
__LINE__);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
SLOT(label) = IPQMAGIC;
|
|
|
|
atomic_add_int(&init_count_ipq, 1);
|
2003-03-26 15:12:03 +00:00
|
|
|
return (0);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
mac_test_init_mbuf_label(struct label *label, int flag)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
2003-04-15 21:20:34 +00:00
|
|
|
if (flag & M_WAITOK)
|
|
|
|
WITNESS_WARN(WARN_GIANTOK | WARN_SLEEPOK, NULL,
|
|
|
|
"mac_test_init_mbuf_label() at %s:%d", __FILE__,
|
|
|
|
__LINE__);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
SLOT(label) = MBUFMAGIC;
|
|
|
|
atomic_add_int(&init_count_mbuf, 1);
|
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
mac_test_init_mount_label(struct label *label)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
SLOT(label) = MOUNTMAGIC;
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
atomic_add_int(&init_count_mount, 1);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
mac_test_init_mount_fs_label(struct label *label)
|
|
|
|
{
|
|
|
|
|
|
|
|
SLOT(label) = MOUNTMAGIC;
|
|
|
|
atomic_add_int(&init_count_mount_fslabel, 1);
|
|
|
|
}
|
|
|
|
|
2002-10-05 21:23:47 +00:00
|
|
|
static int
|
|
|
|
mac_test_init_socket_label(struct label *label, int flag)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
2003-04-15 21:20:34 +00:00
|
|
|
if (flag & M_WAITOK)
|
|
|
|
WITNESS_WARN(WARN_GIANTOK | WARN_SLEEPOK, NULL,
|
|
|
|
"mac_test_init_socket_label() at %s:%d", __FILE__,
|
|
|
|
__LINE__);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
SLOT(label) = SOCKETMAGIC;
|
|
|
|
atomic_add_int(&init_count_socket, 1);
|
2002-10-05 21:23:47 +00:00
|
|
|
return (0);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
2002-10-05 21:23:47 +00:00
|
|
|
static int
|
|
|
|
mac_test_init_socket_peer_label(struct label *label, int flag)
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
{
|
|
|
|
|
2003-04-15 21:20:34 +00:00
|
|
|
if (flag & M_WAITOK)
|
|
|
|
WITNESS_WARN(WARN_GIANTOK | WARN_SLEEPOK, NULL,
|
|
|
|
"mac_test_init_socket_peer_label() at %s:%d", __FILE__,
|
|
|
|
__LINE__);
|
|
|
|
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
SLOT(label) = SOCKETMAGIC;
|
|
|
|
atomic_add_int(&init_count_socket_peerlabel, 1);
|
2002-10-05 21:23:47 +00:00
|
|
|
return (0);
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_init_pipe_label(struct label *label)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
|
|
|
SLOT(label) = PIPEMAGIC;
|
|
|
|
atomic_add_int(&init_count_pipe, 1);
|
|
|
|
}
|
|
|
|
|
2003-03-25 01:20:56 +00:00
|
|
|
static void
|
|
|
|
mac_test_init_proc_label(struct label *label)
|
|
|
|
{
|
|
|
|
|
|
|
|
SLOT(label) = PROCMAGIC;
|
|
|
|
atomic_add_int(&init_count_proc, 1);
|
|
|
|
}
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
static void
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
mac_test_init_vnode_label(struct label *label)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
|
|
|
SLOT(label) = VNODEMAGIC;
|
|
|
|
atomic_add_int(&init_count_vnode, 1);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
mac_test_destroy_bpfdesc_label(struct label *label)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
|
|
|
if (SLOT(label) == BPFMAGIC || SLOT(label) == 0) {
|
|
|
|
atomic_add_int(&destroy_count_bpfdesc, 1);
|
|
|
|
SLOT(label) = EXMAGIC;
|
|
|
|
} else if (SLOT(label) == EXMAGIC) {
|
|
|
|
Debugger("mac_test_destroy_bpfdesc: dup destroy");
|
|
|
|
} else {
|
|
|
|
Debugger("mac_test_destroy_bpfdesc: corrupted label");
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
mac_test_destroy_cred_label(struct label *label)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
|
|
|
if (SLOT(label) == CREDMAGIC || SLOT(label) == 0) {
|
|
|
|
atomic_add_int(&destroy_count_cred, 1);
|
|
|
|
SLOT(label) = EXMAGIC;
|
|
|
|
} else if (SLOT(label) == EXMAGIC) {
|
|
|
|
Debugger("mac_test_destroy_cred: dup destroy");
|
|
|
|
} else {
|
|
|
|
Debugger("mac_test_destroy_cred: corrupted label");
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
mac_test_destroy_devfsdirent_label(struct label *label)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
|
|
|
if (SLOT(label) == DEVFSMAGIC || SLOT(label) == 0) {
|
|
|
|
atomic_add_int(&destroy_count_devfsdirent, 1);
|
|
|
|
SLOT(label) = EXMAGIC;
|
|
|
|
} else if (SLOT(label) == EXMAGIC) {
|
|
|
|
Debugger("mac_test_destroy_devfsdirent: dup destroy");
|
|
|
|
} else {
|
|
|
|
Debugger("mac_test_destroy_devfsdirent: corrupted label");
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
mac_test_destroy_ifnet_label(struct label *label)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
|
|
|
if (SLOT(label) == IFNETMAGIC || SLOT(label) == 0) {
|
|
|
|
atomic_add_int(&destroy_count_ifnet, 1);
|
|
|
|
SLOT(label) = EXMAGIC;
|
|
|
|
} else if (SLOT(label) == EXMAGIC) {
|
|
|
|
Debugger("mac_test_destroy_ifnet: dup destroy");
|
|
|
|
} else {
|
|
|
|
Debugger("mac_test_destroy_ifnet: corrupted label");
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
mac_test_destroy_ipq_label(struct label *label)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
|
|
|
if (SLOT(label) == IPQMAGIC || SLOT(label) == 0) {
|
|
|
|
atomic_add_int(&destroy_count_ipq, 1);
|
|
|
|
SLOT(label) = EXMAGIC;
|
|
|
|
} else if (SLOT(label) == EXMAGIC) {
|
|
|
|
Debugger("mac_test_destroy_ipq: dup destroy");
|
|
|
|
} else {
|
|
|
|
Debugger("mac_test_destroy_ipq: corrupted label");
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
mac_test_destroy_mbuf_label(struct label *label)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
2003-04-15 21:20:34 +00:00
|
|
|
/*
|
|
|
|
* If we're loaded dynamically, there may be mbufs in flight that
|
|
|
|
* didn't have label storage allocated for them. Handle this
|
|
|
|
* gracefully.
|
|
|
|
*/
|
|
|
|
if (label == NULL)
|
|
|
|
return;
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
if (SLOT(label) == MBUFMAGIC || SLOT(label) == 0) {
|
|
|
|
atomic_add_int(&destroy_count_mbuf, 1);
|
|
|
|
SLOT(label) = EXMAGIC;
|
|
|
|
} else if (SLOT(label) == EXMAGIC) {
|
|
|
|
Debugger("mac_test_destroy_mbuf: dup destroy");
|
|
|
|
} else {
|
|
|
|
Debugger("mac_test_destroy_mbuf: corrupted label");
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
mac_test_destroy_mount_label(struct label *label)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
if ((SLOT(label) == MOUNTMAGIC || SLOT(label) == 0)) {
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
atomic_add_int(&destroy_count_mount, 1);
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
SLOT(label) = EXMAGIC;
|
|
|
|
} else if (SLOT(label) == EXMAGIC) {
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
Debugger("mac_test_destroy_mount: dup destroy");
|
|
|
|
} else {
|
|
|
|
Debugger("mac_test_destroy_mount: corrupted label");
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
mac_test_destroy_mount_fs_label(struct label *label)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
if ((SLOT(label) == MOUNTMAGIC || SLOT(label) == 0)) {
|
|
|
|
atomic_add_int(&destroy_count_mount_fslabel, 1);
|
|
|
|
SLOT(label) = EXMAGIC;
|
|
|
|
} else if (SLOT(label) == EXMAGIC) {
|
|
|
|
Debugger("mac_test_destroy_mount_fslabel: dup destroy");
|
|
|
|
} else {
|
|
|
|
Debugger("mac_test_destroy_mount_fslabel: corrupted label");
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_destroy_socket_label(struct label *label)
|
|
|
|
{
|
|
|
|
|
|
|
|
if ((SLOT(label) == SOCKETMAGIC || SLOT(label) == 0)) {
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
atomic_add_int(&destroy_count_socket, 1);
|
|
|
|
SLOT(label) = EXMAGIC;
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
} else if (SLOT(label) == EXMAGIC) {
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
Debugger("mac_test_destroy_socket: dup destroy");
|
|
|
|
} else {
|
|
|
|
Debugger("mac_test_destroy_socket: corrupted label");
|
|
|
|
}
|
|
|
|
}
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_destroy_socket_peer_label(struct label *label)
|
|
|
|
{
|
|
|
|
|
|
|
|
if ((SLOT(label) == SOCKETMAGIC || SLOT(label) == 0)) {
|
|
|
|
atomic_add_int(&destroy_count_socket_peerlabel, 1);
|
|
|
|
SLOT(label) = EXMAGIC;
|
|
|
|
} else if (SLOT(label) == EXMAGIC) {
|
|
|
|
Debugger("mac_test_destroy_socket_peerlabel: dup destroy");
|
|
|
|
} else {
|
|
|
|
Debugger("mac_test_destroy_socket_peerlabel: corrupted label");
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
static void
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
mac_test_destroy_pipe_label(struct label *label)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
|
|
|
if ((SLOT(label) == PIPEMAGIC || SLOT(label) == 0)) {
|
|
|
|
atomic_add_int(&destroy_count_pipe, 1);
|
|
|
|
SLOT(label) = EXMAGIC;
|
|
|
|
} else if (SLOT(label) == EXMAGIC) {
|
|
|
|
Debugger("mac_test_destroy_pipe: dup destroy");
|
|
|
|
} else {
|
|
|
|
Debugger("mac_test_destroy_pipe: corrupted label");
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2003-03-25 01:20:56 +00:00
|
|
|
static void
|
|
|
|
mac_test_destroy_proc_label(struct label *label)
|
|
|
|
{
|
|
|
|
|
|
|
|
if ((SLOT(label) == PROCMAGIC || SLOT(label) == 0)) {
|
|
|
|
atomic_add_int(&destroy_count_proc, 1);
|
|
|
|
SLOT(label) = EXMAGIC;
|
|
|
|
} else if (SLOT(label) == EXMAGIC) {
|
|
|
|
Debugger("mac_test_destroy_proc: dup destroy");
|
|
|
|
} else {
|
|
|
|
Debugger("mac_test_destroy_proc: corrupted label");
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
static void
|
Begin another merge from the TrustedBSD MAC branch:
- Change mpo_init_foo(obj, label) and mpo_destroy_foo(obj, label) policy
entry points to mpo_init_foo_label(label) and
mpo_destroy_foo_label(label). This will permit the use of the same
entry points for holding temporary type-specific label during
internalization and externalization, as well as for caching purposes.
- Because of this, break out mpo_{init,destroy}_socket() and
mpo_{init,destroy}_mount() into seperate entry points for socket
main/peer labels and mount main/fs labels.
- Since the prototype for label initialization is the same across almost
all entry points, implement these entry points using common
implementations for Biba, MLS, and Test, reducing the number of
almost identical looking functions.
This simplifies policy implementation, as well as preparing us for the
merge of the new flexible userland API for managing labels on objects.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 15:10:00 +00:00
|
|
|
mac_test_destroy_vnode_label(struct label *label)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
|
|
|
if (SLOT(label) == VNODEMAGIC || SLOT(label) == 0) {
|
|
|
|
atomic_add_int(&destroy_count_vnode, 1);
|
|
|
|
SLOT(label) = EXMAGIC;
|
|
|
|
} else if (SLOT(label) == EXMAGIC) {
|
|
|
|
Debugger("mac_test_destroy_vnode: dup destroy");
|
|
|
|
} else {
|
|
|
|
Debugger("mac_test_destroy_vnode: corrupted label");
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2003-11-16 18:28:58 +00:00
|
|
|
static void
|
|
|
|
mac_test_copy_mbuf_label(struct label *src, struct label *dest)
|
|
|
|
{
|
|
|
|
|
|
|
|
ASSERT_MBUF_LABEL(src);
|
|
|
|
ASSERT_MBUF_LABEL(dest);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_copy_pipe_label(struct label *src, struct label *dest)
|
|
|
|
{
|
|
|
|
|
|
|
|
ASSERT_PIPE_LABEL(src);
|
|
|
|
ASSERT_PIPE_LABEL(dest);
|
|
|
|
}
|
|
|
|
|
Implement sockets support for __mac_get_fd() and __mac_set_fd()
system calls, and prefer these calls over getsockopt()/setsockopt()
for ABI reasons. When addressing UNIX domain sockets, these calls
retrieve and modify the socket label, not the label of the
rendezvous vnode.
- Create mac_copy_socket_label() entry point based on
mac_copy_pipe_label() entry point, intended to copy the socket
label into temporary storage that doesn't require a socket lock
to be held (currently Giant).
- Implement mac_copy_socket_label() for various policies.
- Expose socket label allocation, free, internalize, externalize
entry points as non-static from mac_net.c.
- Use mac_socket_label_set() in __mac_set_fd().
MAC-aware applications may now use mac_get_fd(), mac_set_fd(), and
mac_get_peer() to retrieve and set various socket labels without
directly invoking the getsockopt() interface.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-16 23:31:45 +00:00
|
|
|
static void
|
|
|
|
mac_test_copy_socket_label(struct label *src, struct label *dest)
|
|
|
|
{
|
|
|
|
|
|
|
|
ASSERT_SOCKET_LABEL(src);
|
|
|
|
ASSERT_SOCKET_LABEL(dest);
|
|
|
|
}
|
|
|
|
|
2003-11-16 18:28:58 +00:00
|
|
|
static void
|
|
|
|
mac_test_copy_vnode_label(struct label *src, struct label *dest)
|
|
|
|
{
|
|
|
|
|
|
|
|
ASSERT_VNODE_LABEL(src);
|
|
|
|
ASSERT_VNODE_LABEL(dest);
|
|
|
|
}
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
static int
|
2002-10-22 14:31:34 +00:00
|
|
|
mac_test_externalize_label(struct label *label, char *element_name,
|
2003-06-23 01:26:34 +00:00
|
|
|
struct sbuf *sb, int *claimed)
|
2002-10-22 14:31:34 +00:00
|
|
|
{
|
|
|
|
|
|
|
|
atomic_add_int(&externalize_count, 1);
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
KASSERT(SLOT(label) != EXMAGIC,
|
|
|
|
("mac_test_externalize_label: destroyed label"));
|
|
|
|
|
2002-10-22 14:31:34 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
static int
|
2002-10-30 18:10:46 +00:00
|
|
|
mac_test_internalize_label(struct label *label, char *element_name,
|
|
|
|
char *element_data, int *claimed)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
|
|
|
atomic_add_int(&internalize_count, 1);
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
KASSERT(SLOT(label) != EXMAGIC,
|
|
|
|
("mac_test_internalize_label: destroyed label"));
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Labeling event operations: file system objects, and things that look
|
|
|
|
* a lot like file system objects.
|
|
|
|
*/
|
Slightly change the semantics of vnode labels for MAC: rather than
"refreshing" the label on the vnode before use, just get the label
right from inception. For single-label file systems, set the label
in the generic VFS getnewvnode() code; for multi-label file systems,
leave the labeling up to the file system. With UFS1/2, this means
reading the extended attribute during vfs_vget() as the inode is
pulled off disk, rather than hitting the extended attributes
frequently during operations later, improving performance. This
also corrects sematics for shared vnode locks, which were not
previously present in the system. This chances the cache
coherrency properties WRT out-of-band access to label data, but in
an acceptable form. With UFS1, there is a small race condition
during automatic extended attribute start -- this is not present
with UFS2, and occurs because EAs aren't available at vnode
inception. We'll introduce a work around for this shortly.
Approved by: re
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-26 14:38:24 +00:00
|
|
|
static void
|
|
|
|
mac_test_associate_vnode_devfs(struct mount *mp, struct label *fslabel,
|
|
|
|
struct devfs_dirent *de, struct label *delabel, struct vnode *vp,
|
|
|
|
struct label *vlabel)
|
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_MOUNT_LABEL(fslabel);
|
|
|
|
ASSERT_DEVFS_LABEL(delabel);
|
|
|
|
ASSERT_VNODE_LABEL(vlabel);
|
Slightly change the semantics of vnode labels for MAC: rather than
"refreshing" the label on the vnode before use, just get the label
right from inception. For single-label file systems, set the label
in the generic VFS getnewvnode() code; for multi-label file systems,
leave the labeling up to the file system. With UFS1/2, this means
reading the extended attribute during vfs_vget() as the inode is
pulled off disk, rather than hitting the extended attributes
frequently during operations later, improving performance. This
also corrects sematics for shared vnode locks, which were not
previously present in the system. This chances the cache
coherrency properties WRT out-of-band access to label data, but in
an acceptable form. With UFS1, there is a small race condition
during automatic extended attribute start -- this is not present
with UFS2, and occurs because EAs aren't available at vnode
inception. We'll introduce a work around for this shortly.
Approved by: re
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-26 14:38:24 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_associate_vnode_extattr(struct mount *mp, struct label *fslabel,
|
|
|
|
struct vnode *vp, struct label *vlabel)
|
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_MOUNT_LABEL(fslabel);
|
|
|
|
ASSERT_VNODE_LABEL(vlabel);
|
Slightly change the semantics of vnode labels for MAC: rather than
"refreshing" the label on the vnode before use, just get the label
right from inception. For single-label file systems, set the label
in the generic VFS getnewvnode() code; for multi-label file systems,
leave the labeling up to the file system. With UFS1/2, this means
reading the extended attribute during vfs_vget() as the inode is
pulled off disk, rather than hitting the extended attributes
frequently during operations later, improving performance. This
also corrects sematics for shared vnode locks, which were not
previously present in the system. This chances the cache
coherrency properties WRT out-of-band access to label data, but in
an acceptable form. With UFS1, there is a small race condition
during automatic extended attribute start -- this is not present
with UFS2, and occurs because EAs aren't available at vnode
inception. We'll introduce a work around for this shortly.
Approved by: re
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-26 14:38:24 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_associate_vnode_singlelabel(struct mount *mp,
|
|
|
|
struct label *fslabel, struct vnode *vp, struct label *vlabel)
|
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_MOUNT_LABEL(fslabel);
|
|
|
|
ASSERT_VNODE_LABEL(vlabel);
|
Slightly change the semantics of vnode labels for MAC: rather than
"refreshing" the label on the vnode before use, just get the label
right from inception. For single-label file systems, set the label
in the generic VFS getnewvnode() code; for multi-label file systems,
leave the labeling up to the file system. With UFS1/2, this means
reading the extended attribute during vfs_vget() as the inode is
pulled off disk, rather than hitting the extended attributes
frequently during operations later, improving performance. This
also corrects sematics for shared vnode locks, which were not
previously present in the system. This chances the cache
coherrency properties WRT out-of-band access to label data, but in
an acceptable form. With UFS1, there is a small race condition
during automatic extended attribute start -- this is not present
with UFS2, and occurs because EAs aren't available at vnode
inception. We'll introduce a work around for this shortly.
Approved by: re
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-26 14:38:24 +00:00
|
|
|
}
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
static void
|
2002-12-09 03:44:28 +00:00
|
|
|
mac_test_create_devfs_device(struct mount *mp, dev_t dev,
|
|
|
|
struct devfs_dirent *devfs_dirent, struct label *label)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_DEVFS_LABEL(label);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
2002-12-09 03:44:28 +00:00
|
|
|
mac_test_create_devfs_directory(struct mount *mp, char *dirname,
|
|
|
|
int dirnamelen, struct devfs_dirent *devfs_dirent, struct label *label)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_DEVFS_LABEL(label);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
2002-10-05 18:56:25 +00:00
|
|
|
static void
|
2002-12-09 03:44:28 +00:00
|
|
|
mac_test_create_devfs_symlink(struct ucred *cred, struct mount *mp,
|
|
|
|
struct devfs_dirent *dd, struct label *ddlabel, struct devfs_dirent *de,
|
|
|
|
struct label *delabel)
|
2002-10-05 18:56:25 +00:00
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_DEVFS_LABEL(ddlabel);
|
|
|
|
ASSERT_DEVFS_LABEL(delabel);
|
2002-10-05 18:56:25 +00:00
|
|
|
}
|
|
|
|
|
Slightly change the semantics of vnode labels for MAC: rather than
"refreshing" the label on the vnode before use, just get the label
right from inception. For single-label file systems, set the label
in the generic VFS getnewvnode() code; for multi-label file systems,
leave the labeling up to the file system. With UFS1/2, this means
reading the extended attribute during vfs_vget() as the inode is
pulled off disk, rather than hitting the extended attributes
frequently during operations later, improving performance. This
also corrects sematics for shared vnode locks, which were not
previously present in the system. This chances the cache
coherrency properties WRT out-of-band access to label data, but in
an acceptable form. With UFS1, there is a small race condition
during automatic extended attribute start -- this is not present
with UFS2, and occurs because EAs aren't available at vnode
inception. We'll introduce a work around for this shortly.
Approved by: re
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-26 14:38:24 +00:00
|
|
|
static int
|
|
|
|
mac_test_create_vnode_extattr(struct ucred *cred, struct mount *mp,
|
|
|
|
struct label *fslabel, struct vnode *dvp, struct label *dlabel,
|
|
|
|
struct vnode *vp, struct label *vlabel, struct componentname *cnp)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_MOUNT_LABEL(fslabel);
|
|
|
|
ASSERT_VNODE_LABEL(dlabel);
|
|
|
|
|
Slightly change the semantics of vnode labels for MAC: rather than
"refreshing" the label on the vnode before use, just get the label
right from inception. For single-label file systems, set the label
in the generic VFS getnewvnode() code; for multi-label file systems,
leave the labeling up to the file system. With UFS1/2, this means
reading the extended attribute during vfs_vget() as the inode is
pulled off disk, rather than hitting the extended attributes
frequently during operations later, improving performance. This
also corrects sematics for shared vnode locks, which were not
previously present in the system. This chances the cache
coherrency properties WRT out-of-band access to label data, but in
an acceptable form. With UFS1, there is a small race condition
during automatic extended attribute start -- this is not present
with UFS2, and occurs because EAs aren't available at vnode
inception. We'll introduce a work around for this shortly.
Approved by: re
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-26 14:38:24 +00:00
|
|
|
return (0);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_create_mount(struct ucred *cred, struct mount *mp,
|
|
|
|
struct label *mntlabel, struct label *fslabel)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_MOUNT_LABEL(mntlabel);
|
|
|
|
ASSERT_MOUNT_LABEL(fslabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_create_root_mount(struct ucred *cred, struct mount *mp,
|
|
|
|
struct label *mntlabel, struct label *fslabel)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_MOUNT_LABEL(mntlabel);
|
|
|
|
ASSERT_MOUNT_LABEL(fslabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_relabel_vnode(struct ucred *cred, struct vnode *vp,
|
|
|
|
struct label *vnodelabel, struct label *label)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(vnodelabel);
|
|
|
|
ASSERT_VNODE_LABEL(label);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
Slightly change the semantics of vnode labels for MAC: rather than
"refreshing" the label on the vnode before use, just get the label
right from inception. For single-label file systems, set the label
in the generic VFS getnewvnode() code; for multi-label file systems,
leave the labeling up to the file system. With UFS1/2, this means
reading the extended attribute during vfs_vget() as the inode is
pulled off disk, rather than hitting the extended attributes
frequently during operations later, improving performance. This
also corrects sematics for shared vnode locks, which were not
previously present in the system. This chances the cache
coherrency properties WRT out-of-band access to label data, but in
an acceptable form. With UFS1, there is a small race condition
during automatic extended attribute start -- this is not present
with UFS2, and occurs because EAs aren't available at vnode
inception. We'll introduce a work around for this shortly.
Approved by: re
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-26 14:38:24 +00:00
|
|
|
mac_test_setlabel_vnode_extattr(struct ucred *cred, struct vnode *vp,
|
|
|
|
struct label *vlabel, struct label *intlabel)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(vlabel);
|
|
|
|
ASSERT_VNODE_LABEL(intlabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
2002-12-09 03:44:28 +00:00
|
|
|
mac_test_update_devfsdirent(struct mount *mp,
|
|
|
|
struct devfs_dirent *devfs_dirent, struct label *direntlabel,
|
|
|
|
struct vnode *vp, struct label *vnodelabel)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_DEVFS_LABEL(direntlabel);
|
|
|
|
ASSERT_VNODE_LABEL(vnodelabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Labeling event operations: IPC object.
|
|
|
|
*/
|
|
|
|
static void
|
|
|
|
mac_test_create_mbuf_from_socket(struct socket *so, struct label *socketlabel,
|
|
|
|
struct mbuf *m, struct label *mbuflabel)
|
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_SOCKET_LABEL(socketlabel);
|
|
|
|
ASSERT_MBUF_LABEL(mbuflabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_create_socket(struct ucred *cred, struct socket *socket,
|
|
|
|
struct label *socketlabel)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_SOCKET_LABEL(socketlabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_create_pipe(struct ucred *cred, struct pipe *pipe,
|
|
|
|
struct label *pipelabel)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_PIPE_LABEL(pipelabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_create_socket_from_socket(struct socket *oldsocket,
|
|
|
|
struct label *oldsocketlabel, struct socket *newsocket,
|
|
|
|
struct label *newsocketlabel)
|
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_SOCKET_LABEL(oldsocketlabel);
|
|
|
|
ASSERT_SOCKET_LABEL(newsocketlabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_relabel_socket(struct ucred *cred, struct socket *socket,
|
|
|
|
struct label *socketlabel, struct label *newlabel)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_SOCKET_LABEL(newlabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_relabel_pipe(struct ucred *cred, struct pipe *pipe,
|
|
|
|
struct label *pipelabel, struct label *newlabel)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_PIPE_LABEL(pipelabel);
|
|
|
|
ASSERT_PIPE_LABEL(newlabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_set_socket_peer_from_mbuf(struct mbuf *mbuf, struct label *mbuflabel,
|
|
|
|
struct socket *socket, struct label *socketpeerlabel)
|
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_MBUF_LABEL(mbuflabel);
|
|
|
|
ASSERT_SOCKET_LABEL(socketpeerlabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Labeling event operations: network objects.
|
|
|
|
*/
|
|
|
|
static void
|
|
|
|
mac_test_set_socket_peer_from_socket(struct socket *oldsocket,
|
|
|
|
struct label *oldsocketlabel, struct socket *newsocket,
|
|
|
|
struct label *newsocketpeerlabel)
|
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_SOCKET_LABEL(oldsocketlabel);
|
|
|
|
ASSERT_SOCKET_LABEL(newsocketpeerlabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_create_bpfdesc(struct ucred *cred, struct bpf_d *bpf_d,
|
|
|
|
struct label *bpflabel)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_BPF_LABEL(bpflabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_create_datagram_from_ipq(struct ipq *ipq, struct label *ipqlabel,
|
|
|
|
struct mbuf *datagram, struct label *datagramlabel)
|
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_IPQ_LABEL(ipqlabel);
|
|
|
|
ASSERT_MBUF_LABEL(datagramlabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_create_fragment(struct mbuf *datagram, struct label *datagramlabel,
|
|
|
|
struct mbuf *fragment, struct label *fragmentlabel)
|
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_MBUF_LABEL(datagramlabel);
|
|
|
|
ASSERT_MBUF_LABEL(fragmentlabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_create_ifnet(struct ifnet *ifnet, struct label *ifnetlabel)
|
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_IFNET_LABEL(ifnetlabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_create_ipq(struct mbuf *fragment, struct label *fragmentlabel,
|
|
|
|
struct ipq *ipq, struct label *ipqlabel)
|
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_MBUF_LABEL(fragmentlabel);
|
|
|
|
ASSERT_IPQ_LABEL(ipqlabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_create_mbuf_from_mbuf(struct mbuf *oldmbuf,
|
|
|
|
struct label *oldmbuflabel, struct mbuf *newmbuf,
|
|
|
|
struct label *newmbuflabel)
|
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_MBUF_LABEL(oldmbuflabel);
|
|
|
|
ASSERT_MBUF_LABEL(newmbuflabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_create_mbuf_linklayer(struct ifnet *ifnet, struct label *ifnetlabel,
|
|
|
|
struct mbuf *mbuf, struct label *mbuflabel)
|
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_IFNET_LABEL(ifnetlabel);
|
|
|
|
ASSERT_MBUF_LABEL(mbuflabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_create_mbuf_from_bpfdesc(struct bpf_d *bpf_d, struct label *bpflabel,
|
|
|
|
struct mbuf *mbuf, struct label *mbuflabel)
|
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_BPF_LABEL(bpflabel);
|
|
|
|
ASSERT_MBUF_LABEL(mbuflabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_create_mbuf_from_ifnet(struct ifnet *ifnet, struct label *ifnetlabel,
|
|
|
|
struct mbuf *m, struct label *mbuflabel)
|
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_IFNET_LABEL(ifnetlabel);
|
|
|
|
ASSERT_MBUF_LABEL(mbuflabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_create_mbuf_multicast_encap(struct mbuf *oldmbuf,
|
|
|
|
struct label *oldmbuflabel, struct ifnet *ifnet, struct label *ifnetlabel,
|
|
|
|
struct mbuf *newmbuf, struct label *newmbuflabel)
|
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_MBUF_LABEL(oldmbuflabel);
|
|
|
|
ASSERT_IFNET_LABEL(ifnetlabel);
|
|
|
|
ASSERT_MBUF_LABEL(newmbuflabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_create_mbuf_netlayer(struct mbuf *oldmbuf,
|
|
|
|
struct label *oldmbuflabel, struct mbuf *newmbuf,
|
|
|
|
struct label *newmbuflabel)
|
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_MBUF_LABEL(oldmbuflabel);
|
|
|
|
ASSERT_MBUF_LABEL(newmbuflabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_fragment_match(struct mbuf *fragment, struct label *fragmentlabel,
|
|
|
|
struct ipq *ipq, struct label *ipqlabel)
|
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_MBUF_LABEL(fragmentlabel);
|
|
|
|
ASSERT_IPQ_LABEL(ipqlabel);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (1);
|
|
|
|
}
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
static void
|
|
|
|
mac_test_reflect_mbuf_icmp(struct mbuf *m, struct label *mlabel)
|
|
|
|
{
|
|
|
|
|
|
|
|
ASSERT_MBUF_LABEL(mlabel);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_reflect_mbuf_tcp(struct mbuf *m, struct label *mlabel)
|
|
|
|
{
|
|
|
|
|
|
|
|
ASSERT_MBUF_LABEL(mlabel);
|
|
|
|
}
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
static void
|
|
|
|
mac_test_relabel_ifnet(struct ucred *cred, struct ifnet *ifnet,
|
|
|
|
struct label *ifnetlabel, struct label *newlabel)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_IFNET_LABEL(ifnetlabel);
|
|
|
|
ASSERT_IFNET_LABEL(newlabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_update_ipq(struct mbuf *fragment, struct label *fragmentlabel,
|
|
|
|
struct ipq *ipq, struct label *ipqlabel)
|
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_MBUF_LABEL(fragmentlabel);
|
|
|
|
ASSERT_IPQ_LABEL(ipqlabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Labeling event operations: processes.
|
|
|
|
*/
|
|
|
|
static void
|
|
|
|
mac_test_create_cred(struct ucred *cred_parent, struct ucred *cred_child)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred_parent->cr_label);
|
|
|
|
ASSERT_CRED_LABEL(cred_child->cr_label);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_execve_transition(struct ucred *old, struct ucred *new,
|
2002-11-05 17:52:42 +00:00
|
|
|
struct vnode *vp, struct label *filelabel,
|
2002-11-08 18:04:36 +00:00
|
|
|
struct label *interpvnodelabel, struct image_params *imgp,
|
|
|
|
struct label *execlabel)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(old->cr_label);
|
|
|
|
ASSERT_CRED_LABEL(new->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(filelabel);
|
|
|
|
ASSERT_VNODE_LABEL(interpvnodelabel);
|
|
|
|
if (execlabel != NULL) {
|
|
|
|
ASSERT_CRED_LABEL(execlabel);
|
|
|
|
}
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_execve_will_transition(struct ucred *old, struct vnode *vp,
|
2002-11-05 17:52:42 +00:00
|
|
|
struct label *filelabel, struct label *interpvnodelabel,
|
2002-11-08 18:04:36 +00:00
|
|
|
struct image_params *imgp, struct label *execlabel)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(old->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(filelabel);
|
|
|
|
if (interpvnodelabel != NULL) {
|
|
|
|
ASSERT_VNODE_LABEL(interpvnodelabel);
|
|
|
|
}
|
|
|
|
if (execlabel != NULL) {
|
|
|
|
ASSERT_CRED_LABEL(execlabel);
|
|
|
|
}
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_create_proc0(struct ucred *cred)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_create_proc1(struct ucred *cred)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
mac_test_relabel_cred(struct ucred *cred, struct label *newlabel)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-11-15 00:26:02 +00:00
|
|
|
ASSERT_CRED_LABEL(newlabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
}
|
|
|
|
|
2003-03-25 01:20:56 +00:00
|
|
|
static void
|
|
|
|
mac_test_thread_userret(struct thread *td)
|
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
printf("mac_test_thread_userret(process = %d)\n",
|
|
|
|
curthread->td_proc->p_pid);
|
2003-03-25 01:20:56 +00:00
|
|
|
}
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
/*
|
|
|
|
* Access control checks.
|
|
|
|
*/
|
|
|
|
static int
|
|
|
|
mac_test_check_bpfdesc_receive(struct bpf_d *bpf_d, struct label *bpflabel,
|
|
|
|
struct ifnet *ifnet, struct label *ifnetlabel)
|
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_BPF_LABEL(bpflabel);
|
|
|
|
ASSERT_IFNET_LABEL(ifnetlabel);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_cred_relabel(struct ucred *cred, struct label *newlabel)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_CRED_LABEL(newlabel);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_cred_visible(struct ucred *u1, struct ucred *u2)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(u1->cr_label);
|
|
|
|
ASSERT_CRED_LABEL(u2->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_ifnet_relabel(struct ucred *cred, struct ifnet *ifnet,
|
|
|
|
struct label *ifnetlabel, struct label *newlabel)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_IFNET_LABEL(ifnetlabel);
|
|
|
|
ASSERT_IFNET_LABEL(newlabel);
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_ifnet_transmit(struct ifnet *ifnet, struct label *ifnetlabel,
|
|
|
|
struct mbuf *m, struct label *mbuflabel)
|
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_IFNET_LABEL(ifnetlabel);
|
|
|
|
ASSERT_MBUF_LABEL(mbuflabel);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
2003-03-25 01:20:56 +00:00
|
|
|
static int
|
|
|
|
mac_test_check_kenv_dump(struct ucred *cred)
|
|
|
|
{
|
|
|
|
|
2003-11-12 17:21:57 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
|
2003-03-25 01:20:56 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_kenv_get(struct ucred *cred, char *name)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
|
2003-03-25 01:20:56 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_kenv_set(struct ucred *cred, char *name, char *value)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
|
2003-03-25 01:20:56 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_kenv_unset(struct ucred *cred, char *name)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
|
2003-03-25 01:20:56 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_kld_load(struct ucred *cred, struct vnode *vp,
|
|
|
|
struct label *label)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
2003-03-25 01:20:56 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_kld_stat(struct ucred *cred)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
|
2003-03-25 01:20:56 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_kld_unload(struct ucred *cred)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
|
2003-03-25 01:20:56 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
static int
|
|
|
|
mac_test_check_mount_stat(struct ucred *cred, struct mount *mp,
|
|
|
|
struct label *mntlabel)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_MOUNT_LABEL(mntlabel);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_pipe_ioctl(struct ucred *cred, struct pipe *pipe,
|
|
|
|
struct label *pipelabel, unsigned long cmd, void /* caddr_t */ *data)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_PIPE_LABEL(pipelabel);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
2002-08-19 16:59:37 +00:00
|
|
|
mac_test_check_pipe_poll(struct ucred *cred, struct pipe *pipe,
|
|
|
|
struct label *pipelabel)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_PIPE_LABEL(pipelabel);
|
|
|
|
|
2002-08-19 16:59:37 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_pipe_read(struct ucred *cred, struct pipe *pipe,
|
|
|
|
struct label *pipelabel)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_PIPE_LABEL(pipelabel);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_pipe_relabel(struct ucred *cred, struct pipe *pipe,
|
|
|
|
struct label *pipelabel, struct label *newlabel)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_PIPE_LABEL(pipelabel);
|
|
|
|
ASSERT_PIPE_LABEL(newlabel);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
2002-08-19 16:59:37 +00:00
|
|
|
static int
|
|
|
|
mac_test_check_pipe_stat(struct ucred *cred, struct pipe *pipe,
|
|
|
|
struct label *pipelabel)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_PIPE_LABEL(pipelabel);
|
|
|
|
|
2002-08-19 16:59:37 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_pipe_write(struct ucred *cred, struct pipe *pipe,
|
|
|
|
struct label *pipelabel)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_PIPE_LABEL(pipelabel);
|
|
|
|
|
2002-08-19 16:59:37 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
static int
|
|
|
|
mac_test_check_proc_debug(struct ucred *cred, struct proc *proc)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
|
|
|
ASSERT_CRED_LABEL(proc->p_ucred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_proc_sched(struct ucred *cred, struct proc *proc)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
|
|
|
ASSERT_CRED_LABEL(proc->p_ucred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
Move to C99 sparse structure initialization for the mac_policy_ops
structure definition, rather than using an operation vector
we translate into the structure. Originally, we used a vector
for two reasons:
(1) We wanted to define the structure sparsely, which wasn't
supported by the C compiler for structures. For a policy
with five entry points, you don't want to have to stick in
a few hundred NULL function pointers.
(2) We thought it would improve ABI compatibility allowing modules
to work with kernels that had a superset of the entry points
defined in the module, even if the kernel had changed its
entry point set.
Both of these no longer apply:
(1) C99 gives us a way to sparsely define a static structure.
(2) The ABI problems existed anyway, due to enumeration numbers,
argument changes, and semantic mismatches. Since the going
rule for FreeBSD is that you really need your modules to
pretty closely match your kernel, it's not worth the
complexity.
This submit eliminates the operation vector, dynamic allocation
of the operation structure, copying of the vector to the
structure, and redoes the vectors in each policy to direct
structure definitions. One enourmous benefit of this change
is that we now get decent type checking on policy entry point
implementation arguments.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-30 18:48:51 +00:00
|
|
|
mac_test_check_proc_signal(struct ucred *cred, struct proc *proc, int signum)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
|
|
|
ASSERT_CRED_LABEL(proc->p_ucred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_socket_bind(struct ucred *cred, struct socket *socket,
|
|
|
|
struct label *socketlabel, struct sockaddr *sockaddr)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_SOCKET_LABEL(socketlabel);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_socket_connect(struct ucred *cred, struct socket *socket,
|
|
|
|
struct label *socketlabel, struct sockaddr *sockaddr)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_SOCKET_LABEL(socketlabel);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
2002-08-15 18:51:27 +00:00
|
|
|
mac_test_check_socket_deliver(struct socket *socket, struct label *socketlabel,
|
|
|
|
struct mbuf *m, struct label *mbuflabel)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_SOCKET_LABEL(socketlabel);
|
|
|
|
ASSERT_MBUF_LABEL(mbuflabel);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
2002-08-15 18:51:27 +00:00
|
|
|
mac_test_check_socket_listen(struct ucred *cred, struct socket *socket,
|
Move to C99 sparse structure initialization for the mac_policy_ops
structure definition, rather than using an operation vector
we translate into the structure. Originally, we used a vector
for two reasons:
(1) We wanted to define the structure sparsely, which wasn't
supported by the C compiler for structures. For a policy
with five entry points, you don't want to have to stick in
a few hundred NULL function pointers.
(2) We thought it would improve ABI compatibility allowing modules
to work with kernels that had a superset of the entry points
defined in the module, even if the kernel had changed its
entry point set.
Both of these no longer apply:
(1) C99 gives us a way to sparsely define a static structure.
(2) The ABI problems existed anyway, due to enumeration numbers,
argument changes, and semantic mismatches. Since the going
rule for FreeBSD is that you really need your modules to
pretty closely match your kernel, it's not worth the
complexity.
This submit eliminates the operation vector, dynamic allocation
of the operation structure, copying of the vector to the
structure, and redoes the vectors in each policy to direct
structure definitions. One enourmous benefit of this change
is that we now get decent type checking on policy entry point
implementation arguments.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-30 18:48:51 +00:00
|
|
|
struct label *socketlabel)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_SOCKET_LABEL(socketlabel);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_socket_visible(struct ucred *cred, struct socket *socket,
|
|
|
|
struct label *socketlabel)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_SOCKET_LABEL(socketlabel);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_socket_relabel(struct ucred *cred, struct socket *socket,
|
|
|
|
struct label *socketlabel, struct label *newlabel)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_SOCKET_LABEL(socketlabel);
|
|
|
|
ASSERT_SOCKET_LABEL(newlabel);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
2003-03-25 01:20:56 +00:00
|
|
|
static int
|
|
|
|
mac_test_check_sysarch_ioperm(struct ucred *cred)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
|
2003-03-25 01:20:56 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_system_acct(struct ucred *cred, struct vnode *vp,
|
|
|
|
struct label *label)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
|
2003-03-25 01:20:56 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_system_reboot(struct ucred *cred, int how)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
|
2003-03-25 01:20:56 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_system_settime(struct ucred *cred)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
|
2003-03-25 01:20:56 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_system_swapon(struct ucred *cred, struct vnode *vp,
|
|
|
|
struct label *label)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
2003-03-25 01:20:56 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_system_swapoff(struct ucred *cred, struct vnode *vp,
|
|
|
|
struct label *label)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
2003-03-25 01:20:56 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_system_sysctl(struct ucred *cred, int *name, u_int namelen,
|
|
|
|
void *old, size_t *oldlenp, int inkernel, void *new, size_t newlen)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
|
2003-03-25 01:20:56 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
static int
|
|
|
|
mac_test_check_vnode_access(struct ucred *cred, struct vnode *vp,
|
2002-10-30 17:56:57 +00:00
|
|
|
struct label *label, int acc_mode)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_vnode_chdir(struct ucred *cred, struct vnode *dvp,
|
|
|
|
struct label *dlabel)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(dlabel);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_vnode_chroot(struct ucred *cred, struct vnode *dvp,
|
|
|
|
struct label *dlabel)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(dlabel);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_vnode_create(struct ucred *cred, struct vnode *dvp,
|
|
|
|
struct label *dlabel, struct componentname *cnp, struct vattr *vap)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(dlabel);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_vnode_delete(struct ucred *cred, struct vnode *dvp,
|
|
|
|
struct label *dlabel, struct vnode *vp, struct label *label,
|
|
|
|
struct componentname *cnp)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(dlabel);
|
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_vnode_deleteacl(struct ucred *cred, struct vnode *vp,
|
|
|
|
struct label *label, acl_type_t type)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_vnode_deleteextattr(struct ucred *cred, struct vnode *vp,
|
|
|
|
struct label *label, int attrnamespace, const char *name)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_vnode_exec(struct ucred *cred, struct vnode *vp,
|
2002-11-08 18:04:36 +00:00
|
|
|
struct label *label, struct image_params *imgp,
|
|
|
|
struct label *execlabel)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
if (execlabel != NULL) {
|
|
|
|
ASSERT_CRED_LABEL(execlabel);
|
|
|
|
}
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_vnode_getacl(struct ucred *cred, struct vnode *vp,
|
|
|
|
struct label *label, acl_type_t type)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_vnode_getextattr(struct ucred *cred, struct vnode *vp,
|
|
|
|
struct label *label, int attrnamespace, const char *name, struct uio *uio)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
2002-10-05 18:25:48 +00:00
|
|
|
static int
|
|
|
|
mac_test_check_vnode_link(struct ucred *cred, struct vnode *dvp,
|
|
|
|
struct label *dlabel, struct vnode *vp, struct label *label,
|
|
|
|
struct componentname *cnp)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(dlabel);
|
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_vnode_listextattr(struct ucred *cred, struct vnode *vp,
|
|
|
|
struct label *label, int attrnamespace)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
2002-10-05 18:25:48 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
static int
|
2003-07-05 01:24:36 +00:00
|
|
|
mac_test_check_vnode_lookup(struct ucred *cred, struct vnode *dvp,
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
struct label *dlabel, struct componentname *cnp)
|
|
|
|
{
|
2002-10-06 02:46:26 +00:00
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(dlabel);
|
|
|
|
|
2002-10-06 02:46:26 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_vnode_mmap(struct ucred *cred, struct vnode *vp,
|
|
|
|
struct label *label, int prot)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
2002-10-06 02:46:26 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_vnode_mprotect(struct ucred *cred, struct vnode *vp,
|
|
|
|
struct label *label, int prot)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
2002-10-06 02:46:26 +00:00
|
|
|
return (0);
|
|
|
|
}
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_vnode_open(struct ucred *cred, struct vnode *vp,
|
2002-10-30 17:56:57 +00:00
|
|
|
struct label *filelabel, int acc_mode)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(filelabel);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
2002-08-19 16:43:25 +00:00
|
|
|
static int
|
2002-08-19 19:04:53 +00:00
|
|
|
mac_test_check_vnode_poll(struct ucred *active_cred, struct ucred *file_cred,
|
|
|
|
struct vnode *vp, struct label *label)
|
2002-08-19 16:43:25 +00:00
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(active_cred->cr_label);
|
|
|
|
ASSERT_CRED_LABEL(file_cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
2002-08-19 16:43:25 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
2002-08-19 19:04:53 +00:00
|
|
|
mac_test_check_vnode_read(struct ucred *active_cred, struct ucred *file_cred,
|
|
|
|
struct vnode *vp, struct label *label)
|
2002-08-19 16:43:25 +00:00
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(active_cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
if (file_cred != NULL) {
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(file_cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
}
|
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
2002-08-19 16:43:25 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
static int
|
|
|
|
mac_test_check_vnode_readdir(struct ucred *cred, struct vnode *dvp,
|
|
|
|
struct label *dlabel)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(dlabel);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_vnode_readlink(struct ucred *cred, struct vnode *vp,
|
|
|
|
struct label *vnodelabel)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(vnodelabel);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_vnode_relabel(struct ucred *cred, struct vnode *vp,
|
|
|
|
struct label *vnodelabel, struct label *newlabel)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(vnodelabel);
|
|
|
|
ASSERT_VNODE_LABEL(newlabel);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_vnode_rename_from(struct ucred *cred, struct vnode *dvp,
|
|
|
|
struct label *dlabel, struct vnode *vp, struct label *label,
|
|
|
|
struct componentname *cnp)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(dlabel);
|
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_vnode_rename_to(struct ucred *cred, struct vnode *dvp,
|
|
|
|
struct label *dlabel, struct vnode *vp, struct label *label, int samedir,
|
|
|
|
struct componentname *cnp)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(dlabel);
|
|
|
|
|
|
|
|
if (vp != NULL) {
|
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
}
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_vnode_revoke(struct ucred *cred, struct vnode *vp,
|
|
|
|
struct label *label)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_vnode_setacl(struct ucred *cred, struct vnode *vp,
|
|
|
|
struct label *label, acl_type_t type, struct acl *acl)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_vnode_setextattr(struct ucred *cred, struct vnode *vp,
|
|
|
|
struct label *label, int attrnamespace, const char *name, struct uio *uio)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_vnode_setflags(struct ucred *cred, struct vnode *vp,
|
|
|
|
struct label *label, u_long flags)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_vnode_setmode(struct ucred *cred, struct vnode *vp,
|
|
|
|
struct label *label, mode_t mode)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_vnode_setowner(struct ucred *cred, struct vnode *vp,
|
|
|
|
struct label *label, uid_t uid, gid_t gid)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mac_test_check_vnode_setutimes(struct ucred *cred, struct vnode *vp,
|
|
|
|
struct label *label, struct timespec atime, struct timespec mtime)
|
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
2002-08-19 19:04:53 +00:00
|
|
|
mac_test_check_vnode_stat(struct ucred *active_cred, struct ucred *file_cred,
|
|
|
|
struct vnode *vp, struct label *label)
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(active_cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
if (file_cred != NULL) {
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(file_cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
}
|
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
2002-08-19 16:43:25 +00:00
|
|
|
static int
|
2002-08-19 19:04:53 +00:00
|
|
|
mac_test_check_vnode_write(struct ucred *active_cred,
|
|
|
|
struct ucred *file_cred, struct vnode *vp, struct label *label)
|
2002-08-19 16:43:25 +00:00
|
|
|
{
|
|
|
|
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(active_cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
if (file_cred != NULL) {
|
Modify the MAC Framework so that instead of embedding a (struct label)
in various kernel objects to represent security data, we embed a
(struct label *) pointer, which now references labels allocated using
a UMA zone (mac_label.c). This allows the size and shape of struct
label to be varied without changing the size and shape of these kernel
objects, which become part of the frozen ABI with 5-STABLE. This opens
the door for boot-time selection of the number of label slots, and hence
changes to the bound on the number of simultaneous labeled policies
at boot-time instead of compile-time. This also makes it easier to
embed label references in new objects as required for locking/caching
with fine-grained network stack locking, such as inpcb structures.
This change also moves us further in the direction of hiding the
structure of kernel objects from MAC policy modules, not to mention
dramatically reducing the number of '&' symbols appearing in both the
MAC Framework and MAC policy modules, and improving readability.
While this results in minimal performance change with MAC enabled, it
will observably shrink the size of a number of critical kernel data
structures for the !MAC case, and should have a small (but measurable)
performance benefit (i.e., struct vnode, struct socket) do to memory
conservation and reduced cost of zeroing memory.
NOTE: Users of MAC must recompile their kernel and all MAC modules as a
result of this change. Because this is an API change, third party
MAC modules will also need to be updated to make less use of the '&'
symbol.
Suggestions from: bmilekic
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-12 03:14:31 +00:00
|
|
|
ASSERT_CRED_LABEL(file_cred->cr_label);
|
2003-08-21 17:28:45 +00:00
|
|
|
}
|
|
|
|
ASSERT_VNODE_LABEL(label);
|
|
|
|
|
2002-08-19 16:43:25 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
Move to C99 sparse structure initialization for the mac_policy_ops
structure definition, rather than using an operation vector
we translate into the structure. Originally, we used a vector
for two reasons:
(1) We wanted to define the structure sparsely, which wasn't
supported by the C compiler for structures. For a policy
with five entry points, you don't want to have to stick in
a few hundred NULL function pointers.
(2) We thought it would improve ABI compatibility allowing modules
to work with kernels that had a superset of the entry points
defined in the module, even if the kernel had changed its
entry point set.
Both of these no longer apply:
(1) C99 gives us a way to sparsely define a static structure.
(2) The ABI problems existed anyway, due to enumeration numbers,
argument changes, and semantic mismatches. Since the going
rule for FreeBSD is that you really need your modules to
pretty closely match your kernel, it's not worth the
complexity.
This submit eliminates the operation vector, dynamic allocation
of the operation structure, copying of the vector to the
structure, and redoes the vectors in each policy to direct
structure definitions. One enourmous benefit of this change
is that we now get decent type checking on policy entry point
implementation arguments.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-30 18:48:51 +00:00
|
|
|
static struct mac_policy_ops mac_test_ops =
|
|
|
|
{
|
|
|
|
.mpo_destroy = mac_test_destroy,
|
|
|
|
.mpo_init = mac_test_init,
|
|
|
|
.mpo_syscall = mac_test_syscall,
|
|
|
|
.mpo_init_bpfdesc_label = mac_test_init_bpfdesc_label,
|
|
|
|
.mpo_init_cred_label = mac_test_init_cred_label,
|
|
|
|
.mpo_init_devfsdirent_label = mac_test_init_devfsdirent_label,
|
|
|
|
.mpo_init_ifnet_label = mac_test_init_ifnet_label,
|
|
|
|
.mpo_init_ipq_label = mac_test_init_ipq_label,
|
|
|
|
.mpo_init_mbuf_label = mac_test_init_mbuf_label,
|
|
|
|
.mpo_init_mount_label = mac_test_init_mount_label,
|
|
|
|
.mpo_init_mount_fs_label = mac_test_init_mount_fs_label,
|
|
|
|
.mpo_init_pipe_label = mac_test_init_pipe_label,
|
2003-03-25 01:20:56 +00:00
|
|
|
.mpo_init_proc_label = mac_test_init_proc_label,
|
Move to C99 sparse structure initialization for the mac_policy_ops
structure definition, rather than using an operation vector
we translate into the structure. Originally, we used a vector
for two reasons:
(1) We wanted to define the structure sparsely, which wasn't
supported by the C compiler for structures. For a policy
with five entry points, you don't want to have to stick in
a few hundred NULL function pointers.
(2) We thought it would improve ABI compatibility allowing modules
to work with kernels that had a superset of the entry points
defined in the module, even if the kernel had changed its
entry point set.
Both of these no longer apply:
(1) C99 gives us a way to sparsely define a static structure.
(2) The ABI problems existed anyway, due to enumeration numbers,
argument changes, and semantic mismatches. Since the going
rule for FreeBSD is that you really need your modules to
pretty closely match your kernel, it's not worth the
complexity.
This submit eliminates the operation vector, dynamic allocation
of the operation structure, copying of the vector to the
structure, and redoes the vectors in each policy to direct
structure definitions. One enourmous benefit of this change
is that we now get decent type checking on policy entry point
implementation arguments.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-30 18:48:51 +00:00
|
|
|
.mpo_init_socket_label = mac_test_init_socket_label,
|
|
|
|
.mpo_init_socket_peer_label = mac_test_init_socket_peer_label,
|
|
|
|
.mpo_init_vnode_label = mac_test_init_vnode_label,
|
|
|
|
.mpo_destroy_bpfdesc_label = mac_test_destroy_bpfdesc_label,
|
|
|
|
.mpo_destroy_cred_label = mac_test_destroy_cred_label,
|
|
|
|
.mpo_destroy_devfsdirent_label = mac_test_destroy_devfsdirent_label,
|
|
|
|
.mpo_destroy_ifnet_label = mac_test_destroy_ifnet_label,
|
|
|
|
.mpo_destroy_ipq_label = mac_test_destroy_ipq_label,
|
|
|
|
.mpo_destroy_mbuf_label = mac_test_destroy_mbuf_label,
|
|
|
|
.mpo_destroy_mount_label = mac_test_destroy_mount_label,
|
|
|
|
.mpo_destroy_mount_fs_label = mac_test_destroy_mount_fs_label,
|
|
|
|
.mpo_destroy_pipe_label = mac_test_destroy_pipe_label,
|
2003-03-25 01:20:56 +00:00
|
|
|
.mpo_destroy_proc_label = mac_test_destroy_proc_label,
|
Move to C99 sparse structure initialization for the mac_policy_ops
structure definition, rather than using an operation vector
we translate into the structure. Originally, we used a vector
for two reasons:
(1) We wanted to define the structure sparsely, which wasn't
supported by the C compiler for structures. For a policy
with five entry points, you don't want to have to stick in
a few hundred NULL function pointers.
(2) We thought it would improve ABI compatibility allowing modules
to work with kernels that had a superset of the entry points
defined in the module, even if the kernel had changed its
entry point set.
Both of these no longer apply:
(1) C99 gives us a way to sparsely define a static structure.
(2) The ABI problems existed anyway, due to enumeration numbers,
argument changes, and semantic mismatches. Since the going
rule for FreeBSD is that you really need your modules to
pretty closely match your kernel, it's not worth the
complexity.
This submit eliminates the operation vector, dynamic allocation
of the operation structure, copying of the vector to the
structure, and redoes the vectors in each policy to direct
structure definitions. One enourmous benefit of this change
is that we now get decent type checking on policy entry point
implementation arguments.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-30 18:48:51 +00:00
|
|
|
.mpo_destroy_socket_label = mac_test_destroy_socket_label,
|
|
|
|
.mpo_destroy_socket_peer_label = mac_test_destroy_socket_peer_label,
|
|
|
|
.mpo_destroy_vnode_label = mac_test_destroy_vnode_label,
|
2003-11-16 18:28:58 +00:00
|
|
|
.mpo_copy_mbuf_label = mac_test_copy_mbuf_label,
|
|
|
|
.mpo_copy_pipe_label = mac_test_copy_pipe_label,
|
Implement sockets support for __mac_get_fd() and __mac_set_fd()
system calls, and prefer these calls over getsockopt()/setsockopt()
for ABI reasons. When addressing UNIX domain sockets, these calls
retrieve and modify the socket label, not the label of the
rendezvous vnode.
- Create mac_copy_socket_label() entry point based on
mac_copy_pipe_label() entry point, intended to copy the socket
label into temporary storage that doesn't require a socket lock
to be held (currently Giant).
- Implement mac_copy_socket_label() for various policies.
- Expose socket label allocation, free, internalize, externalize
entry points as non-static from mac_net.c.
- Use mac_socket_label_set() in __mac_set_fd().
MAC-aware applications may now use mac_get_fd(), mac_set_fd(), and
mac_get_peer() to retrieve and set various socket labels without
directly invoking the getsockopt() interface.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2003-11-16 23:31:45 +00:00
|
|
|
.mpo_copy_socket_label = mac_test_copy_socket_label,
|
2003-11-16 18:28:58 +00:00
|
|
|
.mpo_copy_vnode_label = mac_test_copy_vnode_label,
|
Move to C99 sparse structure initialization for the mac_policy_ops
structure definition, rather than using an operation vector
we translate into the structure. Originally, we used a vector
for two reasons:
(1) We wanted to define the structure sparsely, which wasn't
supported by the C compiler for structures. For a policy
with five entry points, you don't want to have to stick in
a few hundred NULL function pointers.
(2) We thought it would improve ABI compatibility allowing modules
to work with kernels that had a superset of the entry points
defined in the module, even if the kernel had changed its
entry point set.
Both of these no longer apply:
(1) C99 gives us a way to sparsely define a static structure.
(2) The ABI problems existed anyway, due to enumeration numbers,
argument changes, and semantic mismatches. Since the going
rule for FreeBSD is that you really need your modules to
pretty closely match your kernel, it's not worth the
complexity.
This submit eliminates the operation vector, dynamic allocation
of the operation structure, copying of the vector to the
structure, and redoes the vectors in each policy to direct
structure definitions. One enourmous benefit of this change
is that we now get decent type checking on policy entry point
implementation arguments.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-30 18:48:51 +00:00
|
|
|
.mpo_externalize_cred_label = mac_test_externalize_label,
|
|
|
|
.mpo_externalize_ifnet_label = mac_test_externalize_label,
|
|
|
|
.mpo_externalize_pipe_label = mac_test_externalize_label,
|
|
|
|
.mpo_externalize_socket_label = mac_test_externalize_label,
|
|
|
|
.mpo_externalize_socket_peer_label = mac_test_externalize_label,
|
|
|
|
.mpo_externalize_vnode_label = mac_test_externalize_label,
|
|
|
|
.mpo_internalize_cred_label = mac_test_internalize_label,
|
|
|
|
.mpo_internalize_ifnet_label = mac_test_internalize_label,
|
|
|
|
.mpo_internalize_pipe_label = mac_test_internalize_label,
|
|
|
|
.mpo_internalize_socket_label = mac_test_internalize_label,
|
|
|
|
.mpo_internalize_vnode_label = mac_test_internalize_label,
|
|
|
|
.mpo_associate_vnode_devfs = mac_test_associate_vnode_devfs,
|
|
|
|
.mpo_associate_vnode_extattr = mac_test_associate_vnode_extattr,
|
|
|
|
.mpo_associate_vnode_singlelabel = mac_test_associate_vnode_singlelabel,
|
|
|
|
.mpo_create_devfs_device = mac_test_create_devfs_device,
|
|
|
|
.mpo_create_devfs_directory = mac_test_create_devfs_directory,
|
|
|
|
.mpo_create_devfs_symlink = mac_test_create_devfs_symlink,
|
|
|
|
.mpo_create_vnode_extattr = mac_test_create_vnode_extattr,
|
|
|
|
.mpo_create_mount = mac_test_create_mount,
|
|
|
|
.mpo_create_root_mount = mac_test_create_root_mount,
|
|
|
|
.mpo_relabel_vnode = mac_test_relabel_vnode,
|
|
|
|
.mpo_setlabel_vnode_extattr = mac_test_setlabel_vnode_extattr,
|
|
|
|
.mpo_update_devfsdirent = mac_test_update_devfsdirent,
|
|
|
|
.mpo_create_mbuf_from_socket = mac_test_create_mbuf_from_socket,
|
|
|
|
.mpo_create_pipe = mac_test_create_pipe,
|
|
|
|
.mpo_create_socket = mac_test_create_socket,
|
|
|
|
.mpo_create_socket_from_socket = mac_test_create_socket_from_socket,
|
|
|
|
.mpo_relabel_pipe = mac_test_relabel_pipe,
|
|
|
|
.mpo_relabel_socket = mac_test_relabel_socket,
|
|
|
|
.mpo_set_socket_peer_from_mbuf = mac_test_set_socket_peer_from_mbuf,
|
|
|
|
.mpo_set_socket_peer_from_socket = mac_test_set_socket_peer_from_socket,
|
|
|
|
.mpo_create_bpfdesc = mac_test_create_bpfdesc,
|
|
|
|
.mpo_create_ifnet = mac_test_create_ifnet,
|
|
|
|
.mpo_create_datagram_from_ipq = mac_test_create_datagram_from_ipq,
|
|
|
|
.mpo_create_fragment = mac_test_create_fragment,
|
|
|
|
.mpo_create_ipq = mac_test_create_ipq,
|
|
|
|
.mpo_create_mbuf_from_mbuf = mac_test_create_mbuf_from_mbuf,
|
|
|
|
.mpo_create_mbuf_linklayer = mac_test_create_mbuf_linklayer,
|
|
|
|
.mpo_create_mbuf_from_bpfdesc = mac_test_create_mbuf_from_bpfdesc,
|
|
|
|
.mpo_create_mbuf_from_ifnet = mac_test_create_mbuf_from_ifnet,
|
|
|
|
.mpo_create_mbuf_multicast_encap = mac_test_create_mbuf_multicast_encap,
|
|
|
|
.mpo_create_mbuf_netlayer = mac_test_create_mbuf_netlayer,
|
|
|
|
.mpo_fragment_match = mac_test_fragment_match,
|
2003-08-21 17:28:45 +00:00
|
|
|
.mpo_reflect_mbuf_icmp = mac_test_reflect_mbuf_icmp,
|
2003-08-22 12:32:07 +00:00
|
|
|
.mpo_reflect_mbuf_tcp = mac_test_reflect_mbuf_tcp,
|
Move to C99 sparse structure initialization for the mac_policy_ops
structure definition, rather than using an operation vector
we translate into the structure. Originally, we used a vector
for two reasons:
(1) We wanted to define the structure sparsely, which wasn't
supported by the C compiler for structures. For a policy
with five entry points, you don't want to have to stick in
a few hundred NULL function pointers.
(2) We thought it would improve ABI compatibility allowing modules
to work with kernels that had a superset of the entry points
defined in the module, even if the kernel had changed its
entry point set.
Both of these no longer apply:
(1) C99 gives us a way to sparsely define a static structure.
(2) The ABI problems existed anyway, due to enumeration numbers,
argument changes, and semantic mismatches. Since the going
rule for FreeBSD is that you really need your modules to
pretty closely match your kernel, it's not worth the
complexity.
This submit eliminates the operation vector, dynamic allocation
of the operation structure, copying of the vector to the
structure, and redoes the vectors in each policy to direct
structure definitions. One enourmous benefit of this change
is that we now get decent type checking on policy entry point
implementation arguments.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-30 18:48:51 +00:00
|
|
|
.mpo_relabel_ifnet = mac_test_relabel_ifnet,
|
|
|
|
.mpo_update_ipq = mac_test_update_ipq,
|
|
|
|
.mpo_create_cred = mac_test_create_cred,
|
|
|
|
.mpo_execve_transition = mac_test_execve_transition,
|
|
|
|
.mpo_execve_will_transition = mac_test_execve_will_transition,
|
|
|
|
.mpo_create_proc0 = mac_test_create_proc0,
|
|
|
|
.mpo_create_proc1 = mac_test_create_proc1,
|
|
|
|
.mpo_relabel_cred = mac_test_relabel_cred,
|
2003-03-25 01:20:56 +00:00
|
|
|
.mpo_thread_userret = mac_test_thread_userret,
|
Move to C99 sparse structure initialization for the mac_policy_ops
structure definition, rather than using an operation vector
we translate into the structure. Originally, we used a vector
for two reasons:
(1) We wanted to define the structure sparsely, which wasn't
supported by the C compiler for structures. For a policy
with five entry points, you don't want to have to stick in
a few hundred NULL function pointers.
(2) We thought it would improve ABI compatibility allowing modules
to work with kernels that had a superset of the entry points
defined in the module, even if the kernel had changed its
entry point set.
Both of these no longer apply:
(1) C99 gives us a way to sparsely define a static structure.
(2) The ABI problems existed anyway, due to enumeration numbers,
argument changes, and semantic mismatches. Since the going
rule for FreeBSD is that you really need your modules to
pretty closely match your kernel, it's not worth the
complexity.
This submit eliminates the operation vector, dynamic allocation
of the operation structure, copying of the vector to the
structure, and redoes the vectors in each policy to direct
structure definitions. One enourmous benefit of this change
is that we now get decent type checking on policy entry point
implementation arguments.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-30 18:48:51 +00:00
|
|
|
.mpo_check_bpfdesc_receive = mac_test_check_bpfdesc_receive,
|
|
|
|
.mpo_check_cred_relabel = mac_test_check_cred_relabel,
|
|
|
|
.mpo_check_cred_visible = mac_test_check_cred_visible,
|
|
|
|
.mpo_check_ifnet_relabel = mac_test_check_ifnet_relabel,
|
|
|
|
.mpo_check_ifnet_transmit = mac_test_check_ifnet_transmit,
|
2003-03-25 01:20:56 +00:00
|
|
|
.mpo_check_kenv_dump = mac_test_check_kenv_dump,
|
|
|
|
.mpo_check_kenv_get = mac_test_check_kenv_get,
|
|
|
|
.mpo_check_kenv_set = mac_test_check_kenv_set,
|
|
|
|
.mpo_check_kenv_unset = mac_test_check_kenv_unset,
|
|
|
|
.mpo_check_kld_load = mac_test_check_kld_load,
|
|
|
|
.mpo_check_kld_stat = mac_test_check_kld_stat,
|
|
|
|
.mpo_check_kld_unload = mac_test_check_kld_unload,
|
Move to C99 sparse structure initialization for the mac_policy_ops
structure definition, rather than using an operation vector
we translate into the structure. Originally, we used a vector
for two reasons:
(1) We wanted to define the structure sparsely, which wasn't
supported by the C compiler for structures. For a policy
with five entry points, you don't want to have to stick in
a few hundred NULL function pointers.
(2) We thought it would improve ABI compatibility allowing modules
to work with kernels that had a superset of the entry points
defined in the module, even if the kernel had changed its
entry point set.
Both of these no longer apply:
(1) C99 gives us a way to sparsely define a static structure.
(2) The ABI problems existed anyway, due to enumeration numbers,
argument changes, and semantic mismatches. Since the going
rule for FreeBSD is that you really need your modules to
pretty closely match your kernel, it's not worth the
complexity.
This submit eliminates the operation vector, dynamic allocation
of the operation structure, copying of the vector to the
structure, and redoes the vectors in each policy to direct
structure definitions. One enourmous benefit of this change
is that we now get decent type checking on policy entry point
implementation arguments.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-30 18:48:51 +00:00
|
|
|
.mpo_check_mount_stat = mac_test_check_mount_stat,
|
|
|
|
.mpo_check_pipe_ioctl = mac_test_check_pipe_ioctl,
|
|
|
|
.mpo_check_pipe_poll = mac_test_check_pipe_poll,
|
|
|
|
.mpo_check_pipe_read = mac_test_check_pipe_read,
|
|
|
|
.mpo_check_pipe_relabel = mac_test_check_pipe_relabel,
|
|
|
|
.mpo_check_pipe_stat = mac_test_check_pipe_stat,
|
|
|
|
.mpo_check_pipe_write = mac_test_check_pipe_write,
|
|
|
|
.mpo_check_proc_debug = mac_test_check_proc_debug,
|
|
|
|
.mpo_check_proc_sched = mac_test_check_proc_sched,
|
|
|
|
.mpo_check_proc_signal = mac_test_check_proc_signal,
|
|
|
|
.mpo_check_socket_bind = mac_test_check_socket_bind,
|
|
|
|
.mpo_check_socket_connect = mac_test_check_socket_connect,
|
|
|
|
.mpo_check_socket_deliver = mac_test_check_socket_deliver,
|
|
|
|
.mpo_check_socket_listen = mac_test_check_socket_listen,
|
|
|
|
.mpo_check_socket_relabel = mac_test_check_socket_relabel,
|
|
|
|
.mpo_check_socket_visible = mac_test_check_socket_visible,
|
2003-03-25 01:20:56 +00:00
|
|
|
.mpo_check_sysarch_ioperm = mac_test_check_sysarch_ioperm,
|
|
|
|
.mpo_check_system_acct = mac_test_check_system_acct,
|
|
|
|
.mpo_check_system_reboot = mac_test_check_system_reboot,
|
|
|
|
.mpo_check_system_settime = mac_test_check_system_settime,
|
|
|
|
.mpo_check_system_swapon = mac_test_check_system_swapon,
|
|
|
|
.mpo_check_system_swapoff = mac_test_check_system_swapoff,
|
|
|
|
.mpo_check_system_sysctl = mac_test_check_system_sysctl,
|
Move to C99 sparse structure initialization for the mac_policy_ops
structure definition, rather than using an operation vector
we translate into the structure. Originally, we used a vector
for two reasons:
(1) We wanted to define the structure sparsely, which wasn't
supported by the C compiler for structures. For a policy
with five entry points, you don't want to have to stick in
a few hundred NULL function pointers.
(2) We thought it would improve ABI compatibility allowing modules
to work with kernels that had a superset of the entry points
defined in the module, even if the kernel had changed its
entry point set.
Both of these no longer apply:
(1) C99 gives us a way to sparsely define a static structure.
(2) The ABI problems existed anyway, due to enumeration numbers,
argument changes, and semantic mismatches. Since the going
rule for FreeBSD is that you really need your modules to
pretty closely match your kernel, it's not worth the
complexity.
This submit eliminates the operation vector, dynamic allocation
of the operation structure, copying of the vector to the
structure, and redoes the vectors in each policy to direct
structure definitions. One enourmous benefit of this change
is that we now get decent type checking on policy entry point
implementation arguments.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-30 18:48:51 +00:00
|
|
|
.mpo_check_vnode_access = mac_test_check_vnode_access,
|
|
|
|
.mpo_check_vnode_chdir = mac_test_check_vnode_chdir,
|
|
|
|
.mpo_check_vnode_chroot = mac_test_check_vnode_chroot,
|
|
|
|
.mpo_check_vnode_create = mac_test_check_vnode_create,
|
|
|
|
.mpo_check_vnode_delete = mac_test_check_vnode_delete,
|
|
|
|
.mpo_check_vnode_deleteacl = mac_test_check_vnode_deleteacl,
|
2003-08-21 17:28:45 +00:00
|
|
|
.mpo_check_vnode_deleteextattr = mac_test_check_vnode_deleteextattr,
|
Move to C99 sparse structure initialization for the mac_policy_ops
structure definition, rather than using an operation vector
we translate into the structure. Originally, we used a vector
for two reasons:
(1) We wanted to define the structure sparsely, which wasn't
supported by the C compiler for structures. For a policy
with five entry points, you don't want to have to stick in
a few hundred NULL function pointers.
(2) We thought it would improve ABI compatibility allowing modules
to work with kernels that had a superset of the entry points
defined in the module, even if the kernel had changed its
entry point set.
Both of these no longer apply:
(1) C99 gives us a way to sparsely define a static structure.
(2) The ABI problems existed anyway, due to enumeration numbers,
argument changes, and semantic mismatches. Since the going
rule for FreeBSD is that you really need your modules to
pretty closely match your kernel, it's not worth the
complexity.
This submit eliminates the operation vector, dynamic allocation
of the operation structure, copying of the vector to the
structure, and redoes the vectors in each policy to direct
structure definitions. One enourmous benefit of this change
is that we now get decent type checking on policy entry point
implementation arguments.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-30 18:48:51 +00:00
|
|
|
.mpo_check_vnode_exec = mac_test_check_vnode_exec,
|
|
|
|
.mpo_check_vnode_getacl = mac_test_check_vnode_getacl,
|
|
|
|
.mpo_check_vnode_getextattr = mac_test_check_vnode_getextattr,
|
|
|
|
.mpo_check_vnode_link = mac_test_check_vnode_link,
|
2003-08-21 17:28:45 +00:00
|
|
|
.mpo_check_vnode_listextattr = mac_test_check_vnode_listextattr,
|
Move to C99 sparse structure initialization for the mac_policy_ops
structure definition, rather than using an operation vector
we translate into the structure. Originally, we used a vector
for two reasons:
(1) We wanted to define the structure sparsely, which wasn't
supported by the C compiler for structures. For a policy
with five entry points, you don't want to have to stick in
a few hundred NULL function pointers.
(2) We thought it would improve ABI compatibility allowing modules
to work with kernels that had a superset of the entry points
defined in the module, even if the kernel had changed its
entry point set.
Both of these no longer apply:
(1) C99 gives us a way to sparsely define a static structure.
(2) The ABI problems existed anyway, due to enumeration numbers,
argument changes, and semantic mismatches. Since the going
rule for FreeBSD is that you really need your modules to
pretty closely match your kernel, it's not worth the
complexity.
This submit eliminates the operation vector, dynamic allocation
of the operation structure, copying of the vector to the
structure, and redoes the vectors in each policy to direct
structure definitions. One enourmous benefit of this change
is that we now get decent type checking on policy entry point
implementation arguments.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
2002-10-30 18:48:51 +00:00
|
|
|
.mpo_check_vnode_lookup = mac_test_check_vnode_lookup,
|
|
|
|
.mpo_check_vnode_mmap = mac_test_check_vnode_mmap,
|
|
|
|
.mpo_check_vnode_mprotect = mac_test_check_vnode_mprotect,
|
|
|
|
.mpo_check_vnode_open = mac_test_check_vnode_open,
|
|
|
|
.mpo_check_vnode_poll = mac_test_check_vnode_poll,
|
|
|
|
.mpo_check_vnode_read = mac_test_check_vnode_read,
|
|
|
|
.mpo_check_vnode_readdir = mac_test_check_vnode_readdir,
|
|
|
|
.mpo_check_vnode_readlink = mac_test_check_vnode_readlink,
|
|
|
|
.mpo_check_vnode_relabel = mac_test_check_vnode_relabel,
|
|
|
|
.mpo_check_vnode_rename_from = mac_test_check_vnode_rename_from,
|
|
|
|
.mpo_check_vnode_rename_to = mac_test_check_vnode_rename_to,
|
|
|
|
.mpo_check_vnode_revoke = mac_test_check_vnode_revoke,
|
|
|
|
.mpo_check_vnode_setacl = mac_test_check_vnode_setacl,
|
|
|
|
.mpo_check_vnode_setextattr = mac_test_check_vnode_setextattr,
|
|
|
|
.mpo_check_vnode_setflags = mac_test_check_vnode_setflags,
|
|
|
|
.mpo_check_vnode_setmode = mac_test_check_vnode_setmode,
|
|
|
|
.mpo_check_vnode_setowner = mac_test_check_vnode_setowner,
|
|
|
|
.mpo_check_vnode_setutimes = mac_test_check_vnode_setutimes,
|
|
|
|
.mpo_check_vnode_stat = mac_test_check_vnode_stat,
|
|
|
|
.mpo_check_vnode_write = mac_test_check_vnode_write,
|
Introduce support for Mandatory Access Control and extensible
kernel access control.
Provide implementations of some sample operating system security
policy extensions. These are not yet hooked up to the build as
other infrastructure is still being committed. Most of these
work fairly well and are in daily use in our development and (limited)
production environments. Some are not yet in their final form,
and a number of the labeled policies waste a lot of kernel memory
and will be fixed over the next month or so to be more conservative.
They do give good examples of the flexibility of the MAC framework
for implementing a variety of security policies.
mac_biba: Implementation of fixed-label Biba integrity policy,
similar to those found in a number of commercial
trusted operating systems. All subjects and objects
are assigned integrity levels, and information flow
is controlled based on a read-up, write-down
policy. Currently, purely hierarchal.
mac_bsdextended: Implementation of a "file system firewall",
which allows the administrator to specify a series
of rules limiting access by users and groups to
objects owned by other users and groups. This
policy is unlabeled, relying on existing system
security labeling (file permissions/ownership,
process credentials).
mac_ifoff: Secure interface silencing. Special-purpose module
to limit inappropriate out-going network traffic
for silent monitoring scenarios. Prevents the
various network stacks from generating any output
despite an interface being live for reception.
mac_mls: Implementation of fixed-label Multi-Level Security
confidentiality policy, similar to those found in
a number of commercial trusted operating systems.
All subjects and objects are assigned confidentiality
levels, and information flow is controlled based on
a write-up, read-down policy. Currently, purely
hiearchal, although non-hierarchal support is in the
works.
mac_none: Policy module implementing all MAC policy entry
points with empty stubs. A good place to start if
you want all the prototypes types in for you, and
don't mind a bit of pruning. Can be loaded, but
has no access control impact. Useful also for
performance measurements.
mac_seeotheruids: Policy module implementing a security service
similar to security.bsd.seeotheruids, only a slightly
more detailed policy involving exceptions for members
of specific groups, etc. This policy is unlabeled,
relying on existing system security labeling
(process credentials).
mac_test: Policy module implementing basic sanity tests for
label handling. Attempts to ensure that labels are
not freed multiple times, etc, etc.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
2002-07-31 18:07:45 +00:00
|
|
|
};
|
|
|
|
|
2003-03-27 19:26:39 +00:00
|
|
|
MAC_POLICY_SET(&mac_test_ops, mac_test, "TrustedBSD MAC/Test",
|
2003-04-15 20:51:18 +00:00
|
|
|
MPC_LOADTIME_FLAG_UNLOADOK | MPC_LOADTIME_FLAG_LABELMBUFS, &test_slot);
|